ePA-AIR POLLUTION-LESSON

2.0-2/98 1-1

Lesson 1Electrostatic Precipitator Operation

Goal

To familiarize you with the operation of electrostatic precipitators (ESPs).

Objectives

At the end of this lesson, you will be able to do the following:

1. Describe the theory of precipitation

2. Describe how an ESP operates to collect particulate matter

3. Describe the two ESP designs for particle charging and collection: high voltage single-stageand low voltage two-stage

4. Distinguish between cold-side and hot-side ESPs

5. Briefly describe wet ESP operation

Introduction

As you may know, particulate matter (particles) is one of the industrial air pollution problemsthat must be controlled. It's not a problem isolated to a few industries, but pervasive across awide variety of industries. That's why the U.S. Environmental Protection Agency (EPA) hasregulated particulate emissions and why industry has responded with various control devices.Of the major particulate collection devices used today, electrostatic precipitators (ESPs) areone of the more frequently used. They can handle large gas volumes with a wide range of inlettemperatures, pressures, dust volumes, and acid gas conditions. They can collect a wide rangeof particle sizes, and they can collect particles in dry and wet states. For many industries, thecollection efficiency can go as high as 99%. ESPs aren't always the appropriate collectiondevice, but they work because of electrostatic attraction (like charges repel; unlike chargesattract). Let's see how this law of physics works in an ESP.

Theory of Precipitation

Every particle either has or can be given a charge—positive or negative. Let's suppose weimpart a negative charge to all the particles in a gas stream. Then suppose we set up agrounded plate having a positive charge. What would happen? The negatively charged particlewould migrate to the grounded collection plate and be captured. The particles would quicklycollect on the plate, creating a dust layer. The dust layer would accumulate until we removed

Lesson 1

1-2 2.0-2/98

it, which we could do by rapping the plate or by spraying it with a liquid. Charging, collecting,and removing—that's the basic idea of an ESP, but it gets more complicated. Let's look at atypical scenario using a common ESP construction.

Particle Charging

Our typical ESP as shown in Figure 1-1 has thin wires called discharge electrodes, whichare evenly spaced between large plates called collection electrodes, which are grounded.Think of an electrode as something that can conduct or transmit electricity. A negative,high-voltage, pulsating, direct current is applied to the discharge electrode creating a neg-ative electric field. You can mentally divide this field into three regions (Figure 1-2). Thefield is strongest right next to the discharge electrode, weaker in the areas between the dis-charge and collection electrodes called the inter-electrode region, and weakest near thecollection electrode. The region around the discharge electrode is where the particle charg-ing process begins.

Figure 1-1. Typical dry electrostatic precipitator

Figure 1-2. ESP electric field

WeakestWeakest Strongest

Inter-electroderegion

Electric field strength

Electrostatic Precipitator Operation

2.0-2/98 1-3

Corona Discharge: Free Electron GenerationSeveral things happen very rapidly (in a matter of a millisecond) in the small areaaround the discharge electrode. The applied voltage is increased until it produces acorona discharge, which can be seen as a luminous blue glow around the dischargeelectrode. The free electrons created by the corona are rapidly fleeing the negativeelectric field, which repulses them. They move faster and faster away from the dis-charge electrode. This acceleration causes them to literally crash into gas molecules,bumping off electrons in the molecules. As a result of losing an electron (which isnegative), the gas molecules become positively charged, that is, they become positiveions (Figure 1-3). So, this is the first thing that happens—gas molecules are ionized,and electrons are liberated. All this activity occurs very close to the discharge elec-trode. This process continues, creating more and more free electrons and more posi-tive ions. The name for all this electron generation activity is avalanchemultiplication (Figure 1-4).

Figure 1-3. Corona generation

Figure 1-4. Avalanche multiplication of gas molecules

The electrons bump into gas molecules and create additional ionized molecules. Thepositive ions, on the other hand, are drawn back toward the negative discharge elec-trode. The molecules are hundreds of times bigger than the tiny electrons and move

Lesson 1

1-4 2.0-2/98

slowly, but they do pick up speed. In fact, many of them collide right into the metaldischarge electrode or the gas space around the wire causing additional electrons to beknocked off. This is called secondary emission. So, this is the second thing that hap-pens. We still have positive ions and a large amount of free electrons.

Ionization of Gas MoleculesAs the electrons leave the strong electrical field area around the discharge electrode,they start slowing down. Now they're in the inter-electrode area where they are stillrepulsed by the discharge electrode but to a lesser extent. There are also gas moleculesin the inter-electrode region, but instead of violently colliding with them, the electronskind of bump up to them and are captured (Figure 1-5). This imparts a negative chargeto the gas molecules, creating negative gas ions. This time, because the ions are nega-tive, they too want to move in the direction opposite the strong negative field. Now wehave ionization of gas molecules happening near the discharge electrode and in theinter-electrode area, but with a big difference. The ions near the discharge electrodeare positive and remain in that area. The ions in the middle area are negative and moveaway, along the path of invisible electric field lines, toward the collection electrode.

Figure 1-5. Negative gas ions formed in the inter-electrode region

Charging of ParticlesThese negative gas ions play a key role in capturing dust particles. Before the dustparticles can be captured, they must first acquire a negative charge. This is when andwhere it happens. The particles are traveling along in the gas stream and encounternegative ions moving across their path. Actually, what really happens is that the parti-cles get in the way of the negatively charged gas ions. The gas ions stick to the parti-cles, imparting a negative charge to them. At first the charge is fairly insignificant asmost particles are huge compared to a gas molecule. But many gas ions can fit on aparticle, and they do. Small particles (less than 1 µm diameter) can absorb “tens” ofions. Large particles (greater than 10 µm) can absorb "tens of thousands" of ions(Turner et al. 1992). Eventually, there are so many ions stuck to the particles, the par-ticles emit their own negative electrical field. When this happens, the negative fieldaround the particle repulses the negative gas ions and no additional ions are acquired.This is called the saturation charge. Now the negatively-charged particles are feelingthe inescapable pull of electrostatic attraction. Bigger particles have a higher satura-tion charge (more molecules fit) and consequently are pulled more strongly to the col-lection plate. In other words, they move faster than smaller particles. Regardless of

Tocollectionplate

Negativegas ion

GasmoleculeElectron


2.0-2/98 1-5

size, the particles encounter the plate and stick, because of adhesive and cohesiveforces.

Let's stop here and survey the picture. Gas molecules around the discharge electrodeare positively ionized. Free electrons are racing as fast as they can away from thestrong negative field area around the discharge electrode. The electrons are capturedby gas molecules in the inter-electrode area and impart a negative charge to them.Negative gas ions meet particles and are captured (Figure 1-6). And all this happens inthe blink of an eye. The net result is negatively charged particles that are repulsed bythe negative electric field around the discharge electrode and are strongly attracted tothe collection plate. They travel toward the grounded collection plate, bump into it,and stay there.

More and more particles accumulate, creating a dust layer. This dust layer builds untilit is somehow removed. Charging, collecting, and removing—isn't that what we saidit's all about?

Figure 1-6. Particle charging

Particle Charging MechanismsParticles are charged by negative gas ions moving toward the collection plate by oneof these two mechanisms: field charging or diffusion charging. In field charging (themechanism described above), particles capture negatively charged gas ions as the ionsmove toward the grounded collection plate. Diffusion charging, as its name implies,depends on the random motion of the gas ions to charge particles.

Negativegas ion

Negativelychargedparticle

Lesson 1

1-6 2.0-2/98

In field charging (Figure 1-7), as particles enter the electric field, they cause a localdislocation of the field. Negative gas ions traveling along the electric field lines col-lide with the suspended particles and impart a charge to them. The ions will continueto bombard a particle until the charge on that particle is sufficient to divert the electriclines away from it. This prevents new ions from colliding with the charged dust parti-cle. When a particle no longer receives an ion charge, it is said to be saturated. Satu-rated charged particles then migrate to the collection electrode and are collected.

Figure 1-7. Field charging

Diffusion charging is associated with the random Brownian motion of the negativegas ions. The random motion is related to the velocity of the gas ions due to thermaleffects: the higher the temperature, the more movement. Negative gas ions collidewith the particles because of their random thermal motion and impart a charge on theparticles. Because the particles are very small (submicrometer), they do not cause theelectric field to be dislocated, as in field charging. Thus, diffusion charging is the onlymechanism by which these very small particles become charged. The charged parti-cles then migrate to the collection electrode.

Each of these two charging mechanisms occurs to some extent, with one dominatingdepending on particle size. Field charging dominates for particles with a diameter>1.0 micrometer because particles must be large enough to capture gas ions. Diffusioncharging dominates for particles with a diameter less than 0.1 micrometer. A combina-tion of these two charging mechanisms occurs for particles ranging between 0.2 and1.0 micrometer in diameter.

A third type of charging mechanism, which is responsible for very little particle charg-ing is electron charging. With this type of charging, fast-moving free electrons thathave not combined with gas ions hit the particle and impart a charge.

b.) Saturated particle migrates towardcollection electrode

Saturatedcharged particle

a.) Field lines distorted by particle

negativegas ion

Collectionelectrode

particle


2.0-2/98 1-7

Electric Field StrengthIn the inter-electrode region, negative gas ions migrate toward the grounded collectionelectrode. A space charge, which is a stable concentration of negative gas ions, formsin the inter-electrode region because of the high electric field applied to the ESP.Increasing the applied voltage to the discharge electrode will increase the fieldstrength and ion formation until sparkover occurs. Sparkover refers to internal spark-ing between the discharge and collection electrodes. It is a sudden rush of localizedelectric current through the gas layer between the two electrodes. Sparking causes animmediate short-term collapse of the electric field (Figure 1-8.)

For optimum efficiency, the electric field strength should be as high as possible. Morespecifically, ESPs should be operated at voltages high enough to cause some sparking,but not so high that sparking and the collapse of the electric field occur too frequently.The average sparkover rate for optimum precipitator operation is between 50 and 100sparks per minute. At this spark rate, the gain in efficiency associated with increasedvoltage compensates for decreased gas ionization due to collapse of the electric field.

Figure 1-8. Spark generation profile

Particle Collection

When a charged particle reaches the grounded collection electrode, the charge on the par-ticle is only partially discharged. The charge is slowly leaked to the grounded collectionplate. A portion of the charge is retained and contributes to the inter-molecular adhesiveand cohesive forces that hold the particles onto the plates (Figure 1-9). Adhesive forcescause the particles to physically hold on to each other because of their dissimilar surfaces.Newly arrived particles are held to the collected particles by cohesive forces; particles areattracted and held to each other molecularly. The dust layer is allowed to build up on theplate to a desired thickness and then the particle removal cycle is initiated.

Lesson 1

1-8 2.0-2/98

Figure 1-9. Particle collection at collection electrode

Particle Removal

Dust that has accumulated to a certain thickness on the collection electrode is removed byone of two processes, depending on the type of collection electrode. As described ingreater detail in the next section, collection electrodes in precipitators can be either platesor tubes, with plates being more common. Tubes are usually cleaned by water sprays,while plates can be cleaned either by water sprays or a process called rapping.

Rapping is a process whereby deposited, dry particles are dislodged from the collectionplates by sending mechanical impulses, or vibrations, to the plates. Precipitator plates arerapped periodically while maintaining the continuous flue-gas cleaning process. In otherwords, the plates are rapped while the ESP is on-line; the gas flow continues through theprecipitator and the applied voltage remains constant. Plates are rapped when the accumu-lated dust layer is relatively thick (0.08 to 1.27 cm or 0.03 to 0.5 in.). This allows the dustlayer to fall off the plates as large aggregate sheets and helps eliminate dust reentrainment.Most precipitators have adjustable rappers so that rapper intensity and frequency can bechanged according to the dust concentration in the flue gas. Installations where the dustconcentration is heavy require more frequent rapping.

Dislodged dust falls from the plates into the hopper. The hopper is a single collection binwith sides sloping approximately 50 to 70° to allow dust to flow freely from the top of thehopper to the discharge opening. Dust should be removed as soon as possible to avoid(dust) packing. Packed dust is very difficult to remove. Most hoppers are emptied by sometype of discharge device and then transported by a conveyor.

In a precipitator using liquid sprays to remove accumulated liquid or dust, the sludge col-lects in a holding basin at the bottom of the vessel. The sludge is then sent to settlingponds or lined landfills for proper ultimate disposal.

Spraying occurs while the ESP is on-line and is done intermittently to remove the col-lected particles. Water is generally used as the spraying liquid although other liquids couldbe used if absorption of gaseous pollutants is also being accomplished.


2.0-2/98 1-9

Types of Electrostatic Precipitators

ESPs can be grouped, or classified, according to a number of distinguishing features in theirdesign. These features include the following:

• The structural design and operation of the discharge electrodes (rigid-frame, wires orplate) and collection electrodes (tubular or plate)

• The method of charging (single-stage or two-stage)

• The temperature of operation (cold-side or hot-side)

• The method of particle removal from collection surfaces (wet or dry)

These categories are not mutually exclusive. For example, an ESP can be a rigid-frame, sin-gle-stage, cold-side, plate-type ESP as described below.

Tubular and Plate ESPs

TubularTubular precipitators consist of cylindrical collection electrodes (tubes) with dis-charge electrodes (wires) located in the center of the cylinder (Figure 1-10). Dirty gasflows into the tubes, where the particles are charged. The charged particles are thencollected on the inside walls of the tubes. Collected dust and/or liquid is removed bywashing the tubes with water sprays located directly above the tubes. The tubes maybe formed as a circular, square, or hexagonal honeycomb with gas flowing upward ordownward. A tubular ESP is tightly sealed to minimize leaks of collected material.Tube diameters typically vary from 0.15 to 0.31 m (0.5 to 1 ft), with lengths usuallyvarying from 1.85 to 4.0 m (6 to 15 ft).

Figure 1-10. Gas flow through a tubular precipitator

Tubular precipitators are generally used for collecting mists or fogs, and are mostcommonly used when collecting particles that are wet or sticky. Tubular ESPs havebeen used to control particulate emissions from sulfuric acid plants, coke oven by-product gas cleaning (tar removal), and iron and steel sinter plants.

Dischargeelectrode

Collectionelectrodes

Lesson 1

1-10 2.0-2/98

PlatePlate electrostatic precipitators primarily collect dry particles and are used more oftenthan tubular precipitators. Plate ESPs can have wire, rigid-frame, or occasionally,plate discharge electrodes. Figure 1-11 shows a plate ESP with wire discharge elec-trodes. Dirty gas flows into a chamber consisting of a series of discharge electrodesthat are equally spaced along the center line between adjacent collection plates.Charged particles are collected on the plates as dust, which is periodically removed byrapping or water sprays. Discharge wire electrodes are approximately 0.13 to 0.38 cm(0.05 to 0.15 in.) in diameter. Collection plates are usually between 6 and 12 m (20and 40 ft) high. For ESPs with wire discharge electrodes, the plates are usually spacedfrom 15 to 30 cm (6 to 12 in.) apart. For ESPs with rigid-frame or plate discharge elec-trodes, plates are typically spaced 30 to 38 cm (12 to 15 in.) apart and 8 to 12 m (30 to40 ft) in height.

Plate ESPs are typically used for collecting fly ash from industrial and utility boilersas well as in many other industries including cement kilns, glass plants and pulp andpaper mills.

Figure 1-11. Gas flow through a plate precipitator

Single-stage and Two-stage ESPs

Another method of classifying ESPs is by the number of stages used to charge and removeparticles from a gas stream. A single-stage precipitator uses high voltage to charge theparticles, which are then collected within the same chamber on collection surfaces ofopposite charge. In a two-stage precipitator, particles are charged by low voltage in onechamber, and then collected by oppositely charged surfaces in a second chamber.

Single StageMost ESPs that reduce particulate emissions from boilers and other industrial pro-cesses are single-stage ESPs (these units will be emphasized in this course). Single-stage ESPs use very high voltage (50 to 70 kV) to charge particles. After beingcharged, particles move in a direction perpendicular to the gas flow through the ESP,

Dischargeelectrode

Collectionplate


2.0-2/98 1-11

and migrate to an oppositely charged collection surface, usually a plate or tube. Parti-cle charging and collection occurs in the same stage, or field; thus, the precipitatorsare called single-stage ESPs. The term field is used interchangeably with the termstage and is described in more detail later in this course. Figure 1-10 shows a single-stage tubular precipitator. A single-stage plate precipitator is shown in Figure 1-11.

Two StageThe two-stage precipitator differs from the single-stage precipitator in both design andamount of voltage applied. The two-stage ESP has separate particle charging and col-lection stages (Figure 1-12). The ionizing stage consists of a series of small, positivelycharged wires equally spaced 2.5 to 5.1 cm (1 to 2 in.) from parallel grounded tubes orrods. A corona discharge between each wire and a corresponding tube charges the par-ticles suspended in the air flow as they pass through the ionizer. The direct-currentpotential applied to the wires is approximately 12 to 13 kV.

Figure 1-12. Representation of gas flow in a two-stage precipitator

The second stage consists of parallel metal plates less than 2.5 cm (1 in.) apart. Theparticles receive a positive charge in the ionizer stage and are collected at the negativeplates in the second stage. Collected smoke or liquids drain by gravity to a pan locatedbelow the plates, or are sprayed with water mists or solvents that remove the particlesand cause them to fall into the bottom pan.

Two-stage precipitators were originally designed for air purification in conjunctionwith air conditioning systems. (They are also referred to as electronic air filters). Two-stage ESPs are used primarily for the control of finely divided liquid particles. Con-trolling solid or sticky materials is usually difficult, and the collector becomes ineffec-tive for dust loadings greater than 7.35 x 10-3g/m3 (0.4 gr/dscf). Therefore, two-stageprecipitators have limited use for particulate-emission control. They are used almostexclusively to collect liquid aerosols discharged from sources such as meat smoke-houses, pipe-coating machines, asphalt paper saturators, high speed grindingmachines, welding machines, and metal-coating operations.

Ionizer(to charge particles)

Collection plate

Baffle(to distributeair uniformly)

Cleanair

Precipitated(collected)particles

Positivelychargedparticles

Unchargedparticles

Lesson 1

1-12 2.0-2/98

Cold-side and Hot-side ESPs

Electrostatic precipitators are also grouped according to the temperature of the flue gasthat enters the ESP: cold-side ESPs are used for flue gas having temperatures of approxi-mately 204°C (400°F) or less; hot-side ESPs are used for flue gas having temperaturesgreater than 300°C (572°F).

In describing ESPs installed on industrial and utility boilers, or municipal waste combus-tors using heat recovery equipment, cold side and hot side also refer to the placement ofthe ESP in relation to the combustion air preheater. A cold-side ESP is located behind theair preheater, whereas a hot-side ESP is located in front of the air preheater. The air pre-heater is a tube section that preheats the combustion air used for burning fuel in a boiler.When hot flue gas from an industrial process passes through an air preheater, a heatexchange process occurs whereby heat from the flue gas is transferred to the combustionair stream. The flue gas is therefore "cooled" as it passes through the combustion air pre-heater. The warmed combustion air is sent to burners, where it is used to burn gas, oil,coal, or other fuel including garbage. APTI Course SI:428A Introduction to Boiler Opera-tion describes boilers and heat recovery equipment in greater detail.

Cold SideCold-side ESPs (Figure 1-13) have been used for over 50 years with industrial andutility boilers, where the flue gas temperature is relatively low (less than 204°C or400°F). Cold-side ESPs generally use plates to collect charged particles. Becausethese ESPs are operated at lower temperatures than hot-side ESPs, the volume of fluegas that is handled is less. Therefore, the overall size of the unit is smaller, making itless costly. Cold-side ESPs can be used to remove fly ash from boilers that burn high-sulfur coal. As explained in later lessons, cold-side ESPs can effectively remove flyash from boilers burning low-sulfur coal with the addition of conditioning agents.

Figure 1-13. Cold-side ESP

Boiler

ESP

Fan

Combustionair preheater


2.0-2/98 1-13

Hot SideHot-side ESPs (Figure 1-14) are placed in locations where the flue gas temperature isrelatively high. Their collection electrodes can be either tubular or plate. Hot-sideESPs are used in high-temperature applications, such as in the collection of cement-kiln dust or utility and industrial boiler fly ash. A hot-side precipitator is locatedbefore the combustion air preheater in a boiler. The flue gas temperature for hot-sideprecipitators is in the range of 320 to 420°C (608 to 790°F).

The use of hot-side precipitators help reduce corrosion and hopper plugging. How-ever, these units (mainly used on coal-fired boilers) have some disadvantages.Because the temperature of the flue gas is higher, the gas volume treated in the ESP islarger. Consequently, the overall size of the precipitator is larger making it morecostly. Other major disadvantages include structural and mechanical problems thatoccur in the precipitator shell and support structure as a result of differences in ther-mal expansion.

For years, cold-side ESPs were used successfully on boilers burning high-sulfur coal.However, during the 1970s when utilities switched to burning low-sulfur coal, cold-side ESPs were no longer effective at collecting the fly ash. Fly ash produced fromlow sulfur coal-fired boilers has high resistivity (discussed in more detail later in thecourse), making it difficult to collect. As you will learn later, high temperatures canlower resistivity. Consequently, hot-side ESPs became very popular during the 1970sfor removing ash from coal-fired boilers burning low sulfur coal. However, many ofthese units did not operate reliably, and therefore, since the 1980s, operators have gen-erally decided to use cold-side ESPs along with conditioning agents when burning lowsulfur coal.

Hot-side ESPs are also used in industrial applications such as cement kilns and steelrefining furnaces. In these cases, combustion air preheaters are generally not used andhot side just refers to the high flue gas temperature prior to entering the ESP.

Figure 1-14. Hot-side ESP

Boiler

ESP

Fan

Combustionair preheater

Lesson 1

1-14 2.0-2/98

Wet and Dry ESPs

Wet ESPsAny of the previously described ESPs can be operated with a wet spray to remove col-lected particles. Wet ESPs are used for industrial applications where the potential forexplosion is high (such as collecting dust from a closed-hood Basic Oxygen Furnacein the steel industry), or when dust is very sticky, corrosive, or has very high resistiv-ity. The water flow may be applied continuously or intermittently to wash the col-lected particles from the collection electrodes into a sump (a basin used to collectliquid). The advantage of using a wet ESP is that it does not have problems with rap-ping reentrainment or with back corona which are discussed in more detail in Lesson3.

Figures 1-15 and 1-16 show two different wet ESPs. The casing of wet ESPs is madeof steel or fiberglass and the discharge electrodes are made of carbon steel or specialalloys, depending on the corrosiveness of the flue gas stream.

In a circular-plate wet ESP, shown in Figure 1-15, the circular collection plates aresprayed with liquid continuously. The liquid provides the electrical ground for attract-ing the particles and for removing them from the plates. These units can handle gasflow rates of 30,000 to 100,000 cfm. Preconditioning sprays located at the inletremove some particulate matter prior to the charging stage. The operating pressuredrop across these units is typically 1 to 3 inches of water.

Figure 1-15. Circular-plate wet EPSReproduced with permission of Fluid Ionics Systems, adivision of Dresser Industries, Inc.

Gas inlet

Water distributor

Insulator Concentriccollection surfaces

Emittingelectrodes

Venturi/draingutters

Straighteningvanes

Preconditionersprays

Clean gas discharge


2.0-2/98 1-15

Rectangular flat-plate wet ESPs, shown in Figure 1-16, operate similarly to circular-plate wet ESPs. Water sprays precondition the gas stream and provide some particleremoval. Because the water sprays are located over the top of the electrical fields, thecollection plates are continuously irrigated. The collected particulate matter flowsdownward into a trough that is sloped to a drain.

Figure 1-16. Flat-plate modular wet ESPReproduced with permission of Fluid Ionics Systems, adivision of Dresser Industries, Inc.

Dry ESPsMost electrostatic precipitators are operated dry and use rappers to remove the col-lected particulate matter. The term dry is used because particles are charged and col-lected in a dry state and are removed by rapping as opposed to water washing which isused with wet ESPs. The major portion of this course covers dry ESPs that are usedfor collecting dust from many industries including steel furnaces, cement kilns andfossil-fuel-fired boilers.

Water manifolds

Gas outlet

Water Inlet

Discharge electrode

Water outlet

Gas inlet

Access manway

Turning vanes

Perforated plates

Collectionplate

Lesson 1

1-16 2.0-2/98

Summary

All ESPs, no matter how they are grouped, have similar components and operate by chargingparticles or liquid aerosols, collecting them, and finally removing them from the ESP beforeultimate disposal in a landfill or reuse in the industrial process.

ESPs are occasionally referred to as cold-side, tubular, or by some other descriptor. ESPdesigns usually incorporate a number of ESP features into one unit. For example, a typicalESP used for removing particulate matter from a coal-fired boiler will be a cold-side, single-stage, plate ESP. On the other hand, a hot-side, single-stage, tubular ESP may be used to cleanexhaust gas from a blast furnace in a steel mill.

Remember that an ESP is specifically designed to collect particulate matter or liquids for anindividual industrial application. Vendors use those features, i.e., tubes, plates, etc., that mostreadily enhance the removal of the pollutants from the flue gas. These features are described inmore detail in the remaining lessons.


2.0-2/98 1-17

Review Exercise

1. In an electrostatic precipitator, the ____________________ electrode is normally a small-diametermetal wire or a rigid frame containing wires.

2. The charged particles migrate to the ____________________ ____________________.

3. In a single-stage, high-voltage ESP, the applied voltage is increased until it produces a(an)

a. Extremely high alternating current for particle chargingb. Corona discharge, which can be seen as a blue glow around the discharge electrodec. Corona spark that occurs at the collection electrode

4. True or False? Particles are usually charged by negative gas ions that are migrating toward the col-lection electrode.

5. True or False? Large particles move more slowly towards the collection plate than small particles.

6. The average sparkover rate (in sparks per minute) for optimum precipitator operation is between:

a. 1 - 25b. 50 - 100c. 100 - 150d. 500 - 1,000

7. As dust particles reach the grounded collection electrode, their charge is:

a. Immediately transferred to the collection plateb. Slowly leaked to the grounded collection electrodec. Cancelled out by the strong electric field

8. Particles are held onto the collection plates by:

a. A strong electric force fieldb. A high-voltage, pulsating, direct currentc. Intermolecular cohesive and adhesive forcesd. Electric sponsors

9. Dust that has accumulated on collection electrodes can be removed either by____________________ ____________________ or a process called ____________________.

10. True or False? During the rapping process, the voltage is turned down to about 50% of the normaloperating voltage to allow the rapped particles to fall freely into the hopper.

11. ____________________ electrostatic precipitators are used for removing particulate matter fromflue gas that usually has a temperature range of 320 to 420° C (608 to 790° F).

Lesson 1

1-18 2.0-2/98

12. In a boiler, hot-side ESPs are located ____________________ air preheaters, whereas cold-sideESPs are located ____________________ air preheaters.

a. In front of, behindb. Behind, in front of

13. True or False? Wet electrostatic precipitators are used when collecting dust that is sticky or hashigh resistivity.

14. ____________________ ESPs are units where particle charging occurs in the first stage, followedby collection in the second stage.


2.0-2/98 1-19

Review Exercise Answers

1. DischargeIn an electrostatic precipitator, the discharge electrode is normally a small-diameter metal wire or arigid frame containing wires.

2. Collection electrodeThe charged particles migrate to the collection electrode.

3. b. Corona discharge, which can be seen as a blue glow around the discharge electrodeIn a single-stage, high-voltage ESP, the applied voltage is increased until it produces a corona dis-charge, which can be seen as a blue glow around the discharge electrode.

4. TrueParticles are usually charged by negative gas ions that are migrating toward the collection elec-trode.

5. FalseLarge particles move faster towards the collection plate than small particles. Large particles have ahigher saturation charge than small particles; consequently, large particles are pulled more stronglyto the collection plate.

6. b. 50 - 100The average sparkover rate for optimum precipitator operation is between 50 - 100 sparks perminute.

7. b. Slowly leaked to the grounded collection electrodeAs dust particles reach the grounded collection electrode, their charge is slowly leaked to thegrounded collection electrode.

8. c. Intermolecular cohesive and adhesive forcesParticles are held onto the collection plates by intermolecular cohesive and adhesive forces.

9. Water spraysRappingDust that has accumulated on collection electrodes can be removed either by water sprays or a pro-cess called rapping.

10. FalseDuring the rapping process, the voltage is NOT turned down. Rapping occurs while the ESPremains on-line.

11. Hot-sideHot-side electrostatic precipitators are used for removing particulate matter from flue gas that usu-ally has a temperature range of 320 to 420°C (608 to 790°F).

Lesson 1

1-20 2.0-2/98

12. a. In front of, behindIn a boiler, hot-side ESPs are located in front of air preheaters, whereas cold-side ESPs are locatedbehind air preheaters. Recall that flue gas is cooled as it passes through the combustion air pre-heater.

13. TrueWet electrostatic precipitators are used when collecting dust that is sticky or has high resistivity.

14. Two-stageTwo-stage ESPs are units where particle charging occurs in the first stage, followed by collectionin the second stage.


2.0-2/98 1-21

Bibliography

Beachler, D. S., J. A. Jahnke, G. T. Joseph and M. M. Peterson. 1983. Air Pollution Control Systems forSelected Industries-Self-Instructional Guidebook. (APTI Course SI:431). EPA 450/2-82-006. U.S.Environmental Protection Agency.

Bethea, R. M. 1978. Air Pollution Control Technology-an Engineering Analysis Point of View. NewYork: Van Nostrand Reinhold.

Katz, J. 1979. The Art of Electrostatic Precipitators. Munhall, PA: Precipitator Technology.

Nichols, G. B. 1976, September. Electrostatic Precipitation. Seminar presented to the U.S. Environ-mental Protection Agency. Research Triangle Park, NC.

Richards, J.R. 1995. Control of Particulate Emissions-Student Manual. (APTI Course 413). U.S. Envi-ronmental Protection Agency.

Turner, J. H., P. A. Lawless, T. Yamamoto, D. W. Coy, G. P. Greiner, J. D. McKenna, and W. M. Vata-vuk. 1992. Electrostatic precipitators. In A. J. Buonicore and W. T. Davis (Eds.), Air PollutionEngineering Manual (pp. 89-113). Air and Waste Management Association. New York: Van Nos-trand Reinhold.

U.S. Environmental Protection Agency. 1973. Air Pollution Engineering Manual. 2d ed. AP-40.

U.S. Environmental Protection Agency. 1985. Operation and Maintenance Manual for ElectrostaticPrecipitators. EPA 625/1-85/017.

White, H. J. 1977. Electrostatic precipitation of fly ash. APCA Reprint Series. Journal of Air PollutionControl Association. Pittsburgh, PA.

1-22 2.0-2/98

2.0-2/98 2-1

Lesson 2Electrostatic Precipitator Components

Goal

To familiarize you with the components of an ESP.

Objectives


1. Identify six major components of an ESP2. Describe typical discharge electrode designs3. Describe typical collection electrode designs4. Identify how discharge electrodes and collection plates are installed in an ESP5. List three types of rappers and briefly describe how they operate6. Describe how the high-voltage equipment operates7. Describe two factors that are important in hopper design8. Identify two discharge devices used to remove dust from hoppers, and three types of conveyors9. State the purpose for installing insulation on an ESP

Video Presentation (optional): If you have acquired the video titled, Electrostatic Precipitators:Operating Principles and Components, please view it at the end of this lesson.

Precipitator Components

All electrostatic precipitators, regardless of their particular designs, contain the followingessential components:

• Discharge electrodes

• Collection electrodes

• High voltage electrical systems

• Rappers

• Hoppers

• Shell

Discharge electrodes are either small-diameter metal wires that hang vertically (in the electro-static precipitator), a number of wires attached together in rigid frames, or a rigid electrode-made from a single piece of fabricated metal. Discharge electrodes create a strong electricalfield that ionizes flue gas, and this ionization charges particles in the gas.

Lesson 2

2-2 2.0-2/98

Collection electrodes collect charged particles. Collection electrodes are either flat plates ortubes with a charge opposite that of the discharge electrodes.

High voltage equipment provides the electric field between the discharge and collection elec-trodes used to charge particles in the ESP.

Rappers impart a vibration, or shock, to the electrodes, removing the collected dust. Rappersremove dust that has accumulated on both collection electrodes and discharge electrodes.Occasionally, water sprays are used to remove dust from collection electrodes.

Hoppers are located at the bottom of the precipitator. Hoppers are used to collect and tempo-rarily store the dust removed during the rapping process.

The shell provides the base to support the ESP components and to enclose the unit.

Figure 2-1 shows a typical ESP with wires for discharge electrodes and plates for collectionelectrodes. This ESP is used to control particulate emissions in many different industries.

Figure 2-1. Typical dry electrostatic precipitator

Discharge Electrodes

The discharge electrodes in most U.S. precipitator designs (prior to the 1980s) are thin,round wires varying from 0.13 to 0.38 cm (0.05 to 0.15 in.) in diameter. The most com-mon size diameter for wires is approximately 0.25 cm (0.1 in.). The discharge electrodesare hung vertically, supported at the top by a frame and held taut and plumb by a weight atthe bottom. The wires are usually made from high-carbon steel, but have also been con-structed of stainless steel, copper, titanium alloy, and aluminum. The weights are made ofcast iron and are generally 11.4 kg (25 lb) or more.

Discharge wires are supported to help eliminate breakage from mechanical fatigue. Thewires move under the influence of aerodynamic and electrical forces and are subject tomechanical stress. The weights at the bottom of the wire are attached to guide frames tohelp maintain wire alignment and to prevent them from falling into the hopper in the eventthat the wire breaks (Figure 2-2).

Dischargeelectrodes

Flue gas in

Rappers

Hoppers

Cleangasout

Electrostatic Precipitator Components

2.0-2/98 2-3

Weights that are 11.4 kg (25 lb) are used with wires 9.1 m (30 ft) long, and 13.6 kg (30 lb)weights are used with wires from 10.7 to 12.2 m (35 to 40 ft) long. The bottom and top ofeach wire are usually covered with a shroud of steel tubing. The shrouds help minimizesparking and consequent metal erosion by sparks at these points on the wire.

Figure 2-2. Guide frames and shrouds fordischarge wires

The size and shape of the electrodes are governed by the mechanical requirements for thesystem, such as the industrial process on which ESPs are installed and the amount andproperties of the flue gas being treated. Most U.S. designs have traditionally used thin,round wires for corona generation. Some designers have also used twisted wire, squarewire, barbed wire, or other configurations, as illustrated in Figure 2-3.

Figure 2-3. Typical wire dischare electrodes

European precipitator manufacturers and most of the newer systems (since the early1980s) made by U.S. manufacturers use rigid support frames for discharge electrodes. Theframes may consist of coiled-spring wires, serrated strips, or needle points mounted on asupporting strip. A typical rigid-frame discharge electrode is shown in Figure 2-4. The

Upperguide frame

Top shroud

Bottom shroud

Guide loop

Weight

Lowerguide frame

Lesson 2

2-4 2.0-2/98

purpose of the rigid frame is to eliminate the possible swinging of the discharge wires.Another type of discharge electrode is a rigid electrode that is constructed from a singlepiece of fabricated metal and is shown in Figure 2-5. Both designs are occasionallyreferred to as rigid-frame electrodes. They have been used as successfully as the olderU.S. wire designs. One major disadvantage of the rigid-frame design is that a broken wirecannot be replaced without removing the whole frame.

Figure 2-4. Rigid frame discharge electrode design

Figure 2-5. Typical rigid discharge electrode

Dischargeelectrode

Dischargeelectrode

Supportinsulator

Rapperanvil


2.0-2/98 2-5

One U.S. manufacturer (United McGill) uses flat plates instead of wires for dischargeelectrodes. The flat plates, shown in Figure 2-6, increase the average electric field that canbe used for collecting particles and provide an increased surface area for collecting parti-cles, both on the discharge and collection plates. The corona is generated by the sharp-pointed needles attached to the plates. These units generally use positive polarity forcharging the particles. The units are typically operated with low flue gas velocity to pre-vent particle reentrainment during the rapping cycle (Turner, et al. 1992).

Figure 2-6. Flat-plate discharge electrode(United McGill design)

Collection Electrodes

Most U.S. precipitators use plate collection electrodes because these units treat large gasvolumes and are designed to achieve high collection efficiency. The plates are generallymade of carbon steel. However, plates are occasionally made of stainless steel or an alloysteel for special flue-gas stream conditions where corrosion of carbon steel plates wouldoccur. The plates range from 0.05 to 0.2 cm (0.02 to 0.08 in.) in thickness. For ESPs withwire discharge electrodes, plates are spaced from 15 to 30 cm apart (6 to 12 in.). Normalspacing for high-efficiency ESPs (using wires) is 20 to 23 cm (8 to 9 in.). For ESPs usingrigid-frame or plate discharge electrodes, collection plates are typically spaced 30 to 38cm (12 to 15 inches) apart. Plates are usually between 6 and 12 m (20 to 40 ft) high.

Dischargeelectrodeplate

Collectionplate

Lesson 2

2-6 2.0-2/98

Collection plates are constructed in various shapes, as shown in Figure 2-7. These platesare solid sheets that are sometimes reinforced with structural stiffeners to increase platestrength. In some cases, the stiffeners act as baffles to help reduce particle reentrainmentlosses. This design minimizes the amount of excess rapping energy required to dislodgethe dust from the collection plates, because the energy is distributed evenly throughout theplate. The baffles also provide a "quiet zone" for the dislodged dust to fall while mini-mizing dust reentrainment.

Figure 2-7. Typical collection plates

As stated in Lesson 1, tubes are also used as collection electrodes, but not nearly as often.Tubes are typically used to collect sticky particles and when liquid sprays are used toremove the collected particles.

High-Voltage Equipment

High-voltage equipment determines and controls the strength of the electric field gener-ated between the discharge and collection electrodes. This is accomplished by usingpower supply sets consisting of three components: a step-up transformer, a high-voltagerectifier, and control metering and protection circuitry (automatic circuitry). The powersystem maintains voltage at the highest level without causing excess sparkover betweenthe discharge electrode and collection plate. These power sets are also commonly calledtransformer-rectifier (T-R) sets.

In a T-R set, the transformer steps up the voltage from 400 volts to approximately 50,000volts. This high voltage ionizes gas molecules that charge particles in the flue gas. Therectifier converts alternating current to direct current. Direct (or unidirectional current) isrequired for electrical precipitation. Most modern precipitators use solid-state silicon rec-tifiers and oil-filled, high-voltage transformers. The control circuitry in a modern precipi-tator is usually a Silicon-controlled Rectifier (SCR) automatic voltage controller with alinear reactor in the primary side of the transformer. Meters, also included in the control


2.0-2/98 2-7

circuitry, monitor the variations in the electrical power input. A simplified drawing of thecircuitry from the primary control cabinet to the precipitator field is shown in Figure 2.8

Figure 2-8. Schematic diagram of circuitry associated with precipitators

The most commonly used meters are the following:

Primary voltmeter. This meter measures the input voltage, in a.c. volts, coming intothe transformer. The input voltage ranges from 220 to 480 volts; however, most mod-ern precipitators use 400 to 480 volts. The meter is located across the primary windingof the transformer.Primary ammeter. This meter measures the current drawn acrossthe transformer in amperes. The primary ammeter is located across the primary wind-ing (wires wound in the coil) of the transformer. The primary voltage and currentreadings give the power input to a particular section of the ESP.

Secondary voltmeter. This meter measures, in d.c. volts, the operating voltage deliv-ered to the discharge electrodes. The meter is located between the output side of therectifier and the discharge electrodes.

Secondary ammeter. This meter measures the current supplied to the discharge elec-trodes in milliamperes. The secondary ammeter is located between the rectifier outputand the automatic control module. The combination of the secondary voltage and cur-rent readings gives the power input to the discharge electrodes.

Sparkmeter. This meter measures the number of sparks per minute in the precipitatorsection. Sparks are surges of localized electric current between the discharge elec-trodes and the collection plate.

Lesson 2

2-8 2.0-2/98

The terms primary and secondary refer to the side of the transformer being monitored bythe meter. Figure 2-9 shows the typical meters used on each ESP field and are located inthe control cabinet.

Figure 2-9. Typical gauges (meters) installed on control cabinet foreach precipitator field

The transformer-rectifier set ios connected to the discharge electrodes by a bus line. A busline is electric cable that carries high voltage from the transformer-rectifier to the dis-charge electrodes (Figure 2-10). The bus line is encased in a pipe, or bus duct, to protectthe high-voltage line from the environment and to prevent the line from becoming a poten-tial hazard to humans. The high-voltage bus lines are separated, or isolated, from the ESPframe and shells by insulators. The insulators are made of nonconducting plastic orceramic material.

Figure 2-10. High-voltage system

90100

8070605040302010

0

D.C. Kilovolts

7525100

50

0

Sparks/Minute

21.51.5

0

D.C. Amps

20015010050

0

A.C. Amps500

400300200

100

0

A.C. Volts

Secondary Voltage

AlarmPower

OffPower

On

On

StartOff

Spark meterSecondary Current

Primary CurrentPrimary Voltage

Bus lineSupportinsulatorhousing

High voltage bus duct

Transformerrectifier


2.0-2/98 2-9

Rappers

Dust that has accumulated on collection and discharge electrodes is removed by rapping.Dust deposits are generally dislodged by mechanical impulses, or vibrations, imparted tothe electrodes. A rapping system is designed so that rapping intensity and frequency canbe adjusted for varying operational conditions. Once the operating conditions are set, thesystem must be capable of maintaining uniform rapping for a long time.

Collection electrodes are rapped by hammer/anvil or magnetic impulse systems. Rigidframe discharge electrodes are rapped by tumbling hammers and wires are rapped byvibrators. As stated previously, liquid sprays are also used (instead of rapping) to removecollected particles form both tubes and plates.

Hammer/AnvilCollection plates are rapped by a number of methods. One rapper system uses ham-mers mounted on a rotating shaft, as shown in Figure 2-11. As the shaft rotates, thehammers drop (by gravity) and strike anvils that are attached to the collection plates.Rappers can be mounted on the top or on the side of collection plates. European pre-cipitator manufacturers use hammer and anvil rappers for removing particles fromcollection plates.

Rapping intensity is controlled by the weight of the hammers and the length of thehammer mounting arm. The frequency of rapping can be changed by adjusting thespeed of the rotating shafts. Thus, rapping intensity and frequency can be adjusted forthe varying dust concentration of the flue gas.

Figure 2-11. Typical hammer/anvil rappers forcollection plates

Hammer

Anvil

Lesson 2

2-10 2.0-2/98

Magnetic ImpulseAnother rapping system used for many U.S. designs consists of magnetic-impulse rap-pers to remove accumulated dust layers from collection plates. A magnetic-impulserapper has a steel plunger that is raised by a current pulse in a coil. The raised plungerthen drops back, due to gravity, striking a rod connected to a number of plates withinthe precipitator as shown in Figure 2-12. Rapper frequency and intensity are easilyregulated by an electrical control system. The frequency could be one rap every fiveminutes or one rap an hour with an intensity of 10 to 24 g's (Katz 1979). Magnetic-impulse rappers usually operate more frequently, but with less intensity, than rotatinghammer and anvil rappers.

Figure 2-12. Typical magnetic-impulse rappersfor collection plates

Tumbling Hammersfor Rigid Frame Discharge Electrodes

Rigid frame discharge electrodes are rapped by tumbling hammers. The tumblinghammers operate similarly to the hammers used to remove dust from collection elec-trodes. The hammers are arranged on a horizontal shaft. As the shaft rotates, thehammers hit an impact beam which transfers the shock, or vibration, to the centertubes on the discharge system, causing the dust to fall (Figure 2-13).

Electric VibratorWire discharge (or corona) electrodes must also be rapped to prevent excessive dustdeposit buildup that will interfere with corona generation. This is usually accom-plished by the use of air or electric vibrators that gently vibrate the discharge wires.Vibrators are usually mounted externally on precipitator roofs and are connected byrods to the high-tension frames that support the corona electrodes (Figure 2-14). Aninsulator, located above the rod, electrically insulates the rapper while mechanicallytransmitting the rapping force.

Rapper rods


2.0-2/98 2-11

Figure 2-13. Tumbling hammers for rigid-framedischarge electrode

Figure 2-14. Typical electric-vibrator rappers usedfor wire discharge electrodes

Impactbeam

Tumblinghammer

Dischargewire

Centertube

Rapper

Highvoltage

frame

Rapperinsulator

Wiresupportchannel

Lesson 2

2-12 2.0-2/98

Hoppers

When the electrodes are rapped, the dust falls into hoppers and is stored temporarilybefore it is disposed in a landfill or reused in the process. Dust should be removed as soonas possible to avoid packing, which would make removal very difficult. Hoppers are usu-ally designed with a 50 to 70° (60° is common) slope to allow dust to flow freely from thetop of the hopper to the bottom discharge opening.

Some manufacturers add devices to the hopper to promote easy and quick discharge.These devices include strike plates, poke holes, vibrators, and rappers. Strike plates aresimply pieces of flat steel that are bolted or welded to the center of the hopper wall. If dustbecomes stuck in the hopper, rapping the strike plate several times with a mallet will freethis material. Hopper designs also usually include access doors, or ports. Access portsallow easier access for cleaning, inspection, and maintenance of the hopper (Figure 2-15).

Figure 2-15. Hopper

Hopper vibrators are occasionally used to help remove dust from the hopper walls. Hoppervibrators are electrically operated devices that cause the side walls of the hopper tovibrate, thereby removing the dust from the hopper walls. These devices must be care-fully designed and chosen so that they do not cause dust to be firmly packed against thehopper walls, and thereby plug the hopper. Before installing vibrators to reduce hopperplugging, make sure they have been successfully used in other, similar industrial applica-tions.

Hopper Discharge DevicesA discharge device is necessary for emptying the hopper and can be manual or auto-matic. The simplest manual discharge device is the slide gate, a plate held in place bya frame and sealed with gaskets (Figure 2-16). When the hopper needs to be emptied,the plate is removed and the material is discharged. Other manual discharge devices

Accessport

Strikeplate

Dischargedevice

Conveyor


2.0-2/98 2-13

include hinged doors and drawers. The collector must be shut down before openingany manual discharge device. Thus, manual discharge devices are used only on verysmall units that operate on a periodic basis.

Figure 2-16. Slide-gate

Automatic continuous discharge devices are installed on ESPs that operate continuously.Some devices include double-dump valves (also called double flap or trickle valves), androtary airlock valves. Double-dump valves are shown in Figure 2-17. As dust collects inthe hopper, the weight of the dust pushes down the counterweight of the top flap and dustdischarges downward. The top flap then closes, the bottom flap opens, and the materialfalls out. This type of valve is available in gravity-operated and motorized versions.

Figure 2-17. Double-dump discharge device

Rotary airlock valves are used on medium or large-sized ESPs. The valve is designedwith a paddle wheel that is shaft mounted and driven by a motor (Figure 2-18). Therotary valve is similar to a revolving door; the paddles or blades form an airtight sealwith the housing, and the motor slowly moves the blades to allow the dust to dischargefrom the hopper.

Lesson 2

2-14 2.0-2/98

Figure 2-18. Rotary airlock discharge device

After the dust leaves the discharge device it is transported to the final disposal destina-tion by screw, drag, or pneumatic conveyers. Screw conveyors can be used as dis-charge devices when located in the bottom of the hopper as shown in Figure 2-19 oras a separate conveyor to move dust after it is discharged. Screw conveyers employ arevolving screw feeder to move the dust through the conveyor. Drag conveyors usepaddles, or flaps, that are connected to a drag chain to pull the dust through the con-veyor trough (Figure 2-20). Drag conveyors are used frequently for conveying stickyor hygroscopic dusts such as calcium chloride dust generated from municipal wastecombustors (collected fly ash/acid gas products). Pneumatic conveyers use blowersto blow or move the dust through the conveyor (Figure 2-21). Pneumatic conveyorscan be positive pressure (dust is moved by a blower) or vacuum type systems (dust ispulled by a vacuum).

In large ESPs, dust is usually discharged from hoppers by using a combination ofdevices. Either rotary airlock or double dump valves empty dust into screw, drag, orpneumatic conveyers that move dust for final disposal into trucks or storage bins.


2.0-2/98 2-15

Figure 2-19. Screw conveyor

Figure 2-20. Drag conveyor

Figure 2-21. Pneumatic conveyor for transporting dust from ESP

Lesson 2

2-16 2.0-2/98

Shell

The shell structure encloses the electrodes and supports the precipitator components in arigid frame to maintain proper electrode alignment and configuration (Figure 2-22). Thesupport structure is especially critical for hot-side precipitators because precipitator com-ponents can expand and contract when the temperature differences between the ESP(400°C or 752°F) and the ambient atmosphere (20°C or 68°F) are large. Excessive temper-ature stresses can literally tear the shell and hopper joints and welds apart. The outer sheetor casing wall is usually made of low-carbon or mild-grade steel that is 0.5 to 0.6 cm (3/16to 1/4 in.) thick.

Figure 2-22. ESP shell

Collection plates and discharge electrodes are normally attached to the frame at the top sothat the elements hang vertically due to gravity. This allows the elements to expand orcontract with temperature changes without binding or distorting.

Shells, hoppers, and connecting flues should be covered with insulation to conserve heat,and to prevent corrosion resulting from water vapor and acid condensation on internal pre-cipitator components. If the ESP is installed on a coal-fired boiler, the flue gas temperatureshould be kept above 120°C (250°F) at all times to prevent any acid mists in the flue gasfrom condensing on ESP internal components. Insulation will also help minimize temper-ature-differential stresses, especially on hot-side precipitators. Ash hoppers should beinsulated and heated because cold fly ash has a tendency to cake, making it extremely dif-ficult to remove. Insulation material is usually 10 to 15 cm (4 to 6 in.) thick.


2.0-2/98 2-17

Summary

The precipitator should be designed to provide easy access to strategic points of the collectorfor internal inspection of electrode alignment, for maintenance, and for cleaning electrodes,hoppers, and connecting flues during outages. Vendors typically design the ESPs for a spe-cific particulate emission removal efficiency. The overall design, including the specific com-ponents, is based on engineering specifications and/or previous experience with the industrialapplication. These components have an effect on the overall performance and ease of opera-tion of the ESP. These topics are discussed in more detail in the following lessons.

Please view the video titled Electrostatic Precipitators: Operating Principles and Componentsbefore proceeding to the next lesson.

Lesson 2

2-18 2.0-2/98


2.0-2/98 2-19

Review Exercise

1. List the six major components of an ESP.

___________________________ ___________________________

___________________________ ___________________________

___________________________ ___________________________

2. In many U.S. precipitators, the discharge electrodes are thin wires that are approximately____________________ in diameter.

a. 2.0 in.b. 0.1 in.c. 0.01 in.d. 15.0 in.

3. The discharge wires are hung vertically in the ESP and are held taut and plumb at the bottom by:

a. A 25-lb weightb. Two 25-lb weightsc. A 50-lb weightd. A 5-lb weight

4. True or False? Accumulated dust can be removed from discharge and collection electrodes by rap-ping.

5. European precipitators and most new U.S.-designed ESPs use ____________________ for dis-charge electrodes.

a. Wiresb. Rigid framesc. Plates with stiffeners

6. Normal spacing for plates used on high-efficiency wire/plate ESPs is generally:

a. 0.2 to 0.8 in.b. 2 to 4 in.c. 8 to 9 in.d. 24 to 36 in.

7. Normal spacing for plates used on high-efficiency rigid-frame ESPs is generally:

a. 2-4 in.b. 5-7 in.c. 8-9 in.d. 12-15 in.

Lesson 2

2-20 2.0-2/98

8. In ESPs, plates are usually between ____________________ high.

a. 4 to 12 in.b. 20 to 40 ftc. 40 to 60 ft

9. Collection electrodes can be constructed in two general shapes: ____________________ and____________________.

10. Collected dust is removed from tubular ESPs using:

a. Magnetic impulse rappersb. Water spraysc. Hammer and anvil rappersd. Electric vibrator rappers

11. ESPs control the strength of the electric field generated between the discharge and collection elec-trodes by using:

a. Transformer-rectifier setsb. Metersc. Capacitorsd. Insulators

12. In a T-R set, the transformer ____________________ while the rectifier____________________.

a. Steps down the voltage, converts direct current into alternating currentb. Converts alternating current into direct current, steps up the voltagec. Steps up the voltage, converts alternating current into direct current

13. In the control circuitry on an ESP, the primary voltmeter measures the:

a. Number of sparksb. Input voltage (in a.c. volts) coming into the transformerc. Output voltage from the rectifierd. Operating d.c. voltage delivered to the discharge electrodes

14. The combination of the ____________________ voltage and current readings gives the powerinput to the discharge electrodes.

a. Primaryb. Sparkingc. Secondaryd. Tertiary

15. An electric cable that carries high voltage from the T-R set to the discharge electrode is calleda(an):

a. Bus lineb. Pipec. Ductd. Electric vibrator


2.0-2/98 2-21

16. Most precipitators use ____________________ or ____________________ to remove accumu-lated dust from collection plates.

a. Air-vibrator rappers (or) water spraysb. Hammer and anvil (or) magnetic-impulse rappersc. Electric-vibrator (or) magnetic-impulse rappers

17. Which rappers are commonly used for removing dust from discharge electrodes?

a. Hammerb. Electric-vibrator and tumbling-hammerc. Magnetic-impulsed. Water-spray

18. The dust is temporarily stored in a ____________________.

19. A ____________________ ____________________ discharge device works similarly to arevolving door.

20. A ____________________ ____________________ uses a screw feeder located at the bottom ofthe hopper to remove dust from the bin.

21. A ____________________ ____________________ uses a blower or compressed air to removedust from the hopper.

22. A ____________________ ____________________ uses paddles or flaps connected to a dragchain to move dust from the ESP to its final destination.

23. In a precipitator, shells and hoppers should be covered with ____________________ to conserveheat and prevent corrosion.

Lesson 2

2-22 2.0-2/98


2.0-2/98 2-23


1. discharge electrodescollection electrodeshigh voltage electrical systemsrappershoppersshellThe six major components of an ESP are discharge electrodes, collection electrodes, high voltageelectrical systems, rappers, hoppers, and the shell.

2. b. 0.1 in.In many U.S. precipitators, the discharge electrodes are thin wires that are approximately 0.1 inchin diameter.

3. a. A 25-lb weightThe discharge wires are hung vertically in the ESP and are held taut and plumb at the bottom by a25-lb weight.

4. TrueAccumulated dust can be removed from discharge and collection electrodes by rapping.

5. b. Rigid framesEuropean precipitators and most new U.S.-designed ESPs use rigid frames for discharge elec-trodes.

6. c. 8 to 9 in.Normal spacing for plates used on high-efficiency wire/plate ESPs is generally 8 to 9 inches.

7. d. 12 to 15 in.Normal spacing for plates used on high-efficiency rigid-frame ESPs is generally 12 to 15 inches.

8. b. 20 to 40 ftIn ESPs, plates are usually between 20 to 40 ft high.

9. PlatesTubesCollection electrodes can be constructed in two general shapes: plates and tubes.

10. b. Water spraysCollected dust is removed from tubular ESPs using water sprays.

11. a. Transformer-rectifier setsESPs control the strength of the electric field generated between the discharge and collection elec-trodes by using transformer-rectifier sets.

Lesson 2

2-24 2.0-2/98

12. c. Steps up the voltage, converts alternating current into direct currentIn a T-R set, the transformer steps up the voltage while the rectifier converts alternating currentinto direct current.

13. b. Input voltage (in a.c. volts) coming into the transformerIn the control circuitry on an ESP, the primary voltmeter measures the input voltage (in a.c. volts)coming into the transformer.

14. c. SecondaryThe combination of the secondary voltage and current readings gives the power input to the dis-charge electrodes.

15. a. Bus lineAn electric cable that carries high voltage from the T-R set to the discharge electrode is called abus line.

16. b. Hammer and anvil (or) magnetic-impulse rappersMost precipitators use hammer and anvil or magnetic-impulse rappers to remove accumulated dustfrom collection plates.

17. b. Electric-vibrator and tumbling-hammerFor removing dust from discharge electrodes, electric-vibrator rappers (for wires) and tumbling-hammer rappers (for rigid frames) are commonly used.

18. HopperThe dust is temporarily stored in a hopper.

19. Rotary airlockA rotary airlock discharge device works similarly to a revolving door.

20. Screw conveyorA screw conveyor uses a screw feeder located at the bottom of the hopper to remove dust from thebin.

21. Pneumatic conveyorA pneumatic conveyor uses a blower or compressed air to remove dust from the hopper.

22. Drag conveyorA drag conveyor uses paddles or flaps connected to a drag chain to move dust from the ESP to itsfinal destination.

23. InsulationIn a precipitator, shells and hoppers should be covered with insulation to conserve heat and preventcorrosion.

2.0-2/98 2-25

Bibliography

Beachler, D. S., J. A. Jahnke, G. T. Joseph and M. M. Peterson. 1983. Air Pollution Control Systemsfor Selected Industries, Self-Instructional Guidebook. (APTI Course SI:431). EPA 450/2-82-006.U.S. Environmental Protection Agency.

Bethea, R. M. 1978. Air Pollution Control Technology-an Engineering Analysis Point of View. NewYork: Van Nostrand Reinhold.

Cheremisinoff, P. N., and R. A. Young. (Eds.) 1977. Air Pollution Control and Design Handbook,Part 1. New York: Marcel Dekker.

Gallaer, C. A. 1983. Electrostatic Precipitator Reference Manual. Electric Power Research Institute.EPRI CS-2809, Project 1402-4.

Hall, H. J. 1975. Design and application of high voltage power supplies in electrostatic precipitation.Journal of Air Pollution Control Association. 25:132.

Hesketh, H. E. 1979. Air Pollution Control. Ann Arbor: Ann Arbor Science Publishers.


Richards, J.R. 1995. Control of Particulate Emissions, Student Manual. (APTI Course 413). U.S.Environmental Protection Agency.

Szabo, M. F., Y. M. Shah, and S. P. Schliesser. 1981. Inspection Manual for Evaluation of Electro-static Precipitator Performances. EPA 340/1-79-007.

Turner, J. H., P. A. Lawless, T. Yamamoto, D. W. Coy, G. P. Greiner, J. D. McKenna, and W. M. Vata-vuk. 1992. Electrostatic precipitators. In A. J. Buonicore and W. T. Davis (Eds.), Air PollutionEngineering Manual (pp. 89-113). Air and Waste Management Association. New York: Van Nos-trand Reinhold.


White, H. J. 1963. Industrial Electrostatic Precipitation. Reading, MA: Addison-Wesley.White, H. J.1977. Electrostatic precipitation of fly ash. APCA Reprint Series. Journal of Air Pollution Con-trol Association. Pittsburgh, PA.

Lesson 2

2-26 2.0-2/98

2.0-2/98 3-1

Lesson 3ESP Design Parameters andTheir Effects on Collection Efficiency

Goal

To familiarize you with the variables used by vendors to optimally design ESP systems.

Objectives


1. Define the term migration velocity2. Explain the difference between the Deutsch-Anderson equation and the Matts-Ohnfeldt equa-

tion for estimating collection efficiency3. Define the term resistivity4. List three ways to reduce high resistivity and two ways to combat low resistivity5. Explain how sectionalization and increasing corona power improves collection efficiency6. Define aspect ratio and specific collection area and describe their importance for achieving

collection efficiency7. Calculate the aspect ratio and specific collection area of an ESP given a set of design informa-

tion

Introduction

Because of legislation such as the Clean Air Act and the 1977 and 1990 Clean Air Act Amend-ments, ESPs have been carefully designed to collect more than 99.5% of particles in the fluegas from many industries. ESPs efficiently collect particles of various sizes: large particles of3 to 10 µm in diameter, and smaller particles of less than 1 µm in diameter.

An ESP is designed for a particular industrial application. Building an ESP is a costlyendeavor, so a great deal of time and effort is expended during the design stage. Manufacturersuse various methods to design ESPs. They also consider a variety of operating parameters thataffect collection efficiency including resistivity, electrical sectionalization, specific collectionarea, aspect ratio, gas flow distribution, and corona power. This lesson focuses on these meth-ods and operating parameters.

Lesson 3

3-2 2.0-2/98

Design Methods

Manufacturers use mathematical equations to estimate collection efficiency or collection area.In addition, they may build a pilot-plant to determine the parameters necessary to build thefull-scale ESP. They may also use a mathematical model or computer program to test thedesign features and operating parameters in a simulation of the final design. Once the basis ofthe ESP design is completed, the vendor can design the unit using various individual parame-ters that are appropriate for each specific situation.

Using Estimates of Collection Efficiency

Collection efficiency is the primary consideration of ESP design. The collection effi-ciency and/or the collection area of an ESP can be estimated using several equations.These equations give a theoretical estimate of the overall collection efficiency of the unitoperating under ideal conditions. Unfortunately, a number of operating parameters canadversely affect the collection efficiency of the precipitator. A discussion of collection-efficiency equations and operating parameters affecting collection-efficiency equationsfollows.

Particle-Migration VelocityBefore determining the collection area and the collection efficiency, the designer mustestimate or measure (if possible) the particle-migration velocity. This is the speed atwhich a particle, once charged, migrates toward the grounded collection electrode.Variables affecting particle velocity are particle size, the strength of the electric field,and the viscosity of the gas. How readily the charged particles move to the collectionelectrode is denoted by the symbol, w, called the particle-migration velocity, or driftvelocity. The migration-velocity parameter represents the collectability of the particlewithin the confines of a specific ESP. The migration velocity is expressed in Equation3-1.

(3-1)

Where: dp = diameter of the particle, µmEo = strength of field in which particles are charged

(represented by peak voltage), V/m (V/ft)Ep = strength of field in which particles are collected

(normally the field close to the collecting plates), V/m (V/ft)µ = gas viscosity, Pa • s (cp)π = 3.14

As shown in Equation 3-1, migration velocity depends on the voltage strength of boththe charging and collection fields. Therefore, the precipitator must be designed usingthe maximum electric field voltage for maximum collection efficiency. The migrationvelocity also depends on particle size; larger particles are collected more easily thansmaller ones.

w =d E Ep o p

4πµ

ESP Design Parameters and Their Effects on Collection Efficiency

2.0-2/98 3-3

Particle-migration velocity can also be determined by Equation 3-2.

(3-2)

Where: q = particle charge(s)Ep = strength of field in which particles are collected, V/m (V/ft)µ = gas viscosity, Pa • s (cp)r = radius of the particle, µmπ = 3.14

The particle-migration velocity can be calculated using either Equations 3-1 or 3-2,depending on the information available on the particle size and electric field strength.However, most ESPs are designed using a particle-migration velocity based on fieldexperience rather than theory. Typical particle migration velocity rates, such as thoselisted in Table 3-1, have been published by various ESP vendors.

Table 3-1. Typical effective particle-migrationvelocity rates for various applications

ApplicationMigration velocity

(ft/sec) (cm/s)

Utility fly ashPulverized coal fly ashPulp and paper millsSulfuric acid mistCement (wet process)Cement (dry process)GypsumSmelterOpen-hearth furnaceBlast furnaceHot phosphorousFlash roasterMultiple-hearth roasterCatalyst dustCupola

0.13-0.670.33-0.440.21-0.310.19-0.250.33-0.370.19-0.230.52-0.640.060.16-0.190.20-0.460.090.250.260.250.10-0.12

4.0-20.410.1-13.46.4-9.55.8-7.6210.1-11.36.4-7.015.8-19.51.84.9-5.86.1-14.02.77.67.97.63.0-3.7

Sources: Theodore and Buonicore 1976; U.S. EPA 1979.

w =qE

rp

6πµ

Lesson 3

3-4 2.0-2/98

Deutsch-Anderson EquationProbably the best way to gain insight into the process of electrostatic precipitation isto study the relationship known as the Deutsch-Anderson equation. This equation isused to determine the collection efficiency of the precipitator under ideal conditions.The simplest form of the equation is given below.

(3-3)

Where: η = collection efficiency of the precipitatore = base of natural logarithm = 2.718w = migration velocity, cm/s (ft/sec)A = the effective collecting plate area of the precipitator, m2 (ft2)Q = gas flow through the precipitator, m3/s (ft3/sec)

_____________________Source: Deutsch 1922; Anderson 1924.

This equation has been used extensively for many years to calculate theoretical collec-tion efficiencies. Unfortunately, while the equation is scientifically valid, a number ofoperating parameters can cause the results to be in error by a factor of 2 or more. TheDeutsch-Anderson equation neglects three significant process variables. First, it com-pletely ignores the fact that dust reentrainment may occur during the rapping process.Second, it assumes that the particle size and, consequently, the migration velocity areuniform for all particles in the gas stream. As stated previously, this is not true; largerparticles generally have higher migration velocity rates than smaller particles do.Third, it assumes that the gas flow rate is uniform everywhere across the precipitatorand that particle sneakage (particles escape capture) through the hopper section doesnot occur. Particle sneakage can occur when the flue gas flows down through the hop-per section instead of through the ESP chambers, thus preventing particles from beingsubjected to the electric field. Therefore, this equation should be used only for makingpreliminary estimates of precipitator collection efficiency.

More accurate estimates of collection efficiency can be obtained by modifying theDeutsch-Anderson equation. This is accomplished either by substituting the effectiveprecipitation rate, we, in place of the migration velocity, w, or by decreasing the cal-culation of collection efficiency by a factor of k, which is constant (Matts-Ohnfeldtequation). These calculations are used in establishing preliminary design parametersof ESPs.

Modified Deutsch-Anderson EquationUsing the Effective-Precipitation Rate

To make the Deutsch-Anderson equation more accurate in cases where all particlesare not uniform in size, a parameter called the effective precipitation rate (we) can besubstituted for the migration velocity in the equation. Therefore, Dr. Harry White pro-posed modifying the Deutsch-Anderson equation by using the term we instead of w inthe Deutsch-Anderson equation (White 1982).

η 1 e w A Q⁄( )––=


2.0-2/98 3-5

(3-4)

Where: η = collection efficiency of the precipitatore = base of natural logarithm = 2.718we = effective migration velocity, calculated from field experienceA = collecting area, m2 (ft2)Q = gas flow rate, m3/s (ft3/sec)

In contrast to the migration velocity (w), which refers to the speed at which an indi-vidual charged particle migrates to the collection electrode, the effective precipitationrate (we) refers to the average speed at which all particles in the entire dust mass movetoward the collection electrode. The variable, we, is calculated from field experiencerather than from theory; values for we are usually determined using data banks accu-mulated from ESP installations in similar industries or from pilot-plant studies. Insummary, the effective precipitation rate represents a semi-empirical parameter thatcan be used to determine the total collection area necessary for an ESP to achieve aspecified collection efficiency required to meet an emission limit.

Using the Deutsch-Anderson equation in this manner could be particularly usefulwhen trying to determine the amount of additional collection area needed to upgradean existing ESP to meet more stringent regulations or to improve the performance ofthe unit. However, other operating parameters besides collection area play a majorrole in determining the efficiency of an ESP.

Matts-Ohnfeldt EquationAnother modification to the Deutsch-Anderson equation that accounts for non-idealeffects was devised by Sigvard Matts and Per-Olaf Ohnfeldt of Sweden (SvenskaFlaktfabriken) in 1964. The Matts-Ohnfeldt equation is

(3-5)

Where: η = collection efficiency of the precipitatore = base of natural logarithm = 2.718wk = average migration velocity, cm/s (ft/sec)k = a constant, usually 0.4 to 0.6A = collection area, m2 (ft2)Q = gas flow rate, m3/s (ft3/sec)

The term, wk, the average migration velocity in equation 3-5, is determined frominformation obtained from similar installations. The terms wk and we (in equations 3-5and 3-4 respectively) are similar in that both are average migration velocities. Theconstant, k, in the equation is usually between 0.4 and 0.6, depending on the standarddeviation of the particle size distribution and other dust properties affecting collectionefficiency. However, most people who have used this equation report that a value of kequal to 0.5 gives satisfactory results (Gallaer 1983 and U.S. EPA 1985). In an Elec-tric Power Research Institute (EPRI) study, a table was constructed to show the rela-tionship of predicting collection efficiency using the Deutsch-Anderson and Matts-Ohnfeldt equations. This information is given in Table 3-2.

η 1 ewe A Q⁄( )–

–=

η 1 ewk A Q⁄( )k–

–=

Lesson 3

3-6 2.0-2/98

When k = 1.0, the Matts-Ohnfeldt equation is the same as the Deutsch-Andersonequation. To predict the collection efficiency of an existing ESP when the collectionarea or gas flow rate is varied, using lower values for k gives more conservativeresults. From Table 3-2, you can see that the efficiency estimates calculated using theMatts-Ohnfeldt equation are more conservative than those estimated using theDeutsch-Anderson equation, and may more likely predict how efficiently the ESP willactually operate.

Using Pilot Plants

Probably the most reliable method for designing ESPs is to construct and operate a pilotplant. However, time limitations and the expense of construction may make this impossi-ble; a pilot plant can easily cost one million dollars or more. A pilot ESP project can beconstructed on an existing industrial process. In this case, a side stream of flue gas is sentto the small pilot ESP. Flue gas sampling gives valuable information such as gas tempera-ture, moisture content, and dust resistivity. Relating these parameters to the measured col-lection efficiency of the pilot project will help the design engineers plan for scale-up to afull-sized ESP.

Using Computer Programs and Models

Engineers can also use mathematical models or computer programs to design precipita-tors. A mathematical model that relates collection efficiency to precipitator size and vari-ous operating parameters has been developed by Southern Research Institute (SoRI) forEPA. The (SoRI/EPA) model is used to do the following:

• Design a full-scale ESP from fundamental principles or in conjunction with a pilot-plant study·

• Evaluate ESP bids submitted by various manufacturers

• Troubleshoot and diagnose operating problems for existing ESPs

• Evaluate the effectiveness of new ESP developments and technology, such as flue gasconditioning and pulse energizing.

Table 3-2. Collection-efficiency estimationsusing the Deutsch-Anderson andMatts-Ohnfeldt equations

Relative sizeof ESP (A/Q)

Deutschk = 1.0

Matts-Ohnfeldt

k = 0.4 k = 0.5 k = 0.6

12345

909999.999.9999.999

9095.197.298.198.7

9096.298.19999.6

9097.298.899.599.76

Source: Gallaer 1983.


2.0-2/98 3-7

Details of this model are given in EPA publications A Mathematical Model of ElectrostaticPrecipitation (Revision 1), Volumes I and II.

Table 3-3 lists the input data used in the SoRI/EPA Model. Assuming that accurate inputdata are available for use, the model usually can estimate emissions within ± 20 percent ofmeasured values (U.S. EPA 1985). The computer model goes through an iterative compu-tational process to refine its predictions of emission levels for a particular ESP. First, themodel uses secondary voltage and current levels (corona power) to predict emission levelsleaving the ESP. Then, actual emission levels are measured and compared to the predictedemission levels. Empirical factors are then adjusted and the process repeats itself until thepredicted emission levels of the model agree with the actual, measured levels. This modelcan be used to obtain reasonable estimates of emission levels for other ESP operating con-ditions (U.S. EPA 1985). For example, once you create a good, working computer modelfor a particular ESP design under one set of operating conditions, you can run the modelfor different scenarios by altering one or more of the parameters (precipitator length, num-ber of fields, etc.) to obtain reasonably accurate emission level predictions.

Table 3-3. Input data for EPA/SORI ESP computermodel

ESP Specifications Gas/particulatespecifications

Estimated efficiencyPrecipitator lengthSuperficial gas velocityFraction of sneakage/reentrainmentNormalized standard deviation of gas velocity

distributionNumber of stages for sneakage/reentrainmentNumber of electrical sections in direction of gas

flowFor each electrical section

LengthAreaApplied voltageCurrentCorona wire radiusCorona wire lengthWire-to-wire spacing (1/2)Wire-to-plate spacingNumber of wires per linear section

Gas flow rateGas pressureGas temperatureGas viscosityParticulate concentrationParticulate resistivityParticulate densityParticle size distributionDielectric constantIon speed

Source: U.S. EPA 1985.

Lesson 3

3-8 2.0-2/98

Another model, the EPA/RTI model, has been developed by the Research Triangle Insti-tute (RTI) for EPA (Lawless 1992). The EPA/RTI model is based on the localized electricfield strengths and current densities prevailing throughout the precipitator. These data canbe input based on actual readings from operating units, or can be calculated based on elec-trode spacing and resistivity. The data are used to estimate the combined electrical charg-ing on each particle size range due to field-dependent charging and diffusional charging.Particle size-dependent migration velocities are then used in a Deutsch-Anderson typeequation to estimate particle collection in each field of the precipitator. This model takesinto account a number of the site specific factors including gas flow maldistribution, parti-cle size distribution, and rapping reentrainment.

These performance models require detailed information concerning the anticipated config-uration of the precipitator and the gas stream characteristics. Information needed to oper-ate the EPA/RTI model is provided below. It is readily apparent that all of these parametersare not needed in each case, since some can be calculated from the others. The followingdata is data utilized in the EPA/RTI computerized performance model for electrostatic pre-cipitators.

ESP Design

• Specific collection area

• Collection plate area

• Collection height and length

• Gas velocity

• Number of fields in series

• Number of discharge electrodes

• Type of discharge electrodes

• Discharge electrode-to-collection plate spacing

Particulate Matter and Gas Stream Data

• Resistivity

• Particle size mass median diameter

• Particle size distribution standard deviation

• Gas flow rate distribution standard deviation

• Actual gas flow rate

• Gas stream temperature

• Gas stream pressure

• Gas stream composition


2.0-2/98 3-9

Design Parameters

Once the basis of the ESP design has been set, the vendor will complete the design by incorpo-rating a number of parameters that can be adjusted for each specific industrial application.However, before starting this design phase, the vendor must take into account the effect thatparticle resistivity can have on the actual collection efficiency.

Resistivity

Resistivity, which is a characteristic of particles in an electric field, is a measure of a parti-cle's resistance to transferring charge (both accepting and giving up charges). Resistivity isa function of a particle's chemical composition as well as flue gas operating conditionssuch as temperature and moisture. Particles can have high, moderate (normal), or lowresistivity.

In an ESP, where particle charging and discharging are key functions, resistivity is animportant factor that significantly affects collection efficiency. While resistivity is animportant phenomenon in the inter-electrode region where most particle charging takesplace, it has a particularly important effect on the dust layer at the collection electrodewhere discharging occurs. Particles that exhibit high resistivity are difficult to charge. Butonce charged, they do not readily give up their acquired charge on arrival at the collectionelectrode. On the other hand, particles with low resistivity easily become charged andreadily release their charge to the grounded collection plate. Both extremes in resistivityimpede the efficient functioning of ESPs. ESPs work best under normal resistivity condi-tions.

Resistivity is the electrical resistance of a dust sample 1.0 cm2 in cross-sectional area, 1.0cm thick, and is recorded in units of ohm-cm. A method for measuring resistivity will bedescribed later in this lesson. Table 3-4 gives value ranges for low, normal, and high resis-tivity.

Dust Layer ResistivityLet’s take a closer look at the way resistivity affects electrical conditions in the dustlayer. A potential electric field (voltage drop) is formed across the dust layer as nega-tively charged particles arrive at the dust layer surface and leak their electrical chargesto the collection plate. At the metal surface of the electrically grounded collectionplate, the voltage is zero. Whereas at the outer surface of the dust layer, where newparticles and ions are arriving, the electrostatic voltage caused by the gas ions can bequite high. The strength of this electric field depends on the resistivity and thicknessof the dust layer.

Table 3-4. Low, normal, and high resistivity

Resistivity Range of measurement

Low

Normal

High

between 104 and 107 ohm • cm

between 107 and 1010 ohm • cmabove 1010 ohm • cm

(usually between 1010 and 1014 ohm • cm)

Lesson 3

3-10 2.0-2/98

In high resistivity dust layers, the dust is not sufficiently conductive, so electricalcharges have difficulty moving through the dust layer. Consequently, electricalcharges accumulate on and beneath the dust layer surface, creating a strong electricfield. Voltages can be greater than 10,000 volts. Dust particles with high resistivitiesare held too strongly to the plate, making them difficult to remove and causing rappingproblems.

In low resistivity dust layers, the corona current is readily passed to the grounded col-lection electrode. Therefore, a relatively weak electric field, of several thousand volts,is maintained across the dust layer. Collected dust particles with low resistivity do notadhere strongly enough to the collection plate. They are easily dislodged and becomereentrained in the gas stream.

The following discussion of normal, high, and low resistivity applies to ESPs operatedin a dry state; resistivity is not a problem in the operation of wet ESPs because of themoisture concentration in the ESP. The relationship between moisture content andresistivity is explained later in this lesson.

Normal ResistivityAs stated above, ESPs work best under normal resistivity conditions. Particles withnormal resistivity do not rapidly lose their charge on arrival at the collection electrode.These particles slowly leak their charge to grounded plates and are retained on the col-lection plates by intermolecular adhesive and cohesive forces. This allows a particu-late layer to be built up and then dislodged from the plates by rapping. Within therange of normal dust resistivity (between 107 and 1010 ohm-cm), fly ash is collectedmore easily than dust having either low or high resistivity.

High ResistivityIf the voltage drop across the dust layer becomes too high, several adverse effects canoccur. First, the high voltage drop reduces the voltage difference between the dis-charge electrode and collection electrode, and thereby reduces the electrostatic fieldstrength used to drive the gas ion - charged particles over to the collected dust layer.As the dust layer builds up, and the electrical charges accumulate on the surface of thedust layer, the voltage difference between the discharge and collection electrodesdecreases. The migration velocities of small particles are especially affected by thereduced electric field strength.

Another problem that occurs with high resistivity dust layers is called back corona.This occurs when the potential drop across the dust layer is so great that corona dis-charges begin to appear in the gas that is trapped within the dust layer. The dust layerbreaks down electrically, producing small holes or craters from which back coronadischarges occur. Positive gas ions are generated within the dust layer and are acceler-ated toward the "negatively charged" discharge electrode. The positive ions reducesome of the negative charges on the dust layer and neutralize some of the negativeions on the "charged particles" heading toward the collection electrode. Disruptions ofthe normal corona process greatly reduce the ESP's collection efficiency, which insevere cases, may fall below 50% (White 1974).


2.0-2/98 3-11

The third, and generally most common problem with high resistivity dust is increasedelectrical sparking. When the sparking rate exceeds the "set spark rate limit," the auto-matic controllers limit the operating voltage of the field. This causes reduced particlecharging and reduced migration velocities toward the collection electrode.

High resistivity can generally be reduced by doing the following:

• Adjusting the temperature

• Increasing moisture content

• Adding conditioning agents to the gas stream

• Increasing the collection surface area

• Using hot-side precipitators (occasionally)

Figure 3-1 shows the variation in resistivity with changing gas temperature for six dif-ferent industrial dusts (U.S. EPA 1985). For most dusts, resistivity will decrease as theflue gas temperature increases. However, as can be seen from Figure 3-1, the resistiv-ity also decreases for some dusts (cement and ZnO) at low flue gas temperatures.

Figure 3-1. Resistivity of six different dusts at varioustemperaturesSource: U.S. EPA 1985.

Lesson 3

3-12 2.0-2/98

The moisture content of the flue gas stream also affects particle resistivity. Increasingthe moisture content of the gas stream by spraying water or injecting steam into theduct work preceding the ESP lowers the resistivity. In both temperature adjustmentand moisture conditioning, one must maintain gas conditions above the dew point toprevent corrosion problems in the ESP or downstream equipment. Figure 3-2 showsthe effect of temperature and moisture on the resistivity of cement dust. As the per-centage of moisture in the dust increases from 1 to 20%, the resistivity of the dust dra-matically decreases. Also, raising or lowering the temperature can decrease cementdust resistivity for all the moisture percentages represented.

Figure 3-2. Effect of temperature and moisture on theresistivity of cement dustSources: Schmidt 1949, White 1977.

The presence of SO3 in the gas stream has been shown to favor the electrostatic pre-cipitation process when problems with high resistivity occur. Most of the sulfur con-tent in the coal burned for combustion sources converts to SO2. However,approximately 1% of the sulfur converts to SO3. The amount of SO3 in the flue gasnormally increases with increasing sulfur content of the coal. The resistivity of theparticles decreases as the sulfur content of the coal increases (Figure 3-3).


2.0-2/98 3-13

Figure 3-3. Fly ash resistivity versus coal sulfur contentfor several flue gas temperature bandsSource: White 1977.

The use of low-sulfur western coal for boiler operations has caused fly ash resistivityproblems for ESP operators. For coal fly ash dusts, the resistivity can be loweredbelow the critical level by the injection of as little as 10 to 30 ppm SO3 into the gasstream. The SO3 is injected into the duct work preceding the precipitator. Figure 3-4shows the flow diagram of a sulfur-burning flue gas conditioning system used tolower resistivity at a coal-fired boiler.

Figure 3-4. Flow diagram of sulfur-burning flue gas conditioning systemCourtesy of Wahlco, Inc.

Liquidsulfur

storageMetering

pump

Ambientairin

Liquid sulfur

250° - 300°F

Airheater

Controlled to800° - 825°F

Sulfurburner

Converter

Air/SO3

800° - 1100°FBoiler

flue

Injectionprobes

Conditionedflue gas toprecipitator

Lesson 3

3-14 2.0-2/98

Other conditioning agents, such as sulfuric acid, ammonia, sodium chloride, and sodaash, have also been used to reduce particle resistivity (White 1974). Therefore, thechemical composition of the flue gas stream is important with regard to the resistivityof the particles to be collected in the ESP. Table 3-5 lists various conditioning agentsand their mechanisms of operation (U.S. EPA 1985).

.

Two other methods that reduce particle resistivity include increasing the collectionsurface area and handling the flue gas at higher temperatures. Increasing the collectionarea of the precipitator will increase the overall cost of the ESP, which may not bedesirable. Hot-side precipitators, which are usually located in front of the combustionair preheater section of the boiler, are also used to combat resistivity problems. How-ever, the use of conditioning agents has been more successful and very few hot-sideESPs have been installed since the 1980s.

Table 3-5. Reaction mechanisms of majorconditioning agents

Conditioning agent Mechanism(s) of action

Sulfur trioxide andsulfuric acid

Ammonia

Ammonium sulfate1

Triethylamine

Sodium compounds

Compounds of transitionmetals

Potassium sulfate andsodium chloride

Condensation and adsorption on fly ash surfaces;may also increase cohesiveness of fly ash.

Reduces resistivity.

Mechanism is not clear; various ones proposed:Modifies resistivityIncreases ash cohesivenessEnhances space charge effect

Little is known about the actual mechanism; claimsare made for the following:Modifies resistivity (depends upon injection

temperature)Increases ash cohesivenessEnhances space charge effect

Experimental data lacking to substantiate which ofthese is predominant

Particle agglomeration claimed; no supporting data

Natural conditioner if added with coal.Resistivity modifier if injected into gas stream

Postulated that they catalyze oxidation of SO2 toSO3; no definitive tests with fly ash to verify thispostulation

In cement and lime kiln ESPs:Resistivity modifiers in the gas streamNaCl - natural conditioner when mixed with coal

1 If injection occurs at a temperature greater than about 600°F, dissociation into ammonia and sulfurtrioxide results. Depending upon the ash, SO2 may preferentially interact with fly ash as SO3

conditioning. The remainder recombines with ammonia to add to the space charge as well asincrease the cohesiveness of the ash.



2.0-2/98 3-15

Low ResistivityParticles that have low resistivity are difficult to collect because they are easilycharged (very conductive) and rapidly lose their charge on arrival at the collectionelectrode. The particles take on the charge of the collection electrode, bounce off theplates, and become reentrained in the gas stream. Thus, attractive and repulsive elec-trical forces that are normally at work at higher resistivities are lacking, and the bind-ing forces to the plate are considerably lessened. Examples of low-resistivity dusts areunburned carbon in fly ash and carbon black.

If these conductive particles are coarse, they can be removed upstream of the precipi-tator by using a device such as a cyclone. Baffles are often installed on the collectionplates to help eliminate this precipitation-repulsion phenomenon.

The addition of liquid ammonia (NH3) into the gas stream as a conditioning agent hasfound wide use in recent years. It is theorized that ammonia reacts with H2SO4 con-tained in the flue gas to form an ammonium sulfate compound that increases the resis-tivity of the dust. Ammonia vapor is injected into the duct leading to the precipitator atconcentrations of 15 to 40 ppm by volume. The injection of NH3 has improved theresistivity of fly ash from coal-fired boilers with low flue gas temperatures (Katz1979).

Table 3-6 summarized the characteristics associated with low, normal and high resis-tivity dusts.

Measuring ResistivityParticle resistivity is determined by measuring the leakage current through a dust layerto which a high voltage is applied using conductivity cells. A number of conductivitycells have been used in particle-resistivity measurements. For a good review of thedifferent kinds of cells employed, see White (1974). Resistivity can be measured by anumber of methods in either the laboratory or the field. In the lab method, dust sam-ples are first extracted from the flue gas leaving the industrial process and collected ona filter as described in EPA Reference Method 5. The samples are then taken back tothe laboratory and analyzed.

Resistivity measurements are made in the field using an in-situ resistivity probe. Theprobe is inserted into the duct leaving the industrial process and a dust sample isextracted into the probe. High voltage is applied across a point and plate electrode sys-tem inside the probe. Particles are charged and then collected on the plate. After a suf-ficiently thick layer of dust has collected on the plate, the power to the point is turnedoff and a disc is lowered onto the collected dust sample. The thickness of the dustlayer is first measured. Increasing voltages are then applied to the disc, and the corre-sponding current is recorded until the dust layer breaks down and sparkover occurs.The resistivity is calculated from the last set of voltage and current readings obtainedbefore sparkover occurs. Since these resistivity measurements are made at the indus-trial process conditions, these data are generally more useful than data obtained fromthe laboratory methods. A good review of in-situ resistivity measuring techniques isgiven by White (1974) and Gallaer (1983).

Lesson 3

3-16 2.0-2/98

Electrical Sectionalization

Field SectionalizationAn electrostatic precipitator is divided into a series of independently energized bussections or fields (also called stages) in the direction of the gas flow. Precipitator per-formance depends on the number of individual bus sections, or fields, installed. Figure3-5 shows an ESP consisting of four fields, each of which acts as an independent pre-cipitator.

Table 3-6. ESP characterististics associated withdifferent levels of resistivity

Resistivity Level,ohm-cm ESP Characteristics

Less than 107

(Low Resistivity)1. Normal operating voltage and current levels

unless dust layer is thick enough to reduce plateclearances and cause higher current levels

2. Reduced electrical force component retainingcollected dust, vulnerable to high reentrainmentlosses

3. Negligible voltage drop across dust layer4. Reduced collection performance due to (2)

107 to 1010

(Normal Resistivity)

1011

1. Normal operating voltage and current levels2. Negligible voltage drop across dust layer3. Sufficient electrical force component retaining

collected dust4. High collection performance due to (1), (2), and

(3)

1. Reduced operating voltage and current levelswith high spark rates

2. Significant voltage loss across dust layer3. Moderate electrical force component retaining

collected dust4. Reduced collection performance due to (1) and

(2)

Greater than 1012

(High Resistivity)1. Reduced operating voltage levels; high operating

current levels if power supply controller is notoperating properly

2. Very significant voltage loss across dust layer3. High electrical force component retaining

collected dust4. Seriously reduced collection performance due to

(1), (2), and probable back corona

Typical values

Operating voltage: 30 to 70 kV, dependent on design factorsOperating current density: 5 to 50 nA/cm2

Dust layer thickness: 1/4 to 1 inch

Source: Adapted from U.S. EPA 1985.


2.0-2/98 3-17

Figure 3-5. Field sectionalization

Each field has individual transformer-rectifier sets, voltage-stabilization controls, andhigh-voltage conductors that energize the discharge electrodes within the field. Thisdesign feature, called field electrical sectionalization, allows greater flexibility forenergizing individual fields to accommodate different conditions within the precipita-tor. This is an important factor in promoting higher precipitator collection efficiency.Most ESP vendors recommend that there be at least three or more fields in the precip-itator. However, to attain a collection efficiency of more than 99%, some ESPs havebeen designed with as many as seven or more fields. Previous experience with a par-ticular industry is the best factor for determining how many fields are necessary tomeet the required emission limits.

The need for separate fields arises mainly because power input requirements differ atvarious locations within a precipitator. The maximum voltage at which a given fieldcan be maintained depends on the properties of the gas and dust being collected. Theparticulate matter concentration is generally high at the inlet fields of the precipitator.High dust concentrations tend to suppress corona current, requiring a great deal ofpower to generate corona discharge for optimum particle charging. In the downstreamfields of a precipitator, the dust loading is usually lighter, because most of the dust iscollected in the inlet fields. Consequently, corona current flows more freely in down-stream fields. Particle charging will more likely be limited by excessive sparking inthe downstream than in the inlet fields. If the precipitator had only one power set, theexcessive sparking would limit the power input to the entire precipitator, thus reduc-ing the overall collection efficiency. The rating of each power set in the ESP will varydepending on the specific design of the ESP.

Modern precipitators have voltage control devices that automatically limit precipitatorpower input. A well-designed automatic control system keeps the voltage level atapproximately the value needed for optimum particle charging by the discharge elec-trodes. The voltage control device increases the primary voltage applied to the T-R setto the maximum level. As the primary voltage applied to the transformer increases, thesecondary voltage applied to the discharge electrodes increases. As the secondaryvoltage is increased, the intensity and number of corona discharges increase. The volt-age is increased until any of the set limits (primary voltage, primary current, second-ary voltage, secondary current, or spark rate limits) is reached. Occurrence of a sparkcounteracts high ESP performance because it causes an immediate, short-term col-lapse of the precipitator electric field. Consequently, power that is applied to capture

Lesson 3

3-18 2.0-2/98

particles is used less efficiently. There is, however, an optimum sparking rate wherethe gains in particle charging are just offset by corona-current losses from sparkover.

Measurements on commercial precipitators have determined that the optimum spark-ing rate is between 50 and 150 sparks per minute per electrical section. The objectivein power control is to maintain corona power input at this optimum sparking rate bymomentarily reducing precipitator power whenever excessive sparking occurs.

Besides allowing for independent voltage control, another major reason for having anumber of fields in an ESP is that electrical failure may occur in one or more fields.Electrical failure may occur as a result of a number of events, such as over-filling hop-pers, discharge-wire breakage, or power supply failure. These failures are discussed inmore detail later in this course. ESPs having a greater number of fields are less depen-dent on the operation of all fields to achieve a high collection efficiency.

Parallel SectionalizationIn field sectionalization, the precipitator is designed with a single series of indepen-dent fields following one another consecutively. In parallel sectionalization, theseries of fields is electrically divided into two or more sections so that each field hasparallel components. Such divisions are referred to as chambers and each individualunit is called a cell. A precipitator such as the one shown in Figure 3-6 has two paral-lel sections (chambers), four fields, and eight cells. Each cell can be independentlyenergized by a bus line from its own separate transformer-rectifier set.

Figure 3-6. Parallel sectionalization (with two parallelsections, eight cells, and four fields)

One important reason for providing sectionalization across the width of the ESP is toprovide a means of handling varying levels of flue gas temperature, dust concentra-tion, and problems with gas flow distribution. When treating flue gas from a boiler, anESP may experience gas temperatures that vary from one side of the ESP to the other,especially if a rotary air preheater is used in the system. Since fly ash resistivity is afunction of the flue gas temperature, this temperature gradient may cause variations inthe electrical characteristics of the dust from one side of the ESP to the other. The gasflow into the ESP may also be stratified, causing varying gas velocities and dust con-centrations that can also affect the electrical characteristics of the dust. Buildingnumerous fields and cells into an ESP design can provide a means of coping with vari-

Chamber 1

CellChamber 2


2.0-2/98 3-19

ations in the flue gas. In addition, the more cells provided in an ESP, the greater thechance that the unit will operate at its designed collection efficiency.

Lesson 3

3-20 2.0-2/98

Specific Collection Area

The specific collection area (SCA) is defined as the ratio of collection surface area to thegas flow rate into the collector. This ratio represents the A/Q relationship in the Deutsch-Anderson equation and consequently is an important determinant of collection efficiency.The SCA is given in Equation 3-6.

(3-6)

Expressed in metric units,

Expressed in English units,

For example, if the total collection area of an ESP is 600,000 ft2 and the gas flow ratethrough the ESP is 1,000,000 ft3/min (acfm), the SCA is 600 ft2 per 1000 acfm as calcu-lated below.

Increases in the SCA of a precipitator design will, in most cases, increase the collectionefficiency of the precipitator. Most conservative designs call for an SCA of 20 to 25 m2

per 1000 m3/h (350 to 400 ft2 per 1000 acfm) to achieve collection efficiency of more than99.5%. The general range of SCA is between 11 and 45 m2 per 1000 m3/hr (200 and 800ft2 per 1000 acfm), depending on precipitator design conditions and desired collectionefficiency.

Aspect Ratio

The aspect ratio, which relates the length of an ESP to its height, is an important factor inreducing rapping loss (dust reentrainment). When particles are rapped from the electrodes,the gas flow carries the collected dust forward through the ESP until the dust reaches thehopper. Although the amount of time it takes for rapped particles to settle in the hoppers isshort (a matter of seconds), a large amount of "collected dust" can be reentrained in the gasflow and carried out of the ESP if the total effective length of the plates in the ESP is smallcompared to their effective height. For example, the time required for dust to fall from thetop of a 9.1-m plate (30-ft plate) is several seconds. Effective plate lengths must be at least10.7 to 12.2 m (35 to 40 ft) to prevent a large amount of "collected dust" from being car-ried out of the ESP before reaching the hopper.

SCAtotal collection surface

gas flow rate------------------------------------------------------=

SCAtotal collection surface in m2

1000 m3 h⁄---------------------------------------------------------------------=

SCAtotal collection surface in ft2

1000 ft3 min⁄--------------------------------------------------------------------=

SCA600,000 ft2

1000 (1000 acfm)-------------------------------------------=

600 ft2

1000 acfm---------------------------=


2.0-2/98 3-21

The aspect ratio is the ratio of the effective length to the effective height of the collectorsurface. The aspect ratio can be calculated using Equation 3-7.

(3-7)

The effective length of the collection surface is the sum of the plate lengths in each con-secutive field and the effective height is the height of the plates. For example, if an ESPhas four fields, each containing plates that are 10 feet long, the effective length is 40 feet.If the height of each plate is 30 feet, the aspect ratio is 1.33 as shown below:

Aspect ratios for ESPs range from 0.5 to 2.0. However, for high-efficiency ESPs (thosehaving collection efficiencies of > 99%), the aspect ratio should be greater than 1.0 (usu-ally 1.0 to 1.5) and in some installations may approach 2.0.

Gas Flow Distribution

Gas flow through the ESP chamber should be slow and evenly distributed through theunit. Gas velocity is reduced by the expansion, or diverging, section of the inlet plenum(Figure 3-7). The gas velocities in the duct leading into the ESP are generally between 12and 24 m/s (40 and 80 ft/sec). The gas velocity into the ESP must be reduced to0.6-2.4 m/s (2-8 ft/sec) for adequate particle collection. With aspect ratios of 1.5, the opti-mum gas velocity is generally between 1.5 and 1.8 m/s (5 and 6 ft/sec).

Figure 3-7. Gas inlet with perforated diffuser plates

AReffective length, m (ft)effective height, m (ft)------------------------------------------------------=

AR10 ft 10 ft 10 ft 10 ft+ + +

30 ft---------------------------------------------------------------=

40 ft30 ft-----------=

1.33=

Perforateddiffuserplates

Lesson 3

3-22 2.0-2/98

In order to use all of the discharge and collection electrodes across the entire width of theESP, the flue gas must be evenly distributed. The inlet plenum contains perforated open-ings, called diffuser plate openings to evenly distribute the gas flow into the chambersformed by the plates in the precipitator.

Corona Power

As stated previously, a strong electric field is needed for achieving high collection effi-ciency of dust particles. The strength of the field is based on the rating of the T-R set. Thecorona power is the power that energizes the discharge electrodes and thus creates thestrong electric field. The corona power used for precipitation is calculated by multiplyingthe secondary current by the secondary voltage and is expressed in units of watts. In ESPdesign specifications, the corona power is usually given in units of watts per 1000 m3/h(watts per 1000 acfm). Corona power expressed in units of watts/1000 acfm is also calledthe specific corona power. Corona power for any bus section of an ESP can be calculatedby the following approximate relation:

(3-8)

Where: Pc = corona power, wattsVp = peak voltage, voltsVm = minimum voltage, voltsIc = average corona current, amperes

As you can see, corona power increases as the voltage and/or current increases. The totalcorona power of the ESP is the sum of the corona power for all of the individual T-R sets.In an ESP, the collection efficiency is proportional to the amount of corona power suppliedto the unit, assuming the corona power is applied effectively (maintains a good sparkingrate).

(3-9)

Where: η = collection efficiencye = base of natural logarithm = 2.718k = a constant, usually between 0.5 and 0.7Pc/Q = corona power density in units of watts per 1000 m3/hr

(watts per 1000 acfm)

From equation 3-9, you can see that for a given exhaust flow rate, the collection efficiencywill increase as the corona power is increased. This efficiency will depend on the operat-ing conditions of the ESP and on whether the amount of power has been applied effec-tively. For high collection efficiency, corona power is usually between 59 and 295 wattsper 1000 m3/h (100 and 500 watts per 1000 acfm). Recent ESP installations have beendesigned to use as much as 470 to 530 watts per 1000 m3/h (800 to 900 watts per 1000acfm).

Pc 1 2⁄ Vp Vm+( )Ic=

η 1 ekPc Q⁄–

–∝


2.0-2/98 3-23

The terms current density and power density are also used to characterize the design of theESP. Current density is the secondary current supplied by the T-R set for the given platearea and expressed in units of mA/ft2 of plate area. Power density is the corona powersupplied to the plate area and is expressed in units of watts per ft2 of plate area.

The size of the individual power sets (T-R sets) in the ESP will vary depending on theirspecific location and the conditions of the flue gas such as particle size, dust concentra-tion, dust resistivity, and flue gas temperature. In an ESP, the T-R sets are selected to pro-vide lower current density at the inlet sections, where the dust concentration will tend tosuppress the corona current, and to provide higher current density at the outlet sections,where there is a greater percentage of fine particles.

Summary

ESPs can be designed using a number of techniques including mathematical equations, pilotplant studies, and computer modeling programs. The use of pilot plant studies is very effectivebut is not often used because of time limitations and the expense of construction. Use of com-puter models is therefore becoming more common for both the initial design and for trouble-shooting of existing ESPs.

During this lesson we covered a number of equations. The equation for particle migrationvelocity depends on the voltage strength of both the charging and collection fields and on theparticle size. The Deutsch-Anderson and Matts-Ohnfeldt equations can be used to estimatecollection efficiency in an ESP. The Deutsch-Anderson equation assumes that all particles inthe flue gas have the same migration velocity, and that particles do not become reentrained ordo not sneak through the hopper sections. The Deutsch-Anderson equation can be modified byusing field data to determine the effective migration velocity.

The Matts-Ohnfeldt equation also uses information obtained from similar ESP field installa-tions. Use of both the modified Deutsch-Anderson and the Matts-Ohnfeldt equations will typi-cally yield more accurate estimates for collection efficiency.

We also covered operating parameters that affect the collection efficiency of the ESP includingthe following:

• Resistivity

• Sectionalization

• Corona power

• Aspect ratio

• Specific collection area (SCA)

These parameters will be discussed in more detail in Lessons 4 and 6.

Careful design of the ESP involves consideration of the important operating parameters tokeep the unit operating efficiently and effectively. Not only will this help an industry complywith air pollution regulations, but a good design up-front will also reduce plant downtime andkeep maintenance problems to a minimum.

Lesson 3

3-24 2.0-2/98

Review Exercise

1. A charged particle will migrate toward an oppositely charged collection electrode. The velocity atwhich the charged particle moves toward the collection electrode is called the____________________ ____________________ and is denoted by the symbol w.

2. What is the name of the equation given below?

a. Johnstone equationb. Matts-Ohnfeldt equationc. Deutsch-Anderson equationd. Beachler-Joseph equation

3. The symbol η in the Deutsch-Anderson equation is the:

a. Collection areab. Migration velocityc. Gas flow rated. Collection efficiency

4. The Deutsch-Anderson equation does not account for:

a. Dust reentrainment that may occur as a result of rappingb. Varying migration velocities due to various-sized particles in the flue gasc. Uneven gas flow through the precipitatord. All of the above

5. True or False? Using the Matts-Ohnfeldt equation to estimate the collection efficiency of an ESPwill give less conservative results than using the Deutsch-Anderson equation.

6. Resistivity is a measure of a particle’s resistance to ____________________ and____________________ charge.

7. Dust resistivity is a characteristic of the particle in the flue gas that can alter the____________________ of an ESP.

a. Gas flow rateb. Collection efficiencyc. Gas velocity

8. Dust particles with ____________________ resistivity are difficult to remove from collectionplates, causing rapping problems.

a. Lowb. Normalc. High

( )η = −1 e-w A/Q


2.0-2/98 3-25

9. High dust resistivity can be reduced by:

a. Adjusting the flue gas temperatureb. Increasing the moisture content of the flue gasc. Injecting SO3 into the flue gasd. All of the above

10. True or False? Fly ash that results from burning high-sulfur coal generally has high resistivity.

11. A precipitator is divided into a series of independently energized bus sections called:

a. Hoppersb. Fieldsc. Stagesd. b and c, above

12. In the following figure there are ____________________ fields and ____________________cells.

a. Two, fourb. Four, eightc. Eight, twod. Eight, four

13. A precipitator should be designed with at least ____________________ field(s) to attain a highcollection efficiency.

a. Oneb. Twoc. Three or fourd. Ten

14. Electrical sectionalization improves collection efficiency by:

a. Improving resistivity conditionsb. Allowing for independent voltage control of different fieldsc. Allowing continued ESP operation in the event of electrical failure in one of the fieldsd. b and c, above

Lesson 3

3-26 2.0-2/98

15. If the design of the precipitator states that 500,000 ft2 of plate area is used to remove particles fromflue gas flowing at 750,000 ft3/min, what is the SCA of the unit?

a. 0.667 ft2/1000 acfmb. 667 ft2/1000 acfmc. 667 acfm/1000 ft2

d. 1.5 acfm/ft2

16. To achieve a collection efficiency greater than 99.5%, most ESPs have a SCA:

a. Less than 250 ft2/1000 acfmb. Between 350 and 400 ft2/1000 acfmc. Always greater than 800 ft2/1000 acfm

17. To improve the aspect ratio of an ESP design, the ____________________ of the collection sur-face should be increased relative to the ____________________ of the plate.

a. Height; lengthb. Length; height

18. Given an ESP having a configuration as shown below, what is the aspect ratio of this unit?

a. 0.33b. 1.5c. 0.75d. 1.33

19. What should the aspect ratio be for high-efficiency ESPs?

a. Less than 0.8b. Greater than 1.0c. Always greater than 1.5

20. In a properly designed ESP, the gas velocity through the ESP chamber will be:

a. Between 2 and 8 ft/secb. Greater than 20 ft/secc. Approximately between 20 and 80 ft/secd. At least 400 ft2/1000 acfm

10 ft 15 ft 15 ft

30 ft30 ft


2.0-2/98 3-27

21. In an ESP, the collection efficiency is proportional to the amount of ________________________________________ supplied to the unit.

Lesson 3

3-28 2.0-2/98


2.0-2/98 3-29

Review Exercise Answers1. Migration velocity (or drift velocity)

The velocity at which the charged particle moves toward the collection electrode is called themigration velocity (or drift velocity) and is denoted by the symbol w.

2. c. Deutsch-Anderson equation

The following equation, , is the Deutsch-Anderson equation.

3. d. Collection efficiencyThe symbol η in the Deutsch-Anderson equation is the collection efficiency.

4. d. All of the aboveThe Deutsch-Anderson equation does not account for the following:

• Dust reentrainment that may occur as a result of rapping

• Varying migration velocities due to various-sized particles in the flue gas

• Uneven gas flow through the precipitator

5. FalseUsing the Matts-Ohnfeldt equation to estimate the collection efficiency of an ESP will give moreconservative results than using the Deutsch-Anderson equation because the Matts-Ohnfeldt equa-tion accounts for non-ideal effects.

6. AcceptingReleasingResistivity is a measure of a particle’s resistance to accepting and releasing charge.

7. b. Collection efficiencyDust resistivity is a characteristic of the particle in the flue gas that can alter the collection effi-ciency of an ESP.

8. c. HighDust particles with high resistivity are difficult to remove from collection plates, causing rappingproblems.

9. d. All of the aboveHigh dust resistivity can be reduced by the following:

• Adjusting the flue gas temperature

• Increasing the moisture content of the flue gas

• Injecting SO3 into the flue gas

10. FalseFly ash that results from burning high-sulfur coal generally has low resistivity. SO3, which lowersthe resistivity of fly-ash, normally increases as the sulfur content of the coal increases.

( )η = −1 e-w A/Q

2.0-2/98 3-30

11. d. b and c, aboveA precipitator is divided into a series of independently energized bus sections called fields orstages.

12. b. Four, eight

In the above figure there are four fields and eight cells.

13. c. Three or fourA precipitator should be designed with at least three or four fields to attain a high collection effi-ciency.

14. d. b and c, aboveElectrical sectionalization improves collection efficiency by allowing the following:

• Independent voltage control of different fields• Continued ESP operation in the event of electrical failure in one of the fields

15. b. 667 ft2/1000 acfmIf the design of the precipitator states that 500,000 ft2 of plate area is used to remove particles fromflue gas flowing at 750,000 ft3/min, the SCA of the unit is as follows:

16. b. Between 350 and 400 ft2/1000 acfmTo achieve a collection efficiency greater than 99.5%, most ESPs have a SCA between 350 and400 ft2/1000 acfm.

17. b. Length; heightTo improve the aspect ratio of an ESP design, the length of the collection surface should beincreased relative to the height of the plate.

SCA500,000 ft2( )

750 1000 acfm( )-----------------------------------------=

667 ft2 1000 acfm⁄=

3-31 2.0-2/98

Bibliography

18. d. 1.33

An ESP with the above configuration has the following aspect ratio:

19. b. Greater than 1.0The aspect ratio for high-efficiency ESPs should be greater than 1.0.

20. a. Between 2 and 8 ft/secIn a properly designed ESP, the gas velocity through the ESP chamber will be between2 and 8 ft/sec, and most often between 4 and 6 ft/sec.

21. Corona powerIn an ESP, the collection efficiency is proportional to the amount of corona power supplied to theunit.

10 ft 15 ft 15 ft

30 ft30 ft

AR10 15 15+ +

30------------------------------=

4030------=

1.33=

Lesson 3

3-32 2.0-2/98


2.0-2/98 3-33

Bibliography

Anderson, E. 1924. Report, Western Precipitator Co., Los Angeles, CA. 1919. Transactions of theAmerican Institute of Chemical Engineers. 16:69.

Beachler, D. S., J. A. Jahnke, G. T. Joseph and M. M. Peterson. 1983. Air Pollution Control Systems forSelected Industries, Self-Instructional Guidebook. (APTI Course SI:431). EPA 450/2-82-006. U.S.Environmental Protection Agency.

Deutsch, W. 1922. Annals of Physics. (Leipzig) 68:335.


Hall, H. J. 1975. Design and application of high voltage power supplies in electrostatic precipitation.Journal of Air Pollution Control Association. 25:132.

Hesketh, H. E. 1979. Air Pollution Control. Ann Arbor: Ann Arbor Science Publishers.


Lawless, P. 1992. ESPVI 4.0, Electrostatic Precipitator V-I and Performance Model: Users’ Manual.EPA 600/R-29-104a.

Matts, S., and P. O. Ohnfeldt. 1964. Efficient Gas Cleaning with SF Electrostatic Precipitators. Flak-ten.

Richards, J. R. 1995. Control of Particulate Emissions, Student Manual. (APTI Course 413). U.S.Environmental Protection Agency.

Rose, H. E., and A. J. Wood. An Introduction to Electrostatic Precipitation in Theory and Practice.London: Constable and Company.

Schmidt, W. A. 1949. Industrial and Engineering Chemistry. 41:2428.

Theodore, L., and A. J. Buonicore. 1976. Industrial Air Pollution Control Equipment forParticulates. Cleveland: CRC Press.

U.S. Environmental Protection Agency. 1978, June. A Mathematical Model of Electrostatic Precipita-tion (Revision 1). Vol. 1, Modeling and Programming. EPA 600/7-78-llla.

U.S. Environmental Protection Agency. 1978, June. A Mathematical Model of Electrostatic Precipita-tion (Revision 1). Vol. II, User Manual. EPA 600/7-78-lllb.

Lesson 3

3-34 2.0-2/98

U.S. Environmental Protection Agency. 1979. Particulate Control by Fabric Filtration on Coal-FiredIndustrial Boilers. EPA 625/2-79-021.

U.S. Environmental Protection Agency. 1980, May. TI-59 Programmable Calculator Programs for In-stack Opacity, Venturi Scrubbers, and Electrostatic Precipitators. EPA 600/8-80-024.


White, H. J. 1963. Industrial Electrostatic Precipitation. Reading, MA: Addison-Wesley.

White, H. J. 1974. Resistivity problems in electrostatic precipitation. Journal of Air Pollution ControlAssociation 24:315-338.


White, H. J. 1982. Review of the state of the technology. Proceedings of the International Conferenceon Electrostatic Precipitation. Monterey, CA, October 1981. Air Pollution Control Corporation,Pittsburgh, PA.

2.0-2/98 4-1

Lesson 4ESP Design Review

Goal

To familiarize you with the factors to be considered when reviewing ESP design plans for the per-mit process.

Objectives


1. Explain how each of the following dust properties affects ESP performance:

• Dust type (chemical composition)• Size• Concentration in gas stream• Resistivity

2. Explain how each of the following flue gas properties affects ESP performance:

• Gas flow rate• Temperature• Moisture content• Chemical properties (dew point, corrosiveness, and combustibility)

3. Identify important design considerations for discharge electrodes, collection electrodes, andhopper and discharge devices

4. Explain how each of the following factors contributes to good ESP design:

• Electrical sectionalization• Specific collection area• Aspect ratio• Distribution of gas flow

5. Estimate the collection area and the collection efficiency for a given process flow rate andmigration velocity

6. Estimate the capital and operating cost of an ESP using tables and figures

Lesson 4

4-2 2.0-2/98

Introduction

As discussed in Lessons 2 and 3, finalizing the design of an electrostatic precipitator and itscomponents involves consideration of many factors. Air pollution control agency officers whoreview ESP design plans should consider these factors during the review process. Some ofthese factors relate to the properties of the dust and flue gas being filtered, while others applyto the specific ESP design:

• Type of discharge electrode

• Type of collection electrode

• Electrical sectionalization (number of fields and individual power supplied used


• Aspect ratio

Construction details, such as shell insulation, inlet location, hopper design, and dust dischargedevices are also important.

This lesson reviews the ESP design parameters, along with typical ranges for these variables.It also familiarizes you with cost information for various ESP designs so that you can be awareof cost when reviewing design plans and making recommendations.

Review of Design Variables

The principal design variables are the dust concentration, measured in g/m3 (lb/ft3 or gr/ft3)and the gas flow rate to the ESP, measured in m3/min (ft3/min or acfm). The gas volume anddust concentration (loading) are set by the process exhaust gas flow rate. Once these variablesare known, the vendor can begin to design the precipitator for the specific application. A thor-ough review of ESP design plans should consider the factors presented below.

Physical and chemical properties of the dust such as dust type, size of the dust particles, andaverage and maximum concentrations in the gas stream are important ESP design consider-ations. The type of dust to be collected in the ESP refers to the chemical characteristics of thedust such as explosiveness. For example, a dry ESP should not be used to collect explosivedust. In this case, it might be a better idea to use a baghouse or scrubber. Particle size is impor-tant; small particles are more difficult to collect and become reentrained more easily thanlarger particles. Additional fields may be required to meet regulatory limits. The dust loadingcan affect the operating performance. If the dust concentration is too high, the automatic volt-age controller may respond by totally suppressing the current in the inlet fields. Suppressedcurrent flow drives the voltage up, which can cause sparking. For this reason, it might be agood idea to install a cyclone or multicyclone to remove larger particles and reduce the dustconcentration from the flue gas before it enters the ESP. The facility could install a larger ESP(with more plate area), however, this technique would be more costly.

Resistivity is a function of the chemical composition of the dust, the flue gas temperature andmoisture concentration. For fly ash generated from coal-fired boilers, the resistivity dependson the temperature and moisture content of the flue gas and on the sulfur content of the coalburned; the lower the sulfur content, the higher the resistivity, and vice versa. If a boiler burnslow-sulfur coal, the ESP must be designed to deal with potential resistivity problems. As pre-viously stated in Lesson 3, high resistivity can be reduced by spraying water, SO3 or someother conditioning agent into the flue gas before it enters the ESP.

ESP Design Review

2.0-2/98 4-3

Predicting the gas flow rate and gas stream properties is essential for proper ESP design.The average and maximum gas flow rates through the ESP, the temperature, moisture content,chemical properties such as dew point, corrosiveness, and combustibility of the gas should beidentified prior to final design. If the ESP is going to be installed on an existing source, a stacktest should be performed to determine the process gas stream properties. If the ESP is beinginstalled on a new source, data from a similar plant or operation may be used, but the ESPshould be designed conservatively (with a large SCA, a high aspect ratio, and high coronapower). Once the actual gas stream properties are known, the designers can determine if theprecipitator will require extras such as shell insulation for hot-side ESPs, corrosion-proof coat-ings, and installation of heaters in hoppers or ductwork leading into and out of the unit.

The type of discharge electrodes and electrode support are important. Small-diameter wiresshould be firmly supported at the top and connected to a weight heavy enough (11.4-kgweights for 9.1-m wires) to keep the wires from swaying. The bottom and top of each wireshould be covered with shrouds to help minimize sparking and metal erosion at these points.Newer ESPs are generally using rigid-frame or rigid-electrode discharge electrodes.

Collection electrodes—type (either tube or plate), shape of plates, size, and mechanicalstrength—are then chosen. Plates are usually less than 9 m (30 ft) high for high-efficiencyESPs. For ESPs using wires, the spacing between collection plate electrodes usually rangesfrom 15 to 30 cm (6 to 12 in.). For ESPs using rigid-frame or rigid electrodes, the spacing istypically 30 to 38 cm (12 to 15 inches). Equal spacing must be maintained between platesthroughout the entire precipitator. Stiffeners may be used to help prevent the plates from warp-ing, particularly when hot-side precipitators are used.

Proper electrical sectionalization is important to achieve high collection efficiency in theESP. Electrical sectionalization refers to the division of a precipitator into a number of differ-ent fields and cells, each powered by its own T-R set. ESPs should have at least three to fourfields to attain a high collection efficiency. In addition, the greater the number of fields the bet-ter the chance that the ESP will achieve the designed collection efficiency. There should beapproximately one T-R set for every 930 to 2970 m2 (10,000 to 30,000 ft2) of collection-platearea.

The specific collection area (SCA) is the collection area, in m2 per 1000 m3/h (ft2 per 1000ft3/min), of flue gas through the precipitator. The typical range for SCA is between 11 and 45m2 per 1000 m3/h (200 and 800 ft2 per 1000 acfm). The SCA must be large enough to effi-ciently collect particles (99.5% collection efficiency), but not so large that the cost of the ESPis too high. If the dust has a high resistivity, vendors will generally design the ESP with ahigher SCA [usually greater than 22 m2 per 1000 m3/h (400 ft2 per 1000 acfm)] to help reduceresistivity problems.

Aspect ratio is the ratio of effective length to height of the collector surface. The aspect ratioshould be high enough to allow the rapped particles to settle in the hopper before they are car-ried out of the ESP by the gas flow. The aspect ratio is usually greater than 1.0 for high-effi-ciency ESPs. Aspect ratios of 1.3 to 1.5 are common, and they are occasionally as high 2.0.

Even distribution of gas flow across the entire precipitator unit is critical to ensure collectionof the particles. To assure even distribution, gas should enter the ESP through an expansioninlet plenum containing perforated diffuser plates (see Figure 3-7). In addition, the ducts lead-ing into the ESP unit should be straight as shown in Figure 4-1. For ESPs with straight-line

Lesson 4

4-4 2.0-2/98

inlets, the distance of A should be at least as long as the distance of B in the inlet (Katz 1979).In situations where a straight-line inlet is not possible and a curved inlet must be used (see Fig-ure 4-2), straightening vanes should be installed to keep the flue gas from becoming stratified.The gas velocity through the body of the ESP should be approximately 0.6 to 2.4 m/s (2 to 8 ft/sec). For ESPs having aspect ratios of 1.5, the optimum gas velocity is usually between 1.5and 1.8 m/s (5 and 6 ft/sec). The outlet of the ESP should also be carefully designed to provideeven flow of the gas from the ESP to the stack without excessive pressure buildup. This can bedone by using an expansion outlet, as shown in Figure 4-3. Figures 4-1 and 4-2 also haveexpansion outlets.

Figure 4-1. Straight-line inlet

Figure 4-2. Straightening vanes in a curved inlet

Figure 4-3. ESP with expansion outlet

B

A

Expansioninlet

Straighteningvanes

Expansionoutlet

ESP Design Review

2.0-2/98 4-5

The hopper and discharge device design including geometry, size, dust storage capacity,number, and location are important so that dust is removed on a routine basis. A well-designeddust hopper is sloped (usually 60°) to allow dust to flow freely to discharge devices. Itincludes access ports and strike plates to help move dust that becomes stuck. Dust should beonly temporarily stored in the hopper and removed periodically by the discharge devices toprevent it from backing up into the ESP where it can touch the plates, possibly causing a cellto short out. In addition to the amount of fly ash present, there are a couple of special consider-ations to keep in mind when ESPs are used on coal-fired boilers. First, the amount of fly ash inthe flue gas can vary depending on what type of coal is burned and the ash content of the coal.Coal having a higher ash content will produce more fly ash than coal having lower ash values.Consequently, the discharge device must be designed so that the operator can adjust the fre-quency of fly ash removal. Second, hoppers need to be insulated to prevent ash from "freez-ing," or sticking, in the hopper.

Finally, emission regulations in terms of opacity and dust concentration (grain-loading)requirements will ultimately play an important role in the final design decisions. Electrostaticprecipitators are very efficient; collection efficiency can usually be greater than 99% if theESP is properly designed and operated.

Typical Ranges of Design Parameters

While reviewing a permit for ESP installation, check whether the design specifications arewithin the range that is typically used by that industry. The ranges of basic design parametersfor fly ash precipitators are given in Table 4-1.

Table 4-1. Typical ranges of design parameters for fly ashprecipitators

Parameter Range (metric units) Range (English units)

Distance between plates(duct width)

Gas velocity in ESP

SCA

Aspect ratio (L/H)

Particle migration velocity

Number of fields

Corona power/flue gasvolume

Corona current/ft2 platearea

Plate area per electrical (T-R) set

20-30 cm (20-23 cm optimum)

1.2-2.4 m/s (1.5-1.8 m/s optimum)

11-45 m2/1000 m3/h(16.5-22.0 m2/1000 m3/h optimum)

1-1.5 (keep plate height less than9 m for high efficiency)

3.05-15.2 cm/s

4-8

59-295 watts/1000 m3/h

107-860 microamps/m2

465-7430 m2/T-R set(930-2790 m2/T-R set optimum)

8-12 in. (8-9 in. optimum)

4-8 ft/sec (5-6 ft/sec optimum)

200-800 ft2/1000 cfm(300-400 ft2/1000 cfm optimum)

1-1.5 (keep plate height less than30 ft for high efficiency)

0.1-0.5 ft/sec

4-8

100-500 watts/1000 cfm

10-80 microamps/ft2

5000-80,000 ft2/T-R set (10,000-30,000 ft2/T-R set optimum)

Source: White 1977.

Lesson 4

4-6 2.0-2/98

Estimating Collection Efficiency and Collection Area

The manufacturer designs and sizes the electrostatic precipitator. However, the operator (orreviewer) needs to check or estimate the collection efficiency and the amount of collectionarea required for a given process flow rate. You can compute these estimates by using theDeutsch-Anderson or Matts-Ohnfeldt equations (see Lesson 3). These equations are repeatedin Table 4-2.

Table 4-2. Equations used to estimate collection efficiencyand collection area

Calculation Deutsch-Anderson Matts-Ohnfeldt

Collection efficiency

Collection area (to meet arequired efficiency)

Where: η = collection efficiencyA = collection areaw = migration velocityQ = gas flow rateln = natural logarithm

η = collection efficiencyA = collection areawk = average migration

velocityk = constant (usually 0.5)ln = natural logarithm

η 1 e w A Q⁄( )––=

AQ–

w------- ln 1 η–( )[ ]=

η 1 ewk A Q⁄( )k

––=

AQwk

------ k

– ln 1 η–( )[ ]1 k⁄

=

ESP Design Review

2.0-2/98 4-7

Example Estimation

The exhaust rate of the gas being processed is given as 1,000,000 ft3/min. The inlet dustconcentration in the gas as it enters the ESP is 8 gr/ft3. If the emission regulations statethat the outlet dust concentration must be less than 0.04 gr/ft3, how much collection area isrequired to meet the regulations? Use the Deutsch-Anderson equation for this calculationand assume the migration velocity is 0.3 ft/sec.

1. From Table 4-2, use this version of the Deutsch-Anderson equation to solve theproblem:

Where: A = collection area, ft2

Q = gas flow rate, ft3/secw = migration velocity, ft/secη = collection efficiencyln = natural logarithm

In this example,

Q = 1,000,000 ft3/min × 1 min/60 sec= 16,667 ft3/sec

w = 0.3 ft/sec

2. Calculate the collection efficiency, η.

3. Calculate the collection area, A, in ft2.

AQ–

w------- ln 1 η–( )[ ]=

ηdustin dustout–

dustin

-----------------------------------=

8 gr ft3⁄ 0.04 gr ft3⁄–

8 gr ft3⁄------------------------------------------------------=

0.995 or 99.5%=

A16,667– ft3 sec⁄

0.3 ft sec⁄--------------------------------------- 1 0.995–( )ln[ ]=

55,557 ft– 2( ) 5.2983–[ ]×=

294,358 ft2=

Lesson 4

4-8 2.0-2/98

Estimating Capital and Operating Costs

This section contains generalized cost data for ESP systems described throughout this guide-book. These data should be used only as an estimate to determine system cost. The total capitalinvestment (TCI) includes costs for the ESP structure, the internals, rappers, power supply,and auxiliary equipment, and the usual direct and indirect costs associated with installing orerecting new structures. These costs, given in second-quarter 1987 dollars, are described in thefollowing subsections.

ESP Equipment Cost

Most of the following cost discussion is taken from the EPA OAQPS Cost Control Manual(1990). Costs for rigid-electrode, wire and plate, and flat-plate ESPs can be estimatedusing Figure 4-4. Costs for two-stage precipitators are given later.

Figure 4-4 represents two cost curves (the two in the middle) along with their respectiveequations (outer lines with arrows). Each curve requires two equations for calculatingcost: one for total plate areas between 10,000 and 50,000 ft2 and another for total plateareas between 50,000 and 1,000,000 ft2. The lower curve shows the cost for the basic unitwithout the standard options. It represents the flange-to-flange, field-erected price for arigid-electrode design. The upper curve includes all of the standard options (listed in Table4-3) that are normally used in a modern system. All units (both curves) include the ESPcasing, pyramidal hoppers, rigid electrodes and internal collection plates, transformer-rec-tifier (T-R) sets and microprocessor controls, rappers, and stub supports (legs) for 4-footclearance below the hopper discharges. The costs are based on a number of actual quotesthat have been fitted to lines using the “least squares” method. Don’t be surprised if youobtain quotes that differ from these curves by as much as ±25%. (Significant savings canbe obtained by solicitating multiple quotes.) The equations should not be used to extrapo-late costs for total plates areas below 10,000 or above 1,000,000 ft2. The standard optionsincluded in the upper curve add approximately 45% to the basic cost of the flange-to-flange hardware. Insulation costs are for 3 inches of field-installed glass fiber encased in ametal skin and applied on the outside of all areas in contact with the exhaust gas stream.Calculate insulation for ductwork, fan casings, and stacks separately. To obtain more accu-rate results, solve the equations for the lines instead of reading the values from the graph.

ESP Design Review

2.0-2/98 4-9

Figure 4-4. Dry-type rigid electrode ESP flange-to-flange purchase priceversus plate area

Impact of Alternative Electrode Designs

All three designs—rigid electrode, weighted wire, and rigid frame—can be employed inmost applications. Any cost differential between designs will depend on the combinationof vendor experience and site-specific factors that dictate equipment size factors. Therigid-frame design will cost up to 25% more than the wire and plate design if the plateheight is restricted to that used in wire/plate designs. Several vendors can now providerigid-frame ESPs with taller plates, and thus the cost differential can approach zero.

The weighted wire design uses narrower plate spacings and more internal discharge elec-trodes. This design is being used less; therefore, its cost is increasing and currently is

Table 4-3. Standard options for basic equipment

Item Cost Adder, %

1. Inlet and outlet nozzles and diffuser plates2. Hopper auxiliaries/heaters, level detectors3. Weather enclosure and stair access4. Structural supports5. InsulationTotal options 1 to 5

8 to 108 to 108 to 1058 to 101.37 to 1.45 × base

Lesson 4

4-10 2.0-2/98

approximately the same as that for the rigid electrode ESP. Below about 15,000 ft2 of platearea, ESPs are not normally field-erected (erected at the installation site), and the costswill probably be higher than values extrapolated from Figure 4-4.

Impact of Materials of Construction:Metal Thickness and Stainless Steel

Corrosive or other adverse operating conditions may require specifications of thickermetal sections in the precipitator. Metal thickness can be moderately increased with mini-mal cost increases. For example, collection plates are typically constructed of 18-gaugemild steel. Most ESP manufacturers can increase the section thickness by 25% withoutsignificant design changes or increases in manufacturing costs of more than a few percent.

Changes in the type of material can increase the purchase cost of the ESP significantly.Using type 304 stainless steel instead of 18-gauge mild steel for collection plates and pre-cipitator walls can increase costs 30-50%. Using even more expensive materials for allelements of the ESP can increase costs up to several hundred percent. Based on the carbonsteel 18-gauge cost, the approximate factors given below can be used for other materials.

Recent Trends

Most of today's market (1987) is in the 50,000 to 200,000 ft2 plate area size range. ESPselling prices have increased very little over the past 10 years because of more effectivedesigns, increased competition from European suppliers, and a shrinking utility market.

Design improvements have allowed wider plate spacings that reduce the number of inter-nal components and higher plates and masts that provide additional plate area at a lowcost. Microprocessor controls and energy management systems have lowered operatingcosts.

Few, if any, hot-side ESPs (those used upstream from an air preheater on a combustionsource) are being specified for purchase. Recognition that low-sodium coals tend to buildresistive ash layers on the collection plates, thus reducing ESP efficiency, has almost elim-inated sales of hot-side units. Of the 150 existing units, about 75 are candidates for con-version to cold-side units (using resistivity conditioning agents) over the next 10 years(U.S. EPA 1990).

Table 4-4. ESP costs using various materials

Factor Material

1.01.31.71.92.33.24.5

Carbon Steel, 18-gaugeStainless Steel, 304Stainless Steel, 316Carpenter 20 CB-3Monel-400Nickel-200Titanium


ESP Design Review

2.0-2/98 4-11

Specific industry application has little impact on either ESP design or cost, with the fol-lowing three exceptions: paper mills, sulfuric acid manufacturing plants, and coke by-product plants. Because paper mills have dust that can be sticky and difficult to remove,paper mill ESPs use drag conveyer hoppers. These hoppers increase the cost by approxi-mately 10 percent of the base flange-to-flange equipment cost. For emissions control insulfuric acid plants and coke by-product ovens, wet ESPs are used. In sulfuric acid manu-facture, wet ESPs are used to collect acid mist. These precipitators usually are small anduse lead for all interior surfaces; hence, they normally cost $65 to $95/ft2 of collectingarea installed (mid-1987 dollars) and up to $120/ft2 in special situations. Using Figure 4-4,the standard cost for a rigid-frame ESP ranges from $7 to $14/ft2 of collecting area. Inaddition, a wet circular ESP is typically used to control emissions from a coke oven off-gas detarring operation. These precipitators are made from high-alloy stainless steels andtypically cost $90 to $120/ft2 installed. Because of the small number of sales, small size ofunits sold, and dependency of site-specific factors, more definitive costs are not available.

Retrofit Cost Factor

Retrofit installations increase the cost of an ESP because of the frequent need to removesomething to make way for the new ESP. Also, the ducting usually is much more expen-sive as a retrofit application because the ducting path is often constrained by existingstructures, additional supports are required, and the confined areas make erection morelabor intensive and lengthy. Costs are site-specific; however, for estimating purposes, aretrofit multiplier of 1.3 to 1.5 applied to the total capital investment can be used. Themultiplier should be selected within this range based on the relative difficulty of the instal-lation.

A special case is the conversion of a hot-side to a cold-side ESP for coal-fired boiler appli-cations. The magnitude of the conversion is very site-specific, but most projects will con-tain the following elements:

• Relocating the air preheater and the ducting to it

• Resizing the ESP inlet and outlet duct to the new air volume and rerouting it

• Upgrading the ID (induced draft) fan size or motor to accommodate the higher staticpressure and horsepower requirements

• Adding or modifying foundations for fan and duct supports

• Assessing the required SCA and either increasing the collecting area or installing anSO3 gas-conditioning system

• Adding hopper heaters

• Upgrading the analog electrical controls to microprocessor-type controls

• Increasing the number of collecting plate rappers and perhaps the location of rappers

In some installations, it may be cost-effective to gut the existing collector totally, utilizeonly the existing casing and hoppers, and upgrade the ESP using modern internal compo-nents. The cost of conversion is a multimillion dollar project typically running at least 25to 35 percent of the total capital investment of a new unit.

Lesson 4

4-12 2.0-2/98

Costs for Two-Stage Precipitators

Purchase costs for modular, two-stage precipitators should be considered separately fromlarge-scale, single-stage ESPs (see Figure 4-5). To be consistent with industry practice,costs are given as a function of flow rate through the system. The lower cost curve is for atwo-cell unit without a precooler, installed cell washer, and a fan. The upper curve is foran engineered package system with the following components: inlet diffuser plenum, pre-filter, cooling coils with coating, coil plenums with access, water-flow controls, triple-passconfiguration, system exhaust fan with accessories, outlet plenum, and in-place foamcleaning system with semiautomatic control and programmable controller. All equipmentis fully assembled mechanically and electrically, and it is mounted on a steel structuralskid.

Figure 4-5. Purchase costs for two-stage, two-cell precipitators

Total Purchase Cost

The total purchase cost of an ESP system is the sum of the costs of the ESP, options, aux-iliary equipment, instruments and controls, taxes, and freight. The last three items gener-ally are taken as percentages of the estimated total cost of the first three items. Typicalvalues are 10% for instruments and controls, 3% for taxes, and 5% for freight.

Costs of standard and other options can vary from 0% to more than 150% of ESP basecost, depending on site and application requirements. Other factors that can increase ESPcosts are given in Table 4-5.

ESP Design Review

2.0-2/98 4-13

Total Capital Investment

Total capital investment (TCI) is estimated from a series of factors applied to the pur-chased equipment cost (PEC) to obtain direct and indirect costs for installation. The TCI isthe sum of the direct costs (equipment and installation) and indirect costs. The requiredfactors are given in Table 4-6. Because ESPs can vary from small units attached to exist-ing buildings to large, separate structures, specific factors for site preparation or for build-ings are not given. However, costs for buildings and materials may be obtained fromreferences such as Means Square Foot Costs 1987. Land, working capital, and off-sitefacilities are excluded from the table because they are required only for very large installa-tions. However, they can be estimated on an as-needed basis.

Note that the factors given in Table 4-6 are for average installation conditions, and forexample, include no unusual problems with site earthwork, access, shipping, or interferingstructures. Considerable variation may be seen with other-than-average installation cir-cumstances. For two-stage precipitators purchased as packaged systems, several of thecosts in Table 4-6 would be greatly reduced or eliminated. These include instruments andcontrols, foundations and supports, erection and handling, painting, and model studies. Aninstallation factor of 0.25 of the PEC (instead of 0.67 PEC) would be more nearly appro-priate for the two-stage ESPs.

Table 4-5. Items that increase ESP costs

Item Factor orTotal Cost

Applied to

Rigid-frame electrode with restricted plate height

Type 304 stainless-steel collector plates andprecipitator walls

All-stainless construction

ESP with drag conveyor hoppers (paper mill)

Retrofit installations

Wet ESPSulfuric acid mist

Sulfuric acid mist (special installation)

Coke oven off-gas

1.0-1.25

1.3-1.5

2-3

1.1

1.3-1.5

$65-$95/ft2

Up to $120/ft2

$90-$120/ft2

ESP base cost

ESP base cost

ESP base cost

ESP base cost

ESP base cost

-

-

-


Lesson 4

4-14 2.0-2/98

ExampleA basic, flat-plate, rigid-electrode ESP, requiring a plate area of 40,800 ft2, is pro-posed. The manufacturer recommends using 304 stainless steel for the dischargeelectrodes and collection plates due to the corrosive nature of the flue gas.Assume that the auxiliary equipment costs $10,000.

Using Figure 4-4 and Tables 4-4 and 4-6, estimate the following:

1. Equipment cost (EC)

2. Purchased equipment cost (PEC)

3. Total capital cost of purchasing and installing the ESP

Table 4-6. Capital cost factors for ESPs

Cost Item Factor

Direct CostsPurchased equipment costs

ESP + auxiliary equipmentInstrumentsSales taxesFreight

Purchased equipment cost, PEC

Direct installation costsFoundation and supportsHandling and erectionElectricalPipingInsulation for ductwork1

PaintingDirect installation costs

Site preparationBuildings

Total Direct Costs DC

Indirect Costs (installation)EngineeringConstruction and field expenseContractor feesStart-up feePerformance testModel studyContingencies

Total Indirect Costs IC

Total Capital Cost = DC + IC

As estimated, EC0.10 EC0.03 EC0.05 EC

PEC = 1.18 EC

0.04 PEC0.50 PEC0.08 PEC0.01 PEC0.02 PEC0.02 PEC0.67 PEC

As required, SPAs required, Bldg.

1.67 PEC + SP + Bldg.

0.20 PEC0.20 PEC0.10 PEC0.01 PEC0.01 PEC0.02 PEC0.03 PEC0.57 PEC

2.24 PEC + SP + Bldg.1If ductwork dimensions have been established, cost may be estimated based on $10 to $12/ft2 of

surface for field application.Source: U.S. EPA 1990.

ESP Design Review

2.0-2/98 4-15

1. Estimate the equipment cost. Because the ESP is a basic, rigid-frame ESPwithout the standard options, the lower line from Figure 4-4 is used to obtainthe capital cost. Using a collection area of 40,800 ft2, a cost of $470,000 canbe read from Figure 4-4. But this cost figure assumes that the ESP dischargeelectrodes and collection plates are made out of carbon steel material. Asstated in Table 4-4, the cost factor for 304 stainless steel is 1.3. The equipmentcost is:

$470,000 × 1.3 = $611,000

Auxiliary equipment cost = $10,000

Equipment cost (EC) = $621,000

2. Estimate the purchased equipment cost (PEC) using the cost factors inTable 4-6 (some calculations are rounded).

Equipment cost (EC) = $621,000

Instrumentation (0.10 × 621,000) = $62,100

Sales Tax (0.03 × 621,000) = $18,600

Freight (0.05 × 621,000) = $31,100

Purchased equipment cost (PEC) = $732,800

3. Estimate the total capital cost. Knowing the PEC and using the cost factorsin Table 4-6, you can estimate the remaining direct and indirect costs, whichmake up the total capital cost. A summary of these costs are provided in Table4-7.

Lesson 4

4-16 2.0-2/98

Summary

Some key factors that affect the design of an ESP include the following:

• Type of discharge electrode

• Type of collection electrode

• Electrical sectionalization


• Aspect ratio

We also covered how to estimate the cost of ESPs. These estimates can be used as budgetaryestimates by facilities planning to install an ESP or by agency engineers for reviewing permitapplications.

Table 4-7. Example case capital costs

Cost Item Factor Cost(s)

Direct CostsPurchased equipment costs

ESP + auxiliary equipmentInstrumentsSales taxesFreight

Purchased equipment cost, PEC

Direct installation costsFoundation and supportsHandling and erectionElectricalPipingInsulation for ductwork1

PaintingDirect installation costs


Total Direct Cost, DC

Indirect Costs (installation)EngineeringConstruction and field expenseContractor feesStart-up feePerformance testModel studyContingencies

Total Indirect Cost, IC

Total Capital Cost = DC + IC

As estimated, EC0.10 EC0.03 EC0.05 EC

PEC = 1.18 EC

0.04 PEC0.50 PEC0.08 PEC0.01 PEC0.02 PEC0.02 PEC0.67 PEC

As required, SPAs required, Bldg.


0.20 PEC0.20 PEC0.10 PEC0.01 PEC0.01 PEC0.02 PEC0.03 PEC0.57 PEC


$621,00062,10018,60031,100

$732,800

$29,300367,00058,600

7,33014,70014,700

$491,630

$1,224,430

$147,000147,00073,300

7,3307,330

14,70022,000

$418,660

$1,643,0901If ductwork dimensions have been established, cost may be estimated based on $10 to $12/ft2 of surface for

field application.

ESP Design Review

2.0-2/98 4-17

Review Exercise

1. Two important process variables to consider when designing an ESP are the gas____________________ ____________________ and the dust ____________________.

2. In an ESP, the amount of dust coming into the ESP is important. If the dust loading is very high itwill:

a. Suppress the current in the inlet field and cause the controller to drive up the voltageb. Increase the current in the inlet field and cause the controller to decrease the voltagec. Cause an increase in the dust resistivityd. Have no effect on the ESP performance

3. If coal burned in a boiler has a low sulfur content, the resulting dust will usually have____________________ resistivity.

a. Highb. Low

4. Which of the drawings below shows a good design of an inlet into the ESP?

a.

b.

c.

Lesson 4

4-18 2.0-2/98

5. True or False? Dust can be stored in hoppers for any length of time without causing problems.

6. An ESP has a collection area of 750,000 ft2 and filters fly ash from flue gas flowing at 1,500,000ft3/min. The migration velocity of the dust is 0.25 ft/sec. Estimate the collection efficiency of theESP using the Deutsch-Anderson equation.

7. The design plan states that an ESP will filter fly ash from flue gas that has a dust loading of 2 gr/ft3

and a flow rate of 2,000,000 acfm (ft3/min). The dust migration velocity is 0.3 ft/sec. If the regula-tions state that the emissions must be less than 0.02 gr/ft3, what is the total collection area neededfor the ESP design? Use the Deutsch-Anderson equation.

η 1 e w A Q⁄( )––=

ESP Design Review

2.0-2/98 4-19


1. Flow rateConcentrationTwo important process variables to consider when designing an ESP are gas flow rate and dustconcentration.

2. a. Suppress the current in the field and cause the controller to drive up the voltageIn an ESP, the amount of dust coming into the ESP is important. If the dust loading is very high itwill suppress the current in the inlet field and cause the controller to drive up the voltage.

3. a. HighIf coal burned in a boiler has a low sulfur content, the resulting dust will usually have high resistiv-ity.

4. c.

The figure in option “c” shows the best inlet design because it has a straight-on inlet and an inletplenum with a distance of A as long as (or longer than) B. Option "b" is fine if there are straighten-ing vanes in the duct.

5. FalseDust can NOT be stored in hoppers for any length of time without causing problems. Dust shouldbe stored temporarily in the hopper and removed periodically by the discharge device to preventthe dust from backing up into the ESP.

6. 99.94%Solution:Calculate the collection efficiency using the Deutsch-Anderson equation:

Where: w = 0.25 ft/sec × 60 sec/min = 15 ft/minA = 750,000 ft2

Q = 1,500,000 ft3/min

A

B

η 1 e w A Q⁄( )––=

η 1 e 15ft min⁄ 750,000 ft2 /1,500,000 ft3 /min( )––=

1 0.00055–=

0.9994 or 99.94%=

Lesson 4

4-20 2.0-2/98

7. 512,000 ft2

Solution:1. Using equation 4-1, calculate the collection efficiency required to meet emissions regulations.

2. Calculate the total collection area needed, using the following form of the Deutsch-Andersonequation:

Where: w = 0.3 ft/sec × 60 sec/min = 18 ft/minQ = 2,000,000 ft3/minη = 0.99

A =

= 512,000 ft2

η 2gr ft3 0.02gr ft3⁄–⁄2gr ft3⁄

---------------------------------------------------=

0.99 or 99%=

AQ–

w------- ln 1 η–( )[ ]=

2,000,000 ft– 3 min⁄18 ft min⁄

------------------------------------------------ ln 1 0.99–( )[ ]

2.0-2/98 4-21

Bibliography

Beachler, D. S., J. A. Jahnke, G. T. Joseph and M. M. Peterson. 1983. Air Pollution Control Systems forSelected Industries, Self-instructional Guidebook. (APTI Course SI:431). EPA 450/2-82-006. U.S.Environmental Protection Agency.



Neveril, R. B., J. U. Price, and K. L. Engdahl. 1978. Capital and operating costs of selected air pollu-tion control systems - I. Journal of Air Pollution Control Association. 28:829-836.


U.S. Environmental Protection Agency. 1990, January. OAQPS Cost Control Manual. 4th ed. EPA450/3-90-006.

U.S. Environmental Protection Agency. 1991. Control Technology for Hazardous Air PollutantsHandbook. EPA 625/6-91/014.


Lesson 4

4-22 2.0-2/98

2.0-2/98 5-1

Lesson 5Industrial Applications of ESPs

Goal

To familiarize you with the many ways ESPs are used by various industries to reduce emissions.

Objectives


1. List five major industries that use ESPs to reduce particulate emissions2. Describe how ESPs are used with dry flue gas desulfurization systems to reduce SO2 emis-

sions from boilers3. Identify two operating problems that can occur when using ESPs on cement kilns4. List two operating problems associated with ESPs in the steel industry5. Briefly describe how ESPs are used along with acid gas control systems to control particulate

and acid gas emissions from municipal solid waste and hazardous waste incinerators6. Identify two processes in the lead, zinc and copper smelting industries that use ESPs to control

particulate emissions

Introduction

Because ESPs can collect dry particles, sticky or tarry particles, and wet mists, they are usedby many different industries, as diverse as chemical production and food processing. This les-son reviews the following industries that use ESPs to reduce air pollutant emissions: fossil-fuel-fired boilers, cement plants, steel mills, petroleum refineries, municipal waste incinera-tors, hazardous waste incinerators, kraft pulp and paper mills, and lead, zinc, and coppersmelters.

Boilers

Particulate Matter Control System

ESPs are most widely used for the control of fly ash from industrial and utility boilers andhave been used on coal-fired boilers for over 50 years. Particulate matter is generated fromboilers when fossil fuels (coal and oil) are burned to generate steam for industrial pro-cesses or to produce electric power. Both hot-side and cold-side precipitators are used tocontrol particulate emissions. Other than some construction modifications to account forthe temperature difference of the flue gas handled, hot-side and cold-side ESPs are essen-tially the same. Cold-side ESPs are used most often for collecting fly ash from coal-fired

Lesson 5

5-2 2.0-2/98

boilers. If the dust has high resistivity, cold-side units are used along with a conditioningagent such as sulfur trioxide (see Lesson 3).

Dry Sulfur Dioxide (SO2) Control System

One technology for reducing sulfur dioxide (SO2) emissions from boilers is dry flue gasdesulfurization (FGD). In dry FGD, the flue gas containing SO2 is contacted with analkaline material to produce a dry waste product for disposal. This technology consists ofthree different FGD methods:

• Injection of wet alkaline material (slurry) into a spray dryer with collection of dry par-ticles in an electrostatic precipitator or baghouse,

• Injection of dry alkaline material into the flue gas stream with collection of dry parti-cles in an ESP or baghouse, or

• Addition of alkaline material to the fuel prior to combustion

Spray dryers used in dry FGD are similar to those that have been used for over 40 yearsin the chemical, food-processing, and mineral preparation industries. Spray dryers are ves-sels where hot flue gas is contacted with a finely atomized, wet alkaline spray (see Figure5-1). Flue gas enters the top of the spray dryer and is swirled by a fixed vane ring to causeintimate contact with the slurry spray. Sodium carbonate solutions and lime slurries are themost common alkaline material used. The slurry is atomized into extremely fine dropletsby rotary atomizers or two-fluid nozzles. In a rotary atomizer, slurry is broken into drop-lets by centrifugal force as the atomizer wheel spins at a very high speed. In two-fluid noz-zles, slurry is mixed with compressed air, which forms the very small droplets. The hightemperature of the flue gas, 120 to 204°C (250 to 400°F), evaporates the moisture from thewet alkaline sprays, leaving a dry, powdered product. The dry product is then collected inan ESP or baghouse (Joseph and Beachler 1981).

Figure 5-1. Spray dryer with ESP

ESPSpray dryer

absorber

Gas inlet

Fly ashhandlingsystem

Industrial Applications of ESPs

2.0-2/98 5-3

A number of spray dryer FGD systems have been installed on industrial and utility boilers.They are particularly useful in meeting New Source Performance Standards (NSPS) thatrequire only 70% SO2 removal efficiency for utility boilers burning low-sulfur coal and asretrofit applications for units having to meet the standards required by the 1990 Clean AirAct Amendments (see Table 5-1).

Spray dryer absorbers systems can reduce SO2 emissions by 60 to 90%. They have beenused on boilers burning low-sulfur coal (usually less than 2% sulfur content) and areattractive alternatives to wet scrubbing technology, particularly in the arid western U.S.

In dry injection systems, a dry alkaline material (sorbent) is injected pneumatically intothe gas stream by nozzles located in the ductwork prior to the flue gas entering the ESP.Sodium-based sorbents are used more frequently than lime for industrial coal-fired boilersbut hydrated lime is prevalent for waste burning incinerators. Sodium bicarbonate is fre-quently used because it is highly reactive with SO2. Sodium carbonate (soda ash),although not as reactive as sodium bicarbonate, is also used (U.S. EPA 1980). SO2

removal efficiency for these systems is typically between 70 and 80%.

A third way to apply dry FGD is by adding alkaline material to the fuel (coal) prior tocombustion. In fluidized bed boilers, limestone or sometimes lime is added to the coal inthe fluidized burning bed. These systems are capable of removing more than 90% of SO2

from the boiler flue gas. Alkaline material can also be injected into the furnace throughports or directly into the fuel burners. The SO2 removal is typically greater than 70% inthese systems.

Table 5-1. Commercial spray dryer FGD systems using an ESP or a baghouse

Station or plant Size(MW)

Installationdate

System description Sorbent

Coalsulfur

content(%)

SO2

emissionremoval

efficiency(%)

Otter Tail PowerCompany: CoyoteStation No. 1,Beulah, ND

410 6/81 Rockwell/Wheelabrator-Frye system: four spraytowers in parallel with 3atomizers in each:reverse air-shakerbaghouse with Dacronbags

Soda ash(sodiumcarbonate)

0.78 70

Basin Electric:Laramie River StationNo. 3, Wheatland,WY

500 Spring1982

Babcock and Wilcox:four spray reactors with12 "Y-jet" nozzles ineach: electrostaticprecipitator

Lime 0.54-0.81

85-90

Strathmore Paper Co.:Woronco, MA

14 12/79 Mikropul: spray dryerand pulse jet baghouse

Lime 2-2.5 75

Celanese Corp.:Cumberland, MD

31 2/80 Rockwell/Wheelabrator-Frye system: one spraytower followed by abaghouse

Lime 1-2 85

Source: U.S. EPA February 1980.

Lesson 5

5-4 2.0-2/98

Cement Plants

ESPs are used in cement plants to control particulate emissions from cement kilns and clinkercoolers. In a cement plant, raw materials are crushed, ground, blended, and fed into a kiln,where they are heated. The kiln is fired with coal, oil, or gas. The material is heated to a tem-perature above 1595°C (2900°F), which causes it to fuse. The fused material is called cementclinker. The temperature of the hot, marble-sized, glass-hard clinker is cooled by the clinkercooler. The cooled clinker is then sent to the final grinding mills.

ESPs are frequently used to control kiln emissions because of their ability to handle high-tem-perature gases. These ESPs are usually hot-side ESPs with collection plates that are rapped orsprayed with water to remove collected dust. The dust generated in the cement kiln frequentlyhas high resistivity. High resistivity can be reduced by conditioning the flue gas with moisture.Many of the newer cement plants send the high temperature kiln flue gas that contains particu-late matter through a cyclone and conditioning tower (uses water to cool the gas temperature)prior to ducting the flue gas to the ESP. The ESP is then operated at a temperature of approxi-mately 150°C (302°F).

A special problem arises during kiln startup due to the fact that the temperature of the kilnmust be raised slowly to prevent damage to the heat-resistant (refractory) lining in the kiln.While kilns (especially coal-fired ones) are warming up and temperatures are below those forsteady-state operating conditions, complete combustion of the fuels cannot occur, giving riseto combustible gases in the exhaust stream leading into the ESP. Electrostatic precipitatorscannot be activated in the presence of combustibles, because the internal arcing of the precipi-tator could cause a fire or explosion. Use of a cyclone preceding the precipitator helps to min-imize the excessive emissions during startup. Periods of excessive emissions during startup,malfunction, or shutdown are specifically exempted from the federal New Source Perfor-mance Standards for cement kilns.

ESPs can also be used on clinker coolers. However, the ESP must be carefully designed to pre-vent moisture in the flue gas from condensing. Condensed moisture can combine with clinkerdust to coat the ESP internals with cement. (A case history of an ESP used on a cement kiln isgiven in Szabo et al. 1981.)

Steel Mills

ESPs are used in steel mills for reducing particulate emissions from blast furnaces, basic oxy-gen furnaces, and sinter plants.

In a blast furnace, iron ore is reduced to molten iron, commonly called pig iron. Blast furnacesare large, refractory-lined steel shells. Limestone, iron ore, and coke are charged into the top ofthe furnace. The gases produced during the melting process contain carbon monoxide and par-ticulate matter. Particulate matter is removed from the blast furnace gas by wet ESPs or scrub-bers, so that the gas (CO) can be burned "cleanly" in blast furnace stoves or other processes.Both plate and tube-type ESPs having water sprays to remove dust from collection electrodesare commonly used for cleaning blast furnace gas.

Basic oxygen furnaces (BOFs) refine iron from the blast furnace into steel. A BOF is a pear-shaped steel vessel that is lined with refractory brick. The vessel is charged with molten ironand steel scrap. A water-cooled oxygen lance is lowered into the vessel, where oxygen isblown to agitate the liquid, add intense heat to the process, and oxidize any impurities still


2.0-2/98 5-5

contained in the liquid metal. The hot gases generated during the oxygen blow are approxi-mately 1090 to 1650°C (2000 to 3000°F). These are usually cooled by water sprays located inthe hood and ducting above the BOF. The cooled gases are then sent to an ESP or scrubber toremove the particulate matter (iron oxide dust). The iron oxide dust can have high resistivity,making the dust difficult to collect in an ESP. This problem can usually be reduced by condi-tioning the flue gas with additional moisture. Plate ESPs that are rapped or sprayed with waterto remove dust from collection plates are commonly installed on BOFs.

In a sinter plant, materials such as flue dusts, iron ore fines (small particles), coke fines, millscale (waste that occurs from various processing steps), and small scrap are converted into ahigh-quality blast furnace feed. These materials are first fed onto a traveling grate. The bed ofmaterials is ignited by burning gas in burners located at the inlet of the traveling grate. As thebed moves along the traveling grate, air is pulled down through the bed to burn it, forming afused, porous, red-hot sinter. The resulting gases are usually sent to an electrostatic precipita-tor to remove any particulate matter. If oily scrap is used as a feed material, care must be takento prevent ESP collection plates from being coated with tarry particulate matter. Controllingthe amount of oily mill scale and small scrap processed in the sinter plant can help alleviatethis problem. Plate ESPs are commonly used in sinter plants.

Petroleum Refineries

ESPs are used in petroleum refineries to control particulate emissions from fluid-catalyticcracking units and boilers. In a refinery, heavy crude is broken down into lighter componentsby various distilling, cracking, and reforming processes. One common process is to "crack"the high-molecular-weight, high-boiling-point compounds (heavy fuel oils) into smaller, low-molecular-weight, low-boiling-point compounds (gasoline). This is usually done in a fluid-catalytic cracking (FCC) unit.

In an FCC unit, the feed stream (heavy gas oils) is heated and then mixed with a hot catalystthat causes the gas oils to vaporize and crack into smaller hydrocarbon-chain compounds.During this process, the catalyst becomes coated with coke. The coke deposits are eventuallyremoved from the catalyst by a catalyst-regeneration step.

In the regenerator, a controlled amount of air is added to burn the coke deposits off the catalystwithout destroying it. The gases in the regenerator pass through cyclones to separate large cat-alyst particles. The gases can sometimes go to a waste heat boiler to burn any carbon monox-ide and organic emissions present in the gas stream. The boiler's exhaust gas still has a highconcentration of fine catalyst particles. This flue gas is usually sent to an electrostatic precipi-tator to remove the very fine catalyst particles.

ESPs can also reduce particulate emissions from boiler exhausts. Oil-fired and, occasionally,coal-fired boilers generate steam that is used in many processes in the refinery. The flue gasfrom boilers is frequently sent to ESPs to remove particulate matter before the gas is exhaustedinto the atmosphere. ESPs designed similarly to those used on industrial and utility boilers areused on FCC units and petroleum refinery boilers.

Municipal Waste Incinerators

Electrostatic precipitators have been successfully used for many years to reduce particulateemissions from municipal waste incinerators. Municipal incinerators, also commonly calledmunicipal waste combustors (MWCs) are used to reduce the volume of many different solid

Lesson 5

5-6 2.0-2/98

and liquid wastes. Generally, municipal wastes are composed of combustible materials (e.g.paper, wood, rags, food, yard clippings, and plastic and rubber materials) and noncombustiblematerials (e.g. rocks, metal, and glass). MWCs burn waste and produce ash residue that is dis-posed of in landfills.

Both dry and wet plate ESPs are commonly used on municipal incinerators. Collected dust canbe removed from collection plates by rapping or by using water sprays. Plate ESPs havingrigid frame discharge electrodes are currently being used on MWCs (installed after 1982). Thedesigned collection efficiency is usually in the range of 96 to 99.6%. Dust resistivity can be aproblem, particularly if the refuse contains a large quantity of paper products. The dust in theflue gas in this case usually has low resistivity. Resistivity can be adjusted by carefully con-trolling the temperature and the amount of moisture in the flue gas.

Since the mid-1980s a number of large MWCs (plants having a capacity of 250 to 3000 tonsper day) with heat recovery devices have been built. More recent installations have been builtwith acid gas control systems along with an ESP or baghouse. The ESP (or baghouse) collectsacid gas reaction products (mainly calcium chloride and calcium sulfate), unused sorbentmaterial, and fly ash. ESPs are typically designed with 3 to 5 fields and are capable of meetingparticulate emission limits of 0.015 gr/dscf and occasionally can achieve limits as low as 0.01gr/dscf. These units have successfully reduced SO2 by 80% (24 hr avg) and HCl by 90 to 95%.

The acid gas is removed by using dry sorbent injection or spray dryer absorbers. In dry injec-tion systems sorbent is injected (usually hydrated lime) into the furnace or into the ductingprior to the flue gas entering the ESP. Acid gas removal efficiencies of 50% for SO2 and 75%for HCl are routinely achieved (Beachler 1992).

A more commonly used acid gas control system is a spray dryer absorber placed ahead of theESP. These systems have been able to achieve 80% removal (24 hr avg) for SO2 and 90%removal for HCl. A wet calcium hydroxide slurry is injected into a spray dryer by a rotaryatomizer or two-fluid nozzle. The slurry is made by slaking pebble lime (CaO) with water in apaste or detention slaker. The heat of the flue gas evaporates the liquid slurry in the spray dryerand the dry acid gas reaction products along with the particulate matter are collected in theESP. Background information and data prepared as part of the promulgated NSPS and Emis-sion Guidelines (U.S. EPA 1991) shows very good acid gas removal and particulate emissioncontrol for these systems.

Hazardous Waste Incinerators

ESPs are used in combination with a number of other air pollution control (APC) devicesincluding wet scrubbers and dry scrubbers (also called spray dryer absorbers) to clean the fluegas generated by burning hazardous wastes. Some facilities have been designed to use spraydryers to remove the acid gases including HCl, HF, and SO2 followed by the ESP to removethe acid gas reaction salts, any unused sorbent, and particulate matter. Other facilities havebeen designed with an APC system consisting of a spray dryer, baghouse, wet scrubber, and awet ESP (Figure 5-2). The spray dryer cools the flue gas and reduces some of the acid gascomponents. The baghouse collects the particulate matter (including metals) and the wetscrubber removes HCl (> 99%) and other acid gases. The wet ESP collects any particulatematter not removed by the baghouse. The wet scrubbing system is a closed loop. The effluentproduced in the scrubbers is ultimately sent to the spray dryer to evaporate the liquid, therefore


2.0-2/98 5-7

Incineration Process

Gas Flow

Liquid Flow

eliminating the need for a waste water treatment system. A number of facilities using this APCsystem configuration are permitted to burn PCBs and other Toxic Substance Control Act(TSCA) and Resource Conservation and Recovery Act (RCRA) wastes.

Figure 5-2. APC system for a hazardous waste incinerator consisting of aspray dryer, baghouse, wet scrubbers, and wet ESPs

Kraft Pulp and Paper Mills

Plate or tube-type ESPs are used in the kraft pulp and paper industry to reduce particulateemissions from the recovery furnace. In the kraft process of making pulp and paper, chemicalsare recovered by using evaporators, recovery furnaces, and reaction tanks. As part of the pulp-ing process, a waste product, black-liquor, is produced. After it is concentrated, the black-liquor concentrate is burned in the recovery furnace to provide heat and steam to various pro-cesses in the plant. The recovery furnace is essentially a boiler designed to effectively burn theblack-liquor concentrate. The resulting flue gas contains particulate matter that is usuallyremoved by an ESP before it is exhausted into the atmosphere. Dust can be removed from col-lection electrodes by rapping or by using water sprays.

Lead, Zinc, and Copper Smelters

Plate ESPs are used to reduce particulate emissions from a number of processes in the smelt-ing of lead, zinc, and copper metals. Since lead, zinc, and copper are found in sulfide oredeposits, the release of sulfur compounds is a problem during the smelting process. Beforebeing smelted, ore concentrates are often treated, or prepared, by two processes called sinter-ing and roasting. Sintering changes the physical form of a material, usually by taking an oremixture of large and fine pieces and fusing them into strong, porous products that can be usedin the smelting processes. ESPs are commonly used to reduce emissions from lead and zincsinter plants. ESPs are also effective in reducing emissions from zinc and copper roasters.Roasters prepare zinc and copper ores by removing unwanted materials such as sulfur. Theroasted ore is then sent to other refining processes to produce zinc and copper metals.

Lesson 5

5-8 2.0-2/98

Other Industries

ESPs are used in many other large and small industrial processes including glass melting, sul-furic acid production, food processing, and chemical manufacturing. Glass melting furnacesusually use hot-side ESPs because the flue gas temperature in this process is approximately230 to 260°C (450 to 500°F). Sulfuric acid plants usually use plate or tube-type ESPs to col-lect sulfuric acid mists. Collected mists are removed from collection electrodes by watersprays. Some smaller industries that produce coatings, resins, asphalt, rubber, textiles, plastics,vinyl, and carpet frequently use a small two-stage precipitator to control particles and smoke.The two-stage ESPs use liquid sprays to remove collected particles, smokes, and oils from thecollection plates.

Summary

Table 5-2 summarizes the information presented in this lesson for various industries that useESPs to reduce emissions.

Table 5-2. Summary of typical ESP applications (by industry)

Industry Process

MaterialCollected by

ESPESP Collection

Efficiency ESP FeaturesPotentialProblems

1. Industrial &utility boilers

Burning fossilfuels

Dry SO2controlsystems

Fly ash

Dry, alkalineproduct

> 99%

> 99%(particles);70-80% (SO2)

Hot-side andcold-sideESPs

Cold-side ESP(usually rigidelectrode orrigid frame)

Fly ash fromlow sulfurcoals hashighresistivity

2. Cement plants Cement kilns

Clinker coolers

Particulateemissions

Particulateemissions

> 99%

> 99%

Usually hot-side ESPswithcollectionplates.Rapped orsprayed withwater.

Hot-side orcold-sidedepending ongastemperature.

Dust often hashighresistivity.Combustiblegases arepresent whenkiln iswarming up.

Must preventmoisture influe gas fromcondensing

3. Steel mills Blast furnaces

Basic oxygenfurnaces

Carbonmonoxideandparticulatematter

Iron oxide dust

> 99%Particulatematter

> 99%

Wet ESPs.Both plateand tube withwater sprays.

Wet or dryplate ESPs

Iron oxide dustcan have highresistivity

Cont. on next page


2.0-2/98 5-9

Table 5-2. (continued)Summary of typical ESP applications (by industry)

Industry Process

MaterialCollected by

ESPESP Collection

Efficiency ESP FeaturesPotentialProblems

3. Steel mills(cont’d)

Sinter plant Particulatematter

> 99% Wet or dryplate ESPs

Oily scrapused as feedmaterial cancoat plateswith tarrysubstance

4. Petroleumrefineries

Fluid-catalyticcracking

Boileroperations

Catalystparticles

Particulatematter

> 99%

> 99%

Usually dryESPs

Usually dryESPs

5. Municipal wasteincinerators

Incinerationand heatrecovery

Acid controlsystems(spray dryeralong withESP)

Particulatematter

Acid gasreactionproducts,unusedsorbents

96-99.6%

> 99.5 (0.015gr/dscf) forparticulatematter. SO2

and HClreduced 80and 90%respectively

Wet and dryplate ESPs

Usually rigid-electrodesystems(newerfacilities)

Low resistivityof dust frompaperproducts

6. Hazardouswasteincinerators

Acid controlsystems

(1) Spray dryerandbaghousefollowed bywet ESPs

(2) Spray dryerfollowed byan ESP

Acid gasreactionproducts,unusedsorbents

> 99% (0.015gr/dscf)HCl removalefficiency> 95%

Wet ESPs ordry ESPswhen usedwith spraydryer

7. Kraft pulp andpaper mills

Recoveryfurnaceboilers

Particulatematter

> 99% Wet or dryESPs

8. Lead, zinc,copper smelters

Sinter plants

Roasting

Particulatematter

Particulatematter

> 99%

> 99%

Usually plateESPs

Usually plateESPs

Lesson 5

5-10 2.0-2/98

Suggested Reading

For more information about the specific industries discussed in this lesson see:

Beachler, D. S., J. A. Jahnke, G. T. Joseph and M. M. Peterson. 1983. Air Pollution Control Sys-tems for Selected Industries, Self-Instructional Guidebook. (APTI Course SI:431). EPA 450/2-82-006. U.S. Environmental Protection Agency.

U.S. Environmental Protection Agency. 1985. Operation and Maintenance Manual for Electro-static Precipitators. EPA 625/1-85/017.

U.S. Environmental Protection Agency. 1981. Inspection Manual for Evaluation of ElectrostaticPrecipitator Performance. EPA 340/1-79-007.


2.0-2/98 5-11

Review Exercise

1. ESPs reduce particulate emissions from which of the following industries?

a. Utility boilersb. Cement kilnsc. Steel furnaces (basic oxygen furnace) and sinter plantsd. Municipal waste incineratorse. All of the above

2. One technology for reducing both SO2 gas and particulate emissions involves the injection of a(an)____________________slurry in a spray ____________________ with dry particle collection inan electrostatic precipitator.

3. In a spray dryer, moisture is ____________________ from the wet alkaline sprays, leaving a____________________ powdered product.

4. Acid gas and particulate emissions can be controlled by using ____________________.

a. Spray dryer absorber and ESPb. Dry injection and ESPc. a and b, above

5. ESPs should not be activated during the startup of a(an) ________________________________________ because of the possibility of a fire or explosion.

6. In a steel mill, which of the following processes would not likely use an ESP to control particulateemissions?

a. Blast furnace meltingb. Sinter processc. Ingot pouringd. Basic oxygen furnace melting and tapping

7. In a municipal incinerator where the burned refuse contains a large quantity of paper products, theresulting dust usually has a ____________________ resistivity.

a. Highb. Low

8. True or False? ESPs are used in petroleum refineries to control particulate emissions from thefluid-catalytic-cracking unit and boiler exhausts.

Lesson 5

5-12 2.0-2/98

9. When a spray dryer absorber is used with an ESP to control acid gas and particulate emissionsfrom municipal waste combustors, which of the following is (are) true?

a. The control system can reduce particulate emissions to a level of less than 0.015 gr/dscf.b. The control system can reduce SO2 by 80%.c. The dust collected in the ESP hoppers contains calcium chloride which is very hygroscopic

(sticky).d. All of the above.

10. True or False? Both wet and dry ESPs are used in the pulp and paper industries to remove greaterthan 99% of the particulate matter from recovery furnace.


2.0-2/98 5-13

Review Exercise Answers1. e. All of the above

ESPs reduce particulate emissions from the following industries:

• Utility boilers• Cement kilns• Steel furnaces (basic oxygen furnace) and sinter plants• Municipal waste incinerators

2. AlkalineDryerOne technology for reducing both SO2 gas and particulate emissions involves the injection of analkaline slurry in a spray dryer with dry particle collection in an electrostatic precipitator.

3. EvaporatedDryIn a spray dryer, moisture is evaporated from the wet alkaline sprays, leaving a dry powderedproduct.

4. c. a and b, aboveAcid gas and particulate emissions can be controlled by using either a spray dryer absorber andESP or dry injection and ESP.

5. Cement kilnESPs should not be activated during the startup of a cement kiln because of the possibility of a fireor explosion.

6. c. Ingot pouringIn a steel mill, ingot pouring would not likely use an ESP to control particulate emissions.

7. b. LowIn a municipal incinerator where the burned refuse contains a large quantity of paper products, theresulting dust usually has a low resistivity.

8. TrueESPs are used in petroleum refineries to control particulate emissions from the fluid-catalytic-cracking unit and boiler exhausts.

9. d. All of the aboveWhen a spray dryer absorber is used with an ESP to control acid gas and particulate emissionsfrom municipal waste combustors, the following are true:

• The control system can reduce particulate emissions to a level of less than 0.015 gr/dscf.• The control system can reduce SO2 by 80%.• The dust collected in the ESP hoppers contains calcium chloride which is very hygroscopic

(sticky).

Lesson 5

5-14 2.0-2/98

10. TrueBoth wet and dry ESPs are used in the pulp and paper industries to remove greater than 99% of theparticulate matter from recovery furnaces.


2.0-2/98 5-15

Bibliography

Beachler, D. S. 1992. Coming clean on waste-to-energy emissions. Chemical Processing TechnologyInternational. London.

Beachler, D. S., G. T. Joseph, and M. Pompelia. 1995. Fabric Filter Operation Review. (APTI CourseSI:412A). U.S. Environmental Protection Agency.

Beachler, D. S., J. A. Jahnke, G. T. Joseph and M. M. Peterson. 1983. Air Pollution Control Systems forSelected Industries, Self-Instructional Guidebook. (APTI Course SI:431). EPA 450/2-82-006. U.S.Environmental Protection Agency.

Kaplan, S. M., and K. Felsvang. 1979, April. Spray Dryer Absorption of SO2 from Industrial BoilerFlue Gas. Paper presented at 86th National AICHE Meeting. Houston, TX.

Pezze, J. 1983. Personal Communication. Pennsylvania Department of Environmental Resources,Pittsburgh, PA.


Richards, J. R. 1995. Control of Gaseous Emissions, Student Manual. (APTI Course 415). U.S. Envi-ronmental Protection Agency.

Szabo, M. F., Y. M. Shah, and S. P. Schliesser. 1981. Inspection Manual for Evaluation of ElectrostaticPrecipitator Performances. EPA 340/1-79-007.

U.S. Environmental Protection Agency. 1976. Capital and Operating Costs of Selected Air PollutionControl Systems. EPA 450/3-76-014.

U.S. Environmental Protection Agency. 1980, February. Survey of Dry SO2 Control Systems. EPA 600/7-80-030.

U.S. Environmental Protection Agency. 1991. Requirements for preparation, adoption, and submittalof implementation plans. In Code of Federal Regulations—Protection of the Environment. 40 CFR51. Washington, D.C.: U.S. Government Printing Office.

U.S. Environmental Protection Agency. 1991. Approval and promulgation of implementation plans. InCode of Federal Regulations—Protection of the Environment. 40 CFR 52. Washington, D.C.: U.S.Government Printing Office.

U.S. Environmental Protection Agency. 1991. Standards of performance for new stationary sources—general provisions. In Code of Federal Regulations—Protection of the Environment. 40 CFR 60.Washington, D.C.: U.S. Government Printing Office.

5-16 2.0-2/98

2.0-2/98 6-1

Lesson 6ESP Operation and Maintenance

Goal

To familiarize you with typical operation and maintenance problems associated with ESPs.

Objectives


1. Identify typical ESP components which require inspection prior to startup2. Identify the major steps in ESP startup and shutdown procedures3. Explain the importance of monitoring each of the following parameters:

• Voltage/current

• Opacity

• Gas temperature

• Gas flow rate and distribution

• Gas composition and moisture

4. Describe the function of air-load and gas-load voltage-current curves5. Identify typical maintenance steps that ensure proper ESP functioning6. Identify and describe seven common problems that affect ESP performance7. Describe how evaluating the current, voltage, and spark rate trends from inlet to outlet fields

provides information about general resistivity conditions8. Identify important safety precautions to take when operating ESPs

Introduction

As with any air pollution control system, an ESP must be operated and maintained accordingto the manufacturer's recommendations. Plant personnel must be properly trained to performthese activities with confidence and efficiency. This lesson reviews some of the key functionsthat must be completed to keep the ESP operating as it was intended including installation,startup and shutdown procedures, performance monitoring, routine maintenance and record-keeping and problem evaluation.

Lesson 6

6-2 2.0-2/98

ESP Installation

Depending on the electrostatic precipitator chosen, production, installation and operation star-tup may take from a few months to one or two years. In any case, proper installation proce-dures will save time and money, and will also help in future operation and maintenance(O&M) of the ESP.

Good coordination between the ESP designer (vendor) and the installation and maintenancecrews will help keep the ESP running smoothly for years. Occasionally this coordination isoverlooked. Because they are so large, ESPs are usually installed by skilled craftsmen who donot work for the ESP vendor, and, therefore, may not be informed of specific installationinstructions. Since all design tolerances are critical (especially those affecting discharge andcollection electrode alignment), it is imperative that information about the proper installationprocedures be transferred from designers to installers.

Some key considerations during installation are:

• Easy access to all potential maintenance areas—fans, motors, hoppers, discharge devices,dampers, flue gas flow rate and temperature monitors, insulators, rappers, T-R sets, anddischarge and collection electrodes

• Easy access to all inspection and test areas—stack testing ports and continuous emissionmonitors (opacity monitors)

• Weather conditions—the ESP must be able to withstand inclement weather such as rain orsnow

During installation, the customer purchasing the ESP should be responsible for checking thecriteria presented below. The regulatory agency review engineer also should review the pro-cess on which the ESP will be installed and verify that these items are being addressed.

1. Uniform flue gas distribution across the entire unit. Ductwork, turning vanes, baffleplates, and inlets with perforated diffuser plates all affect flue gas distribution. These itemsare usually installed in the field and should be checked visually. If improperly installed,they induce high airflow regions that decrease collection efficiency and cause reentrain-ment of collected dust, especially during rapping cycles.

2. Complete seal of ESP system from dust pickup to stack outlet. Air inleakage or outleakageat flanges or collector access points either adds additional airflow to be processed or forcesthe process gases to bypass the collector. Inleakage to a high-temperature system (hot-sideESP) is extremely damaging, as it creates cold spots which can lead to moisture or acidcondensation and possible corrosion. If severe, it can cause the entire process gas temper-ature to fall below the gas dew point, causing moisture or acid to condense on the hopperwalls, the discharge electrode, or collection plates. In addition, air inleakage and moisturecondensation can cause caking of fly ash in the hopper, making normal dust removal bythe discharge device very difficult. The best way to check for leaks is an inspection of thewalls from inside the system during daylight. Light penetration from outside helps to iso-late the problem areas.

3. Proper installation of discharge electrodes and collection plates. Collection electrodes areusually installed first, and the discharge wires or rigid frames are positioned relative tothem. Check each section of electrodes to ensure that the electrodes are plumb, level, andproperly aligned.

ESP Operation and Maintenance

2.0-2/98 6-3

4. Proper installation of rappers. Collection-plate rappers and discharge-electrode rappersshould be installed and aligned according to vendor specifications. Check magnetic-impulse rappers to see if they strike the support frame on the collection plates. Check ham-mer and anvil rappers to see if the hammers strike the anvils squarely. Check vibrator rap-pers installed on discharge wires to make sure they operate when activated. Rapperfrequency and intensity can be adjusted later when the unit is brought on-line.

5. Proper insulation. Most ESPs use some type of insulation to keep the flue gas temperaturehigh. This prevents any moisture or acids present in the flue gas from condensing on thehoppers, electrodes, or duct surfaces. Because most ESPs are installed in the field, checkthat all surfaces and areas of potential heat loss are adequately covered.

6. Proper installation and operation of discharge devices. It is important to check the opera-tion of the discharge devices before bringing the ESP on-line to see if they are properlyinstalled. Make sure that the discharge devices are moving in the right direction so theycan remove the dust freely from the hopper. A backward-moving screw conveyor can packdust so tightly that it can bend the screw.

Overfilled hoppers are common operating problems that can be avoided by proper installa-tion and maintenance of discharge devices. Installed as maintenance tools, dust-leveldetectors in the hoppers can help alert ESP operators that hoppers are nearly full.

7. Smoothly running fans. Check fans for proper rotation, drive component alignments, andvibration. Fans should be securely mounted to a component of sufficient mass to eliminateexcessive vibration.

In addition to the above items, each ESP installation should have its own checklist reflect-ing the unique construction features of that unit. The installation crew should prepare achecklist before beginning final inspection and initial startup. A prestartup checklist forthe initial startup suggested by Peter Bibbo (1982) is shown in Table 6-1.

Lesson 6

6-4 2.0-2/98

ESP Startup and Shutdown

A specific startup and shutdown procedure should be supplied by the ESP vendor. Improperstartup and shutdown can damage the collector. It is imperative for the operator (source) tohave a copy of these procedures. Review agency engineers may want to assure that these pro-

Table 6-1. Prestartup checklist for electrostaticprecipitators

Collecting plates1. Free of longitudinal and horizontal bows2. Free of burrs and sharp edges3. Support system square and level4. Spacer bars and corner guides free5. Free of excessive dust buildup6. Gas leakage baffles in place and not binding

Discharge electrodes1. No breaks or slack wires2. Wires free in guides and suspension weight free on pin3. Rigid frames square and level4. Rigid electrodes plumb and straight5. Free of excessive dust buildup and grounds6. Alignment within design specifications

Hoppers1. Scaffolding removed2. Discharge throat and poke holes clear3. Level detector unobstructed4. Baffle door and access door closed5. Heaters, vibrators, and alarms operational

Top housing or insulator compartments1. Insulators and bushing clear and dry with no carbon tracks2. All grounding chains in storage brackets3. Heaters intact, seal-air system controls, alarms, dampers, and filters

in place and operational4. Seal-air fan motor rotation correct, or vent pipes free5. All access doors closed

Rappers1. All swing hammers or drop rods in place and free2. Guide sleeves and bearings intact3. Control and field wiring properly terminated4. Indicating lights and instrumentation operational5. All debris removed from precipitator6. All personnel out of unit and off clearances7. All interlocks operational and locked out

a. No broken or missing keysb. Covers on all locks

Transformer-rectifier sets1. Surge arrestor not cracked or chipped and gap set2. Liquid level satisfactory3. High-voltage connections properly made4. Grounds on: precipitator, output bushings, bus ducts, conduits

Rectifier control units1. Controls grounded2. Power supply and alarm wiring properly completed3. Interlock key in transfer block

Source: Bibbo 1982.


2.0-2/98 6-5

cedures exist at the sites and that the operators follow the procedures or document reasons fordeviations.

Startup

Startup of an electrostatic precipitator is generally a routine operation. It involves heatinga number of components such as support insulators and hoppers. If possible, the ESPshould not be turned on until the process reaches steady-state conditions. As described inLesson 5, this is particularly important for ESPs used on cement kilns burning coal as fuel.The internal arcing of the ESP could cause a fire or an explosion. When ESPs are used onoil-burning boilers, the boiler should be started with gas or #2 fuel oil. Heavy oil (#6 fueloil) is not a good fuel for startup because tarry particulate emissions can coat collectionplates and be difficult to remove. If an ESP is used on a coal-fired boiler, the ESP shouldnot be started until coal firing can be verified. This will help prevent combustible gasesfrom accumulating in the unit and causing explosive conditions. A typical startup proce-dure for an ESP used on a boiler is given in Table 6-2 (Bibbo 1982).

Table 6-2. Typical startup procedures forelectrostatic precipitators

Normal Operation

Startup (preoperational checks - at least 2 hours prior to gas load):1. Complete all maintenance/inspection items.2. Remove all debris from ESP.3. Safety interlocks should be operational and all keys accounted for.4. No personnel should be in ESP.5. Lock out ESP and insert keys in transfer blocks.

Prestart (at least one hour prior to gas load):6. Check hoppers.

a. Level-indicating system should be operational.b. Ash-handling system operating and sequence check - leave in

operational mode.c. Hopper heaters should be on.

7. Check top housing seal-air system.a. Check operation of seal-air fan—leave running.b. Bushing heaters should be on.

8. Check rappers.a. Energize control, run rapid sequence, ensure that all rappers are

operational.b. Set cycle time and intensity adjustments, using installed

instrumentation—leave rappers operating.9. Check T-R sets.

a. Check half-wave/full-wave operation (half-wave operation isrecommended for filtering fly ash when lignite is burned and acold-side ESP is used.)

b. Keys should be in all breakers.c. Test-energize all T-R sets and check local control alarm functions.d. Set power levels and de-energize all T-R controls.

e. Lamp and function-test all local and remote alarms.Continued on next page

Lesson 6

6-6 2.0-2/98

Shutdown

When an industrial process is shut down temporarily, the ESP system should be de-ener-gized to save energy. The shutdown of the ESP is usually done by reversing the order ofthe startup steps. Begin with de-energizing the ESP fields starting with the inlet field tomaintain appropriate opacity levels from the stack. The rappers should be run for a shorttime after the ESP is de-energized so that accumulated dust from the collection plates anddischarge wires can be removed. All hoppers should be emptied completely before bring-ing the unit back on line. A typical shutdown and emergency shutdown procedure forESPs used on industrial sources is given in Table 6-3 (Bibbo 1982).

Table 6-2. (continued)Typical startup procedures forelectrostatic precipitators

Normal OperationGas load:

10. After gas at temperature of 200°F has entered ESP for 2 hours -a. Energize T-R sets.b. Check for normal operation of T-R control.c. Check all alarm functions in local and remote.d. Within 2 hours, check proper operation of ash removal system.e. De-energize bushing heaters after 2 hours (hopper heaters

optional).Cold start (when it is not possible to admit flue gas at 200°F for 2 hours prior

to energizing controls), proceed as follows:1. Perform steps 1-9 above. Increase rapping intensity 50%.2. Energize T-R sets, starting with inlet field, setting Powertrac voltage

to a point just below sparking.3. Successively energize successive field as load picks up to maintain

opacity, keeping voltage below normal sparking (less than 10flashes/min on spark indicator).

4. Perform step 10d above.5. After flue gas at 200°F has entered ESP for 2 hours, perform steps

10b, c, d, and e above.Set normal rapping.

Source: Bibbo 1982.


2.0-2/98 6-7

Performance Monitoring

As with the operation of any piece of equipment, performance monitoring and recordkeepingare essential to establishing a good operation and maintenance program. The key to any moni-toring program is establishing an adequate baseline of acceptable ranges that is used as a refer-ence point. Then, by monitoring and recording key operating parameters, the operator canidentify performance problems, need for maintenance, and operating trends.

Typical parameters that can be monitored include:

• Voltage/current

• Opacity

• Gas temperature

• Gas flow rate and distribution

• Gas composition and moisture

In addition, site-specific data on process operating rates and conditioning system (if used)should also be documented. Operators should not rely on just one parameter as an indicator ofperformance—trends for a number of parameters gives a clearer picture. Let's briefly look atthe ways these parameters affect performance and the techniques used to measure them. Muchof this information was extracted from Operation and Maintenance Manual for ElectrostaticPrecipitators (U.S. EPA 1985).

Voltage and Current

Voltage and current values for each T-R set should be recorded; they indicate ESP perfor-mance more than any other parameter. Most modern ESPs are equipped with primary volt-age and current meters on the low-voltage (a.c.) side of the transformer and secondaryvoltage and current meters on the high-voltage rectified (d.c.) side of the transformer.When both voltage and current meters are available on the T-R control cabinet, these val-ues can be multiplied to estimate the power input to the ESP. (Note that the primary cur-

Table 6-3. Typical shutdown and emergency shutdownprocedures for electrostatic precipitators

Typical shutdown1. When boiler load drops and total ash quantity diminishes:

a. De-energize ESP by field, starting with inlet field to maintain opacitylimit.

b. De-energize outlet field when all fuel flow ceases and combustion airflow falls below 30% of rated flow.

c. Leave rappers, ash removal system, seal-air system, and hopperheaters operational.

d. Four hours after boiler shutdown, de-energize seal-air system andhopper heaters. Secure ash removal system.

e. Eight hours after boiler shutdown, de-energize rappers.Note: Normal shutdown is a convenient time to check operation of

alarms.

Emergency shutdown1. De-energize all T-R sets.2. Follow steps 1c, d, and e above (shutdown).

Source: Bibbo 1982.

Lesson 6

6-8 2.0-2/98

rent reading is multiplied by the primary voltage reading and the secondary currentreading is multiplied by the secondary voltage reading). These values (current times volt-age) represent the number of watts being drawn by the ESP and is referred to as the coronapower input. In addition, whenever a short term spark occurs in a field it can be detectedand counted by a spark rate meter. ESPs generally have spark rate meters to aid in the per-formance evaluation.

The power input on the primary versus the secondary side of the T-R set will differbecause of the circuitry and metering of these values. The secondary power outlet (inwatts) is always less than the primary power input to the T-R. The ratio of the secondarypower to the primary power will range from 0.5 to 0.9 and average from 0.70 to 0.75 (U.S.EPA 1985).

Voltage and current values for each individual T-R set are useful because they inform theoperators how effectively each field is operating. However, the trends noted within theentire ESP are more important. T-R set readings for current, voltage, and sparking ratesshould follow certain patterns from the inlet to the outlet fields. For example, coronapower density should increase from inlet to outlet fields as the particulate matter isremoved from the gas stream.

The electrical meters on the T-R cabinets are always fluctuating. Normal sparking withinthe ESP causes these fluctuations in the meter readings. These short term movements ofthe gauges indicate that the automatic voltage controller is restoring the maximum voltageafter shutting down for several milliseconds to quench the spark. When recording valuesof the electrical data from the T-R meters it is important to note the maximum value that issustained for at least a fraction of a second.

Opacity

In many situations, ESP operation is evaluated in terms of the opacity monitored by atransmissometer (opacity monitor) on a real-time basis. Under optimum conditions theESP should be able to operate at some base-level opacity with a minimum of opacity spik-ing from rapper reentrainment. A facility can have one or more monitors that indicateopacity from various ESP outlet ducts and from the stack.

An opacity monitor compares the amount of light generated and transmitted by the instru-ment on one side of the gas stream with the quantity measured by the receiver on the otherside of the gas stream. The difference, which is caused by absorption, reflection, refrac-tion, and light scattering by the particles in the gas stream, is the opacity of the gas stream.Opacity is expressed as a percent from 0 to 100% and is a function of particle size, con-centration, and path length.

Most of the opacity monitors being installed today are double-pass monitors; that is, thelight beam is passed through the gas stream and reflected back across to a transceiver. Thisarrangement is advantageous for several reasons:

1. Automatic checking of the zero and span of the monitor is possible when the processis operational.

2. The monitor is more sensitive to slight variations in opacity because the path length islonger.

3. The entire electronics package is located on one side of the stack as a transceiver.


2.0-2/98 6-9

Although single-pass transmissometers are available at a lower cost (and sensitivity), thesingle-pass monitors cannot meet the requirements in EPA Performance Specification 1,Appendix B, 40 CFR 60.

For many sources, dust concentration and opacity correlations can be developed to pro-vide a relative indication of ESP performance. These correlations are very site-specific,but can provide plant and agency personnel with an indication of relative performance forgiven opacity levels. In addition, site-specific opacity charts can be used to predict deteri-oration of ESP performance that requires attention by plant personnel. Readings fromopacity monitors can also be used to optimize spark rate, voltage/current levels, and rap-ping cycles, even though the conditions within the ESP are not static. In high-efficiencyESPs, reentrainment may account for 50 to 70% of the total outlet emissions. Therefore,optimization of the rapping pattern may prove more beneficial than trying to optimize thevoltage, current and sparking levels. Dust reentrainment from rapping must be observedby using the opacity monitor operating in a real-time or nonintegrating mode because rap-ping spikes tend to get smoothed out in integrated averages such as the 6-minute averagecommonly in use. However, the integrated average does provide a good indication of aver-age opacity and emissions.

When parallel ESPs or chambers are used, an opacity monitor is often placed in each out-let duct, as well as on the stack, to measure the opacity of the combined emissions.Although the stack monitor is commonly used to indicate stack opacity (averaging opaci-ties from different ducts can be difficult), the individual duct monitors can be useful inindicating the performance of each ESP or chamber and in troubleshooting. Although thisoption is often not required and represents an additional expense, it can be very useful,particularly on relatively large ESPs.

New systems, such as the digital microprocessor design, are available in which the opacitymonitor data can be used as input for the T-R controller. In this case, the data are used tocontrol power input throughout the ESP to maintain an opacity level preselected by thesource. If the opacity increases, the controller increases power input accordingly until theopacity limit, spark limit, current limit, or voltage limit is reached. This system (often soldas an energy saver because it uses only the power required) can save a substantial quantityof energy:

1. On large, high-efficiency ESPs

2. For processes operating at reduced gas loads.

In many cases, however, reduction of ESP power does not significantly alter ESP perfor-mance because dust reentrainment and gas sneakage constitute the largest sources of emis-sions; additional power often does not reduce these emissions significantly. In someobserved cases, reducing power by one-half did not change the performance. For unitstypically operated at 1000 to 1500 watts/1000 acfm, operating the ESPs at power levels of500 to 750 watts/1000 afcm still provide acceptable collection efficiencies.

Gas Temperature

Monitoring the temperature of the gas stream can provide useful information concerningESP performance. Temperature is measured using a thermocouple in conjunction with adigital, analog, or strip chart recorder. Temperature is usually measured using a single-point probe or thermocouple. This method has a major limitation in that the probe may be

Lesson 6

6-10 2.0-2/98

placed at an unrepresentative (stratified) point—one that is not representative of the bulkgas flow. Most ESPs are designed with a minimum of three fields. The gas temperature foreach field should be measured at both the inlet and outlet, if possible. Significant tempera-ture changes between the inlet and outlet values may indicate air inleakage problems thatshould be confirmed by measurement of gas composition.

Changes in gas temperature can have profound effects on ESP performance. The tempera-ture variation can be very small (in some cases as little as 15oF) and yet cause a significantchange in ESP power levels and opacity. Although gas temperature variations may havesome effect on corona discharge characteristics and physical characteristics of the ESP(corrosion, expansion/contraction), their most important effect is on particle resistivity.For sources with the potential for high resistivity, temperature changes can cause dramaticchanges in performance, even when all other parameters seem to be the same. The gastemperature should be checked once per shift for smaller sources and measured continu-ously on larger sources and on those sources with temperature-sensitive performance.

Temperature measurement can also be a useful tool in finding excessive inleakage orunequal gas flow through the ESP. Both of these conditions can affect localized gas veloc-ity patterns without noticeably affecting the average velocity within the ESP. Yet, local-ized changes in gas velocities can reduce ESP performance even though the average gasvelocity seems adequate.

Gas Flow Rate and Distribution

Gas flow rate determines most of the key design and operating parameters such as spe-cific collection area (ft2/1000 acfm), gas velocity (ft/sec) and treatment time within theESP, and specific corona power (watts/1000 acfm). The operator should calculate the fluegas flow rate if the ESP is not operating efficiently. For example, significant variations inoxygen may indicate large swings in the gas flow rate that may decrease ESP performanceand indicate the need to routinely determine ESP gas volume. Low SCA values, highvelocities, short gas treatment times (5 seconds or less), and much higher oxygen levels atnearly full load conditions are indicators that excess flue gas flow rate may be causingdecreased ESP performance.

Presently, most sources do not continuously measure gas velocities or flow rates. Gasvelocities are generally only measured during emission compliance testing or when thereis a perceived problem. Manual pitot tube traverses are normally used to measure gasvelocity (EPA Reference Methods 1 and 2). Because of new technologies and regulations,some of the larger sources are beginning to install continuous flow measurement systems.Multi-point pitot devices, ultrasonic devices, and temperature-based flow devices can beused to continuously measure gas velocity. These devices must be calibrated to the indi-vidual stack where they are installed. Most existing facilities currently use indirect indica-tors to estimate gas flow rate; these include fan operating parameters, production rates oroxygen/carbon dioxide gas concentration levels. However, EPA is now requiring largecoal-fired utility boilers to install and certify flow monitors (EPA Acid Rain Program, Part75 Regulations).

Another important parameter is gas flow distribution through the ESP. Ideally, the gasflow should be uniformly distributed throughout the ESP (top to bottom, side to side).Actually, however, gas flow through the ESP is not evenly distributed, and ESP manufac-turers settle for what they consider an acceptable variation. Standards recommended by


2.0-2/98 6-11

the Industrial Gas Cleaning Institute have been set for gas flow distribution. Based on avelocity sampling routine, 85% of the points should be within 15% of the average velocityand 99% should be within 1.4 times the average velocity. Generally, uneven gas flowthrough the ESP results in reduced performance because the reduction in collection effi-ciency in areas of high gas flow is not compensated for by the improved performance inareas of lower flow. Also, improper gas distribution can also affect gas sneakage throughthe ESP. As stated earlier, good gas distribution can be accomplished by using perforatedplates in the inlet plenum and turning vanes in the ductwork.

Measurement of gas flow distribution through the ESP is even less common than measur-ing flue gas flow rate. Because the flow measurements are obtained in the ESP rather thanthe ductwork (where total gas volumetric flow rates are usually measured), more sensitiveinstrumentation is needed for measuring the low gas velocities. The instrument typicallyspecified is a calibrated hot-wire anemometer. The anemometer test is usually performedat some mid-point between the inlet and outlet (usually between two fields). Care must betaken to assure that internal ESP structural members do not interfere with the samplingpoints.

Gas flow distribution tests are conducted when the process is inoperative, and the ESP andductwork are relatively cool. This often limits the amount of gas volume that can be drawnthrough the ESP to less than 50% of the normal operating flow; however, the relativevelocities at each point are assumed to remain the same throughout the normal operatingrange of the ESP. A large number of points are sampled by this technique. The actual num-ber depends upon the ESP design, but 200 to 500 individual readings per ESP are notunusual. By using a good sampling protocol, any severe variations should become readilyapparent.

Gas Composition and Moisture

The chemical composition of both the particulate matter and flue gas can affect ESP per-formance. In many applications, key indicators of gas composition are often obtained byusing continuous emission monitors. However, particle concentration and composition aredetermined by using intermittent grab sampling.

The operation of an ESP depends on the concentration of electronegative gases O2, H2O,CO and SO2/SO3 to generate an effective corona discharge. Often, sources use continuousmonitors to measure these gas concentrations to meet regulatory requirements, or in thecase of combustion sources to determine excess air levels (CO2 or O2).

Evaluating Air-Load/Gas-Load Voltage-Current (V-I) Curves

In addition to the routine panel meter readings, other electrical tests of interest to personnelresponsible for evaluating and maintaining ESPs include the air-load and gas-load V-I (volt-age-current) tests, which may be conducted on virtually all ESPs. Air-load and gas-load curvesare graphs of the voltage (kV) versus the current (mA) values obtained at a set condition (testpoint). These curves are developed to evaluate ESP performance by comparing the graphsfrom inlet field to outlet field and over periods in time. Deviation from the normal or previousresults can indicate that a problem exists.

Lesson 6

6-12 2.0-2/98

Air-Load Curves

Air-load tests are generally conducted on cool, inoperative ESPs through which no gas isflowing. This test should be conducted when the ESP is new, after the first shutdown, andevery time off-line maintenance is performed on the ESP. These air-load V-I curves serveas the basis for comparison in the evaluation of ESP maintenance and performance. A typ-ical air-load curve is shown in Figure 6-1.

Figure 6-1. Typical air-load voltage and current readings

An air-load V-I curve can be generated with readings from either primary or secondarymeters. The following procedures can be used by the ESP operator to develop an air-loadcurve.

1. Energize a de-energized T-R set on manual control (but with zero voltage and current),and increase the power to the T-R set manually.

2. At corona initiation the meters should suddenly jump and the voltage and near zerocurrent levels should be recorded. It is sometimes difficult to identify this point pre-cisely, so the lowest practical value should be recorded.

3. After corona initiation is achieved, increase the power at predetermined increments[for example, every 50 or 100 milliamps of secondary current or every 10 volts of ACprimary voltage (the increment is discretionary)], and record the values for voltageand current.

4. Continue this procedure until one of the following occurs:

• Sparking

• Current limit is achieved

• Voltage limit is achieved

5. Repeat this procedure for each T-R set.

When the air-load tests have been completed for each field, plot each field's voltage/cur-rent curves. When ESPs are equipped with identical fields throughout, the curves for eachfield should be nearly identical. In most cases, the curves also should be similar to those


2.0-2/98 6-13

generated when the unit was new, but shifted slightly to the right due to residual dust onthe wires (or rigid frames) and plates of older units. These curves should become part ofthe permanent record of the ESP.

The use of air-load curves enables plant personnel to identify which field(s) may not beperforming as designed. Also, comparison of air-load curves from test runs taken justbefore and after a unit is serviced will confirm whether the maintenance work correctedthe problem(s).

One major advantage of air-load tests is that they are performed under nearly identicalconditions each time, which means the curves can be compared. One disadvantage is thatthe internal ESP conditions are not always the same as during normal operation. Forexample, misalignment of electrodes may appear or disappear when the ESP is cooled(expansion/contraction), and dust buildup may be removed by rapping during ESP shut-down.

Gas-Load Curves

The gas-load V-I curve, on the other hand, is generated during the normal operation of theprocess while the ESP is energized. The procedure for generating the gas-load V-I curve isthe same as for the air load except that gas-load V-I curves are always generated from theoutlet fields first and move toward the inlet. This prevents the upstream flow that is beingchecked from disturbing the V-I curve of the downstream field readings. Although suchdisturbances would be short-lived (usually 2 minutes, but sometimes lasting up to 20 min-utes), working from outlet to inlet speeds up the process.

The curves generated under gas-load conditions will be similar to air-load curves. Gas-load curves will generally be shifted to the left however, because sparking occurs at lowervoltage and current when particles are present. The shape of the curve will be different foreach field depending on the presence of particulate matter in the gas stream (seeFigure 6-2).

Figure 6-2. Comparison of typical air-load and gas-loadV-I curves Source:EPA 1985

Gas-load

Air-load

I

V

Lesson 6

6-14 2.0-2/98

Also, gas-load curves vary from day to day, even minute to minute. Curve positions maychange as a result of fluctuations in the following:

• Amount of dust on plates

• Gas flow

• Particulate chemistry loading

• Temperature

• Resistivity

Nonetheless, they still should maintain a characteristic pattern. Gas-load curves are nor-mally used to isolate the cause of a suspected problem rather than being used on a day-to-day basis; however, they can be used daily if necessary.

Routine Maintenance and Recordkeeping

While the overall performance of the ESP is continuously monitored by devices such as volt-age meters and transmissometers, the components of the ESP and their operation are periodi-cally inspected by plant personnel as part of a preventive maintenance program. In this way,problems are detected and corrected before they cause a major shutdown of the ESP. Ofcourse, good recordkeeping should be an integral part of any maintenance program.

The frequency of inspection of all ESP components should be established by a formal in-housemaintenance procedure. Vendors' recommendations for an inspection schedule should be fol-lowed. A listing of typical periodic maintenance procedures for an ESP used to collect fly ashis given in Table 6-4 (Bibbo 1982).

In addition to the daily monitoring of meters and the periodic inspection of ESP components,some operational checks should be performed every shift and the findings should be recordedon a shift data sheet. At the end of every shift, these shift data sheets should be evaluated formaintenance needs. These once-per-shift checkpoints include an inspection of rappers, dustdischarge systems, and T-R sets for proper functioning and an indication of which T-R sets arein the "off" position. Rappers that are not functioning should be scheduled for maintenance,particularly if large sections of rappers are out of service. Dust discharge systems should havehighest priority for repair; dust should not accumulate in the bottom of the ESP for long peri-ods of time because of the potential for causing severe plate misalignment problems. Hopperheaters can usually be repaired with little difficulty after removing weather protection andinsulation. Insulator heaters may be difficult to repair except during short outages. Hopperheaters keep condensation on the insulators to a minimum and help keep the dust warm andfree-flowing.

In addition to performing maintenance, keeping records of the actions taken is also important.For example, wire replacement diagrams should be kept. Although an ESP can operate effec-tively with up to 10% of its wires removed, care must be taken that no more than 5 to 10 wiresin any one gas lane are removed. The loss of wires down any one lane can result in a substan-tial increase in emissions. The only way to adequately track where wires have failed or slippedout of the ESP is with a wire replacement chart. Also, any adjustments to the rapper frequencyand intensity should be recorded along with any repairs. These same recordkeeping practicesshould be followed for any repairs or replacements made on T-R sets, insulator/heaters, align-ment, and the dust discharge systems.


2.0-2/98 6-15

Problem Evaluation

Good site specific records of both the design and operating history will enable operating per-sonnel to better evaluate ESP performance. Design parameters built into the ESP include thefollowing: the specific collection area (SCA), number of fields, number of T-R sets, sectional-ization, T-R set capacity, design velocity and treatment time, aspect ratio and particle charac-teristics (resistivity). Design records indicate the specific conditions under which the ESP wasdesigned to operate. A comparison between design records and operating records indicatewhether operating parameters have changed significantly from the design conditions. Sec-

Table 6-4. Preventive maintenance checklist for a typicalfly ash precipitator

Daily1. Take and record electrical readings and transmissometer data.2. Check operation of hoppers and ash removal system.3. Examine control room ventilation system.4. Investigate cause of abnormal arcing in T-R enclosures and bus duct.

Weekly1. Check rapper operation.2. Check and clean air filter.3. Inspect control set interiors.

Monthly1. Check operation of standby top-housing pressurizing fan and thermostat.2. Check operation of hopper heaters.3. Check hopper level alarm operation.

Quarterly1. Check and clean rapper and vibrator switch contacts.2. Check transmissometer calibration.

Semiannual1. Clean and lubricate access-door dog bolt and hinges.2. Clean and lubricate interlock covers.3. Clean and lubricate test connections.4. Check exterior for visual signs of deterioration, and abnormal vibration, noise, leaks.5. Check T-R liquid and surge-arrestor spark gap.

Annual1. Conduct internal inspection.2. Clean top housing or insulator compartment and all electrical insulating surfaces.3. Check and correct defective alignment.4. Examine and clean all contactors and inspect tightness of all electrical connections.5. Clean and inspect all gasketed connections.6. Check and adjust operation of switchgear.7. Check and tighten rapper insulator connections.8. Observe and record areas of corrosion.

Situational1. Record air-load and gas-load readings during and after each outage.2. Clean and check interior of control sets during each outage of more than 72 hours.3. Clean all internal bushings during outages of more than 5 days.4. Inspect condition of all grounding devices during each outage over 72 hours.5. Clean all shorts and hopper buildups during each outage.6. Inspect and record amount and location of residual dust deposits on electrodes

during each outage of 72 hours or longer.7. Check all alarms, interlocks, and all other safety devices during each outage.

Source: Bibbo 1982.

Lesson 6

6-16 2.0-2/98

ondly, maintaining proper operating records establishes good baseline information to bracketnormal ranges of operation.

Evaluating ESP operating problems can be difficult and no single parameter can identify allpotential problems; a combination of factors should be considered to accurately pinpoint prob-lems. For example, although most ESP problems are reflected in the electrical readings, manydifferent problems produce the same characteristics on the meters. In addition, an initial failureor problem can cause a "domino effect" bringing about even more problems and making it dif-ficult to identify the original cause. Table 6-5 contains a typical troubleshooting cycle (Szaboand Gerstle 1977) that is useful as a general guide.

The EPA (1985) categorized the major performance problems associated with electrostaticprecipitators into the following seven areas: resistivity, dust buildup, wire breakage, hopperpluggage, misalignment of ESP components, changes in particle size distribution, and airinleakage. These problems are related to design limitations, operational changes, and/or main-tenance procedures. The following discussion about the identification of these problems andtheir effect on ESP performance is excerpted from the EPA document titled Operation andMaintenance Manual for Electrostatic Precipitators (1985).

Problems Related to Resistivity

The resistivity of the collected dust on the collection plate affects the acceptable currentdensity through the dust layer, dust removal from the plates, and indirectly, the coronacharging process. High resistivity conditions in utility fly ash applications have receivedmuch attention. The optimum resistivity range for ESP operation is relatively narrow; bothhigh and low resistivity cause problems. Excursions outside the optimum resistivity rangeare particularly a problem when a unit is designed with a modest amount of plate area, sec-tionalization, and power-input capabilities. At industrial sources where resistivity changesare intermittent, modification of operating procedures may improve performance tempo-rarily. Expensive retrofitting or modifications may be required if the dust resistivity isvastly different than the design range.

High ResistivityHigh dust resistivity is a more common problem than low dust resistivity. Particleshaving high resistivity are unable to release or transfer electrical charge. At the collec-tion plate, the particles neither give up very much of their acquired charge nor easilypass the corona current to the grounded collection plates. High dust resistivity condi-tions are indicated by low primary and secondary voltages, suppressed secondary cur-rents and high spark rates in all fields. This condition makes it difficult for the T-Rcontroller to function adequately.

Severe sparking can cause excessive charging off-time, spark "blasting" of particulateon the plate, broken wires due to electrical erosion, and reduced average current lev-els. The reduced current levels generally lead to deteriorated performance. Becausethe current level is indicative of the charging process, the low current and voltage lev-els that occur inside an ESP operating with high resistivity dust generally reflectslower charging rates and lower particle migration velocities to the plate. Particle col-lection is reduced; consequently, the ESP operates as though it were "undersized." Ifhigh resistivity is expected to continue, the operating conditions can be modified or


2.0-2/98 6-17

conditioning agents can be used to accommodate this problem and thereby improveperformance.

High resistivity also tends to promote rapping problems, as the electrical properties ofthe dust tend to make it very tenacious. High voltage drop through the dust layer andthe retention of electrical charge by the particles make the dust difficult to removebecause of its strong attraction to the plate. The greater rapping forces usually requiredto dislodge the dust may also aggravate or cause a rapping reentrainment problem.Important items to remember are (1) difficulty in removing the high-resistivity dust isrelated to the electrical characteristics, not to the sticky or cohesive nature of the dust;and (2) the ESP must be able to withstand the necessary increased rapping forceswithout sustaining damage to insulators or plate support systems.

Low ResistivityLow dust resistivity, although not as common, can be just as detrimental to the perfor-mance of an ESP as high resistivity. When particles with low resistivity reach the col-lection plate, they release much of their acquired charge and pass the corona currentquite easily to the grounded collection plate. Without the attractive and repulsive elec-trical forces that are normally present at normal dust resistivities, the binding forcesbetween the dust and the plate are considerably weakened. Therefore, particle reen-trainment is a substantial problem at low resistivity, and ESP performance appears tobe very sensitive to contributors of reentrainment, such as poor rapping or poor gasdistribution.

Since there is lower resistance to current flow for particles with low resistivity (com-pared to normal or high), lower operating voltages are required to obtain substantialcurrent flow. Operating voltages and currents are typically close to clean plate condi-tions, even when there is some dust accumulation on the plate. Low-resistivity condi-tions, are typically characterized by low operating voltages, high current flow, and lowspark rates.

Despite the large flow of current under low-resistivity conditions, the correspondinglow voltages yield lower particle migration velocities to the plate. Thus, particles of agiven size take longer to reach the plate than would be expected. When combined withsubstantial dust reentrainment, the result is poor ESP performance. In this case, thelarge flow of power to the ESP represents a waste of power.

Low-resistivity problems typically result from the chemical characteristics of the par-ticulate and not from flue gas temperature. The particulate may be enriched with com-pounds that are inherently low in resistivity, either due to poor operation of the processor to the inherent nature of the process. Examples of such enrichment include exces-sive carbon levels in fly ash (due to poor combustion), the presence of naturally occur-ring alkalis in wood ash, iron oxide in steel-making operations, or the presence ofother low-resistivity materials in the dust. Over-conditioning may also occur in someprocess operations, such as the burning of high-sulfur coals or the presence of highSO3 levels in the gas stream, which lower the inherent resistivity of the dust. In someinstances, large ESPs with SCAs greater than 750 ft2/1000 acfm have performedpoorly because of the failure to fully account for the difficulty involved in collecting alow-resistivity dust. Although some corrective actions for low resistivity are available,they are sometimes more difficult to implement than those for high resistivity.

Lesson 6

6-18 2.0-2/98

Typical High, Normal and Low Resistivity CurvesEvaluating the current and spark rate trends from the inlet to the outlet fields providesa means of evaluating the general resistivity conditions. Moderate dust resistivity con-ditions, under which ESPs work very well, are indicated by low secondary currents inthe inlet field and progressively higher values going toward the outlet. Spark ratesunder moderate resistivity are moderate in the inlet fields and decrease to essentiallyzero in the outlet field. High resistivity conditions are indicated by low secondary cur-rents in all of the fields coupled with very high spark rates. Conversely, low resistivityhas very high currents and low spark rates in all the fields.

Figure 6-3 shows the typical trend lines for moderate (normal) and high resistivitydusts. As the resistivity goes from moderate to high, the currents decrease dramati-cally in all of the fields. This is due to the suppressing effect caused by the strong elec-trostatic field created on the dust layer, and to increased electrical sparking. Thedecrease in currents is most noticeable in the outlet fields which previously had rela-tively high currents. Spark rates increase dramatically during high resistivity. Oftenmost of the fields will hit the spark rate limits programmed in by the plant operators.Once the spark rate limit is sensed by the automatic voltage controllers, it no longerattempts to drive up the voltage. This causes a reduction in the operating voltages ofthese fields. The overall impact on the opacity is substantially increased emissions. Insome cases, puffing again occurs during rapping. This is due to reduced capability ofthe precipitator fields to collect the slight quantities of particles released during rap-ping of high resistivity dust.

Figure 6-4 shows the typical trend lines for moderate (normal) and low resistivitydusts in a four-field ESP. The moderate resistivity dust shows a steady increase of cur-rent from the first field to the fourth field, while the secondary current increases rap-idly for all fields when the dust exhibits low resistivity. This effect is especiallynoticeable in the inlet fields which previously had the lowest currents. This increase incurrent is due simply to the fact that the dust layer’s electrostatic field is too weak tosignificantly impede the charging field created by the discharged electrodes. At lowresistivity, the spark rates are generally very low or zero. The voltages in all of thefields are a little lower than normal since the automatic voltage controllers sense thatthe power supply is at its current limit; therefore, the controller does not attempt todrive the voltage up any further. While the low resistivity conditions persist, there canbe frequent and severe puffs (opacity increase) which occur after each collection platerapper activates.


2.0-2/98 6-19

Figure 6-3. Typical T-R set plots - high resistivity versus moderate (normal)resistivity

Lesson 6

6-20 2.0-2/98

Figure 6-4. Typical T-R set plots - low resistivity versus moderate (normal)resistivity


2.0-2/98 6-21

Using the current, voltage, and spark rate plots is a very good way to use readily avail-able information to evaluate the impossible-to-directly monitor but neverthelessimportant resistivity conditions. It is possible to differentiate between problemscaused by mechanical faults in a single field (such as insulator leakage) and resistivityconditions which inherently affect all of the fields in varying degrees. However, thesetrend lines are not a perfect analysis tool for evaluating resistivity. A few precipitatorsnever display typical electrical trend lines since they have undersized T-R sets, under-sized fields, improperly set automatic voltage controllers, or severe mechanical prob-lems affecting most of the fields.

Dust Accumulation

There are three primary causes of dust accumulation on electrodes:

• Inadequate rapping system

• Sticky dust

• Operation at temperatures below the dew point level

The usual cause for buildup of dust on the collection plates or discharge wires is failure ofthe rapping system or an inadequate rapping system. The rapping system must providesufficient force to dislodge the dust without damaging the ESP or causing excessive reen-trainment. The failure of one or two isolated rappers does not usually degrade ESP perfor-mance significantly. The failure of an entire rapper control system or all the rappers in onefield, however, can cause a noticeable decrease in ESP performance, particularly withhigh-resistivity dust. Therefore, rapper operation should be checked at least once per day,or perhaps even once per shift. A convenient time to make this check is during routine T-Rset readings.

Rapper operation may be difficult to check on some ESPs because the time periodsbetween rapper activation can range from 1 to 8 hours on the outlet field. One method ofchecking rapper operation involves installing a maintenance-check cycle that allows acheck of all rappers in 2 to 5 minutes by following a simple rapping pattern. The cycle isactivated by plant personnel, who interrupt the normal rapping cycle and note any rappersthat fail to operate. After the check cycle, the rappers resume their normal operation.Maintenance of rapper operation is important to optimum ESP performance.

Excessive dust buildup also may result from sticky dusts or operation at gas dew pointconditions. In some cases, the dusts may be removed by increasing the temperature, but inmany cases the ESP must be entered and washed out. If sticky particulates are expected(such as tars and asphalts), a wet-wall ESP is usually used because problems can occurwhen large quantities of sticky particles enter a dry ESP.

Sticky particulates can also become a problem when the flue gas temperature falls belowthe dew point level. Although acid dew point is usually of greater concern in most applica-tions, moisture dew point is important, too. When moisture dew point conditions arereached, liquid droplets tend to form that can bind the particulate to the plate and wire.These conditions also accelerate corrosion. Carryover of water droplets or excessive mois-ture can also cause this problem (e.g., improper atomization of water in spray cooling ofthe gas or failure of a waterwall or economizer tube in a boiler). In some instances the dustlayer that has built up can be removed by increasing the intensity and frequency of the rap-ping while raising the temperature to "dry out" the dust layer. In most cases, however, it is

Lesson 6

6-22 2.0-2/98

necessary to shutdown the unit and wash out or "chisel out" the buildup to clean the plates.Localized problems can occur where inleakage causes localized decreases in gas tempera-ture.

In an operating ESP, differences in the V-I curves can be used to evaluate if a dust buildupproblem exists. Buildup of material on the discharge electrodes often means an increase involtage to maintain a given operating current. The effect of dust buildup on discharge elec-trodes is usually equivalent to increasing the effective wire diameter. Since the coronastarting voltage is strongly a function of wire diameter, the corona starting voltage tends toincrease and the whole V-I curve tends to shift to the right (see Figure 6-5). Sparking tendsto occur at about the same voltage as it does without dust buildup, unless resistivity ishigh. This effect on corona starting voltage is usually more pronounced when straightwires are uniformly coated with a heavy dust, and less pronounced on barbed wires andrigid electrodes or when the dust layer is not uniform. Barbed wires and rigid electrodestend to keep the "points" relatively clean and to maintain a small effective wire diameterand, therefore, a low corona starting voltage. Nevertheless, a higher voltage would still berequired to spread the corona discharge over the wire when dust buildup occurs. Thus,buildup on the discharge electrodes would still be characterized by a higher voltage tomaintain a given current level.

Figure 6-5. V-I curve for a field with excessive wirebuildup

Wire Breakage

Some ESPs operate for 10 to 15 years without experiencing a single wire breakage.Whereas others experience severe wire breakage problems causing one or more sections tobe out of service nearly every day of operation. Much time and effort have been expended


2.0-2/98 6-23

to determine the causes of wire breakage. One of the advantages of rigid-frame and rigid-electrode ESPs is their use of shorter wires or no wires at all. Although most new ESPshave either rigid frames or rigid electrodes, and some weighted-wire systems have beenretrofitted to rigid electrodes, the most common ESP in service today is still the weighted-wire.

Wires usually fail in one of three areas: at the top of the wire, at the bottom of the wire,and wherever misalignment or slack wires reduce the clearance between the wire andplate. Wire failure may be due to electrical erosion, mechanical erosion, corrosion, orsome combination of these. When wire failures occur, they usually short-out the fieldwhere they are located. In some cases, they may short-out an adjacent field as well. Thus,the failure of one wire can cause the loss of particle collection in an entire field or bus sec-tion. In some smaller ESP applications, this can represent one-third to one-half of thecharging/collecting area and thus substantially limit ESP performance. One of the advan-tages of higher sectionalization is that wire failure is confined to smaller areas so overallESP performance does not suffer as much. Some ESPs are designed to meet emissionstandards with some percentage of the ESP de-energized, whereas others may not haveany margin to cover downtime. Because they receive and remove the greatest percentageof particulate matter, inlet fields are usually more important to ESP operation than outletfields.

Electrical erosion is caused by excessive sparking. Sparking usually occurs at pointswhere there is close clearance within a field due to a warped plate, misaligned guidanceframes, or bowed wires. The maximum operating voltage is usually limited by these closetolerance areas because the spark-over voltage depends on the distance between the wireand the plate. The smaller the distance between the wire and plate, the lower the spark-over voltage. Under normal circumstances random sparking does little damage to the ESP.During sparking, most of the power supplied to energize the field is directed to the loca-tion of the spark, and the voltage field around the remaining wires collapses. The consid-erable quantity of energy available during the spark is usually sufficient to vaporize asmall quantity of metal. When sparking continues to occur at the same location, the wireusually "necks down" because of electrical erosion until it is unable to withstand the ten-sion and breaks. Misalignment of the discharge electrodes relative to the plates increasesthe potential for broken wires, decreases the operating voltage and current because ofsparking, and decreases the performance potential of that field in the ESP.

Although the breakage of wires at the top and bottom where the wire passes through thefield can be aggravated by misalignment, the distortion of the electrical field at the edgesof the plate tends to be the cause of breakage. This distortion of the field, which occurswhere the wire passes the end of the plate, tends to promote sparking and gradual electri-cal erosion of the wires.

Design faults and the failure to maintain alignment generally contribute to mechanicalerosion (or wear) of the wire. In some designs, the lower guide frame guides the wires ortheir weight hooks (not the weights themselves) into alignment with the plates. Whenalignment is good, the guide frame or grid allows the wires or weight hooks to float freelywithin their respective openings. When the position of the wire guide frame shifts, how-ever, the wire or weight hook rubs the wire frame within the particulate-laden gas stream.Failures of this type usually result from a combination of mechanical and electrical ero-sion. Corrosion may also contribute to this failure. Microsparking action between the

Lesson 6

6-24 2.0-2/98

guide frame and the wire or weight hook apparently causes the electrical erosion. Thesame type of failure also can occur in some rigid frame designs where the wires ride in theframe.

Another mechanical failure that sometimes occurs involves crossed wires. When replac-ing a wire, maintenance personnel must make sure that the replacement wire does notcross another wire. Eventually, the resulting wearing action breaks one or both wires. Ifone of the wires does survive, it is usually worn down enough to promote greater sparkingat the point of contact until it finally does break. Any wires that are found to be exception-ally long and slack should be replaced; they should not be crossed with another wire toachieve the desired length.

Corrosion of the wires can also lead to wire failures. Corrosion, an electrochemical reac-tion, can occur for several reasons, the most common being acid dew point. When the rateof corrosion is slow and generally spread throughout the ESP, it may not lead to a singlewire failure for 5 to 10 years. When the rate of corrosion is high because of long periods ofoperating the ESP below the acid dew point, failures are frequent. In these cases the corro-sion problem is more likely to be a localized one (e.g., in places where cooling of the gasstream occurs, such as inleakage points and the walls of the ESP). Corrosion-related wirefailures can also be aggravated by startup-shutdown procedures that allow the gas streamsto pass through the dew point many times. Facilities have mainly experienced wire break-age problems during the initial process shakedown period when the process operation maynot be continuous. Once steady operation has been achieved, wire breakage problems tendto diminish at most plants.

Wire crimping is another cause of wire failure. Crimps usually occur at the top and bot-tom of the wires where they attach to the upper wire frame or bottle weight; however, acrimp may occur at any point along the wire. Because a crimp creates a residual stresspoint, all three mechanisms (electrical erosion, mechanical erosion, and corrosion) may beat work in this situation. A crimp can:

1. Distort the electric field along the wire and promote sparking;

2. Mechanically weaken the wire and make it thinner;

3. Subject the wire to a stress corrosion failure (materials under stress tend to corrodemore rapidly than those not under stress).

Wire failure should not be a severe maintenance problem or operating limitation in a well-designed ESP. Excessive wire failures are usually a symptom of a more fundamental prob-lem. Plant personnel should maintain records of wire failure locations. Although ESP per-formance will generally not suffer with up to approximately 10% of the wires removed,these records should be maintained to help avoid a condition in which entire gas lanes maybe de-energized. Improved sectionalization helps to minimize the effect of a broken wireon ESP performance, but performance usually begins to suffer when a large percentage ofthe ESP fields are de-energized.

Hopper Pluggage

Perhaps no other problem (except fire or explosion) has the potential for degrading ESPperformance as much as hopper pluggage. Hopper pluggage can permanently damage anESP and severely affect both short-term and long-term performance. Hopper pluggage is


2.0-2/98 6-25

difficult to diagnose because its effect is not immediately apparent on the T-R set panelmeters. Depending on its location, a hopper can usually be filled in 4 to 24 hours. In manycases, the effect of pluggage does not show up on the electrical readings until the hopper isnearly full.

The electrical reaction to most plugged hoppers is the same as that for internal misalign-ment, a loose wire in the ESP, or excessive dust buildup on the plates. Typical symptomsinclude heavy or "bursty" sparking in the field(s) over the plugged hopper and reducedvoltage and current in response to the reduced clearance and higher spark rate. Inweighted-wire designs, high dust levels in the hopper may raise the weight and cause slackwires and increased arcing within the ESP. In many cases, this will trip the T-R set off-linebecause of overcurrent or undervoltage protection circuits. In some situations, the spark-ing continues even as the dust level exceeds hopper capacity and builds up between theplate and the wire; whereas in others, the voltage continues to decrease as the currentincreases and little or no sparking occurs. This drain of power away from corona genera-tion renders the field performance virtually useless. The flow of current also can cause theformation of a dust clinker (solidified dust) resulting from the heating of the dust betweenthe wire and plate.

The buildup of dust under and into the collection area can cause the plate or dischargeelectrode guide frames to shift. The buildup can also place these frames under enoughpressure to distort them or to cause permanent warping of the collection plate(s). If thishappens, performance of the affected field remains diminished by misalignment, evenafter the hopper is cleared.

Hopper pluggage can be caused by the following:

• Obstructions due to fallen wires and/or bottle weights

• Inadequately sized solids-removal equipment

• Use of hoppers for dust storage

• Inadequate insulation and hopper heating

• Air inleakage through access doors

Most dusts flow best when they are hot, therefore, cooling the dusts can promote a hopperpluggage problem.

Hopper pluggage can begin and perpetuate a cycle of failure in the ESP. For example,there was a case where a severely plugged hopper misaligned both the plates and the wireguide grid in one of the ESP fields. Because the performance of this field had decreased,the ESP was taken off-line and the hopper was cleared. But no one noticed the deterioratedcondition of the wire-guide grid. The misalignment had caused the wires and weighthooks to rub the lower guide and erode the metal. When the ESP was brought back on-line, the guide-grid metal eventually wore through. Hopper pluggage increased as weights(and sometimes wires) fell into the hopper, plugging the discharge opening and causingthe hopper to fill again and cause more misalignment. The rate of failure continued toincrease until it was almost an everyday occurrence. This problem, which has occurredmore than once in different applications, demonstrates how one relatively simple problemcan lead to more complicated and costly ones.

In most pyramid-shaped hoppers, the rate of buildup lessens as the hopper is filled due tothe geometry of the inverted pyramid. Hopper level indicators or alarms should provide

Lesson 6

6-26 2.0-2/98

some margin of safety so that plant personnel can respond before the hopper is filled.When the dust layer rises to a level where it interferes with the electrical characteristics ofthe field, less dust is collected and the collection efficiency is reduced. Also, reentrainmentof the dust from the hopper can limit how high into the field the dust can go. Althoughbuildups as deep as 4 feet have been observed, they usually are limited to 12 - 18 inchesabove the bottom of the plates.

Misalignment

As mentioned several times in the previous sections, electrode misalignment is both a con-tributor to and a result of component failures. In general, most ESPs are not affected by amisalignment of less than about 3/16 inches. Indeed, some tolerance must be provided forexpansion and contraction of the components. Beyond this limit, however, misalignmentcan become a limiting factor in ESP performance and is visually evident during an internalinspection of the ESP electrodes. Whether caused by warped plates, misaligned or skeweddischarge electrode guide frames, insulator failure, or failure to maintain ESP "box-squareness," misalignment reduces the operating voltage and current required for spark-ing. The V-I curve would indicate a somewhat lower voltage to achieve a low current levelwith the sparking voltage and current greatly reduced. Since the maximum operating volt-age/current levels depend on the path of least resistance in a field, any point of close toler-ance will control these operating levels.

Changes in Particle Size

Unusually fine particles present a problem under the following circumstances:

1. When the ESP is not designed to handle them

2. When a process change or modification shifts the particle size distribution into therange where ESP performance is poorest.

A shift in particle size distribution tends to alter electrical characteristics and increase thenumber of particles emitted in the light-scattering size ranges (opacity).

As stated in Lesson 1, there are two principal charging mechanisms: field charging anddiffusion charging. Although field charging tends to dominate in the ESP and acts on par-ticles greater than 1 micrometer in diameter, it cannot charge and capture smaller particles.Diffusion charging, on the other hand, works well for particles smaller than 0.1 microme-ter in diameter. ESP performance diminishes for particulates in the range of 0.2 - 0.9micrometer because neither charging mechanism is very effective for particles in thisrange. These particles are more difficult to charge and once charged, they are easilybumped around by the gas stream, making them difficult to collect. Depending upon thetype of source being controlled, the collection efficiency of an ESP can drop from as highas 99.9% on particles sized above 1.0 micrometer or below 0.1 micrometer, to only 85 to90% on particles in the 0.2 - 0.9 micrometer diameter range. If a significant quantity ofparticles fall into this size range, the ESP design must be altered to accommodate the fineparticles.

When heavy loadings of fine particles enter the ESP, two significant electrical effects canoccur: space charge and corona quenching. At moderate resistivities, the space-chargeeffects normally occur in the inlet or perhaps the second field of ESPs. Because it takes a


2.0-2/98 6-27

longer time to charge fine particles and to force them to migrate to the plate, a cloud ofnegatively charged particles forms in the gas stream. This cloud of charged particles iscalled a space charge. It interferes with the corona generation process and impedes theflow of negatively charged gas ions from the wire to the collection plate. The interferenceof the space charge with corona generation is called corona quenching. When this occurs,the T-R controller responds by increasing the operating voltage to maintain current flowand corona generation. The increase in voltage usually causes increased spark rates, whichmay in turn signal the controller to reduce the voltage and current in an attempt to main-tain a reasonable spark rate. Under moderate resistivity conditions, the fine dust particlesare usually collected by the time they reach the third field of the ESP which explains thedisappearance of the space charge in these later fields. The T-R controller responds to thecleaner gas in these later fields by decreasing the voltage level, but the current levels willincrease markedly. When quantities of fine particles being processed by the ESP increase,the space charging effect may progress further into the ESP.

Air Inleakage

Inleakage is often overlooked as an operating problem. In some instances, it can be benefi-cial to ESP performance, but in most cases its effect is detrimental. Inleakage may occurwithin the process itself or in the ESP and is caused by leaking access doors, leaking duct-work, and even open sample ports.

Inleakage usually cools the gas stream, and can also introduce additional moisture. Airinleakage often causes localized corrosion of the ESP shell, plates, and wires. The temper-ature differential also can cause electrical disturbances (sparking) in the field. Finally, theintroduction of ambient air can affect the gas distribution near the point of entry. The pri-mary entrance paths are through the ESP access and hopper doors. Inleakage through hop-per doors may reentrain and excessively cool the dust in the hopper, which can cause bothreentrainment in the gas stream and hopper pluggage. Inleakage through the access doorsis normally accompanied by an audible in-rush of air.

Inleakage is also accompanied by an increase in gas volume. In some processes, a certainamount of inleakage is expected. For example, application of Lungstrom regenerative airheaters on power boilers or recovery boilers is normally accompanied by an increase influe gas oxygen. For utility boilers the increase may be from 4.5% oxygen at the inlet to6.5% at the boiler outlet. For other boilers the percentage increase may be smaller whenmeasured by the O2 content, but 20 to 40% increases in gas volumes are typical and theESP must be sized accordingly. Excessive gas volume due to air inleakage, however, cancause an increase in emissions due to higher velocities through the ESP and greater reen-trainment of particulate matter. For example, at a kraft recovery boiler, an ESP that wasdesigned for a superficial velocity of just under 6 ft/s was operating at over 12 ft/s to han-dle an increased firing rate, increased excess air, and inleakage downstream of the boiler.Because the velocities were so high through the ESP, the captured material was blown offthe plate and the source was unable to meet emission standards.

Table 6-5 summarizes the problems associated with electrostatic precipitators, along withcorrective actions and preventive measures.

Lesson 6

6-28 2.0-2/98

Table 6-5. Summary of problems associated with electrostatic precipitators

Malfunction Cause

Effect onelectrostaticprecipitatorefficiency1 Corrective action

Preventivemeasures

1. Poor electrodealignment

1. Poor design2. Ash buildup on

frame hoppers3. Poor gas flow

Can drastically affectperformance andlower efficiency

Realign electrodesCorrect gas flow

Check hoppersfrequently for properoperation

2. Broken electrodes 1. Wire not rappedclean, causes anarc whichembroglios andburns through thewire

2. Clinkered wire.Causes:a. Poor flow area,

distributionthrough unit isuneven

b. Excess freecarbon due toexcess air abovecombustionrequirements orfan capacityinsufficient fordemand required

c. Wires notproperly centered

d. Ash buildup,resulting in bentframe, same as(c)

e. Clinker bridgesthe plates andwire shorts out

f. Ash buildup,pushes bottleweight upcausing sag inthe wire

g. "J" hooks haveimproperclearances to thehanging wire

h. Bottle weighthangs up duringcooling causing abuckled wire

i. Ash buildup onbottle weight tothe frame forms aclinker and burnsoff the wire

Reduction in efficiencydue to reducedpower input, bussection unavailability

Replace electrode Boiler problems;check spacebetween recordingsteam and air flowpens, pressuregauges, fouledscreen tubes

Inspect hoppers;check electrodesfrequently for wear;inspect rappersfrequently

3. Distorted or skewedelectrode plates

1. Ash buildup inhoppers

2. Gas flowirregularities

3. High temperatures

Reduced efficiency Repair or replaceplates

Correct gas flow

Check hoppersfrequently for properoperation; checkelectrode platesduring outages

4. Vibrating or swingingelectrodes

1. Uneven gas flow2. Broken electrodes

Decrease in efficiencydue to reducedpower input

Repair electrode Check electrodesfrequently for wear

Continued on next page


2.0-2/98 6-29

Table 6-5. (continued)Summary of problems associated with electrostatic precipitators

Malfunction Cause


Preventivemeasures

5. Inadequate level ofpower input (voltagetoo low)

1. High dust resistivity2. Excessive ash on

electrodes3. Unusually fine

particle size4. Inadequate power

supply5. Inadequate

sectionalization6. Improper rectifier

and controloperation

7. Misalignment ofelectrodes

Reduction in efficiency Clean electrodes;gas conditioningor alterations intemperature toreduce resistivity;increasesectionalization

Check range ofvoltages frequentlyto make sure theyare correct; check in-situ resistivitymeasurements

6. Back corona 1. Ash accumulatedon electrodescauses excessivesparking requiringreduction in voltagecharge

Reduction in efficiency Same as above Same as above

7. Broken or crackedinsulator or flower potbushing leakage

1. Ash buildup duringoperation causesleakage to ground

2. Moisture gatheredduring shutdown orlow-load operation

Reduction in efficiency Clean or replaceinsulators andbushings

Check frequently;clean and dry asneeded; check foradequatepressurization of tophousing

8. Air inleakage throughhoppers

1. From dust conveyor Lower efficiency; dustreentrained throughelectrostaticprecipitator

Seal leaks Identify early byincrease in ashconcentration atbottom of exit toelectrostaticprecipitator

9. Air inleakage throughelectrostaticprecipitator shell

1. Flange expansion Same as above; alsocauses intensesparking

Seal leaks Check for large fluegas temperaturedrop across the ESP

10.Gas bypass aroundelectrostaticprecipitator• dead passage

above plates• around high

tension frame

1. Poor design;improper isolationof active portion ofelectrostaticprecipitator

Only few percent dropin efficiency unlesssevere

Baffling to directgas into activeelectrostaticprecipitatorsection

Identify early bymeasurement of gasflow in suspectedareas

11.Corrosion 1. Temperature goesbelow dew point

Negligible untilprecipitation interiorplugs or plates areeaten away; air leaksmay develop causingsignificant drops inperformance

Maintain flue gastemperatureabove dew point

Energize precipitatorafter boiler systemhas been on line forample period to raiseflue gas temperatureabove acid dew point


Lesson 6

6-30 2.0-2/98

Table 6-5. (continued)Summary of problems associated with electrostatic precipitators

Malfunction Cause


Preventivemeasures

12.Hopper pluggage 1. Wires, plates,insulators fouledbecause of lowtemperature

2. Inadequate hopperinsulation

3. Impropermaintenance

4. Boiler leaks causingexcess moisture

5. Ash conveyingsystem malfunction(gasket leakage,blower malfunction,solenoid valves)

6. Misjudgments ofhopper vibrators

7. Material droppedinto hopper frombottle weights

8. Solenoid, timermalfunction

9. Suction blower filternot changed

Reduction in efficiency Provide proper flowof ash

Frequent checks foradequate operationof hoppers. Provideheater thermalinsulation to avoidmoisturecondensation

13. Inadequate rapping,vibrators fail

1. Ash buildup2. Poor design3. Rappers

misadjusted

Resulting buildup onelectrodes mayreduce efficiency

Adjust rappers withoptical dustmeasuringinstrument inelectrostaticprecipitator exitstream

Frequent checks foradequate operationof rappers

14.Too intense rapping 1. Poor design2. Rappers

misadjusted3. Improper rapping

force

Reentrains ash,reduces efficiency

Same as above Same as above;reduce vibrating orimpact force

15.Control failures 1. Power failure inprimary systema. Insulation

breakdown intransformer

b. Arcing intransformerbetween high-voltage switchcontacts

c. Leaks or shorts inhigh-voltagestructure

d. Insulating fieldcontamination

Reduced efficiency Find source offailure and repairor replace

Pay close attention todaily readings ofcontrol roominstrumentation tospot deviations fromnormal readings

16.Sparking 1. Inspection door ajar2. Boiler leaks3. Plugging of hoppers4. Dirty insulators

Reduced efficiency Close inspectiondoors; repair leaksin boiler; unplughoppers; cleaninsulators

Regular preventivemaintenance willalleviate theseproblems

1The effects of precipitation problems can be discussed only on a qualitative basis. There are no known emission tests of precipitators to determineperformance degradation as a function of operational problems.

Sources: Szabo and Gerstle 1977, and Englebrecht 1980.


2.0-2/98 6-31

Safety

Persons who will be operating and maintaining an ESP must be well trained on all safetyaspects to avoid injury. One person at the plant should be assigned the responsibility of con-stantly checking safety standards and equipment and to train or procure safety training for allthose who will work with the ESP. A suggested list of important safety precautions is listed inTable 6-6 (Bibbo 1982).

Table 6-6. Important safety precautions

Wiring and controls1. Prior to startup, double-check that field wiring between controls and devices

(T-R sets, rapper prime motors, etc.) is correct, complete, and properly labeled.2. Never touch exposed internal parts of control system. Operation of the

transformer-rectifier controls involves the use of dangerous high voltage.Although all practical safety control measures have been incorporated into thisequipment, always take responsible precautions when operating it.

3. Never use fingers or metal screwdrivers to adjust uninsulated control devices.

Access1. Use a positive method to ensure that personnel are out of the precipitator, flues,

or controls prior to energization. Never violate established plant clearancepractices.

2. Never bypass the safety key interlock system. Destroy any extra keys. Alwayskeep lock caps in place. Use powdered graphite only to lubricate lock systemparts; never use oil or grease. Never tamper with a key interlock.

3. Use grounding chains whenever entering the precipitator, T-R switch enclosure,or bus ducts. The precipitator can hold a high static charge, up to 15 kV, after it isde-energized. The only safe ground is one that can be seen.

4. Never open a hopper door unless the dust level is positively below the door. Donot trust the level alarm. Check from the upper access in the precipitator. Hot dustcan flow like water and severely burn or kill a person standing below the door.Wear protective clothing.

5. Be on firm footing prior to entering the precipitator. Clear all trip hazards. Use theback of the hand to test for high metal temperatures.

6. Avoid ozone inhalation. Ozone is created any time the discharge electrodes areenergized. Wear an air-line mask when entering the precipitator, flues, or stackwhen ozone may be present. Do not use filters, cartridge, or canister respirators.

7. Never poke hoppers with an uninsulated metal bar. Keep safety and danger signsin place. Clean, bright signs are obeyed more than deteriorated signs.

Fire/explosion1. In case of boiler malfunction that could permit volatile gases and/or heavy carbon

carryover to enter the precipitator, immediately shut down all transformer-rectifiersets. Volatile gases and carbon carryover could be ignited by sparks in theprecipitator, causing fire or explosion, damaging precipitator internals.

2. If high levels of carbon are known to exist on the collecting surface or in thehoppers, do not open precipitator access doors until the precipitator has cooledbelow 52°C (125°F). Spontaneous combustion of the hot dust may be caused bythe inrush of air.

3. If a fire is suspected in the hoppers, empty the affected hopper. If unable to emptythe hopper immediately, shut down the transformer-rectifier sets above thehopper until it is empty. Use no other method to empty the hopper. Never usewater or steam to control this type of fire. These agents can release hydrogen,increasing the possibility of explosion.

Source: Bibbo 1982.

Lesson 6

6-32 2.0-2/98

Summary

Successful longtime operation of an ESP ultimately depends on effective inspection, startupand shutdown and operation and maintenance procedures. Regardless of how well the ESP isdesigned, if these procedures are not developed and routinely followed the ESP will deterio-rate resulting in a decrease of its particulate emission removal efficiency.

The lesson discusses the importance of monitoring key operating parameters including voltageand current readings of each T-R set, opacity, flue gas flow rate and flue gas composition andmoisture levels. We also covered how evaluating current, voltage and spark rate trends canhelp provide information on dust resistivity conditions. A change in dust resistivity can drasti-cally alter the performance of the ESP and will likely lead to emission compliance problems ifnot rectified.

Suggested Reading

Bibbo, P. P. 1982. Electrostatic precipitators. In L. Theodore and A. Buonicore (Eds.), Air PollutionControl Equipment-Selection, Design, Operation and Maintenance (pp.3-44). Englewood Cliffs,NJ: Prentice Hall.

Englebrecht, H. L. 1980. Mechanical and electrical aspects of electrostatic precipitator O&M. In R. A.Young and F. L. Cross (Eds.), Operation and Maintenance for Air Particulate Control Equipment(pp. 283-354). Ann Arbor, MI: Ann Arbor Science.



2.0-2/98 6-33

Review Exercise

1. Air inleakage at flanges or collector access points in high-temperature systems (hot-side ESPs)may:

a. Allow dust to settle out quickly into hoppersb. Cause acids and moisture to condense on internal components of the ESPc. Increase the overall collection efficiency of the unit

2. Gas streams of high temperature should be maintained above the:

a. Ignition temperatureb. Gas dew pointc. Concentration limit

3. Since most ESPs are installed in the field, it is important to check that all surfaces and areas ofpotential heat loss are adequately covered with:

a. Paintb. Plastic coatingc. Insulationd. Aluminum siding

4. Before the ESP is started, the installation crew should prepare and use a____________________.

5. Which of the following ESP components should be checked before starting the collector?

a. Hoppers and discharge devicesb. Rappersc. Discharge and collection electrodesd. All of the above

6. Two very important parameters monitored by meters on T-R sets and used to evaluate ESP perfor-mance are ____________________ and____________________.

7. True or False? Individual T-R set values for voltage and current are important; however, the trendsfor voltage and current noted within an entire ESP are more valuable in assessing performance.

8. As particulate matter is removed from the gas stream, the ____________________ shouldincrease from the inlet to the outlet fields.

a. Opacityb. Current densityc. Rapper intensityd. Amperage

Lesson 6

6-34 2.0-2/98

9. An opacity monitor (transmissometer) measures:

a. Particle weightb. Particle sizec. Light differentiald. Primary current

10. True or False? Opacity monitors are useful tools to aid in optimization of spark rate, power levelsand rapping cycles in ESPs.

11. True or False? Changes in flue gas temperature generally have little or no effect on particle resis-tivity.

12. Operating parameters such as specific collection area, superficial velocity, and treatment time aredependent on the ____________________ ____________________ ____________________.

13. True or False? Because of their open design, gas flow distribution through ESPs are generally veryevenly distributed.

14. ____________________ tests are generally conducted on cool, inoperative ESPs through whichno gas is flowing.

a. Air Load V-I Curveb. Gas Load V-I Curvec. Complianced. All of the above

15. True or False? When ESPs are equipped with identical fields, the air-load curves for each fieldshould be very similar.

16. Air Load V-I curves for a given ESP field will generally shift to the ____________________ ifplates are dirty compared to previous tests.

a. Leftb. Rightc. a and b, above

17. Gas-load curves are similar to air-load curves except the gas-load curves are shifted to the____________________ compared to the air-load curves.

a. Leftb. Right

18. True or False? Gas-load curves generally are identical for a given ESP field on a day-to-day basis.

19. True or False? High dust resistivity is characterized by the tendency toward high spark rates at lowcurrent levels.


2.0-2/98 6-35

20. Excessive dust buildup on the collecting plates or discharge wires can be caused by failure of the:

a. Primary and secondary voltageb. Rapping systemc. Back coronad. All the above

21. Wire failure can be caused by:

a. Electrical erosionb. Mechanical erosionc. Corrosiond. All of the above

22. True or False? Unlike baghouses, ESPs are not affected by operating temperatures falling belowthe acid or moisture dew point.

23. True or False? In general, a well-designed ESP can operate effectively with a small percentage(less than 10) of its wires out-of-service.

24. True or False? Dust discharge hopper pluggage is not a major concern for ESPs.

Lesson 6

6-36 2.0-2/98


2.0-2/98 6-37


1. b. Cause acids and moisture to condense on internal components of the ESPAir inleakage at flanges or collector access points in high-temperature systems (hot-side ESPs)may cause acids and moisture to condense on internal components of the ESP.

2. b. Gas dew pointGas streams of high temperature should be maintained above the gas dew point. When the temper-ature falls below the gas dew point, moisture or acid can condense on ESP components and possi-bly cause corrosion.

3. c. InsulationSince most ESPs are installed in the field, it is important to check that all surfaces and areas ofpotential heat loss are adequately covered with insulation.

4. ChecklistBefore the ESP is started, the installation crew should prepare and use a checklist.

5. d. All of the aboveThe following are some ESP components that should be checked before starting the collector:

• Hoppers and discharge devices

• Rappers

• Discharge and collection electrodes

6. VoltageCurrentTwo very important parameters monitored by meters on T-R sets and used to evaluate ESP perfor-mance are voltage and current.

7. TrueIndividual T-R set values for voltage and current are important; however, the trends for voltageand current noted within an entire ESP are more valuable in assessing performance. T-R set read-ings for current, voltage, and sparking should follow certain patterns from the inlet to the outletfields.

8. b. Current densityAs particulate matter is removed from the gas stream, the current density should increase from theinlet to the outlet fields. The dust concentration in the inlet sections will suppress the current.Increased current density is needed in the outlet sections where there is a greater percentage ofvery small particles.

9. c. Light differentialAn opacity monitor (transmissometer) measures light differential. An opacity monitor comparesthe amount of light generated and transmitted by the instrument on one side of the gas stream withthe quantity measured on the other side of the gas stream.

Lesson 6

6-38 2.0-2/98

10. TrueOpacity monitors are useful tools to aid in optimization of spark rate, power levels and rappingcycles in ESPs.

11. FalseChanges in flue gas temperature have an important effect on particle resistivity. In fact, while gastemperature variations may have some effect on corona discharge characteristics and physicalcharacteristics of the ESP (corrosion, expansion/contraction), their most important effect is on par-ticle resistivity. See Figure 3-1.

12. Gas flow rateOperating parameters such as specific collection area, superficial velocity, and treatment time aredependent on the gas flow rate.

13. FalseActually, gas flow through the ESP is not evenly distributed. ESP manufacturers settle for whatthey consider to be an acceptable variation.

14. a. Air Load V-I CurveAir-Load V-I Curve tests are generally conducted on cool, inoperative ESPs through which no gasis flowing.

15. TrueWhen ESPs are equipped with identical fields, the air-load curves for each field should be verysimilar.

16. b. RightAir Load V-I curves for a given ESP field will generally shift to the right if plates are dirty com-pared to previous tests. Dirty plates suppress the current. It takes a higher voltage to generate thesame amount of current as with a “clean plate” condition.

17. a. LeftGas-load curves are similar to air-load curves except the gas-load curves are shifted to the leftcompared to the air-load curves. Gas-load curves are generated while the unit is on-line. Thecurves are generally shifted to the left because sparking occurs at lower voltage and current whenparticles are present.

18. FalseGas-load curves for a given ESP field generally vary on a day-to-day basis. Curve positions canchange due to fluctuations in the amount of dust on the plates, gas flow, particulate loadings, tem-perature, and resistivity.

19. TrueHigh dust resistivity is characterized by the tendency toward high spark rates at low current levels.

20. b. Rapping systemExcessive dust buildup on the collecting plates or discharge wires can be caused by failure of therapping system.


2.0-2/98 6-39

21. d. All of the aboveWire failure can be caused by the following:

• Electrical erosion

• Mechanical erosion

• Corrosion.

22. FalseLike baghouses, ESPs are affected by operating temperatures falling below the acid or moisturedew point. At temperatures below the acid or moisture dew point, acid or moisture can condenseon ESP components and cause corrosion.

23. TrueIn general, a well-designed ESP can operate effectively with a small percentage (less than 10) ofits wires out-of-service.

24. FalseDust discharge hopper pluggage is a major concern for ESPs. Hopper pluggage can permanentlydamage an ESP.

Lesson 6

6-40 2.0-2/98

2.0-2/98 6-41

Bibliography

Bibbo, P. P. 1982. Electrostatic precipitators. In L. Theodore and A. Buonicore (Eds.), Air PollutionControl Equipment-Selection, Design, Operation and Maintenance (pp.3-44). Englewood Cliffs,NJ: Prentice Hall.

Cross, F. L., and H. E. Hesketh. (Eds.) 1975. Handbook for the Operation and Maintenance of Air Pol-lution Control Equipment. Westport, CT: Technomic Publishing.

Englebrecht, H. L. 1980. Mechanical and electrical aspects of electrostatic precipitator O&M. In R. A.Young and F. L. Cross (Eds.), Operation and Maintenance for Air Particulate Control Equipment(pp. 283-354). Ann Arbor, MI: Ann Arbor Science.



Szabo, M. F., and R. W. Gerstle. 1977. Electrostatic Precipitator Malfunctions in the Electric UtilityIndustry. EPA 600/2-77-006.

Szabo, M. F., Y. M. Shah, and S. P. Schliesser. 1981. Inspection Manual for Evaluation of ElectrostaticPrecipitator Performances. EPA 340/1-79-007.


U.S. Environmental Protection Agency. 1987, August. Recommended Recordkeeping Systems for AirPollution Control Equipment. Part I, Particulate Matter Controls. EPA 340/1-86-021.

U.S. Environmental Protection Agency. 1993. Monitoring, Recordkeeping, and Reporting Require-ments for the Acid Rain Program. In Code of Federal Regulations - Protection of the Environment.40 CFR 75. Washington, D.C.; U.S. Government Printing Office.

Lesson 6

6-42 2.0-2/98

2.0-3/95 2-1

Lesson 2Bag Cleaning

Goal

To familiarize you with the mechanisms to clean collected dust from the bags.

Objectives


1. Name two bag cleaning sequences and briefly discuss the conditions under which they areused

2. List three major cleaning methods and briefly describe how each method is used to removedust from bags

3. Describe how bags are attached and supported in the three different bag-cleaning designs

4. Identify the major parameters associated with each of the three major bag cleaning methods

Cleaning Sequences

Two basic sequences are used for bag cleaning: intermittent (or periodic) cleaning and contin-uous cleaning.

Intermittently cleaned baghouses consist of a number of compartments or sections. Onecompartment at a time is removed from service and cleaned on a regular rotational basis. Thedirty gas stream is diverted from the compartment being cleaned to the other compartments inthe baghouse, so it is not necessary to shut down the process. Occasionally, the baghouse isvery small and consists of a single compartment. The flow of dirty air into these baghouses isstopped during bag cleaning. These small, single-compartment baghouses are used on batchprocesses that can be shut down for bag cleaning.

Continuously cleaned baghouses are fully automatic and can constantly remain on-line forfiltering. The filtering process is momentarily interrupted by a blast of compressed air thatcleans the bag, called pulse-jet cleaning. In continuous cleaning, a row of bags is always beingcleaned somewhere in the baghouse. The advantage of continuous cleaning is that it is not nec-essary to take the baghouse or a compartment out of service for bag cleaning. Small continu-ously cleaned baghouses only have one compartment and are cleaned by pulse-jet cleaningdescribed in detail later in this lesson. Large continuous cleaning baghouses are built withcompartments to help prevent total baghouse shutdown for bag maintenance and failures to thecompressed air cleaning system or hopper conveyers. This allows the operator to take onecompartment off-line to perform necessary maintenance.

Lesson 2

2-2 2.0-3/95

Types of Bag Cleaning

A number of cleaning mechanisms are used to remove caked particles from bags. The fourmost common are shaking, reverse air, pulse jet, and sonic. Another mechanism called blowring or reverse jet is normally not used in modern bag cleaning systems and is not discussed inthis course. Note that several manufacturers use the term reverse jet to mean pulse jet.

Shaking

Shaking can be done manually but is usually performed mechanically in industrial-scalebaghouses. Small baghouses handling exhaust streams less than 500 cfm (14.2 m3/min)are frequently cleaned by hand levers. However, thorough cleaning is rarely achievedsince a great amount of effort must be used for several minutes to remove dust cakes fromthe bags. In addition, these small units do not usually have a manometer installed on themto give pressure drop readings across the baghouse. These readings are used to determinewhen bag cleaning is necessary. Therefore, manual shaker baghouses are not recom-mended for use in controlling particulate emissions from industrial sources.

Mechanical shaking is accomplished by using a motor that drives a shaft to move a rodconnected to the bags. It is a low energy process that gently shakes the bags to removedeposited particles. The shaking motion and speed depends on the vendor’s design and thecomposition of dust deposited on the bag (see Figure 2-1). The shaking motion is gener-ally in the horizontal direction.

Figure 2-1. Shaking

The tops of the bags in shaker baghouses are sealed or closed and supported by a hook orclasp (see Figure 2-2). Bags are open at the bottom and attached to a cell plate. The bagsare shaken at the top by moving the frame where the bags are attached. This causes thebags to ripple and release the dust. The flow of dirty gas is stopped during the cleaningprocess. Therefore the baghouse must be compartmentalized to be usable on a continuousbasis. Shaker baghouses always use interior filtration (dust collected on the inside of thebags).

Bag Cleaning

2.0-3/95 2-3

Figure 2-2. Bag attachment for shaker cleaning baghouses

Figure 2-3. Typical shaker baghouse

In a typical shaker baghouse, bags are attached to a shaft that is driven by an externallymounted motor (Figure 2-3). The bags are shaken, and the dust falls into a hopper locatedbelow the bags. The duration of the cleaning cycle can last from 30 seconds to as long as afew minutes, but generally lasts around 30 seconds.

Frequency of bag cleaning depends on the type of dust, the concentration, and the pressuredrop across the baghouse. The baghouse usually has two or more compartments to allowone compartment to be shut down for cleaning.

Lesson 2

2-4 2.0-3/95

Figure 2-4 shows a typical shaking mechanism of a shaker baghouse. The bags areattached in sets of two rows to mounting frames across the width of the baghouse. A motordrives the shaking lever, which in turn causes the frame to move and the bags to shake.

Figure 2-4. Detail of a shaking lever system

Shaking should not be used when collecting sticky dusts. The force needed to removesticky dust can tear or rip the bag.

Bag wear can occur at the top of the bag where the support loop attaches; it can also be aproblem at the bottom of the bag where it is attached to the cell plate. Proper frequency ofbag cleaning is therefore important to prevent premature bag failure.

Typical design parameters for shaking cleaning are given in Table 2-1. Occasionally shak-ing cleaning is used along with reverse-air cleaning to promote thorough bag cleaning forapplications such as coal-fired utility boilers.

Reverse Air

Reverse-air cleaning baghouses are compartmentalized to permit a section to be off-linefor cleaning. In a reverse-air baghouse, the flow of dirty gas into the compartment isstopped and the compartment is backwashed with a low pressure flow of air. Dust is

Table 2-1. Shaker cleaning parameters

Frequency Usually several cycles per second; adjustable

Motion Simple harmonic or sinusoidal

Peak acceleration 1 to 10 g

Bag movement (amplitude) Fraction of an inch to a few inches

Operation mode Compartment off-stream for cleaning

Duration 10 to 100 cycles; 30 seconds to a few minutes

Common bag dimensions 5, 8, or 12-inch diameters; 8, 10, 22, or 30-foot lengthsSources: McKenna and Greiner 1982.

McKenna and Turner 1989.Adapted and reproduced by permission of ETS, Inc.

Bag Cleaning

2.0-3/95 2-5

removed by merely allowing the bags to collapse, thus causing the dust cake to break andfall into the hopper. Cleaning air is supplied by a separate fan which is normally muchsmaller than the main system fan, since only one compartment is cleaned at a time (seeFigure 2-5). The cleaning action is very gentle, allowing the use of less abrasion resistantfabrics such as fiberglass.

Figure 2-5. Typical reverse-air baghouse

During the filtering mode, the compartment’s outlet gas damper and inlet gas damper areboth open. When bag cleaning begins, the outlet damper is closed to block the flow of gas.The bags are allowed to relax for a short time and the reverse air damper located at the topof the compartment is opened to bring reverse air for bag cleaning into the compartment.The reverse air flow usually lasts from about 30 seconds to as long as several minutes.During this time, dust falls into the hopper. Reverse-air baghouses also have by-passdampers that allow the dirty gas to by-pass the compartments during malfunctions andstart up periods.

Lesson 2

2-6 2.0-3/95

In reverse-air baghouses, dust is collected on the inside of the bag. The bag is open at thebottom and sealed by a metal cap at the top (see Figure 2-6). Bags are connected to a ten-sion spring that is attached to the frame located above to hold them in place. The tensionspring allows the bags to move slightly during the cleaning process. The tension springcan be adjusted to make sure the bags do not sag too much, thus preventing the bags fromcreasing and eventually wearing out. The bottom of the bag fits over a thimble and the bagis attached snugly to the thimble by a clasp or clamp (see Figure 2-7).

Figure 2-6. Bag attachment for reverse-air baghouses

The bag contains rings to keep it from completely collapsing during the cleaning cycle.Complete collapse of the bag would prevent the dust from falling into the hopper. Bags aresupported by small steel rings sewn to the inside of the bag (see Figure 2-7). Rings areusually made of 3/16 inch carbon steel. Depending on flue gas conditions, they can also becomposed of cadmium-plated galvanized, or stainless steel. The rings are placed every 2to 4 feet apart throughout the bag length depending on the length and diameter of the bag.Usually, the spacing between anti-collapse rings is larger at the top of the bag and issmaller near the bottom of the bag. Reverse-air baghouses use very large bags (as com-pared to shaker or pulse-jet baghouses) ranging from 8 to 18 inches in diameter and from20 to 40 feet in length.

Bag Cleaning

2.0-3/95 2-7

Figure 2-7. Bag construction for a reverse-air baghouses

Reverse-air cleaning is generally used for cleaning woven fabrics. Cleaning frequencyvaries from 30 minutes to several hours, depending on the inlet dust concentration and thepressure drop of the baghouse. The cleaning duration is approximately 10 to 30 seconds;the total time is 1 to 2 minutes including time for valve opening and closing, and dust set-tling. Typical design parameters for reverse-air cleaning are given in Table 2-2.

Pulse Jet

The most commonly used cleaning method is the pulse-jet or pressure-jet cleaning. Bag-houses using pulse-jet cleaning make up approximately 40 to 50% of the new baghouseinstallations in the U.S. today. The pulse-jet cleaning mechanism uses a high pressure jetof air (compressed air-induced pulse) to remove the dust from the bag. Bags in the bag-

Table 2-2. Reverse-air cleaning parameters

Frequency Cleaned one compartment at a time, sequencing onecompartment after another; can be continuous or initiated by amaximum-pressure-drop switch

Motion Gentle collapse of bag (concave inward) upon deflation; slowlyrepressurize a compartment after completion of a back-flush

Operation mode Compartment taken off-stream for cleaning

Duration 1 to 2 minutes, including valve opening, closing and dust settlingperiod; reverse-air flow normally 10 to 30 seconds

Common bag dimensions 8, 12, and 18 inch-diameters; 22, 30, 40 foot-lengths

Bag tension 50 to 75 lbs typical - optimum value varies; bag tension adjustedafter unit is on-stream

Sources: McKenna and Greiner 1982.McKenna and Furlong 1992.Adapted and reproduced by permission of ETS, Inc.

Lesson 2

2-8 2.0-3/95

house compartment are supported internally by rings or metal cages. Bags are held firmlyin place at the top by clasps and usually have an enclosed bottom (the bag is sewn closedat the bottom). In another design, a snap ring is sewn into the top of the bag which fits intothe tube sheet opening. The cage slides inside the bag and the top of the cage sits on thetube sheet (see Figure 2-8). Dust-laden gas is filtered through the bag, depositing dust onthe outside surface of the bag. Pulse-jet cleaning is used for cleaning bags in an exteriorfiltration system (See Figure 2-9).

Figure 2-8. Snap-ring bag design for pulse-jet systems

Figure 2-9. Typical pulse-jet baghouse with pulsing air supply

Bag Cleaning

2.0-3/95 2-9

The dust is removed from the bag by a blast of compressed air injected into the top of thebag tube. The blast of high pressure air stops the normal flow of air through the bag filter.However, during pulse-jet cleaning, the flow of dirty air into the baghouse compartment isnot stopped. The air blast develops into a standing or shock wave that causes the bag toflex or expand as the shock wave travels down the bag tube. As the bag flexes, the cakefractures, and deposited particles are discharged from the bag (Figure 2-10). The shockwave travels down and back up the tube in approximately 0.5 seconds.

Pulse-jet units are usually operated in a “non-dust cake” mode. Bags are pulsed frequentlyto prevent the formation of a thick cake and to keep the unit from having a high pressuredrop across the dust cake and felted filter. However, sometimes a dust cake is desired incases where woven bags are used in a pulse-jet baghouse.

Figure 2-10. Pulse-jet cleaning

Lesson 2

2-10 2.0-3/95

The blast of compressed air must be strong enough for the shock wave to travel the lengthof the bag and shatter or crack the dust cake. Pulse-jet units use air supplies from a com-mon header which feeds pulsing air through a separate blow pipe located above each rowof bags in a compartment. Pulsing air is directed into the bags through nozzles or orificeslocated on the blow pipe (Figure 2-11). A diaphragm valve on each blow pipe provides thevery brief pulse of compressed air. The opening and closing of the diaphragm is controlledby an electrically operated solenoid valve.

Figure 2-11. Pulse-jet cleaning system

In some baghouse designs, a venturi sealed at the top of each bag (see Figure 2-12) or justinside the top of each bag is used to create a large enough pulse to travel down and up thebag. Vendors using venturis in pulse-jet units claim that the venturis can help increase thecleaning pressure, and thereby improve bag cleaning. In other pulse-jet designs, venturisare not used, but the bags are still cleaned effectively. The importance of the venturis isdebatable. The use of venturis has in some cases directed an increased air flow to a spe-cific spot on the bag, and actually caused the bag to wear a hole very quickly. The criticalfactor to providing thorough bag cleaning is to make sure that the blow pipe and nozzle areproperly aligned above the bag tubes.

Bag Cleaning

2.0-3/95 2-11

Figure 2-12. Venturis used with pulse-jet cleaning

The bag cleaning by the pulse occurs in approximately 0.3 to 0.5 seconds. The pressuresinvolved are commonly between 60 and 100 psig (414 kPa and 689 kPa). Some vendorshave developed systems to use a lower pressure pulsing air (40 psi).

Most pulse-jet baghouses use bag tubes that are 4 to 6 in. (10.2 to 15.2 cm) in diameter.The length of the bag is usually around 10 to 12 ft (3.05 to 3.66 m), but can be as long as20 ft (6.1 m). The shaker and reverse-air baghouses use larger bags than the pulse-jetunits. The bags in shaker and reverse-air units are 6 to 18 in. (15.2 to 45.7 cm) in diameterand up to 40 ft (12.2 m) in length. Typical design parameters for pulse-jet cleaning aregiven in Table 2-3.

Compartmentalized Pulse-Jet Baghouses

Pulse-jet baghouses can also be compartmentalized. In this case poppet valves located inthe clean air plenum are used to stop the flow of dirty air into the compartment. Each com-partment can be equipped either with a single pulse valve that supplies compressed air tothe group of bags, or have separate pulsing valves that direct pulsing air into the blow

Table 2-3. Pulse-jet cleaning parameters

Frequency Usually a row of bags at a time, sequenced one row after another;can be sequenced such that no adjacent rows are cleaned oneafter another; initiation of bag cleaning can be triggered bymaximum pressure-drop set-point, be timed, or continuous

Motion Shock wave passes down bag, bag distends from bag cagemomentarily

Operation mode Cleaning can be done while unit is on-stream; cleaning can also bedone off-stream (off-line) for difficult to clean applications such ascoal-fired boilers or MSW incinerators

Duration Compressed air 60 to 100 psi for on-line cleaning and 40 to 100 psifor off-line cleaning. Pulse duration is 0.1 sec.

Common bag dimensions 5 to 6 inch diameters; 8, 10, 12, 14, 16, and 20- foot lengthsSources: McKenna and Greiner 1982.

Beachler and Greiner 1989.Adapted and reproduced by permission of ETS, Inc.

Bagcage

Venturi fitsinside cage

Venturi sits ontop of tube sheet

Tube sheet

Bagcage

Lesson 2

2-12 2.0-3/95

pipes above the bag rows in the compartment. During the cleaning cycle the poppet valvecloses, stopping the air flow through the compartment. The pulse valve opens for about0.1 second, supplying a burst of air into the bags for cleaning. The compartment remainsoff-line for approximately 30 seconds, although this time period can be longer or shorter ifdesired. The poppet valve then automatically reopens, bringing the compartment back onstream. Alternate compartments are cleaned successively until all the bags in the baghousehave been cleaned (Figure 2-13). The cleaning cycle in each compartment lasts about 40to 120 seconds. This cleaning is called off-line cleaning. It is frequently used on fabric fil-ters installed on coal-fired boilers and municipal waste incinerators, allowing very thor-ough bag cleaning while the baghouse continuously achieves very low emission levels(less than 0.015 gr/dscf).

Figure 2-13. Compartmentalized pulse-jet baghouse (plenum-pulsebaghouse)

Bag Cleaning

2.0-3/95 2-13

Sonic

In a few systems, shaking is accomplished by sonic vibration (Figure 2-14). A sound gen-erator is used to produce a low frequency sound that causes the bags to vibrate. The noiselevel produced by the generator is barely discernible outside the baghouse. Sonic cleaningis generally used along with one of the other cleaning techniques to help thoroughly cleandirty bags.

Figure 2-14. Sonic vibrations, usually usedalong with another bagcleaning mechanism

Lesson 2

2-14 2.0-3/95

Review Exercise

1. Two basic sequences for bag cleaning are ____________________ and ____________________cleaning.

2. True or False? Intermittent baghouses consist of compartments that are all cleaned simultaneously.

3. True or False? It is not necessary to take a continuously cleaned baghouse off-line for bagcleaning.

4. Mechanical shaking is accomplished by using a(an) ____________________ that drives a shaft toshake the dust-laden bags.

5. True or False? Bags are not sealed or closed at the top in a shaker baghouse.

6. True or False? The flow of dirty air into a compartment is shut down for bag cleaning in a shakerbaghouse.

7. The shaking motion causes the dust cake to break and fall into the ____________________.

8. Bag cleaning frequency for shaker baghouse depends on dust type, dust concentration, and the____________________ ____________________ across the baghouse.

9. True or False? Reverse-air cleaning is accomplished by a blast of air into each bag.

10. Reverse-air cleaning is very gentle allowing the use of less abrasion-resistant fabrics such aswoven ____________________ (or ____________________).

11. In reverse-air cleaning units, dust is collected on the ____________________ of the bags.

12. Cleaning air in reverse-air baghouses is usually supplied by a ________________________________________.

13. True or False? During reverse-air cleaning the flow of dirty air into the compartment is stopped.

14. The bags are attached at the top in a reverse-air cleaning baghouse by a spring and a metal____________________.

15. In a reverse-air baghouse, rings are usually sewn into the inside of the bag every:

a. 36 to 60 in.b. 1 to 2 in.c. 4 to 18 in.d. 2 to 4 ft

Bag Cleaning

2.0-3/95 2-15

16. Reverse-air baghouses use large bags whose lengths range from:

a. 3 to 5 ftb. 20 to 40 ftc. 5 to 10 ftd. 75 to 100 ft

17. True or False? In reverse-air cleaning baghouses, the bags are attached at the bottom to the cellplate by a rubber gasket.

18. Reverse-air cleaning duration is approximately:

a. 1 to 2 hoursb. 10 to 20 minutesc. 10 to 30 secondsd. Less than 1 second

19. Pulse-jet cleaning is accomplished by:

a. Shaking each bag in the compartment while the damper is closedb. Injecting a blast of compressed air into each bagc. Reversing the flow of air into the baghouse compartment and gently shaking the bags

20. In a pulse-jet baghouse, dust is removed from the ____________________ of the bag when thebag is cleaned.

21. In a pulse-jet baghouse, the dust collects on the outside of the bag, therefore the bag must besupported, usually by a ____________________ ____________________.

22. True or False? In pulse-jet cleaning, the flow of dirty air into the compartment must be stoppedbefore cleaning is initiated.

23. True or False? Pulse-jet air is supplied from a common header which feeds into a nozzle locatedabove each bag.

24. In pulse-jet cleaning, the shock wave travels down and then back up the bag tube inapproximately:

a. 1 to 2 minutesb. 10 to 30 secondsc. 0.5 seconds

25. Pulse-jet baghouses use bags that are usually:

a. 12 to 16 in. in diameter and 20 to 40 ft longb. 4 to 6 in. in diameter and 10 to 12 ft longc. 16 to 24 in. in diameter and 15 to 25 ft long

Lesson 2

2-16 2.0-3/95

Review Answers

1. IntermittentContinuousTwo basic sequences for bag cleaning are intermittent and continuous cleaning.

2. FalseIntermittent baghouses consist of compartments that are NOT all cleaned simultaneously. Onecompartment at a time is removed from service and cleaned on a rotational basis.

3. TrueIt is not necessary to take a continuously cleaned baghouse off-line for bag cleaning.

4. MotorMechanical shaking is accomplished by using a motor that drives a shaft to shake the dust-ladenbags.

5. FalseBags are sealed or closed at the top in a shaker baghouse.

6. TrueThe flow of dirty air into a compartment is shut down for bag cleaning in a shaker baghouse.

7. HopperThe shaking motion causes the dust cake to break and fall into the hopper.

8. Pressure dropBag cleaning frequency for a shaker baghouse depends on dust type, dust concentration, and thepressure drop across the baghouse.

9. FalseIn reverse-air cleaning, the flow of dirty gas into the compartment is stopped and the compartmentis backwashed with a low pressure flow of air.

10. Glass (or fiberglass)Reverse-air cleaning is very gentle allowing the use of less abrasion-resistant fabrics such aswoven glass (or fiberglass).

11. InsideIn reverse-air cleaning units, dust is collected on the inside of the bags.

12. Separate fanCleaning air in reverse-air baghouses is usually supplied by a separate fan.

13. TrueDuring reverse-air cleaning, the flow of dirty air into the compartment is stopped.

Bag Cleaning

2.0-3/95 2-17

14. CapThe bags are attached at the top in a reverse-air cleaning baghouse by a spring and a metal cap.

15. d. 2 to 4 ftIn a reverse-air baghouse, rings are usually sewn into the inside of the bag every 2 to 4 ft.

16. b. 20 to 40 ftReverse-air baghouses use large bags whose lengths range from 20 to 40 ft.

17. FalseIn reverse-air cleaning baghouses, the bags are attached at the bottom to the cell plate by a clamp.

18. c. 10 to 30 secondsReverse-air cleaning duration is approximately 10 to 30 seconds.

19. b. Injecting a blast of compressed air into each bagPulse-jet cleaning is accomplished by injecting a blast of compressed air into each bag.

20. OutsideIn a pulse-jet baghouse, dust is removed from the outside of the bag when the bag is cleaned.

21. Metal cageIn a pulse-jet baghouse, the dust collects on the outside of the bag, therefore the bag must besupported, usually by a metal cage.

22. FalseIn pulse-jet cleaning, the flow of dirty air into the compartment is NOT stopped before cleaning isinitiated.

23. TruePulse-jet air is supplied from a common header which feeds into a nozzle located above each bag.

24. c. 0.5 secondsIn pulse-jet cleaning, the shock wave travels down and then back up the bag tube in approximately0.5 seconds.

25. b. 4 to 6 in. in diameter and 10 to 12 ft longPulse-jet baghouses use bags that are usually 4 to 6 in. in diameter and 10 to 12 ft long.

2-18 2.0-3/95

Bibliography

Beachler, D. S. and G. P. Greiner. 1989, April. Design considerations and selection of an emissioncontrol system operating at low temperatures for a MSW combustion facility. Paper presented atInternational Conference on Municipal Waste Combustion. Hollywood, FL.

Beachler, D. S., and J. A. Jahnke. 1981. Control of Particulate Emissions. (APTI Course 413). EPA450/2-80-066. U.S. Environmental Protection Agency.

Bethea, R. M. 1978. Air Pollution Control Technology: An Engineering Analysis Point of View. NewYork: Van Nostrand Reinhold.

Billings, C. E. and J. Wilder. 1970. Fabric Filter Systems Study. Vol. 1, Handbook of Fabric FilterTechnology. Springfield, VA: HRD Press.

Cheremisinoff, P. N. and R. A. Young, (Eds.). 1977. Air Pollution Control and Design Handbook, PartI. New York: Marcel Dekker.

McKenna, J. D. and D. Furlong. 1992. Fabric filters. In A. J. Buonicore and W. T. Davis (Eds.), AirPollution Engineering Manual. New York: Van Nostrand Reinhold.

McKenna, J. D. and G. P. Greiner. 1982. Baghouses. In L. Theodore and A. J. Buonicore (Eds.), AirPollution Control Equipment - Selection, Design, Operation and Maintenance. Englewood Cliffs,NJ: Prentice-Hall.

McKenna, J. D. and J. H. Turner. 1989. Fabric Filter-Baghouses I, Theory, Design, and Selection.Roanoke, VA: ETS.

Stern, A. C. (Ed.). 1977. Engineering Control of Air Pollution. Vol. 4, Air Pollution. 3rd ed. NY:Academic Press.

Theodore, L. and A. J. Buonicore. 1976. Industrial Air Pollution Control Equipment for Particulates.Cleveland: CRC Press.

2.0-3/95 3-1

Lesson 3Fabric Filter Design Variables

Goal

To familiarize you with the variables used by vendors to design fabric filter systems.

Objectives


1. Define pressure drop and recognize the equations used to calculate pressure drop

2. Define the term filter drag

3. Define the terms air-to-cloth ratio and filtration velocity

4. Identify the typical air-to-cloth ratios for shaker, reverse-air, and pulse-jet baghouses

Video Presentation (optional): If you have acquired the video titled, Pulse-Jet and Reverse-AirFabric Filters: Operating Principles and Components, please view it at the end of this lesson.

Introduction

Baghouses are designed by considering a number of variables: pressure drop, filter drag,air-to-cloth ratio, and collection efficiency. Although rarely done because it may not be pos-sible or practical, it is a good idea to use a pilot-scale baghouse during the initial stages of thebaghouse design. However, previous vendor experience with the same or similar process to becontrolled will generally be adequate for design purposes. Careful design will reduce the num-ber of operating problems and possible air pollution violations.

Pressure Drop

Pressure drop (∆p), a very important baghouse design variable, describes the resistance to airflow across the baghouse: the higher the pressure drop, the higher the resistance to air flow.Pressure drop is usually expressed in millimeters of mercury or inches of water. The pressuredrop of a system (fabric filter) is determined by measuring the difference in total pressure attwo points, usually the inlet and outlet. The total system pressure drop can be related to thesize of the fan that would be necessary to either push or pull the exhaust gas through the bag-house. A baghouse with a high pressure drop would need more energy or possibly a larger fanto move the exhaust gas through the baghouse.

Many different relationships have been used to estimate the pressure drop across a fabric filter.In a baghouse, the total pressure drop is a function of the pressure drop across both the filter

Lesson 3

3-2 2.0-3/95

and the deposited dust cake. Some pressure losses due to friction also occur as the gas streammoves through the baghouse.

The simplest equation used to predict pressure drop across a filter is derived from Darcy's lawgoverning the flow of fluids through porous materials and given as:

(3-1)

Where: ∆pf = pressure drop across the clean fabric, in. H2O (cm H2O)k1 = fabric resistance, in. H2O/(ft/min) [cm H2O/(cm/sec)]vf = filtration velocity, ft/min (cm/sec)

The term k1 is the fabric resistance (also called drag) and is a function of exhaust gas viscosityand filter characteristics such as thickness and porosity. Porosity describes the amount of voidvolume in the filter.

The pressure drop across the deposited dust cake can be estimated by using Equation 3-2 (Bill-ings and Wilder 1970). This formula is also derived from Darcy's law and the simplified formis given as:

(3-2)

Where: ∆pc = pressure drop across the cake, in. H2O (cm H2O)k2 = resistance of the cake, in. H2O/(lb/ft2-ft/min)

[cm H2O/(g/cm2-cm/sec)]ci = dust concentration loading, lb/ft3 (g/cm3)vf = filtration velocity, ft/min (cm/sec)t = filtration time, min (sec)

The term k2 is the dust-fabric filter resistance coefficient and is determined experimentally.This coefficient depends on gas viscosity, particle density and dust porosity. The dust porosityis the amount of void volume in the dust cake. The porosity is related to the permeability. Per-meability for the fabric only is defined in American Society of Testing and Materials (ASTM)standard D737-69 as the volume of air which can be passed through one square foot of filtermedium with a pressure drop of no more than 0.5 inches of water. The term k2 is dependent onthe size of the particles in the gas stream. If the particles are very small (< 2µm) k2 is high. Ifk2 is high, then the pressure drop will tend to increase and the bags will have to be cleanedmore frequently.

Filtration velocity also has an effect on k2. In more recent tests, conducted in the late 1980'sunder controlled conditions, the relationships of k2, particle size, and velocity have beenshown more clearly. Researchers including Dennis, Cass, and Cooper (1977) and Davis andKurzyske (1979) showed that both particle size and velocity have an effect on k2.

∆p vf f= k1

∆p c v tc i f= k22

Fabric Filter Design Variables

2.0-3/95 3-3

The total pressure drop equals the pressure drop across the filter plus the pressure drop acrossthe cake and is given as:

∆pt = ∆pf +∆pc (3-3)

∆pt = k1vf + k2 ci vf2 t (3-4)

Use equations 3-3 and 3-4 only as an estimate of pressure drop across shaker and reverse-aircleaning baghouses. In the industrial filtration process, complicated particle-fabric interactionsare occurring just after the filtration cycle begins. In addition, the filter resistance factor k1 cantake on two values; one value for the filter before it is brought on-line and another after the fil-ter has been cleaned. When the dust cake builds up to a significant thickness, the pressure dropwill become exceedingly high (> 10 in. H2O or 25 cm H2O). At this time the filter must becleaned. Some dust will remain on the cloth even after cleaning; therefore, the filter resistancelevel will be higher than during original conditions. A baghouse is normally operated with apressure drop across the unit of 4 to 10 in. H2O. But many units operate at less than 6 in. ofH2O. Bag cleaning is usually initiated when the pressure drop approaches this point.

Filter Drag

Filter drag is the filter resistance across the fabric-dust layer. The equation for filter dragessentially gives the pressure drop occurring per unit velocity. It is a function of the quantity ofdust accumulated on the fabric and is given as:

(3-5)

Where: S = filter drag, in. H2O/(ft/min) [cm H2O/(cm/sec)]∆p = pressure drop across the fabric and dust cake, in.

H2O (cm H2O)vf = filtration velocity, ft/min (cm/sec)

The true filtering surface of a woven filter is not the bag itself, but the dust layer. Dust bridgesthe pores or openings in the weave, plugging the openings with particles, increasing the dragrapidly.

Single Bag

A filter performance curve of a single bag of a fabric is shown in Figure 3-1. The drag isplotted versus the dust mass, or cake, deposited on the filter.

S =p

vf

∆

Lesson 3

3-4 2.0-3/95

Figure 3-1. Performance curve for a single woven bag

The point cr on the graph is the residual drag of the clean filter medium. The filter dragincreases exponentially up to a constant rate of increase. This is the period of cake repairand initial cake buildup. Effective filtration takes place while the filter drag increases at aconstant rate. When the total pressure drop reaches a value set by the system design, bagcleaning is initiated. At this point, the pressure drop decreases (almost vertically on theperformance curve) to the initial point. Cake repair begins when the cleaning cycle stopsand the cycle repeats. Baghouses are designed to remove most of the dust cake during thecleaning process. However, shaking or reverse-air baghouses are designed so that duringthe cleaning cycle some dust will remain on the bags. Therefore, a dust layer will not haveto be built up again on the openings in the weave of the fabric. If the fabric is cleaned tooefficiently, the cake repair cycle would be as long as the initial cake buildup, lessening theoverall effective filtration time of the baghouse.

Multicompartment Baghouse

In multicompartment baghouses where the various compartments are cleaned one at atime, the performance curve takes on a different shape. In this case the change in the curveis less pronounced than in Figure 3-1. The performance curve has a slight saw tooth shapefor the net pressure drop across the entire baghouse (Figure 3-2). Each of the minimumpoints on the curve represents the cleaning of an entire compartment. The average pres-sure drop would be represented by the dotted line. For optimum filtration rate and collec-tion efficiency, the baghouse should be designed to operate at a pressure drop thatapproaches a constant value. This involves careful selection of fabrics and cleaning mech-anisms for the baghouse. The weave, and any pretreatment of the fabric can affect the cakerepair time. Poor cleaning will increase the filter drag; therefore, the bags must be thor-oughly cleaned to reduce the filter drag effect. If cake repair time can be minimized, thepressure drop will be lower. Consequently, the effective filtration rate will be longer foroptimum filtering use.


2.0-3/95 3-5

Figure 3-2. Overall pressure drop of a multi-compartment baghouse

Pulse-Jet Baghouse

In a pulse-jet baghouse, felted filters are typically used as bag material (although wovenfabrics can also be used). Since there are no openings in the fabric material, there is no ini-tial cake buildup period. Effective filtration begins immediately as the dust is filtered bythe bag. The performance curve of a pulse-jet bag (or row of bags) is given in Figure 3-3.The pressure drop across the bags is slightly higher than with woven filters. The baghouseis usually operated with pressure drops of 4 to 6 in. of H2O and occasionally as high as 10in. of H2O. In a pulse-jet baghouse one row of bags is cleaned at a time. Some of the dustis knocked off the bags being cleaned while adjacent rows are still filtering. Bag cleaningcycles are initiated to keep the overall pressure drop across the baghouse within thedesigned range. If off-line cleaning is used, a compartment is taken out of service and bagcleaning is initiated in that compartment (module).

Figure 3-3. Performance curve of a pulse-jet bag or a row of bags

To test your knowledge of the preceding section, answer the questions in Part 1 of theReview Exercise.

Lesson 3

3-6 2.0-3/95

Filtration Velocity: Air-To-Cloth Ratio

The terms filtration velocity and air-to-cloth (A/C) ratio can be used interchangeably. The for-mula used to express filtration velocity is:

(3-6)

Where: vf = filtration velocity, ft/min (cm/sec)Q = volumetric air flow rate, ft3/min (cm3/sec)Ac = area of cloth filter, ft2 (cm2)

The air-to-cloth ratio (also called the gas-to-cloth ratio) is defined as the ratio of gas filteredin cubic feet per minute (cfm) to the area of filtering media in square feet. Typical units used toexpress the A/C ratio are:

(ft3/min)/ft2 or (cm3/sec)/cm2

These A/C ratio units reduce to velocity units. The units for filtration velocity are ft/min orcm/sec.

The term gross air-to-cloth ratio refers to the total amount of cloth area used to filter theentire flue gas stream. The term net air-to-cloth ratio is used to describe the net amount ofcloth available for filtering when one baghouse compartment is taken off-line for maintenanceor bag cleaning. The term net, net air-to-cloth ratio describes the amount of cloth availablewhen 2 compartments are taken off-line. In Lesson 5, you will learn how to calculate theseratios.

Bag Cleaning Comparisons

Air-to-cloth ratios describe how much dirty gas passes through a given surface area of filter ina given time. A high air-to-cloth ratio means a large volume of air passes through the fabricarea. A low air-to-cloth ratio means a small volume of air passes through the fabric. Whenusing the A/C ratios for comparison purposes the units are (ft3/min)/ft2 or (cm3/sec)/cm2. Like-wise, when using filtration velocities the units are ft/min or cm/sec.

Reverse-air cleaning baghouses generally have very low air-to-cloth ratios. For reverse-airbaghouses, the filtering velocity (filtration velocity) range is usually between 1 and 4 ft/min(0.51 and 2.04 cm/sec).

For shaker baghouses, the filtering velocity ranges between 2 and 6 ft/min (1.02 and 3.05cm/sec). More cloth is generally needed for a given flow rate in a reverse-air baghouse than ina shaker baghouse. Hence, reverse-air baghouses tend to be larger in size.

Occasionally, baghouse cleaning is accomplished by two methods in combination. Many bag-houses have been designed with both reverse-air and gentle shaking to remove the dust cakefrom the bag. This cleaning is called shake and deflate.

Pulse-jet baghouses are designed with filtering velocities between 2 and 15 ft/min (1 to 7.5cm/sec), with many velocities falling in the 2.0 to 2.5 ft/min range. Therefore, these units typ-

vQ

Afc

=


2.0-3/95 3-7

ically use felted fabrics as bag material. Felted material holds up very well under the high fil-tering rate and vigorous pulse-jet cleaning. Due to their typically higher A/C ratios, pulse-jetbaghouses are generally smaller in size than reverse-air and shaker baghouses. Pulse-jet clean-ing methods have the advantage of having no moving parts within the compartments. In addi-tion, pulse-jet units can clean bags on a continuous basis without isolating a compartmentfrom service. The duration of the cleaning time is short (< 1.0 sec) when compared to thelength of time between cleaning intervals (approximately 20 min to several hours). The majordisadvantage of high pressure cleaning methods is that the bags are subjected to more mechan-ical stress. Fabrics with higher dimensional stability and high tensile strength are required forthese units. Air-to-cloth ratios for the various cleaning methods are given in Table 3-1. Com-parisons of the cleaning methods are given in Table 3-2.

The A/C ratio (filtering velocity) is a very important factor used in the design and operation ofa baghouse. Improper ratios can contribute to inefficient operation of the baghouse. Operatingat an A/C ratio that is too high may lead to a number of problems. Very high ratios can causecompaction of dust on the bag resulting in excessive pressure drops. In addition, breakdown ofthe dust cake could also occur, which in turn results in reduced collection efficiency. Themajor problem of a baghouse using a very low A/C ratio is that the baghouse will be larger insize, and therefore have a higher capital cost.

Collection Efficiency

Extremely small particles (less than 1 µm in diameter) can be efficiently collected in a bag-house. Emission regulations for various industries including municipal waste combustors andhazardous waste incinerators require emission limits of 0.010 gr/dscf. Baghouse unitsdesigned with overall collection efficiencies of 99.9% (varying particle sizes) are common.Exhaust air from many baghouses can even be recirculated back into the plant for heating pur-poses, as long as the gas stream is not toxic.

Baghouses are not normally designed with the use of fractional efficiency curves as are someof the other particulate emission control devices. Vendors design and size the units strictly onexperience. The baghouse units are designed to meet particulate emission outlet loading andopacity regulations. There is no one formula that can determine the collection efficiency of abaghouse. Some theoretical formulas for determining collection efficiency have been sug-gested, but these formulas contain numerous (3 to 4) experimentally determined coefficients inthe equations. Therefore, these efficiency equations give at best only an estimate of baghouseperformance.

Table 3-1. Typical air-to-cloth ratio (filtration velocity) comparisonsfor three cleaning mechanisms

Cleaning Air-to-cloth ratio Filtration velocity

mechanisms (cm3/sec)/cm2 (ft3/min)/ft2 cm/sec ft/min

Shaking 1 to 3:1 2 to 6:1 1 to 3:1 2 to 6:1

Reverse-air 0.5 to 2:1 1 to 4:1 0.5 to 2:1 1 to 4:1

Pulse-jet 1 to 7.5:1 2 to 15:1 1 to 7.5:1 2 to 15:1Note: These may vary for specific applications.

Lesson 3

3-8 2.0-3/95

If you have acquired the video titled, Pulse-Jet and Reverse-Air Fabric Filters: OperatingPrinciples and Components, please view it before proceeding to the next lesson.

To test your knowledge of the preceding section, answer the questions in Part 2 of the ReviewExercise.

Table 3-2. Comparison of bag cleaning parameters

Parameter Shake cleaning Reverse-air cleaning Pulse-jet cleaning

Frequency Usually several cycles/second; adjustable

Cleaned onecompartment at a time,sequencing onecompartment afteranother; can becontinuous or initiatedby a maximum-pressure-drop switch

Usually, a row of bags ata time, sequenced onerow after another; cansequence such that noadjacent rows cleanone after another;initiation of cleaningcan be triggered bymaximum-pressure-drop switch or may becontinuous

Motion Simple harmonic orsinusoidal

Gentle collapse of bag(concave inward) upondeflation: slowlyrepressurize acompartment aftercompletion of abackflush

Shock wave passesdown bag; bagdistends from cagemomentarily

Peak acceleration 4 to 8 g 1 - 2 g 30 - 60 g

Amplitude Fraction of an inch tofew inches

NA NA

Mode Off-stream Off-stream On-stream: in difficult-to-clean applicationssuch as coal-firedboilers, off-streamcompartment cleaningbeing studied

Duration 10 to 100 cycles, 30 secto few minutes

1 to 2 min. includingvalve opening andclosing and dustsettling periods:reverse-air flow itselfnormally 10-30 sec

Compressed-air (40 -100 psi) pulse duration0.1 sec: bag roweffectively off-line

Common bag dimensions 5, 8, 12 in. diam; 8 to 10,22, 30 ft length

8, 12 in. diam; 22, 30, 40ft length

5 to 6 in. diam; 8 to 20 ftlength

Bag tension NA 50 to 120 lbs typical,optimum varies;adjusted after on-stream

NA

Sources: McKenna and Greiner 1982.Dennis and Klemm 1980.Morris 1984.


2.0-3/95 3-9

Review Exercise

Part 1

1. The ____________________ ____________________ of a system is determined by measuringthe difference in total pressure at two points.

2. True or False? Compared to a baghouse with a high pressure drop, a baghouse with a low pressuredrop would need a large fan and require more energy to move the gas through the baghouse.

3. What is the formula used to estimate the pressure drop across the clean fabric?

a. ∆pf = k1 vf

b. ∆pc = k2 vf

c. ∆pf = vc2 ci t

4. In the formula, pc = k2 ci vf2 t, used to estimate the pressure drop across the dust cake, the term k2

is the dust-fabric filter resistance coefficient. If the dust particles are very small (< 2 µm), k2 islarge. In this case, the pressure drop will:

a. Generally decreaseb. Generally increasec. Stay the same

5. Many baghouses operate with a pressure drop:

a. Between 15 and 20 in. H2Ob. Greater than 20 in. H2Oc. Of approximately 4 to 6 in. H2O

6. The filter resistance across a fabric-dust layer is called ________________________________________.

7. In a reverse-air or shaker baghouse, bags are cleaned:

a. To remove all dust completelyb. To leave a small amount of dust on the bagc. To leave approximately 60% of the dust cake on the bag

8. True or False? The pressure drop across a pulse-jet baghouse is generally higher than across areverse-air baghouse.

Part 2

9. True or False? The terms filtration velocity, (vf), and air-to-cloth ratio (A/C) can be used inter-changeably.

Lesson 3

3-10 2.0-3/95

10. Air-to-cloth ratios:

a. Describe how much dirty gas passes through a given surface area of filter in a given timeb. Describe how efficiently bags are cleaned by a pulse of reverse airc. Indicate how fast the dirty air passes through a square foot of cloth material

11. Air-to-cloth ratios are usually expressed in units of:

a. ft2/min.b. (ft3/min)/ft2

c. (ft/min)/ft2

12. A high air-to-cloth ratio means that a ____________________ volume of air passes through thefabric.

13. The air-to-cloth ratios for shaker baghouses are typically less than ____________________(cm3/sec)/cm2.

14. What are the usual air-to-cloth ratios for reverse-air baghouses?

a. Less than 4:1 (ft3/min)/ft2

b. Greater than 5:1 (ft3/min)/ft2

c. Between 3:1 and 8:1 (ft3/min)/ft2

15. The baghouses that usually have the highest air-to-cloth ratios are:

a. Pulse-jetb. Reverse-airc. Shaker

16. True or False? For a given exhaust flow rate, pulse-jet baghouses are usually smaller than reverse-air baghouses.

17. Operating the baghouse at air-to-cloth ratios ____________________ than the designed valuescan cause problems in the baghouse.

a. Greaterb. Less


2.0-3/95 3-11

Lesson 3

3-12 2.0-3/95

Review Answers

Part 1

1. Pressure dropThe pressure drop of a system is determined by measuring the difference in total pressure at twopoints.

2. FalseBaghouses with low pressure drops need less energy to move the exhaust gas than baghouses withhigh pressure drops.

3. a. ∆pf = k1 vfThe formula for estimating the pressure drop across the clean fabric is: ∆pf = k1 vf.

4. b. Generally increaseIn the formula, pc = k2 ci vf

2 t, used to estimate the pressure drop across the dust cake, the term k2

is the dust-fabric filter resistance coefficient. If the dust particles are very small (< 2 µm), k2 islarge. In this case, the pressure drop will generally increase.

5. c. Of approximately 4 to 6 in. H2OMany baghouses operate with a pressure drop of approximately 4 to 6 in. H2O, but the pressuredrop in some baghouses can sometimes be as high as 10 in. of H2O.

6. Filter dragThe filter resistance across a fabric-dust layer is called filter drag.

7. b. To leave a small amount of dust on the bagIn a reverse-air or shaker baghouse, bags are cleaned to the point where a small amount of dust isleft on the bag.

8. TrueThe pressure drop across a pulse-jet baghouse is generally higher than across a reverse-airbaghouse.

Part 2

9. TrueThe terms filtration velocity, (vf), and air-to-cloth ratio (A/C) can be used interchangeably.

10. a. Describe how much dirty gas passes through a given surface area of filter in a given time.Air-to-cloth ratios describe how much dirty gas passes through a given surface area of filter in agiven time.

11. b. (ft3/min)/ft2

Air-to-cloth ratios are usually expressed in units of (ft3/min)/ft2.


2.0-3/95 3-13

12. LargeA high air-to-cloth ratio means that a large volume of air passes through the fabric.

13. 3:1 (cm3/sec)/cm2 [6:1 (ft3/min)/ft2]The air-to-cloth ratios for shaker baghouses are typically less than 3:1 (cm3/sec)/cm2

[6:1 (ft3/min)/ft2].

14. a. Less than 4:1 (ft3/min)/ft2

Air-to-cloth ratios for reverse-air baghouses are usually less than 4:1 (ft3/min)/ft2.

15. a. Pulse-jetPulse-jet baghouses usually have the highest air-to-cloth ratios.

16. TrueFor a given exhaust flow rate, pulse-jet baghouses are usually smaller than reverse-air baghouses.

17. a. GreaterOperating the baghouse at air-to-cloth ratios greater than the designed values can cause problemsin the baghouse.

3-14 2.0-3/95

Bibliography

Beachler, D. S., and J. A. Jahnke. 1981. Control of Particulate Emissions. (APTI Course 413). EPA450/ 2-80-066. U.S. Environmental Protection Agency.


Billings, C. E. and J. Wilder. 1970. Fabric Filter Systems Study. Vol. 1, Handbook of Fabric FilterTechnology. Springfield, VA: HRD Press.

Cheremisinoff, P. N. and R. A. Young, (Eds.). 1977. Air Pollution Control and Design Handbook, PartI. New York: Marcel Dekker.

Davis, W. T. and F. R. Kurzyske. 1979. The effect of cyclonic precleaners on the pressure drop offabric filters. Filtration & Separation. 16(5): 451-454.

Dennis, R., R. W. Cass, and W. Cooper. 1977. Filtration model for coal fly ash with glass fabrics. EPA600-7-77-084. U.S. Environmental Protection Agency.

Dennis, R. and H. A. Klemm. 1980. Modeling concepts for pulse jet filtration. Journal of the AirPollution Control Association. 30(1):38-43.

Kraus, M. N. 1979. Baghouses: separating and collecting industrial dusts. Chemical Engineering.86:94-106.



Morris, W. J. 1984. Cleaning mechanisms in pulse jet fabric filters. Filtration and Separation.21(1):50-54.

Sittig, M. 1977. Particulates and Fine Dust Removal Processes and Equipment. Park Ridge, NJ:Noyes Data Corporation.

Stern, A. C. (Ed.). 1977. Engineering Control of Air Pollution. Vol. 4, Air Pollution. 3rd ed. NY:Academic Press.

2.0-3/95 4-1

Lesson 4Fabric Filter Material

Goal

To familiarize you with the construction of fabric filter material, fibers used, and problems affect-ing fabric life.

Objectives


1. Name two ways filters are constructed

2. List at least seven natural or synthetic fibers used to make filters and identify the conditionsunder which they are used

3. Define five fabric treatment processes

4. Name three failure mechanisms that reduce filter life

5. Briefly describe four types of fabric tests that are used for troubleshooting bag problems

Filter Construction

Bag filters can be made of woven or nonwoven materials. Nonwoven materials can further bedivided as felted or membrane. Most bags are either completely or partially made by weavingsince nonwoven fabrics are generally attached to a woven base called a scrim. Woven filtersare made of yarn with a definite repeated pattern. Felted filters are composed of randomlyplaced fibers compressed into a mat and attached to loosely woven backing material. A mem-brane filter is a special treatment where a thin, porous membrane (expanded polyfluorocar-bon) is bonded to the scrim, or support fabric. Woven filters are generally used with lowenergy cleaning methods such as shaking and reverse-air. Felted fabrics are usually used withhigher energy cleaning systems such as pulse-jet cleaning. Membrane filters were developedin efforts to achieve high efficiency particle capture and to handle flue gas conditions wherehigh moisture and resulting high pressure drop problems frequently occur.

Lesson 4

4-2 2.0-3/95

Woven Filters

Woven filters have open spaces around the fibers. The weave design used will depend onthe intended application of the woven filter. The simplest weave is the plain weave. Theyarn is woven over and under to form a checkerboard pattern (Figure 4-1). This weave isusually the tightest, having the smallest pore openings in the fabric. Consequently, itretains particles very quickly. This weave is not frequently used, because the bags tend tohave a high pressure drop (even without any dust cake).

Figure 4-1. Plain weave or checkerboard

Other weaves include the twill and sateen (satin). In the twill weave, yarn is woven overtwo and under one for a 2/1 twill and over three and under one for a 3/1 twill weave (seeFigure 4-2).

Figure 4-2. Twill weave patterns (2/1 and 3/1)

Fabric Filter Material

2.0-3/95 4-3

The twill weave does not retain particles as well as the plain weave, but does not tend toblind as fast. Bag blinding is a condition where the particles (dust) becomes embedded inthe filter over time and are not removed by the bag cleaning process. The twill weaveallows good flow rates through the filter and high resistance to abrasion. In the satinweave, yarn is woven over one and under four in both directions. Satin weave does notretain particles as well as the plain twill weave, but has the best (easiest) cake releasewhen the fabric is cleaned (Figure 4-3).

Figure 4-3. Sateen weave (satin weave)

Different weaving patterns increase or decrease the open spaces between the fibers. Thiswill affect both fabric strength and permeability. Fabric permeability affects the amountof air passing through the filter at a specified pressure drop. A tight weave, for instance,has low permeability and is better for the capture of small particles at the cost of increasedpressure drop.

Lesson 4

4-4 2.0-3/95

The true filtering surface for the woven filter is not the bag itself, but the dust layer or fil-ter cake. The bag simply provides the surface for capture of larger particles. Particles arecollected by impaction or interception as the open areas in the weave are closed. This pro-cess is referred to as sieving (Figure 4-4). Some particles escape through the filter until thecake is formed. Once the cake builds up, effective filtering will occur until the bagbecomes plugged and cleaning is required. At this point, the pressure drop will be exceed-ingly high and filtering will no longer be cost effective. The effective filtering time willvary from approximately 15 to 20 minutes to as long as a number of hours, depending onthe concentration of particulate matter in the gas stream.

Figure 4-4. Sieving (on a woven filter)

Felted Filters

Felted filters are made by needle punching fibers onto a woven backing called a scrim.The fibers are randomly placed as opposed to the definite repeated pattern of the wovenfilter. The felts are attached to the scrim by chemical, heat, resin, or stitch-bondingmethods.

To collect fine particles, the felted filters depend to a lesser degree on the initial dustdeposits than do woven filters. The felted filters are generally 2 to 3 times thicker thanwoven filters. Each individual randomly oriented fiber acts as a target for particle captureby impaction and interception. Small particles can be collected on the outer surface of thefilter (Figure 4-5).


2.0-3/95 4-5

Figure 4-5. Felted fabric filter

Felted filters are usually used in pulse-jet baghouses. A pulse-jet baghouse generally fil-ters more air per cloth area (higher air-to-cloth ratio) than a shaker or reverse-air unit.Felted bags should not be used in high humidity situations, especially if the particles arehygroscopic (these particles have an affinity to absorb moisture and thus become sticky).Clogging or blinding could result in such situations.

Fibers

The fibers used for fabric filters vary depending on the industrial application to be controlled.Early filters were mostly made from natural fibers such as cotton or wool. These fibers are rel-atively inexpensive but have temperature limitations (< 212°F or 100°C) and only averageabrasion resistance. Cotton is readily available making it very popular for low temperaturesimple applications. Wool withstands moisture very well and can be made into thick felteasily.

Synthetic fibers are more widely used today than natural fibers because they can operate athigher temperatures and better resist chemical attack. The synthetic fiber most often used forhigh temperature application is fiberglass or glass fibers. Fiberglass is the generic substancefound in Fiberglas . For low temperature applications polypropylene is the most inexpensivesynthetic fiber and is used in many industrial applications such as foundries, coal crushers, andfood industries. Nylon is the most abrasion-resistant synthetic fiber, making it useful in appli-cations filtering abrasive dusts. Polyesters such as Dacron fibers have good overall qualities toresist acids, alkalines, and abrasion and are relatively inexpensive, making them useful formany industrial processes such as smelters, foundries, and other metal industries.

Nomex fibers are widely used for fabric filter bags because of their resistance to relativelyhigh temperatures and to abrasion. Nomex is used for filtering dusts from cement coolers,asphalt batch plants, ferroalloy furnaces, and coal dryers.

Other registered trademark fibers such as Teflon, Fiberglas, Ryton, and P84, as well as carbonfibers can be used in very high temperature situations. Teflon has very good resistance to acidattack (except fluorine) and can withstand continuous temperatures up to 445°F (230°C).

Lesson 4

4-6 2.0-3/95

Fiberglas or glass is often used in baghouses that handle very high temperatures (up to 500°For 260°C) for continuous operation. About 90% of the baghouses currently operating on coalfired utility boilers use bags made with glass fibers (McKenna and Furlong 1992). Glass fibersare usually lubricated in some fashion so they will slide over one another without breaking orcutting during the cleaning cycle. Graphite is commonly used as a lubricant and will helpretain the upper service temperature limits. Glass fibers can break easily and require a verygentle cleaning cycle. Ryton is a felted filter made from polyphenylene sulfide fibers generallyattached to a polyfluorocarbon scrim. Ryton can operate at high temperatures (350°F or177°C) and shows good resistance to acids and alkalis. Fiberglas, Teflon, Nomex and Rytonhave been used to remove particulate emissions generated from industrial and utility coal-firedboilers (Belba et al. 1992).

Another material used to make bags is Gore-tex membrane manufactured by W. G. Gore andAssociates, Inc. The Gore-tex membrane is an expanded polytetrafluoroethylene (PTFE)membrane that is laminated with a variety of fibers such as Fiberglas, polyester, and Nomex toproduce felt and woven filters. Some test reports have indicated very good emission reduction(99.9+%), low pressure drops, increased bag life and higher air-to-cloth ratios using this mate-rial in metal industries, chemical industries, food industries, and coal-fired boilers. However,other fabrics have been able to obtain similar results.

Finally, for very high temperature applications (> 500°C), ceramic filters are now available(McKenna and Turner 1989). These filters show promise for high temperature applicationssuch as using the filters ahead of boiler superheater tube sections to remove particles andimprove heat transfer in the boiler tubes.

Table 4-1 lists a number of typical fibers used for fabric filters. The properties of the listedfibers include temperature limits, acid and alkali resistance, abrasion resistance, and relativebag costs. Table 4-1 is only a general guide since bag filters can be made of two or more layersof materials to achieve specific effects (i.e. strength, stability, filtering etc.) The cost (1992) ofa fiberglass bag 14 feet long and 6 inches in diameter is approximately $35 to $40. From Table4-1 the price of a Teflon bag of the same size is approximately $115 to $135.


2.0-3/95 4-7

Fabric Treatment

Fabrics are usually pretreated to improve their mechanical and dimensional stability. They canbe treated with silicone to give them better cake release properties. Natural fabrics (wool andcotton) are usually preshrunk to eliminate bag shrinkage during operation. Both synthetic andnatural fabrics usually undergo processes such as calendering, napping, singeing, glazing, orcoating.

These processes increase fabric life, improve dimensional stability (so that the bags retaintheir shape or fit after long use), and facilitate bag cleaning.

Calendering is the high pressure pressing of the fabric by rollers to flatten or smooth thematerial. Calendering pushes the surface fibers down onto the body of the filter medium. Thisis done to increase surface life and dimensional stability and to give a more uniform surface tobag fabric.

Napping is the scraping of the filter surface across metal points or burrs on a revolvingcylinder. Napping raises the surface fibers, creating a "fuzz", that provides a large number of

Table 4-1. Typical fabrics used for bags

Genericname

Fiber Maximum temperature Acidresistance

Alkaliresistance

Flexabrasion

resistance

Relativecost

Continuous Surges

°F °C °F °C

Natural fibercellulose

Cotton 180 82 225 107 poor excellent average 0.4

Polyolefin Polypro-pylene

190 88 200 93 excellent excellent good 0.5

Natural fiberprotein

Wool 200 93 250 121 good poor average 0.8

Polyamide Nylon 200 93 250 121 poor to fair excellent excellent 0.6

Acrylic Orlon® 240 116 260 127 very good fair average 0.7

Polyester Dacron® 275 135 325 163 good fair excellent 0.5

Aromaticpolyamide

Nomex® 400 204 425 218 fair very good very good 2.0

Fluoro-carbon

Teflon® 450 232 500 260 excellentexceptpoor forfluorine

excellentexceptpoor fortrifluoride,chlorine,andmoltenalkalinemetals

fair 6.7

Glass Fiberglas®

or glass500 260 550 288 good poor poor to fair 1.0

Polymer P84® 450 232 500 260 good fair fair 2.5

Polymer Ryton® 375 191 450 232 excellent excellent good 2.5-4.0

Sources: McKenna and Turner 1989.Greiner 1993.

Lesson 4

4-8 2.0-3/95

sites for particle collection by interception and diffusion. Fabrics used for collecting sticky oroily dusts are occasionally napped to provide good collection and bag cleaning ease.

Singeing is done by passing the filter material over an open flame, removing any straggly sur-face fibers. This provides a more uniform surface.

Glazing is the high pressure pressing of the fiber at elevated temperatures. The fibers are fusedto the body of the filter medium. Glazing improves the mechanical stability of the filter andhelps reduce bag shrinkage that occurs from prolonged use.

Coating, or resin treating, involves immersing the filter material in natural or synthetic resinsuch as polyvinyl chloride, cellulose acetate, or urea-phenol. This is done to lubricate thewoven fibers, or to provide high temperature durability or chemical resistance for variousfabric material. For example, glass bags are occasionally coated with Teflon or silicon graphiteto prevent abrasion during bag cleaning and aid in acid resistance. The Teflon coating is gener-ally applied at 10% of finished weight level.

Bag Failure Mechanisms

Three failure mechanisms can shorten the operating life of a bag. They are related to thermaldurability, abrasion, and chemical attack.

The chief design variable is the upper temperature limit of the fabric, or thermal durability.As shown in Table 4-1, fabrics have upper temperature limits which they can withstand contin-uously. The table also shows surge limits which are temperatures at which the baghouse can beoperated for short durations. Consult the fabric supplier for the length of time that the surgetemperature can be tolerated. The process exhaust temperature will determine which fabricmaterial should be used for dust collection. Exhaust gas cooling may be feasible, but theexhaust gas must be kept hot enough to prevent moisture or acid from condensing on the bags.

Another problem frequently encountered in baghouse operation is abrasion. Bag abrasion canresult from bags rubbing against each other, from the type of bag cleaning used, or where dustenters the bag and contacts the fabric material. For instance, in a shaker baghouse, vigorous

Table 4-2. Summary of pretreatment processes

Pretreatment Method Result Reason for use

Calendering High pressurepressing by rollers

Flattens, smooths, ordecorates

Increases surface lifeIncreases dimensional stabilityProvides more uniform fabric

surface

Napping Scraping acrossmetal points

Raises surface fibers Provides extra areas forinterception and diffusion

Singeing Passing over openflame

Removes stragglysurface fibers

Provides uniform surface area

Glazing High pressurepressing at elevatedtemperatures

Fibers fused to filtermedium

Improves mechanical stability

Coating Immersing in naturalor synthetic resin

Lubricates wovenfibers

Provides high temperaturedurability

Provides chemical resistancefor various fabric material

Source: McKenna and Greiner 1982.


2.0-3/95 4-9

shaking may cause premature bag deterioration, particularly at the points where the bags areattached. In pulse-jet units, the continual, slight motion of the bags against the supportingcages can also seriously affect bag life. As a result, a 25% per year bag replacement rate iscommon. This can be the single biggest maintenance problem associated with baghouses(Greiner 1992).

Bag failure can also occur from chemical attack to the fabric. Changes in dust compositionand exhaust gas temperatures from industrial processes can greatly affect the bag material. Ifthe exhaust gas stream is lowered to its dew point (either water or acid dew point), the designof the baghouse (fabric choice) may be completely inadequate. Proper fabric selection andgood process operating practices can help eliminate bag deterioration caused by chemicalattack. Lesson 6 discusses bag failures in more detail.

Fabric Testing

A number of standard ASTM tests can be conducted on bag filters either to verify the bag fil-ter's conformity with purchase specifications or to use as a troubleshooting tool for problembag failures. As with all measurement techniques, the results of these bag tests are relative.Often for these tests to be useful, they must be conducted over time in order to compare rela-tive degradation. In addition, with some of the newer fabrics, some of these tests may not bemeaningful. These tests can be used to indicate bag strength and flow loss. Four of the stan-dard tests performed are: permeability, MIT flex, Mullen burst strength, and tensile strength(McKenna and Turner 1989). These tests can be conducted if the installed baghouse is havingproblems with bag life or unusually high pressure drop.

Permeability

The permeability test is used to determine the amount of air that can flow through a givencloth area. Permeability is defined in ASTM Standard D-737-69 as the volume of air thatcan flow through one square foot of cloth at a pressure drop of no more than 0.5 in. w.g.(125 Pa). Because air permeability is not a linear function of the pressure difference mea-sured across fabric surfaces, the ASTM method prescribes that permeability tests be madeat a pressure drop of 0.5 in. w.g. (125 Pa). Certain fabrics may be too dense or too open tomaintain this pressure drop. In these cases, the ASTM method states that measured pres-sure drop be given in the test report.

The permeability of clean felts usually ranges between 15-35 ft/min (8-18 cm/s), whilelighter-weight woven materials have permeability values greater than 50 ft/min (25 cm/s).Permeability can be measured on clean or dirty bags. Dirty bags are usually tested in the"as received" state. They are then cleaned by vacuuming or washing and retested. Thesemeasured values can be compared to the original clean permeability of the fabric to deter-mine if bags that have been in service have become blinded. It is also possible that thepores in the fabric will open wider after extended use, which is shown by permeability val-ues higher than the original values. This condition, however, does not occur as frequentlyas blinding.

Lesson 4

4-10 2.0-3/95

MIT Flex

The MIT Flex Test is used to measure the ability of fabrics to withstand self-abrasion fromflexing. This test method is described in ASTM Standard D-2176-69, which is the stan-dard method for testing the endurance of paper with the MIT test apparatus.

The flex test has frequently been used to help determine the rate of deterioration of glassbags used in baghouses installed on coal-fired utility boilers. This test also helps provideinsight into the effect of bag tensioning on bag life. Flex testing is occasionally performedafter exposing the fabric to heat and/or acid in order to simulate conditions in utility boilerbaghouses. The test cannot be done with a continuous dust load on the fabric, which limitsthe comparison to actual field conditions.

Mullen Burst Strength

The Mullen burst strength test, described in ASTM Standard D-231, is designed to showthe relative total strength of fabrics to withstand pulsing or pressure.

For new glass fabrics, the Mullen burst test provides a good indication of whether the fab-ric has been weakened by the heat cleaning given the fabric before coating it with materi-als such as Teflon or silicon graphite.

Tensile Strength

The tensile strength test provides data on fabric stretch, elongation, and tear. This testmethod is described in ASTM Standard D-1682-64 for breaking load and elongation oftextile fabrics.

Tensile strength varies, depending on fabric type and weight. Synthetic fabrics generallytend to stretch or show greater elongation than natural fabrics. Glass materials usuallyhave high tensile strengths. The tensile test, used in combination with the Mullen burst testto compare strengths of new and used bags, can indicate the deterioration in strength ofused bags.


2.0-3/95 4-11

Review Exercise

1. Bag filters (bags) are made from ____________________ materials.

a. Wovenb. Feltedc. Membraned. All of the above

2. ____________________ filters are made from yarn with a definite repeated pattern.

3. The ____________________ and ____________________ weaves have better cake release thanthe simple weave.

4. In a woven filter, the woven material is not the true filtering surface. The dust____________________ provides the surface for filtering particles.

5. ____________________ filters are made by needle punching fibers onto a woven backing called ascrim.

6. The layer of woven material used for strength and support of felted or membrane material isreferred to as the ____________________.

7. Two natural fibers used for fabric filters are ____________________ and____________________.

8. Wool and cotton are inexpensive but are susceptible to failure at ________________________________________.

9. The fabric that is most often used in high temperature (> 200°C) industrial processes is:

a. Fiberglasb. Nylonc. Cottond. Polypropylene

10. True or False? Fabrics are pretreated to improve their mechanical and dimensional stability.

11. The filter surface of fabric material is sometimes scraped with metal points or burrs on a revolvingcylinder to create a "fuzz" on the material. This treatment is called:

a. Singeingb. Glazingc. Nappingd. Resin treating

12. True or False? Glass bags are occasionally coated with Teflon or silicon graphite to preventabrasion during bag cleaning.

Lesson 4

4-12 2.0-3/95

13. When fabric material is passed over an open flame to remove straggly fibers, the treatment iscalled ____________________.

14. Failure mechanisms that shorten bag operating life are:

a. Abrasionb. Temperature excursionsc. Chemical attackd. Varying particle size in flue gase. a, b, and c only

15. True or False? The chief design variable for prolonged bag life is the upper temperature limit ofthe bag.

16. The amount of air that can flow through a given cloth area is the ____________________ of thecloth.


2.0-3/95 4-13

Review Answers

1. d. All of the aboveBag filters (bags) are made from woven, felted, or membrane materials.

2. WovenWoven filters are made from yarn with a definite repeated pattern.

3. TwillSateenThe twill and sateen weaves have better cake release than the simple weave.

4. CakeIn a woven filter, the woven material is not the true filtering surface. The dust cake provides thesurface for filtering particles.

5. FeltedFelted filters are made by needle punching fibers onto a woven backing called a scrim.

6. ScrimThe layer of woven material used for strength and support of felted or membrane material isreferred to as the scrim.

7. WoolCottonTwo natural fibers used for fabric filters are wool and cotton.

8. High temperatureWool and cotton are inexpensive but are susceptible to failure at high temperature.

9. a. FiberglasThe fabric that is most often used in high temperature (> 200°C) industrial processes is Fiberglas.

10. TrueFabrics are pretreated to improve their mechanical and dimensional stability.

11. c. NappingThe filter surface of fabric material is sometimes scraped with metal points or burrs on a revolvingcylinder to create a "fuzz" on the material. This treatment is called napping.

12. TrueGlass bags are occasionally coated with Teflon or silicon graphite to prevent abrasion during bagcleaning.

13. SingeingWhen fabric material is passed over an open flame to remove straggly fibers, the treatment iscalled singeing.

Lesson 4

4-14 2.0-3/95

14. e. a, b, and c onlyThree failure mechanisms that shorten bag operating life are abrasion, temperature excursions, andchemical attack.

15. TrueThe chief design variable for prolonged bag life is the upper temperature limit of the bag.

16. PermeabilityThe amount of air that can flow through a given cloth area is the permeability of the cloth.

2.0-3/95 4-15

Bibliography

Beachler, D. S., and J. A. Jahnke. 1981. Control of Particulate Emissions. (APTI Course 413). EPA450/2-80-066. U.S. Environmental Protection Agency.

Belba, V. H., W. T. Grubb, and R. L. Chang. 1992. The potential of pulse-jet baghouse for utilityboilers. Part 1: A world-wide survey of users. Journal of the Air and Waste ManagementAssociation. 42(2):209-218.


Eggerstedt, P. M., J. F. Zievers and E. C. Zievers. 1993. Choose the right ceramic for filtering hotgases. Chemical Engineering Progress. 89(1):62-68.

Greiner, G. P. 1993. Fabric Filter - Baghouses II. Operation, Maintenance, and Trouble Shooting(A User’s Manual). Salem, VA: Valley Printers.




Proceedings: The user and fabric filtration equipment III. October 1-3, 1978. Air Pollution ControlAssociation Specialty Conference. Niagara Falls: NY.

Theodore, L. and A. J. Buonicore. 1976. Industrial Air Pollution Control Equipment for Particulates.Cleveland: CRC Press.

4-16 2.0-3/95

Bibliography

2.0-3/95 5-1

Lesson 5Fabric Filter Design Review

Goal

To familiarize you with the factors to be considered when reviewing baghouse design plans for airpollution control programs.

Objectives


1. List and explain at least six factors important in good baghouse design

2. Estimate the cloth area needed for a given gas process flow rate

3. Calculate the number of bags required in a baghouse for a given process flow rate

4. Calculate the gross air-to-cloth ratio, the net air-to-cloth ratio, and the net,net air-to-cloth ratiofor a baghouse design

Introduction

The design of an industrial baghouse involves consideration of many factors including spacerestriction, cleaning method, fabric construction, fiber, air-to-cloth ratio; and many construc-tion details such as inlet location, hopper design, and dust discharge devices. Air pollutioncontrol agency personnel who review baghouse design plans should consider these factorsduring the review process.

A given process might often dictate a specified type of baghouse for particulate emission con-trol. The manufacturer’s previous experience with a particular industry is sometimes the keyfactor. For example, a pulse-jet baghouse with its higher filter rates would take up less spaceand would be easier to maintain than a shaker or reverse-air baghouse. But if the baghouse wasto be used in a high temperature application (500°F or 260°C), a reverse-air cleaning baghousewith woven fiberglass bags might be chosen. This would prevent the need of exhaust gas cool-ing for the use of Nomex felt bags (on the pulse-jet unit), which are more expensive than fiber-glass bags. All design factors must be weighed carefully in choosing the most appropriatebaghouse design.

Review of Design Criteria

The first step in reviewing design criteria is determining the flow rate of the gas being filteredby the baghouse, which is measured in cubic meters (cubic feet) per minute. The gas volume

Lesson 5

5-2 2.0-3/95

to be treated is set by the process exhaust, but the filtration velocity or air-to-cloth ratio isdetermined by the baghouse vendor's design. The air-to-cloth ratio that is finally chosendepends on specific design features including fabric type, fibers used for the fabric, bag clean-ing mechanism, and the total number of compartments, to mention a few. Figure 5-1 depicts anumber of these design features. A thorough review of baghouse design plans should considerthe following factors.

Figure 5-1. Design considerations for a pulse-jet baghouse

Physical and chemical properties of the dust are extremely important for selecting the fabricthat will be used. These include size, type, shape, and density of dust; average and maximumconcentrations; chemical and physical properties such as abrasiveness, explosiveness, electro-static charge, and agglomerating tendencies. For example, abrasive dusts will deteriorate fab-rics such as cotton or glass very quickly. If the dust has an electrostatic charge, the fabricchoice must be compatible to provide maximum particle collection yet still be able to becleaned without damaging the bags.

Predicting the gas flow rate is essential for good baghouse design. The average and maxi-mum flow rate, temperature, moisture content, chemical properties such as dew point, corro-siveness, and combustibility should be identified prior to the final design. If the baghouse isgoing to be installed on an existing source, a stack test could be performed by the industrialfacility to determine the process gas stream properties. If the baghouse is being installed on anew source, data from a similar plant or operation may be used, but the baghouse should bedesigned conservatively (large amount of bags, additional compartments, etc.). Sometimes,

Fabric Filter Design Review

2.0-3/95 5-3

heavy dust concentrations are handled by using a baghouse in conjunction with a cyclone pre-cleaner, instead of building a larger baghouse. Once the gas stream properties are known, thedesigners will be able to determine if the baghouse will require extras such as shell insulation,special bag treatments, or corrosion-proof coatings on structural components.

Fabric construction design features are then chosen. The design engineers must determinethe following: woven or felt filters, filter thickness, fiber size, fiber density, filter treatmentssuch as napping, resin and heat setting, and special coatings. Once dust and gas stream proper-ties have been determined, filter choice and special treatment of the filter can be properlymade. For example, if the process exhaust from a coal-fired boiler is 400°F (204°C), with afairly high sulfur oxide concentration, the best choice might be to go with woven glass bagsthat are coated with silicon graphite or other lubricating material such as Teflon.

Along with choosing the filter type the designer must select the appropriate fiber type. Fiberstypically used include cotton, nylon, fiberglass, Teflon, Nomex, Ryton, etc. The design shouldinclude a fiber choice dictated by any gas stream properties that would limit the life of the bag.(See Lesson 4 for typical fabrics and fibers used for bags.) For more information about fabricconstruction, see McKenna and Turner (1989).

Proper air-to-cloth (A/C) ratio is the key parameter for proper design. As stated previously,reverse-air fabric filters have the lowest A/C ratios, then shakers, and pulse-jet baghouses havethe highest. For more information about air-to-cloth ratios, see McKenna and Turner (1989).

Once the bag material is selected, the bag cleaning methods must be properly matched withthe chosen bags. The cost of the bag, filter construction, and the normal operating pressuredrop across the baghouse help dictate which cleaning method is most appropriate. For exam-ple, if felted Nomex bags are chosen for gas stream conditions that are high in temperature andsomewhat alkaline (see Table 4-1), pulse-jet cleaning would most likely be used.

The ratio of filtering time to cleaning time is the measure of the percent of time the filtersare performing. This general, “rule-of-thumb” ratio should be at least 10:1 or greater(McKenna and Furlong 1992). For example, if the bags need shaking for 2 minutes every 15minutes they are on-line, the baghouse should be enlarged to handle this heavy dust concentra-tion from the process. If bags are cleaned too frequently, their life will be greatly reduced.

Cleaning and filtering stress is very important to minimize bag failures. The amount of flex-ing and creasing to the fabric must be matched with the cleaning mechanism and the A/C ratio;reverse-air is the gentlest, shaking and pulse-jet place the most vigorous stress on the fabric.For example, it would probably not be advisable to use woven glass bags on a shaker bag-house. These bags would normally not last very long due to the great stress on them during thecleaning cycle. However, fiberglass bags are used on reverse-air baghouses that use shake-and-deflate cleaning. Also, some heavy woven glass bags (16 to 20 oz) are used on pulse-jetunits (which also have high cleaning stress).

Bag spacing is very important for good operation and ease of maintenance. Bag spacingaffects the velocity at which the flue gas moves through the baghouse compartment. If bagsare spaced too close together, the gas velocity would be high because there is very little areabetween the bags for the gas stream to pass through. Settling of dust particles during bagcleaning would become difficult at high velocities. Therefore, it is preferable to space bags far

Lesson 5

5-4 2.0-3/95

enough apart to minimize this potential problem but not so far apart as to increase the size ofthe baghouse shell and associated costs.

For pulse-jet baghouses, bag spacing is important to prevent bag abrasion. Bag-to-bag abra-sion can occur at the bottom of the bags because the bags are attached to the tube sheet only attheir tops which allows them to hang freely. Slight bows in the bag support cages or a slightwarping in the tube sheet can cause bag-to-bag contact at the bottom of the bags.

Finally, access for bag inspection and replacement is important. For example, in a reverse-airunit, sufficient space between bags should be used so that maintenance personnel can checkeach bag visually for holes. The bag can either be replaced or a cap can be placed on the tubesheet opening to seal off the bag until it is later changed. The bag layout should allow the bagmaintenance technician to reach all the bags from the walkway. One measure of bag accessi-bility is called bag reach and is the maximum number of rows from the nearest walkway.There is no single value for bag reach, but typical units have a value of 3 or 4.

The compartment design should allow for proper cleaning of bags. The design shouldinclude an extra compartment to allow for reserve capacity and inspection and maintenance ofbroken bags. Shaker and reverse-air cleaning baghouses that are used in continuous operationrequire an extra compartment for cleaning bags while the other compartments are still on-linefiltering. Compartmentalized pulse-jet units are frequently being used on municipal solidwaste and hazardous waste incinerators for controlling particulate and acid gas emissions.

The design of baghouse dampers (also called baghouse valves) is important. Reverse-air bag-houses use inlet and outlet dampers for gas filtering and bag cleaning sequences. As describedin Lesson 2, during the filtering mode, the compartment’s outlet gas damper and inlet dampersare both open. During the cleaning sequence, the outlet damper is closed to block the flow ofgas through the compartment. The reverse-air damper is then opened to allow the air for bagcleaning to enter the compartment.

Dampers are occasionally installed in by-pass ducts. By-pass ducts, which allow the gasstream to by-pass the baghouse completely, are a means of preventing significant damage tothe bags and/or baghouse. Dampers in by-pass ducts are opened when the pressure drop acrossthe baghouse or the gas temperature becomes too high. However, many state regulatory agen-cies have outlawed the use of baghouse by-pass ducts and dampers to prevent the release ofunabated particulate emissions into the atmosphere.

Space and cost requirements are also considered in the design. Baghouses require a gooddeal of installation space; initial costs, and operating and maintenance costs can be high. Bagreplacement per year can average between 25 and 50% of the original number installed, partic-ularly if the unit is operated continuously and required to meet emission limits less than 0.010gr/dscf. This can be very expensive if the bags are made of Teflon which are approximately$100 for a 5-inch, 9-foot long bag, or Gore-tex which are approximately $140 for a 6-inch, 12-foot long bag.

The emission regulations in terms of grain-loading and opacity requirements will ulti-mately play an important role in the final design decisions. Baghouses usually have a collec-tion efficiency of greater than 99%. Many emission regulations (and permit limits) require thatindustrial facilities meet opacity limits of less than 10% for six minutes, thus requiring thebaghouse to operate continuously at optimum performance.


2.0-3/95 5-5

Typical Air-To-Cloth Ratios

During a permit review for baghouse installations, the reviewer should check the A/C ratio.Typical A/C ratios for shakers, reverse-air, and pulse-jet baghouses are listed in Table 3-1,Lesson 3.

Baghouses should be operated within a reasonable design A/C ratio range. For example,assume a permit application was submitted indicating the use of a reverse-air cleaning bag-house using woven fiberglass bags for reducing particulate emissions from a small foundryfurnace. If the information supplied indicated that the baghouse would operate with an A/Cratio of 6 (cm3/sec)/cm2 [12 (ft3/min)/ft2] of fabric material, you should question this informa-tion. Reverse-air units should be operated with a much lower A/C ratio, typically 1 (cm3/sec)/cm2 [2 (ft3/min)/ft2] or lower. The fabric would probably not be able to withstand the stressfrom such high filtering rates and could cause premature bag deterioration. Too high an A/Cratio results in excessive pressure drops, reduced collection efficiency, blinding, and rapidwear. In this case a better design might include reducing the A/C ratio within the acceptablerange by adding more bags. Another alternative would be to use a pulse-jet baghouse with theoriginal design A/C ratio of 6 (cm3/ sec)/cm2 [12 (ft3/min)/ft2] and use felted bags made ofNomex fibers. However, Nomex is not very resistant to acid attack and should not be usedwhere a high concentration of SO2 or acids are in the exhaust gas. Either alternative would bemore acceptable to the original permit submission.

Typical air-to-cloth ratios for baghouses used in industrial processes are listed in Tables 5-1and 5-2. Use these values as a guide only. Actual design values may need to be reduced if thedust loading is high or the particle size is small. When compartmental baghouses are used, thedesign A/C ratio must be based upon having enough filter cloth available for filtering whileone or two compartments are off-stream for cleaning.

Table 5-1. Typical A/C ratios [(ft3/min)/ft2] for selected industries1

Industry Fabric filter air-to-cloth ratio

Reverse air Pulse jet Mechanicalshaker

Basic oxygen furnaces 1.5-2 6-8 2.5-3

Brick manufacturing 1.5-2 9-10 2.5-3.2

Castable refractories 1.5-2 8-10 2.5-3

Clay refractories 1.5-2 8-10 2.5-3.2

Coal-fired boilers 1-1.5 3-5 -

Conical incinerators - - -

Cotton ginning - - -

Detergent manufacturing 1.2-1.5 5-6 2-2.5

Electric arc furnaces 1.5-2 6-8 2.5-3

Feed mills - 10-15 3.5-5

Ferroalloy plants 2 9 2

Glass manufacturing 1.5 - -

Grey iron foundries 1.5-2 7-8 2.5-3

Iron and steel (sintering) 1.5-2 7-8 2.5-3

Kraft recovery furnaces - - -Continued on next page

Lesson 5

5-6 2.0-3/95

Table 5-1. (continued)Typical A/C ratios [(ft3/min)/ft2] for selected industries1

Industry Fabric filter air-to-cloth ratio

Reverse air Pulse jet Mechanicalshaker

Lime kilns 1.5-2 8-9 2.5-3

Municipal and medical waste incinerators 1-2 2.5-4 -

Petroleum catalytic cracking - - -

Phosphate fertilizer 1.8-2 8-9 3-3.5

Phosphate rock crushing - 5-10 3-3.5

Polyvinyl chloride production - 7 -

Portland cement 1.2-1.5 7-10 2-3

Pulp and paper (fluidized bed reactor) - - -

Secondary aluminum smelters - 6-8 2

Secondary copper smelters - 6-8 -

Sewage sludge incinerators - - -

Surface coatings spray booth - - -1. High efficiency: a sufficiently low grain loading to expect a clear stack.Source: EPA 1976, revised 1992.


2.0-3/95 5-7

Table 5-2. Typical A/C ratios for fabric filters used for control ofparticulate emissions from industrial boilers.

Size of boiler(103 lb steam per hour)

Temperature (°F) Air-to-cloth ratio[(ft3/min)/ft2]

Cleaningmechanism

Fabricmaterial

260 (3 boilers) 400° 4.4:1 On- or off-linepulse-jet orreverse-air

Glass with 10%Teflon coating(24 oz/yd2)

170 (5 boilers) 500° 4.5:1 Reverse-airwith pulse-jetassist

Glass with 10%Teflon coating

140 (2 boilers) 360° 2:1 Reverse-air No. 0004Fiberglas withsilicone-graphite Teflonfinish

250 338° 2.3:1 Shake anddeflate

WovenFiberglas withsiliconegraphite finish

200 (3 boilers) 300° 3.6:1 Shake anddeflate

WovenFiberglas withsilicone-graphite finish

400 (2 boilers) Stoker, 285° to300°; pulverizedcoal, 350°

2.5:1 Reverse-air Glass withTeflon finish

75 150° 2.8:1 Reverse-air Fiberglas withTeflon coating

50 350° 3:1 On-line pulse-jet

Glass withTeflon finish

270 (2 boilers) 330° 3.7:1 On-line pulse-jet

Teflon felt(23 oz)

450 (4 boilers) 330° 3.7:1 On-line pulse-jet

Teflon felt(23 oz)

380 NA 2:1 Reverse-airvibratorassist


645 NA 2:1 Reverse-airvibratorassist


1440 (3 boilers) 360° 3.4:1 Shake anddeflate

WovenFiberglas withsilicone-graphite finish

Source: EPA 1979.

Lesson 5

5-8 2.0-3/95

Simple Cloth Size Check

Baghouse sizing is done by the manufacturer. This example will show you how to verify themanufacture’s measurements by doing a simple cloth size check. Given the process gasexhaust rate and the filtration velocity, you can estimate the amount of cloth required by thebaghouse. Once you know the total amount of cloth required and the dimensions of a bag, youcan calculate the number of bags in the baghouse.

Problem

Calculate the number of bags required for an 8-compartment pulse-jet baghouse with thefollowing process information and bag dimensions.

Q, process gas exhaust rate 100,000 ft3/min

A/C, gross air-to-cloth ratio 4 (ft3/min)/ft2

Bag dimensions:bag diameter 6 in.bag height 12 ft

Solution

1. Calculate the total gross cloth area. Use equation 3-6 (in Lesson 3):

Where: Ac = cloth area, ft2

Q = process exhaust rate, ft3/minvf = filtration velocity, ft/min

2. Determine the amount of fabric required per bag. Use the formula:

Where: Ab = area of bag, ft2

π = 3.14

Given: d = 0.5 ft, bag diameterh = 12 ft, bag height

Ab = 3.14 × 0.5 ft × 12 ft

= 18.84 ft2 required per bag

v =Q

Aor A =

Q

vfc

cf

A c =100,000 ft / min

4 ft / min

= 25,000 ft

3

2

A b = dhπ


2.0-3/95 5-9

3. Calculate the number of bags required in the baghouse.

From step 1: Ac = 25,000 ft2

From step 2: Ab = 18.84 ft2

So there will be an even number of bags in each of the 8 compartments, round thevalue 1326.96 up to the next highest multiple of 8 (i.e. 1,328). Thus, there will be 166bags (1,328/8) in each compartment.

4. Calculate the net air-to-cloth ratio. As you recall from Lesson 3, the net air-to-clothratio is the A/C ratio when one compartment is taken off-line for bag cleaning ormaintenance. Use the formula:

Given: Q = 100,000 ft3/min, process exhaust gas rateThe total number of compartments is 8.

From step 1: Ac = 25,000 ft2, total cloth area

Or, you can simply divide the gross air-to-cloth ratio by 7/8.

Number of bags =A

Ac

b

Number of bags =25,000 ft

18.84 ft1,326.96 bags

or 1,328 bags

2

2

=

( )A Cnet

/ =Q

Atotal # of compartments 1

total # of compartmentsc−

( )( )

( )A C

ft

net/

/ min

, /

/ min /

=

=

100,000 ft

ft

4.57 ft

3

3

25 000 7 82

2

( ) ( )

( )A C

ft

ft

net/

/ min /

/

/ min /

=ft

4.57 ft

3

3

4

7 8

2

2=

Lesson 5

5-10 2.0-3/95

5. Calculate the net, net air-to-cloth ratio (when two compartments are off-line).

( )( )

( )[ ]A Ctotal #net

gross/

, net=

A / C

of compartments 2

total # of compartments

−

( ) ( )

( )A C

ft

ft

net/

/ min /

/ min /

, net

3

3

4 ft

6 / 8

5.33 ft

=

=

2

2


2.0-3/95 5-11

Review Exercise

1. From the baghouses listed below, which would take up less space because of high filter rates?

a. Shakerb. Pulse-jetc. Reverse-air

2. True or False? Gas and dust stream properties influence filter choice.

3. An appropriate “rule of thumb” ratio of filtering time to cleaning time should be at least:

a. 3:1b. 1.5:1c. 5:1d. 10:1

4. True or False? An air-to-cloth ratio that is too high results in reduced pressure drops.

5. Nomex is not very resistant to:

a. HClb. CO2

c. SO2

d. Leade. a and c, only

6. Calculate the area of a bag (Ab) given a bag diameter of 15 inches and a bag height of 20 feet.

a. 942 ft2

b. 70.5 in.2

c. 78.5 ft2

d. 25 ft2

7. If the cloth area (Ac) is known to be 4,050 ft2, how many bags would be used in a baghouse withthe bag area (Ab) given above?

a. 52 bagsb. 519 bagsc. 120 bagsd. 10 bags

8. A baghouse has 8 compartments and a gross air-to-cloth ratio of 2.0 (ft3/min)/ft2. What is the netair-to-cloth ratio?

a. 1.75 (ft3/min)/ft2

b. 2.29 (ft3/min)/ft2

c. 2.66 (ft3/min)/ft2

d. 16.0 (ft3/min)/ft2

Lesson 5

5-12 2.0-3/95

9. For the baghouse information given in question 8 above, what is the net, net air-to-cloth ratio?

a. 1.75 (ft3/min)/ft2

b. 2.29 (ft3/min)/ft2

c. 2.67 (ft3/min)/ft2

d. 16.0 (ft3/min)/ft2


2.0-3/95 5-13

Review Answers

1. b. Pulse-jetDue to their high filter rates, pulse-jet baghouses take up less space than shaker and reverse-airbaghouses.

2. TrueGas and dust stream properties influence filter choice.

3. d. 10:1An appropriate “rule of thumb” ratio of filtering time to cleaning time should be at least 10:1. Ifthe ratio is much lower, the bags would be cleaned too frequently and may wear out too quickly.

4. FalseAn air-to-cloth ratio that is too high results in higher pressure drops.

5. e. a and c, onlyNomex is not very resistant to HCl and SO2 (acid gases).

6. c. 78.5 ft2

Solution:

1. Calculate the area of a bag (Ab).

Given: π= 3.14d = 15 in., diameter of bagh = 20 ft, height of bag

A b = dhπ

A b = 3.14 15 in.1 ft

12 in.20 ft

= 78.5 ft 2

× × ×

Lesson 5

5-14 2.0-3/95

7. a. 52 bags

Solution:

1. Calculate the number of bags.

Given: Ac = 4,050 ft2, the total cloth areaAb = 78.5 ft2, the area of a bag

8. b. 2.29 (ft3/min)/ft2

Solution:

1. Calculate the net air-to-cloth ratio using the following equation:

Given: (A/C)gross = 2.0 (ft3/min)/ft2

The total # of compartments is 8.

Number of bags =A

Ac

b

Number of bags =4,050 ft

78.5 ft= 52 bags

2

2

( )( )

( )[ ]A Ctotal #net

gross/ =

A / C

of compartments 1


−

( ) ( )

( )A C

ft

ft

net/

/ min /

/

/ min /

=ft

2.29 ft

3

3

2

7 8

2

2=


2.0-3/95 5-15

9. c. 2.67 (ft3/min)/ft2

Solution:

1. Calculate the net, net air-to-cloth ratio using the following equation:

Given: (A/C)gross = 2.0 (ft3/min)/ft2

The total # of compartments is 8.

( )( )

( )[ ]A Ctotal #net

gross/

, net=

A / C

of compartments 2


−

( ) ( )

( )A C

ft

ft

net/

/ min /

/

/ min /

, net

3

3

=ft

2.67 ft

2

6 8

2

2=

5-16 2.0-3/95

Bibliography

Fine particle fabric filtration. Proceedings: Symposium on the use of fabric filters for the control ofsubmicron particulates. April 8-10, 1974. Boston, MA. Journal of the Air Pollution ControlAssociation. 24(12):1139-1197.



Proceedings: The user and fabric filtration equipment III. October 1-3, 1978. Air Pollution ControlAssociation Specialty Conference. Niagara Falls: NY.



2.0-3/95 6-1

Lesson 6Fabric Filter Operation and Maintenance

Goal

To familiarize you with typical baghouse operation and maintenance problems.

Objectives


1. Identify typical steps for baghouse inspection prior to starting up

2. Identify typical parameters that a facility operator should monitor while operating the bag-house

3. Describe typical operating problems associated with shaker, reverse-air, and pulse-jetbaghouses

Introduction

This lesson provides a general overview of common operating problems and maintenancepractices for fabric filter systems. The text is written as a general guide for both baghouseoperators and air pollution regulatory agency inspectors and permit reviewers. For the bag-house system operators, there are checklists and general guidelines on what to look for or toavoid during the installation phases, instrumentation and recordkeeping suggestions for evalu-ating the operating systems, and examples of some common operating problems that canoccur.

For the agency inspectors and permit reviewers, this lesson provides information that will beuseful for performing field inspections, or for reviewing operation and maintenance (O&M)plans that many state agencies require as part of air permit applications for air pollution con-trol systems. The lesson is intended to provide a general compilation of typical baghouse oper-ating problems and typical checklists used during installation, startup, and operation. Thelesson also provides agency permit review engineers with sufficient technical information todetermine if the facility baghouse operators have adequate O&M plans in place to assureproper operation of the baghouse and subsequent compliance with the regulations and/or per-mit limits.

A number of sections of this lesson were extracted from the sources listed in the SuggestedReadings section at the end of this lesson. These sources provide much greater detail on fabricfilter system O&M procedures.

Lesson 6

6-2 2.0-3/95

Installation

Depending on the baghouse chosen, installation and initial operation startup may take from afew days to a few months. In any case, proper installation procedures will save time andmoney and will also help in future operation and maintenance of the baghouse.

Good coordination between the baghouse designer and the installation and maintenance per-sonnel will help keep the baghouse running smoothly for years. Occasionally this coordinationis overlooked. The baghouse is installed, turned on, and forgotten about until it stops workingcompletely. By then it may be too late to keep the unit going, and the baghouse may have to berebuilt or even scrapped. Some key features for the facility operator to evaluate during theinstallation period are listed here:

• Easy access to all potential maintenance areas - fans, motors, conveyors, discharge valves,dampers, pressure and temperature monitors, and bags

• Easy access to all inspection and test areas - stack testing ports and continuous emissionmonitors (opacity monitors)

• Weather conditions - the baghouse must be able to withstand inclement weather such asrain or snow

The following features have been suggested for a properly designed and installed baghouse(McKenna and Greiner 1982):

1. Uniform air and dust distribution to all filters. Duct design, turning vanes, and deflectionplates all contribute to uniform gas distribution. Often, this equipment arrives loose and isfield-installed. If improperly installed, it can induce high airflow regions that will abradethe duct or bag filters or cause reentrainment and induce high-dust-concentration regionsthat can produce uneven hopper loading and uneven filter bag dust cake.

2. Total seal of system from dust pickup to stack outlet. Inleakage of air at flanges or collec-tor access points either adds additional airflow to be processed or short-circuits the processgases. Inleakage to a high-temperature system is extremely damaging, as it creates coldspots and can lead to dew point excursions (gas temperature falls below the dew point) andcorrosion. If severe, it can cause the entire process gas temperature to pass through thedew point and result in condensate on the bags. Early bag failure and high pressure dropwill generally result. The best check for leaks is for the installation technician to inspectthe walls from inside the system during daylight. Light penetration from outside isolatesthe problem areas. It is particularly important to seal the dust discharge points in negativesystems. Inleakage here will result in incomplete or no discharge, which can lead to reen-trainment problems, yielding high pressure drop and hopper fires.

3. Effective coatings and paint. Most systems are painted on the exterior surfaces only.Extra care should be taken to touch up damaged areas with a good primer and if equipmentis not delivered finish-painted, apply it as soon as possible following erection. Unpro-tected primers allow corrosion to occur and require sandblasting and costly repairs for thefacility operators. If the system has been internally protected with a coating, it should bethoroughly inspected for cracks and chips, particularly in corners, and repaired beforeoperation begins. A poor interior coating can be worse than none at all because it will trapcorrosive elements between the coating and the surface it was intended to protect.

Fabric Filter Operation and Maintenance

2.0-3/95 6-3

4. Properly installed filter bags. The filter bags are the heart of any fabric filter collectionsystem. Improper installation can result in early bag failure, loss of cleaning effectiveness,and thus high pressure drop and operating costs or increased stack emission. Each manu-facturer provides instructions on the proper filter bag installation and tensioning (whererequired). These must be explicitly followed. Very often, early bag failures can be tracedto improper installation. It is much easier for the installation technicians to check andrecheck bag connections, tensioning, locations, and so on, in a clean, cool, dry collectorthan it will be one day after startup. Bag maintenance usually accounts for 70% of annualmaintenance time and money. Extra efforts in this area during installation can have a sig-nificant effect.

5. Proper insulation installation. Insulation is typically used to prevent O&M problems onhigh-temperature collector systems. When handling high-temperature gases, it is impor-tant to maintain the temperature of the gas and all collector components coming in contactwith it above the gas dew point. Much of the time, all or a part of the insulation is field-installed. The installers should check to see that all surfaces and areas of potential heatloss are adequately covered. In particular, they should check to see that field flashing alsohas insulation beneath it. Cold spots cause local corrosion. Gross heat loss may causeexcessive warm-up time or lower the gas temperature below the dew point.

6. Total seal between dirty side and clean side of collector. Remember, the primary purposeof the dust collector is to separate the particulate matter from the gas by means of fabricfiltration. This means that all the gas must pass through the fabric. Any leaks bypassingthe fabric filters will directly emit dust to the stack and therefore reduce the collection effi-ciency of the system. The time to inspect "bypass leaks" is before startup, when everythingis clean and accessible. The best technique is to use a bright light on one side of the ple-num and visually observe for light penetration on the other. This is the most effective intotal darkness. The installers should take extra time to check this important area. Trackingdown stack emissions not associated with bag failures can be extremely difficult afterstartup.

7. Properly installed and operating dampers. Most systems employ several dampers to iso-late areas of the system or control the volume of air flow. Proper alignment of both inter-nal blades and the operating linkage is important. In high-temperature applications, specialcare must be taken to allow for proper operation and sealing at the operating temperatures.Some dampers may require readjusting after reaching high-temperature operation. Inmodular systems, single modules are normally isolated for bag cleaning and maintenance.Leakage through these isolation dampers can cause improper bag cleaning. It will also cre-ate a very poor ambient condition for maintenance workers to work in. This, in someapplications, can pose a health hazard, and in all applications results in lower-qualityworkmanship or incomplete maintenance.

8. Properly operating mechanical components. Most mechanical components are designedwith a normal operating direction. Cylinder rod location, motor rotation, and so on, mustbe checked. Remember, when hooking up an AC motor, the installer has a 50% chance ofbeing correct on the first try. Not only will a backward-moving conveyor produce no dis-charge, but it can pack material so tightly that it bends the screw. Left uncorrected, areversed screw conveyor will result in a full hopper. The industry abounds with horror sto-ries where full hoppers have led to burned bags, or dust that has set up, requiring jackham-mers to remove it.

Lesson 6

6-4 2.0-3/95

9. Smoothly running fans. Fans must be checked for proper rotation, drive component align-ments, and vibration. Fans should be securely mounted to a sufficient mass to preventexcessive vibration.

10. Clean, dry compressed air. Most systems employ compressed air to operate dampers, con-trols, instruments, and so on. Probably more systems suffer shutdowns and maintenanceproblems due to poor-quality compressed air than for any other reason. Clean, dry air isnecessary to maintain proper operation of the pneumatic components. In installationswhere the ambient temperature drops below 32°F, a desiccant dryer system is generallyemployed. Sometimes, insulation of air lines and pneumatic components will be required.Often, these considerations are not included in the dust collector system, with "clean, drycompressed air to be supplied by the owner." Remember the air must be clean and drywhen it reaches the pneumatic component.

Each baghouse installation should have its own checklist reflecting the unique constructioncomponents of the unit. The installation crew should prepare a checklist before beginning thefinal inspection and initial startup. Table 6-1 shows an example of a typical inspection andstartup checklist. This checklist would be useful for the facility engineer to make sure that thebaghouse is properly installed.

Installation errors can have a disastrous effect on the operation and maintenance of the bag-house. Typical installation errors and their effect on O&M are given in Table 6-2.

Table 6-1. Inspection and startup checklist

1. Visually inspect:Structural connections for tightnessDuct flanges for proper sealFilter bags for proper seating in tube sheetDampers for operation and sequenceSystem fan, reverse-air fan, and conveyors - check for proper rotationElectrical controls for proper operationRotary valves or slide gates for operation

2. Remove inspection door and check conveyor for loose items or obstructions.

3. Adjust ductwork dampers - open or at proper setting.

4. Remove any temporary baffles.

5. Test horn alarm system, if included, by jumping connected sensors.

6. Start screw conveyors and check for proper operation.

7. Start reverse-air fan, if included.

8. Start system fan.

9. Log manometer and temperature (if appropriate) readings at 15-minute intervals; log readings.

10.Check to see that reverse-air dampers are cycling.

11. Adjust pressure drop cleaning initiation switch, if included.

12.Determine system air volume and adjust dampers, as required.

13.Check cell plates for dust leaks.

14.Check to see that dust is being discharged from hopper.Source: McKenna and Greiner 1982.

Reproduced by permission of ETS, Inc.


2.0-3/95 6-5

Operation and Maintenance Training

Before starting up the baghouse, the plant engineer should schedule training sessions for plantemployees that operate and maintain the baghouse. In these training sessions the followingsubjects should be covered: systems design, system controls, critical limits of equipment,function of each baghouse component, operating parameters that should be monitored, goodoperating practices, preventive maintenance, startup and shutdown procedures, emergencyshutdown procedures, and safety considerations.

Supervisors, operators, and maintenance people should attend O&M training sessions. Thetraining could be provided by the baghouse vendor or by a consulting company specializing inbaghouses. Many companies have in-house expertise to provide training. The length of train-ing would vary depending on the complexity of the system design. Average training will ordi-narily take at least 40 hours for full-time maintenance people.

Table 6-2. Typical installation errors and their effects on O&M

Item Immediate potential effect Long term effect

Baffle plates and turning vanes - improperinstallation or left out

Uneven dust distribution;uneven hopper loading;higher pressure loss

Bag wear, duct wear,hopper fires

Poor seal of flanges and access areas Inleakage resulting in:reduced inlet volumehigher fan volumehigher operating costslower baghouse temperature

Localized cold spotsresulting in:

component corrosionbag degradation

Poor seal at dust discharge flanges Incomplete discharge,reentrainment; hopper fires

Reentrainment; creepingpressure drop

Cracked or chipped paint and coatings Aesthetics Corrosion

Improper bag tensioning(reverse-air bags)

Ineffective cleaning; bagcollapse

Bag wear; high pressuredrop

Improper bag seating Stack emission Compliance failure; bagwear; high pressure drop

Incomplete insulation Cold spots Corrosion

Seal between dirty and clean aircompartments

Stack emission; dirtying ofclean side of plenum

Compliance failure

Duct damper alignment Loss of flow control Poor maintenance/ambientair inleakage

Screw conveyor direction reversed No discharge Bent screw(s); full hopper;fires

Fan mount Noise - vibration Broken components

Fan belt alignment Noise - improper fan volume Broken belts

Exposed compressed-air lines without drier Freeze-up - condensation Damaged downstreamcomponents

Lack of inspection access Lack of early warning signs Major problems

Lack of maintenance access Lack of regular preventivemaintenance

Major breakdowns

Source: McKenna and Greiner 1982.Reproduced by permission of ETS, Inc.

Lesson 6

6-6 2.0-3/95

Baghouse Startup and Shutdown

A specific startup and shutdown procedure should be supplied by the baghouse vendor.Improper startup and shutdown can damage the equipment. If hot moist gases are to be fil-tered, the baghouse must be preheated to raise the interior temperature in the baghouse abovethe dew point to prevent condensation and potential corrosion problems. This can be done byusing heaters in each compartment or by burning a clean fuel such as natural gas before filter-ing gases from a coal-fired boiler.

The baghouse must also be brought on-line slowly to avoid permanent damage to the fabric.Clean filters do not have a protective dust cake on them and are sensitive to dust abrasion andpenetration by fine particles. Penetration can lead to permanent residual pressure drop. Insome applications, bags are precoated with a protective dust layer prior to bringing the unit on-line. This protective dust can be the same dust from the process or other material such as pul-verized limestone. In all cases, the filter velocity should always be kept low until a sufficientdust cake is built up on the bags. This is indicated by a pressure drop of 1 to 2 inches H2O. Thegas flow can then be slowly increased to the designed rate (McKenna and Greiner 1982).

A suggested startup and shutdown list for baghouse system operators is given below.

Startup

1. Make sure all collector components are in working order and in proper mode.

2. Do not allow higher-than-design filtering velocities or air flow.

3. Avoid passing through (below) the dew point within the baghouse when dirty gasesare present. The system should be preheated to above the dew point with clean, hot airbefore the introduction of flue gas. During normal operation, maintain the temperatureapproximately 25 degrees above the dew point level. The gas dew point level can beobtained by making process exhaust gas measurements (acid concentration, moisture,and gas temperature) and appropriate calculations or by looking it up in literature suchas The Handbook of Chemistry and Physics.

4. Operate the bypass system to assure its readiness in an emergency situation.

5. Check all indicating and monitoring devices for proper operation.

Shutdown

1. Purge the collector with clean (hot when necessary) dry air before allowing the gastemperature to descend below the dew point. This is imperative when bringing a unitoff-line.

2. Do not store dust in the collector. Many maintenance workers have resigned afterspending a day with pick and shovel inside a dust collector hopper.

3. The bags should be "cleaned down" after dust flow ends, but not overcleaned. Theoperator should allow for one or two cleaning cycles then stop the cleaning process.


2.0-3/95 6-7

4. Finally, the operator should check to see that all components are in the proper shut-down mode.


Performance Monitoring

To determine if a baghouse is operating properly and to aid in troubleshooting when failuresoccur, the operator must monitor certain operating parameters. Routine monitoring of keyparameters, either on a continuous or periodic basis, is imperative for performance evaluationand problem diagnosis. An adequate baseline must be developed to determine when futurechanges in performance occur. Some typical parameters that are monitored are: inlet and outletgas temperature (only on units operated above ambient temperature), pressure drop, opacity,and gas velocity. In addition to these parameters that can be routinely measured, it can beimportant to periodically evaluate the chemical composition of the gas stream, including mois-ture, acid dew point, and particle loading and size distribution. The following describes howthe above parameters affect performance and the techniques used to measure each. In addition,there is also some common auxiliary equipment that should be monitored or periodicallychecked. These include receiver air pressure, bag tension, fan amperage, and high hopperlevel.

Gas Temperature

Gas temperature is important because fabrics are designed to operate within a given range.(See Lesson 4 for details on fabric operating conditions.) Exceedances of these fabric tem-perature limits, even for short periods of time, can weaken or damage the bags. Exposureof the fabric to temperatures above the maximum limits can cause immediate failure dueto loss of strength or elongation from melting. Minimum temperatures are related to thedew point temperature of the gas stream. Operation of the baghouse below these dew pointtemperatures can result in moisture or acid condensation and cause bag blinding or chemi-cal attack of the fabric. Condensation problems are one of the major causes of bag failures.

Temperature measurements are also used to indicate inleakage into the gas stream. Tem-perature drops across baghouses can range from 1 to 2 degrees on small units to up to 25degrees on large baghouses (EPA 1984). The facility must establish an acceptable or nor-mal operating range. If this range is exceeded, it indicates that a problem is occurring andneeds to be addressed.

To measure temperature, a thermocouple with digital, analog, or strip-chart display isused. The temperature signals are often tied to an alarm limit indicator to notify the opera-tor of trouble. Temperature measurements are generally made at the inlet and outlet of theunit with the inlet being the primary focus.

Pressure Drop

Baghouses are designed to operate within a certain pressure drop range, based on a spe-cific gas volumetric flow rate. Within this range during normal operation, the pressuredrop fluctuates with the cyclic cleaning process. The average baghouse pressure dropgradually increases as the filter cake builds on the bags and then takes a step decreaseimmediately after the compartment has been cleaned. The pressure drop across the bag-

Lesson 6

6-8 2.0-3/95

house gives an indication of the resistance to gas flow (drag) and the effectiveness of thecleaning system. Changes in pressure drop (either gradual or especially sudden) can indi-cate the need for maintenance. In addition, changes in the shape of the cleaning cycle pres-sure drop curve (i.e. pressure drop vs. cleaning cycle time) can also indicate the need formaintenance or change in system operation.

At a minimum the pressure differential across the baghouse should be continuouslyrecorded by the operator. Static pressure taps are connected to a transmitter/recorder sothat the differential can be monitored preferably from a central control room. The mostcommon problem with measuring the pressure drop is plugging of the static tap lines. Thepressure sensors should be shielded from direct impact of the dirty gas stream, and ameans to clean the lines should also be installed.

Opacity

Opacity is a measurement of the amount of light scattering that occurs because of the par-ticles in a gas stream. Although opacity is not a direct measurement of particle concentra-tion, it is a very good indicator of the amount of dust leaving the baghouse, and thusprovides a performance measure. Once a unit is operating at normal conditions, the opac-ity value for the system should be maintained within a narrow range. A continued elevatedopacity level indicates operating problems, such as bag failures. The opacity monitor (alsocalled transmissometer) can be used to identify the problem area. For multicompartmentbaghouses, each compartment can be isolated to identify the compartment where problemsare occurring.

There are a number of vendors who sell continuous opacity monitoring systems. Many ofthese monitors are double-pass opacity monitors where the light source is on one side ofthe stack while the reflector is on the other side of the stack. Continuous opacity monitor-ing systems provide continuous feedback on a real-time basis and for set averaging peri-ods. Coupled with a strip-chart or data acquisition system, they provide excellent trendinformation on baghouse operation. See Jahnke (1993) for more information on this topic.

Some facilities use broken bag detectors that give a relative indication of the dust loadingleaving the baghouse. Broken bag detectors are single-pass opacity monitors where thelight source is on one side of the stack and the detector is on the other side (there is noreflector). These are less expensive than double-pass opacity monitors and don’t meet theEPA performance specifications for opacity monitors.

Gas Volumetric Flow Rate

As discussed in Lesson 5, baghouses are designed to accommodate a range of gas flows. Ifgas flow rates increase, the operating pressure drop and air-to-cloth ratio will increase.This in effect means that the baghouse has to work harder and the bag life can be short-ened due to more frequent cleaning and high particle velocity.

Presently most sources do not continuously measure gas flow rates. Gas flow rates aregenerally only measured during emission compliance testing or when there is a perceivedproblem. Manual pitot tube traverses are normally used to measure gas flow (EPA Method1 and 2, see Code of Federal Regulations, Part 60). Because of new technologies and reg-ulations, some of the larger sources are beginning to install continuous flow measurementsystems. Multipoint pitot devices are being used to continuously measure gas velocity.


2.0-3/95 6-9

These devices generally consist of two tubes (in the same structure) with two sets of holes;one to sense the impact pressure and the other to measure static pressure. These devicesmust be calibrated to the individual stacks where they are installed.

Composition of Flue Gas

Baghouses are designed based on the composition of the flue gas they treat. Important fluegas parameters are moisture level, acid dew point, particle size, and concentration. If theoperating temperature falls below the condensation point, either during startup/shutdownor normal operation, blinding of the bags can occur. Similarly, if the temperature fallsbelow the acid dew point, there is a substantial risk of corrosion. These parameters aregenerally only measured during a diagnostic test or emission compliance stack test. How-ever, it is important to identify both of these minimum temperature points and have operat-ing procedures for startup/shutdown that minimize the condensation potential. Particlesize distribution and loading must be considered during design and also during operation;however, within certain limits (± 10 to 20%) changes in these parameters do not seriouslyaffect baghouse efficiency (EPA 1984). Unless there is a defined problem such as bagblinding or abrasion from particles these parameters are rarely measured.

Typical monitoring devices are listed in Table 6-3.

Recordkeeping and Routine Maintenance

Every operation and maintenance manual ever written states that "good recordkeeping is thekey to an effective operating system." In the real world, recordkeeping practices range fromnone to extensive computerized logging and retrieval systems. As stated previously, it is veryimportant to develop a baseline for both the baghouse operation and the process that it controlsto evaluate future performance and maintenance trends.

Table 6-3. Typical baghouse monitoring devices

Parameter Method of Measurement Function

Pilot lights Electronic on/off signals Show motors operating, compartmentson- or off-line, number of bags beingpulsed, etc.

Temperature indicators Thermocouple Alert operators of high or low temperatureconditions.

Pressure drop Manometer, magnehelic orphotohelic gauges

Determine pressure drop of variouspoints in the baghouse - across eachcomponent or the entire baghouse.

Opacity Transmissometer or visualobservation

Indicator of potential problems. Also,broken bags can be located by isolatingeach compartment to determine whichone causes the high opacity.

Gas flow Calibrated orifice (pitot tube)or an installed flow monitorusing an ultrasonic, thermal,or pressure differentialmeasurement technique

Indication of process change

Fan motor current(amps)

Ammeter Indication of gas flow and early warningsigns of potential fan failure if fan is notoperating at design levels.

Lesson 6

6-10 2.0-3/95

Although most operators agree that recordkeeping is imperative, the specifics on what param-eters are monitored and at what frequency are very site-specific. A number of performanceparameters were listed in the previous section. In addition to these parameters, the baghousevendor will generally provide some checklists for performing routine inspections. Thesechecklists should be used as templates to develop forms for the operators to fill out when mak-ing their rounds.

In addition to documenting the routine inspections, the operator should document all mainte-nance performed on the baghouse; especially bag replacement. A majority of the larger plantshave computerized work order systems that should be used to develop a special file for bag-house maintenance. In addition, since the most common and expensive failures are for bagreplacement, maintaining a trend of bag failures is imperative. A typical bag replacementrecord as shown in Figure 6-1 should be used. Using this type of tool can help identify failurepatterns due to design or operating practices.

Figure 6-1. Bag failure location record

Inspection frequencies of all baghouse components should be established by maintenanceengineers. Vendors' recommendations of an inspection schedule should be followed. A listingof typical periodic maintenance follows.

Bag Replacement RecordTube Sheet Layout

Unit__________, Compartment______________

A B C D E F G H I J K11 0 0 0 0 0 0 0 0 0 0 0

10 0 0 0 0 0 0 0 0 0 0 0

9 0 0 0 0 0 0 0 0 0 0 0

8 0 0 0 0 0 0 0 0 0 0 0

7 0 0 0 0 0 0 0 0 0 0 0

6 0 0 0 0 0 0 0 0 0 0 0

5 0 0 0 0 0 0 0 0 0 0 0

4 0 0 0 0 0 0 0 0 0 0 0

3 0 0 0 0 0 0 0 0 0 0 0

2 0 0 0 0 0 0 0 0 0 0 0

1 0 0 0 0 0 0 0 0 0 0 0

Access Door

Mark Failed Bags with X.Reason for Failure__________________________________Date________________________


2.0-3/95 6-11

Daily Maintenance

1. Check pressure drop.

2. Monitor gas flow rate.

3. Observe stack outlet visually or with a continuous monitor.

4. Monitor cleaning cycle, pilot lights, or meters on control panel.

5. Check compressed air on pulse-jet baghouses.

6. Monitor discharge system; make sure dust is removed as needed.

7. Walk through baghouse to check for normal or abnormal visual and audible condi-tions.

Weekly Maintenance

1. Check all moving parts on the discharge system including screw-conveyor bearings.

2. Check damper operation; bypass, isolation, etc.

3. Spot check bag tensioning for reverse-air and shaker bags.

4. Check compressed air lines including line oilers and filters.

5. Blow out any dust from manometer lines.

6. Verify temperature-indicating equipment.

7. Check bag-cleaning sequence to see that all valves are seating properly.

8. Check drive components on fan.

Monthly Maintenance

1. Spot check bag-seating condition.

2. Check all moving parts on shaker baghouses.

3. Check fan for corrosion and blade wear.

4. Check all hoses and clamps.

5. Spot check for bag leaks and holes.

6. Inspect baghouse housing for corrosion.

Quarterly Maintenance

1. Thoroughly inspect bags.

2. Check duct for dust buildup.

3. Observe damper valves for proper seating.

4. Check gaskets on all doors.

Lesson 6

6-12 2.0-3/95

5. Inspect paint on baghouse.

6. Calibrate opacity monitor.

7. Inspect baffle plate for wear.

Annual Maintenance

1. Check all welds and bolts.

2. Check hopper for wear.

3. Replace high-wear parts on cleaning system.

Sources: Reigel and Applewhite 1980; McKenna and Greiner 1982.



2.0-3/95 6-13

Bag Maintenance

Inspecting and changing bags takes a long time and are the highest maintenance costs in a bag-house. Bag failures occur at varying times depending on the operation of the collector. Thelonger the time before bag changeout, the lower the maintenance cost to the owner. Typicalbag life is from two to five years. Table 6-4 lists some common causes and reasons for bagfailures.

Bag failures can be spotted through daily monitoring and inspection. Stack opacity is a goodindication of bag failure. If the plume is dirty, then some problem exists, either in a singlecompartment or throughout the baghouse. In a compartmentalized baghouse it is possible tomonitor the stack while isolating a compartment. Stack emissions would be reduced if thecompartment with broken bags were taken off-line. In a noncompartmentalized baghouse itmay be necessary to check the entire unit for broken bags.

Three ways to search for broken bags are (Reigel and Applewhite 1980):

1. Hunt for the hole.

2. Hunt for the accumulation of dust which can be related to a nearby hole.

3. Use a detecting device.

In shaker and reverse-air baghouses where dust is collected on the inside of the bags, bag fail-ures occur frequently at the bottom of bags. Accumulation of dust on the cell plate is some-

Table 6-4. Common causes of fabric failures

Cause Result Reason

Improper bag installation Holes or tears in bagsReduce bag strength

Lack of proper vendor instructionsPoor access to bagsImproper tensioning, rough handling such

as bending or stepping on bagsBags too snug for cagesSharp edges on cages

High temperatures Loss of fabric strengthAttack finish of bag

causing self abrasion

Improper fabric for serviceNo high temperature alarmContinual operation at close to fabric

temperature limits

Condensation Alters adhesioncharacteristics of dustresulting in mudding orblinding

Chemical attack

Unit not preheated or purged properlyAir inleakageInadequate insulation

Chemical degradation Attack fibers and loss ofstrength

Improper fabric for service

High A/C ratio Increase in bag abrasion Change in process

High pressure drop Increase in bag abrasionBag tears

Poor cleaningBlind bagsIncrease in gas velocity

Bag abrasion Worn or torn bags Contact between bag and another surfaceHigh gas volumes or particle loadingLarge particle inspection on bag

Lesson 6

6-14 2.0-3/95

times visible, making it relatively easy to spot the failure. It may be necessary to inspect theentire circumference and length of the bag if the hole is higher up on the bag tube. In reverse-air baghouses, other bag failures can also occur near the anti-collapse rings and at the top cuffwhere the bags are attached. In shaker baghouses, bags tend to fail at the top where they areattached to hooks or clamps.

In pulse-jet baghouses it is normally very difficult to locate bags that have failed. However, inmany baghouses dust accumulation on the top tube sheet or in the blow pipe above the failedbag will be readily noticeable (Reigel and Applewhite 1980).

A technique for locating torn bags is to use fluorescent powder and a black light. Fluorescentpowder is injected in the inlet to the baghouse. An ultraviolet light is used to scan the clean airinside of the baghouse. Leaks can be detected by the glow of the powder getting through a tornbag. This technique is useful for spotting broken welds or leaks in the cell plates, tube sheetsor housing.

The importance of detecting broken bags depends on the baghouse design. In reverse-air andshaker units, leaks in the bags can cause air streams or jets of dust to abrade adjacent bags.This causes what is known as the "domino effect", where one torn bag creates another tornbag. In pulse-jet baghouses however, torn bags generally do not cause tears in adjacent bagssince the dust leaves the inside (clean side) of the bags. If opacity limits are exceeded beyondthe permit level, corrective action should be initiated immediately and the bag(s) should bechanged. It may take several broken bags to cause an opacity violation.

In the past, bags were usually replaced as they failed. However, a new bag in the vicinity ofold ones will be forced to take on more dust (air will tend to follow the path of least resistance)and will become worn-out quicker than the old "seasoned" bags (Reigel and Applewhite1980). It has become accepted practice in reverse-air and shaker baghouses to simply tie off atorn bag and stuff it into the cell plate. If the failure is close to the cell plate then the holeshould be plugged by using steel plate plugs with gaskets or sand bags to seal off the hole. Inpulse-jet baghouses with top access, a plug is placed over the tube sheet hole of the failed bag.

The operator should keep track of the bag failure rate of individual bags to correct any condi-tions that would cause premature bag failure. In addition, the tracking is helpful to determinethe scheduling of a complete changeout of bags at a convenient time.

Common Operating Problems

When a baghouse begins to have problems that cannot be readily identified, the operatorshould contact the vendor to identify and correct the problem. Problems and/or failure of com-ponents within a baghouse can occur for a number of reasons. Some problems may be uniqueto a particular type of baghouse design while others are generic to all fabric filters. The follow-ing is a summary of some of these problems (EPA 1984).

Dust Discharge Failures

Hopper pluggage can cause serious problems in a fabric filter. Many dusts flow less easilywhen they are cold. Thus, insulation, hopper heaters, air tight seals, and continuous dustremoval may be necessary to minimize the hopper pluggage problems. Regardless of thereason (cooling of the dust, inleakage, failure of the discharge system operation, or simplyusing the hoppers for storage), failure to remove the dust from the hopper usually results


2.0-3/95 6-15

in having to open up the hoppers to clean them out. The fugitive emissions generated by asingle cleaning out of the hoppers may be greater than the emissions emanating from thefabric filter outlet for an entire year. Therefore, the occurrences of hopper pluggage shouldbe minimized. Air inleakage is most common through the dust discharge valves and hop-per access doors.

Shaker Cleaning System Failures

Several problems are characteristic of shaker type baghouses:

1. Failure of the shaker motor may lead to excessive dust cake buildup on the bags andan increase in pressure drop. In some applications, when the gas flow is stopped byclosing the dampers, the dust will slide off the bag. In most applications, however, theshaker system is needed for adequate removal of the dust and maintenance of a rea-sonable pressure drop.

2. Shaker linkages must be maintained in a manner that allows the energy provided bythe shaker motor to be distributed through the shaking system to the bags. Becausethese systems are mechanical, periodic lubrication, checking for wear or loose parts,and replacement of broken parts are required to maintain their cleaning effectiveness.The only way to evaluate this system is to watch it in operation to ascertain that all thebags are being cleaned at approximately the same intensity.

3. Bag tension changes with the age of the bag and with the amount of material collectedon the dust layer. Bags that are too tight may not transfer the shaker energy effectivelyand may be damaged during shaking. Bags that are too loose may sag on the tubesheet, and bag abrasion may result from the bag being placed in the gas stream orbeing contacted by the thimble or other bags. Loose bags also may not use the clean-ing energy effectively and may block the flow of dust out of the bags if they sag, fold,or close off above the tube sheet.

Reverse-Air Cleaning Systems

Common problems associated with reverse-air cleaning baghouses include isolationdampers, bag tensioning, and corrosion. The reverse-air system is a low-energy systemand no gas flow can be present in the module or compartment being cleaned. The dampersystems for fabric filters with this cleaning mechanism tend to be complex because areverse flow of gas is used to collapse the bag, to break and release the dust cake, and toallow it to be collected and removed from the fabric filter. This requires a positive seal onthe reverse-air isolating damper (a poppet damper is often used). Without proper sealing,the bags may not collapse properly and the cleaning action may be ineffective. Unlike theother cleaning systems, relatively little energy is available to clean the fabric, as thereverse flow of gas through the bags is usually small compared with normal, on-line gasflow.

Failure of the isolation dampers is usually easily detected, as the actuators are generallypneumatically or hydraulically operated and the movement of the piston is visible. Too lit-tle movement of the piston usually indicates that the damper is not sealing properly. Insome situations, the failure of the damper system can be detected by a missing spike andsubsequent decrease in pressure drop after the affected module comes off-line for clean-ing. Moisture and oil in the compressed-air supply lines can cause blockage during freez-

Lesson 6

6-16 2.0-3/95

ing weather and result in the failure of these pneumatically operated systems. Damperoperation failures, however, usually result from failures of the controlling timers or pres-sure drop sensors that are used to activate the cleaning cycle at certain intervals or at cer-tain pressure-drop thresholds.

Buildup of materials around the dampers or deformation of the dampers or their seals cancause problems with proper isolation of a compartment for cleaning. Confirmation of poordamper sealing is only possible by internal examination of the equipment. Even internalinspection of the damper system may be inconclusive because the system must be cooledsufficiently for safe entry. An internal inspection, however, may indicate the presence oflight leaks, warped dampers and seals, or buildup or wear of the dampers caused by mate-rial passing through the fabric filter. The damper operation and seal should be checkedperiodically as part of a preventive maintenance program.

As with shaker baghouses, proper bag tension is essential to provide effective bag clean-ing. Bags that are too tight may not collapse enough to allow effective flexing of the dustcake. Too much tension can also damage the fabric. On the other hand, insufficient bagtension may cause the bags to collapse to the point where they are closed down during thereverse-air cleaning cycle (even when anticollapse rings are used). Loose bags also maysuffer abrasion from being sucked down into the thimble. Thimbles should be rounded andfree of sharp edges to prevent tears, if this should occur.

Proper bag tension is a function of attention to detail during the initial installation. Bagsmust be hung properly, without damage, to achieve the proper bag life expectancy. Bagtension will vary with the age of the bag and also within any given cleaning cycle as mate-rial builds up on the bags. Poor bag tension can increase bag wear, cause high pressuredrop, and shorten bag life.

Corrosion also can be a problem in this type of fabric filter. In some applications, mostnotably where acid dew point conditions have not been adequately considered, corrosionof the metal anticollapse rings has resulted in abrasion and wear of the bag at the site ofbag ring contact. Special alloy metals or coatings also can be used to minimize or elimi-nate corrosion problems.

Pulse-Jet Cleaning Systems

Common operating problems associated with pulse-jet cleaning systems include bag abra-sion, bag misalignment, and failure of the pulsing system. Pulse-jet fabric filters arewidely used because of their smaller size and their higher available cleaning energy whichallows for higher A/C ratios. The higher A/C ratios on this fabric filter type increase thepotential for fabric abrasion.

Typically, the bags in a pulse-jet fabric filter are suspended from a tube sheet and sup-ported by a cage. This single-point method of attachment allows the bag to move aroundduring normal operation. One source of bag abrasion is bag-to-bag contact due toimproper installation, poor alignment of the bag/cage assemblies with the tube sheet, orbent/warped cages. The rubbing together of the bags (usually at the bottom) can wear ahole in one or more of the bags.

The misalignment of bag/cage assemblies can also cause other problems. In some designs,the misalignment of the cage will prevent proper sealing of the bag with the tube sheet.


2.0-3/95 6-17

This may allow some of the dust to bypass the filter area, which decreases performancebut probably causes little or no change in pressure drop. Particularly abrasive dust hasbeen known to wear the bags and the tube sheet so severely at the point of the leak thatachieving an adequate seal may be impossible without replacing the tube sheet.

Another abrasion-related problem concerns poor distribution of inlet gas flow such thatthe larger particles strike the bottom of the bags opposite the inlet. Some designs areequipped with a blast or diffuser plate, which is designed to bring the gas flow below thebottom of the bags. When failure of the bags occurs within about 18 inches of the bottomon the side opposite the inlet, the presence and/or integrity of the blast plate or diffuserplate should be checked.

The pressure supplied by the compressed-air system must be high enough to clean theentire length of the bag during the pulse, but not so high that it damages the upper portionof the bag. Insufficient cleaning of the bag may gradually increase pressure drop andreduce the useful bag life. Too low compressed-air pressure, which is usually more com-mon than excessive pressure, may be caused by wear of the compressor rings, leakage ofdiaphragms, or excessive draining of the reserve of the compressors by other equipmenttied to a common supply line.

The leakage around a diaphragm, which can usually be detected by a continuous audibleleak, affects the cleaning effectiveness for all the bags. Although it may take several hoursor several days, the pressure drop usually will increase eventually if the leak is severeenough.

Failure of the solenoid(s) or the timer circuit may cause one or more rows not to becleaned. Effects on fabric filter performance may range from indiscernible to completecutoff of gas flow, depending upon the percentage area of the bags affected and the dustcharacteristics. Both mechanical and electronic timers are still in use, and both have cer-tain advantages and disadvantages. Both types must be kept in a dust-free, dry environ-ment and relatively free from the shocks and jolts that can accompany normal operations.Solenoid failures affect the row that has experienced the failure whereas timer failurestend to affect most, if not all, of the fabric filter system.

Several problems may result from improper operation of the pulse pipe cleaning system.First, the pulse pipe may not be properly aligned to provide effective cleaning to that row.Second, the alignment may be such that the nozzles are aimed directly at the bags and canblow holes in them. Lastly, a loose pipe may damage the tube sheet or even the fabric filterenclosure, which would necessitate additional repairs.

Although all of these problems are relatively common in most pulse-jet systems and mayproduce bag abrasion or shorten bag life, the one problem that seems to occur with great-est frequency is the presence of water and/or oil in the pulse-jet compressed-air supply.Water and oil that are blown into the bags during cleaning tend to absorb through the bagand cause bag blinding as the dust cake becomes wet. The result is an increase in pressuredrop and ultimate replacement of the blinded bags. The oil usually comes from leakage ofthe compressor rings and seals and the moisture from the atmosphere. Compressed-air

Lesson 6

6-18 2.0-3/95

systems can be equipped with small water and oil traps that work well if the system ismaintained and the humidity is not excessive.

A typical troubleshooting guide is listed in Table 6-5 and should be used only as a generalguide. When a baghouse begins to have problems that cannot be readily identified, theoperator should contact the vendor to assist in correcting the problem.

Table 6-5. Troubleshooting guide

Symptom Possible cause Remedy

High collector pressuredrop

Malfunction of bag-cleaning system

Ineffective cleaning

Reentrainment of dust in collectordue to low-density material orinleakage at discharge

Wetting of bags

Too high A/C ratio either throughadded capacity or improper originaldesign

Change in inlet loading or particledistribution

Check all cleaning-systemcomponents

Modify cleaning cycleReview with designer

Check discharge valvesLower A/C ratio

Control dew point excursionsDry bags with clean airClean bags with vacuum or wet

washVerify gas volumeReduce inlet volume if possibleReview with designer

TestReview with designerCheck for changes in process

operation or feed malfunction

Abnormally low pressuredrop

Manometer line(s) plugged

Manometer line(s) broken oruncoupled

Overcleaning of bags

Blow back through linesProtect sensing point from dust or

water buildupIncorporate autopurging system in

sensing lines

Verify with local manometerInspect and repair

Reduce cleaning energy and/orcycle time

Stack emission Broken bag

Bag permeability increase

Clean-to-dirty plenum leakage

Change of inlet conditions

See bag maintenance section

Test bagCheck cleaning energy/cycle and

reduce if possible

Inspect and repair

Test and reviewContinued on next page


2.0-3/95 6-19


Suggested Readings


Table 6-5. (continued)Troubleshooting guide

Symptom Possible cause Remedy

Puffing High pressure drop across baghouse

Low system fan speed

Improper duct balancing

Plugged duct lines

Poor hood design

Improper system fan damperposition

See above

Check drive systemIncrease speed

Rebalance system

Clean out

Evaluate temporary modificationsand implement

Check and adjust

Low dust discharge Inleakage at discharge points

Malfunction of discharge valve, screwconveyor or material transferequipment

Reentrainment of dust withincollector

Reentrainment of dust on filter bags

Inspect and repair seals or valves

Inspect and repair

Lower A/C ratio

Increase cleaning

Loud or unusual noises Vibrations

Banging of moving parts

Squealing of belt drives

Check source and make appropriatechanges



Corrosion Improper paint material orapplication

Improper insulation

Dew point excursions

Improper shutdowns

Repaint with appropriate material

Add insulation

Carefully monitor and controlprocess

Follow proper shutdown proceduresSource: McKenna and Greiner 1982.

Reproduced by permission of ETS, Inc.

Lesson 6

6-20 2.0-3/95

U.S. Environmental Protection Agency. 1984, December. Operation and Maintenance Manual forFabric Filters. Contract # 68-02-3919.

U.S. Environmental Protection Agency. 1987. Recommended Recordkeeping Systems for AirPollution Control Equipment, Part 1: Particulate Matter Control. Technical AssistanceDocument. EPA 340/1-86-021.


2.0-3/95 6-21

Review Exercise

Part 1

1. Inleakage at flanges or collector access points can cause condensate on the bags which may resultin early ____________________ failure and high ________________________________________.

2. True or False? A poor interior coating is worse than none at all.

3. Gas streams of high temperature should be maintained above the:

a. Ignition temperatureb. Gas dew pointc. Concentration limit

4. Cold spots in the baghouse can cause:

a. Local corrosionb. Firesc. Explosions

5. Many systems suffer shutdown and maintenance problems due to:

a. Low pressure dropb. Low air-to-cloth ratioc. Low dew pointd. Poor-quality compressed air

6. Before the baghouse is started up, the installation crew should prepare and use a____________________.

7. Who should supply a specific startup and shutdown procedure for baghouses?

a. The inspection teamb. The baghouse vendorc. The process plant ownerd. The air pollution agency

8. True or False? Bringing a baghouse on-line quickly helps seal woven bags and prevents damage tothe fabric.

9. To operate properly, bags must be coated sufficiently with:

a. Paintb. Condensatec. Dustd. All of the above

Lesson 6

6-22 2.0-3/95

10. During shutdown, before allowing the collector temperature to descend below the dew point,purge it with:

a. Clean dry airb. Cool spraysc. An alcohol cleanerd. All of the above

Part 2

11. It is important to monitor the operating temperature of the baghouse to avoid and/or document:

a. Exposure of bags to excessive temperatureb. Excessive air inleakagec. Condensation occurrencesd. All of the above

12. Measuring the ____________________ ____________________ across the baghouse gives anindication of resistance to flow and effectiveness of the cleaning system.

a. Temperature dropb. Pressure dropc. Opacity increased. Collection efficiency

13. An opacity monitor is useful to baghouse maintenance because:

a. Inspectors can monitor bag cleaning inside the baghouseb. Inspectors can monitor the process stack gas plumec. Inspectors can monitor operations of motors and on- and off-line compartments

14. If the gas velocity____________________ the operating pressure drop and air-to-cloth ratio willincrease.

a. Increasesb. Decreases

15. True or False? Pressure drop can be very easily measured merely by using two static pressure taps.

Part 3

16. The longer the time before the bag changeout, the ____________________ the maintenance costto the owner.

17. Bag failure can often be indicated by observing ________________________________________.

18. Broken bags can be discovered by:

a. Using a detecting deviceb. Visually searching out holesc. Looking for an accumulation of dustd. All of the above


2.0-3/95 6-23

19. In reverse-air baghouses, bag failures occur most frequently:

a. At the bag bottom and around the anti-collapse ringsb. Near the hookc. Along the internal support caged. All of the above

20. True or False? In reverse-air and shaker baghouse design, the "domino effect" means that one tornbag creates another torn bag.

21. True or False? In a pulse-jet baghouse, the opacity limits are exceeded when one bag is torn.

22. For processes that operate at elevated temperature, dust hopper pluggage can be caused by:

a. Lack of insulationb. Air inleakage through discharge valvec. Lack of hopper heatersd. All of the above

23. Poor cleaning in reverse-air systems can be caused by:

a. Compressed airb. Motor linkagesc. Isolation dampersd. All of the above

24. Bag tension is very important in ____________________ and ________________________________________ cleaning systems to assure proper operation.

25. For a pulse-jet cleaning system, excessive bag wear can be caused by:

a. Bent or warped cagesb. Poor inlet gas distributionc. High compressed-air pressured. All of the above

26. A very common problem of bag failure in pulse-jet systems is:

a. Oil or water in compressed air supplyb. Improper bag tensionc. Failure of isolation dampersd. All of the above

Lesson 6

6-24 2.0-3/95

Review Answers

Part 1

1. BagPressure dropInleakage at flanges or collector access points can cause condensate on the bags which may resultin early bag failure and high pressure drop.

2. TrueA poor interior coating is worse than none at all.

3. b. Gas dew pointGas streams of high temperature should be maintained above the gas dew point.

4. a. Local corrosionCold spots in the baghouse can cause local corrosion.

5. d. Poor-quality compressed airMany systems suffer shutdown and maintenance problems due to poor-quality compressed air.

6. ChecklistBefore the baghouse is started up, the installation crew should prepare and use a checklist.

7. b. The baghouse vendorThe baghouse vendor should supply a specific startup and shutdown procedure for baghouses.

8. FalseThe baghouse should be brought on-line slowly to avoid permanent damage to the fabric.

9. c. DustTo operate properly, bags must be coated sufficiently with dust.

10. a. Clean dry airDuring shutdown, before allowing the collector temperature to descend below the dew point,purge it with clean dry air.

Part 2

11. d. All of the aboveIt is important to monitor the operating temperature of the baghouse to avoid and/or document:

• Exposure of bags to excessive temperature

• Excessive air inleakage

• Condensation occurrences


2.0-3/95 6-25

12. b. Pressure dropMeasuring the pressure drop across the baghouse gives an indication of resistance to flow andeffectiveness of the cleaning system.

13. b. Inspectors can monitor the process stack gas plumeAn opacity monitor is useful to baghouse maintenance because inspectors can monitor the processstack gas plume.

14. a. IncreasesIf the gas velocity increases, the operating pressure drop and air-to-cloth ratio will increase.

15. TruePressure drop can be very easily measured merely by using two static pressure taps.

Part 3

16. LowerThe longer the time before the bag changeout, the lower the maintenance cost to the owner.

17. Stack opacityBag failure can often be indicated by observing stack opacity.

18. d. All of the aboveBroken bags can be discovered by doing the following:

• Using a detecting device

• Visually searching out holes

• Looking for an accumulation of dust

19. a. At the bag bottom and around the anti-collapse ringsIn reverse-air baghouses, bag failures occur most frequently at the bag bottom and around theanti-collapse rings.

20. TrueIn reverse-air and shaker baghouse designs, the "domino effect" means that one torn bag createsanother torn bag.

21. FalseIn a pulse-jet baghouse, the opacity limits may not be exceeded when one bag is torn.

22. d. All of the aboveFor processes that operate at elevated temperature, dust hopper pluggage can be caused by thefollowing:

• Lack of insulation

• Air inleakage through discharge valve

• Lack of hopper heaters

23. c. Isolation dampersPoor cleaning in reverse-air systems can be caused by the isolation dampers.

Lesson 6

6-26 2.0-3/95

24. ShakerReverse-airBag tension is very important in shaker and reverse-air cleaning systems to assure properoperation.

25. d. All of the aboveFor a pulse-jet cleaning system excessive bag wear can be caused by the following:

• Bent or warped cages

• Poor inlet gas distribution

• High compressed-air pressure

26. a. Oil or water in compressed air supplyA very common problem of bag failure in pulse-jet systems is oil or water in the compressed airsupply.

2.0-3/95 6-27

Bibliography

Cross, F. L. and H. E. Hesketh, (Eds.). 1975. Handbook for the Operation and Maintenance of AirPollution Control Equipment. Westport, CN: Technomic Publishing.



Jahnke, J. A. 1993. Continuous Emission Monitoring. Van Nostrand Reinhold: New York.



Reigel, S. A. and G. D. Applewhite. 1980. Operation and maintenance of fabric filter systems. In R. A.Young and F. L. Cross (Eds.), Operation and Maintenance for Air Particulate Control Equipment.Ann Arbor: Ann Arbor Science Publishers.

U.S. Environmental Protection Agency. 1973. Air Pollution Engineering Manual. 2nd ed. AP-40.

U.S. Environmental Protection Agency. 1984, December. Operation and Maintenance Manual forFabric Filters. Contract # 68-02-3919.

U.S. Environmental Protection Agency. 1987. Recommended Recordkeeping Systems for Air PollutionControl Equipment, Part 1: Particulate Matter Control. Technical Assistance Document. EPA340/1-86-021.

U.S. Environmental Protection Agency. 1991. Standards of performance for new stationary sources–general provisions. In Code of Federal Regulations–Protection of the Environment. 40 CFR 60.Washington, D.C.: U.S. Government Printing Office.

6-28 2.0-3/95

Bibliography

2.0-3/95 7-1

Lesson 7Industrial Applications of Fabric Filters

Goal

To familiarize you with the typical industrial uses and basic cost estimates of fabric filters.

Objectives


1. List six process industries that use fabric filters to control particulate emissions

2. Describe the specific uses and design features of fabric filters used in conjunction with acidgas control systems

3. Identify how to use charts and figures to estimate the cost of fabric filters

Introduction

Fabric filters are used for particulate emission reduction for many industrial applications. Fab-ric filters can be designed to collect particles in the submicrometer range with 99.9% controlefficiency. They are occasionally used to remove particles from exhaust air streams generatedby industrial processes where the clean air is recirculated back into the plant to help offsetspace heating needs. Fabric filters are used in the power generation, incineration, chemical,steel, cement, food, pharmaceutical, metal working, aggregate, and carbon black industries.Shaker, reverse-air, and pulse-jet fabric filters are used in a number of industrial applicationsas shown in Table 7-1.

Lesson 7

7-2 2.0-3/95

Fabric filters have been used for filtering fly ash in fossil-fuel fired boilers, municipal and haz-ardous waste incinerators, and a number of other industrial processes. In many industries fab-ric filters are becoming as popular as electrostatic precipitators for removing up to 99.9% ofthe particulate matter from particulate laden exhaust gas streams. The rapid growth in the useof fabric filters for particulate control has been aided by EPA's changing the definition of par-ticulate matter from total particulate matter to that fraction with a mean aerodynamic diameterof 10 micrometers or less (PM10). This is due to the fact that fabric filters are considered to besuperior collection devices for fine particulate control. Electrostatic precipitators (ESPs) arealso efficient at collecting fine particles. EPA course SI:412B Electrostatic Precipitator Oper-ation Review discusses these control devices.

Fossil-fuel Fired Boilers

Utility companies have been using fabric filters on coal-fired boilers since the mid 1970s andbecause of the advances in their design and operation, fabric filters have become a preferredtechnology for the control of particulate matter (Cushing 1990). Utility use of fabric filters isexpected to increase as emission limits become more stringent and regulatory attention to airtoxics increases. Fabric filters can also be integrated with acid gas controls providing an addeddimension not possible with some other forms of particulate control.

Based on a survey conducted by the Electric Power Research Institute (EPRI) in 1989, therewere 99 fabric filters operating on utility boilers representing 21,359 MW of generating capac-ity (Cushing 1990). Since the mid 1980s the application of fabric filters downstream of acidgas control equipment has increased substantially. Worldwide, industrial and utility use of fab-ric filters is even more dramatic as over 300 pulse-jet fabric filters are treating exhaust gasfrom coal-fired boilers alone (Belba 1992).

Table 7-2 lists some coal-fired boilers that use fabric filters for controlling particulate matteremissions that use either the reverse gas or shake/deflate cleaning method. The fabric mostcommonly used in the applications depicted on Table 7-2 is woven glass. Fabric coatings usedinclude Teflon, silicon graphite, and other proprietary acid resistant coatings.

Table 7-1. Typical industrial applications for baghouses

Shaker Reverse-air Pulse-jet

Screening, crushing, and conveying of rockproducts

Low temperature steel applicationsMetal workingMining operationsTextilesWoodworking processesChemical industryFood industryCoal-fired boilers

Cement kilnsLime kilnsElectric steel furnacesGypsum calciningOre smelters and roastersSintering plantsRock dryersFoundriesCarbon blackMagnesium oxide kilnsCoal-fired boilers

PharmaceuticalsFood industryWoodworkingSinter plantsMetal workingFoundriesTextilesChemical industryCoal-fired boilersAsphalt batch plantsMunicipal waste

incinerators

Industrial Applications of Fabric Filters

2.0-3/95 7-3

Design efficiencies of the fabric filters depicted on Table 7-2 ranged from 98 to 99.9%. Thelowest particulate emission rates were found on units using reverse-gas cleaning and rangedfrom 0.005 to 0.03 lb/MMBtu. Particulate emissions from fabric filters using reverse-gas

Table 7-2. Fabric filter performance data

Plantgenerating

capacity(MW)

Coaltype

Coal sulfurcontent %

Bagcleaningmethod

Gas temp°F

Flange toflange

pressuredrop in.

H2O

Tube sheetpressure

drop in. H2O

Gross air/cloth ratio

ft/min

Dustcakedensity

lb/ft2

Emissionrate

(lb/MMBtu)Stack

opacity %

Pulverized coal boilers

150852232234054478402452411015029530

56556525457044100166447391851851857935019119187.587.5

384384593593(79)

WSWSWSWSWSWSWSWSWBWSWSWSWS

WSWSWSWSWSWSWSWSWSEBEBEBAPEBEBEBAPAP

WSWSTLTLAP

0.240.370.370.370.410.430.430.470.490.520.520.520.61

0.30.30.330.450.520.520.60.610.690.850.860.871.791.832.22.42.62.7

0.350.360.430.491.79

RGRGRGRGRGRGRGRGRGRGRGRGRG

RG/SRG/SRG/SRG/SRG/SRG/SRG/SRG/SRG/SRG/SRG/SRG/SRG/SRG/SRG/SRG/SRG/SRG/S

S/DS/DS/DS/DS/D

325-270282267-305273-306260-280320360290283-296309290

275275230325290290315290240-280301305300325-304303400400

305320350350350

7.5-86765-5.53.5-55-64-566-7.15-5.24-56-7

885.5-6.85.5-6.56.5-8.24.2-65.5-6.56-84.8-5.55-6.55.75-6.565-9773.53.5

97.59-139-136

--6-4.6-4.82.5-3.53.8-4.5---4.2-4.73.5-4.5-

5.6-5.8-4.5-5.54-5----4-4.42.5-3.52.72.5-3.5-4-8----

8-7-11.57-11.5-

1.951.771.581.811.651.721.891.461.651.801.491.971.90(D)

1.71.71.98(D)1.912.09(D)1.98(D)2.0 (D)1.93(D)1.501.761.871.911.711.83(D)1.51.51.89(D)1.89(D)

3.22.82.62.61.9

----0.780.350.24---0.860.35-

0.350.280.290.19----0.640.32--------

0.23----

--0.012--0.0045-0.01-0.0150.013--

0.030.023-0.0080.016---0.0230.0290.0180.036--0.0390.1250.0850.085

0.030.051--0.01-0.07

----0.5-2-2-3---3-43-

3-53-511-2-----33-53-5------

2-4----

Pulverized coal boilers with dry FGD systems

27931944860415

WSWSWSWSNDL

0.310.360.520.61.08

RGRGRGRGS/D

185165180165200

4669.84-8

-----

1.581.60(D)1.54(D)2.00(D)2.24(D)

-0.09---

--0.030.0240.018

-----

Fluidized bed combustion boilers

160110

EBWS

0.330.39

RGS/D

290294

7.25-6.5

6.83.7-5.2

1.532.4-2.9

-0.23

< 0.030.0072

--

Coal Type: WS (Western Subbituminous); WB (Western Bituminous); AP (Anthracite/Petroleum Coke); TL (Texas Lignite); EB (Eastern Bituminous); NDL(North Dakota Lignite).

Cleaning Method: RG (Reverse Gas); RG/S (Reverse Gas with Sonic Assistance); S/D (Shake/Deflate).(D): Design Air-to-Cloth Ratio.Source: Cushing 1990.

Reproduced by permission of The Journal of the Air and Waste Management Association.

Lesson 7

7-4 2.0-3/95

cleaning with sonic assistance ranged from 0.008 to 0.125 lb/MMBtu. The units using shake/deflate cleaning had particulate emissions of 0.007 to 0.07 lb/MMBtu.

Table 7-3 lists some coal-fired boilers that use fabric filters with pulse-jet cleaning. This tablegives you an idea of the different combinations of bag material and A/C ratios that are beingused successfully at different sites. Woven glass and felted fabrics are the most common bagmaterials used. Fabric filters using 16 oz/yd2 woven fiberglass bags were found to be less effi-cient in particulate matter collection than fabric filters using 22 oz/yd2 bags. Fabric filtersusing 22 oz/yd2 bags achieved particulate emission levels consistently less than 0.02 lb/MMBtu (Belba 1992).

Table 7-3. Pulse-jet fabric filter performance data

SiteNo.

DesignVolume(Kacfm)

Boilertype

Coalsulfur (%)

Flue gas/ashmodifications Fabric

DesignA/C

(ft/min)

ActualA/C

(ft/min)

Particulateemissions(lb/MMBtu)

1 320 PC 0.35 Dralon T Felt 6.74 6.74 0.0808

2 320 PC 0.35 Dralon T Felt 6.74 6.74 0.0808

3 192 PC 0.48 Ryton Felt 3.71 4.17

4 178 PC 0.50 SCR:NH3 27 oz WG 3.84 2.71 0.0080

5 96 PC 2.20 DFSDA:Lime Glass Felt 4.00 0.0849

6 96 PC 1.00 DFSDA:Lime 22 oz WG 4.00 0.0636

7 96 PC 2.20 DFSDA:Lime Nomex/Ryton Felt 4.00 0.0446

8 132 PC 0.76 DFSDA:Lime/2yrs PC Ryton Felt 5.52 3.85

9 82 PC 0.68 16 oz WG 3.66 1.99 0.0534

10 205 PC 0.70 16 oz. WG 3.83 2.83 0.0280

11 60 PC 0.66 22 oz WG 3.70 2.01

12 50 PC MC 16 oz WG 3.42 0.0170

13 48 PC 0.82 MC 22 oz WG 2.50 1.56 0.0210

14 84 PC 0.58 16 oz WG 3.23 2.82 0.0159

15 860 PC 0.16 MC Nomex Felt 6.44 4.04 0.1050

16 530 PC 0.26 Nomex Felt 6.09 3.18 0.0180

17 1017 PC 0.38 Dralon T-Felt 3.94 3.80 0.0050

18 127 PC 0.40 Teflon Felt 4.46 4.40

19 127 PC 0.40 Teflon Felt 3.35 3.89 0.0695

20 127 PC 0.40 Teflon Felt 3.35 3.95 0.1981

21 127 PC 0.40 Teflon Felt 3.35 3.69 0.0735

22 127 PC 0.40 Teflon Felt 3.35 3.55 0.1263

23 463 PC 0.80 FSI:LS/FUI/ESP Nomex Felt 5.53 5.53 0.0162

24 220 PC 0.5-0.6 DFSDA:Lime Nomex Felt 5.15 5.15 0.0032

25 297 PC 0.7-1.5 Dralon T Felt 6.69 2.14 0.0106

26 297 PC 0.7-1.5 Dralon T Felt 6.69 0.0106

27 297 PC 0.7-1.5 Teflon Felt 6.69 0.0127

28 729 PC 0.75 SCR:NH3 Nomex Felt 5.44 0.0269

29 729 PC 0.75 SCR:NH3 Teflon Felt 5.44 0.0269

30 729 PC 0.75 SCR:NH3 Ryton Felt 4.77 2.66 0.0269

31 492 PC 0.60 FSI:LS/DSI:Na/SCR:U Nomex Felt 5.23 4.10 0.0180

32 320 PC 0.51 Daytex Felt 5.56 0.0230

33 320 PC 0.51 ASI Daytex Felt 5.56 4.65 0.0230

34 1017 PC 0.38 Dralon T Felt 3.35

35 180 PC 0.76 Daytex/Ryton Felt 5.30 3.38 0.0920

36 194 PC/WB ESP/RASDA:Lime Polyester Felt 4.88 4.13

37 194 PC/WB Tefaire Felt 4.88 0.0241

38 194 PC/WB Glass Felt 4.88 0.0241

39 178 Stoker 0.32 FSI:Dolomite/Cyc Dralon T Felt 4.92 2.98 0.0062

40 178 Stoker 0.32 FSI:Dolomite/Cyc Dralon T Felt 4.92 2.98

41 94 Stoker 1.30 RASDA:Lime Nomex Felt 2.76 0.0026

42 94 Stoker 1.30 RASDA:Lime Ryton Felt 2.76 1.80 0.0020



2.0-3/95 7-5

Dry Sulfur Dioxide (SO2) Control Systems

One technology for reducing sulfur dioxide (SO2) emissions from combustion sources thatdoes not generate any liquid sidestreams is dry flue gas desulfurization (FGD). This technol-ogy is prevalent in treating acid gas emissions from waste incinerators. In dry FGD, the fluegas containing SO2 is contacted with an alkaline material to produce a dry waste product fordisposal. This technology includes the following:

• Injection of an alkaline slurry in a spray dryer with collection of dry particles in a fabricfilter or electrostatic precipitator (ESP)

• Dry injection of alkaline material into the flue-gas stream with collection of dry particlesin a fabric filter or ESP

• Addition of alkaline material to the fuel prior to or during combustion

These technologies are capable of SO2 and hydrogen chloride (HCl) emission reduction rang-ing from 60 to 90% and 70 to 90+% respectively depending on which system is used. Typical

Table 7-3. (continued)Pulse-jet fabric filter performance data

SiteNo.

DesignVolume(Kacfm)

Boilertype

Coalsulfur (%)

Flue gas/ashmodifications Fabric

DesignA/C

(ft/min)

ActualA/C

(ft/min)

Particulateemissions(lb/MMBtu)

43 133 Stoker 0.40 Teflon Felt 4.33 5.09 0.0024



46 110 Stoker 0.51 Ryton Felt 5.73 6.37

47 91 AFBC 1.19 LS/2ndary MC 16 oz WG 3.16

48 91 AFBC 1.19 LS/2ndary MC Nomex Felt 3.16 0.0128

49 91 AFBC 1.19 LS/2ndary MC Nomex Felt 3.16 0.0168

50 91 AFBC 1.19 LS/2ndary MC Ryton Felt 3.16 3.56 0.0185

51 146 AFBC 3.11 LS 16 oz WG/G-T 4.52 2.57

52 59 AFBC 0.90 LS Nomex Felt 2.82 3.57 0.0041

53 161 AFBC 1.2-3.2 LS/FAR Nomex Felt 3.57 1.84 0.0095

54 56 AFBC 0.3-0.4 Sand Ryton Felt 2.97 3.29 0.0057

55 203 CFBC 0.63 LS/NH3/FAR 22 oz WG 3.60 3.98 0.0064

56 182 CFBC 4.28 LS 22 oz WG 3.15 2.68 0.0030

57 182 CFBC 4.28 LS 22 oz WG 3.15 0.0007

58 111 CFBC 0.84 LS Ryton Felt 4.59 0.0114

59 111 CFBC 0.84 LS Ryton Felt 4.59 3.40 0.0189

60 165 CFBC LS Ryton Felt 3.94 2.37 0.0095

61 165 CFBC LS Ryton Felt 3.94

62 99 CFBC 8.00 LS Ryton Felt 3.54 3.54

63 99 CFBC 8.00 LS P84 Felt 3.54

64 99 CFBC 8.00 LS 16 oz WG 3.54 0.3200

65 128 CFBC LS Nomex Felt 3.12Boiler Type: PC (Pulverized Coal); PC/WB (PC w/ Wet Bottom); AFBC (Bubbling Fluidized Bed Combustor); CFBC (Circulating Fluidized Bed Combustor).Flue Gas/Ash Modifications (Upstream of PJFF): ASI (Alcohol & Sludge Incineration); MC (Mechanical Collector); LS (Limestone in FBC Bed or Injected Into

Furnace); (Sand in FBC Bed); SCR:NH3 (Selective Catalytic DeNOX w/ Ammonia Injection); FAR (PJFF Fly ash Reinjection into FBC); DFSDA:Lime (Dual FluidSpray Dryer Absorber w/ Lime sorbent); FSI:LS (Furnace Sorbent Injection of Limestone); FUI (Furnace Urea Injection for NOX Control); ESP (ElectrostaticPrecipitator); Cyc (Cyclone); RASDA (Rotary Atomizer Spray Dryer Absorber); DSI:Na (Duct Sorbent Injection of Sodium Bicarbonate); SCR:U (SCR DeNOX w/Urea Injection).

Fabric: 16 oz WG (16 oz/square yard Woven Fiberglass); 22 oz WG (22 oz/square yard Woven Fiberglass); G-T (Gore-Tex Membrane); Nomex/Ryton (Nomex andRyton Felt Bags).

Source: Belba et al. 1992.Reproduced by permission of The Journal of the Air and Waste Management Association.

Lesson 7

7-6 2.0-3/95

reagents used with these technologies include lime, limestone (only in furnace injection),sodium carbonate, sodium bicarbonate, and nahcolite. These technologies have been used onboilers burning low sulfur coal (usually less than 2%), municipal waste incinerators, and haz-ardous waste incinerators and are attractive alternatives to wet scrubbing technology, particu-larly in the arid western U.S.

Spray Dryer with a Fabric Filter or ESP

One type of dry FGD installation is a spray dryer (sometimes referred to as a dry scrubber)and can be used on utility boilers and waste incinerators. Alkaline material is injected intoa spray dryer with dry particle collection in a fabric filter or ESP. Spray dryers have beenused in the chemical, food processing, and mineral preparation industries over the past 40years. Spray dryers are vessels where hot flue gases are contacted with a finely atomizedwet alkaline spray. The high temperatures of the flue gas, 250 to 400°F (121 to 204°C),evaporate the water from the wet alkaline sprays, leaving a dry powdered product. The dryproduct is collected in a fabric filter or ESP (Figure 7-1).

Figure 7-1. Spray dryer absorber and baghouse system


2.0-3/95 7-7

Flue gas enters the top of the spray dryer and is swirled by a fixed vane ring to cause inti-mate contact with the slurry spray (Figure 7-2). The slurry is atomized into extremely finedroplets by rotary atomizers or spray nozzles. The turbulent mixing of the flue gas withthe fine droplets results in rapid SO2 absorption and evaporation of the moisture. A smallportion of the hot flue gas may be added to the spray-dryer-discharge duct to maintain thetemperature of the gas above the dew point. Reheat prevents condensation and corrosionin the duct. Reheat also prevents bags in the fabric filter from becoming plugged or cakedwith moist particles.

Figure 7-2. Spray dryer

Sodium carbonate solutions and lime slurries are the most common absorbents used. Asodium carbonate solution will generally achieve a higher level of SO2 removal than limeslurries (EPA 1980). When sodium carbonate is used, SO2 removal efficiencies areapproximately 75 to 90%, lime removal efficiencies are 70 to 85% (EPA 1980). However,vendors of dry scrubbing systems claim that their units are capable of achieving 90% SO2

reduction using a lime slurry in a spray dryer. Lime is very popular for two reasons: (1) itis less expensive than sodium carbonate and (2) sodium carbonate and SO2 form sodiumsulfite and sodium sulfate, which are very soluble causing leaching problems when land-filled.

Some of the evaporated alkaline spray will fall into the bottom of the spray dryer. In coal-fired units where appreciable quantities of HCl do not exist, this material can be recycled.In municipal and hazardous waste incinerators, this spray dryer product is not recycleddue to the presence of calcium chloride. Calcium chloride is formed when HCl in the fluegas reacts with calcium hydroxide (lime slurry). Calcium chloride is very hygroscopic andcan plug bags, hoppers and conveyors if the material is not kept dry and the exhaust gasstream conveying this material is not kept well above the dew point. The majority of thespray reacts with SO2 in the flue gas to form powdered sulfates and sulfites. These parti-cles, along with fly ash in the flue gas, are then collected in a fabric filter or ESP. Fabric

Lesson 7

7-8 2.0-3/95

filters have an advantage because unreacted alkaline material collected on the bags canreact with any remaining SO2 in the flue gas. Some process developers have reported SO2

removal on bag surfaces on the order of 10% (Kaplan and Felsvang 1979). However, sincebags are sensitive to wetting, a 35 to 50°F (2.5 to 10°C) margin above the saturation tem-perature of the flue gas must be maintained in coal-fired installations (EPA 1980). Withwaste incineration facilities this margin must be increased to around 100°F (38°C) due tothe presence of calcium chloride. ESPs have the advantage of not being as sensitive tomoisture as fabric filters. However, SO2 removal is not quite as efficient when using ESPs.

In a spray dryer, finely atomized alkaline droplets are contacted with flue gas, which is atair preheater outlet temperatures of 250 to 400°F (121 to 204°C). The flue gas is humidi-fied to within 50 to 100°F (28 to 56°C) of its saturation temperature by the moisture evap-orating from the alkaline slurry. Reaction of SO2 with the alkaline material proceeds bothduring and following the drying process. However, sodium-based sorbents are more reac-tive in the dry state than calcium-based sorbents are. Since the flue gas temperature andhumidity are set by air preheater outlet conditions, the amount of moisture that can beevaporated into the flue gas is also set. This means that the amount of alkaline slurry thatcan be evaporated in the dryer is limited by flue gas conditions. Alkaline slurry sprayedinto the dryer must be carefully controlled to avoid moisture in the flue gas from condens-ing in the ducting, particulate emission control equipment, or the stack.

Many spray dryer systems have been installed on industrial and utility boilers. Some arelisted in Table 7-4. Additional experience in using FGD systems in combination withpulse-jet fabric filters is noted on Table 7-3 (see column “Flue Gas/Ash Modifications”).Permit reviewers should review the EPA BACT Clearinghouse for additional informationon spray dryers and baghouse systems. Spray dryers will be particularly useful in meetingNew Source Performance Standards (NSPS) for utility boilers burning low sulfur coal thatrequire only 70% SO2 scrubbing in addition to achieving the requirements of the acid rainprovisions included in Title IV of the 1990 Clean Air Act Amendments.


2.0-3/95 7-9

Dry Injection

In dry injection systems, a dry alkaline material is injected into a flue gas stream. This isaccomplished by pneumatically injecting the dry sorbent into a flue gas duct, or by pre-coating or continuously feeding sorbent onto a fabric filter surface. Most dry injection sys-tems use pneumatic injection of dry alkaline material in the boiler furnace area or in theduct that precedes the ESP or baghouse. Sodium-based sorbents are used more frequentlythan lime for coal-fired installations but hydrated lime is prevalent in waste burning incin-erators. Many dry injection systems have used nahcolite, a naturally occurring mineralwhich is 80% sodium bicarbonate found in large reserves in Colorado. Sodium carbonate(soda ash) is also used but is not as reactive as sodium bicarbonate (EPA 1980). The majorproblem of using nahcolite is that it is not presently being mined on a commercial scale.Large investments must be made before it will be mined commercially. Other naturalminerals such as raw trona have been tested; trona contains sodium bicarbonate andsodium carbonate.

Table 7-4. Commercial spray dryer FGD systems using abaghouse or an ESP

Station or plantSize(MW)

Installationdate System description Sorbent

Coalsulfur

content(%)

SO2

emissionremoval

efficiency(%)

Otter Tail PowerCompany: CoyoteStation No. 1,Beulah, ND

410 6/81 Rockwell/Wheelabrator-Fryesystem: four spraytowers in parallelwith 3 atomizers ineach: reverse-air-shaker baghousewith Dacron bags

Soda ash(sodiumcarbonate)

0.78 70

Basin Electric:Laramie RiverStation No. 3,Wheatland, WY

500 Spring1982

Babcock andWilcox: four sprayreactors with 12"Y-jet" nozzles ineach: electrostaticprecipitator

Lime 0.54-0.81

85-90

Strathmore PaperCo.: Woronco, MA

14 12/79 Mikropul: spraydryer and pulse-jetbaghouse

Lime 2-2.5 75

Celanese Corp.:Cumberland, MD

31 2/80 Rockwell/Wheelabrator-Frye: one spraytower followed by abaghouse

Lime 1-2 85

Source: EPA February 1980.

Lesson 7

7-10 2.0-3/95

Municipal Waste Incinerators

Spray dryers followed by fabric filters have become the control option of choice for municipalwaste incineration facilities. A survey conducted in 1990 by the Institute of Resource Recov-ery (IRR) reported that of 158 municipal waste combustion facilities, 47 used fabric filters forparticulate control. Almost all of these were preceded by a spray dryer. In fact spray dryers fol-lowed by fabric filters are typically considered best available control technology for municipalwaste incinerators since this equipment is effective in removing acid gases, particulate matter,and a number of hazardous air pollutants.

Modern municipal waste incinerators recover waste heat by using boilers to generate steamand electricity. After passing through the heat recovery equipment, the flue gas typically entersthe air pollution control system at 350 to 400°F (177 to 204°C). Emission controls typicallyconsist of a spray dryer absorber to remove acid gases followed by a fabric filter to removeparticulate matter, which includes acid gas reaction products, unreacted reagent, fly ash, andtrace metals. A survey of spray dryer applications on municipal waste incinerators in the U.S.shows that lime is used exclusively as the reagent. Onsite lime slaking systems are typicallyused to prepare the lime slurry.

A calcium hydroxide [Ca(OH)2] slurry, frequently referred to as lime slurry, is injected into thespray dryer reaction vessel as a finely atomized spray. Acid gases (mainly HCl and SO2) areabsorbed into the atomized lime slurry. The hot flue gas causes the water in the droplets toevaporate and leave behind dry reaction products (calcium salts).

Spray dryers must be operated at flue gas temperatures adequate to produce a dry reactantproduct. Spray dryers are typically designed to operate with an inlet (flue gas) temperature ofapproximately 350 to 400°F (177 to 204°C) and outlet temperature of 260 to 300°F (127 to149°C). Some major benefits can be realized when operating at these temperatures, includingincreased boiler efficiency, lime utilization, and trace metal and organic removal efficiency.

Potential operating problems can occur when handling the reaction products that contain cal-cium chloride (CaCl2). This material is hygroscopic, and can cause caked deposits on reactorwalls, bag plugging or blinding problems in the baghouse, and/or caking and plugging prob-lems in the fly ash removal equipment. The spray dryer and fabric filter must be operatedwithin the above specified design temperature limits, be well insulated, and be designed tominimize air inleakage to prevent these potential problems from occurring.

A fabric filter is used downstream from the spray dryer to collect reactant products, unreactedsorbent, and fly ash. Fabric filters applied to incinerators often use woven fiberglass bags toremove particulate matter from the flue gas stream. Fabric filters can act as secondary acid gascollectors because the dust cake that builds on the bags contains some unreacted sorbent thatprovides a surface to neutralize some of the acid gases passing through the cake. Many recentfabric filter designs applied to municipal waste incinerators use pulse-jet cleaning and haveeasily achieved the NSPS of 0.015 gr/dscf corrected to 7% O2 (Pompelia and Beachler 1991).Use of fabric filters on municipal waste incinerators is also effective in removing heavy metalsand organics (Brna and Kilgroe 1990).

Performance of this equipment has been studied in depth since the mid 1980s in support ofrevising the NSPS for Municipal Waste Combustors (58 FR 5488). Typically, use of a spraydryer followed by a fabric filter has shown to remove 75 to 85% of SO2 and 90 to 95% of HCl.Higher removal efficiencies have been achieved when calculating removal efficiencies overlong term time periods (i.e. long term averages) (EPA 1989; Beachler and Joseph 1992).


2.0-3/95 7-11

Other Fabric Filter Applications

Examples of typical baghouse installations are given in Table 7-5. This table lists the industry,exhaust gas temperature, dust concentration, baghouse cleaning method, fabrics, and air-to-cloth ratios. This list is by no means inclusive of the industries using baghouses for controllingparticulate emissions. Typical air-to-cloth ratios of shaker, reverse-air, and pulse-jet baghousesfor various industries are also given in Table 5-2.

Table 7-5. Typical baghouse installations

Industry

Process dustconcentration

(gr/ft3)Baghouse

cleaning method FabricsTemperature

(°F)Air-to-cloth

ratio (cfm/ft2)Aluminum

furnacesscrap conveyor

6 to 20 ShakerPulse-jet

Nomex, OrlonPolyester

250 to 375100

2 to 2.5:17 to 8:1

Asphalt batch plants Pulse-jet Nomex 250 4 to 6:1

Coal-fired boilers(1.5% sulfur coal)

Reverse-airPulse-jet

GlassFelt/Glass

350 to 450300 to 450

2:12 to 5:1

Coal processingpulverizing milldryerroller millcrusher

Pulse-jetPulse-jetPulse-jetPulse-jet

Nomex feltNomex feltPolyester feltPolypropylene felt

240400225100

4 to 6:15 to 7:16:17 to 8:1

Carbon black Reverse-air Glass-Teflon(treated) or Teflon

1.5:1

Cementclinker coolercrusher venting

kiln 10 to 12

Pulse-jetReverse-air and

shakeReverse-air

Nomex feltPolyester felt, Gore-tex

Glass 400 to 500

5:15:1

2:1

Claycalcining kiln or dryers 25 Pulse-jet Glass felt, Nomex 300 to 400 6:1

Copper smelter < 2 Shaker Dacron, Teflon 130

Cupola furnace (gray iron) 1 to 2 Reverse-air shaker Glass-Teflontreated Nomex

550 1.9:1

Chemicalpolyvinyl chloride (PVC)spray dryer

Reverse-air Acrylic, Gore-tex 350-425 2 to 3.6:1

Foodsugar storage bin Pulse-jet Polyester,

Gore-tex10:1

Ferro alloy plantsilicon metalelectric arc furnace

< 1.0 Reverse-air withshaker assist

Nomex 350

Foundrysand castingoperation

5 to 10 Pulse-jet Polyester felt 275 6 to 7:1

Glass meltingfurnaces

Reverse-airReverse-air and

shake

GlassNomex

400 to 500375 to 400

< 2:1

Gypsum building materials Pulse-jet Nomex

Lead smelting (battery lead) Pulse-jet Nomex, Teflon 320 to 325

Lime calcining Pulse-jet Nomex 280

Metalslead oxide processing Shaker Dacron, Gore-tex 1.5 to 3:1

Municipal incinerators 0.5 to5.0

Reverse-airPulse-jet

GlassGlass, Teflon

300300

2:12 to 3:1

Steelelectric arc furnacecanopy hood over steelfurnace

0.1 to 0.50.1 to 0.51.0 or less

ShakerReverse-airPulse-jet

DacronDacronPolyester felt

275125 to 250250 8:1

Secondary copper and brassrotary kiln

Shaker Nomex 350

Woodworkingfurniture manufacturing Pulse-jet Polyester 10:1

Zinc refiningcoker (zinc oxide) Pulse-jet Glass felt, Nomex 350 to 450 4 to 6:1

Lesson 7

7-12 2.0-3/95

Capital and Operating Cost Estimation

This section contains generalized cost data for baghouse systems described throughout thismanual. These data should be used only as an estimate to determine systems costs. In somecases the cost of the control device may represent only a very small portion (< 20%) of thetotal installed cost; in other cases it may represent the total cost. Variations in the total cost canbe attributed to a number of variable factors such as cost of auxiliary equipment, new or retro-fitted installation, local labor costs, engineering overhead, location and accessibility of plantsite, and installation procedure (factory or field assembled).

These cost estimation data are from an EPA publication, OAQPS Cost Control Manual, (EPA1990). Refer to this publication for additional information concerning this subject. These esti-mations represent equipment costs based on a reference date of third quarter, 1986.

Total Capital Costs

Total capital costs include costs for the baghouse structure, the initial complement of bags,auxiliary equipment, and the usual direct and indirect costs associated with installing orerecting new structures. These costs are described below, and may be escalated if desired.See EPA's OAQPS Cost Control Manual (EPA 1990) for escalation techniques.

Structure CostA guide to estimate the structural costs of six types of bare fabric filter systems (EPA1990), is provided in Table 7-6.

Table 7-6 associates a figure (found later in this lesson) with each of the six types offabric filters listed. Each figure consists of a graph that plots the following three struc-tural costs as a function of gross cloth area:

1. Cost of the filter structure (without bags)

2. Additional cost for 304 stainless steel construction

3. Additional cost for insulation

Extrapolation of these lines is not recommended. All units include unit and exhaustmanifolds, supports, platforms, handrails, and hopper discharge devices. The indi-cated prices are flange-to-flange. Note that the scales on axes differ.

Table 7-6. Guide to estimate costs of bare fabric filter systems

Operation Cleaning Mechanism Figure

Preassembled UnitsIntermittentContinuousContinuousContinuousContinuous

ShakerShakerPulse-jet (common housing)Pulse-jet (modular)Reverse-air

7-37-47-57-67-7

Field-assembled unitsContinuous Any method 7-8


2.0-3/95 7-13

The 304 stainless steel add-on cost is used when stainless steel is necessary to preventthe exhaust stream from corroding the interior of the baghouse. Stainless steel is sub-stituted for all metal surfaces that are in contact with the exhaust gas stream.

Insulation costs are for 3 inches of shop-installed glass fiber encased in a metal skin.One exception is the custom baghouse, which has field-installed insulation. Costs forinsulation include only the flange-to-flange baghouse structure on the outside of allareas in contact with the exhaust gas stream. Insulation for ductwork, fan casings, andstacks must be calculated separately.

The costs for intermittent service, mechanical shaker baghouses (including the shakermechanism) as a function of gross cloth area are presented in Figure 7-3. Becauseintermittent service baghouses do not require an extra compartment for cleaning, grossand net fabric areas are the same.

The same costs for a continuously operated baghouse cleaned by mechanical shaker asa function of the gross cloth area are presented in Figure 7-4. As in Figure 7-3, theunits are modular in construction. Costs for these units, on a square foot basis, arehigher because of increased complexity and generally heavier construction.

Costs of common-housing pulse-jet units and modular pulse-jet units are presented inFigures 7-5 and 7-6. Modular units are constructed of separate modules that may bearranged for off-line cleaning, while common-housing units have all bags within onehousing. The cleaning system compressor is not included. Because the common hous-ing is relatively inexpensive, the stainless steel add-on is proportionately higher thanfor modular units. Added material costs and set-up and labor charges associated withthe less workable stainless steel account for most of the added expense.

The costs for the reverse-air baghouses are shown in Figure 7-7. The construction ismodular and the reverse-air fan is included. Costs for custom baghouses which mustbe field assembled because of their large size are given in Figure 7-8. These unitsoften are used on power plants, steel mills, or other applications too large for the fac-tory-assembled baghouses.

Figure 7-3. Structure costs for intermittent shaker filters

Lesson 7

7-14 2.0-3/95

Figure 7-4. Structure costs for continuous shaker filters

Figure 7-5. Structure costs for pulse-jet filters (common housing)

Figure 7-6. Structure costs for pulse-jet filters (modular)


2.0-3/95 7-15

Figure 7-7. Structure costs for reverse-air filters

Figure 7-8. Structure costs for custom-built filters

Bag Costs/CB

The price per square foot (in 3rd quarter 1986 dollars) of bags by type of fabric andtype of cleaning system used is given in Table 7-7. The prices represent about a 10percent range. In calculating the cost, use the gross area as determined from Table 7-8.Gore-Tex fabric costs are a combination of the base fabric cost and a premium for thePTFE laminate and its application. As fiber market conditions change, the costs offabrics relative to each other also change. The bag prices are based on typical fabricweights, in ounces/square yard, for the fabric being priced. Sewn-in snap rings areincluded in the price, but other mounting hardware, such as clamps or cages, is anadded cost. See the notes on Table 7-7 for these costs. EPA's OAQPS Cost ControlManual can be used to obtain additional information on the cost (EPA 1990).

Lesson 7

7-16 2.0-3/95

Table 7-7. Bag prices (3rd quarter 1986 $/ft2)

Type of Materials1

Type of Cleaning

BagDiameter(Inches)

PE PP NO HA FG CO TF

Pulse-jet, TR2 4-1/2 to 5-1/86 to 8

0.590.43

0.610.44

1.881.56

0.920.71

1.291.08

NANA

9.056.80

Pulse-jet, BBR 4-1/2 to 5-1/86 to 8

0.370.32

0.400.33

1.371.18

0.660.58

1.240.95

NANA

8.786.71

ShakerStrap topLoop top

55

0.450.43

0.480.45

1.281.17

0.750.66

NANA

0.440.39

NANA

Reverse-air with rings 8 0.46 NA 1.72 NA 0.99 NA NAReverse-air w/o rings3 8

11-1/20.320.32

NANA

1.201.16

NANA

0.690.53

NANA

NANA

NA = Not applicable.1. Materials:

PE = 16-oz polyesterPP = 16-oz polypropyleneNO = 14-oz NomexHA = 16-oz homopolymer acrylicFG = 16-oz fiberglass with 10% TeflonCO = 9-oz cottonTF = 22-oz Teflon felt

2. Bag removal methods:TR = Top bag removal (snap in)BBR= Bottom bag removal

3. Identified as reverse-air bags, but used in low pressure pulse applications.

Note: For pulse-jet baghouses, all bags are felts except for the fiberglass, which is woven. For bottom access pulse-jets, the mild steel cageprice for one cage can be calculated from the single-bag fabric area using:$ = 4.941 + 0.163 ft2 in 50 cage lots$ = 4.441 + 0.163 ft2 in 100 cage lots$ = 3.941 + 0.163 ft2 in 500 cage lots

These costs apply to 4-1/2-in. or 5-5/8-in. diameter, 8-ft and 10-ft cages made of 11 gauge mild steel and having 10 vertical wires and "Roll Band"tops. For flanged tops, add $1 per cage. If flow control venturis are used (as they are in about half of the pulse-jet manufacturers' designs),add $5 per cage. For stainless steel cages use:$ = 23.335 + 0.280 ft2 in 50 cage lots$ = 21.791 + 0.263 ft2 in 100 cage lots$ = 20.564 + 0.248 ft2 in 500 cage lots

For shakers and reverse-air baghouses, all bags are woven. All prices are for finished bags, and prices can vary from one supplier to another.For Gore-Tex bag prices, multiply base fabric price by factors of 3 to 4.5.

Source: EPA 1990.

Table 7-8. Factors to obtain gross cloth area from net cloth area

Net Cloth Area, Anc (ft2)Factor to Obtain

Gross Cloth Area, Atc (ft2)

1 - 4,0004,001 - 12,000

12,001 - 24,00024,001 - 36,00036,001 - 48,00048,001 - 60,00060,001 - 72,00072,001 - 84,00084,001 - 96,00096,001 - 108,000

108,001 - 132,000132,001 - 180,000

Multiply by 2Multiply by 1.5

Multiply by 1.25Multiply by 1.17

Multiply by 1.125Multiply by 1.11Multiply by 1.10Multiply by 1.09Multiply by 1.08Multiply by 1.07Multiply by 1.06Multiply by 1.05

Source: EPA 1990.


2.0-3/95 7-17

Purchased Equipment Cost (PEC) and Total Capital Costs (TCC)The purchased equipment cost (PEC) of the fabric filter system is the sum of the costsof the baghouse, bags, auxiliary equipment, instruments and controls, taxes, andfreight costs. The factors necessary to estimate these costs are presented in Table 7-9.The factors necessary to estimate the remaining direct and indirect capital costs toobtain total capital costs are also provided in Table 7-9. EPA's OAQPS Cost ControlManual can be used to estimate the cost of auxiliary equipment (EPA 1990).

Table 7-9. Capital cost factors for fabric filters

Direct CostsPurchased Equipment Costs:

Fabric FilterBagsAuxiliary equipment

Instruments & controlsTaxesFreight

Purchased Equipment Cost, PEC

Installation Direct CostsFoundation & supportsErection & handlingElectricalPipingInsulation for ductwork1

Painting2


Total Direct Costs, DC

Indirect CostsEngineering and supervisionConstruction and field expenseConstruction feeStart-up feesPerformance testContingencies

Total Indirect Costs, IC

Total Capital Cost (TCC) = DC + IC

Factor

As estimatedAs estimatedAs estimated

EC = Sum of estimatedvalues

0.10 EC0.03 EC0.05 EC

PEC = 1.18 EC

0.04 PEC0.50 PEC0.08 PEC0.01 PEC0.07 PEC0.02 PEC

SP (as required)Bldg. (as required)




0.45 PEC

2.17 PEC + SP + Bldg.1. If ductwork dimensions have been established, cost may be established based on $10 to $12/ft2 of surface for

field application. Fan housings and stacks may also be insulated.2. The increased use of special coatings may increase this factor to 0.06 PEC or higher.Source: EPA 1990.

Lesson 7

7-18 2.0-3/95

Example Case

This problem will show you how to calculate the total capital cost of a baghouse using thefigures and tables in this lesson.

Problem:A facility is proposing to build a reverse-air baghouse that will operate with a net air-to-cloth ratio of 2.5:1 (ft3/min)/ft2 and an exhaust gas flow rate of 110,000 acfm.Eight-inch diameter fiberglass bags with Teflon backing and rings are proposed. Thestructure requires stainless steel add-on and insulation. Auxiliary equipment is esti-mated to cost $10,000. Calculate the total capital cost of the baghouse.

Solution:1. Calculate the total net cloth area using a variation of equation 3-6 (lesson 3).

Where: Anc = net cloth area, ft2

Q = process exhaust rate, acfmvf = filtration velocity, ft/min

Since the filtration velocity (vf) equals the air-to-cloth ratio:

2. Calculate the total gross cloth area. Use Table 7-8 to find the factor needed toconvert the total net cloth area to the total gross cloth area. For a net cloth area of44,000 ft2, the factor is 1.125.

Atc = Anc × 1.125= 44,000 ft2 × 1.125 = 49,500 ft2

3. Calculate the structure cost of the baghouse. Knowing that the total cloth areais 49,500 ft2 and using Figure 7-7 (structure costs for reverse-air filters), you cancalculate the structure cost as follows:

Base cost $ 380,000

Stainless steel add-on 270,000

Insulation add-on + 40,000

Structure cost $ 690,000

vfQ

Anc

--------=

AncQvf

----=

vf 2.5 ft3 min⁄( ) ft2⁄ 2.5 ft min⁄==

Anc110,000 acfm

2.5 ft min⁄--------------------------------- 44,000 ft2= =


2.0-3/95 7-19

4. Calculate the total bag cost (CB). From Table 7-7 (Bag Prices), the cost of fiber-

glass bags for a reverse-air baghouse with rings is $0.99/ft2.

Total bag cost = $ 0.99/ft2 × 49,500 ft2

= $ 49,0005. Calculate the total capital cost of the baghouse. Based on the above informa-

tion, the equipment cost (EC) can be calculated to be $749,000. See Table 7-10.

Use the factors given in Table 7-9 to calculate the following:

1. Purchased equipment costs (PEC)

2. Installation direct costs

3. Indirect costs

4. Total capital cost (TCC)

A summary of these costs is provided in Table 7-10.

Lesson 7

7-20 2.0-3/95

Table 7-10. Example case capital costs1

Direct CostsPurchased Equipment Costs:Fabric FilterBagsAuxiliary equipment

Instruments & controlsTaxesFreight

Purchased Equipment Cost, PEC

Installation Direct CostsFoundation & supportsErection & handlingElectricalPipingInsulation for ductwork1

Painting2


Total Direct Costs, DC

Indirect CostsEngineering and supervisionConstruction and field expenseConstruction feeStart-up feesPerformance testContingencies

Total Indirect Costs, IC

Total Capital Cost (TCC) = DC + IC

Factor

As estimatedAs estimatedAs estimated

EC = Sum of estimatedvalues

0.10 EC0.03 EC0.05 EC

PEC = 1.18 EC


SP (as required)Bldg. (as required)




0.45 PEC


Cost(s)

$690,00049,00010,000

$749,000

$74,90022,50037,500

$883,900

$35,400442,000

70,7008,840

61,90017,700

__

$636,540

$1,520,440

$88,400177,000

88,4008,8408,840

26,500

$397,980

$1,918,420

1. If ductwork dimensions have been established, cost may be established based on $10 to $12/ft2 of surface for field application. Fanhousings and stacks may also be insulated.

2. The increased use of special coatings may increase this factor to 0.06 PEC or higher.


2.0-3/95 7-21

Review Exercise

1. True or False? Fabric filters cannot be used for the collection of fly ash from coal-fired boilerssince the flue gas deteriorates the bags.

2. For fabric filters used on coal-fired boilers, the bags are usually made of:

a. Cottonb. Glassc. Wool

3. One technology for reducing both SO2 gas and particulate emissions involves the injection of a(an)___________________ slurry in a spray ____________________ with dry particle collection in abaghouse.

4. True or False? Fabric filters preceded by spray dryers are commonly applied to municipal wasteincinerators.

5. In a spray dryer, moisture is ____________________ from the wet alkaline sprays, leaving a(an)____________________ powdered product.

6. Which one of the following materials is hygroscopic and can cause bag plugging or blinding prob-lems?

a. Calcium carbonateb. Calcium chloridec. Calcium sulfate

7. True or False? Dry FGD systems using lime injected in a spray dryer and a baghouse for dry parti-cle collection are capable of 70% SO2 reduction and 99+% particulate matter removal efficiency.

8. In dry sulfur dioxide control systems for coal-fired boilers using a spray dryer, the most commonalkaline absorbents used are:

a. Sodium citrate and magnesium oxideb. Sodium carbonate and limec. Sodium bisulfate and sodium hydroxide

9. Fabric filters with bags made of woven glass usually have air-to-cloth ratios:

a. Greater than 6:1b. Approximately 7.5:1c. Less than 4:1

10. True or False? Pulse-jet fabric filters with polyester felt bags cannot be used to collect iron oxidedusts from steel furnaces.

11. True or False? Fabric filters have been used for filtering dust-laden gas from cement kilns, clinkercoolers, and crushing operations.

Lesson 7

7-22 2.0-3/95

Review Answers

1. FalseFabric filters can be used to collect fly ash from coal-fired boilers.

2. b. GlassFor fabric filters used on coal-fired boilers, the bags are usually made of glass.

3. AlkalineDryerOne technology for reducing both SO2 gas and particulate emissions involves the injection of analkaline slurry in a spray dryer with dry particle collection in a baghouse.

4. TrueFabric filters preceded by spray dryers are commonly applied to municipal waste incinerators.

5. EvaporatedDryIn a spray dryer, moisture is evaporated from the wet alkaline sprays, leaving a dry powderedproduct.

6. b. Calcium chlorideCalcium chloride is hygroscopic and can cause bag plugging or blinding problems.

7. TrueDry FGD systems using lime injected in a spray dryer and a baghouse for dry particle collectionare capable of 70% SO2 reduction and 99+% particulate matter removal efficiency.

8. b. Sodium carbonate and limeIn dry sulfur dioxide control systems for coal-fired boilers using a spray dryer, the most commonalkaline absorbents used are sodium carbonate and lime.

9. c. Less than 4:1Fabric filters with bags made of woven glass usually have air-to-cloth ratios less than 4:1.

10. FalsePulse-jet fabric filters with polyester felt bags can be used to collect iron oxide dusts from steelfurnaces (see Table 7-5).

11. TrueFabric filters have been used for filtering dust-laden gas from cement kilns, clinker coolers, andcrushing operations (see Table 7-5).

2.0-3/95 7-23

Bibliography

Beachler, D. S. and G. P. Greiner. 1989, April. Design considerations and selection of an emissioncontrol system operating at low temperatures for a MSW combustion facility. Paper presented atInternational Conference on Municipal Waste Combustion. Hollywood, FL.

Beachler, D. S. and G. T. Joseph. 1992, November. Air emission test results from two operating waste-to-energy facilities. Paper presented at the Air and Waste Management Association SpecialtyConference: Environmental Aspects of Cogeneration. Pittsburgh, PA.

Belba, V. H., W. T. Grubb, and R. L. Chang. 1992. The potential of pulse-jet baghouse for utilityboilers. Part 1: A world-wide survey of users. Journal of the Air and Waste ManagementAssociation. 42(2):209-218.

Brna, T. G. and J. D. Kilgroe. 1990. The impact of particulate emissions control on the control of otherMWC air emissions. Journal of the Air and Waste Management Association. 40(9).

Cushing, K. M., R. L. Merritt, and R. L. Chang. 1990. Operating history and current status of fabricfilters in the utility industry. Journal of the Air and Waste Management Association.40(7):1051-1058.



Kaplan, S. M. and K. Felsvang. 1979, April. Spray dryer absorption of SO2 from industrial boiler fluegas. Paper presented at the 86th National Meeting of the American Institute of ChemicalEngineers. Houston, TX.

Neveril, R. B., J. U. Price, and K. L. Engdahl. 1978. Capital and operating costs of selected airpollution control systems. Journal of the Air Pollution Control Association. 28:829-836.

Pompelia, D. M. and D. S. Beachler. 1991, January. Designing and operating a waste-to-energy facilityto comply with federal particulate matter requirements. Paper presented at the 6th Annual Waste-to-Energy Symposium of the Governmental Refuse Collection and Disposal Association/SolidWaste Association of North America. Arlington, VA.

U.S. Environmental Protection Agency. 1973. Air Pollution Engineering Manual. 2nd ed. AP-40.


7-24 2.0-3/95

Bibliography


U.S. Environmental Protection Agency. 1980, February. Survey of Dry SO2 Control Systems.EPA 600/7-80-030.

U.S. Environmental Protection Agency. 1989. Municipal Waste Combustors - Background Informationfor Proposed Standards: Post-Combustion Technology Performance. EPA 450/3-89-27c.

U.S. Environmental Protection Agency. 1990, January. OAQPS Control Cost Manual, 4th ed.EPA 450/3-90-006.

Scrubber Systems Operation Review

Self-Instructional Manual APTI Course SI:412C Second Edition

__________________________________________________________________________ Industrial Extension Service College of Engineering North Carolina State University

Scrubber Systems Operation Review

Self-Instructional Manual APTI Course SI:412C Second Edition

Authors Gerald T. Joseph, P.E., DMG Environmental, Inc. David S. Beachler, DMG Environmental, Inc.

Instructional Designer (First Edition) Marilyn M. Peterson, Northrop Services, Inc.

Instructional Designer (Second Edition) Nancy Tusa, North Carolina State University

__________________________________________________________________________ Developed by North Carolina State University EPA Cooperative Assistance Agreement CT-902765

This project has been funded wholly or in part by the United States Environmental Protection Agency under Cooperative Assistance Agreement CT-902765 to North Carolina State University. The contents of this document do not necessarily reflect the views and policies of the Environmental Protection Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use.

1998 North Carolina State University

All rights reserved, including the right of reproduction in whole or in part in any form.

Printed on recycled paper in the United States of America

North Carolina State University

Raleigh, NC 27695

2.0-7/98 iii

Contents Figures..........................................................................................................................................ix

Tables............................................................................................................................................xi

Notation......................................................................................................................................xiii

English Symbols................................................................................................................xiii Greek Symbols...................................................................................................................xv Units of Measurement........................................................................................................xv Acronyms ......................................................................................................................... xvii

Course Description Objectives .........................................................................................................................xix Audience ............................................................................................................................xx Course Length and CEUs ..................................................................................................xx Suggested Prerequisites....................................................................................................xx Required Materials .............................................................................................................xx Supplemental Materials .....................................................................................................xx Taking the Course.............................................................................................................xxi Completing the Course .....................................................................................................xxi

Lesson 1

Introduction to Scrubbing Systems ...................................................................................1-1 Introduction ......................................................................................................................1-1 Wet Scrubbers .................................................................................................................1-1 Wet Scrubber Systems ....................................................................................................1-5 Categorization of Wet Scrubbers .....................................................................................1-6 Dry Scrubbing Systems ...................................................................................................1-7 Design Evaluation ............................................................................................................1-9 Summary........................................................................................................................1-10 Review Exercise.............................................................................................................1-13 Review Answers.............................................................................................................1-15

Bibliography........................................................................................................................1-17

APTI Course SI:412C ___________________________________________________________________________________

iv 2.0-7/98

Lesson 2

Operating Principles of Scrubbers .....................................................................................2-1 Introduction ......................................................................................................................2-1 Particle Collection ............................................................................................................2-2

Impaction ...................................................................................................................2-3 Diffusion.....................................................................................................................2-4 Other Collection Mechanisms ...................................................................................2-5

Gas Collection..................................................................................................................2-6 Acid Gas Removal mechanisms ......................................................................................2-7 Pressure Drop ..................................................................................................................2-8 Liquid-To-Gas Ratio.........................................................................................................2-9 Summary........................................................................................................................2-11 Review Exercise.............................................................................................................2-13 Review Exercise Answers..............................................................................................2-15

Bibliography........................................................................................................................2-17

Lesson 3

Gas-Phase Contacting Scrubbers ......................................................................................3-1 Introduction ......................................................................................................................3-1 Venturi Scrubbers ............................................................................................................3-2

Particle Collection......................................................................................................3-9 Gas Collection .........................................................................................................3-10 Summary .................................................................................................................3-13

Plate Towers ..................................................................................................................3-16 Particle Collection....................................................................................................3-19 Gas Collection .........................................................................................................3-20 Maintenance Problems............................................................................................3-20

Orifice Scrubbers ...........................................................................................................3-23 Particle Collection....................................................................................................3-25 Gas Collection .........................................................................................................3-25 Maintenance Problems............................................................................................3-25 Summary .................................................................................................................3-26

Review Exercise.............................................................................................................3-27 Review Exercise Answers..............................................................................................3-33

Bibliography........................................................................................................................3-37

Lesson 4

Liquid-Phase Contacting Scrubbers ..................................................................................4-1 Introduction ......................................................................................................................4-1 Spray Towers ...................................................................................................................4-1

Particle Collection......................................................................................................4-3 Gas Collection ...........................................................................................................4-4 Summary ...................................................................................................................4-4

Scrubber Systems Operation Review ___________________________________________________________________________________

2.0-7/98 v

Ejector Venturis................................................................................................................4-6 Particle Collection......................................................................................................4-7 Gas Collection ...........................................................................................................4-7 Maintenance Problems..............................................................................................4-7 Summary ...................................................................................................................4-7

Review Exercise...............................................................................................................4-9 Review Exercise Answers..............................................................................................4-11

Bibliography........................................................................................................................4-13

Lesson 5

Wet-Film (Packed Tower)Scrubbers ...................................................................................5-1 Introduction ......................................................................................................................5-1 Gas Collection..................................................................................................................5-2 Tower Designs .................................................................................................................5-2 Packing Material...............................................................................................................5-6 Exhaust Gas Distribution .................................................................................................5-7 Liquid Distribution.............................................................................................................5-7 Maintenance Problems ....................................................................................................5-9 Summary........................................................................................................................5-11 Review Exercise.............................................................................................................5-15 Review Exercise Answers..............................................................................................5-17

Bibliography........................................................................................................................5-19

Lesson 6

Combination Devices – Liquid-Phase and Gas-Phase Contacting Scrubbers ..............6-1 Introduction ......................................................................................................................6-1 Cyclonic Spray Scrubbers................................................................................................6-2

Particle Collection......................................................................................................6-3 Gas Collection ...........................................................................................................6-4 Maintenance Problems..............................................................................................6-4 Summary ...................................................................................................................6-4

Mobile-Bed Scrubbers......................................................................................................6-5 Particle Collection......................................................................................................6-6 Gas Collection ...........................................................................................................6-7 Maintenance Problems..............................................................................................6-7 Summary ...................................................................................................................6-8

Baffle Spray Scrubbers ....................................................................................................6-8 Particle Collection......................................................................................................6-9 Gas Collection ...........................................................................................................6-9 Summary .................................................................................................................6-10

Mechanically Aided Scrubbers.......................................................................................6-10 Particle Collection....................................................................................................6-12 Gas Collection .........................................................................................................6-12 Maintenance Problems............................................................................................6-12

APTI Course SI:412C ___________________________________________________________________________________

vi 2.0-7/98

Summary .................................................................................................................6-13 Review Exercise.............................................................................................................6-15 Review Exercise Answers..............................................................................................6-19

Bibliography........................................................................................................................6-23

Lesson 7

Dry Scrubbing Systems .......................................................................................................7-1 Introduction ......................................................................................................................7-1 Gas Removal Mechanisms ..............................................................................................7-2 Stoichiometry ...................................................................................................................7-3 Dry Injection .....................................................................................................................7-5 Spray Dryer Systems .......................................................................................................7-8

Operating and Design Parameters..........................................................................7-10 Spray Drying Equipment..........................................................................................7-13

Atomizers ..........................................................................................................7-13 Spray-Dryer Chamber.......................................................................................7-16

Particulate Matter Collection ..........................................................................................7-18 Maintenance Problems ..................................................................................................7-18 Summary........................................................................................................................7-19 Review Exercise.............................................................................................................7-21 Review Exercise Answers..............................................................................................7-25

Bibliography........................................................................................................................7-27

Lesson 8

Equipment Associated with Scrubbing Systems..............................................................8-1 Introduction ......................................................................................................................8-1 Transport Equipment for Exhaust Gases and Scrubbing Liquids....................................8-2

Fans...........................................................................................................................8-2 Ducts..........................................................................................................................8-3 Pumps........................................................................................................................8-4 Pipes..........................................................................................................................8-4 Quenchers .................................................................................................................8-6 Spray Nozzles ...........................................................................................................8-6 Entrainment Separators.............................................................................................8-9

Construction Materials ...................................................................................................8-13 Monitoring Equipment ....................................................................................................8-15

Pressure Drop .........................................................................................................8-16 Temperature ............................................................................................................8-18 Liquid Flow Monitors................................................................................................8-18 pH Monitors .............................................................................................................8-18

Recordkeeping ...............................................................................................................8-19 Summary........................................................................................................................8-20 Review Exercise.............................................................................................................8-23 Review Exercise Answers..............................................................................................8-27

Bibliography........................................................................................................................8-31


2.0-7/98 vii

Lesson 9

Flue Gas Desulfurization (Acid Gas Removal) Systems ..................................................9-1 Introduction ......................................................................................................................9-1 Nonregenerable FGD Processes.....................................................................................9-6

Lime Scrubbing..........................................................................................................9-6 Process Chemistry..............................................................................................9-6 System Description .............................................................................................9-7 Operating Experience .......................................................................................9-11

Limestone Scrubbing...............................................................................................9-12 Process Chemistry............................................................................................9-12 System Description ...........................................................................................9-12 Operating Experience .......................................................................................9-16

Dual-Alkali Scrubbing ..............................................................................................9-17 Process Chemistry............................................................................................9-17 System Description ...........................................................................................9-18 Operating Experience .......................................................................................9-19

Sodium-Based Once-Through Scrubbing ...............................................................9-21 Process Chemistry............................................................................................9-21 System Description ...........................................................................................9-22 Operating Experience .......................................................................................9-23

Regenerable FGD Processes ........................................................................................9-26 Emerging Technologies .................................................................................................9-26 Summary........................................................................................................................9-29 Review Exercise.............................................................................................................9-31 Review Exercise Answers..............................................................................................9-37 Bibliography .................................................................................................................9-41

Lesson 10

Design Evaluation of Particulate Wet Scrubbing Systems ............................................10-1 Introduction ....................................................................................................................10-1 Particulate Scrubber Design Factors .............................................................................10-2 Estimating Collection Efficiency and Pressure Drop......................................................10-4

Collection Efficiency ................................................................................................10-4 The Infinite Throat Model for Estimating Ventrui Scrubber Efficiency ..............10-5

Example 10-1 ...........................................................................................10-11 Pressure Drop .......................................................................................................10-23

Using Pilot Methods to Design Scrubbers ...................................................................10-24 Summary......................................................................................................................10-25 Review Exercise...........................................................................................................10-27 Review Exercise Answers............................................................................................10-37 Bibliography ...............................................................................................................10-47

APTI Course SI:412C ___________________________________________________________________________________

viii 2.0-7/98

Lesson 11

Design Review of Absorbers Used for Gaseous Pollutants ..........................................11-1 Introduction ....................................................................................................................11-1 Review of Design Criteria ..............................................................................................11-2 Absorption ......................................................................................................................11-3 Solubility.........................................................................................................................11-4

Example 11-1 ..........................................................................................................11-7 Absorber Design ..........................................................................................................11-10

Theory....................................................................................................................11-10 Mass-Transfer Models ....................................................................................11-14

Procedures ............................................................................................................11-15 Material Balance .............................................................................................11-16 Determining the Liquid Requirement ..............................................................11-19 Example 11-2..................................................................................................11-22

Sizing a Packed Tower ................................................................................................11-25 Packed Tower Diameter ........................................................................................11-25

Example 11-3..................................................................................................11-29 Packed Tower Height ............................................................................................11-33

Example 11-4..................................................................................................11-38 Sizing a Plate Tower ....................................................................................................11-39

Plate Tower Diameter............................................................................................11-39 Example 11-5..................................................................................................11-41

Number of Theoretical Plates................................................................................11-43 Example 11-6..................................................................................................11-44

Summary......................................................................................................................11-46 Review Exercise...........................................................................................................11-47 Review Exercise Answers............................................................................................11-63 Bibliography ...............................................................................................................11-77

Appendix A

Mole Fraction and Part Per Million (ppm) ...........................................................................A1


2.0-7/98 ix

Figures

Figure 1-1. An example of a venturi scrubber design ...........................................................1-2 Figure 1-2. An example of a tower scrubber design .............................................................1-3 Figure 1-3. An example of a wet scrubbing system ..............................................................1-6 Figure 1-4. Dry sorbent injection scrubber system................................................................1-8 Figure 1-5. Spray dryer absorber system..............................................................................1-9 Figure 2-1. Impaction.............................................................................................................2-3 Figure 2-2. Diffusion ..............................................................................................................2-4 Figure 2-3. Hypothetical curve illustrating relationship between particle size and collection

efficiency for typical wet scrubber .......................................................................2-5 Figure 2-4. Absorption ...........................................................................................................2-6 Figure 2-5. Measuring pressure drop across a venturi scrubber ..........................................2-8 Figure 3-1. Venturi configuration ...........................................................................................3-2 Figure 3-2. Venturi scrubber with a wetted throat .................................................................3-3 Figure 3-3. Venturi with throat sprays ...................................................................................3-4 Figure 3-4. Spray venturi with rectangular throat ..................................................................3-5 Figure 3-5. Adjustable-throat venturi with plunger ................................................................3-6 Figure 3-6. Adjustable-throat venturi with movable plate ......................................................3-7 Figure 3-7. Venturi-rod scrubber ...........................................................................................3-8 Figure 3-8. Flooded elbow leading into cyclonic separator ...................................................3-9 Figure 3-9. Plate tower ........................................................................................................3-16 Figure 3-10. Sieve plate ........................................................................................................3-17 Figure 3-11. Impingement plate ............................................................................................3-18 Figure 3-12. Bubble-cap plate ...............................................................................................3-18 Figure 3-13. Valve plate ........................................................................................................3-19 Figure 3-14. Detail of orifice action........................................................................................3-23 Figure 3-15. Self-induced spray orifice scrubber...................................................................3-24 Figure 4-1. Countercurrent-flow spray tower.........................................................................4-2 Figure 4-2. Crosscurrent-flow spray tower ............................................................................4-3 Figure 4-3. Ejector venturi scrubber ......................................................................................4-6 Figure 5-1. Countercurrent-flow packed tower......................................................................5-3 Figure 5-2. Cocurrent-flow packed tower ..............................................................................5-4 Figure 5-3. Crossflow packed tower......................................................................................5-5 Figure 5-4. Three-bed crossflow packed tower .....................................................................5-5 Figure 5-5. Fiber-bed scrubber..............................................................................................5-6 Figure 5-6. Common packing materials ................................................................................5-6 Figure 5-7. Two types of liquid distributors (trough and weir, and perforated tube)..............5-8 Figure 6-1. Irrigated cyclone scrubber...................................................................................6-2 Figure 6-2. Cyclonic spray scrubber......................................................................................6-3 Figure 6-3. Flooded-bed scrubber.........................................................................................6-5 Figure 6-4. Fluidized-bed scrubber .......................................................................................6-6

APTI Course SI:412C ___________________________________________________________________________________

x 2.0-7/98

Figure 6-5. Baffle spray scrubber ..........................................................................................6-9 Figure 6-6. Centrifugal-fan scrubber ...................................................................................6-11 Figure 6-7. Mechanically induced spray scrubber...............................................................6-11 Figure 7-1. Components of a dry injection system................................................................7-5 Figure 7-2. Spray dryer absorber ..........................................................................................7-8 Figure 7-3. Components of a spray dryer absorber system..................................................7-9 Figure 7-4. Example of rotary atomizer used in spray-dryer FGD systems ........................7-13 Figure 7-5. Two-fluid nozzle atomizer (nozzle body) ..........................................................7-14 Figure 7-6. Two-fluid nozzle atomizer (high pressure air stream).......................................7-15 Figure 7-7. Two types of spray-dryer chambers (rotary-atomizer and two-fluid pneumatic

nozzle)...............................................................................................................7-17 Figure 8-1. Four types of centrifugal fans..............................................................................8-2 Figure 8-2. Impingement nozzle............................................................................................8-7 Figure 8-3. Solid cone nozzle................................................................................................8-7 Figure 8-4. Helical spray nozzle ............................................................................................8-8 Figure 8-5. Cyclonic separator ............................................................................................8-10 Figure 8-6. Mesh-pad separator..........................................................................................8-11 Figure 8-7. Two types of blade separators (chevron and impingement).............................8-11 Figure 8-8. Two methods for measuring static pressure (copper tube and pitot tube) .......8-17 Figure 9-1. Typical process flow for a lime or limestone FGD system..................................9-8 Figure 9-2. Typical process flow for a double-alkali FGD system.......................................9-19 Figure 9-3. Typical process flow for a sodium-based throwaway (single-alkali) FGD system ......................................................................................................9-22 Figure 10-1. Overall penetration, Pt, versus B with Kpg as a parameter, with different

geometric standard deviations σgm .................................................................10-10 Figure 10-2. Overview of steps for completing Example 10-1 ............................................10-12 Figure 10-3. Overall penetration, Pt, for Example 10-1, where the standard deviation σgm is equal to 2.5............................................................................10-12 Figure 11-1. Equilibrium lines for SO2 - H2O systems at various temperatures..................10-16 Figure 11-2. Equilibrium diagram for SO2 - H2O system for the data given in Example 11-1....................................................................................................11-5 Figure 11-3. Visualization of two-film theory .........................................................................11-6 Figure 11-4. Resistance to motion encountered by a molecule being absorbed................11-10 Figure 11-5. Comparison of overall absorption coefficient for SO2 in water .......................11-12 Figure 11-6. Material balance for countercurrent-flow absorber .........................................11-15 Figure 11-7. Typical operating line diagram........................................................................11-16 Figure 11-8. Graphic determination of liquid flow rate ........................................................11-19 Figure 11-9. Material balance for Example 11-2 .................................................................11-21 Figure 11-10. Graphical solution to Example 11-2................................................................11-22 Figure 11-11. Generalized flooding and pressure drop correlation.......................................11-26 Figure 11-12. Generalized flooding and pressure drop correlation for Example 11-3 ..........11-30 Figure 11-13. Generalized flooding and pressure drop correlation for Example 11-3 ..........11-33 Figure 11-14. Colburn diagram .............................................................................................11-36


2.0-7/98 xi

Figure 11-15. Column packing comparison for ammonia and water system........................11-37 Figure 11-16. Tray spacing correction factor.........................................................................11-41 Figure 11-17. Tray spacing correction factor for Example 11-5 ............................................11-42 Figure 11-18. Graphic determination of the number of theoretical plates.............................11-43

Tables

Table 1-1. Relative advantages and disadvantages of wet scrubbers compared to other control devices ....................................................................................................1-4

Table 1-2. Categories of wet collectors by energy source used for contact ........................1-7 Table 2-1. Particle collection mechanisms for wet scrubbing systems................................2-3 Table 2-2. Scrubbing systems used on utility boilers.........................................................2-10 Table 3-1. Operational problems associated with venturi scrubbers .................................3-11 Table 3-2. Operating characteristics of venturi scrubbers .................................................3-14 Table 3-3. Performance data of typical venturi scrubbers .................................................3-14 Table 3-4. Operational problems associated with plate towers .........................................3-21 Table 3-5. Operating characteristics of plate towers..........................................................3-22 Table 3-6. Operating characteristics of orifice scrubbers...................................................3-26 Table 4-1. Operating characteristics of spray towers...........................................................4-5 Table 4-2. Operating characteristics of ejector venturis.......................................................4-8 Table 5-1. Liquid distributors for packed towers ..................................................................5-9 Table 5-2. Operating problems associated with packed towers ........................................5-10 Table 5-3. Operating characteristics of wet-film scrubbers................................................5-12 Table 6-1. Operating characteristics of cyclonic scrubbers .................................................6-4 Table 6-2. Operating characteristics of mobile-bed scrubbers ............................................6-8 Table 6-3. Operating characteristics of baffle spray scrubbers..........................................6-10 Table 6-4. Operating characteristics of mechanically aided scrubbers .............................6-13 Table 7-1. Examples of dry injection systems on medical and municipal waste incinerators..........................................................................................................7-7 Table 7-2. Summary of spray-dryer applications ...............................................................7-11 Table 8-1. Pipe materials for scrubber systems advantages and disadvantages.............8-5 Table 8-2. Typical operational characteristics of entrainment separators .........................8-12 Table 8-3. Construction materials for wet scrubber components.......................................8-13 Table 8-4. Monitoring equipment for wet scrubbing systems.............................................8-16 Table 8-5. Scrubber operation data ...................................................................................8-19 Table 9-1. Summary of FGD systems by process (percentage of total MW) ......................9-3 Table 9-2. Operational data for lime FGD systems on utility boilers....................................9-9 Table 9-3. Operational data for limestone FGD systems on utility boilers.........................9-13 Table 9-4. Operational data for double-alkali FGD systems on utility and industrial boilers................................................................................................................9-20 Table 9-5. Operational data for sodium-based once-through FGD systems on utility and industrial boilers................................................................................9-24 Table 9-6. SO2 and SO2/NOx control technologies for coal-fired boilers ...........................9-27 Table 10-1. Ranges of pressure drops and liquid-to-gas (L/G) ratios for various wet scrubbers ..........................................................................................................10-3 Table 10-2. Parameters α and β for the contact power theory ..........................................10-20

APTI Course SI:412C ___________________________________________________________________________________

xii 2.0-7/98

Table 10-3. Methods for predicting venturi scrubber pressure requirements ....................10-22 Table 11-1. Partial pressure of SO2 in aqueous solution, mm Hg .......................................11-4 Table 11-2. Henry’s law constants for gases in H2O............................................................11-7 Table 11-3. Equilibrium data ................................................................................................11-7 Table 11-4. Equilibrium data for Example 11-1....................................................................11-9 Table 11-5. Packing data ...................................................................................................11-28 Table 11-6. Empirical constants for Equation 11-26 ..........................................................11-40

2.0-7/98 xiii

Notation English Symbols

Symbol Definition

A - Cross-sectional area of Column A

a - Interfacial contact area

B - Parameter characterizing the liquid-to-gas ratio

CD - Drag coefficient for the liquid at the throat entrance

Cc - Cunningham slip correction factor *Ac - Equilibrium concentration of solute A at operating conditions

cAI - Concentration of solute A at the interface

cAL - Concentration of solute A in the liquid

dd - Droplet diameter

dp - Particle aerodynamic resistance diameter

dpg - Particle aerodynamic geometric mean diameter

dps - Particle physical, or Stokes, diameter

dt - Diameter of column

F - Packing factor

f - The percent of flooding velocity

G′ - Gas mass flow rate per cross-sectional area of tower

Gm - Gas molar flow rate

gc - Gravitational constant

H - Henry’s law constant, Pa/mole fraction

H′ - Henry’s law constant, mole fraction in vapor per mole fraction of liquid

HOG - Height of a transfer unit based on an overall gas-film coefficient, m

HOL - Height of a transfer unit based on an overall liquid-film coefficient, m

HTU - Height of a transfer unit

APTI Course SI:412C ___________________________________________________________________________________

xiv 2.0-7/98

KOG - Overall mass-transfer coefficient based on gas phase

KOL - Overall mass-transfer coefficient based on liquid phase

Kpo - Inertial parameter at venturi throat entrance

Kpg - Inertial parameter for mass-median diameter

kg - Mass-transfer coefficient for gas film (two-film theory)

kl - Mass-transfer coefficient for liquid film (two-film theory)

Lm - Liquid molar flow rate

L/G - Liquid-to-gas ratio

Lm/Gm - Liquid-to-gas ratio (mass flow rates)

l - Venturi throat length parameter, dimensionless

l t - Venturi throat length

m - Henry’s law constant for the equilibrium diagram

NA - Rate of transfer of component A (two-film theory)

NOG - Number of transfer units based on an overall gas-film coefficient, KOG

NOL - Number of transfer units based on an overall liquid-film coefficient, KOL

Np - Number of theoretical plates

NReo - Reynolds Number

Nt - Number of transfer units (plate towers)

NTU - Number of transfer units

P - Pressure

PG - Power input from gas stream

PL - Power input from liquid injection

PT - Total contacting power

Pt - Penetration

p - Partial pressure of solute at equilibrium *Ap - Equilibrium partial pressure of solute A at operating conditions

pAG - Partial pressure of solute A in the gas

pAI - Partial pressure of solute A at the interface

pL - Liquid-inlet pressure

QG - Gas flow rate

QL - Liquid flow rate

R - Ideal gas constant

T - Temperature


2.0-7/98 xv

Tg - Gas temperature

Tl - Water temperature

vgt - Gas velocity in the venturi throat

X - Mole fraction of solute in pure liquid

Y - Mole fraction of solute in inert gas

Z - Height of packing

Greek Symbols

Symbol Definition

α - Empirical constant used in contact power theory

β - Empirical constant used in contact power theory

∆p - Pressure drop

φ - Ratio of specific gravity of scrubbing liquid to that of water

η - Collection efficiency

ψ - Empirical correlation used to size a plate tower

µg - Gas viscosity

µl - Liquid viscosity

νg - Gas kinematic viscosity

π - Pi, value = 3.14

ρg - Gas density

ρl - Liquid density

ρp - Particle density

σgm - Geometric standard deviation

Units of Measurement

Abbreviation Unit of Measurement

acfm - Actual cubic feet per minute

cfm - Cubic feet per minute

cm - Centimeter

ft - Foot

APTI Course SI:412C ___________________________________________________________________________________

xvi 2.0-7/98

g - Gram

gal - Gallon

g-mol - Gram mole

h - Hour (metric)

hp - Horse power

hr - Hour (English)

in. - Inch

kg - Kilogram

kPa - Kilopascal

L - Liter

lb - Pound

lb-mole - Pound mole

MW - Megawatt

m - Meter

min - Minute

mm - Millimeter

mol - Mole

mph - Miles per hour

Pa - Pascal

ppm - Parts per million

psi - Pounds per square inch

psig - Pounds per square inch (gauge)

s - Second (metric)

scf - Standard cubic feet

sec - Second (English)

µm - Micrometer

µmA - Micrometer, aerodynamic diameter

cmA - Centimeter, aerodynamic diameter


2.0-7/98 xvii

Acronyms

ADVACATE - Advance Silicate

FGD - Flue Gas Desulfurization

SDA - Spray dryer absorber

DSI - Dry sorbent injector

ESP - Electrostatic precipitator

EPRI - Electric Power Research Institute

EPA - U.S. Environmental Protection Agency

DOE - U.S. Department of Energy

LIMB - Limestone Injection Multistage Burners

SNRB - Slow NOx Reduction Burners

HETP - Height Equivalent to Theoretical Plate

APTI Course SI:412C ___________________________________________________________________________________

xviii 2.0-7/98

2.0-7/98 xix

Course Description In this course, you will learn how various wet and dry scrubbers operate and how to evaluate the effectiveness of scrubber designs in reducing particulate and gaseous emissions from industrial sources. Major topics include the following:

• General description of various wet and dry scrubber designs

• Particle collection and absorption theory

• Operation and maintenance problems associated with scrubbers

• Scrubber components

• Use of scrubbers in flue gas desulfurization (FGD)

• Estimating collection efficiency of wet scrubbing systems

• Determining the liquid requirements for gas absorbers

• Sizing plate towers and packed towers (height and diameter)

Objectives Upon completion of this course, you will be able to do the following:

1. Identify various scrubber designs (both wet and dry) and briefly describe their operation

2. Briefly describe the mechanisms for particle collection and gas absorption in a scrubber

3. List three key design parameters affecting particle and gaseous pollutant removal

4. Identify which scrubbers are used mainly for particle collection and which are used mainly for gaseous pollutant removal

5. Describe typical operation and maintenance problems associated with various wet and dry scrubbers

6. Briefly describe four FGD systems used for removing sulfur dioxide emissions from boilers

7. Use estimating techniques and typical "rules of thumb" to evaluate scrubber plan designs for collection efficiency, adequate liquid flow rates, and proper sizing

APTI Course SI:412C ___________________________________________________________________________________

xx 2.0-7/98

Audience This course is intended primarily for air permit reviewers and air quality inspectors employed by state and local agencies. The course also provides useful training for technical personnel in private industry who prepare permit applications and are responsible for operating scrubbers in compliance with air quality regulations.

Course Length and CEUs This course will take approximately 40 hours to complete. The number of Continuing Education Units (CEUs) awarded with successful completion of the course is 4.

Suggested Prerequisites Prior to taking this course, completion of the following U.S. EPA APTI courses or the equivalent of one year’s experience in the air pollution control field is recommended:

SI:422 Air Pollution Control Orientation Course or 452 Principles and Practices of Air Pollution Control

SI:100 Mathematics Review for Air Pollution Control

You should also be able to use a calculator with various math functions.

Required Materials • Self-Instructional Manual, Scrubber Systems Operation Review

• Final examination

• Calculator

Supplemental Materials • Video titled, Venturi Scrubbers: Operating Principles and Components


2.0-7/98 xxi

Taking the Course Proceed sequentially through the manual until you have completed Lesson 11. The review exercises located at the end of every lesson test your mastery of the objectives covered in that lesson. The review exercises are designed so that you can either complete them as you finish a section of the lesson material or wait until you have completed the entire lesson. The review exercises in Lessons 10 and 11 give you the opportunity to test your abilities in solving problems similar to the example problems presented in those lessons. The video, Venturi Scrubbers: Operating Principles and Components is optional. If you acquire a copy, view it at the end of Lesson 3.

Each lesson contains the following:

Learning goal and objectives

Text material

Review exercises and exercise answers

For each lesson, follow these steps:

1. Do the assigned reading and view the assigned videotapes.

2. Complete the review exercise.

3. Check your answers against the key.

4. Review the instruction for any questions that you answered incorrectly.

Completing the Course A final examination accompanies this book (provided in a separate envelope). Take the final exam after you have finished the course. The exam is a closed book exam. Do not use your notes or books.

The final examination counts as 100% of your grade. To receive your certificate of completion and 4 Continuing Education Units (CEUs), you must score 70 or above on the exam. Follow these procedures:

1. Arrange for someone to be your test supervisor and give him/her the envelope.

2. Complete the final exam under the supervision of your test supervisor according to the test directions.

3. After you have finished the exam, ask the test supervisor to sign a statement on the answer sheet certifying that you took the exam in accordance with the specified test directions.

APTI Course SI:412C ___________________________________________________________________________________

xxii 2.0-7/98

4. Have your test supervisor mail the exam and answer sheet to the appropriate registrar below:

Registrar - Private Sector

NCSU Environmental Programs Box 7902 Raleigh, NC 27695 - 7902 Phone: (919) 515-5875 Fax: (919) 515-4386

or

Registrar - EPA/State Agency

U.S. Environmental Protection Agency MD-17 Research Triangle Park, NC 27711 Phone: (919) 541-2497 Fax: (919) 541-5598

Your exam and grade results will be mailed to you.


2.0-7/98 xxiii

2.0-7/98 1-1

Lesson 1 Introduction to Scrubbing Systems

Goal To familiarize you with scrubbers, the reasons they are used for emission reduction, and the review process for evaluating scrubber design.

Objectives At the end of this lesson, you will be able to do the following:

1. Briefly describe the purpose and process of wet and dry scrubbing

2. List four advantages and disadvantages of using wet scrubbers to collect particles and gases compared to using other air pollution control devices

3. Describe at least five components of a wet scrubbing system

4. Describe the major differences between wet and dry scrubbers

5. Identify two different types of dry scrubbers

Introduction

Scrubber systems are a diverse group of air pollution control devices that can be used to remove particles and/or gases from industrial exhaust streams. Traditionally, scrubbers have referred to pollution control devices that used liquid to "scrub" unwanted pollutants from a gas stream. Recently, the term scrubber is also used to describe systems that inject a dry reagent or slurry into a dirty exhaust stream to "scrub out" acid gases. Scrubbers are one of the primary devices that control gaseous emissions, especially acid gases.

Wet Scrubbers

Wet scrubber is a term used to describe a variety of devices that use liquid to remove pollutants. In a wet scrubber, the dirty gas stream is brought into contact with the scrubbing liquid by spraying it with the liquid, by forcing it through a pool of liquid, or by some other contact method.

Of course the design of any air pollution control device (wet scrubbers are no exception) depends on the industrial process conditions and the nature of the air pollutants involved.

Lesson 1 ___________________________________________________________________________________

1-2 2.0-7/98

Exhaust gas characteristics and dust properties, if particles are present, are of primary importance. Scrubbers can be designed to collect particulates and/or gaseous pollutants. Wet scrubbers remove particles by capturing them in liquid droplets. Wet scrubbers remove pollutant gases by dissolving or absorbing them into the liquid. Any droplets that are in the flue gas must then be separated from the clean exhaust stream by means of another device referred to as a mist eliminator or entrainment separator (these terms are interchangeable). Also, the resultant scrubbing liquid must be treated prior to any ultimate discharge or reused in the plant.

There are numerous configurations of scrubbers and scrubbing systems all designed to provide good contact between the liquid and dirty gas stream. Figures 1-1 and 1-2 show two examples of wet scrubber designs, including their mist eliminators. Figure 1-1 is a venturi scrubber design, which is discussed in greater detail in Lesson 3. The mist eliminator for a venturi scrubber is often a separate device called a cyclonic separator. Figure 1-2 has a tower design where the mist eliminator is built into the top of the structure. Various tower designs are covered in lessons 3, 4, 5, and 6. Entrainment separators and mist eliminators are covered in more detail in Lesson 8.

Figure 1-1. An example of a venturi scrubber design

��

��

Introduction to Scrubbing Systems ___________________________________________________________________________________

2.0-7/98 1-3

Figure 1-2. An example of a tower scrubber design

A wet scrubber's ability to collect small sized particulates is often directly proportional to the power input into the scrubber. Low energy devices such as spray towers are used to collect particulate matter larger than 5 micrometers. To obtain high efficiency removal of 1 micrometer (or less) particles generally requires high energy devices such as venturis or augmented devices such as condensation scrubbers. Additionally, a properly designed and operated entrainment separator/mist eliminator is important to achieve high removal efficiencies: the greater the number of liquid droplets that are not captured by the mist eliminator the higher the potential emission levels.

Wet scrubbers that remove gaseous pollutants are referred to as absorbers. Good gas-to-liquid contact is essential to obtain high removal efficiencies in absorbers. A number of wet scrubber designs are used to remove gaseous pollutants, with the packed tower and the plate tower being the most common.

If the exhaust stream contains both particles and gases, wet scrubbers are generally the only single air pollution control device that can remove both types of pollutants. Wet scrubbers can achieve high removal efficiencies for either particles or gases and, in some instances, can achieve a high removal efficiency for both pollutants in the same system. However, in many cases, the best operating conditions for particle collection are the poorest for gas removal. In general, obtaining high simultaneous gas and particle removal efficiencies requires that one

��

��

��

Lesson 1 ___________________________________________________________________________________

1-4 2.0-7/98

of them be easily collected (i.e., that the gases are very soluble in the liquid or that the particles are large and readily captured).

For particulate control, wet scrubbers (also referred to as wet collectors) are evaluated against fabric filters and electrostatic precipitators (ESPs). Some advantages of wet scrubbers over these devices are as follows:

• Wet scrubbers have the ability to handle high temperatures and moisture.

• In wet scrubbers, flue gases are cooled, resulting in smaller overall size of equipment.

• Wet scrubbers can remove both gases and particles.

• Wet scrubbers can neutralize corrosive gases.

Some disadvantages of wet scrubbers include corrosion, the need for mist removal to obtain high efficiencies, the need for treatment or reuse of spent liquid, and reduced plume buoyancy. Table 1-1 summarizes these advantages and disadvantages. Wet scrubbers have been used in a variety of industries such as acid plants, fertilizer plants, steel mills, asphalt plants, and large power plants.

Table 1-1. Relative advantages and disadvantages of wet scrubbers compared to other control devices

Advantages Disadvantages

Small space requirements Scrubbers reduce the temperature and volume of the unsaturated exhaust stream. Therefore, vessel sizes, including fans and ducts downstream, are smaller than those of other control devices. Smaller sizes result in lower capital costs and more flexibility in site location of the scrubber.

No secondary dust sources Once particles are collected, they cannot escape from hoppers or during transport.

Handles high-temperature, high-humidity gas streams No temperature limits or condensation problems can occur as in baghouses or ESPs.

Minimal fire and explosion hazards Various dry dusts are flammable. Using water eliminates the possibility of explosions.

Ability to collect both gases and particles

Corrosion problems Water and dissolved pollutants can form highly corrosive acid solutions. Proper construction materials are very important. Also, wet-dry interface areas can result in corrosion.

High power requirements High collection efficiencies for particles are attainable only at high pressure drops, resulting in high operating costs.

Water-disposal problems Settling ponds or sludge clarifiers may be needed to meet waste-water regulations.

Difficult product recovery Dewatering and drying of scrubber sludge make recovery of any dust for reuse very expensive and difficult.

Meteorological problems The saturated exhaust gases can produce a wet, visible steam plume. Fog and precipitation from the plume may cause local meteorological problems.


2.0-7/98 1-5

Wet Scrubber Systems

Wet scrubber systems generally consist of the following components:

• Ductwork and fan system

• A saturation chamber (optional)

• Scrubbing vessel

• Mist eliminator

• Pumping (and possible recycle system)

• Spent scrubbing liquid treatment and/or reuse system

• Exhaust stack

Figure 1-3 illustrates a typical wet scrubbing process. Hot flue gas enters the saturator where gases are cooled and humidified prior to entering the scrubbing area. The saturator removes a small percentage of the particles present in the flue gas. Next, the gas enters the venturi scrubber where approximately half of the gases are removed. By the time the gas exits the venturi, 95% of the particles have been removed. The gas flows through a second scrubber, a packed bed absorber, where the rest of the gases (and particles) are collected. The mist eliminator removes any liquid droplets that may have become entrained in the flue gas. The recirculation pump moves some of the spent scrubbing liquid back to the venturi scrubber where it is recycled and the remainder is sent to a treatment system. Treated scrubbing liquid is recycled back to the saturator and the packed bed absorber. Fans and ductwork move the flue gas stream through the system and eventually out the stack.

Lesson 1 ___________________________________________________________________________________

1-6 2.0-7/98

Figure 1-3. An example of a wet scrubbing system

Categorization of Wet Scrubbers

Since wet scrubbers vary greatly in complexity and method of operation, devising categories into which all of them neatly fit is extremely difficult. Scrubbers for particle collection are usually categorized by the gas-side pressure drop of the system. Gas-side pressure drop refers to the pressure difference, or pressure drop, that occurs as the


2.0-7/98 1-7

exhaust gas is pushed or pulled through the scrubber, disregarding the pressure that would be used for pumping or spraying the liquid into the scrubber. In this manual, the terms pressure drop and gas-side pressure drop will be used interchangeably. Classification of scrubbers by pressure drop is as follows:

• Low-energy scrubbers have pressure drops of less than 12.7 cm (5 in.) of water.

• Medium-energy scrubbers have pressure drops between 12.7 and 38.1 cm (5 and 15 in.) of water.

• High-energy scrubbers have pressure drops greater than 38.1 cm (15 in.) of water.

However, most scrubbers operate over a wide range of pressure drops, depending on their specific application, thereby making this type of categorization difficult.

Another way to classify wet scrubbers is by their use to primarily collect either particles or gaseous pollutants. Again, this distinction is not always clear since scrubbers can often be used to remove both types of pollutants.

In this course, wet scrubbers are categorized by the manner in which the gas and liquid phases are brought into contact. Scrubbers are designed to use power, or energy, from the gas stream or the liquid stream, or some other method to bring the pollutant gas stream into contact with the liquid. These categories are given in Table 1-2.

Table 1-2. Categories of wet collectors by energy source used for contact

Wet collector Energy source used for gas-liquid contact

Gas-phase contacting

Liquid-phase contacting

Wet film

Combination

• Liquid phase and gas phase

• Mechanically aided

Gas stream

Liquid stream

Liquid and gas streams

Liquid and gas streams

Mechanically driven rotor

Each of the wet collectors listed in Table 1-2 will be discussed in this course. For each design category, the following topics will be discussed: operation, collection efficiency, industrial applications, prominent maintenance problems and, when applicable, primary use.

Dry Scrubbing Systems

A dry or semi-dry scrubbing system, unlike the wet scrubber, does not saturate with moisture the flue gas stream that is being treated. In some cases no moisture is added; while in other designs only the amount of moisture that can be evaporated in the flue gas without condensing is added. Therefore, dry scrubbers do not have a stack steam plume or wastewater handling/disposal requirements. Dry scrubbing systems are used to remove acid gases (such as SO2 and HCl) primarily from combustion sources. There are a number of dry

Lesson 1 ___________________________________________________________________________________

1-8 2.0-7/98

type scrubbing system designs. However, all consist of two main sections or devices: (1) a device to introduce the acid gas sorbent material into the gas stream and (2) a particulate-matter control device to remove reaction products, excess sorbent material as well as any particulate matter already in the flue gas. Dry scrubbing systems can be categorized as dry sorbent injectors (DSIs) or as spray dryer absorbers (SDAs). Spray dryer absorbers are also called semi-dry scrubbers or spray dryers. Figures 1-4 and 1-5 illustrate both the DSI and spray dryer processes.

Dry sorbent injection involves the addition of an alkaline material (usually hydrated lime or soda ash) into the gas stream to react with the acid gases. The sorbent can be injected directly into several different locations: (1) the combustion process, (2) the flue gas duct (ahead of the particulate control device), or (3) an open reaction chamber (it one exists). The acid gases react with the alkaline sorbents to form solid salts which are removed in the particulate control device. These simple systems can achieve only limited acid gas (SO2 and HCl) removal efficiencies.

Higher collection efficiencies can be achieved by increasing the flue gas humidity (i.e., cooling using water spray). These devices have been used on medical waste incinerators and a few municipal waste combustors.

In spray dryer absorbers, the flue gases are introduced into an absorbing tower (dryer) where the gases are contacted with a finely atomized alkaline slurry. Acid gases are absorbed by the slurry mixture and react to form solid salts which are removed by the particulate control device. The heat of the flue gas is used to evaporate all the water droplets, leaving a non-saturated flue gas to exit the absorber tower. Spray dryers are capable of achieving high (80+%) acid gas removal efficiencies. These devices have been used on industrial and utility boilers and municipal waste combustors.

Figure 1-4. Dry sorbent injection scrubber system


2.0-7/98 1-9

Figure 1-5. Spray dryer absorber system

Design Evaluation

In evaluating a new scrubber design, especially from a regulatory viewpoint, the major issue is whether the proposed design will achieve the required particle and/or gaseous removal levels. There are three basic approaches to evaluating the capability of scrubbing systems: (1) empirical evaluations based on historical data on similar scrubbers, (2) theoretical models based on basic engineering principles and (3) pilot scale test data. Scrubber vendors utilize all three, especially historical data to design their systems. A person reviewing scrubber plans generally will not have access to all necessary data and may be limited to using the theoretical equations. In this workbook, operating information on various scrubbing systems along with basic theoretical models are presented so that a reviewer can utilize both to perform evaluations of scrubbing systems.

The design and operation of these systems are based on the basic laws of physics and general chemical engineering principles. For example, most scrubbers require a certain velocity (or residence time) through the vessel to obtain the required removal efficiency. This parameter is set by the size of the scrubber in relation to the volume of flue gas to be treated. In performing an in-depth evaluation of a scrubbing system there are certain parameters that should be verified such as the velocity, pressure drop or power input, and sorbent or reagent usage rate. For the more popular types of scrubbing systems there are general "rules of thumb" and empirical relationships that can be used to determine if scrubber design and operating parameters are within "normal" ranges. However, because of the variety of scrubber designs and the complex mass and heat balance transfer occurring in a scrubber, there is no one set of simple equations that can be used to do an in-depth evaluation of all scrubbing systems. Scrubber vendors utilize past operating and pilot scale data to refine system designs to specific applications.

Lesson 1 ___________________________________________________________________________________

1-10 2.0-7/98

To evaluate scrubber designs, this document provides both operating information on specific scrubbers and generalized review equations and procedures for common scrubbers. Reviewers can use the operating data to do the following:

• Determine whether the system being evaluated is within normal ranges for similar types of systems

• Make sure common operating problems are being addressed

• Assure that adequate data monitoring devices are specified and that data are being recorded

Utilizing all this information, a reviewer can develop a list of questions to be answered by the vendor/or operator in order to confirm adequate design of the system.

As is the case with any pollution control system, the ultimate proof of system design is in the guarantee and start-up performance test. Vendors will guarantee certain performance (efficiency) parameters, and these guarantee points should be evaluated in an initial review process. The initial stack and performance test will highlight any major design problem. Day-to-day operation will need to be assessed from monitoring instrumentation.

In addition to being useful to persons evaluating a new scrubber design, information presented in this course is also intended to be useful to persons responsible for overseeing the day-to-day operation and/or inspection of a scrubbing system. Evaluation of scrubber system performance begins with developing good baseline information on the important operating parameters. Using the information presented in this course on these operating parameters, a person can develop appropriate recordkeeping or checklists to evaluate scrubber operation. Also, example troubleshooting summaries are presented for certain scrubber designs. Regulatory personnel could use these lists as a guide to formulate questions or identify areas to investigate during a routine inspection of a facility. As with the design review, there is no one set of procedures or guidelines that can be used to evaluate operation and maintenance of all scrubber systems. General procedures presented in this course need to be customized to site specific applications.

Summary

Wet scrubbing systems are devices that use a liquid (generally water) to remove particulate and/or gaseous pollutants from a process exhaust gas stream. There are numerous different configurations of wet scrubbers. All designs attempt to provide good liquid-to-pollutant contact in order to obtain high removal (95% plus) efficiencies. Wet scrubbers saturate the gas stream thereby creating a steam plume and resulting wastewater stream that must be treated or reused in the plant. Also, since the gas stream is saturated with liquid, a mist eliminator or entrainment separator is often an integral part of any wet scrubbing system. Mist eliminators (entrainment separators) remove and/or recycle the scrubbing liquid in addition to providing additional pollutant removal.

Dry and semi-dry scrubbing systems are used to remove acid gases from combustion gas streams. These systems utilize a powder sorbent material, either calcium (lime) or sodium based to react with the acid gases in the flue gas and produce a solid salt that must be removed in a particulate control device. In dry systems, also referred to as dry sorbent injection, the dry powder sorbent material is injected directly into the ductwork, or reaction


2.0-7/98 1-11

chamber. In a semi-dry (or spray dryer absorber) system, the sorbent material is first mixed with water and then injected into a spray drying vessel where all the liquid is totally evaporated by cooling the gas stream while the sorbent reacts with the acid gases. By cooling the hot combustion gas stream, higher acid gas removal efficiencies are achievable than with simple duct injection.

To evaluate scrubber designs, this manual provides both a generalized review of design equations/procedures and operating information on specific scrubbing systems. Reviewers can use this information to determine if the scrubbing system is operating within normal ranges compared to other similar systems. This will provide the reviewer with a starting point to develop a list of questions aimed at vendors or operators that will aid in evaluating the adequacy of the design.

The next lesson provides an overview of the design features of wet and dry scrubbers that enhance collection of pollutants.

Lesson 1 ___________________________________________________________________________________

1-12 2.0-7/98


2.0-7/98 1-13

Review Exercise

1. True or False? Both wet and dry scrubbers can be used to remove both gaseous and particulate emissions.

2. Where is the mist eliminator (entrainment separator) located in relation to the scrubber vessel?

a. The mist eliminator is always separate from the scrubber vessel. b. The mist eliminator is always built into the same structure as the scrubber vessel. c. Depending on the scrubber design, the mist eliminator can either be a separate structure from the scrubber vessel or built into it.

3. A wet scrubber's ability to remove very small sized particles is often related to the:

a. Size of the system b. Flow pattern c. Power input d. Sorbent

4. In general, high removal rates for both particles and gases in the same scrubber are obtained by:

a. The use of large amounts of water b. Having gases that are highly soluble and/or particles that are relatively large c. The use of extremely high pressure drops d. A reagent added to the water

5. For particulate control, wet scrubbers are evaluated against two other types of control devices: ____________________ ____________________ and ____________________ ____________________.

6. True or False? Wet scrubbers used for particulate control have the advantage of being able to handle high-temperature and high-moisture gas streams.

7. Wet scrubbers are ____________________ than fabric filters or electrostatic precipitators used for the same application.

a. Larger b. Smaller c. Less noisy d. b and c, only

8. Compared to fabric filters or electrostatic precipitators, wet scrubbers have the following disadvantage(s):

a. Smaller equipment size b. Reduced plume buoyancy c. Need for treatment or reuse of spent scrubbing liquid d. b and c, only

Lesson 1 ___________________________________________________________________________________

1-14 2.0-7/98

9. Dry scrubbing systems consist of two main sections:

a. A quench tower and mist eliminator b. A device to introduce the sorbent and a particulate control device c. A venturi and a packed scrubber d. A liquid and a dry section

10. The two types of dry scrubbing systems are: ____________________ ____________________ ____________________ and ____________________ ____________________ ____________________.

11. A spray dryer uses a finely atomized ____________________ of alkaline sorbent to remove acid gases.

a. Liquid slurry b. Dry slurry c. Gas stream d. Any of the above


2.0-7/98 1-15


1. False Only wet scrubbers can remove both gaseous and particulate emissions.

2. c. Depending on the scrubber design, the mist eliminator can either be a separate structure from the scrubber vessel or built into it. The location of the mist eliminator (entrainment separator) in relation to the scrubber vessel depends on the scrubber design. The entrainment separator for a venturi scrubber is often a separate device (cyclonic separator). Scrubbers with tower designs have the mist eliminator built into the top of the structure.

3. c. Power input A wet scrubber's ability to remove very small sized particles is often related to the power input.

4. b. Having gases that are highly soluble and/or particles that are relatively large In general, high removal rates for both particles and gases in the same scrubber are obtained by having gases that are highly soluble and/or particles that are relatively large.

5. Fabric filters Electrostatic precipitators For particulate control, wet scrubbers are evaluated against two other types of control devices: fabric filters and electrostatic precipitators.

6. True Wet scrubbers used for particulate control have the advantage of being able to handle high-temperature and high-moisture gas streams.

7. b. Smaller Wet scrubbers are smaller in size than fabric filters or electrostatic precipitators used for the same application.

8. d. b and c, only When compared to fabric filters or electrostatic precipitators, wet scrubbers have the following disadvantages:

• Reduced plume buoyancy

• The need for treatment or reuse of spent scrubbing liquid

9. b. A device to introduce the sorbent and a particulate control device Dry scrubbing systems consist of two main sections: a device to introduce the sorbent and a particulate control device.

10. Dry sorbent injection Spray dryer absorber The two types of dry scrubbing systems are dry sorbent injection and spray dryer absorber.

11. a. Liquid slurry A spray dryer uses a finely atomized liquid slurry of alkaline sorbent to remove acid gases.

Lesson 1 ___________________________________________________________________________________

1-16 2.0-7/98


2.0-7/98 1-17

Bibliography

Bethea, R. M. 1978. Air Pollution Control Technology. New York: Van Nostrand Reinhold.

Perry, J. H. (Ed.). 1973. Chemical Engineers’ Handbook. 5th ed. New York: McGraw-Hill.

Richards, J. R. 1995. Control of Particulate Emissions (APTI Course 413). U.S. Environmental Protection Agency.

Richards, J. R. 1995. Control of Gaseous Emissions. (APTI Course 415). U.S. Environmental Protection Agency.

Semrau, K. T. 1977. Practical process design of particulate scrubbers. Chemical Engineering. 84:87-91.

Lesson 1 ___________________________________________________________________________________

1-18 2.0-7/98

2.0-7/98 2-1

Lesson 2 Operating Principles Of Scrubbers

Goal To introduce you to the operating principles of wet and dry scrubbers that enhance the collection of air pollutants.


1. Name three scrubbing process variables that affect particle collection in a wet scrubber

2. Describe the two most important mechanisms for collecting particles in a wet scrubber

3. List three conditions that will enhance the absorption process for either wet or dry scrubbers

4. Define pressure drop and explain its importance in the scrubbing process

5. Define liquid-to-gas ratio (L/G) ratio and explain its importance in the scrubbing process

Introduction

Depending on the design, scrubbers can collect particles, gases, both particles and gases, or acid gases. This lesson discusses the basic principles on which scrubber designs are based. As discussed in Lesson 1, wet scrubbers remove particles by capturing them in liquid droplets and they remove gases by dissolving or absorbing them into liquid. In this lesson, you will learn which set of conditions enhances particle collection and which set promotes gas collection. Some conditions enhance the collection of both types of pollutants.

Wet scrubbers remove particles from gas streams by capturing the particles in liquid droplets or in sheets of scrubbing liquid (usually water) and then separating the droplets from the gas stream. Several process variables affect particle capture; they include particle size, the size of liquid droplets, and the relative velocity of the particle and the liquid droplets, with particle size being the most important parameter. In general, larger particles are easier to collect than smaller ones. The key to effective particle capture in a wet scrubber is creating a mist of tiny droplets that act as collection targets: usually, the smaller the droplet and the more densely the droplets are packed, the better the ability to capture smaller-sized particles. Particle capture generally improves with higher energy systems because energy is required to produce

Lesson 2 ___________________________________________________________________________________

2-2 2.0-7/98

the mist of tiny droplets. Also, a high relative velocity between particles and liquid droplets (the particles are moving fast compared to the liquid droplets) promotes particle collection.

For gaseous pollutant collection, the pollutant must be soluble in the chosen scrubbing liquid. In addition, the system must be designed to provide good mixing between the gas and liquid phases, and enough time (residence time) for the gaseous pollutants to dissolve.

Other important considerations for both particulate and gaseous pollutant collection are the amount of liquid injected into the scrubber per given volume of gas flow (referred to as the liquid-to-gas ratio) and the removal of any entrained liquid droplets. The liquid-to-gas ratio is important to provide sufficient liquid for effective pollutant removal. Also, the system must be designed to remove entrained mists, or droplets, from the cleaned exhaust gas stream before it leaves the stack. If not removed, the "captured" pollutants could be emitted from the stack.

A dry sorbent injector (DSI) removes acid gases by causing direct contact between the alkaline sorbent and the acid gases. The acid gases are adsorbed onto the solid sorbent particle and the alkaline material reacts with the acid gases to form solid salts. In these simple systems, the degree to which the alkaline material can be brought into intimate contact with the acid gases is the key in obtaining effective removal. Also, the surface area of the sorbent and reactivity with acid gases are primary parameters that control collection.

An enhancement of the DSI is the spray dryer absorber (SDA). In the SDA, the alkaline sorbent is mixed with water and injected into a reaction vessel. The water cools and humidifies the gas stream, increasing the efficiency of the reaction. Also, mechanical atomization of the slurry is provided to produce very fine alkaline droplets to increase contact between the acid gases and sorbent material. In this process the acid gases are absorbed into the liquid droplet where they react with sorbent.

Particle Collection

Wet scrubbers capture relatively small dust particles with large liquid droplets. In most wet scrubbing systems, droplets produced are generally larger than 50 micrometers (in the 150 to 500 micrometer range). As a point of reference, human hair ranges in diameter from 50 to 100 micrometers. The size distribution of particles to be collected is source specific. For example, particles produced by mechanical means (crush or grind) tend to be large (above 10 micrometers); whereas, particles produced from combustion or a chemical reaction will have a substantial portion of small (i.e. less than 5 micrometers) and submicrometer-sized particles. The most critical sized particles are those in the 0.1 to 0.5 micrometer range because they are the most difficult for wet scrubbers to collect.

Droplets are produced by several methods:

1. Injecting liquid at high pressure through specially designed nozzles

2. Aspirating the particle-laden gas stream through a liquid pool

3. Submerging a whirling rotor in a liquid pool.

These droplets collect particles by using one or more of several collection mechanisms. These mechanisms impaction, direct interception, diffusion, electrostatic attraction, condensation,

Operating Principles of Scrubbers ___________________________________________________________________________________

2.0-7/98 2-3

centrifugal force, and gravity are explained in Table 2-1. However, impaction and diffusion are the two primary ones.

Table 2-1. Particle collection mechanisms for wet scrubbing systems

Mechanism Explanation

Impaction Particles too large to follow gas streamlines around a droplet collide with it.

Diffusion Very tiny particles move randomly, colliding with droplets because they are confined in a limited space.

Direct interception An extension of the impaction mechanism. The center of a particle follows the streamlines around the droplet, but a collision occurs if the distance between the particle and droplet is less than the radius of the particle.

Electrostatic attraction Particles and droplets become oppositely charged and attract each other.

Condensation When hot gas cools rapidly, particles in the gas stream can act as condensation nuclei and, as a result, become larger.

Centrifugal force The shape or curvature of a collector causes the gas stream to rotate in a spiral motion, throwing larger particles toward the wall.

Gravity Large particles moving slowly enough will fall from the gas stream and be collected.

Impaction

In a wet scrubbing system, dust particles will tend to follow the streamlines of the exhaust stream. However, when liquid droplets are introduced into the exhaust stream, particles cannot always follow these streamlines as they diverge around the droplet (Figure 2-1). The particle's mass causes it to break away from the streamlines and impact or hit the droplet.

Figure 2-1. Impaction

Impaction increases as the diameter of the particle increases and as the relative velocity between the particle and droplets increases. As particles get larger they are less likely to

Lesson 2 ___________________________________________________________________________________

2-4 2.0-7/98

follow the gas streamlines around droplets. Also, as particles move faster relative to the liquid droplet, there is a greater chance that the particle will hit a droplet. Impaction is the predominant collection mechanism for scrubbers having gas stream velocities greater than 0.3 m/s (1 ft/sec) (Perry 1973). Most scrubbers do operate with gas stream velocities well above 0.3 m/s. Therefore, at these velocities, particles having diameters greater than 1.0 µm are collected by this mechanism.

Impaction also increases as the size of the liquid droplet decreases because the presence of more droplets within the vessel increases the likelihood that particles will impact on the droplets.

Diffusion

Very small particles (less than 0.1 µm in diameter) experience random movement in an exhaust stream. These particles are so tiny that they are bumped by gas molecules as they move in the exhaust stream. This bumping, or bombardment, causes them to first move one way and then another in a random manner, or to diffuse, through the gas. This irregular motion can cause the particles to collide with a droplet and be collected (Figure 2-2). Because of this, diffusion is the primary collection mechanism in wet scrubbers for particles smaller than 0.1 µm.

Figure 2-2. Diffusion

The rate of diffusion depends on the following:

1. The relative velocity between the particle and droplet

2. The particle diameter

3. The liquid-droplet diameter.

For both impaction and diffusion, collection efficiency increases with an increase in relative velocity (liquid- or gas-pressure input) and a decrease in liquid-droplet size.


2.0-7/98 2-5

However, collection by diffusion increases as particle size decreases. This mechanism enables certain scrubbers to effectively remove the very tiny particles (less than 0.1 µm). In the particle size range of approximately 0.1 to 1.0 µm, neither of these two collection mechanisms (impaction or diffusion) dominates. This relationship is illustrated in Figure 2-3.

Figure 2-3. Hypothetical curve illustrating relationship between particle size and collection efficiency for typical wet scrubber

Other Collection Mechanisms

In recent years, some scrubber manufacturers have utilized other collection mechanisms such as electrostatic attraction and condensation to enhance particle collection without increasing power consumption. In electrostatic attraction, particles are captured by first inducing a charge on them. Then, the charged particles are either attracted to each other, forming larger, easier-to-collect particles, or they are collected on a surface. Condensation of water vapor on particles promotes collection by adding mass to the particles. Other mechanisms such as gravity, centrifugal force, and direct interception slightly affect particle collection.

To test your knowledge of the preceding section, answer the questions in Part 1 of the Review Exercise.

Lesson 2 ___________________________________________________________________________________

2-6 2.0-7/98

Gas Collection

The process of dissolving gaseous pollutants in a liquid is referred to as absorption. Absorption is a mass-transfer operation. Mass transfer refers to the process of transferring a component from one phase or stream to another. Mass transfer can be compared to heat transfer in that both occur because a system is trying to reach equilibrium conditions. For example, in heat transfer, if a hot slab of metal is placed on top of a cold slab, heat energy will transfer from the hot slab to the cold slab until both reach the same temperature (equilibrium). In absorption, mass instead of heat is transferred as a result of a concentration difference, rather than a heat-energy difference. Absorption continues as long as a concentration differential exists between the liquid and the gas from which the contaminant is being removed.

To remove a gaseous pollutant by absorption, the exhaust stream must pass through (be in contact with) a liquid. Figure 2-4 illustrates the three steps involved in absorption. In the first step, the gaseous pollutant diffuses from the bulk area of the gas phase to the gas-liquid interface. In the second step, the gas moves (transfers) across the interface to the liquid phase. This step occurs extremely rapidly once the gas molecules (pollutant) arrive at the interface area. In the third step, the gas diffuses into the bulk area of the liquid, thus making room for additional gas molecules to be absorbed. The rate of absorption (mass transfer of the pollutant from the gas phase to the liquid phase) depends on the diffusion rates of the pollutant in the gas phase (first step) and in the liquid phase (third step).

Figure 2-4. Absorption

To enhance gas diffusion and therefore absorption, scrubber designs should do the following:

1. Provide a large interfacial contact area between the gas and liquid phases (i.e. numerous, tiny liquid droplets)

2. Provide good mixing of the gas and liquid phases (turbulence)

3. Allow sufficient residence, or contact, time between the gas and liquid phases


2.0-7/98 2-7

Two of these three gas-collection mechanisms, large contact area and good mixing, are also important for particle collection. The third factor, sufficient residence time, works in direct opposition to efficient particle collection. To increase residence time, the relative velocity of the gas and liquid streams must be reduced. Therefore, achieving a high removal efficiency for both gaseous and particulate pollutants is extremely difficult unless the gaseous pollutant is very soluble in the liquid.

Solubility is a very important factor affecting the amount of a pollutant that can be absorbed. Solubility governs the amount of liquid required (liquid-to-gas ratio) and the necessary contact time. More soluble gases require less liquid. Also, more soluble gases will be absorbed faster. Solubility is a function of both the temperature and, to a lesser extent, the pressure of the system. As temperature increases, the amount of gas that can be absorbed by a liquid decreases. From the ideal gas law: as temperature increases, the volume of a gas also increases; therefore, at a higher temperature, gas volume increases and less gas is absorbed. For this reason, some absorption systems use inlet quench sprays to cool the incoming exhaust stream, thereby increasing absorption efficiency. Pressure affects the solubility of a gas in the opposite manner. When the pressure of a system is increased, the amount of gas absorbed generally increases.

Acid Gas Removal Mechanisms

In dry scrubbing systems, the mechanism of acid gas removal involves both the adsorption and absorption processes. Adsorption occurs when the acid gas molecules adhere to the surface of the solid sorbent. Absorption occurs when the acid gases dissolve in the liquid droplets. In either case, the acid gases eventually react with the alkaline sorbent to form a salt.

For dry injection systems, adsorption is the primary reaction mechanism. The acid gases are brought into contact with the solid sorbent, adsorbed onto the surface and then react to form a solid salt. In a spray dryer, both absorption and adsorption occur. Acid gases are absorbed into the tiny liquid droplets where they react with the dissolved alkaline material. Adsorption occurs when the liquid is evaporated and the acid gases adsorb onto the solid sorbents that are present.

The alkaline sorbent materials used in dry scrubbing systems are chosen because of their ability to react with or neutralize the acid gases. The materials, either calcium- or sodium-based, are tiny, porous solids that have the ability to adsorb the acid gas material.

Both adsorption and absorption are mass transfer processes that involve gas diffusion to provide effective acid gas removal in dry scrubbing systems. Therefore, similar operating conditions enhance both processes; and as discussed in the previous section they are as follows: • Large contact areas for reactions to occur The porous surface area of the solid sorbent

and/or the tiny liquid droplets provide these reaction sites.

• Good gas to solid/liquid mixing The acid gases must be brought into intimate contact with the alkaline sorbent or liquid droplets for reaction to occur.

• Sufficient reaction time Either a reaction chamber or residence time in the ductwork needs to be provided.

Lesson 2 ___________________________________________________________________________________

2-8 2.0-7/98

Temperature also affects both the adsorption and absorption processes: the cooler the flue gas, the more effective both processes become. Therefore, higher removal efficiencies can be achieved by reducing the flue gas temperature.


Pressure Drop

The pressure drop across a scrubbing system, and more importantly across individual components, is an important parameter in evaluating wet scrubber operation especially for particle removal. Pressure drop is a measurement of the resistance to flow as the flue gas passes from one point to another point. The resistance to flow is caused by both friction and turbulence.

In describing air flow through a system, a number of pressure terms are used. The three most important are: static pressure, velocity pressure, and total pressure. Air flowing in a pipe is acted on by static pressure and velocity, which when added together give total pressure. Static pressure acts in all directions and is measured at right angles to the direction of air flow to avoid influence from air velocity. Velocity pressure is the pressure created by air moving at a specific velocity. Total pressure is the sum of the static and velocity pressures. The pressure drop across a device is simply the arithmetic difference between the static pressures (measured at right angles to the flow) at the inlet and outlet of the device (See Figure 2-5). Lesson 8 describes measurement of pressure in more detail.

Figure 2-5. Measuring pressure drop across a venturi scrubber


2.0-7/98 2-9

Pressure drop is an important indicator of wet scrubber performance because for most wet scrubbers, particulate collection increases as the power input (or pressure drop) increases. Therefore, by supplying more power (i.e. creating a higher pressure drop) particle collection can be enhanced. (This theory is discussed in Lesson 10.) Also, as the pressure drop increases, the cost to operate the system increases because more fan or liquid pumping power is required.

As stated previously, pressure losses from air flowing in a pipe are due to friction and turbulence. Friction losses are caused by the air rubbing or abrading against the sides of the duct, scrubber or internal components. Obviously, the more internal components or obstructions there are, the higher the potential friction losses will be. Turbulent losses are caused when the air flow is disturbed. For example, changes in direction and/or reduction in size of the duct or scrubber are main causes of turbulence.

Liquid-To-Gas Ratio

An important parameter in wet scrubbing systems is the rate of liquid flow. It is common in wet scrubber terminology to express the liquid flow as a function of the gas flow rate that is being treated. This is commonly called the liquid-to-gas ratio and uses the units of gallons per 1000 acfm. Expressing the amount of liquid used as a ratio enables systems of different sizes to be readily compared.

For particulate removal, the liquid-to-gas ratio is a function of the mechanical design of the system; while for gas absorption the liquid-to-gas ratio gives an indication of the difficulty of removing a pollutant. Most wet scrubbers used for particulate control operate with liquid-to-gas ratios in the 4 to 20 gallons per 1000 acfm range. Dependent on scrubber design, a minimum volume of liquid is required to "wet" the scrubber internals and create sufficient collection targets. After a certain optimum point, adding excess liquid to a particulate wet scrubber does not help increase efficiency and in fact, could be counter productive by causing excessive pressure loss. Liquid-to-gas ratios for gas absorption are often higher; in the 20 to 40 gallons per 1000 acfm range.

Table 2-2 presents a listing of the types of wet scrubbers that have been used in flue gas desulfurization systems on utility boilers. The table illustrates a number of points about the choice of wet scrubbers used for gas absorption. For example, because flue gas desulfurization systems must deal with heavy particulate loadings, open, simple designs (such as venturi, spray chamber and moving bed) are used.

Also, the liquid-to-gas ratio for the absorption process is higher than for particulate removal and gas velocities are kept low to enhance the absorption process.

Lesson 2 ___________________________________________________________________________________

2-10 2.0-7/98

Table 2-2. Scrubbing systems used on utility boilers

L/G, gpm/1000 acfm

Plant name

No. of modules per unit

Type of module

Presaturator

/scrubber

Absorber

Gas flow per

module, acfm

∆P in. H2O

Gas velocity

fps

Pleasants 1 and 2 4 Venturi/tray

tower

14-17 35 600,000 6 10.4

Four Corners 1, 2,

and 3

2 Venturi 25 - 407,000

(1 & 2)

515,000

(3)

12 NR

R. D. Green 1 and

2

2 Spray tower NR1 45 500,000 3 9.2

Conesville 5 and

6

2 Moving bed 0.55 57 500,000 6 10

Coal Creek 1 and

2

4 Spray tower - 60 685,000 4.5 10.6

Elrama Station 5 Venturi - 40 550,000 11 NR

Phillips Station 5 Venturi - 40 547,000 16 40

Hawthorn 3 and 4 2 Marble bed 26 - 250,000 12 10

Green River

Station

1 Venturi/moving

bed

34 34 288,000 7/4 14

Cane Run 4 2 Venturi/moving

bed

NR 60 368,000 4 11.5

Mill Creek 3 4 Venturi/moving

bed

5 60 400,000 11 10

Paddy's Run 6 2 Marble bed - 16.5 175,000 11.5 9

Clay Boswell 4 4 Venturi/spray

tower

20 50 640,000 20 12

Milton R. Young 2 2 Spray tower - 80 859,000 8 9

Colstrip 1 and 2 3 Venturi/spray

tower

15 18 426,000 17/1 200/8.7

Bruce Mansfield 1

and 2

6 Venturi/venturi 23 20 558,000 18/6 200/100

Bruce Mansfield 3 5 Horizontal

spray

chamber

0.7 88 992,000 2.8 22

Hunter 1 and 2 4 Spray tower - 70 330,000 5-6 9.6

Huntington 1 4 Spray tower - 70 330,000 5-6 9.6

1. NR - Not reported Source: EPRI 1983.


2.0-7/98 2-11

Dry injection systems do not use liquid; therefore, this term is meaningless in describing their operation. In a spray dryer absorber, the amount of liquid injected is dependent on the saturation temperature of the flue gas. Spray dryer absorbers inject water to reduce the flue gas temperature to above the saturation level. Again, lower temperatures promote better acid gas removal efficiencies; however, gases exiting a spray dryer must be maintained above the dew point to protect downstream equipment including the particulate control device. The exit temperature dictates the amount of water injected; while the acid gas levels determine the level of alkaline reagent injected. Therefore, in a spray dryer absorber the important parameter that indicates operating acid gas removal efficiencies is the stoichiometric ratio of alkaline sorbent injected to acid gases removed. This ratio is determined on a molar basis. For example, a stoichiometric ratio of 1.0 would indicate that 1 mole of sorbent was used to neutralize 1 mole of acid gases. Dry scrubbing systems operate with stoichiometric ratios ranging from 1.5 to 4.0.


Summary

Wet scrubbers have been used for many years to remove particles from a gas stream. The primary collection mechanisms that affect particle capture in a wet scrubber are impaction and diffusion. Both impaction and diffusion effects increase as the size of the liquid droplets in the scrubber decreases. Impaction can also be enhanced by increasing the velocity of the particles compared to the velocity of the liquid collection droplets. Therefore, as liquid droplets become smaller and relative velocities between particles and liquid droplets increase, particle collection will improve. In addition, the particle size distribution is important because wet scrubbers can have difficulty collecting particles in the 0.1 to 0.5 µm range.

For wet scrubbers that remove gaseous pollutants, absorption is dependent on the solubility of the gas in the liquid. This solubility relationship will determine the type and amount of scrubbing liquid utilized. Once this is set, the keys to obtaining good removal efficiency are:

• Provide a large contact area between the gas and liquid

• Provide good mixing

• Allow adequate residence or contact time between gas and liquid

In dry scrubbing systems the mechanism of acid gas removal involves both adsorption and absorption. Adsorption occurs when the acid gases are adsorbed (adhere) to the surface of the solid alkaline material. Absorption occurs in the semi-dry systems when the acid gases absorb into the liquid droplets. Both adsorption and absorption are mass transfer processes that are enhanced by the criteria discussed above for wet scrubber gas absorption.

Pressure drop is an important parameter, especially for wet particulate scrubbers. Pressure drop is a measurement of the resistance to flow as the flue gas passes from one point to the next. For most particulate wet scrubbers, as the power input increases (i.e. increase in pressure drop) the collection efficiency increases. However, as the pressure drop increases, the cost to operate the system also increases.

Lesson 2 ___________________________________________________________________________________

2-12 2.0-7/98

Another important parameter in wet scrubbing systems is the liquid-to-gas ratio, expressed in terms of gallons per 1000 acfm. For particulate collection, the liquid-to-gas ratio is a function of the amount of liquid needed to thoroughly "wet" the scrubber. After a certain optimum point, adding more liquid does not enhance particulate removal. For gaseous removal, more liquid is required than for particulate collection. For absorption, the amount of liquid depends on the degree of difficulty in removing the pollutant of concern. In the next several lessons, the operation of specific scrubber designs will be covered.


2.0-7/98 2-13

Review Exercise Part 1

1. What are two primary mechanisms used for collecting particles in a scrubber?

a. Impaction and diffusion b. Direct interception and diffusion c. Impaction and condensation d. Direct interception and gravity

2. ____________________ is/are the predominant collection mechanism(s) for particles larger than 1.0 µm traveling faster than 0.3 m/s (1 ft/sec).

a. Impaction b. Diffusion c. Direct interception d. All of the above

3. For very small particles, below 0.1 µm in diameter, ____________________ is/are the predominant collection mechanism(s) in wet collection.

a. Impaction b. Diffusion c. Gravity d. All of the above

4. Collection efficiency for particles by the impaction mechanism increases as the:

a. Particles' velocity in the exhaust stream increases relative to the liquid droplets' velocity b. Particle size decreases below 0.1 µm in diameter c. Liquid-droplet size increases

5. For particle collection by diffusion, the collection efficiency ____________________ as the size of the particle decreases.

a. Decreases b. Increases

Part 2

6. In absorption, gaseous pollutants are ____________________ in a liquid.

7. Absorption occurs because of a ____________________ difference between the gas phase and liquid phase.

a. Heat b. Mass c. Concentration d. Weight

Lesson 2 ___________________________________________________________________________________

2-14 2.0-7/98

8. Which of the following would not enhance the absorption process?

a. Providing a large contact area between the gas and liquid phases b. Providing a turbulent mixing of the phases c. Increasing the gas velocity relative to the liquid velocity d. Allowing long contact or residence time

9. True or False? The solubility of the gaseous pollutant in the liquid will affect the required liquid-to-gas ratio of the system.

10. Increasing which of the following decreases the solubility of gas in a liquid?

a. Temperature b. Pressure

11. In a dry scrubbing system, the mechanism for acid gas removal involves:

a. Absorption b. Adsorption c. Electrostatic attraction d. a and b

12. In a dry scrubbing system, ____________________ the flue gas stream improves the potential removal efficiency.

a. Heating b. Cooling

Part 3

13. Pressure drop is a measure of the ____________________ to flow as the flue gas moves from point to point.

14. The pressure drop across a scrubbing system is obtained by:

a. Using transmissometers b. Taking the arithmetic difference between the static pressure measured at the inlet and outlet of a component c. Dividing the total pressure by the velocity pressure d. All of the above

15. True or False? Increasing the liquid-to-gas ratio will always increase removal efficiency in any type of scrubbing system.

16. Stoichiometric ratio refers to the ratio of the:

a. Amount of water to acid gases removed b. Amount of sorbent to acid gases removed c. Amount of solids to liquid present in a slurry d. Diameter to height


2.0-7/98 2-15

Review Exercise Answers Part 1

1. a. Impaction and diffusion The two primary mechanisms used for collecting particles in a scrubber are impaction and diffusion.

2. a. Impaction Impaction is the predominant collection mechanism for particles larger than 1.0 µm traveling faster than 0.3 m/s (1 ft/sec).

3. b. Diffusion For very small particles, below 0.1 µm in diameter, diffusion is the predominant collection mechanism in wet collection.

4. a. Particles' velocity in the exhaust stream increases relative to the liquid droplets' velocity Collection efficiency for particles by the impaction mechanism increases as the particles' velocity in the exhaust stream increases relative to the liquid droplets' velocity.

5. b. Increases Collection efficiency for particles by diffusion increases as the size of the particle decreases.

Part 2

6. Dissolved In absorption, gaseous pollutants are dissolved in a liquid.

7. c. Concentration Absorption occurs because of a concentration difference between the gas phase and liquid phase.

8. c. Increasing the gas velocity relative to the liquid velocity The absorption process is NOT enhanced by increasing the gas velocity relative to the liquid velocity (as particle collection is); but it is enhanced by:

• providing a large contact area between the gas and liquid phases

• providing a turbulent mixing of the phases

• allowing long contact or residence time

9. True The solubility of the gaseous pollutant in the liquid will affect the required liquid-to-gas ratio of the system.

10. a. Temperature Increasing the temperature decreases the solubility of gas in a liquid. As the temperature increases, the amount of gas that can be absorbed by the liquid decreases because the gas expands.

11. d. a and b In a dry scrubbing system, the mechanism for acid gas removal involves absorption and adsorption.

Lesson 2 ___________________________________________________________________________________

2-16 2.0-7/98

12. b. Cooling In a dry scrubbing system, cooling the flue gas stream improves the potential removal efficiency.

Part 3

13. Resistance Pressure drop is a measure of the resistance to flow as the flue gas moves from point to point.

14. b. Taking the arithmetic difference between the static pressure measured at the inlet and outlet of a component The pressure drop across a scrubbing system is obtained by taking the arithmetic difference between the static pressure measured at the inlet and outlet of a component.

15. False Increasing the liquid-to-gas ratio will NOT always increase removal efficiency in any type of scrubbing system. For particulate removal, after a certain L/G ratio is achieved, increasing the amount of liquid will not increase removal efficiency.

16. b. Amount of sorbent to acid gases removed Stoichiometric ratio refers to the ratio of the amount of sorbent to acid gases removed.


2.0-7/98 2-17

Bibliography


National Asphalt Pavement Association. 1978. The Maintenance and Operation of Exhaust Systems in the Hot Mix Batch Plant. 2nd ed. Information Series 52.




Schifftner, K. C. 1979, April. Venturi scrubber operation and maintenance. Paper presented at the U.S. EPA Environmental Research Information Center. Atlanta, GA.


U.S. Environmental Protection Agency. 1982, September. Control Techniques for Particulate Emissions from Stationary Sources. Vol. 1. EPA 450/3-81-005a.

Wechselblatt, P. M. 1975. Wet scrubbers (particulates). In F. L. Cross and H. E. Hesketh (Eds.), Handbook for the Operation and Maintenance of Air Pollution Control Equipment. Westport: Technomic Publishing.

Lesson 2 ___________________________________________________________________________________

2-18 2.0-7/98

2.0-7/98 3-1

Lesson 3 Gas-Phase Contacting Scrubbers

Goal To familiarize you with the operation, collection efficiency, and maintenance problems of gas-phase contacting scrubbers.


1. List three gas-phase contacting scrubbers and briefly describe how each operates

2. For each scrubber above, identify the range of operating values for pressure drop, liquid-to-gas ratio, as well as the collection efficiency for both particulate and gaseous pollutants

3. Describe typical operating and maintenance problems associated with each gas-phase contacting scrubber design

Video Presentation (optional): If you have acquired the video titled, Venturi Scrubbers: Operating Principles and Components, please view it at the end of this lesson.

Introduction

As mentioned in Lesson 1, wet scrubbers in this manual are categorized according to the manner in which the gas and liquid phases come into contact. This lesson, on gas-phase contacting scrubbers, begins your introduction to the many different scrubber designs covered in Lessons 3 through 6.

Gas-phase contacting scrubbers use the exhaust gas stream to provide the energy for gas-liquid contact. The exhaust stream moves across or through a liquid surface, shearing it to form tiny droplets. Breaking the liquid into fine droplets helps increase both particle and gas collection because the droplets provide targets on which the particles hit and are collected. They also provide a huge surface area for collecting (absorbing) gaseous pollutants. Thus, gas-phase contacting scrubbers can be used for both particles and gaseous pollutant collection.

A number of methods are used to provide this shearing action. The gas can be forced through constricted passages wetted with liquid, such as in venturi and orifice scrubbers. Or, the gas can be forced through cascades of liquid falling over flat plates as in plate tower

Lesson 3 ___________________________________________________________________________________

3-2 2.0-7/98

scrubbers. This lesson will focus on these collectors that work primarily by gas-phase contacting: venturi scrubbers, plate towers, and orifice scrubbers.

Venturi Scrubbers

A venturi scrubber is designed to effectively use the energy from the exhaust stream to atomize the scrubbing liquid. Venturi devices have been used for over 100 years to measure fluid flow (Venturi tubes derived their name from G. B. Venturi, an Italian physicist). About 35 years ago, Johnstone (1949) and other researchers found that they could effectively use the venturi configuration to remove particles from an exhaust stream. Figure 3-1 illustrates the classic venturi configuration.

Figure 3-1. Venturi configuration

A venturi scrubber consists of three sections: a converging section, a throat section, and a diverging section. The exhaust stream enters the converging section and, as the area decreases, gas velocity increases. Liquid is introduced either at the throat or at the entrance to the converging section. The exhaust gas, forced to move at extremely high velocities in the small throat section, shears the liquid from its walls, producing an enormous number of very tiny droplets. Particle and gas removal occur in the throat section as the exhaust stream mixes with the fog of tiny liquid droplets. The exhaust stream then exits through the diverging

Gas-Phase Contacting Scrubbers ___________________________________________________________________________________

2.0-7/98 3-3

section, where it is forced to slow down. Venturis can be used to collect both particulate and gaseous pollutants, but they are more effective in removing particles than gaseous pollutants.

Liquid can be injected at the converging section or at the throat. Figure 3-2 shows liquid injected at the converging section. Thus, the liquid coats the venturi throat making it very effective for handling hot, dry exhaust gas that contains dust. Otherwise, the dust would have a tendency to cake on or abrade a dry throat. These venturis are sometimes referred to as having a wetted approach.

Figure 3-2. Venturi scrubber with a wetted throat

��

��

��

Lesson 3 ___________________________________________________________________________________

3-4 2.0-7/98

Figure 3-3 shows liquid injected at the venturi throat. Since it is sprayed at or just before the throat, it does not actually coat the throat surface. These throats are susceptible to solids buildup when the throat is dry. They are also susceptible to abrasion by dust particles. These venturis are best used when the exhaust stream is cool and moist. These venturis are referred to as having a non-wetted approach.

Figure 3-3. Venturi with throat sprays

Venturis with round throats (Figures 3-2 and 3-3) can handle exhaust flows as large as 88,000 m3/h (40,000 cfm) (Brady and Legatski 1977). At exhaust flow rates greater than this, achieving uniform liquid distribution is difficult, unless additional weirs or baffles are used. To handle large exhaust flows, scrubbers designed with long, narrow, rectangular throats (Figure 3-4) have been used.

��

��

��


2.0-7/98 3-5

Figure 3-4. Spray venturi with rectangular throat

Simple venturis have fixed throat areas and cannot be used over a wide range of gas flow rates. Manufacturers have developed other modifications to the basic venturi design to maintain scrubber efficiency by changing the throat area for varying exhaust gas rates. Certain types of orifices (throat areas) that create more turbulence than a true venturi were found to be equally efficient for a given unit of energy consumed (McIlvaine Company 1974). Results of these findings led to the development of the annular-orifice, or adjustable-throat, venturi scrubber (Figure 3-5). The size of the throat area is varied by moving a plunger, or adjustable disk, up or down in the throat, thereby decreasing or increasing the annular opening. Gas flows through the annular opening and atomizes liquid that is sprayed onto the plunger or swirled in from the top.

��

Lesson 3 ___________________________________________________________________________________

3-6 2.0-7/98

Figure 3-5. Adjustable-throat venturi with plunger

��

��

��


2.0-7/98 3-7

Another adjustable-throat venturi is shown in Figure 3-6. In this scrubber, the throat area is varied by using a movable plate. A water-wash spray is used to continually wash collected material from the plate.

Figure 3-6. Adjustable-throat venturi with movable plate

��

� ��

��

Lesson 3 ___________________________________________________________________________________

3-8 2.0-7/98

Another modification can be seen in the venturi-rod or rod deck scrubber. By placing a number of pipes parallel to each other, a series of longitudinal venturi openings can be created as shown in Figure 3-7. The area between adjacent rods is a small venturi throat. Water sprays help prevent solids buildup. The principal atomization of the liquid occurs at the rods, where the high-velocity gas moving through spacings creates the small droplets necessary for fine particle collection. These rods must be made of abrasion-resistant material due to the high velocities present.

Figure 3-7. Venturi-rod scrubber

All venturi scrubbers require an entrainment separator because the high velocity of gas through the scrubber will have a tendency to exhaust the droplets with the clean gas stream. Cyclonic, mesh-pad, and blade separators (described in more detail in Lesson 8) are all used to remove liquid droplets from the flue gas and return the liquid to the scrubber water. Cyclonic separators, the most popular for use with venturi scrubbers, are connected to the venturi vessel by a flooded elbow (Figure 3-8). The liquid reduces abrasion of the elbow as the exhaust gas passes at high velocities from the venturi to the separator.


2.0-7/98 3-9

Figure 3-8. Flooded elbow leading into cyclonic separator

Particle Collection

Venturis are the most commonly used scrubber for particle collection and are capable of achieving the highest particle collection efficiency of any wet scrubbing system. As the exhaust stream enters the throat, its velocity increases greatly, atomizing and turbulently mixing with any liquid present. The atomized liquid provides an enormous number of tiny droplets for the dust particles to impact on. These liquid droplets incorporating the particles must be removed from the scrubber exhaust stream, generally by cyclonic separators.

Particle removal efficiency increases with increasing pressure drop because of increased turbulence due to high gas velocity in the throat. Venturis can be operated with pressure drops ranging from 12 to 250 cm (5 to 100 in.) of water. Most venturis normally operate with pressure drops in the range of 50 to 150 cm (20 to 60 in.) of water. At these pressure drops, the gas velocity in the throat section is usually between 30 and 120 m/s (100 to 400 ft/sec), or approximately 270 mph at the high end. These high pressure drops result in high operating costs.

The liquid-injection rate, or liquid-to-gas ratio (L/G), also affects particle collection. The proper amount of liquid must be injected to provide adequate liquid coverage over the throat area and make up for any evaporation losses. If there is insufficient liquid, then

��

��

Lesson 3 ___________________________________________________________________________________

3-10 2.0-7/98

there will not be enough liquid targets to provide the required capture efficiency. Most venturi systems operate with an L/G ratio of 0.4 to 1.3 L/m3 (3 to 10 gal/1000 ft3) (Brady and Legatski 1977). L/G ratios less than 0.4 L/m3 (3 gal/1000 ft3) are usually not sufficient to cover the throat, and adding more than 1.3 L/m3 (10 gal/1000 ft3) does not usually significantly improve particle collection efficiency.

Gas Collection

Venturi scrubbers can be used for removing gaseous pollutants; however, they are not used when removal of gaseous pollutants is the only concern. The high exhaust gas velocities in a venturi result in a very short contact time between the liquid and gas phases. This short contact time limits gas absorption. However, because venturis have a relatively open design compared to other scrubbers, they are very useful for simultaneous gaseous and particulate pollutant removal, especially when:

• Scaling could be a problem

• A high concentration of dust is in the exhaust stream

• The dust is sticky or has a tendency to plug openings

• The gaseous contaminant is very soluble or chemically reactive with the liquid

To maximize the absorption of gases, venturis are designed to operate at a different set of conditions from those used to collect particles. The gas velocities are lower and the liquid-to-gas ratios are higher for absorption. For a given venturi design, if the gas velocity is decreased, then the pressure drop (resistance to flow) will also decrease and vice versa. Therefore, by reducing pressure drop, the gas velocity is decreased and the corresponding residence time is increased. Liquid-to-gas ratios for these gas absorption applications are approximately 2.7 to 5.3 L/m3 (20 to 40 gal/1000 ft3). The reduction in gas velocity allows for a longer contact time between phases and better absorption. Increasing the liquid-to-gas ratio will increase the potential solubility of the pollutant in the liquid.


2.0-7/98 3-11

Maintenance Problems

The primary maintenance problem for venturi scrubbers is wear, or abrasion, of the scrubber shell because of high gas velocities. Gas velocities in the throat can reach speeds of 430 km/h (270 mph). Particles and liquid droplets traveling at these speeds can rapidly erode the scrubber shell. Abrasion can be reduced by lining the throat with silicon carbide brick or fitting it with a replaceable liner. Abrasion can also occur downstream of the throat section. To reduce abrasion here, the elbow at the bottom of the scrubber (leading into the separator) can be flooded (i.e. filled with a pool of scrubbing liquid). Particles and droplets impact on the pool of liquid, reducing wear on the scrubber shell. Another technique to help reduce abrasion is to use a precleaner (i.e., quench sprays or cyclone) to remove the larger particles.

The method of liquid injection at the venturi throat can also cause problems. Spray nozzles are used for liquid distribution because they are more efficient (have a more effective spray pattern) for liquid injection than weirs. However, spray nozzles can easily plug when liquid is recirculated. Automatic or manual reamers can be used to correct this problem. However, when heavy liquid slurries (either viscous or particle-loaded) are recirculated, open-weir injection is often necessary. Table 3-1 summarizes some of the operational problems associated with venturi scrubbers.

Table 3-1. Operational problems associated with venturi scrubbers

Problem Probable cause Corrective action

Low efficiency Low scrubbing-liquid flow rate

Check for plugged pipe or nozzles, incorrectly opened valves, or over-throttled pump-discharge valve.

Low pressure drop Check for low scrubbing liquor, low gas flow; inoperative or uncalibrated variable-throat controller; damaged variable-throat blade or disk.

Partially blocked entrainment separator

Check washdown sprays, spray liquor composition, and pH (for scaling).

Excessive gas flow Check for damper setting or venturi throat setting.

Inlet dust loading or particle size distribution different from that for which scrubber is designed Continued on next page

Lesson 3 ___________________________________________________________________________________

3-12 2.0-7/98

Table 3-1. (continued) Operational problems associated with venturi scrubbers


High exit-gas temperature Low scrubbing-liquid rate Check for plugged pipe or nozzles, incorrectly opened valves, or over-throttled pump-discharge valve.

High water-inlet temperature

Check and adjust makeup or heat-exchanger liquid flow rates.

High inlet-gas temperature

Check quench sprays, if applicable, or upstream equipment.

Exhaust gas liquor entrainment

Plugged entrainment separator

Check washdown sprays and spray pattern; use more flushing periods if necessary. Check liquor chemistry for scaling agents.

Plugged moisture-eliminator drain

Clean drain; add flushing water to continuously irrigate drain pipe.

Excessive gas flow Reduce gas flow.

Plugging or excessive wear of spray nozzles

Nozzle openings too small

Modify strainer/nozzle opening ratio so that nozzle holes are at least twice the diameter of strainer openings.

Solids concentration too high in spray liquor

Check bleed line for malfunctions; check for excessive dust loading. Check strainers.

Abrasives in spray liquor Remove abrasives or install abrasion-resistant linings.

Low pH in combination with abrasives is causing erosion or corrosion

Check separation equipment. Check for excessive dust loading in gas stream and for purge-line malfunctioning. Remove abrasives from liquor stream or install abrasion-resistant linings in wear zones. Add alkali for pH modification.

Excessive throat wear High solids recirculation Check bleed line for malfunctions. Check for excessive dust loading. Check strainers.

Excessive gas velocity Check throat pressure drop.

Corrosion/erosion Check separation equipment. Check for excessive dust loading in gas stream and for purge-line malfunctioning. Add alkali for pH modification. Install abrasion-resistant liners in high-wear zones if liquor modifications are not practical.



2.0-7/98 3-13

Table 3-1. (continued) Operational problems associated with venturi scrubbers


Erratic automatic-throat operation

Throat-mover malfunction

Remove from service; repair or replace

Sensor signal incorrect Check sensor taps for solids buildup. Check transmission tubing for liquid buildup or air leaks. Clean or repair sensor.

Damaged damper-disk mechanism

First make external inspection of drive train. If damaged area is not observed, shut unit down and make internal inspection using a throat-actuator manual override. Check for packing damage and excessively tight packing gland.

Low pressure drop Broken, leaking, or plugged static-tap line

Repair.

Low gas-flow rate Check gas flow against design. Check and, if necessary, adjust fan belt or speed. Check inlet duct for obstructions.

Fan overloads and shuts off

Excessive flow rate through fan

Check fan damper and variable-throat opening.

Low scrubbing-liquor flow rate

Check for plugged pipe or nozzles, incorrectly opened valves, or over-throttled pump-discharge valve.

Wet-dry interface buildup Particle buildup where gas goes from unsaturated to saturated condition

Install special inlets. Reduce dissolved solids in scrubbing liquor. Devote routine maintenance to removal of buildup.

Sources: Kelly 1978 and Anderson 2000 Co.

Summary

Venturi scrubbers can have the highest particle collection efficiencies (especially for very small particles) of any wet scrubbing system. They are the most widely used scrubbers because their open construction enables them to remove most particles without plugging or scaling. Venturis can also be used to absorb pollutant gases; however, they are not as efficient for this as are packed or plate towers (discussed later). The operating characteristics of venturi scrubbers are listed in Table 3-2.

Lesson 3 ___________________________________________________________________________________

3-14 2.0-7/98

Table 3-2. Operating characteristics of venturi scrubbers

Pollutant

Pressure drop

(∆p)

Liquid-to-gas ratio

(L/G)

Liquid-inlet pressure

(pL)

Removal efficiency

(%)

Gases 13-250 cm of water

(5-100 in. of water) 2.7-5.3 L/m3

(20-40 gal/1000 ft3)

< 7-100 kPa

(< 1-15 psig)

30-60% per venturi, depending on pollutant solubility

Particles 50-250 cm of water

[50-100 cm of water is common]

(20-100 in. of water)

[20-60 in. of water is common]

0.4-2.7 L/m3

(3.0-20.0 gal/1000 ft3)

[90-99% is typical]

Venturi scrubbers have been designed to collect particles at very high collection efficiencies, sometimes exceeding 99%. The ability of venturis to handle large exhaust volumes at high temperatures makes them very attractive to many industries; consequently, they are used to reduce particulate emissions in a number of industrial applications. This ability is particularly desirable for cement kiln emission reduction and for control of emissions from basic oxygen furnaces in the steel industry, where the exhaust gas enters the scrubber at temperatures greater than 350°C (660°F). Venturis are also used to control fly ash and sulfur dioxide emissions from industrial and utility boilers. A list of performance data for venturi scrubbers is given in Table 3-3.

Table 3-3. Performance data of typical venturi scrubbers

Application

Dust

Typical particle-size range

(µm)

∆p

(in. H2O)

Average collection efficiency

(%)

Iron and steel

Electric furnace Ferro-manganese and ferro silicon

0.1-1.0 30-50 92-99

BOF Iron oxide 0.5-2.0 40-60 98.5

Gray iron cupola Iron, coke, and silica

0.1-10.0 30-50 95



2.0-7/98 3-15

Table 3-3. (continued) Performance data of typical venturi scrubbers

Application

Dust

Typical particle-size range

(µm)

∆p

(in. H2O)

Average collection efficiency

(%)

Mineral products

Asphalt dryer Limestone and rock

1.0-50.0 10-15 98

Lime kiln Lime 1.0-50.0 15-25 99

Cement kiln Cement 0.5-55.0 10-15 97

Crushing and screening areas

Rock 0.5-100.0 6-20 99.9

Fertilizer manufacturing

Dryers Ammonium chloride fumes

0.05-1.0 10-20 85

Petroleum refining

Catalytic cracking unit

Catalyst dust 0.5-50.0 - 95

Chemical

Spray dryers Fumes and odor - 20-60 -

Phosphoric acid plant

Acid mist - 40-80 98

Pulp and paper

Lime kiln Soda ash 0.1-2.0 20-40 99

Recovery boiler Salt particles - 30-40 90

Boilers

Coal pulverizer Fly ash 20 (mass median)

15-40 97-99

Stoker Fly ash 75 (mass median)

10-15 97-99


Lesson 3 ___________________________________________________________________________________

3-16 2.0-7/98

Plate Towers

A plate tower is a vertical column with one or more plates (trays) mounted horizontally inside. As shown in Figure 3-9, the exhaust stream enters at the bottom and flows upward, passing through openings in the plates. Liquid enters at the top of the tower, traveling across each plate to a downcomer through which it reaches either the next plate below or the bottom of the tower. Pollutant collection occurs on each plate as the exhaust gas stream contacts and then atomizes the liquid flowing over each plate. Plate towers are very effective in removing gaseous pollutants and can be used simultaneously for particle removal. Plate towers may not be appropriate when particle removal is the only consideration.

Figure 3-9. Plate tower

Liquidinlet

Plate

Undersidewater sprays


2.0-7/98 3-17

Plates, or trays, are designed in a variety of ways. The ones most commonly used for industrial sources are the sieve, impingement, bubble-cap, and valve. The sieve, impingement, and bubble-cap plates do not have moving parts, while the valve plates have liftable caps above the opening in the plate. Plate openings for the sieve plate can range from 0.32 to 2.50 cm (0.125 to 1.0 in.) in diameter. Openings for the other plate designs are generally larger.

Sieve plates contain approximately 1800 to 9000 holes per square meter (600 to 3000 per square foot) of surface. Exhaust gas rises through these small holes and contacts the liquid atthe holes. The gas atomizes the liquid, forming a froth with droplets ranging from 10 to 100 µm in diameter. Particle collection efficiency increases as the size of the sieve opening decreases. Smaller openings cause the gas velocity to increase thereby forming smaller droplets (collection targets). Sieve plates with large openings will not become plugged as easily as will other plate designs. Figure 3-10 depicts gas-liquid contact on a sieve plate.

Figure 3-10. Sieve plate

Impingement plates are similar to sieve plates with the addition of an impaction target placed above each hole in the plate (Figure 3-11). The gas coming up through the hole forces the liquid on the plate up against the target (impingement surface). This design increases the mixing of the gas and liquid, provides an additional contact zone, and creates more liquid droplets.

Lesson 3 ___________________________________________________________________________________

3-18 2.0-7/98

Figure 3-11. Impingement plate

In the bubble-cap plate design, the exhaust gas enters each cap through a riser around each hole in the plate and exits from several slots in each cap (Figure 3-12). This combination of caps and risers creates a bubbly froth that allows good gas-liquid mixing, regardless of the liquid-to-gas ratio. In addition, the caps provide a longer gas-liquid contact than either sieves or impingement plates, thus increasing absorption efficiency. Plugging and corrosion can be a problem for bubble-cap plates because of this more complex design.

Figure 3-12. Bubble-cap plate


2.0-7/98 3-19

In the valve plate design, the exhaust gas passes through small holes in the plate, pushing up against a metal valve that covers each hole. The metal valve moves up and down with the gas flow. The valve is limited in its vertical movement by legs attached to the plate (Figure 3-13). Therefore, the liftable valve acts as a variable orifice. Caps are available in different weights to provide flexibility for varying exhaust gas flow rates. Floating valves increase gaseous pollutant collection efficiency by providing adequate gas-liquid contact time, regardless of the exhaust gas flow rate. This design is also suited for very small particle collection; however, valves will plug if large particles are in the exhaust stream. Wear and corrosion are also a problem for the retaining legs. Valve plates are more expensive than sieve and impingement plates, but less expensive than bubble-cap plates.

Figure 3-13. Valve plate

Particle Collection

Particles are collected in plate towers as the exhaust gas atomizes the liquid flowing over the holes in the plates. The atomized droplets serve as impaction targets for the particles. Plate towers are considered to be medium-energy scrubbers having moderate particle collection efficiencies. Collection efficiency does not significantly increase by increasing the number of plates over two or three. Collection efficiency can be improved by decreasing the hole size and increasing the number of holes per plate. This produces more liquid droplets of a smaller size and increases the gas velocity through the plate. However, it also increases the pressure drop of the system.

Lesson 3 ___________________________________________________________________________________

3-20 2.0-7/98

Gas Collection

Plate towers are very effective for removing gaseous pollutants from an exhaust stream. They can easily achieve greater than 98% removal in many applications. Absorption occurs as the exhaust stream bubbles up through the liquid on the plates and contacts the atomized liquid droplets. This action provides intimate contact between the exhaust gas and liquid streams, allowing the liquid on each plate to absorb the pollutant gas. Each plate acts as a separate absorption stage; therefore, absorption efficiency can be increased by adding plates. Absorption efficiency can also be improved by adding more liquid or by increasing the pressure drop across each plate, which increases gas-liquid contact.


Plate towers are susceptible to plugging and/or scale-buildup problems. If the exhaust stream contains a high concentration of dust or sticky materials, plate towers are generally not used. To clean the plates, access ports to each one are usually installed. In some systems, plates can actually be removed for cleaning. In addition, water sprays can be used to spray the underside of the lowest plate in the tower to eliminate the possibility of a wet-dry interface. If surfaces are not totally dry or wet, particles in these transition areas tend to become sticky and cause plugging.

Gas-liquid distribution may also be a problem with plate towers. First, if the plates are not level, gas-liquid contact will be reduced thus reducing collection efficiency. Second, a condition known as flooding (liquid buildup on a plate) can occur if either the liquid-injection or exhaust gas velocity is excessive, which will cause liquid to "stand" on the top of a plate. Flooding causes an increase in pressure drop and a decrease in gas-liquid mixing. Weeping (liquid dripping through the holes in the plates) can occur if the gas velocity is too low. Weeping also reduces gas-liquid contact. Table 3-4 summarizes some of the operational problems associated with plate towers.


2.0-7/98 3-21

Table 3-4. Operational problems associated with plate towers


Weeping Too large an open area (holes) on tray

Try bleeding in excess air or blocking off excess area.

Gas rate lower than design Check and adjust fan belt or speed.

Check inlet duct for obstructions.

Flooding Excessive liquid injected onto a plate

Excessive gas flowing through a plate, causing the liquid to "stand" on a plate

Reduce the liquid-injection rate.

Reduce the gas flow rate, if possible. If the gas flow rate is set because of process conditions (and it is excessive), an increase in tower design (size) may be necessary.

Plugging High solids concentration in scrubbing liquor

Check percentage of solids in recycle liquid. Check solid-separation equipment on recycle liquor. Use spray wash header. Clean trays periodically.

Little or no water flow to trays Check pump output; look for plugged piping, nozzles, incorrectly opened valves, or over throtteld pump-discharge valves.

Higher-than-expected particle content in inlet gas

Add prequench sprays.

Scale buildup Use a low-pH wash periodically to dissolve scale.

Poor liquid distribution Trays not level Check and level.

Liquid flow rate too high or gas flow rate too low

Check pump output; look for plugged piping, nozzles, incorrectly opened valves, or overthrottled pump-discharge valves.

Mechanical problems with trays Check for warped trays, loose fittings, and loose or broken baffle strips or caps.

Sources: Kelly 1978 and Buonicore 1982.

Lesson 3 ___________________________________________________________________________________

3-22 2.0-7/98

Summary

Plate towers are used most often when gaseous pollutant removal is the major concern. They can achieve greater than 98% collection efficiency, depending on the solubility of the gaseous pollutant. They can also be used to collect particles, but plugging and scale buildup problems may occur.

Plate towers are used successfully in flue gas desulfurization systems to remove sulfur dioxide emissions from utility boilers. They are also used to reduce pollutants emitted from petroleum refineries, chemical processes, acid manufacturing plants, and metal smelters. A summary of the operating characteristics of plate towers is given in Table 3-5.

Table 3-5. Operating characteristics of plate towers

Pollutant

Pressure drop (∆p)

Liquid-to-gas ratio (L/G)


(pL)

Removal efficiency

Applications

Gases

2.5-20 cm of water per tray

(1-8 in. of water)

0.7-2.0 L/m3

(5-15 gal/1000 ft3)

< 34.5 kPa

(< 5 psig)

Very effective (> 98%), depending on the solubility of the gaseous pollutant

Coal dryers

Copper roasting

Industrial boilers

Chemical process industries

Petroleum refineries

Particles [Normal pressure drops are 7.6 cm (3 in.) of water]

0.3-0.7 L/m3

(2-5 gal/1000 ft3)

> 1.0 µm diameter

Incineration processes



2.0-7/98 3-23

Orifice Scrubbers

In orifice scrubbers, the exhaust stream from the process is forced through a pool of liquid, usually water. The exhaust stream moves through restricted passages, or orifices, to disperse and atomize the water into droplets. These scrubbers are also called self-induced spray, inertial, or submerged orifice scrubbers.

Several orifice scrubbers designs are typically used. In each, the incoming exhaust stream is directed across or through a pool of water as shown in Figure 3-14. The high exhaust stream velocity, approximately 15.2 m/s (50 ft/sec), creates a large number of liquid droplets. Both particles and gaseous pollutants are collected as they are forced through the liquid pool and impact the droplets. However, these scrubbers are generally used for removing particles. Large particles are collected when they impact the liquid pool or its surface. Small particles are collected when they impact the droplets. Baffles, or air foils, are added to provide turbulent mixing of the exhaust stream and droplets.

Figure 3-14. Detail of orifice action

Lesson 3 ___________________________________________________________________________________

3-24 2.0-7/98

In the self-induced spray scrubber, the exhaust stream enters through a duct as shown in Figure 3-15. The exhaust stream is forced by baffles through a pool of liquid. Particles and gases are collected in the pool and by the droplets. Additional baffles placed in the path of the "clean" exhaust stream as it exits the vessel serve as impingement surfaces to remove entrained droplets.

Figure 3-15. Self-induced spray orifice scrubber

Particulate matter collected in the scrubber forms a sludge that must be disposed of. Sludge disposal involves removing and recycling large amounts of liquid, from 3.5 to 4.2 L/m3 (25 to 30 gal/1000 acfm). Some designs incorporate a sludge separation and removal system inside the scrubber to minimize the amount of water that must be removed. The water level inside the scrubber must be maintained during the sludge separation and removal cycle so that the unit can operate efficiently.


2.0-7/98 3-25

Particle Collection

Large particles in the incoming exhaust stream are collected as they impinge on the surface of the pool. Smaller particles are collected as they impact on the droplets produced by the high-velocity gas skimming over the liquid. Overall particle collection in an orifice scrubber depends on the level of the liquid. The level of the liquid determines the gas velocity (and, thus, the pressure drop) through the orifice. If the liquid level is low, gas velocities decrease because the orifice opening is larger. Lower velocities produce fewer droplets and ones that are larger in size, decreasing particle collection. A turn-down of the system, or reduction in gas volume, will also result in less atomization and produce larger droplets. It is recommended that gas velocities should not fluctuate by more than 10 to 15% of design values to provide maximum effectiveness (Bethea 1978).

Gas Collection

Orifice scrubbers are rarely used for absorption (McIlvaine Company 1974). However, because orifice scrubbers provide both thorough mixing of the gas and liquid, and large liquid-surface contact areas (many tiny droplets), these devices can be effective for removing gaseous pollutants that are already very soluble in the liquid. In reactive scrubbing, the gaseous pollutants chemically react with the scrubbing liquid. These reactions occasionally produce scale or sludge that can plug scrubber internals. The relatively large orifice openings will not plug as easily as those in plate towers.


The greatest problem for orifice scrubbers is maintaining the liquid at the proper level for a constant gas flow rate. Orifice scrubbers are designed to operate with a specific liquid level for a given gas velocity. If the gas flow decreases (or the liquid level decreases), less atomization occurs, thus reducing collection efficiency. If gas flow rate increases too much, it is possible to blow the liquid chamber dry (Bethea 1978). Systems are generally designed to operate at the upper end of the process exhaust rate and to introduce makeup air if the exhaust stream volume (velocity) becomes too low. Controlling the liquid level is much more difficult than maintaining a constant exhaust gas flow rate because of the turbulent condition of the water.

Lesson 3 ___________________________________________________________________________________

3-26 2.0-7/98

Summary

Orifice scrubbers are medium-energy devices with moderate collection efficiencies. The pressure drops across these devices are usually between 5 and 25 cm (2 and 10 in.) of water. The relatively large openings enable them to accommodate exhaust streams with high concentrations of particulate matter. Plugging by sticky or stringy material and scale buildup are not major problems. Because the gas stream is forced through a pool of liquid to create liquid droplets, spray nozzles are not necessary.

Orifice scrubbers are used mainly on metallurgical processes (crushing, screening, grinding, etc.), where the particles generated are mostly above 1 µm in diameter. Removal efficiencies depend on exhaust stream velocities. Reduction in exhaust stream velocities or liquid levels in the device will cause a reduction in collection efficiency. Table 3-6 lists operating characteristics of orifice scrubbers.

Table 3-6. Operating characteristics of orifice scrubbers

Pollutant


Liquid-to-gas ratio

(L/G)


(pL)

Removal efficiency

Applications

Gases

5-25 cm of water

0.07-0.7 L/m3

(0.5-5 gal/1000 ft3)

Not applicable (nozzles are not used)

Limited to very soluble gases or reactive scrubbing

Mining operations

Rock products industries

Foundries

Pulp and paper industries

Particles (2-10 in. of water)

1.3-5.3 L/m3

(10-40 gal/1000 ft3) for sludge removal and recycle

0.8-1 µm diameter




2.0-7/98 3-27


1. Label the following sections of the venturi.

2. In a venturi scrubber, the majority of pollutant removal occurs in the:

a. Converging section b. Throat c. Diverging section

3. A venturi scrubber in which liquid is introduced above the throat section:

a. Increases the likelihood of dust buildup on the throat b. Reduces dust buildup on throat surfaces c. Has the highest gas absorption capabilities of any wet collector d. None of the above

Lesson 3 ___________________________________________________________________________________

3-28 2.0-7/98

4. Many venturi scrubbers have devices by which the throat area can be varied to maintain:

a. Gas velocity through the throat b. Pressure drop c. Turbulence in the throat d. All of the above

5. True or False? Venturis are the most commonly used scrubber for particle collection.

6. In a venturi scrubber, particle collection increases with an increase in ____________________ ____________________.

7. To be effective in collecting particles, venturi scrubbers must operate at a pressure drop of:

a. 10 cm (5 in.) of water b. 50 cm (20 in.) of water c. 150 cm (60 in.) of water d. Any of the above

8. Venturi scrubbers are generally limited in their capability of removing gaseous pollutants because of:

a. The short gas-liquid contact time b. Low L/G ratios c. Small liquid droplets formed in the throat d. All of the above

9. Venturi scrubbers are useful for simultaneous gas and particle removal, especially when:

a. Scale buildup could be a problem b. A high concentration of dust is in the exhaust stream c. The dust is sticky or has a tendency to plug openings d. The gaseous pollutant is very soluble e. All of the above

10. To maximize gas collection in a venturi scrubber, the pressure drop is ____________________ and the L/G ratio is usually ____________________ when compared to operating conditions for particle collection.

a. Increased, increased b. Increased, decreased c. Decreased, increased d. Decreased, decreased

11. The primary maintenance problem for venturis is:

a. Plugging due to the many internal parts b. Weeping due to low gas flows c. Abrasion of the throat due to the high gas velocities d. All of the above


2.0-7/98 3-29

12. What does flooding the elbow between the venturi and the separator reduce?

a. Abrasion of the elbow b. Velocity of the gas stream c. Plugging in the elbow d. Pressure drop across the device

13. In general, venturis are more effective in removing ____________________ than they are in removing ____________________.

a. Gases, particles b. Particles, gases

Part 2

14. Name each of the following designs for plates in plate towers.

(a) (b)

(c)

Lesson 3 ___________________________________________________________________________________

3-30 2.0-7/98

15. True or False? For particle collection, efficiency does not significantly increase by increasing the number of plates in a plate tower.

16. In a plate tower, particle collection can usually be enhanced by ____________________ the hole size and/or ____________________ the number of holes per plate.

a. Increasing, increasing b. Increasing, decreasing c. Decreasing, increasing d. Decreasing, decreasing

17. For gaseous pollutant collection in a plate tower, absorption can usually be enhanced by:

a. Adding plates b. Increasing the amount of liquid c. Both a and b

18. If the plates in a tower are not ____________________, gas-liquid contact can be reduced, thus reducing collection efficiency.

a. The same size b. Level c. Staggered d. Omitted

19. In plate towers, liquid dripping through the holes in the plates, due to a low gas velocity, is referred to as:

a. Flooding b. Dropsy c. Weeping d. Drooling

20. List at least three common operational problems associated with plate towers. ________________________________________ ________________________________________ ________________________________________

Part 3

21. Although orifice scrubbers are produced in a variety of configurations, all are designed so that the exhaust gas stream:

a. Is split into two streams b. Travels concurrently with the liquid stream c. Breaks through a pool of liquid d. None of the above

22. True or False? In an orifice scrubber, all particles are collected as they impinge on the surface of the liquid.


2.0-7/98 3-31

23. The exhaust gas velocity (thus, the pressure drop) in an orifice scrubber is dictated by the:

a. Adjustable throat b. Level of liquid c. The plant foreman d. Precise calculations

24. In orifice scrubbers, a reduction in the design exhaust gas flow rate results in:

a. An increase in gas collection b. An increase in particle collection c. Less atomization and production of larger liquid droplets d. Both a and b

25. True or False? An orifice scrubber is capable of operating over a wide range of gas flow rates.

26. True or False? Orifice scrubbers are not primarily used for gas absorption.

27. The greatest problem with orifice scrubbers is:

a. Maintaining the proper liquid level b. Plugging c. Scale buildup d. Erosion

28. True or False? Plugging and scale buildup are not major operating problems with orifice scrubbers.

Lesson 3 ___________________________________________________________________________________

3-32 2.0-7/98


2.0-7/98 3-33


1. a. Converging b. Throat c. Diverging

2. b. Throat In a venturi scrubber, the majority of pollutant removal occurs in the throat.

3. b. Reduces dust buildup on throat surfaces A venturi scrubber in which liquid is introduced above the throat section reduces dust buildup on throat surfaces.

4. d. All of the above Many venturi scrubbers have devices by which the throat area can be varied to maintain:

• Gas velocity through the throat

• Pressure drop

• Turbulence in the throat

Lesson 3 ___________________________________________________________________________________

3-34 2.0-7/98

5. True Venturis are the most commonly used scrubber for particle collection.

6. Pressure drop In a venturi scrubber, particle collection increases with an increase in pressure drop.

7. d. Any of the above (depending on the specific scrubber design and application) To be effective in collecting particles, venturi scrubbers can operate at the following pressure drops (depending on the specific scrubber design and application):

• 10 cm (5 in.) of water



8. a. The short gas-liquid contact time Venturi scrubbers are generally limited in their capability of removing gaseous pollutants because of the short gas-liquid contact time.

9. e. All of the above Venturi scrubbers are useful for simultaneous gas and particle removal, especially in the following situations:

• Potential problem of scale buildup

• A high concentration of dust is in the exhaust stream

• The dust is sticky or has a tendency to plug openings

• The gaseous pollutant is very soluble

10. c. Decreased, increased To maximize gas collection in a venturi scrubber, the pressure drop is decreased and the L/G ratio is usually increased when compared to operating conditions for particle collection. Decreasing the pressure drop will result in a decrease in velocity and an increase in residence time. Higher liquid rates will provide better potential gas solubility.

11. c. Abrasion of the throat due to the high gas velocities The primary maintenance problem for venturis is abrasion of the throat due to the high gas velocities.

12. a. Abrasion of the elbow Flooding the elbow between the venturi and the separator reduces abrasion of the elbow.

13. b. Particles, gases In general, venturis are more effective in removing particles than they are in removing gases.


2.0-7/98 3-35

Part 2

14. a. Bubble-cap b. Sieve c. Valve

(a) (b)

(c)

15. True For particle collection, efficiency does not significantly increase by increasing the number of plates in a plate tower.

16. c. Decreasing, increasing In a plate tower, particle collection can usually be enhanced by decreasing the hole size and/or increasing the number of holes per plate. Both of these changes result in better gas-to-liquid contact. However, they also add more resistance (increase pressure drop) and therefore higher operating costs.

Lesson 3 ___________________________________________________________________________________

3-36 2.0-7/98

17. c. Both a and b For gaseous pollutant collection in a plate tower, absorption can usually be enhanced by adding plates and increasing the amount of liquid.

18. b. Level If the plates in a tower are not level, gas-liquid contact can be reduced, thus reducing collection efficiency.

19. c. Weeping In plate towers, liquid dripping through the holes in the plates, due to a low gas velocity, is referred to as weeping.

20. Weeping Plugging Flooding Three common operational problems associated with plate towers are weeping, plugging, and flooding.

Part 3

21. c. Breaks through a pool of liquid Although orifice scrubbers are produced in a variety of configurations, all are designed so that the exhaust gas stream breaks through a pool of liquid.

22. False In an orifice scrubber, only large particles are collected as they impinge on the surface of the liquid. Small particles are collected by liquid droplets produced by the exhaust stream.

23. b. Level of liquid The exhaust gas velocity (thus, the pressure drop) in an orifice scrubber is dictated by the level of liquid.

24. c. Less atomization and production of larger liquid droplets In orifice scrubbers, a reduction in the design exhaust gas flow rate results in less atomization and production of larger liquid droplets which could potentially lower the gas and/or particle collection efficiency.

25. False Exhaust gas velocities should not fluctuate greatly in orifice scrubbers.

26. True Orifice scrubbers are NOT primarily used for gas absorption.

27. a. Maintaining the proper liquid level The greatest problem with orifice scrubbers is maintaining the proper liquid level.

28. True Plugging and scale buildup are not major operating problems with orifice scrubbers because of their open design.


2.0-7/98 3-37

Bibliography

Anderson 2000 Company. Venturi scrubbing equipment. Engineering Manual with Operating and Maintenance Instructions. Atlanta: Anderson Company.


Brady, J. D., and L. K. Legatski. 1977. Venturi scrubbers. In P. N. Cheremisinoff and R. A. Young (Eds.), Air Pollution Control and Design Handbook. Part 2. New York: Marcel Dekker.

Buonicore, A. J. 1982. Wet scrubbers. In L. Theodore and A. J. Buonicore (Eds.), Air Pollution Control Equipment, Design, Selection, Operation and Maintenance. Englewood Cliffs: Prentice-Hall.

Calvert, S. 1977. How to choose a particulate scrubber. Chemical Engineering. 84:133-140.

Johnstone, H. F., and M. H. Roberts. 1949. Deposition of aerosol particles from moving gas streams. Industrial and Engineering Chemistry. 41:2417-2423.

Kelly, J. W. 1978, December 4. Maintaining venturi-tray scrubbers. Chemical Engineering.

McIlvaine Company. 1974. The Wet Scrubber Handbook. Northbrook, IL: McIlvaine Company.



Lesson 3 ___________________________________________________________________________________

3-38 2.0-7/98

2.0-7/98 4-1

Lesson 4 Liquid-Phase Contacting Scrubbers

Goal To familiarize you with the operation, collection efficiency, and major maintenance problems of liquid-phase contacting scrubbers.


1. List two liquid-phase contacting scrubbers and briefly describe the operation of each

2. For each scrubber above, identify the range of operating values for pressure drop, liquid-to-gas ratio, as well as the collection efficiency for particles and gases

3. Describe typical operating and maintenance problems associated with each design of liquid-phase contacting scrubbers

Introduction

The previous lesson described scrubbers that use the process gas stream as energy to atomize liquid into collection droplets. Energy can also be applied to a scrubbing system by injecting liquid at high pressure through specially designed nozzles. Nozzles produce droplets that fan out into a spray in the scrubber chamber. Droplets act as targets for collecting particles and/or absorbing gas in a pollutant exhaust stream. In liquid-phase contacting scrubbers, the liquid-inlet pressure provides the major portion of the energy required for contacting the gas (exhaust stream) and liquid phases.

Two liquid-phase contacting scrubbers are the spray tower and the ejector venturi. Many other scrubber designs also incorporate sprays produced by nozzles, but in those scrubbers, the sprays are used to clean trays or to wet scrubber surfaces and orifices, and not to provide the gas-liquid contact in the system.

Spray Towers

Spray towers, or chambers, are constructed very simply consisting of empty cylindrical vessels made of steel or plastic and nozzles that spray liquid into the vessels. The exhaust stream usually enters the bottom of the tower and moves upward, while liquid is sprayed downward from one or more levels. This flow of exhaust gas and liquid in opposite direction

Lesson 4 ___________________________________________________________________________________

4-2 2.0-7/98

is called countercurrent flow. Figure 4-1 shows a typical countercurrent-flow spray tower. Countercurrent flow exposes the exhaust gas with the lowest pollutant concentration to the freshest scrubbing liquid.

Figure 4-1. Countercurrent-flow spray tower

Many nozzles are placed across the tower at different heights to spray all of the exhaust gas as it moves up through the tower. The major purpose of using many nozzles is to form a tremendous amount of fine droplets for impacting particles and to provide a large surface area for absorbing gas. Theoretically, the smaller the droplets formed, the higher the collection efficiency achieved for both gaseous and particulate pollutants. However, the liquid droplets must be large enough to not be carried out of the scrubber by the exhaust stream. Therefore, spray towers use nozzles to produce droplets that are usually 500 to 1000 µm in diameter. Although small in size, these droplets are large compared to those created in the venturi scrubbers that are 10 to 50 µm in size. The exhaust gas velocity is kept low, from 0.3 to 1.2 m/s (1 to 4 ft/sec) to prevent excess droplets from being carried out of the tower. In order to maintain low exhaust velocities, spray towers must be larger than other scrubbers that handle similar exhaust stream flow rates. Another problem occurring in spray towers is that after the droplets fall short distances, they tend to agglomerate or hit the walls of the tower. Consequently, the total liquid surface area for contact is reduced, thus reducing the collection efficiency of the scrubber.

��

Liquid-Phase Contacting Scrubbers ___________________________________________________________________________________

2.0-7/98 4-3

In addition to a countercurrent-flow configuration, the flow in spray towers can be either a cocurrent or crosscurrent configuration. In cocurrent-flow spray towers, the exhaust gas and liquid flow in the same direction. Because the exhaust gas stream does not "push" against the liquid sprays, the exhaust gas velocities through the vessels are higher than in countercurrent-flow spray towers. Consequently, cocurrent-flow spray towers are smaller than countercurrent-flow spray towers treating the same amount of exhaust flow.

In crosscurrent-flow spray towers, also called horizontal-spray scrubbers, the exhaust gas and liquid flow in directions perpendicular to each other (Figure 4-2). In this vessel, the exhaust gas flows horizontally through a number of spray sections. The amount and quality of liquid sprayed in each section can be varied, usually with the cleanest liquid (if recycled liquid is used) sprayed in the last set of sprays.

Figure 4-2. Crosscurrent-flow spray tower

Particle Collection

Spray towers are low-energy scrubbers. Contacting power is much lower than in venturi scrubbers, and the pressure drops across such systems are generally less than 2.5 cm (1 in.) of water. The collection efficiency for small particles is correspondingly lower than in more energy-intensive devices. They are adequate for the collection of coarse particles larger than 10 to 25 µm in diameter, although with increased liquid inlet nozzle pressures, particles with diameters of 2.0 µm can be collected. Smaller droplets can be formed by higher liquid pressures at the nozzle. The highest collection efficiencies are achieved when small droplets are produced and the difference between the velocity of the droplet and the velocity of the upward-moving particles is high. Small droplets, however, have small settling velocities, so there is an optimum range of droplet sizes for scrubbers that work by this mechanism. Stairmand (1956) found this range of droplet sizes to be between 500 to 1000 µm for gravity-spray (counter current) towers. The injection of water at very high pressures, 2070 to 3100 kPa (300 to 450 psi), creates a fog of very fine droplets. Higher particle-collection efficiencies can be achieved in such cases since collection mechanisms other than inertial impaction occur (Bethea 1978). However, these

� ��

��

��

Lesson 4 ___________________________________________________________________________________

4-4 2.0-7/98

spray nozzles may use more power to form droplets than would a venturi operating at the same collection efficiency.

Gas Collection

Spray towers can be used for gas absorption, but they are not as effective as packed or plate towers. (Packed towers will be discussed in the next lesson.) Spray towers can be very effective in removing pollutants if the pollutants are highly soluble or if a chemical reagent is added to the liquid. For example, spray towers are used to remove HCl gas from the tail-gas exhaust in manufacturing hydrochloric acid. In the production of superphosphate used in manufacturing fertilizer, SiF4 and HF gases are vented from various points in the processes. Spray towers have been used to remove these highly soluble compounds. Spray towers are also used for odor removal in bone meal and tallow manufacturing industries by scrubbing the exhaust gases with a solution of KMnO4. Because of their ability to handle large exhaust gas volumes in corrosive atmospheres, spray towers are also used in a number of flue gas desulfurization systems as the first or second stage in the pollutant removal process.

In a spray tower, absorption can be increased by decreasing the size of the liquid droplets and/or increasing the liquid-to-gas ratio (L/G). However, to accomplish either of these, an increase in both power consumed and operating cost is required. In addition, the physical size of the spray tower will limit the amount of liquid and the size of droplets that can be used.


The main advantage of spray towers over other scrubbers is their completely open design; they have no internal parts except for the spray nozzles. This feature eliminates many of the scale buildup and plugging problems associated with other scrubbers. The primary maintenance problems are spray-nozzle plugging or eroding, especially when using recycled scrubber liquid. To reduce these problems, a settling or filtration system is used to remove abrasive particles from the recycled scrubbing liquid before pumping it back into the nozzles.

Summary

Spray towers are inexpensive control devices primarily used for gas conditioning (cooling or humidifying) or for first-stage particle or gas removal. They are also being used in many flue gas desulfurization systems to reduce plugging and scale buildup by pollutants. Many scrubbing systems use sprays either prior to or in the bottom of the primary scrubber to remove large particles that could plug it. Spray towers have been used effectively to remove large particles and highly soluble gases. The pressure drops across the towers are very low [usually less than 2.5 cm (1.0 in.) of water]; thus, the scrubber operating costs are relatively low. However, the liquid pumping costs can be very high.

Spray towers are constructed in various sizes small ones to handle small exhaust flows of 0.05 m3/s (100 cfm) or less, and large ones to handle large exhaust flows of 50 m3/s (100,000 cfm) or greater. Because of the low gas velocity required, units handling large


2.0-7/98 4-5

exhaust flow rates tend to be large in size. Operating characteristics of spray towers are presented in Table 4-1.

Table 4-1. Operating characteristics of spray towers

Pollutant




(pL)

Removal efficiency

Applications

Gases

1.3-7.6 cm of water

0.07-2.70 L/m3

(0.5-20 gal/1000 ft3)

70-2800 kPa

50-90+% (high efficiency only when the gas is very soluble)

Mining industries

Chemical process industry

Particles (0.5-3.0 in. of water)

(5 gal/1000 ft3 is normal; >10 when using pressure sprays)

(10-400 psig)

2-8 µm diameter

Boilers and incinerators

Iron and steel industry


Lesson 4 ___________________________________________________________________________________

4-6 2.0-7/98

Ejector Venturis

The ejector, or jet, venturi scrubber uses a preformed spray, as does the simple spray tower. The difference is that only a single nozzle is used instead of many nozzles. This nozzle operates at higher pressures and higher injection rates than those in most spray chambers. The high-pressure spray nozzle (up to 689 kPa or 100 psig) is aimed at the throat section of a venturi constriction. Figure 4-3 illustrates the ejector venturi design.

Figure 4-3. Ejector venturi scrubber

The ejector venturi is unique among available scrubbing systems since it can move the process gas without the aid of a blower or fan. The liquid spray coming from the nozzle creates a partial vacuum in the side duct of the scrubber. This has the same effect as the water aspirator used in high school chemistry labs to pull a small vacuum for filtering precipitated materials (due to the Bernoulli effect). This partial vacuum can be used to move the process gas through the venturi as well as through the facility's process system. In the case of explosive or extremely corrosive atmospheres, the elimination of a fan in the system can avoid many potential problems.

The energy for the formation of scrubbing droplets comes from the injected liquid. The high-pressure sprays passing through the venturi throat form numerous fine liquid droplets that

��


2.0-7/98 4-7

provide turbulent mixing between the gas and liquid phases. Very high liquid-injection rates are used to provide the gas-moving capability and higher collection efficiencies. As with other types of venturis, a means of separating entrained liquid from the gas stream must be installed. Entrainment separators are commonly used to remove remaining small droplets.

Particle Collection

Ejector venturis are effective in removing particles larger than 1.0 µm in diameter. These scrubbers are not used on submicrometer-sized particles unless the particles are condensable (Gilbert 1977). Particle collection occurs primarily by impaction as the exhaust gas (from the process) passes through the spray.

The turbulence that occurs in the throat area also causes the particles to contact the wet droplets and be collected. Particle collection efficiency increases with an increase in nozzle pressure and/or an increase in the liquid-to-gas ratio. Increases in either of these two operating parameters will also result in an increase in pressure drop for a given system. Therefore, an increase in pressure drop also increases particle collection efficiency. Ejector venturis operate at higher L/G ratios than most other particle scrubbers (i.e. 50 to 100 gal/1000 ft3 compared to 3-20 gal/1000 ft3 for most other designs).

Gas Collection

Ejector venturis have a short gas-liquid contact time because the exhaust gas velocities through the vessel are very high. This short contact time limits the absorption efficiency of the system. Although ejector venturis are not used primarily for gas removal, they can be effective if the gas is very soluble or if a very reactive scrubbing reagent is used. In these instances, removal efficiencies of as high as 95% can be achieved (Gilbert 1977).


Ejector venturis are subject to abrasion problems in the high-velocity areas - nozzle and throat. Both must be constructed of wear-resistant materials because of the high liquid-injection rates and nozzle pressures. Maintaining the pump that recirculates liquid is also very important. In addition, the high gas velocities necessitate the use of entrainment separators to prevent excessive liquid carryover. The separators should be easily accessible or removable so that they can be cleaned if plugging occurs.

Summary

Because of their open design and the fact that they do not require a fan, ejector venturis are capable of handling a wide range of corrosive and/or sticky particles. However, they are not very effective in removing submicrometer particles. They have an advantage in being able to handle small, medium and large exhaust flows. They can be used singly or in multiple stages of two or more in series, depending on the specific application. Multiple-stage systems have been used where extremely high collection efficiency of particles or gaseous pollutants was necessary. Multiple-stage systems provide increased gas-liquid contact time, thus increasing absorption efficiency. Table 4-2 lists the operating parameters for ejector venturis.

Lesson 4 ___________________________________________________________________________________

4-8 2.0-7/98

Table 4-2. Operating characteristics of ejector venturis

Pollutant




(pL)

Removal efficiency

Applications

Gases

1.3-13 cm of water

7-13 L/m3

100-830 kPa

95% for very soluble gases

Pulp and paper industry

Chemical process industry

Particles (0.5-5 in. of water)

(50-100 gal/1000 ft3)

(15-120 psig) 1 µm diameter Food industry

Metals-processing industry



2.0-7/98 4-9


1. The liquid and exhaust gas flow in opposite directions in a ____________________ scrubber.

a. Cocurrent b. Countercurrent c. Crosscurrent d. Crosshatch

2. In a spray tower, the ____________________ the droplet is, the higher the theoretical collection efficiency will be.

a. Smaller b. Larger c. Higher d. Lower

3. Gas velocities in spray towers are usually kept very ____________________ to prevent excessive liquid from becoming entrained in the exhaust gas stream leaving the tower.

a. High b. Low c. Stable d. None of the above

4. True or False? In general, countercurrent-flow spray towers must be larger than crosscurrent- or cocurrent-flow spray towers to accommodate the same volumetric flow rate.

5. In a spray tower, gas collection can be increased by increasing:

a. The size of the liquid droplets b. The liquid-to-gas ratio (L/G) c. The gas velocity d. All of the above

6. Because spray towers contain few internal parts, they:

a. Eliminate many potential problems due to plugging and scale buildup b. Have low pressure drops c. Are relatively simple and inexpensive d. All of the above

7. What are the main maintenance problems with spray towers? ________________________________________ ________________________________________

Lesson 4 ___________________________________________________________________________________

4-10 2.0-7/98

8. In spray towers, the pressure drops across the tower are usually ____________________ and the liquid pumping costs can be very ____________________.

a. High, high b. High, low c. Low, high d. Low, low

Part 2

9. The ejector, or jet, venturi scrubber uses ____________________ to move the process exhaust stream.

a. Multiple nozzles b. A single high-pressure nozzle c. A compressor d. A fan

10. For ejector venturis, particle collection efficiencies increase with an increase in:

a. Nozzle pressure b. Liquid-to-gas ratio (L/G) c. Pressure drop d. All of the above

11. Ejector venturis are subject to abrasion problems in the:

a. Throat b. Nozzle c. Packing area d. a and b, only

12. True or False? Because of their open design and the fact that they do not require a fan, ejector venturis are capable of handling a wide range of corrosive and/or sticky particles.


2.0-7/98 4-11


1. b. Countercurrent The liquid and exhaust gas flow in opposite directions in a countercurrent scrubber.

2. a. Smaller In a spray tower, the smaller the droplet is, the higher the theoretical collection efficiency will be. However, if the droplets are too small and the gas flow up the tower is too fast, then the droplets can be carried out of the tower.

3. b. Low Gas velocities in spray towers are usually kept very low to prevent excessive liquid from becoming entrained in the exhaust gas stream leaving the tower.

4. True In general, countercurrent-flow spray towers must be larger than crosscurrent- or cocurrent-flow spray towers to accommodate the same volumetric flow rate.

5. b. The liquid-to-gas ratio (L/G) In a spray tower, gas collection can be increased by increasing the liquid-to-gas ratio (L/G).

6. d. All of the above Because spray towers contain few internal parts, they:

• Eliminate many potential problems due to plugging and scale buildup

• Have low pressure drops

• Are relatively simple and inexpensive

7. Plugging Erosion of the nozzle The main maintenance problems with spray towers are plugging and erosion of the nozzle.

8. c. Low, high In spray towers, the pressure drops across the tower are usually low and the liquid pumping costs can be very high.

Part 2

9. b. A single high-pressure nozzle The ejector, or jet, venturi scrubber uses a single high-pressure nozzle to move the process exhaust stream.

10. d. All of the above For ejector venturis, particle collection efficiencies increase with an increase in:

• Nozzle pressure

• Liquid-to-gas ratio (L/G)

• Pressure drop

Lesson 4 ___________________________________________________________________________________

4-12 2.0-7/98

11. d. a and b, only Ejector venturis are subject to abrasion problems in the throat and nozzle.

12. True Because of their open design and the fact that they do not require a fan, ejector venturis are capable of handling a wide range of corrosive and/or sticky particles.


2.0-7/98 4-13

Bibliography


Gilbert, J. W. 1977. Jet venturi fume scrubbing. In P. N. Cheremisinoff and R. A. Young (Eds.), Air Pollution Control and Design Handbook. Part 2. New York: Marcel Dekker.




Lesson 4 ___________________________________________________________________________________

4-14 2.0-7/98

2.0-7/98 5-1

Lesson 5 Wet-Film (Packed Tower) Scrubbers

Goal To familiarize you with the operation, collection efficiency, and major maintenance problems of packed tower scrubbers.


1. Describe the operation of packed tower scrubbers

2. Describe at least three different gas-liquid flow arrangements (designs) for packed tower scrubbers

3. Describe major operating and maintenance problems associated with each packed scrubber design

4. Identify the range of operating values for pressure drop, liquid-to-gas ratio, as well as the collection efficiency of packed tower scrubbers for particles and gases

Introduction

In packed tower or wet-film scrubbers, liquid is sprayed or poured over packing material contained between support trays. A liquid film coats the packing through which the exhaust gas stream is forced. Pollutants are collected as they pass through the packing, contacting the liquid film. Therefore, both gas and liquid phases provide energy for the gas-liquid contact.

A wet-film scrubber uses packing to provide a large contact area between the gas and liquid phases, turbulent mixing of the phases, and sufficient residence time for the exhaust gas to contact the liquid. These conditions are ideal for gas absorption. Large contact area and good mixing are also good for particle collection; however, once collected, the particles tend to accumulate and plug the packing bed. The exhaust gas is forced to make many changes in direction as it winds through the openings of the packed material. Large particles unable to follow the streamlines, hit the packing and are collected in the liquid. As this liquid drains through the packing bed, the collected particles may accumulate, thus plugging the void spaces in the packed bed. Therefore, wet-film scrubbers are not used when particle removal is the only concern. Many other scrubber designs achieve better particle removal for the same power input (operating costs).

Lesson 5 ___________________________________________________________________________________

5-2 2.0-7/98

Gas Collection

For gas absorption, packed scrubbers are the most commonly used devices. The wet film covering the packing enhances gas absorption several ways by providing:

• A large surface area for gas-liquid contact

• Turbulent contact (good mixing) between the two phases

• Long residence time and repetitive contact

Because of these features, packed towers are capable of achieving high removal efficiencies for many different gaseous pollutants.

Numerous operating variables affect absorption efficiency. Of primary importance is the solubility of the gaseous pollutants. Pollutants that are readily soluble in the scrubbing liquid can be easily removed under a variety of operating conditions. Some other important operating variables are discussed below.

Gas velocity - The rate of exhaust gas from the process determines the scrubber size to be used. The scrubber should be designed so that the gas velocity through it will promote good mixing between the gas and liquid phases. However, the velocity should not be too fast to cause flooding.

Liquid-injection rate - Generally, removal efficiency is increased by an increase in the liquid-injection rate to the vessel. The amount of liquid that can be injected is limited by the dimensions of the scrubber. Increasing liquid-injection rates will also increase the operating costs. The optimum amount of liquid injected is based on the exhaust gas flow rate.

Packing size - Smaller packing sizes offer a larger surface area, thus enhancing absorption. However, smaller packing fits more tightly, which decreases the open area between packing, thus increasing the pressure drop across the packing bed.

Packing height - As packing height increases, total surface area and residence time increases, enhancing absorption. However, more packing necessitates a larger absorption system, which increases capital cost.

Tower Designs

Packed towers are typically designated by the flow arrangement used for gas-liquid contact or by the material used as packing for the bed. The most common flow configuration for packed towers is countercurrent flow. Figure 5-1 shows a packed tower with this arrangement. The exhaust stream being treated enters the bottom of the tower and flows upward over the packing material. Liquid is introduced at the top of the packing by sprays or weirs, and it flows downward over the packing material. As the exhaust stream moves up through the packing, it is forced to make many winding changes in direction, resulting in intimate mixing of both the exhaust gas and liquid streams. This countercurrent-flow arrangement results in the highest theoretically achievable efficiency. The most dilute gas is contacted with the purest absorbing liquor, providing a maximized concentration difference (driving force) for the entire length of the column. In the other two flow arrangements, the scrubbing liquid could theoretically reach the same concentration as the flue gas (since they are moving in similar directions) and therefore absorption would stop.

Wet-Film (Packed Tower) Scrubbers

2.0-7/98 5-3

Figure 5-1. Countercurrent-flow packed tower

The countercurrent-flow packed tower does not operate effectively if there are large variations in the liquid or gas flow rates. If either the liquid-injection rate or the gas flow rate through the packing bed is too high, a condition called flooding may occur. Flooding is a condition where the liquid is "held" in the pockets, or void spaces, between the packing and does not drain down through the packing. Flooding can be reduced by reducing the gas velocity through the bed or by reducing the liquid-injection rate.

In another flow arrangement used with packed towers, cocurrent flow, both the exhaust gas and liquid phases enter at the top of the absorber and move downward over the packing material. This allows the absorber to operate at higher liquid and gas flow rates since flooding is not a problem. The pressure drop is lower than with countercurrent flow since both streams move in the same direction. The major disadvantage is that removal efficiency is very limited due to the decreasing driving force (concentration differential) as the streams travel down through the column. This limits the areas of application for cocurrent absorbers. They are used almost exclusively in situations where limited equipment space is available, since the tower diameter is smaller than that for countercurrent or plate towers for equivalent flow rates. Cocurrent flow is illustrated in Figure 5-2.

��

��

��

Lesson 5 ___________________________________________________________________________________

5-4 2.0-7/98

Figure 5-2. Cocurrent-flow packed tower

In packed towers using the crossflow arrangement, the exhaust gas stream moves horizontally through the packed bed. The bed is irrigated by the scrubbing liquid flowing down through the packing material. The liquid and exhaust gas flow in directions perpendicular to each other. A typical crossflow packed tower is shown in Figure 5-3. Inlet sprays aimed at the face of the bed may also be included. If included, these sprays scrub both the entering gas and the face of the packed bed. The leading face of the packed bed is slanted in the direction of the oncoming gas stream. This ensures complete wetting of the packing by allowing time for the liquid at the front face of the packing to drop to the bottom before being pushed back by the entering gas.

��

��


2.0-7/98 5-5

Figure 5-3. Crossflow packed tower

Crossflow absorbers can be designed to be smaller and have a lower pressure drop than any other packed or plate tower for the same application (i.e. removal efficiency and flow rates). In addition, they are better suited than other wet-film scrubbers to handle exhaust streams with high particle concentrations. By adjusting the liquid flow rate, incoming particles can be removed and washed away in the front half of the bed. This also results in a liquid savings by enabling the crossflow packed tower to use less liquid in the rear sprays. This practice is carried one step further by actually constructing the tower into sections as shown in Figure 5-4. The front section can be equipped with water sprays and used for particulate matter removal. In the second section, sprays may contain a reagent in the scrubbing liquor for gas removal. The last section can be left dry to act as an entrainment separator. Crossflow packed towers do require complex design procedures since concentration gradients exist in two directions in the liquid: from top to bottom and from front to rear.

Figure 5-4. Three-bed crossflow packed tower

��

��

��

��

��

Lesson 5 ___________________________________________________________________________________

5-6 2.0-7/98

Another crossflow packed tower is the fiber-bed scrubber. The fiber-bed scrubber has packed beds that are made with fibrous material such as fiberglass or plastic (Figure 5-5). Liquid is sprayed onto the fiber beds to provide a wetted surface for pollutant removal and to wash away any collected material.

Figure 5-5. Fiber-bed scrubber

Packing Material

Packing material is the heart of the tower. It provides the surface over which the scrubbing liquid flows, presenting a large area for mass transfer to occur. Packing material represents the largest material cost of the packed tower. Pictured in Figure 5-6 are some of the more commonly used packings. These materials were originally made of stoneware, porcelain, or metal, but presently, a large majority are made of high-density thermoplastics (polyethylene and polypropylene). A specific packing is described by its trade name and overall size. For example, a column can be packed with 5-cm (2-in.) Raschig rings or 2.5-cm (1-in.) Tellerette packing. The overall dimensions of packing materials normally range from 0.6 to 10 cm (0.25 to 4 in.).

Figure 5-6. Common packing materials

��

��

��

!�� !�"��


2.0-7/98 5-7

Specific packing selected for an industrial application depends on the nature of the contaminants, geometric mode of contact, size of the absorber, and scrubbing objectives. The following factors provide a general guide for selecting packing materials (MacDonald 1977):

Cost - Generally, plastic packing is less expensive than metal packing, with ceramic packing being the most expensive. Packing costs are expressed in dollars per cubic meter ($/m3).

Low pressure drop - Pressure drop is a function of the volume of void space in a tower when filled with packing: generally, the larger the packing size for a given bed size, the smaller the pressure drop becomes.

Corrosion resistance - Ceramic or porcelain packings are commonly used in a very corrosive atmosphere.

Large specific area - A large surface area per cubic foot of packing, m2/m3 (ft2/ft3), is desirable for mass transfer.

Structural strength - Packing must be strong enough to withstand normal loads during installation, service, physical handling, and thermal fluctuations. Ceramic packing may crack under sudden temperature changes.

Weight - Heavier packing may require additional support materials or heavier tower construction. Plastics have a big advantage in this area since they are much lighter than either ceramic or metal packings.

Design flexibility - The efficiency of a scrubber changes as the liquid and gas flow rates vary. Packing material must be able to handle the process changes without substantially affecting removal efficiency.

Arrangement - Packing material may be arranged in an absorber in one of two ways. The packing may be dumped into the column randomly or, in certain cases, systematically stacked, as bricks are laid atop each other. Randomly packed towers provide a higher surface area, m2/m3 (ft2/ft3), but also cause a higher pressure drop than stacked packing. In addition to the lower pressure drop, the stacked packing provides better liquid distribution over the entire surface of the packing. However, the large installation costs required to stack the packing material usually make it impractical.

Exhaust Gas Distribution

Uniform distribution of the exhaust gas through the packed beds is very important for efficient pollutant removal. This is accomplished by properly designing the support trays that contain the packing in the bed. The support trays are essentially metal plates, or grids, that support the packing while allowing the exhaust gas to flow evenly into the bed. If the packed tower has multiple packing sections, each support grid acts as a distribution baffle, directing the exhaust gas into the next packing section.

Liquid Distribution

As stated previously, one of the keys to effective packed tower operation is to intimately contact the gas stream with the liquid stream. This contact must be maintained throughout the entire column length. No packing material will adequately distribute liquid poured onto it at

Lesson 5 ___________________________________________________________________________________

5-8 2.0-7/98

only one point. Liquid introduced into the tower in this manner tends to flow down over a relatively small cross section of the tower diameter. Known as liquid channeling, the liquid flows in little streams down through the tower without wetting the entire packing area. Liquid should be distributed over the entire cross-sectional top of the packing.

Once the liquid is distributed over the packing, it flows down (by the force of gravity) through the packing, following the path of least resistance. The liquid tends to flow toward the tower wall, where the void spaces are greater than in the center. Once the liquid hits the wall, it flows straight down the tower from that point (liquid channeling). A way must be provided to redirect the liquid from the tower wall back to the center of the column. This is usually done by using liquid redistributors, which funnel the liquid back over the entire surface of packing. It is recommended that redistributors be placed at intervals of no more than 3 m (10 ft) or 5 tower diameters, whichever is smaller (Zenz 1972).

Liquid can be distributed over the packing material by one of three devices: weirs, tubes, or spray nozzles. Figure 5-7 shows both the commonly used weir and perforated-tube liquid distributors. In the weir design, liquid is introduced into a trough with holes at the top. The liquid fills to the top and "spills" over onto the packing or another trough for redistribution. These weir designs have the advantage of being open and not plugged easily. However when installed they must be level or else the liquid will not be evenly distributed.

The perforated-tube provides good liquid distribution patterns, however the holes are subject to plugging if any particles or contaminants are in the liquid. The drilled tube is often buried within the packing bed. This allows the liquid coming out of the holes to be distributed over the packing without being blown against the side walls of the tower. Burying the tube also allows the packing above the tube to act as an entrainment separator for countercurrent flow towers.

Figure 5-7. Two types of liquid distributors

Packed towers, designed with spray nozzles to distribute liquid, operate better with a few large nozzles than with many small nozzles. Large nozzles are less susceptible to plugging. Small nozzles that produce a finer spray are not needed in a packed tower because pollutant collection occurs on the wetted packing and not by the liquid droplets. The advantages and disadvantages of each liquid distributor are listed in Table 5-1.


2.0-7/98 5-9

Table 5-1. Liquid distributors for packed towers

Distributor Advantages Disadvantages

Weirs Handle dirty liquids with a high solids content

Can use river or unfiltered water

Can be easily inspected and maintained if access is available

Most costly to purchase

Do not distribute liquid as uniformly as other methods

Weirs must be level

Tubes Uniform liquid distribution

Can be buried below packing surface

Generally least expensive to purchase

Easily plugged, must use filter

Difficult to determine if holes are plugged when tube is buried in the packing

Spray nozzles

Uniform liquid distribution

Tower need not be plumb

Can be easily inspected and maintained if access is available

Highest pressure drops and operating costs

Easily plugged, must use filter

Source: Clark 1975.


A serious problem that can drastically affect the operation of a packed tower is the buildup of solids in the packing. This can occur as a result of a number of situations. If the incoming exhaust gas contains a high concentration of particulate matter, the beds can easily become plugged. Precleaning sprays can reduce this problem by removing particles before the exhaust gas enters the packed bed. Solids buildup can also occur as a result of a chemical reaction between the scrubbing liquid and gaseous pollutant, producing a solid compound. In this case, the packing may occasionally be flushed with a cleaning fluid to remove the solids. For example, potassium permanganate is occasionally used in scrubbing solutions to control odors. The use of potassium permanganate results in a residue buildup on the packing that must periodically be cleaned with an acid backwash. No matter what the cause, plugging presents an expensive maintenance problem. Tower internals are not easily accessible; cleaning requires shutting the system down and then removing, cleaning, and, finally, reinstalling the packing material.

Another critical problem in packed tower operation is maintaining the proper liquid and gas flow rates. If the liquid or gas flow rate increases (one in relation to the other), a point is reached where the rising exhaust gas starts to hold up the descending liquid. The liquid fills the upper portion of the packing until its weight causes it to fall. This condition, known as flooding, results in a high pressure drop, a pulsating airflow in the tower, and greatly reduced pollutant removal efficiencies. Optimum operating flow rates are normally at 60 to 75% of the flooding conditions. Conversely, a gas flow rate that is too low can also cause mixing problems, resulting in gas channeling. Gas channeling occurs when the gas does not distribute uniformly through the packing, but moves only through a small portion of the bed (following the path of least resistance). This normally occurs near the walls of the tower,

Lesson 5 ___________________________________________________________________________________

5-10 2.0-7/98

where the void spaces are the greatest. Table 5-2 lists problems that can occur in daily operation of packed towers and some probable causes of these problems.

Table 5-2. Operating problems associated with packed towers

Problem Possible causes

Static pressure drop increases

Liquid flow rate to liquid distributor has increased and should be checked.

Packing in irrigated bed could be partially plugged due to solids deposition, and may require cleaning.

Entrainment separator could be partially plugged and may require cleaning.

Packing support plate at bottom of packed section could be plugged, causing increased pressure drop, which will require cleaning.

Packing could be settling due to corrosion or solids deposition, again requiring cleaning or additional packing.

Airflow rate through absorber could have been increased by a change in damper setting, which may need readjustment.

Pressure drop decreases, slowly or rapidly

Liquid flow rate to distributor has decreased and should be adjusted accordingly.

Airflow rate to scrubber has decreased due to a change in fan characteristics or due to a change in system damper settings.

Partial plugging of spray or liquid distributor, causing channeling through scrubber, could be occurring. Liquid distributor should be inspected to ensure that it is totally operable.

Packing support plate could have been damaged and fallen into bottom of the absorber, allowing packing to fall to bottom and produce a lower pressure drop. This should be checked.

Pressure or flow change in recycled liquid causing reduced liquid flow

Plugged strainer or filter in recycle piping, which may require cleaning.

Plugged spray nozzles, which may require cleaning.

Piping may be becoming partially plugged with solids and need cleaning.

Liquid level in sump could have decreased, causing pump cavitation.

Pump impeller could have been worn excessively.

Valve in either suction or discharge side of pump could have been inadvertently closed.

High liquid flow Break in the internal distributor piping.

Spray nozzle that has been inadvertently "uninstalled."

Spray nozzle that may have come loose or eroded away, creating a low pressure drop.

Change in throttling valve setting on the discharge side of the pump, allowing larger liquid flow; reset to the proper conditions.



2.0-7/98 5-11

Table 5-2. (continued) Operating problems associated with packed towers

Problem Possible causes

Excessive liquid carryover

Partially plugged entrainment separator, causing channeling and reentrainment of the collected liquid droplets.

Airflow rate to absorber could have increased above the design capability, causing reentrainment.

If a packed-type entrainment separator was used, packing may not be level, causing channeling and reentrainment of moisture.

If a packed entrainment separator was used, and a sudden surge of air through the absorber occurred, this could have caused the packing to be carried out of absorber or to be blown aside, creating an open area "hole" through separator.

Velocity through absorber has decreased to a point that absorption does not effectively take place, and low removal is achieved.

Reading indicating low airflow

Packing in absorber may be plugged, causing a restriction to airflow.

Liquid flow rate to absorber could have been increased inadvertently, again causing greater restriction and pressure drop, creating lower gas flow rate.

Fan belts have worn or loosened, reducing airflow to equipment.

Fan impeller could be partially corroded, reducing fan efficiency.

Ductwork to or from absorber could be partially plugged with solids and may need cleaning.

Damper in system has been inadvertently closed or setting changed.

Break or leak in duct could have occurred due to corrosion.

Increase in airflow Sudden opening of damper in system.

Low liquid flow rate to absorber.

Packing has suddenly been damaged and has fallen to bottom of absorber.

Sudden decrease in absorber efficiency

Liquid makeup rate to the absorber has been inadvertently shut off or throttled to a low level, decreasing absorber efficiency.

Set point on pH control may have to be adjusted to allow more chemical feed.

Problem may exist with chemical metering pump, control valve, or line pluggage.

Liquid flow rate to scrubber may be too low for effective removal.

Source: MacDonald 1982.

Summary

Packed towers are mainly used to remove gaseous pollutants. Because of plugging problems, they are not used when particle removal is the only concern, or when a high concentration of particulate matter is in the exhaust gas. Packed towers are capable of very high efficiencies for removing many gaseous pollutants. Packed towers and plate towers are ideal when pollutants are only slightly soluble, or when the gaseous pollutant removal efficiency must be greater than 99%. In a packed tower, the optimum pressure drop through a packing section is 1.7 to 5.0 cm (0.2 to 0.6 in.) of water per foot of installed packing (Clark 1975). The overall

Lesson 5 ___________________________________________________________________________________

5-12 2.0-7/98

pressure drops across packed towers are usually between 5 and 25 cm (2 and 10 in.) of water. Thus, packed towers are generally considered as medium-energy scrubbers.

Packed towers are most suited to applications where a high gas-removal efficiency is required and the exhaust gas is relatively free from particles. These include removing HCl, NH4, and SO2 gases from a variety of process streams such as those from fertilizer manufacturing, chemical processing, acid manufacturing, steel making, and metal pickling operations. One important point that should be noted is that packed towers are not effective in removing submicrometer-sized particles, even if the particles are very soluble. Inorganic salts or fumes such as ammonium chloride or aluminum chloride are prime examples. These particles are usually so small that they flow with the exhaust gas through the packing bed and are not absorbed. Table 5-3 lists the general operating characteristics of wet-film scrubbers.

Table 5-3. Operating characteristics of wet-film scrubbers

Pollutant


Liquid-to-gas ratio(L/G)


(pL)

Removal efficiency

Applications

Gases

2-8.5 cm/m of column packing

0.13-2.0 L/m3

34-100 kPa

Very high, 99+%, depending on operating conditions

Mainly used for gaseous pollutant removal

Metal operations

Particles (0.25-1 in./ft of column packing)

(1-15 gal/1000 ft3, depending on type of flow and packing)

(5-15 psig) 2.0 µm diameter

Acid plants


Plate towers (described in Lesson 3) are used to control emissions from many of the same processes that could use packed towers. Therefore, when gas removal is the only objective, the choice is often between a packed or plate absorber.

The following list gives some factors to consider when comparing plate towers to packed towers:

1. Packed towers are not able to handle particulate matter or other solid materials in the flue gas as well as plate towers.

2. Plate towers are chosen for operations that involve difficult gases to absorb or that must handle large gas volumes. To achieve the same collection efficiency for difficult absorption processes, packed towers must have either deep packed beds or multiple beds. Packed towers can experience liquid channeling problems if the diameter or height of the tower is too large. Redistribution trays must be installed in large-diameter and tall packed towers to avoid channeling.

3. The total weight of a packed tower is more than that of a comparable plate tower.


2.0-7/98 5-13

4. Packed towers are much cheaper to construct than plate towers if corrosive substances are to be handled. Packed towers can be constructed with a fiberglass-reinforced polyester shell which is generally about half the cost of a carbon steel plate tower.

5. Packed towers cannot handle volume and temperature fluctuations as well as plate towers. Expansion or contraction due to temperature changes can crush or melt packing material.

Lesson 5 ___________________________________________________________________________________

5-14 2.0-7/98


2.0-7/98 5-15

Review Exercise 1. True or False? Packed towers have limited application for particulate removal.

2. Packed towers are frequently used for removing gaseous pollutants because:

a. The packing provides a large surface area for gas-liquid contact b. They have relatively low pressure drops compared to plate towers c. The packing provides good mixing of gas and liquid and a long residence time d. All of the above

3. Increasing the liquid flow rate in a packed tower will usually ____________________ the gas removal rate.

a. Increase b. Decrease c. Have no effect on

4. In a ____________________ packed tower, the gas stream being treated enters the bottom and flows upward through the packing while the liquid is introduced over the top of the packing and flows down through it.

a. Cocurrent b. Crossflow c. Countercurrent

5. A ____________________ packed tower cannot handle large variations in liquid or gas flow rates because flooding may occur.

a. Cocurrent b. Countercurrent c. Crossflow d. Fiber-bed

6. Cocurrent packed towers usually have ____________________ pressure drops than countercurrent packed towers.

a. Higher b. Lower

7. True or False? Crossflow packed towers can handle flue gas containing a high concentration of particulate matter because they use liquid sprays that will remove and wash away particles in the front half of the bed.

8. Packing material is usually made of:

a. Porcelain b. Polyethylene c. Polypropylene d. All of the above

Lesson 5 ___________________________________________________________________________________

5-16 2.0-7/98

9. In a packed tower, liquid occasionally flows in little streams straight through the packing without wetting the packing surface. This condition is called:

a. Flooding b. Liquid channeling c. Mixing d. Plugging

10. In packed tower, liquid is distributed over the packing by using:

a. Sprays b. Sprays and small venturis c. Sprays, weirs, and tubes d. Chevron-shaped sheets and sprays

11. If the gas flow rate through a packed tower is too low, ____________________ may occur.

a. Flooding b. Mixing c. Gas channeling d. Plugging

12. True or False? Packed towers are most suitable for industrial processes requiring high gas-removal efficiency, but not having a high concentration of particulate matter in the flue gas.

13. True or False? Packed towers remove particulate matter and other solids more easily and with less maintenance problems than plate towers.

14. In processes having high-temperature flue gas, ____________________ towers are more suitable because their internal components will expand and contract.

a. Plate b. Packed


2.0-7/98 5-17

Review Exercise Answers 1. True

Packed towers have limited application for particulate removal.

2. d. All of the above Packed towers are frequently used for removing gaseous pollutants for the following reasons:

• The packing provides a large surface area for gas-liquid contact

• They have relatively low pressure drops compared to plate towers

• The packing provides good mixing of gas and liquid and a long residence time

3. a. Increase Increasing the liquid flow rate in a packed tower will usually increase the gas removal rate because of increasing the potential solubility of the pollutant in the additional liquid.

4. c. Countercurrent In a countercurrent packed tower, the gas stream being treated enters the bottom and flows upward through the packing while the liquid is introduced over the top of the packing and flows down through it.

5. b. Countercurrent A countercurrent packed tower cannot handle large variations in liquid or gas flow rates because flooding may occur.

6. b. Lower Cocurrent packed towers usually have lower pressure drops than countercurrent packed towers. Because the liquid and gas streams move in the same direction in cocurrent packed towers, there is less resistance to flow.

7. True Crossflow packed towers can handle flue gas containing a high concentration of particulate matter because they use liquid sprays that will remove and wash away particles in the front half of the bed.

8. d. All of the above Packing material is usually made of porcelain, polyethylene, or polypropylene.

9. b. Liquid channeling In a packed tower, liquid occasionally flows in little streams straight through the packing without wetting the packing surface. This condition is called liquid channeling.

10. c. Sprays, weirs, and tubes In packed tower, liquid is distributed over the packing by using sprays, weirs, and tubes.

11. c. Gas channeling If the gas flow rate through a packed tower is too low, gas channeling may occur.

12. True Packed towers are most suitable for industrial processes requiring high gas-removal

Lesson 5 ___________________________________________________________________________________

5-18 2.0-7/98

efficiency, but not having a high concentration of particulate matter in the flue gas. Packed towers are susceptible to plugging.

13. False Packed towers do NOT remove particulate matter and other solids more easily and with less maintenance problems than plate towers. The tops of plates can usually be accessed through openings, while the middle of the packed bed cannot.

14. a. Plate In processes having high-temperature flue gas, plate towers are more suitable because their internal components will expand and contract.


2.0-7/98 5-19

Bibliography


Clark, J. M. 1975. Absorption equipment. In F. L. Cross and H. E. Hesketh (Eds.), Handbook for the Operation and Maintenance of Air Pollution Control Equipment. Westport: Technomic Publishing.

MacDonald, J. W. 1977. Packed wet scrubbers. In P. N. Cheremisinoff and R. A. Young (Eds.), Air Pollution Control and Design Handbook. Part 2. New York: Marcel Dekker.

MacDonald, J. W. 1982. Absorbers. In L. Theodore and A. J. Buonicore (Eds.), Air Pollution Control Equipment, Design, Selection, Operation, and Maintenance. Englewood Cliffs: Prentice-Hall.




Zenz, F. A. 1972. Designing gas absorption towers. Chemical Engineering. 79:120-138.

Lesson 5 ___________________________________________________________________________________

5-20 2.0-7/98

2.0-7/98 6-1

Lesson 6 Combination Devices - Liquid-Phase and Gas-Phase Contacting Scrubbers

Goal To familiarize you with the operation, collection efficiency, and major maintenance problems of scrubbers that use energy from both the liquid and gas phases for contact.


1. List four combination contacting scrubbers

2. Describe the operation of four combination contacting scrubbers

3. For each scrubber above, identify the range of operating values for pressure drop, liquid-to-gas ratio, as well as the collection efficiency for both particles and gaseous pollutants

4. Describe the major operating or maintenance problems associated with each device

Introduction

A number of wet-collector designs use energy from both the gas stream and liquid stream to collect pollutants. Many of these combination devices are available commercially. A seemingly unending number of scrubber designs have been developed by changing system geometry and incorporating vanes, nozzles, and baffles. This lesson will describe four scrubbing systems that incorporate features of both liquid-phase and gas-phase contacting wet collectors:

• Cyclonic spray

• Mobile bed

• Baffle spray

• Mechanically aided scrubbers

Lesson 6 ___________________________________________________________________________________

6-2 2.0-7/98

Cyclonic Spray Scrubbers

Cyclonic spray scrubbers use the features of both the dry cyclone and the spray chamber to collect pollutants. Generally, the exhaust gas enters the chamber tangentially, swirls through the chamber in a corkscrew motion, and exits. At the same time, liquid is sprayed inside the chamber. As the exhaust gas swirls around the chamber, pollutants are captured when they impact on liquid droplets, are thrown to the walls, and washed back down and out.

Cyclonic scrubbers are generally low- to medium-energy devices, with pressure drops of 4 to 25 cm (1.5 to 10 in.) of water. Commercially available designs include the irrigated cyclone scrubber and the cyclonic spray scrubber. In the irrigated cyclone (Figure 6-1), the exhaust gas enters near the top of the scrubber into the water sprays. The exhaust gas is forced to swirl downward, then change directions, and return upward in a tighter spiral. The liquid droplets produced capture the pollutants, are eventually thrown to the side walls, and carried out of the collector. The "cleaned" gas leaves through the top of the chamber.

Figure 6-1. Irrigated cyclone scrubber

��

Combination Devices – Liquid-Phase and Gas-Phase Contacting Scrubbers ___________________________________________________________________________________

2.0-7/98 6-3

The cyclonic spray scrubber (Figure 6-2) forces the exhaust gas up through the chamber from a bottom tangential entry. Liquid sprayed from nozzles on a center post (manifold) is directed toward the chamber walls and through the swirling exhaust gas. As in the irrigated cyclone, liquid captures the pollutant, is forced to the walls, and washes out. The "cleaned" gas continues upward, exiting through the straightening vanes at the top of the chamber.

Figure 6-2. Cyclonic spray scrubber

Particle Collection

Cyclonic spray scrubbers are more efficient than spray towers, but not as efficient as venturi scrubbers, in removing particles from the exhaust gas stream. Particles larger than 5 µm are generally collected by impaction with 90% efficiency. In a simple spray tower, the velocity of the particles in the exhaust gas stream is low: 0.6 to 1.5 m/s (2 to 5 ft/sec). By introducing the exhaust gas tangentially into the spray chamber, the cyclonic scrubber increases exhaust gas velocities (thus, particle velocities) to approximately 60 to 180 m/s (200 to 600 ft/sec). The velocity of the liquid spray is approximately the same in both devices. This higher particle-to-liquid relative velocity increases particle collection efficiency for this device over that of the spray chamber. Exhaust gas velocities of 60 to 180 m/s are equivalent to those encountered in a venturi scrubber. However, cyclonic spray scrubbers are not as efficient as venturis because they are not capable of producing the same degree of useful turbulence.

��

��

��

��

Lesson 6 ___________________________________________________________________________________

6-4 2.0-7/98

Gas Collection

High exhaust gas velocities through these devices reduce the gas-liquid contact time, thus reducing absorption efficiency. Cyclonic spray scrubbers are capable of effectively removing some gases; however, they are rarely chosen when gaseous pollutant removal is the only concern.


The main maintenance problems with cyclonic scrubbers are nozzle plugging and corrosion or erosion of the side walls of the cyclone body. Nozzles have a tendency to plug from particles that are in the recycled liquid and/or particles that are in the exhaust gas stream. The best solution is to install the nozzles so that they are easily accessible for cleaning or removal. Due to high gas velocities, erosion of the side walls of the cyclone can also be a problem. Abrasion-resistant materials may be used to protect the cyclone body, especially at the inlet.

Summary

The pressure drops across cyclonic scrubbers are usually 4 to 25 cm (1.5 to 10 in.) of water; therefore, they are low- to medium-energy devices and are most often used to control large-sized particles. Relatively simple devices, they resist plugging because of their open construction. They also have the additional advantage of acting as entrainment separators because of their shape. The liquid droplets are forced to the sides of the cyclone and removed prior to exiting the vessel. Their biggest disadvantages are that they are not capable of removing submicrometer particles and they do not efficiently absorb most pollutant gases. Table 6-1 lists typical operating characteristics of cyclonic scrubbers.

Table 6-1. Operating characteristics of cyclonic scrubbers

Pollutant


Liquid-to-gas ratio

(L/G)


(pL)

Removal efficiency

Applications

Gases

4-25 cm of water

0.3-1.3 L/m3

280-2800 kPa

Only effective for very soluble gases

Mining operations

Drying operations

Food processing

Particles (1.5-10 in. of water)

(2-10 gal/1000 ft3) (40-400 psig) 2-3 µm diameter

Foundries



2.0-7/98 6-5

Mobile-Bed Scrubbers

Mobile-bed, also called moving-bed, scrubbers use energy from both liquid sprays and the gas stream to provide contact. Mobile-bed scrubbers are similar to packed towers. However, instead of having stationary packing, as in packed towers, they use a bed containing packing that is in constant motion. The gas stream provides the energy to keep the packing in motion while, at the same time, liquid is sprayed over the packing. Mobile-bed scrubbers can be classified as either flooded or fluidized, depending on the degree of packing movement. In a flooded-bed scrubber, the packing gently moves and rotates, whereas in a fluidized scrubber, the packing is suspended, or fluidized, within the bed.

Mobile-bed scrubbers were developed to provide the effective mass-transfer (absorption) characteristics of packed and plate towers, without the plugging problems. The wetted packing provides a large area for gas-to-liquid contact, promoting absorption. The movement of the bed cleans off any deposited particles. Therefore, these devices are primarily used when good collection efficiency for both particulate and gaseous pollutants is required.

A flooded-bed scrubber (Figure 6-3) contains a section of mobile packing (spheres) 10 to 20 cm (4 to 8 in.) deep. The spheres are usually made of plastic; however, glass or marble spheres have been used. The exhaust gas stream enters from the bottom while liquid is sprayed from the top and/or bottom over the packing. Bottom, or inlet, sprays are usually included to saturate the exhaust gas stream and remove any large particles. The gas velocity is such that it causes the packing materials to rotate and rub against each other. This rotating motion acts as a self-cleaning mechanism in addition to enhancing gas and liquid mixing.

Figure 6-3. Flooded-bed scrubber

��

��

� ��

��

� ��

��

Lesson 6 ___________________________________________________________________________________

6-6 2.0-7/98

Bubbles formed in the bed create a layer of froth over the bed approximately twice as deep as the bed itself. This turbulent froth layer provides an additional surface for absorbing pollutant gases and collecting fine particles. Because of the high gas velocities, entrainment separators are required to prevent liquid-mist carryover.

A fluidized-bed scrubber is very similar to a flooded-bed scrubber, except for the degree of movement of the packing. In a fluidized-bed scrubber, the exhaust gas velocity (1.8 to 4.8 m/s, or 6 to 16 ft/sec) is high enough to keep the packing in constant motion between a lower and upper retaining grid. This is shown in Figure 6-4. The packing is made of either polypropylene or polyethylene plastic balls that are hollow, resembling ping pong balls. The packed sections are usually 0.3 to 0.6 m (1 to 2 ft) thick with a froth zone about 0.6 m (2 ft) thick above the packing. These devices can have one to as many as six fluidized packed sections. When used for gas absorption, they are sometimes referred to as turbulent-contact absorbers (TCA).

Figure 6-4. Fluidized-bed scrubber

Particle Collection

In a mobile-bed scrubber, particles can be collected in three locations. First, sprays are used to remove coarse particles in the inlet below the bed. Particles are also captured when they impinge on the wetted surface of the packing. Finally, small particles are captured in the froth, or foam, layers above the bed. These devices will generally remove

��

� ��

��


2.0-7/98 6-7

particles as small as 2 to 3 µm in diameter and have been used extensively when the exhaust stream does not contain a substantial amount of particles in the submicrometer range. These devices usually contain one bed, unless gas absorption is a consideration. Adding additional beds or more packing does not substantially increase the particle collection efficiency (i.e., any particles not captured by the first stage will probably not be collected in any following stages). The pressure drop in mobile-bed scrubbers ranges from 5 to 15 cm (2 to 6 in.) of water per stage of packing.

Gas Collection

Mobile-bed scrubbers are capable of achieving high gaseous-pollutant removal efficiencies. Their operation is very similar to the operation of packed towers. Liquid is sprayed over the mobile packing, providing a huge surface for the pollutant gas to contact the liquid. Movement of both the gas around the packing and the constantly sprayed liquid provides excellent mixing and contact time for absorption to occur. Mobile-bed scrubbers provide the same amount of absorption efficiency as packed and plate towers without the associated plugging problems. Due to the high exhaust gas velocities through mobile-bed scrubbers, these units can handle five to six times the amount of exhaust gas handled by packed or plate towers of similar size (Bethea 1978). However, they are not as efficient as packed or plate towers per unit of energy consumed.

Absorption in mobile-bed scrubbers is enhanced by the same factors that affect packed towers: increasing the liquid-to-gas ratio, increasing the depth of packing, or increasing the number of stages. Increasing these factors increases the gas and liquid contact and the residence time. However, increasing these factors also raises the capital and/or operating costs of the system. As with any system, these process variables are set to achieve the desired removal efficiency at a minimum cost. For gas absorption, multiple stages are used and the liquid-to-gas ratios are high. For example, mobile-bed scrubbers have been used to remove SO2 from boiler flue gas exhausts. Using a lime or limestone slurry, the liquid-injection rates are approximately 8 L/m3 (60 gal/1000 ft3) of flue gas. This is compared to 0.4 L/m3 (3.0 gal/1000 ft3) when these devices are used for particle removal (McIlvaine Company 1974).


Mobile-bed scrubbers are designed to minimize plugging and scale buildup problems through the constant motion of the packing spheres. However, these problems can still occur at the scrubber inlet (wet-dry interface) or on the packing support grid. Scale buildup in these areas can cause an uneven airflow distribution through the bed. Uneven airflows result in some areas of the packing bed having a high gas velocity, while the gas velocity is much lower in other areas. This can result in a decrease in collection efficiency and in excessive liquid carryover. Adjusting the inlet sprays can help solve this problem. As with any spray system, nozzles can also be a major maintenance problem. Nozzle maintenance is a special concern in lime or limestone scrubbing systems because of the large quantities of solids present in the recycled scrubbing liquor.

Deterioration of the spheres can also be a problem. Neither plastic nor marble balls are able to withstand high temperatures. The marble cracks and breaks while the plastic deforms. Most systems have safety mechanisms to prevent a total loss of water that would cause high temperatures. Deterioration of the balls from constant rubbing against

Lesson 6 ___________________________________________________________________________________

6-8 2.0-7/98

each other can also be a problem. Glass balls can generally withstand abrasive conditions, whereas plastic balls cannot; therefore, they wear out quickly.

Summary

Mobile-bed scrubbers are used when high collection efficiency of particulate and gaseous pollutants is required. Typical applications would be for treating flue gases from industrial boilers, smelting operations, and kraft pulp mills. The main advantage of mobile-bed scrubbers is that they are capable of high-efficiency absorption without plugging. The main disadvantage is that they do not efficiently remove particles in the submicrometer range. A major maintenance problem is the effect of abrasive wear and high temperatures on packing balls, causing them to deteriorate.

Mobile-bed scrubbers are generally designed in one stage for particle collection, or in multiple stages for high-efficiency gas absorption. Gas velocities are much higher than those in packed or plate towers; therefore, mobile-bed scrubbers can be much smaller in size than either tower. Because of these high gas velocities, incorporating some type of entrainment separator is mandatory. Table 6-2 lists some general operating characteristics of mobile-bed scrubbers.

Table 6-2. Operating characteristics of mobile-bed scrubbers

Pollutant


Liquid-to-gas ratio

(L/G)

Liquid-inlet

pressure(p

L)

Removal efficiency

Applications

Gases 5-15 cm of water per stage

2.7-8.0 L/m3

(20-60 gal/1000 ft3)

−

99+% of theoretical

Mining operations

Pulp mills

Utility boilers

Particles (2-6 in. of water per stage)

0.4-0.7 L/m3

(3-5 gal/1000 ft3)

2-3 µm diameter

Food industry


Baffle Spray Scrubbers

Baffle spray scrubbers are very similar to spray towers in design and operation. However, in addition to using the energy provided by the spray nozzles, baffles are added to allow the gas stream to atomize some liquid as it passes over them. A simple baffle scrubber system is shown in Figure 6-5. Liquid sprays capture pollutants and also remove collected particles from the baffles. Adding baffles slightly increases the pressure drop of the system.


2.0-7/98 6-9

Figure 6-5. Baffle spray scrubber

Particle Collection

These devices are used much the same as spray towers to preclean or remove particles larger than 10 µm in diameter. However, they will tend to plug or corrode if particle concentration of the exhaust gas stream is high.

Gas Collection

Even though these devices are not specifically used for gas collection, they are capable of a small amount of gas absorption because of their large wetted surface.

��

��

��

Lesson 6 ___________________________________________________________________________________

6-10 2.0-7/98

Summary

These devices are most commonly used as precleaners to remove large particles (>10 µm in diameter). The pressure drops across baffle scrubbers are usually low, but so are the collection efficiencies. Maintenance problems are minimal. The main problem is the buildup of solids on the baffles. Table 6-3 summarizes the operating characteristics of baffle spray scrubbers.

Table 6-3. Operating characteristics of baffle spray scrubbers

Pollutant



Liquid-inlet

pressure (p

L)

Removal efficiency

Applications

Gases 2.5-7.5 cm of water

0.13 L/m3

< 100 kPa

Very low Mining operations

Incineration

Particles (1-3 in. of water)

(1 gal/1000 ft3) (< 15 psig) 10 µm diameter


Mechanically Aided Scrubbers

In addition to using liquid sprays or the exhaust stream, scrubbing systems can use motors to supply energy. The motor drives a rotor or paddles which, in turn, generate water droplets for gas and particle collection. Systems designed in this manner have the advantage of requiring less space than other scrubbers, but their overall power requirements tend to be higher than other scrubbers of equivalent efficiency. Significant power losses occur in driving the rotor. Therefore, not all the power used is expended for gas-liquid contact.

There are fewer mechanically aided scrubber designs available than liquid- and gas-phase contacting collector designs. Two will be discussed here: centrifugal-fan scrubbers and mechanically induced spray scrubbers.

A centrifugal-fan scrubber can serve as both an air mover and a collection device. Figure 6-6 shows such a system, where water is sprayed onto the fan blades cocurrently with the moving exhaust gas. Some gaseous pollutants and particles are initially removed as they pass over the liquid sprays. The liquid droplets then impact on the blades to create smaller droplets for additional collection targets. Collection can also take place on the liquid film that forms on the fan blades. The rotating blades force the liquid and collected particles off the blades. The liquid droplets separate from the gas stream because of their centrifugal motion.

Centrifugal-fan collectors are the most compact of the wet scrubbers since the fan and collector comprise a combined unit. No internal pressure loss occurs across the scrubber, but a power loss equivalent to a pressure drop of 10.2 to 15.2 cm (4 to 6 in.) of water occurs because the blower efficiency is low.


2.0-7/98 6-11

Figure 6-6. Centrifugal-fan scrubber

Another mechanically aided scrubber, the induced-spray, consists of a whirling rotor submerged in a pool of liquid. The whirling rotor produces a fine droplet spray. By moving the process gas through the spray, particles and gaseous pollutants can subsequently be collected. Figure 6-7 shows an induced-spray scrubber that uses a vertical-spray rotor.

Figure 6-7. Mechanically induced spray scrubber

��

� � �

Lesson 6 ___________________________________________________________________________________

6-12 2.0-7/98

Particle Collection

Mechanically aided scrubbers are capable of high collection efficiencies for particles with diameters of 1 µm or greater. However, achieving these high efficiencies usually requires a greater energy input than those of other scrubbers operating at similar efficiencies. In mechanically aided scrubbers, the majority of particle collection occurs in the liquid droplets formed by the rotating blades or rotor.

Gas Collection

Mechanically aided scrubbers are generally not used for gas absorption. The contact time between the gas and liquid phases is very short, limiting absorption. For gas removal, several other scrubbing systems provide much better removal per unit of energy consumed.


As with almost any device, the addition of moving parts leads to an increase in potential maintenance problems. Mechanically aided scrubbers have higher maintenance costs than other wet collector systems. The moving parts are particularly susceptible to corrosion and fouling. In addition, rotating parts are subject to vibration-induced fatigue or wear, causing them to become unbalanced. Corrosion-resistant materials for these scrubbers are very expensive; therefore, these devices are not used in applications where corrosion or sticky materials could cause problems.


2.0-7/98 6-13

Summary

Mechanically aided scrubbers have been used to control exhaust streams containing particulate matter. They have the advantage of being smaller than most other scrubbing systems, since the fan is incorporated into the scrubber. In addition, they operate with low liquid-to-gas ratios. Their disadvantages include their generally high maintenance requirements, low absorption efficiency, and high operating costs. The performance characteristics of mechanically aided scrubbers are given in Table 6-4.

Table 6-4. Operating characteristics of mechanically aided scrubbers

Pollutant


Liquid-to-gas ratio(L/G)


(pL)

Particle

diameter

Applications

Particles 10-20 cm of water

(4.0-8.0 in. of water)

0.07-0.2 L/m3 (centrifugal)

0.5-1.5 gal/1000 ft3 (centrifugal)

20-60 psig (centrifugal)

< 1 µm diameter

Mining operations

Food product industries

Chemical industry

Foundries and steel mills

0.5-0.7 L/m3 (spray rotor)

4-5 gal/1000 ft3 (spray rotor)

Note: These devices are used mainly for particle collection; however, they can also remove gaseous pollutants that are present in the exhaust stream.


Lesson 6 ___________________________________________________________________________________

6-14 2.0-7/98


2.0-7/98 6-15


1. Cyclonic scrubbers are ____________________ energy devices.

a. High b. Low- to medium-

2. In a cyclonic spray scrubber, particles are primarily collected:

a. As they hit the wetted walls b. As they impact the liquid droplets c. Due to gravity d. In the throat

3. Cyclonic spray scrubbers are more efficient than ____________________, but not as efficient as ____________________, in removing particles.

a. Spray towers, venturi scrubbers b. Venturi scrubbers, spray towers

4. True or False? Cyclonic scrubbers are not often used to control gaseous emissions.

5. List two main maintenance problems associated with cyclonic scrubbers. ________________________________________ ________________________________________

6. What are cyclonic scrubbers used most often to control?

a. Micrometer-sized particles b. Large-sized particles c. Gaseous emissions d. Particles and gases simultaneously

Part 2

7. Mobile-bed, or moving-bed, scrubbers were developed to provide the effective mass-transfer characteristics of ____________________ or ____________________ towers without the plugging problems.

a. Spray (or) venturi b. Packed (or) plate c. Cyclonic (or) orifice d. Ejector (or) spray

8. In mobile-bed scrubbers, the moving packing is made of:

a. Glass b. Plastic c. Marble d. Any of the above

Lesson 6 ___________________________________________________________________________________

6-16 2.0-7/98

9. In mobile-bed scrubbers, particles are collected:

a. By using inlet sprays b. As they impinge on the wetted surface of the spheres c. In a froth, or foam, layer above the bed d. All of the above

10. True or False? In mobile-bed scrubbers, adding stages or more packing will usually increase particle collection efficiency.

11. In mobile-bed scrubbers, gas velocities are much ____________________ than in packed towers or plate towers; therefore, mobile-bed scrubbers can be much ____________________ in size.

a. Lower, smaller b. Lower, larger c. Higher, smaller d. Higher, larger

12. Mobile-bed scrubbers provide the gas absorption efficiency of packed or plate towers; however, they consume ____________________ energy for the same unit operation.

a. More b. Less c. The same

13. Gas absorption in mobile-bed scrubbers can be enhanced by:

a. Increasing the L/G ratio b. Adding more packing height c. Adding stages d. All of the above

14. When used for gas absorption, mobile-bed scrubbers operate at ____________________ L/G ratios than when used for particle collection.

a. Much higher b. Much lower c. The same

15. Scale buildup or plugging at the mobile-bed scrubber inlet can cause ____________________ that leads to a decrease in efficiency.

a. A low liquid pH b. Uneven gas flow distribution through the bed c. Excessive liquid carryover d. Low liquid flow


2.0-7/98 6-17

16. The biggest maintenance problem with mobile-bed scrubbers is ball deterioration due to:

a. Abrasive wear b. High temperatures c. Both high temperatures and abrasive wear d. None of the above

17. True or False? A major limitation of mobile-bed scrubbers is that they are not effective in removing submicrometer-sized particles.

Part 3

18. Adding baffles in a spray tower will generally help increase the particle removal efficiency, but also increases the:

a. L/G ratio b. Pressure drop c. Height of the unit d. All of the above

19. Spray towers and baffle spray towers are generally not effective in removing particles smaller than:

a. 10 µm b. 50 µm c. 100 µm d. Any of the above

20. Mechanically aided scrubbers use a rotor to generate water droplets. These devices usually require less ____________________ than other scrubbers, but have ____________________ that tend to be higher.

a. Liquid, gas flows b. Space, power requirements c. Power, liquid requirements

21. True or False? Mechanically aided scrubbers can serve as both an air mover and a collection device.

22. In mechanically aided scrubbers, the majority of particle collection occurs:

a. In liquid droplets formed by the rotating blades b. On the wetted blades c. At the inlet sprays

23. True or False? Mechanically aided scrubbers are generally not used for gas absorption, since several other designs provide better removal.

24. True or False? Mechanically aided scrubbers operate at lower liquid-to-gas ratios than most other scrubbers.

Lesson 6 ___________________________________________________________________________________

6-18 2.0-7/98


2.0-7/98 6-19


1. b. Low- to medium- Cyclonic scrubbers are low- to medium-energy devices with pressure drops ranging from 2 to 10 inches of water.

2. b. As they impact the liquid droplets In a cyclonic spray scrubber, particles are primarily collected as they impact the liquid droplets.

3. a. Spray towers, venturi scrubbers Cyclonic spray scrubbers are more efficient than spray towers but not as efficient as venturi scrubbers in removing particles.

4. True Cyclonic scrubbers are not often used to control gaseous emissions due to limited liquid-to-gas contact.

5. Nozzle plugging Corrosion or erosion of the side walls in the chamber Two main maintenance problems associated with cyclonic scrubbers are nozzle plugging and corrosion or erosion of the side walls in the chamber.

6. b. Large-sized particles Cyclonic scrubbers are used most often to control large-sized particles.

Part 2

7. b. Packed (or) plate Mobile-bed, or moving-bed, scrubbers were developed to provide the effective mass-transfer characteristics of packed or plate towers without the plugging problems.

8. d. Any of the above In mobile-bed scrubbers, the moving packing can be made of:

• Glass

• Plastic

• Marble

9. d. All of the above In mobile-bed scrubbers, particles are collected:

• By using inlet sprays

• As they impinge on the wetted surface of the spheres

• In a froth, or foam, layer above the bed

10. False In mobile-bed scrubbers, adding stages or more packing will usually NOT increase particle collection efficiency. Particle collection is based on particle size. Once a given size range is removed, you need to change the mechanism, not just do more of the same.

Lesson 6 ___________________________________________________________________________________

6-20 2.0-7/98

11. c. Higher, smaller In mobile-bed scrubbers, gas velocities are much higher than in packed towers or plate towers; therefore, mobile-bed scrubbers can be much smaller in size.

12. a. More Mobile-bed scrubbers provide the gas absorption efficiency of packed or plate towers; however, they consume more energy for the same unit operation.

13. d. All of the above Gas absorption in mobile-bed scrubbers can be enhanced by the following:

• Increasing the L/G ratio

• Adding more packing height

• Adding stages

14. a. Much higher When used for gas absorption, mobile-bed scrubbers operate at much higher L/G ratios than when used for particle collection. The added liquid increases the potential solubility of the gases.

15. b. Uneven gas flow distribution through the bed Scale buildup or plugging at the mobile-bed scrubber inlet can cause uneven gas flow distribution through the bed that leads to a decrease in efficiency.

16. c. Both high temperatures and abrasive wear The biggest maintenance problem with mobile-bed scrubbers is ball deterioration due to both high temperatures and abrasive wear.

17. True A major limitation of mobile-bed scrubbers is that they are not effective in removing submicrometer-sized particles.

Part 3

18. b. Pressure drop Adding baffles in a spray tower will generally help increase the particle removal efficiency, but also increases the pressure drop.

19. a. 10 µm Spray towers and baffle spray towers are generally not effective in removing particles smaller than 10 µm.

20. b. Space, power requirements Mechanically aided scrubbers use a rotor to generate water droplets. These devices usually require less space than other scrubbers, but have power requirements that tend to be higher.

21. True Mechanically aided scrubbers can serve as both an air mover and a collection device.

22. a. In liquid droplets formed by the rotating blades In mechanically aided scrubbers, the majority of particle collection occurs in liquid droplets formed by the rotating blades.


2.0-7/98 6-21

23. True Mechanically aided scrubbers are generally not used for gas absorption, since several other designs provide better removal.

24. True Mechanically aided scrubbers operate at lower liquid-to-gas ratios than most other scrubbers.

Lesson 6 ___________________________________________________________________________________

6-22 2.0-7/98


2.0-7/98 6-23

Bibliography





U.S. Environmental Protection Agency. 1969. Control Techniques for Particulate Air Pollutants. AP-51.

Lesson 6 ___________________________________________________________________________________

6-24 2.0-7/98

2.0-7/98 7-1

Lesson 7 Dry Scrubbing Systems

Goal To familiarize you with the types, operating characteristics and collection efficiency associated with dry scrubbing systems.

Objectives At the end of this lesson you will be able to do the following:

1. Name three industrial processes where dry scrubbers are primarily used

2. Briefly describe how dry sorbent and spray dryer absorbers operate to collect gaseous emissions

3. Name two types of atomizers used in spray dryers

4. Name and describe at least three operating parameters that affect the performance of dry scrubbing systems

5. Briefly describe operation and maintenance problems associated with spray dryer absorbers

Introduction

Dry scrubbing systems control acid gas emissions (SO2, HCl, HF, etc.) and are used primarily on utility and industrial boilers, municipal waste combustors, medical waste incinerators, and some refinery processes. Of course, wet scrubbing systems can also function effectively as acid gas collectors. Regardless of whether scrubber acid gas control systems operate wet or dry, they have a mechanism for introducing alkaline material into the exhaust gas to react with the acid gases present. Dry scrubbing systems are discussed in this lesson, while wet flue gas desulfurization systems (wet acid gas control systems that remove SO2) are discussed in Lesson 9.

Up to this point, you have been learning about wet scrubber designs. In wet scrubbers, liquid droplets provide the primary targets for collecting particles and gases. To facilitate this process, gas streams are saturated with moisture; therefore wet scrubbing systems release a steam plume when exiting the stack. Also, wet scrubbers require a system of pipes and pumps for collecting, treating, and recirculating the scrubbing liquid. In contrast, as their name implies, dry scrubbers either operate completely dry or use much smaller amounts of

Lesson 7 ___________________________________________________________________________________

7-2 2.0-7/98

liquid than wet scrubbers. In some semi-dry designs, liquid is added to the alkaline material, creating a slurry. High scrubber temperatures evaporate the moisture before the gases and reaction products leave the scrubber. Therefore, dry scrubbing systems do not have a stack steam plume or waste water handling/disposal requirement.

There are a number of different dry scrubbing systems designs. However, all consist of two main sections or devices: (1) a device to introduce the acid gas sorbent material into the gas stream, and (2) a particulate-matter control device to remove reaction products, excess sorbent material and any other particulate pollutants in the flue gas. Dry scrubbing systems can be categorized as dry sorbent injectors (DSIs) or as spray dryers [also called semi-dry scrubbers or spray dryer absorbers (SDAs)]. Since dry scrubbing systems only remove gases, a separate device is always required to remove particles. The particulate control devices are generally fabric filters or electrostatic precipitators (ESPs).

Dry sorbent injection involves the addition of a dry alkaline material (usually hydrated lime or soda ash) into the gas stream to react with any acid gases that are present. The sorbent can be injected directly into the flue gas duct ahead of the particulate control device or into an open reaction chamber. The acid gases react with alkaline sorbents to form solid salts which are removed in the particulate control device.

In spray dryer absorbers, the flue gases are introduced into an absorbing tower (dryer) where the gases are contacted with a finely atomized alkaline slurry [usually a calcium-based sorbent such as Ca(OH)2 or CaO]. Acid gases are absorbed by the slurry mixture, and react to form solid salts. The heat of the flue gas is used to evaporate all the water droplets leaving a non-saturated (i.e. dry) flue gas exiting the absorber tower. The effect of cooling and humidifying the hot gas stream increases collection efficiency over simple dry injection.

Gas Removal Mechanisms

In dry scrubbing, acid gas is removed by the mechanisms of adsorption and absorption. In dry injection systems, where adsorption is the primary removal mechanism, pollutant gas molecules adhere to the surface area of the alkaline particles. Thus, the reaction between the acid gas and the alkaline material takes place on the surface of these alkaline particles. The alkaline materials are generally calcium hydroxide or sodium-based reagents that have the consistency of a fine powder. These fine particles have large surface areas to aid in adsorbing the acid gases.

In spray dryer systems, absorption is the predominant collection mechanism. Lesson 2 describes the general process of gaseous pollutants being absorbed by liquid droplets. Absorption can occur in conjunction with a chemical reaction if a reagent has been added to the scrubbing liquid. Spray dryer absorbers utilize this principle. First, the acid gas dissolves in the alkaline slurry droplets, then reacts with the alkaline material dissolved therein to form solid salts. Because the acid gases react to form new compounds, additional acid gases can be absorbed by the liquid. Also, when the liquid droplets evaporate, the acid gases continue to react (by adsorption) with the solid alkaline materials remaining in the SDA.

Adsorption and absorption are similar mass transfer processes in that the acid gases must first be brought into contact with the alkaline sorbent material, be provided ample reaction sites and time, and finally, be removed from the gas stream. Intimate contact between the alkaline sorbent and acid gases is important for effective gas removal. With dry injection, solid

Dry Scrubbing Systems ___________________________________________________________________________________

2.0-7/98 7-3

powder-like sorbent is dispersed in either the furnace area, exhaust duct, or in a reaction chamber. Dispersion is generally provided by injecting the sorbent through a venturi device countercurrent to the flow of the gas stream to create turbulence. In spray dryers, the alkaline sorbent slurry is dispersed as a mist of tiny liquid droplets in the reaction or drying vessel. Due to their fine spray mists, spray dryers provide much more contact area than dry injectors for gas absorption to occur. Also, spray dryer absorbers provide more effective mixing of acid gases with the alkaline sorbent than dry sorbent injectors because it is easier to mix a gas with a liquid than with a solid. Spray dryer absorbers have some disadvantages; the injection (atomization) equipment required by spray dryer absorbers is much more complicated and expensive to operate.

Residence or reaction time can be enhanced in these applications in a number of ways. In dry injectors, the sorbent is often injected directly into the furnace or ductwork. To extend the residence time, reaction or holding vessels can be added to the dry sorbent injection system. Spray dryers always have a reaction or drying chamber to assure a dry gas stream leaving the chamber. Also, in both systems, the particulate control device will provide an additional area for the acid gases to further react with the sorbent.

In addition, both the absorption and adsorption processes are temperature dependent: the cooler the flue gas, the more effectively the acid gases will react with the sorbents. Spray dryer absorbers cool the gas stream and therefore, can achieve higher removal efficiencies than dry injection with no cooling.

Stoichiometry

An important parameter in the operation of a dry scrubbing system is the amount of alkaline material feed into the system. The amount of sorbent required is a function of the following:

1. The type of sorbent used

2. The inlet and outlet acid gas levels (the outlet level is determined by removal requirements)

3. The effectiveness of the dry scrubbing system design

The amount of sorbent added is usually reported on a molar basis as the stoichiometric ratio of sorbent to acid gases.

Although the sorbents are either calcium- or sodium-based solids, the exact chemical reaction that occurs depends on the type of sorbent used and the injection point in the process. Presently the most widely used dry scrubbing system is the calcium-based hydrated lime [Ca(OH)2]. A slurry of hydrated lime and water is injected into the spray dryer and reacts with the acid gases in a simplified manner as follows:

Ca(OH)2 + SO2 → CaSO3(s) + H2O (7-1)

Ca(OH)2 + 2HCl → CaCl2(s) + 2H2O (7-2)

As you can see from the above reactions, one mole of calcium hydroxide [Ca(OH)2] will neutralize one mole of SO2, whereas one mole of calcium hydroxide will neutralize two moles of HCl.

Lesson 7 ___________________________________________________________________________________

7-4 2.0-7/98

To compute the pounds of calcium hydroxide required to neutralize a given weight of SO2 or HCl, the molecular weight of each component must be utilized. For example, the molecular weights of SO2, HCl, and Ca(OH)2 are as follows:

Compound Molecular Weight (lb/lb-mole)

SO2 64

HCl 36

Ca(OH)2 74

Therefore, one pound of calcium hydroxide can neutralize 0.86 pounds of SO2 (64 divided by 74) or 0.97 pounds of HCl (36 times 2 divided by 74).

In computing the stoichiometric ratio of a system, all the acid compounds in the waste stream need to be accounted for. Also, the above equations are for the stoichiometric quantities of sorbent. The actual use of sorbent will be above these quantities because of normal inefficiencies in operation; contact of sorbent and acid gases is never ideal and distribution of acid gases in the flue gas is often not uniform (especially in incineration systems). The actual stoichiometric ratios can range from as low as 1.5 to 4.0 dependent on system design and required removal efficiencies.

Similar type reactions occur with sodium-based compounds. For semi-dry systems using caustic soda (NaOH) the following simplified reactions can be written:

SO2 + 1/2 O2 + 2NaOH → Na2SO4 + H2O (7-3)

HCl + NaOH → NaCl + H2O (7-4)

Also, sorbents react with different acids at different rates. For example, sorbents react with chlorides at a faster rate than with SO2. Therefore, in waste streams that have both SO2 and HCl, the HCl is removed at a higher rate than the SO2.



2.0-7/98 7-5

Dry Injection

Dry sorbent injection (DSI) is a process used to control acid gases by injecting a powdered sorbent into the flue gas stream. The sorbent can be injected into the furnace, boiler area or the ductwork/reaction chamber prior to the air pollution control device. The injection point depends on the type of sorbent and required reaction time. For example, some sorbents need to be injected at elevated temperatures to undergo a decomposition reaction before they can effectively remove the acid gas. Figure 7-1 shows a schematic of a typical dry injection system.

Figure 7-1. Components of a dry injection system

The dry sorbent injection system is a very simple system that consists of a dry sorbent storage tank, a weight feeder to meter the required amount of sorbent, a blower and transfer line, and an injection device such as a venturi. The dry sorbent material is blown through a pneumatic line to the injection area where transfer through the pneumatic line provides fluidization of the sorbent material. Injection into the duct is generally done countercurrent to the gas flow

Lesson 7 ___________________________________________________________________________________

7-6 2.0-7/98

to create added turbulence and promote mixing. An expansion/reaction chamber may be included to increase the residence time of the acid gases to react with sorbent.

The simple dry injector process described above is capable of achieving moderate control of acid gases for example 50% SO2 and 90% HCl removal on municipal and medical waste combustors. The acid gas removal efficiencies can be increased by cooling and/or humidifying the flue gas stream. Exhaust gases from industrial boilers or refuse combustors can range from 600oF to 400oF. The flue gases can be cooled (and the humidity increased) by using a heat exchanger or a dry quench chamber upstream of the injection point. Cooling the flue gas temperature increases the rate of reaction between the sorbent and acid gases. But, the temperature must be maintained high enough (300-350oF) to ensure that all the water droplets used to quench are evaporated.

Recycling a portion of the collected particles and unreacted sorbent is another method used to increase overall effectiveness of dry scrubbing systems. As stated previously, it is difficult to mix a dry solid and a gas stream; therefore, additional sorbent (above stoichiometric amount) must be injected. As a result, there is unreacted sorbent captured in the baghouse or electrostatic precipitator. In some instances a portion of this waste stream is recycled back to the injection point.

In order to achieve high removal efficiencies using relatively inexpensive calcium sorbents, most dry injection systems have to operate at higher stoichiometric ratios than a spray dryer would. For example, stoichiometric ratios of 2.0 to 4.0 are used on municipal waste combustors to achieve moderate acid gas control. This increased sorbent usage limits their application to smaller sources such as medical waste incinerators. Table 7-1 lists some facilities that have installed dry injection acid gas control systems.


2.0-7/98 7-7

Table 7-1. Examples of dry injection systems on medical and municipal waste incinerators

Facility location Incinerator Control device Combustor

Facility name City State manufacturer manufacturer size, lb/hr

Baltimore Baltimore MD Consumat Procedaire 6000

Trumbull Hospital Warren OH Joy Joy 765

Erlanger N Chattanooga TN Basic BACT 1176

Evanston Hospital Evanston IL Basic United McGill 1176

Florida Hospital Orlando FL Basic Mikropul FF 1176

Mediwaste West Babylon NY U.S. Waste Systems

Interel 2000

Northwest Hospital Seattle WA Consumat Consumat 1200

Healthcare Incinerators

Fargo ND Consumat Consumat 1200

Incindere Spring Hill LA Consumat Consumat 1500

Biomedical Services

Mathews NC Consumat Consumat 1500

WMI Terrel TX Disc International ERA-Tech 1300

Midway Stroud OK Basic United McGill 6588

Sparrow Hospital Lansing MI Econotherm Airopulse 1200

Thermtec Elyria OH Therm Tec Donalson 1000-1200

Thermtec Cincinnati OH Therm Tec Donalson 1000-1200

WMI Northwood OH Joy 2000 TES ERA-Tech 1525

WMI W. Carrolton OH Joy 2000 TES ERA-Tech 1525

WMI Germantown WI Joy 2000 TES Research Cottrell

1525

WMI Apopka FL Joy 2500 TES United McGill 1910

Morristown Memorial Hospital

Morristown NJ ThermAll, Inc. ThermAll, Inc. 800

Swedish Hospital Med. Ctr.

Seattle WA Therm-tec Mikropul FF 800

Hamot Erie PA BICO >1000

Borgess Kalamazoo MI Cleaver Brooks Cleaver Brooks DI/ Mikropul FF

650

Note: All systems are a dry injector followed by a fabric filter.

Lesson 7 ___________________________________________________________________________________

7-8 2.0-7/98

Spray Dryer Systems

In the spray drying process, an alkaline slurry [usually Ca(OH)2] is injected into a spray dryer chamber through either a rotary atomizer or two-fluid nozzle injectors.

The atomized slurry droplets contact the hot flue gas in the spray dryer chamber (See Figure 7-2). The water in the alkaline (lime) slurry evaporates to cool the flue gas, and the lime reacts with the acid gases in the flue gas to form calcium- or sodium-based salts. The reaction or absorption chamber is designed to provide sufficient contact and residence time to produce a dry product leaving the chamber. The particulate exiting the chamber contains fly ash, calcium salts and unreacted lime that must be sent to a particulate control device, usually a fabric filter or electrostatic precipitator (ESP).

Figure 7-2. Spray dryer absorber


2.0-7/98 7-9

Collected solid reaction products from the system are sometimes recycled to the feed of the spray dryer to reduce alkaline sorbent use. Figure 7-3 provides a diagram of a typical spray drying system. The major components of a typical spray drying system are:

• Alkaline (lime) storage and slaking system

• Alkaline mixing and feed tanks

• Atomizer (rotary or nozzle)

• Spray dryer chamber

• Particulate control device (e.g. baghouse)

• Recycle system (optional)

Figure 7-3. Components of a spray dryer absorber system

Lesson 7 ___________________________________________________________________________________

7-10 2.0-7/98

Operating and Design Parameters

Key design and operating parameters that affect spray dryer design and/or performance are:

• Flue gas flow rate and composition

• Temperature of flue gas exiting the boiler and spray dryer

• Alkaline stoichiometric ratio

• Alkaline (lime or sodium) properties

• Required removal efficiency

The most important parameter in sizing the spray dryer is to ensure an adequate gas residence time at maximum gas flow rate to prevent wet solids at the spray dryer outlet. Gas residence times are generally in the range of 10 to 15 seconds for most commercial systems. Due to the large volume of gases that must be treated from utility boilers, these installations often have multiple dryers per boiler.

The spray dryer outlet temperature is controlled by the amount of water injected either with the alkaline slurry or as makeup water. The key to achieving good SO2 removal is to maintain the temperature of the flue gas exiting the spray dryer as close above its dew point (adiabatic saturation) as possible without actually saturating the flue gas. Generally, 20 - 30°F above the adiabatic saturation point is a good target range. This will enhance the reaction yet still prevent condensation. The amount of water that can evaporate in a spray dryer is dependent on the incoming flue gas temperature and to a lesser extent on the moisture content.

The alkaline feed rate is a function of the incoming acid gas levels and the required removal efficiency. The stoichiometric ratio is defined as the molar ratio of alkaline (i.e. calcium) in the spray dryer feed to the amount of acid gases (SO2 and HCl) present. For example, at a ratio of 1.0 the moles of calcium are equal to the moles of incoming HCl and SO2. However, due to inefficiencies in the mixing process, more than the theoretical amount of alkaline material is required to assure compliance with applicable standards. Thus, stoichiometric feed rates of 1.5 to 2.5 have been used to achieve SO2 removal level in the 75 to 85% range and HCl removal efficiencies of 95% on municipal waste combustors. For utility and industrial boilers, sulfur removal guarantees by spray dryer vendors have ranged from 60 to 90%. Table 7-2 lists information on operating spray dryer systems at utility boilers.


2.0-7/98 7-11

Table 7-2. Summary of spray-dryer applications

Size

Unit

Type1

Atomizers per Dryer/Number of

Dryers

RT2 (s)

T3 (°F)

∆T4(°F)

Spray-Dryer

Diameter (ft)

Utilities (size in Megawatts)

440 (each)

Antelope Valley 1, 2

Basin Electric

R 1/5 12 310 20 46

110 Riverside 6, 7

Northern States Power

R 1/1 10 350 Var.5

46

60 Stanton 10

United Power

R 3/1 10 323 20 38

450 Craig 3

Colorado-Ute

N 12/4 NR6 NR 25 NR

280 Rawhide 1

Platte River Power

R 1/3 11 276 NR 46

320 Holcomb 1

Sunflower Coop.

R 1/3 10.6 249 50 51

44 Shiras 3

City of Marquette

R 1/1 10 265 25 36

270 North Valmy

Sierra Pacific Power

Idaho Power

R 3/3 NR 260-300

30 NR

570 Laramie River 3

Basin Electric

N 12/4 8 286 23 55

370 Springerville 1, 2

Tucson Electric

R 1/3 12 256 20 46

575 GRDA R 3/4 12 310 20 52

Industries (size in acfm)

75,000 Argonne National Lab

Argonne, IL

R 1/1 12 330-340

≈20 25

90,500 Container Corp.

Philadelphia, PA

R 1/1 NR 350 NR NR

46,500

(3 units)

Fairchild Air Force Base

Spokane, WA

R 1/1 NR 375 ≈25 20


Lesson 7 ___________________________________________________________________________________

7-12 2.0-7/98

Table 7-2. (continued) Summary of spray-dryer applications.

Size

Unit

Type1

Atomizers per Dryer/Number of

Dryers

RT2 (s)

T3 (°F)

∆T4(°F)

Spray-Dryer

Diameter(ft)

Industries (size in acfm)

167,000 General Motors

Buick Division

Flint, MI

R 1/1 NR 300 NR 32

48,600 Griffis Air Force Base

Rome, NY

R 1/1 NR 400 NR 22

44,400 Malstrom Air Force Base

Great Falls, MI

R 1/1 NR 325 ≈35 20

40,000 Strathmore Paper

Woronco, MA

N 4/1 NR NR NR NR

62,000

96,000

University of Minnesota

N

N

1/1

1/1

12

375

20

24

97,000 Rockwell International

Columbus, OH

R 1/1 12 450 30 30

205,000 M. M. Carbon

Long Beach, CA

R 1/1 10 405 90-120

36

81,710 Ohio State University

Columbus, OH

N NR/1 NR 400 NR 24

1. R = rotary; N = nozzle. 2. Residence time. 3. Flue-gas temperature at entrance. 4. Approach to saturation at exit. 5. Varies. 6. Not reported. Source: Huang et al. 1988.


2.0-7/98 7-13

Spray Drying Equipment

In a spray drying system, there are a number of system components. Three of the major components are the atomizer, spray dryer chamber and particulate control system. An overview of these systems is provided in the following sections and was adapted from Spray-Dryer Flue-Gas-Cleaning System Handbook (Huang et al. 1988).

Atomizers Currently, two types of atomizers are used in spray dryers for acid gas removal: rotary disks or wheels and dual-fluid nozzles. In either case, the purpose of the atomizer is to break the sorbent slurry into a cloud of fine droplets to promote intimate sorbent contact with the acid bases.

In the rotary atomizer, the slurry is fed into the top of the rotating wheel or disk. Centrifugal force causes the slurry to form a thin film on the internal surface of the cavity. As the slurry emerges from the cavity through abrasion-resistant inserts in the side of the wheel, the liquid is atomized into discrete droplets that are propelled radially outward. These droplets, generally 25-150 µm in diameter, dry rapidly in the hot flue gas within the spray dryer. Figure 7-4 shows an example of a typical atomizer wheel used in spray dryers (Huang 1988).

Figure 7-4. Example of rotary atomizer used in spray-dryer FGD systems

��

� ��

Lesson 7 ___________________________________________________________________________________

7-14 2.0-7/98

For FGD spray dryer applications, atomizer wheels range from 8 to 16 inches in diameter and have rotational speeds from 7,000 to 20,000 revolutions per minute (rpm). Due to the highly abrasive nature of the slurry (which can consist of either slaked lime [Ca(OH)2] or slaked lime plus recycled fly ash/reacted product), the wheels are constructed of corrosion- and abrasion-resistant materials, including ceramic inserts in the vanes or nozzles.

In dual-fluid pneumatic nozzle atomization, the slurry feed is injected into the body of a nozzle and is entrained into a high-velocity, high-pressure air stream as shown in Figures 7-5 and 7-6 (Maurin 1983). The high-velocity air impacts on the slurry-feed stream, resulting in the production of fine droplets. The air stream and slurry comprise the two fluids. The size of liquid droplets produced decreases as the compressed air pressure and relative velocity of the liquid to air increases.

Figure 7-5. Two-fluid nozzle atomizer (nozzle body)

��

��

��

��

��

�� !��


2.0-7/98 7-15

Figure 7-6. Two-fluid nozzle atomizer (high pressure air stream)

The mean droplet size for both atomizing systems has been shown to be the same, indicating that the systems perform similarly. Likewise, the capacity of a nozzle system for atomization of slurries is the same as that for a rotary atomizer. Nevertheless, rotary atomizers and pneumatic nozzles have somewhat different advantages and disadvantages (Huang 1988):

1. Rotary atomizers, with their higher capacity per unit, will have a simpler piping system. In a rotary-atomizer system, usually only one feed pipe per atomizer is used; whereas in a nozzle-type atomizer, there will be an individual feed pipe (and valve) to each nozzle. In very large installations, this results in a complex piping system.

2. Pneumatic nozzle atomizers are much easier to maintain than rotary atomizers while the system is on-line because the individual feed lines have isolation and control valves. With multiple nozzles, it is possible to isolate an individual nozzle, remove it for cleaning or replacement, and then return the cleaned or new nozzle to service without reducing the gas flow to the system or bypassing the gas flow to another spray dryer.

3. The net-energy requirements of a rotary atomizer and a set of pneumatic nozzles are approximately the same, but the method by which this energy is applied is different. For a rotary atomizer, the atomization energy is supplied via a motor coupled to the atomizing wheel with a gear and/or belt drive. For a pneumatic atomizer, the energy of atomization is produced primarily by the pressure of the atomizing air. Hence, the energy is supplied through an air compressor that may also supply air for instrumentation or other purposes.

��

��

"��

#��

$� �� %&�� !��

��!'

( �� !�� )�

�*

Lesson 7 ___________________________________________________________________________________

7-16 2.0-7/98

4. A spare rotary atomizer is often required as a backup in case of failure. In a pneumatic nozzle system, the required spares consist of nozzles and an extra air compressor. For a smaller single rotary-atomizer unit, the relative cost of a spare atomizer would be substantial.

Spray-Dryer Chamber The atomization method chosen will affect the design of the spray-dryer chamber, including the physical dimensions. For a rotary-atomizer type of spray dryer, which projects the droplets radially outward and perpendicular to the gas flow, the length-to-diameter ratio of the dryer (L/D) is typically 0.8:1. Figure 7-7(a) illustrates two typical configurations of rotary atomizer spray dryers. The droplets decelerate rapidly due to the drag forces of the downward-moving flue gas and eventually attain the speed and direction of the flue gas. To avoid wall deposition, the designed radial distance between the atomizer and the dryer wall must be sufficient to allow for adequate drying of the largest droplets. This is accomplished by proper choice of the L/D, droplet size, and residence time.

For a two-fluid pneumatic nozzle spray dryer [shown in Figure 7-7(b)], which atomizes the droplets in the direction of the gas flow (downward), the L/D is typically 2:1. In this case, sidewall deposition is a minor problem.

Typically, industrial boiler spray dryers have diameters of 25-30 ft, whereas utility spray dryers have diameters of 40-50 ft. Currently, the maximum diameter of an installed spray dryer is about 60 ft. In general, if the gas-flow rate is large enough that a single unit greater than 40-50 ft in diameter would be specified, then the installation of multiple spray dryers should be considered. In utility systems where the gas flow can range from 1-2 million acfm, multiple spray dryers are common. Multiple spray dryers are installed for easy maintenance and high reliability.

Flue gas may enter a spray dryer in one of three patterns relative to the slurry direction: cocurrent, countercurrent, or mixed. In cocurrent spray dryers, all of the gas enters through a roof gas disperser in the top of the vessel, where its rotation is controlled by angled vanes that direct the gas around the atomizer [shown in Figure 7-7(a)]. This type of gas distribution precisely controls the exit gas temperature since the gas and slurry travel in the same direction. This is the most common flow pattern used in acid gas control systems.

In countercurrent spray dryers, the gas enters from the bottom of the vessel and is directed at the atomized liquid above. Although uncommon in utility or industrial flue-gas control systems, these spray dryers have the advantage of a much higher drying capacity than the cocurrent system.

Another type of spray dryer, the compound-gas disperser or mixed, is offered by one manufacturer as an option in specific applications. This type of spray dryer is sometimes used on very large units as an alternative to multiple rotary atomizers to obtain efficient contact between the hot gas and the liquid droplets.


2.0-7/98 7-17

Figure 7-7. Two types of spray-dryer chambers

%�' +��,�� ,��

��

-��

.��

.��

.��

-��, ��

.��

-��

/��

.��

%�' #0�� ! �� ,��

�� ,��

.��

-��

.��

Lesson 7 ___________________________________________________________________________________

7-18 2.0-7/98

Particulate-Matter Collection

A spray-dryer system is not complete without a means of particulate-matter collection. Not only is a well-designed particulate-matter control system needed to meet emissions requirements, but it also aids in acid-gas removal. Acid gases are removed when the flue gas comes in contact with lime-containing particles in the fabric filter or ESP. Fabric filters have been used on the majority of acid gas control systems, due to their ability to obtain slightly higher residual acid gas removal than ESPs.

Regardless of the type of particulate control device, an important design feature is to minimize potential heat loss in the fly ash collection system. The fly ash contains unreacted alkaline sorbent along with calcium (or sodium) sulfates, and in the case of waste incinerators, it also contains calcium chlorides. These materials are very hygroscopic and can result in corrosion problems or ash plugging of equipment if condensation occurs. Adding insulation, hopper heaters and reducing air in-leakage are essential to prevent operational problems with the ash handling system.


Except for rotary atomizers in spray dryers, dry injectors and spray dryer absorbers are relatively simple devices with few moving parts. (Note: Maintenance associated with an atomizer is specific to the type and manufacturer of the atomizer and is not covered in this lesson). The primary maintenance problem associated with any dry scrubbing system is potential plugging in the solid or slurry transport systems. Manufacturers of the various systems provide suggested maintenance and inspection schedules for each component. These schedules should be followed and information recorded to aid in documenting the system operation.

Dry scrubbing systems involve transporting a solid or slurry (which can be 10 to 40% solids) in small pipes; therefore, plugging problems could occur in a number of locations. The most common locations of plugging problems are in "dead" areas of the solid or slurry piping, valves and the atomizer. Dead areas of piping are associated with tees going to spare pumps or a cleanout port. In these areas, flow only occurs occasionally and provides an area for solids to buildup and block transport lines. Eliminating the tees is not practical since redundancy is needed (and often mandated) in order to assure continual operation of the scrubbing system. Also, certain tees are installed specifically to allow quick access to piping internals in order that a specific length of pipe can be flushed with water to dislodge buildup. Flexible rubber hosing and quick-type connectors have been used to try and minimize line plugging. Flexible piping is not as susceptible to plugging as solid pipe, and with the use of quick connectors, the flexible piping can be installed or removed quickly to flush out areas or to connect spare components.

Plugging problems associated with valves and atomizers in slurry systems are minimized by using screens in transport lines to remove solids. However, these screens must be periodically checked and cleaned or else they will cause plugging. Atomizing systems are often designed so that they can be flushed with water during operation (this will temporary reduce potential acid gas removal efficiency). Atomizers should also be designed so that they can be replaced in a short timeframe.


2.0-7/98 7-19

Another area of maintenance with semi-dry scrubbing systems is the lime slaking system. Lime slaking is the process of mixing controlled amounts of water and lime in a mixing vessel (slaker). The lime and water react (an exothermic process) to form the lime slurry which is then screened of grit, stored in agitated mixing tanks and then metered to the atomizer.

With the slaking system, plugging and the quality of slurry produced are the two biggest maintenance concerns. Plugging of dead spots and pumps can occur as already discussed. In addition, the grit screening process is of concern because if the screen is damaged, then large quantities of grit can get into the entire lime-slurry transport system causing extensive plugging and/or abrasion wear problems. Slurry quality is dependent on the quality of lime and slaking water utilized in addition to the mechanical action of the slaker. Both the lime and water should be of high quality (limited contaminants or other chemicals present) to prevent adverse reactions that can result in scaling, plugging or reduced acid gas removal efficiencies. The mechanical action of the slaker will determine how efficiently the slaking reaction occurs. The slaker should be frequently inspected to ensure that it is operating as designed.

Plugging of lime slurry transport components can also occur due to a lack of slurry movement (i.e. during standby periods) when solids could settle out or the calcium could have time to react and form scale. During extended downtimes, lines and storage tanks should be drained and flushed where practical. Also, manufacturers recommend periodic cleaning in acid of screens and other components that are prone to plugging problems.

Summary

Dry scrubbing systems are used to control acid gas emissions primarily from combustion sources such as utility and industrial boilers and municipal and medical waste incinerators. Dry scrubbing systems only remove acid gases and therefore must be followed by a particulate control device (ESP or fabric filter) prior to exhausting the gases to the atmosphere.

Dry scrubbing systems can be categorized as dry sorbent injectors (DSI) or as semi-dry scrubbers (also referred to as spray dryer absorbers or spray dryers). Dry sorbent injection involves the addition of a dry alkaline material (usually hydrated lime or soda ash) into the gas stream to react with any acid gases that are present. The sorbent can be injected directly into the flue gas duct ahead of the particulate control device or into an open reaction chamber. The acid gases are adsorbed onto and react with alkaline sorbents to form solid salts which are removed in the particulate control device.

In spray dryer absorbers (SDAs) the flue gases are introduced into an absorbing tower (dryer) where the gases are contacted with a finely atomized alkaline slurry: usually a calcium-based sorbent such as Ca(OH)2 or CaO. Acid gases are absorbed by the slurry droplets and react to form solid salts. The heat of the flue gas is used to evaporate all the water droplets, leaving a non-saturated (i.e. dry) flue gas exiting the absorber tower. The effect of cooling and humidifying the hot gas stream increases collection efficiency over simple dry injection.

The major components of a spray dryer absorber are the atomizer, spray dryer chamber and the particulate control device. Two types of atomizers are currently utilized for acid gas removal: rotary disks (wheel type) and dual-fluid nozzles. In either case, the purpose of the

Lesson 7 ___________________________________________________________________________________

7-20 2.0-7/98

atomizer is to break the sorbent slurry into a cloud of fine droplets. The spray chamber is designed based on the type of atomizer utilized. Spray chambers used with rotary atomizers are shorter but fatter than those used with two-fluid nozzle atomizers. Both ESPs and baghouses have been used with spray dryer absorbers. An important design feature of the particulate control device is to minimize potential heat loss in the fly ash collection system to prevent potential plugging problems.

An important parameter in the operation of a dry scrubbing system is the amount of alkaline material feed into the system. The amount of sorbent required is a function of the type of sorbent used, inlet and outlet (required removal) acid gas levels and the effectiveness of the dry scrubbing system design. The amount of sorbent added is generally reported as the stoichiometric ratio on a molar basis of sorbent to acid gases. A stoichiometric ratio of 1:1 would be used under ideal conditions; in practice more than the theoretical amount must be utilized to assure compliance with required acid gas removal levels.

Except for rotary atomizers in spray dryers, dry scrubbing systems are relatively simple devices with few moving parts. The primary maintenance problem is potential plugging in the solid or slurry transport lines. Plugging can occur whenever there are bends or restrictions in piping.



2.0-7/98 7-21


1. Dry scrubbing systems are used to remove ____________________ ____________________ from flue gas streams.

2. True or False? In dry scrubbing systems no water or slurry is ever used.

3. In dry scrubbing, the following mechanisms are applicable:

a. Absorption b. Adsorption c. Impaction d. a and b, only

4. In general, higher acid gas removal efficiencies are achievable as the operating temperature of the dry scrubbing system:

a. Increases b. Decreases c. Does not change d. All of the above

5. The ratio of the sorbent materials injected into the spray dryer relative to the acid gases present is referred to as the ____________________ ____________________.

6. The alkaline sorbent used in spray drying systems is:

a. Calcium based b. A form of lime or soda ash c. Sodium based d. a and b, only e. a, b, and c

7. In a scrubbing system, HCl reacts ____________________ with the sorbent than SO2 does.

a. Faster b. Slower c. At the same rate d. None of the above

Part 2

8. True or False? Dry sorbent injection is a very simple process that involves injecting a solid into the flue gas.

9. Spray dryer gas residence times are generally in the range of:

a. 1 to 2 seconds b. 10 to 15 seconds c. 1 to 2 minutes d. a or b

Lesson 7 ___________________________________________________________________________________

7-22 2.0-7/98

10. True or False? The spray dryer outlet temperature should be maintained as close above its adiabatic saturation temperature as possible without risking condensation to obtain best acid gas removal rates.

11. The amount of water that can evaporate in a spray dryer is dependent on the:

a. Acid gas levels b. Sorbent type c. Incoming temperature d. All the above

12. For a given system design, what is the alkaline sorbent feed rate a function of? ________________________________________ and ________________________________________

13. True or False? Spray dryers can operate at stoichiometric ratios of less than 1.0 and achieve very high (90+) removal efficiencies.

14. The two types of atomizer systems used on spray dryers are ____________________ ____________________ and ____________________ ____________________ ____________________.

15. True or False? The droplet size produced and power consumption of a rotary atomizer and dual-fluid nozzle system are essentially the same.

16. The spray chamber length for a dual fluid nozzle system should be ____________________ than for a system with a rotary atomizer.

a. Shorter b. Longer c. About the same d. Any of the above

17. The particulate matter control device on spray drying systems removes particles and can aid in additional ____________________ ____________________ removal.

18. Fly ash collection systems on spray dryers must be properly insulated and heated to prevent condensation which could cause:

a. Plugging b. Corrosion c. Reentrainment d. a and b, only

19. The primary maintenance problem with dry scrubbing systems is:

a. Plugging in the sorbent transport system b. Scaling c. Corrosion d. Erosion


2.0-7/98 7-23

20. Lime slurry is dependent on the mechanical action of the slaker and the quality of the:

a. Water b. Lime c. Soda ash d. a and b, only

Lesson 7 ___________________________________________________________________________________

7-24 2.0-7/98


2.0-7/98 7-25


1. Acid gases Dry scrubbing systems are used to remove acid gases from flue gas streams.

2. False In dry scrubbing systems, water or slurry is sometimes used. Semi-dry systems (also called spray dryer absorbers) use an alkaline slurry.

3. d. a and b, only In dry scrubbing, the following mechanisms are applicable: absorption and adsorption.

4. b. Decreases In general, higher acid gas removal efficiencies are achievable as the operating temperature of the dry scrubbing system decreases.

5. Stoichiometric ratio The ratio of the sorbent materials injected into the spray dryer relative to the acid gases present is referred to as the stoichiometric ratio.

6. e. a, b, and c The alkaline sorbent used in spray drying systems can be any of the following:

• Calcium based

• A form of lime or soda ash

• Sodium based

7. a. Faster In a scrubbing system, HCl reacts faster with the sorbent than SO2 does.

Part 2

8. True Dry sorbent injection is a very simple process that involves injecting a solid into the flue gas.

9. b. 10 to 15 seconds Spray dryer gas residence times are generally in the range of 10 to 15 seconds.

10. True The spray dryer outlet temperature should be maintained as close above its adiabatic saturation temperature as possible without risking condensation to obtain best acid gas removal rates.

11. c. Incoming temperature The amount of water that can evaporate in a spray dryer is dependent on the incoming temperature.

12. Incoming acid gas levels Removal rate For a given system design, the alkaline sorbent feed rate is a function of the incoming acid gas levels and removal rate.

Lesson 7 ___________________________________________________________________________________

7-26 2.0-7/98

13. False Spray dryers cannot operate at stoichiometric ratios of less than 1.0 and achieve very high (90+) removal efficiencies. They must operate at ratios above 1.0 to be effective.

14. Rotary atomizers (and) dual-fluid nozzles (or two-fluid nozzles) The two types of atomizer systems used on spray dryers are rotary atomizers and dual-fluid nozzles.

15. True The droplet size produced and power consumption of a rotary atomizer and dual-fluid nozzle system are essentially the same.

16. b. Longer The spray chamber length for a dual-fluid nozzle system should be longer than for a system with a rotary atomizer because of the type of spray pattern required by dual-fluid nozzle.

17. Acid gas The particulate matter control device on spray drying systems removes particles and can aid in additional acid gas removal.

18. d. a and b, only Fly ash collection systems on spray dryers must be properly insulated and heated to prevent condensation which could cause plugging and corrosion.

19. a. Plugging in the sorbent transport system The primary maintenance problem with dry scrubbing systems is plugging in the sorbent transport system.

20. d. a and b, only Lime slurry is dependent on the mechanical action of the slaker and the quality of the water and lime.


2.0-7/98 7-27

Bibliography

Apple, C., and M. E. Kelly. 1982, April. Mechanisms of Dry SO2 Control Processes. EPA-600/7-82-026, NTIS PB 82-196924. U.S. Environmental Protection Agency.

Huang, H., J. W. Allen, C. D. Livengood, W. T. Davis, and P. S. Farber. 1988. Spray-Dryer Flue-Gas-Cleaning System Handbook. U.S. Department of Energy. Publication No. ANL/ESD-7. Energy Systems Division, Argonne National Laboratory.

Maurin, P. G., et al. 1982, April. Two-fluid nozzle vs. rotary atomization for dry-scrubbing systems. Chemical Engineering Progress. (pp. 51-59).

U.S. Environmental Protection Agency. 1982, September. Flue Gas Desulfurization - Spray Dryer Process. Sulfur Oxides Control Technology Series. EPA 625/8-82-009.

Lesson 7 ___________________________________________________________________________________

7-28 2.0-7/98

2.0-7/98 8-1

Lesson 8 Equipment Associated with Scrubbing Systems

Goal To familiarize you with the operation of equipment associated with scrubbing systems.


1. Briefly describe the operation of a centrifugal fan

2. Distinguish between forced- and induced-draft fans

3. List two maintenance problems associated with fans, pumps, ducts, and pipes in wet scrubbing systems

4. List three types of pipe materials used in scrubbing systems and the advantages and disadvantages of each

5. Briefly describe the function of quenchers

6. Describe three spray nozzle designs and identify two maintenance problems associated with nozzles

7. Describe the operation of three mist eliminators and identify two diagnostic monitoring techniques to ensure proper functioning of these components

8. List five important variables that should be monitored in scrubbing systems

Introduction

Many components comprise a complete scrubbing system. In previous lessons we have focused only on the operation of the scrubbing vessel itself. To fully understand the operation of a scrubber, it is important to have a basic knowledge of all the components of the system. For instance, fans and ducts are required to transport exhaust gas while pumps, nozzles, and pipes transport liquid to and from the scrubbing vessel. Water-recirculation and mist-elimination systems are also necessary. In addition, many systems use a quench ahead of the scrubber to humidify and cool the flue gases. Failure of any of these parts can cause problems for the entire scrubbing system. Finally, monitoring and recordkeeping are required not only to document but to prevent potential problems. This lesson presents an overview of

Lesson 8 ___________________________________________________________________________________

8-2 2.0-7/98

the equipment associated with scrubbing systems covering their operation and some typical maintenance problems.

Transport Equipment For Exhaust Gases and Scrubbing Liquids

Fans transport (push or pull) exhaust gases through ducts to and from the scrubber, while pumps transport liquids through pipes. Although not part of the scrubber chamber, both fans and pumps are essential to its operation.

Fans

Fans in scrubbing systems are usually centrifugal. In centrifugal fans, exhaust gas is introduced into the center of a revolving wheel, or rotor, and exits at a right angle (90°) to the rotation of the blades (Figure 8-1). Centrifugal fans are classified by the type and shape of blades used in the fan. The forward-curved fans use blades that are curved toward the direction of the wheel rotation. The blades are smaller and spaced closer together than the blades in other centrifugal fans. These fans are not usually used if the flue gas contains dust or sticky materials. They have been used for heating, ventilating, and air conditioning applications in industrial plants. Backward-curved fans use blades that are curved away from the direction of wheel rotation. The blades will clog when the fan is used to move flue gas containing dust and sticky fumes. They may be used on the clean-air discharge of air pollution control devices or to provide clean combustion air for boilers. Radial fans use straight blades that are attached to the wheel of the rotor. These fans are built for high mechanical strength and can be easily repaired. Airfoil fans use thick teardrop-shaped blades that are curved away from the wheel rotation. Airfoil fans can clog when handling dust or sticky materials.

Fan blades may be constructed of alloys or coated steel to help prevent deterioration when handling abrasive and corrosive exhaust gas. Radial fans are used most frequently for air pollution control applications; however, backward-curved fans are also used on wet scrubbing systems.

Equipment Associated with Scrubbing Systems ___________________________________________________________________________________

2.0-7/98 8-3

Figure 8-1. Types of centrifugal fans

Lesson 8 ___________________________________________________________________________________

8-4 2.0-7/98

Fans in scrubbing systems can be located before or after the scrubber. When located before the scrubber, they are referred to as forced-draft, positive-pressure, or dirty-side fans. These fans normally move dry air, but can move moist air depending on process conditions. They are subject to abrasion and solids buildup when dust concentration is high. Abrasion on the fan can be reduced by using special wear-resistant alloys, by using replaceable liners on the wheel, or by reducing fan speed (using a large fan that moves more slowly). The solids buildup can sometimes be controlled by using a spray wash to periodically clean the wheel. If dirty-side fans are used, a cyclone or knockout chamber can be placed before the fan to reduce dust concentration.

Fans located after scrubbers are always operated wet, and are called induced-draft, negative-pressure, or clean-side fans. These fans are subject to corrosion and solids buildup from mist escaping from the entrainment separator. Corrosion problems can result when the exhaust gas contains acid-forming or soluble electrolytic compounds, especially if the temperature of the gas stream falls below the dew point of these compounds. Corrosion can be reduced by using proper construction materials and careful pH control in the scrubbing system. Solids buildup can occur when the mist escaping from the entrainment separator contains dissolved or settleable solids. As the mist enters the fan, evaporation occurs and some solids deposit on the wheel. If the buildup on the wheel is uniform, no problems occur until the buildup starts to flake off, knocking the fan out of balance (Wechselblatt 1975). Keeping entrainment separators operating efficiently or using clean water sprays on the fan blades will help reduce solids-buildup problems.

Ducts

Ducts, or ductwork, transport exhaust gas to and from the scrubber. Ducts are carefully designed to keep pressure losses and, consequently, operating costs at a minimum. In general, this requires sizing the duct properly and minimizing the number of bends, expansions, and contractions. Sizing the duct to suit the exhaust stream velocity will also reduce the amount of dust that settles in the ductwork.

Abrasion and corrosion are common problems of ductwork. Abrasion is generally more severe on ductwork leading into the scrubber, while corrosion affects ductwork leaving the scrubber. Using proper construction materials or linings greatly reduces corrosion or abrasion. For example, ductwork can be lined partially or fully with brick (especially at elbows) to prevent erosion due to abrasion. For ductwork exiting the scrubber, special alloys resistant to acid attack should be used. Also, ductwork can be insulated to prevent acids in the flue gas from condensing.


2.0-7/98 8-5

Pumps

A wide variety of pumps are used to transport both the scrubbing liquid and the sludge. The proper choice of a pump depends on flow rate, pressure, temperature, and material being pumped. Electric-motor-driven centrifugal pumps are the pumps most frequently used in wet scrubbing systems (Calvert et al. 1972). In a centrifugal pump, the rotating impeller produces a reduction in pressure at the eye (center) of the impeller, causing liquid to flow from the suction pipe into the center of the impeller. The liquid is then forced outward along the blades and discharged generally at a 90 degree angle.

As with fans, abrasion and corrosion are the major maintenance problems associated with pumps in scrubbing systems. The impellers, housing and seals are subject to potential corrosion and abrasion problems. Abrasion is caused by solids buildup in the scrubbing liquid. Bleeding this liquid and removing the solids before recycling it back through the pump (or scrubber) will reduce pump wear. Most vendors suggest that the solids content be less than 15% (EPA 1982). Special alloys or rubber linings can also be used to help reduce abrasion and corrosion.

Pipes

Pipes transport liquid to and from the scrubber. As with pumps, pipes are susceptible to abrasion, corrosion, and plugging. Pipes can be made from a wide variety of materials to reduce these problems. Some advantages and disadvantages of pipe materials commonly used are given in Table 8-1.

To prevent solids from building up in or plugging the pipe, a liquid slurry velocity in the scrubbing system of 1.2 to 2.1 m/s (4 to 7 ft/sec) is recommended as a reasonable compromise (Czuchra 1979).


Lesson 8 ___________________________________________________________________________________

8-6 2.0-7/98

Table 8-1. Pipe materials for scrubber systems advantages and disadvantages

Material Advantages Disadvantages

Metals

Cast iron Flanged, threaded, or welded

Not resistant to corrosion

Steel Inexpensive

Stainless steel Easy to cut and install on site

Copper alloys

Metal pipe linings

Hard rubber Good resistance to many strong acids and alkalis

Cannot be cut to size on site

Soft rubber Resists abrasion Must be precisely manufactured

Glass Resists acid and alkali attack

Fragile

Thermoplastic PVC Polyethylene Polypropylene

Resists corrosion

Easily site-installed

Good resistance to temperature and stress

Not as abrasion resistant as rubber or stainless steel

Nonmetals

Plastic Resists corrosion May not be as heat resistant as other materials

Fiberglass-reinforced pipe (FRP)

Resists chemical corrosion

On-site installation

Less abrasion resistant than rubber-lined pipe

Operates at higher temperatures than a solid plastic pipe

Adapted from Calvert et al. 1972.


2.0-7/98 8-7

Quenchers

Occasionally, hot exhaust gas is quenched or cooled by water sprays before entering the scrubber. Hot gases (those above ambient temperature) are often cooled to near the saturation level. If not cooled, the hot gas stream can evaporate a large portion of the scrubbing liquor, adversely affecting collection efficiency and damaging scrubber internal parts. If the gases entering the scrubber are too hot, some liquid droplets may evaporate before they have a chance to contact pollutants in the exhaust stream, and others may evaporate after contact, causing captured particles to become reentrained. In some cases, quenching can actually save money. Cooling the gases reduces the temperature and, therefore, the volume of gases, permitting the use of less expensive construction materials and a smaller scrubber vessel and fan.

A quenching system can be as simple as spraying liquid into the duct just preceding the main scrubbing vessel, or it can be a separate chamber (or tower) with its own spray system identical to a spray tower.

Quenchers are designed using the same principles as scrubbers. Increasing the gas-liquid contact in them increases their operational efficiency. Small liquid droplets cool the exhaust stream more quickly than large droplets because they evaporate more easily. Therefore, less liquid is required. However, in most scrubbing systems, approximately one-and-a-half to two-and-a-half times the theoretical evaporation demand is required to ensure proper cooling (Industrial Gas Cleaning Institute 1975). Evaporation also depends on time - it does not occur instantaneously. Therefore, the quencher should be sized to allow for an adequate exhaust-stream residence time. Normal residence times range from 0.15 to 0.25 seconds for gases under 540°C (1000°F) to 0.2 to 0.3 seconds for gases hotter than 540°C (Schifftner 1979).

Quenching with recirculated scrubber liquor could potentially reduce overall scrubber performance, since recycled liquid usually contains a high level of suspended and dissolved solids. As the liquid droplets evaporate, these solids could become reentrained in the exhaust gas stream. To help reduce this problem, clean makeup water can be added directly to the quench system rather than adding all makeup water to a common sump (EPA 1982).

Spray Nozzles

Three different nozzle designs are used to produce a fine, cone-patterned spray. In the impingement nozzle (Figure 8-2), highly pressurized liquid passes through a hollow tube in the nozzle and strikes a pin or plate at the nozzle tip. A very fine fog of tiny, uniform-sized droplets approximately 25 to 400 µm in diameter is produced. Because there are no internal parts in the nozzle, it will not plug as long as particles larger than the opening are filtered out by a strainer. These nozzles are usually made of stainless steel or brass. In the solid cone nozzle (Figure 8-3), liquid is forced over an insert to break it up into a cone of fine droplets. Cones can be full, hollow, or square with spray angles from 15° to 140°. These nozzles can be made of stainless steel, brass, alloys, Teflon, and other plastic materials. The helical spray nozzle (Figure 8-4), has a descending spiral impingement surface that breaks up the sprayed liquid into a cone of tiny droplets. The cones can be full or hollow with spray angles from 50° to 180°. There are no internal parts, which helps reduce nozzle plugging. These nozzles can be made of stainless steel, brass, alloys, Teflon, and other plastic materials.

Lesson 8 ___________________________________________________________________________________

8-8 2.0-7/98

Figure 8-2. Impingement nozzle

Figure 8-3. Solid cone nozzle


2.0-7/98 8-9

Figure 8-4. Helical spray nozzle

Different spray nozzles are appropriate for different scrubbing systems. Characteristics of the nozzles and sprays include the following:

1. Droplet size - In general, scrubbers using sprays to provide gas-liquid contact (such as in spray towers) require tiny, uniform-sized droplets to operate effectively. If the sprays are used merely as a method of introducing liquid into the vessel (such as in packed towers), then droplet size is not as critical.

2. Opening size - The actual opening in the nozzle will vary depending on the applications and the amount of liquid required. Openings range from 0.32 to 6.4 cm (0.125 to 2.5 in.).

3. Spray pattern - Nozzles are available that produce sprays in a number of geometric shapes such as square, fan, hollow cone, and full cone. Full-cone sprays are used to provide complete coverage of the areas sprayed.

4. Operating mechanism - Droplets can be produced by a number of methods such as impinging the liquid on a solid surface or atomizing the liquid using air.

5. Power consumption - In general, the finer the liquid droplet, the higher the power consumption.

Lesson 8 ___________________________________________________________________________________

8-10 2.0-7/98

Nozzle plugging is one of the most common malfunctions in scrubbers. Plugged nozzles reduce the gas-liquid contact and can also result in scale buildup on, or heat damage to, the scrubber parts formerly sprayed by the nozzle. Nozzle plugging can be most readily detected by observing the liquid spray pattern; however, if the nozzles are not easily accessible, a decrease in liquid flow is also a telltale sign (EPA 1982). Remedies include the following: (1) replacing the nozzle with one that is more open, (2) cleaning the nozzle frequently, (3) filtering the scrubbing liquid, (4) increasing the bleed rate and makeup water rates.

Another problem that can arise is reduced pressure in the spray header. This can cause a reduction in the spray angle (area covered) and an increase in the size of droplets produced.

Entrainment Separators

As mentioned in Lesson 1, the pollutant must first be contacted with the liquid, then the liquid droplets must be removed from the exhaust gas stream before it is exhausted to the atmosphere. Entrainment separators, also called mist eliminators, are used to remove the liquid droplets prior to exhausting gases to the atmosphere. Although the major function of an entrainment separator is to prevent liquid carryover, it also performs additional scrubbing and recovers the scrubbing liquor, thus saving on operating costs. Therefore, entrainment separators are usually an integral part of any wet scrubbing system.

Entrained liquid droplets vary in size depending on how the droplets were formed. Droplets that are torn from the body of a liquid are large (10 to 100 µm in diameter), whereas droplets that are formed by a chemical reaction or by condensation are on the order of 5 µm or less in diameter. Numerous types of entrainment separators are capable of removing these droplets. Those most commonly used for air pollution control purposes are cyclonic, mesh-pad, and blade separators.

The cyclonic (centrifugal) separator, which is commonly used with venturi scrubbers (see Lesson 3), is a cylindrical tank with a tangential inlet or turning vanes. The tangential inlet or turning vanes impart a swirling motion to the droplet-laden gas stream. The droplets are thrown outward by centrifugal force to the walls of the cylinder. Here they coalesce and drop down the walls to a central location and are recycled to the absorber (Figure 8-5). These units are simple in construction, having no moving parts. Therefore, they have few plugging problems as long as continuous flow is maintained. Good separation of droplets 10 to 25 µm in diameter can be expected. The pressure drop across the separator is 10 to 15 cm (4 to 6 in.) of water for a 98% removal efficiency of droplets in the size range of 20 to 25 µm.


2.0-7/98 8-11

Figure 8-5. Cyclonic separator

In another design, wire or plastic is used to form mesh pads (Figure 8-6). These mesh-pad separators are approximately 10 to 15 cm (4 to 6 in.) thick and fit across the entire diameter of the scrubber. The mesh allows droplets to impact on the material surface, agglomerate with other droplets, and drain off by gravity. The pad is usually slanted (no more than a few degrees) to permit the liquid to drain off. Better than 95% collection of droplets larger than 3 µm is obtained with pressure drops of approximately 1.0 to 15 cm (0.5 to 6 in.) of water. (The pressure drop depends on depth and compaction of fibers). The disadvantage with mesh pads is that their small passages are subject to plugging. Periodically spraying pads from both below and above can remove some trapped material. However, spraying only from beneath will drive entrapped material further into the mesh, necessitating removal of the pads for cleaning or replacement (Schifftner 1979).

��

��

Lesson 8 ___________________________________________________________________________________

8-12 2.0-7/98

Figure 8-6. Mesh-pad separator

Blade separators can be of two types: chevron or impingement. In the chevron separator [Figure 8-7(a)], gas passes between the blades and is forced to travel in a zigzag pattern. The liquid droplets cannot follow the gas streamlines, so they impinge on the blade surfaces, coalesce, and fall back into the scrubber chamber or drain. Special features such as hooks and pockets can be added to the sides of these blades to help improve droplet capture. Chevron grids can be stacked or angled on top of one another to provide a series of separation stages. Pressure drop is approximately 6.4 cm (2.5 in.) of water for capture of droplets as small as 5 µm in diameter. Impingement separators [Figure 8-7(b)], being similar in shape to the common house fan, create a cyclonic motion. As the gas passes over the curved blades, they impart a spinning motion that causes the mist droplets to be directed to the vessel walls, where they are collected. Pressure drop ranges from 5 to 15 cm (2 to 6 in.) of water.

Figure 8-7. Two types of blade separators


2.0-7/98 8-13

The most important diagnostic aid in monitoring separator performance is the pressure drop. By measuring the pressure drop across the separator, the following problems can be identified (Wechselblatt 1975):

• A sudden decrease in pressure drop at constant load indicates that the separators have shifted out of place or are broken.

• An increase in pressure drop, even as little as 0.5 cm (0.2 in.) of water, is an indication of material buildup in the separator.

Another diagnostic measurement is gas velocity. Gas velocity through the separator must be kept below the maximum rate to avoid liquid reentrainment. Maximum velocities depend on operating conditions and the physical properties of the exhaust gas and liquid streams. The gas velocity should be kept below 3 m/s (10 ft/sec) for chevron separators, below 5 m/s (15 ft/sec) for mesh pads, and below 8 m/s (27 ft/sec) for impingement blades to reduce liquid carryover (Schifftner 1979). Table 8-2 summarizes some operating characteristics of entrainment separators.

Table 8-2. Typical operational characteristics of entrainment separators1

Droplet size collected at

99%

Maximum gas velocity

Pressure drop

Type (µm) m/s ft/sec cm H2O in. H

2O

Mesh pads 3.0 5 15 1.0-15 0.5-6

Cyclone 10-25 20 65 10-15 4-6

Blades

Chevron

Impingement vane

35

20

3

8

10

27

6.4

5-15

2.5

2-6

1. Note: Values in this table are given as a general guide only. The collection efficiency for various droplet sizes depends on the gas velocity through the entrainment separators.


Lesson 8 ___________________________________________________________________________________

8-14 2.0-7/98

Construction Materials

By now it should be obvious that scrubbing systems require special materials to prevent or reduce corrosion and abrasion. These are summarized in Table 8-3.

Table 8-3. Construction materials for wet scrubber components

Material Properties/uses Corrosion resistance

Metal

Cast iron High strength; low ductility; brittleness; hardness; low cost

Ordinary cast irons exhibit fair resistance to mildly corrosive environments; high-silicon cast irons exhibit excellent resistance in a variety of environments (hydrofluoric acid is an important exception); cast irons are susceptible to galvanic corrosion when coupled to copper alloys or stainless steels

Carbon steel Good strength, ductility, and workability; low cost

Fair to poor in many environments; low pH and/or high dissolved solids in moist or immersion service leads to corrosion; properly applied protective coatings give appropriate protection in many applications; susceptible to galvanic corrosion when coupled to copper alloys or stainless steels

Martensitic stainless steel (410, 416, 420, 440c)

Chromium alloy, hardenable by heat treatment; typically used for machine parts; costs 2 to 5 times more than carbon steel

Good

Ferritic stainless steel Chromium alloy, not hardenable by heat treatment; costs 2 to 4 times more than carbon steel

Good; better than martensitic stainless steel; resists stress corrosion; better chloride resistance than austenitic stainless steels

405 Modified for weldability

430 General purpose, often used for chimney liners

Good resistance to atmospheric corrosion

442, 446 Used in high-temperature service



2.0-7/98 8-15

Table 8-3. (continued) Construction materials for wet scrubber components


Austenitic stainless steel Chromium and nickel alloy; not hardenable by heat; hardenable by cold working; nonmagnetic

Types 201, 202, 301, 302, 303, 304, and 304L cost 3 to 5 times more than carbon steel; types 310, 316, 316L, and 321 cost 4 to 10 times more than carbon steel

Excellent; better than martensitic or ferritic stainless steel (except for halides)

201, 202 Nitrogen added, used as a substitute for 301 and 302

301 Good hardenability

302 General purpose

304 General purpose

304L Modified for weldability

310 Used in high-temperature service

316

316L

Used in corrosive environments

Improved weldability

Superior corrosion resistance; good acid resistance; resistant to hot organic acids; good pitting resistance

Nickel alloy Good strength; costs over 10 times more than carbon steel

Excellent resistance in most environments; not resistant in strong oxidizing solutions such as ammonium and HNO3

Inconel 1 Good resistance to stress corrosion

Monel 1 Good resistance to hydrofluoric acid

Hastelloy 2 and Chlorimet 3

Excellent overall resistance

Titanium High strength; light weight (60% that of steel); costs over 10 times more than carbon steel

Exceptional resistance at ambient temperatures; excellent resistance at other temperatures, except that crevice corrosion is possible in chloride solutions above 110°C (250°F)


Lesson 8 ___________________________________________________________________________________

8-16 2.0-7/98

Table 8-3. (continued) Construction materials for wet scrubber components


Nonmetal

Glass and glass linings Brittleness, subject to damage by thermal shock; can be protected against breakage by coating with polyester fiberglass

Good resistance to hydrochloric and dilute sulfuric acid

Brick linings Carbon brick Used when fluorides are

present; 540°C (1000°F) temperature limit

Acid resistant and abrasion resistant; also provides thermal protection for inner materials

Acid brick 870°C (1600°F) temperature limit

Silicon carbide brick 1370°C (2500° F) temperature limit; high installation costs

Porcelain and stoneware Same properties but greater strength than glass; easily damaged by thermal shock

Good acid resistance

Rubber Excellent mechanical properties and abrasion resistance; temperature limit of approximately 105°C (220°F)

Resistant to dilute acids, alkalis, and salts, but some oxidizing media will attach to it

Plastics Less resistance to mechanical abuse, lower strength, and higher expansion rates; cannot be used where temperatures constantly exceed 105°C (220°F)

Excellent resistance to weak acids and alkalis; do not corrode and are not affected by slight changes in pH or oxygen content

1. Registered trademark of Huntington Alloys, Inc. 2. Registered trademark of the Stalite Divison of Cabot Corporation. 3. Registered trademark of the Duriron Company, Inc. Sources: EPA 1982 and Perry 1973.

Monitoring Equipment

Having adequate equipment is imperative when monitoring the performance of a scrubber. Instrumentation on a wet scrubber can provide three distinct services:

• Obtaining operational information by recording daily data to help detect any problems or mis-operation that may occur


2.0-7/98 8-17

• Providing operating input for other devices to automatically operate some parts of the system

• Providing for safety by sounding alarms and/or releasing interlocks to protect both the operators and equipment

A monitoring system must be properly installed and maintained to provide reliable data. Monitors should be installed, operated, and calibrated according to the manufacturer's instructions. Because every scrubbing system is unique, the instrumentation and variables measured will vary from source to source. Table 8-4 lists monitors that are typically used in wet scrubbing systems.

Table 8-4. Monitoring equipment for wet scrubbing systems

Monitor Measurements

Manometer Measures pressure drop (inlet and outlet static pressure) across fan, scrubber vessel, and entrainment separator

Thermometer or thermocouple

Measures inlet and outlet temperatures of gas to and from scrubber

Measures inlet and outlet temperatures of liquid to and from scrubber

Flowmeter Measures liquid flow rate to scrubber

Measures the amount of recycled liquid and bleed stream

Measures flow rate of fresh makeup liquid to scrubber

pH meter Measures pH level in chemical feed stream, scrubbing liquid, recycle liquor, and bleed stream

Ammeter Monitors the current of the fans and pumps

For any of these monitors, high and/or low settings can be chosen so that if the set value is exceeded, an alarm sounds, a bypass is opened, or an emergency system is activated. For example, sources that scrub hot gases normally have a high-temperature alarm and/or an interlock system to automatically introduce emergency water or to bypass the scrubber if the high-temperature setting is exceeded.

Pressure Drop

One of the most useful operating parameters monitored on most scrubbing systems (especially venturis) is the static pressure drop (generally just referred to as pressure drop). To provide the most useful information, the pressure drop should be monitored across specific components, (i.e. the scrubber chamber and mist eliminator) instead of across the entire scrubber train. For example, measuring the pressure drop across the mist eliminator will give immediate indications of any plugging or particle buildup.

Static pressure is measured by simply inserting a tube (pressure tap) upstream and downstream of the scrubbing component. Figure 8-8 shows two types of tubes that can be used to measure pressure. Figure 8-8(a) illustrates using a small 1/4 inch copper tube, while Figure 8-8(b) illustrates use of an "s" type pitot tube (EPA 1983). The taps can be connected directly to indicating gauges such as manometers or magnahelics. The devices

Lesson 8 ___________________________________________________________________________________

8-18 2.0-7/98

are located near where the measurements are being made. For information feedback to a central control room, the pressure taps can be connected to a differential pressure transmitter which sends the signal to a monitoring system in the control room. The biggest problems with measuring pressure in a scrubbing system are plugging of the taps and water condensing in the sample lines.

Figure 8-8. Two methods for measuring static pressure


2.0-7/98 8-19

Temperature

Temperature should be monitored both before and after the scrubber. Monitoring the scrubber inlet temperature is important to prevent high inlet gas temperatures. Excessive (higher than design) inlet temperatures could lead to excessive liquid loss due to evaporation, resulting in damage to scrubber components. Many scrubbers are constructed of fiberglass reinforced plastics (FRP) or have corrosion-resistant liners which have maximum gas temperature limits ranging from 200 to 400oF. Emergency flush systems are included in many scrubbers to protect these components.

Measuring outlet temperatures is important for evaluating scrubber operation and protecting downstream equipment from excessive temperatures. Downstream from a wet scrubber, the gas should be saturated; saturation usually occurs at temperatures in the 150 to 160oF range. High outlet temperatures can indicate poor liquid distribution or plugging of the liquid inlet (i.e. reduced heat transfer between liquid and gas).

Temperatures are measured using thermocouples. The main consideration with using thermocouples is that they should be installed in a location that provides an accurate representation of the gas stream temperature being measured.

Liquid Flow Monitors

Liquid flow monitors are used to indicate that flow rates are maintained in the design operating ranges. Liquid flow monitors can be used on the scrubber inlet as well as the makeup and/or blowdown from the scrubber.

The type of instrumentation used to measure flow depends on the size of the scrubber and characteristics of the liquid being monitored. Clean liquid streams can be monitored using orifice or venturi meters, swinging vane meters or a rotameter. All of these devices are in direct contact with the liquid stream and therefore subject to wear and buildup when suspended solids are present. Ultrasonic and magnetic meters, being non-contact devices, are not subject to these problems. However, they are more expensive, do not handle shock as well and require additional maintenance to obtain reliable data.

pH Monitors

The pH of various liquid streams is often manually monitored to prevent corrosion and scaling problems. At low pH levels (below 5) corrosion of metals will become a problem and at high levels, calcium and magnesium compounds can precipitate out of solution and cause scaling problems. The important areas where pH is monitored are the chemical and scrubbing liquor feed streams and the recycle liquor systems.

Occasionally, pH monitors are used to control the flow of alkaline reagent to scrubbing systems. Generally, pH monitors require a substantial amount of maintenance to remain operational. Most successful applications of pH meters for continuous pH monitoring employ sidestream monitors where only a small sample of the water flow is monitored rather than the total flow through the scrubber.

Lesson 8 ___________________________________________________________________________________

8-20 2.0-7/98

Recordkeeping

A comprehensive, site-specific recordkeeping system of both the design and operating history will enable personnel to better evaluate scrubber performance. Design records indicate the specific conditions under which the scrubbing system was built to operate. A comparison between design records and operating records can indicate whether operating parameters have changed significantly from the design conditions. Secondly, maintaining proper operating records establishes a good baseline of information to bracket normal ranges of operation.

There are certain data common to most scrubber types which should be included in any scrubber recordkeeping system. These data elements are listed in Table 8-5 (EPA 1983). Comparing these routinely measured parameter values to the baseline values can provide a very good indication of the performance of a scrubber. In addition, examination of these parameters over time can aid in the detection of component deterioration in the scrubber system. To be effective, recordkeeping should be conducted on a daily basis, if not once per shift.

Table 8-5. Scrubber operation data

Inlet Gas Temperature

Outlet Gas Temperature

Total Static Pressure Drop

Static Pressure Drop of Mist Eliminator

Liquor Feed Rate

Liquor pH

Water Makeup Rate

Fan Current

Fan RPM

Fan Gas Inlet Temperature

Nozzle Pressure

Pump Discharge Pressure

Recycle Bleed Rate

Chemical Addition Rate

Liquid Solids Concentration Source: EPA 1983

It is recommended that whenever possible, the scrubber operation data be obtained using portable instruments (EPA 1983). Tap holes through which a measurement is made should be cleaned prior to every measurement to ensure that a partially or completely plugged hole does not result in an erroneous measurement. This, in fact, is one of the reasons that portable instruments should be used rather than fixed gauges. Often a reading from a fixed gauge will be recorded without checks to see that the gauge's tap hole is not plugged. Regardless of whether the instruments are fixed or portable, each must be calibrated at intervals which are at least as frequent as the manufacturer’s specifications.


2.0-7/98 8-21

In addition to the scrubber data listed in Table 8-5, process data should also be recorded. Variations in process feed rate, capacity of the system and type of material being processed can affect the operation and/or efficiency of the scrubbing system.

Summary

Many components comprise any scrubbing system. All of the individual components must be properly designed and operated or else the scrubbing system may not function.

Fans and ductwork must transport the flue gas from the process through the scrubbing system and exhaust it through the stack. Also, pumps and associated piping carry the scrubbing liquid to and from the system. These components should be designed to minimize friction losses (resistance to flow) and to reduce their susceptibility to abrasion and corrosion problems.

Quenching systems are used to cool and humidify hot gases prior to entering the scrubber vessel. Cooling the hot gases protects the construction materials of the scrubber vessel and also reduces the amount of evaporation that could potentially occur in the scrubber vessel.

For both the quench and scrubber vessels, various spray nozzle designs are utilized. Two important factors in spray nozzle operation are the type of spray pattern produced and the ability to handle solids in the liquid spray. The impingement nozzle, solid core nozzle, and helical spray nozzles are described in this lesson.

Finally, any liquid droplets that become entrained in the gas stream must be removed by an entrainment separator (mist eliminator) before exhausting gas to the atmosphere. The three designs discussed in this lesson are cyclonic separators, mesh-pad separators, and blade separators. Properly designed and operated entrainment separators can help increase pollutant removal efficiencies. These devices must be carefully monitored to prevent potential plugging which could result in excess emissions or cause the system to shutdown.

A comprehensive monitoring and recordkeeping program will enable personnel to readily assess the effectiveness of the scrubbing system in addition to highlighting any potential component failures. The following variables should be monitored in scrubbing systems:

• Inlet and out gas temperatures

• Liquid flow rates

• Pressure drop

• pH levels in chemical bed streams

• The current flowing through fans and pumps

The pressure drop across the scrubber vessel and entrainment separator will give an indication of any potential plugging problems or flow variations. Temperature measurements across the scrubber or quencher can reveal any liquid distribution problems which are indicated by excess gas temperature. Measurement of scrubber liquid pH is important to prevent scaling and/or corrosion problems and also to maintain effective gas pollutant removal.

Lesson 8 ___________________________________________________________________________________

8-22 2.0-7/98



2.0-7/98 8-23

Lesson 8 ___________________________________________________________________________________

8-24 2.0-7/98


1. What are the most popular types of centrifugal fans for wet scrubbing systems?

a. Radial and forward-curved b. Radial and backward-curved c. Vane-axial fans and airfoil d. Forward-curved and backward-curved

2. Fans located before the scrubber are referred to as ____________________ fans.

a. Positive-pressure b. Dirty-side c. Forced-draft d. All of the above

3. Fans located after the scrubber are always operated:

a. Wet b. Dry

4. To reduce pressure losses in ducts, the number of ____________________ should be kept to a minimum.

a. Bends b. Expansions c. Contractions d. All of the above

5. What is/are the primary maintenance problems(s) associated with fans?

a. Abrasion b. Solids buildup c. Corrosion d. All of the above

6. True or False? In general, electric-motor-driven centrifugal pumps are the most frequently used pumps in wet scrubbing systems.

7. What area(s) of the pump is/are most susceptible to abrasion or corrosion?

a. Impeller b. Housing c. Seals d. All of the above


2.0-7/98 8-25

8. What is a/(are) common problem(s) for pipes in most scrubbing systems?

a. Abrasion b. Corrosion c. Plugging d. All of the above

9. True or False? Cast iron and steel pipes are very resistant to attack by corrosive materials.

Part 2

10. As the liquid droplets produced by the quench spray become ____________________, the quencher becomes more efficient in cooling the exhaust gas stream.

a. Smaller b. Larger c. Rounder d. Heavier

11. Quenchers must be sized to provide an adequate ____________________ ____________________ for the exhaust gas, since evaporation does not occur instantaneously.

12. Quenching should be done with the ____________________ water available.

a. Dirtiest b. Cleanest c. Highest-pH d. Lowest-pH

13. List five important characteristics of spray nozzles used in wet scrubbing systems. ________________________________________ ________________________________________ ________________________________________ ________________________________________ ________________________________________

14. True or False? Nozzle plugging is one of the most common malfunctions in wet scrubbers.

15. List five remedies for plugged nozzles. _______________________________________________________ _______________________________________________________ _______________________________________________________ _______________________________________________________ _______________________________________________________

16. Entrainment separators are used to:

a. Prevent liquid carryover b. Recover scrubbing liquor c. Perform additional scrubbing d. All of the above

Lesson 8 ___________________________________________________________________________________

8-26 2.0-7/98

17. Cyclonic separators can remove liquid droplets as small as ____________________ in diameter.

a. 0.01 µm b. 0.1 µm c. 1.0 µm d. 10.0 µm

18. In general, wire-mesh pads should be ____________________ to prevent plugging.

a. Installed at a slant b. Sprayed from the bottom c. Sprayed from the top d. Sprayed from the top and bottom

19. True or False? Wire- or plastic-mesh pads are capable of removing smaller droplets than either cyclonic or blade separators; however, they are also more susceptible to plugging.

Part 3

20. Monitors are used in scrubbing systems to:

a. Obtain operating information to trouble shoot potential problems. b. Provide input signals for other devices. c. Provide a safety feature by sounding alarm when design limits are exceeded. d. All of the above

21. True or False? The best manner to monitor pressure differentials is across individual components in a scrubbing system as opposed to the whole system.

22. Temperature is monitored in a scrubbing system to:

a. Provide a safety feature by indicating potential high temperature. b. Prevent damage to scrubber components and equipment downstream from scrubber. c. Evaluate scrubber operation. d. All of the above

23. The primary problem(s) with measuring pressure differential is(are):

a. The use of complicated and sensitive devices b. Difficulty of obtaining accurate readings c. Plugging of the pressure tap lines d. All of the above

24. True or False? It is impossible to measure liquid flow without having a device in contact with the liquid stream being measured.

25. In a good recordkeeping system, information on which of the following should be kept?

a. Scrubber operating data b. Process operating data c. Design records d. a and b, only e. a, b, and c


2.0-7/98 8-27

26. True or False? In recording data from scrubbing system monitors/gauges, it is important to ensure that the gauge’s tap hole is not plugged.

Lesson 8 ___________________________________________________________________________________

8-28 2.0-7/98


1. b. Radial and backward-curved The most popular types of centrifugal fans for wet scrubbing systems are radial and backward-curved fans.

2. d. All of the above Fans located before the scrubber are referred to as positive-pressure, dirty-side or forced-draft fans.

3. a. Wet Fans located after the scrubber are always operated wet because the airstream is saturated with moisture.

4. d. All of the above To reduce pressure losses in ducts, the number of bends, expansions, and contractions should be kept to a minimum.

5. d. All of the above The primary maintenance problems associated with fans are:

• Abrasion

• Solids buildup

• Corrosion

6. True In general, electric-motor-driven centrifugal pumps are the most frequently used pumps in wet scrubbing systems.

7. d. All of the above The areas of pumps that are most susceptible to abrasion or corrosion are the following:

• Impeller

• Housing

• Seals

8. d. All of the above Common problems for pipes in most scrubbing systems are the following:

• Abrasion

• Corrosion

• Plugging

9. False Cast iron and steel pipes are NOT very resistant to attack by corrosive materials.


2.0-7/98 8-29

Part 2

10. a. Smaller As the liquid droplets produced by the quench spray become smaller, the quencher becomes more efficient in cooling the exhaust gas stream. Smaller liquid droplets increase the surface area of the liquid, thereby facilitating evaporation.

11. Residence time Quenchers must be sized to provide an adequate residence time for the exhaust gas, since evaporation does not occur instantaneously.

12. b. Cleanest Quenching should be done with the cleanest water available.

13. Opening size Droplet size Spray pattern Operating mechanism Power consumption Five important characteristics of spray nozzles used in wet scrubbing systems are:

• Opening size

• Droplet size

• Spray pattern

• Operating mechanism

• Power consumption

14. True Nozzle plugging is one of the most common malfunctions in wet scrubbers.

15. Replace nozzle with one having a more open design Clean nozzles frequently Filter the scrubbing liquor Increase bleed rate Increase makeup water rate Five remedies for plugged nozzles are:

• Replace nozzle with one having a more open design

• Clean nozzles frequently

• Filter the scrubbing liquor

• Increase bleed rate

• Increase makeup water rate

Lesson 8 ___________________________________________________________________________________

8-30 2.0-7/98

16. d. All of the above Entrainment separators do the following:

• Prevent liquid carryover

• Recover scrubbing liquor

• Perform additional scrubbing

17. d. 10.0 µm Cyclonic separators can remove liquid droplets as small as 10.0 µm in diameter.

18. d. Sprayed from the top and bottom In general, wire-mesh pads should be sprayed from the top and bottom to prevent plugging.

19. True Wire- or plastic-mesh pads are capable of removing smaller droplets than either cyclonic or blade separators; however, they are also more susceptible to plugging.

Part 3

20. d. All of the above Monitors are used in scrubbing systems to do the following:

• Obtain operating information to trouble shoot potential problems

• Provide input signals for other devices

• Provide a safety feature by sounding alarm when design limits are exceeded

21. True The best manner to monitor pressure differentials is across individual components in a scrubbing system as opposed to the whole system.

22. d. All of the above Temperature is monitored in a scrubbing system to:

• Provide a safety feature by indicating potential high temperature

• Prevent damage to scrubber components and equipment downstream from scrubber

• Evaluate scrubber operation

23. c. Plugging of the pressure tap lines The primary problem with measuring pressure differential is plugging of the pressure tap lines.

24. False Liquid flow can be measured without having the device in contact with the liquid stream. Ultrasonic and magnetic meters are non-contact devices.


2.0-7/98 8-31

25. e. a, b, and c In a good recordkeeping system, information should be kept on the following:

• Scrubber operating data

• Process operating data

• Design records

26. True In recording data from scrubbing system monitors/gauges, it is important to ensure that the gauge’s tap hole is not plugged.

Lesson 8 ___________________________________________________________________________________

8-32 2.0-7/98

Bibliography

Calvert, S., J. Goldschmid, D. Leith, and D. Mehta. 1972, August. Wet Scrubber System Study. Vol. 1, Scrubber Handbook. EPA-R2-72-118a. U.S. Environmental Protection Agency.

Calvert, S., I. L. Jadmani, S. Young, and S. Stahlberg. 1974, October. Entrainment Separators for Scrubbers - Initial Report. EPA 650/2-74-119a. U.S. Environmental Protection Agency.

Czuchra, P. A. 1979, April. Operation and maintenance of a particulate scrubber system’s ancillary components. Paper presented at the U. S. Environmental Research Information Seminar. Atlanta, GA.

Gleason, T. G. 1977. How to avoid scrubber corrosion. In P. N. Cheremisinoff and R. A. Young (Eds.), Air Pollution Control and Design Handbook. New York: Marcel Dekker.

Industrial Gas Cleaning Institute. 1975. Scrubber System Major Auxiliaries. Publication WS-4. Stamford, CT.

Kalika, P. W. 1969. How water recirculation and steam plumes influence scrubber design. Chemical Engineering. 79:133-138.

Kashdan, E. R., and M. B. Ranada. 1979. Design Guidelines for an Optimum Scrubber System. EPA 600/7-79-018. U.S. Environmental Protection Agency.

MacDonald, J. W. 1982. Absorbers. In L. Theodore and A. J. Buonicore (Eds.), Air Pollution Control Equipment, Design, Selection, Operation, and Maintenance. Englewood Cliffs: Prentice-Hall.

National Asphalt Pavement Association. 1978. The Maintenance and Operation of Exhaust Systems in the Hot Mix Batch Plant. 2nd ed. Information Series 52.



Schifftner, K. C. 1979, April. Venturi scrubber operation and maintenance. Paper presented at the U.S. EPA Environmental Research Information Center. Atlanta, GA.


Wechselblatt, P. M. 1975. Wet scrubbers (particulates). In F. L. Cross and H. E. Hesketh (Eds.), Handbook for the Operation and Maintenance of Air Pollution Control Equipment. Westport: Technomic Publishing.


2.0-7/98 8-33

2.0-7/98 9-1

Lesson 9 Flue Gas Desulfurization (Acid Gas Removal) Systems

Goal To familiarize you with the operation of flue gas desulfurization (FGD) systems that use a scrubbing liquid to absorb SO2 present in the exhaust gas stream.


1. Describe how the following six operating variables affect wet scrubber operation in FGD systems:

• Liquid-to-gas ratio

• pH

• Gas velocity/residence time

• Gas distribution system

• Scrubber design

• Turndown ability

2. Briefly describe four FGD wet scrubbing processes

3. Identify operating problems associated with each FGD process above

4. Identify some of the various scrubber designs and typical operating conditions associated with FGD processes

Introduction

The previous lessons describe various scrubber designs that control emissions of gaseous and particulate pollutants. This lesson discusses a major application for scrubbers in air pollution control: flue gas desulfurization (FGD), which is one of the largest markets for scrubbing systems (in terms of money spent). The term flue gas desulfurization has traditionally referred to wet scrubbers that remove sulfur dioxide (SO2) emissions from large electric utility boilers (mainly coal combustion). However, because of the requirement to control acid

Lesson 9 ___________________________________________________________________________________

9-2 2.0-7/98

emissions from industrial boilers and incinerators and the evolution of different types of acid control systems, the terms FGD, acid gas or acid rain control are used interchangeably to categorize a wide variety of control system designs. FGD systems are also used to reduce SO2 emissions from process plants such as smelters, acid plants, refineries, and pulp and paper mills.

FGD systems can be categorized as dry or wet. In Lesson 7, you learned about dry scrubbing systems that control SO2 and other acid gases from utility and industrial boilers and incinerators. This lesson focuses on the traditional, wet FGD systems that have been installed on operating plants. This lesson will also briefly cover some of the emerging technologies (both wet or dry) that are being developed for FGD (acid rain) control.

In wet FGD scrubbing systems, the scrubbing liquid contains an alkali reagent to enhance the absorption of SO2 and other acid gases. More than a dozen different reagents have been used, with lime and limestone being the most popular. Sodium-based solutions (sometimes referred to as clear solutions) provide better SO2 solubility and less scaling problems than lime or limestone. However, sodium reagents are much more expensive.

Wet FGD scrubbers can further be classified as nonregenerable or regenerable. Nonregenerable processes, also called throwaway processes, produce a sludge waste that must be disposed of properly. It should be noted that in throwaway or nonregenerable processes the scrubbing liquid can still be recycled or regenerated; however, no useful product is obtained from the eventual sludge. Regenerable processes produce a product from the sludge that may be sold to partially offset the cost of operating the FGD system. Regenerated products include elemental sulfur, sulfuric acid and gypsum. Based on the recent capacities listed in Table 9-1, approximately 91% of FGD processes are nonregenerable, or throwaway. The throwaway processes are simpler and presently more economical than those that recover and sell products. Also, Table 9-1 shows that approximately 78% of the FGD systems represented are wet systems using lime or limestone as a reagent.

Flue Gas Desulfurization (Acid Gas Removal) Systems ___________________________________________________________________________________

2.0-7/98 9-3

Table 9-1. Summary of FGD systems by process (percentage of total Megawatts)

Process

By-product

Percent of total MW

(as of 12/89)

Throwaway product

Wet scrubbing

Dual alkali 3.4% Lime 16.3 Lime/alkaline fly ash 7.0 Limestone 48.2 Limestone/alkaline fly ash 2.4 Sodium carbonate 4.0

Spray drying

Lime 8.8 Sodium carbonate 0 Reagent type not selected 0.7

Dry injection

Lime 0.2 Sodium carbonate 0 Reagent type not selected 0

Process not selected 0

Saleable product

Wet scrubbing

Lime Metals/fly ash/other < 0.1 Limestone Gypsum 4.1 Magnesium oxide Sulfuric acid 1.4 Wellman Lord Sulfuric acid 3.1

Spray drying

Lime Dry scrubber waste 0

Process undecided 0

Total 100.0 Source: Hance 1991.

Most FGD systems employ two stages: one for fly ash removal and the other for SO2 removal. Attempts have been made to remove both the fly ash and SO2 in one scrubbing vessel. However, these systems experienced severe maintenance problems and low simultaneous removal efficiencies. In wet scrubbing systems the flue gas normally passes

Lesson 9 ___________________________________________________________________________________

9-4 2.0-7/98

first through a fly ash removal device, either an electrostatic precipitator or a wet scrubber, and then into the SO2 absorber. However, in dry injection or spray drying operations, the SO2 is first reacted with the sorbent and then the flue gas passes through a particulate control device.

Many different types of absorbers have been used in wet FGD systems, including spray towers, venturis, plate towers, and mobile packed beds. Because of scale buildup, plugging, or erosion, which affect FGD dependability and absorber efficiency, the trend is to use simple scrubbers such as spray towers instead of more complicated ones. The configuration of the tower may be vertical or horizontal, and flue gas can flow cocurrently, countercurrently, or crosscurrently with respect to the liquid. The chief drawback of spray towers is that they require a higher liquid-to-gas ratio requirement for equivalent SO2 removal than other absorber designs (Makansi 1982).

Numerous operating variables affect the SO2 removal rate of the absorber. Most of these variables were discussed in previous lessons; however, some are unique to FGD absorbers. The following list contains some of the important parameters affecting the operation of an FGD scrubber (Ponder et al. 1979 and Leivo 1978):

Liquid-to-gas ratio - The ratio of scrubber liquid slurry to gas flow (L/G ratio). For a given set of system variables, a minimum L/G ratio is required to achieve the desired SO2 absorption, based on the solubility of SO2 in the liquid. High L/G ratios require more piping and structural design considerations, resulting in higher costs.

pH - Depending on the particular type of FGD system, pH must be kept within a certain range to ensure high solubility of SO2 and to prevent scale buildup.

Gas velocity - To minimize equipment cost, scrubbers are designed to operate at maximum practicable gas velocities, thereby minimizing vessel size. Maximum velocities are dictated by gas-liquid distribution characteristics and by the maximum allowable liquid entrainment that the mist eliminator can handle. Gas velocities may be 1.5 to 10 m/s (5 to 30 ft/sec) in tower scrubbers and more than 30 m/s (100 ft/sec) in the throat of a venturi scrubber. A common range of the gas velocity for FGD absorbers is 2.0 to 3.0 m/s (7 to 10 ft/sec). The lower the velocity is, the less the entrainment, but the more costly the scrubber will be.

Residence Time - For FGD processes using an alkali slurry for scrubbing, the system should be designed to provide adequate residence time in the absorber vessel for the SO2 to be absorbed by the alkali slurry. The main objective is to make sure that the maximum amount of alkali is utilized in the scrubber. Residence times in packed towers may be as long as 5 seconds. Residence times in venturi scrubbers are a few hundredths of a second, usually too short for high absorption efficiency of SO2 in systems using lime or limestone scrubbing slurries, unless additives or two scrubbing stages are used.

Gas distribution - Maintaining a uniform gas flow is a major problem that occurs in commercial FGD scrubbers. If the flow is not uniform, the scrubber will not operate at design efficiencies. In practice, uniform flow has been difficult to achieve. Typically, turning vanes near the scrubber inlet duct and compartmentalization have been used.

Scrubber designs - To promote maximum gas-liquid surface area and contact time, a number of scrubber designs have been used. Common ones are mobile-bed scrubbers, venturi-rod scrubbers, plate towers, packed towers, and spray towers. Countercurrent packed towers are


2.0-7/98 9-5

infrequently used because they have a tendency to become plugged by collected particles or to scale when lime or limestone scrubbing slurries are used.

Turndown - The ability to operate at less than full load and to adjust to changes in boiler load. The scrubber must provide good gas-liquid distribution, sufficient residence time, and high gas-liquid interfacial area for varying gas flow rates. Some scrubbers can be turned down to 50% of design, while others must be divided into sections that can be closed off. A variable-throat venturi can be used to accommodate turndown. In a large FGD installation, individual modules can be taken out of service.

It is important to note that the above list does not imply that these are the only parameters affecting SO2 absorption efficiency. Each FGD process has a unique set of operating criteria.

In addition to the set of factors just given, the coal properties greatly affect FGD system design for boiler operations. The major coal properties affecting FGD system design and operation are (Leivo 1978):

Heating value of coal - Affects flue gas flow rate. Flow rate is generally higher for lower heating value coals, which also contribute a greater water-vapor content to the flue gas.

Moisture content - Affects the heating value (i.e. since the higher the moisture content the lower the heating value) and contributes directly to the moisture content and volume of the flue gas.

Sulfur content - The sulfur content, together with the allowable emission standards, determines the required SO2 removal efficiency, the FGD system complexity and cost, and also affects sulfite oxidation.

Ash content - May affect FGD system chemistry and increase erosion. In some cases, it may be desirable to remove fly ash upstream from the FGD system.

Chlorine content - May require high-alloy metals or linings to combat corrosion for some process equipment and could affect process chemistry or require prescrubbing.

Another important design consideration associated with wet FGD systems is that the flue gas exiting the absorber is saturated with water and still contains some SO2. (No system is 100% efficient.) Therefore, these gases are highly corrosive to any downstream equipment - i.e., fans, ducts, and stacks. Two methods that minimize corrosion are: (1) reheating the gases to above their dew point and (2) choosing construction materials and design conditions that allow equipment to withstand the corrosive conditions. The selection of a reheating method or the decision not to reheat (thereby requiring the use of special construction materials) are very controversial topics connected with FGD design (Makansi 1982). Both alternatives are expensive and must be considered on a by-site basis.

Four methods used to reheat stack gases:

1. Indirect in-line reheating - The flue gas passes through a heat exchanger that uses steam or hot water.

2. Indirect-direct reheating - Steam is used to heat air (outside the duct) and then the hot air is mixed with the scrubbed gases.

Lesson 9 ___________________________________________________________________________________

9-6 2.0-7/98

3. Direct combustion reheating - Oil or gas is burned either in the duct or in an external chamber, and the resulting hot gases are mixed with the scrubbed gases.

4. Bypass reheating - A portion of the untreated hot flue gas bypasses the scrubber and is mixed with the scrubbed gases.

None of the above methods has a clear advantage over the others (Makansi 1982). Systems using indirect in-line reheating have experienced severe corrosion and plugging problems. Indirect-direct and direct combustion reheating are expensive because of added fuel costs and bypass reheating is limited in the degree of reheating obtainable (due to SO2 emissions in the bypass). Because of the expense and problems associated with reheat, newer FGD designs are utilizing more plastics (fiberglass reinforced plastic) and exotic alloys instead of reheat.

This lesson will discuss four of the more popular FGD systems that are nonregenerable, calcium- and/or sodium-based systems. The process chemistry, system description, and operating experience involved in each will be presented.


Nonregenerable FGD Processes

Nonregenerable FGD processes generate a sludge or waste product. The sludge must be disposed of properly in a pond or landfill. The three most common nonregenerable processes used on utility boilers in the U. S. are lime, limestone, and double-alkali. Although the double-alkali process regenerates the scrubbing reagent, it is classified as throwaway since it does not produce a saleable product and generates solids that must be disposed of in a landfill. The fourth nonregenerable process discussed here, sodium-based throwaway systems (NaOH and Na2CO3), are utilized mostly on industrial boilers.

Lime Scrubbing

Process Chemistry Lime scrubbing uses an alkaline slurry made by adding lime (CaO), usually 90% pure, to water. The alkaline slurry is sprayed in the absorber and reacts with the SO2 in the flue gas. Insoluble calcium sulfite (CaSO3) and calcium sulfate (CaSO4) salts are formed in the chemical reaction that occurs in the scrubber and are removed as sludge.

A number of reactions take place in the absorber. Before the calcium can react with the SO2, both must be broken down into their respective ions. This is accomplished by slaking (dissolving) the lime in water and then spraying the slurry into the flue gas to dissolve the SO2. Simplified reactions occur simultaneously and are illustrated below.


2.0-7/98 9-7

SO2 dissociation: SO2 (gaseous) → SO2 (aqueous)

SO2 + H2O → H2SO3

H SO2 3 → → H + HSO 2H + SO+3- +

3=

Lime (CaO) dissolution: CaO(solid) + H2O → Ca(OH)2 (aqueous)

Ca(OH2) → Ca++ + 2OH-

Now that SO2 and lime are broken into their ions ( SO3= and Ca++), the following

reaction occurs:

Ca O++ → + SO + 2H + 2OH CaSO + 2H3= + -

3 (solid) 2

In addition, the following reactions can also occur when there is excess oxygen:

SO

SO3

4

=

=

→

→

+ 1 / 2 O SO

+ Ca CaSO2 4

=

++4 (solid)

From the above relationships and assuming that the lime is 90% pure, it will take 1.1 moles of lime to remove 1 mole of SO2 gas.

System Description The equipment necessary for SO2 emission reduction comes under four operations:

1. Scrubbing or absorption - Accomplished with scrubbers, holding tanks, liquid-spray nozzles, and circulation pumps.

2. Lime handling and slurry preparation - Accomplished with lime unloading and storage equipment, lime processing and slurry preparation equipment.

3. Sludge processing - Accomplished with sludge clarifiers for dewatering, sludge pumps and handling equipment, and sludge solidifying equipment.

4. Flue-gas handling - Accomplished with inlet and outlet ductwork, dampers, fans, and stack gas reheaters.

Figure 9-1 is a schematic of a typical lime FGD system. Individual FGD systems vary considerably, depending on the FGD vendor and the plant layout. ESPs or scrubbers can be used for particle removal, followed by one of various absorber designs that are effective for SO2 removal. In general, as shown in Figure 9-1, flue gas from the boiler first passes through a particulate emission removal device then into the absorber where the SO2 is removed. The gas then passes through the entrainment separator to a reheater and is finally exhausted out of the stack.

Lesson 9 ___________________________________________________________________________________

9-8 2.0-7/98

Figure 9-1. Typical process flow for a lime or limestone FGD system

A slurry of spent scrubbing liquid and sludge from the absorber then goes to a recirculation tank. From this tank, a fixed amount of the slurry is bled off to process the sludge, and, at the same time, an equal amount of fresh lime is added to the recirculation tank. Sludge is sent to a clarifier, where a large portion of water is removed from the sludge, and sent to a holding tank. Makeup water is added to the process-water holding tank, and this liquid is returned to the recirculation tank. The partially dewatered sludge from the clarifier is sent to a vacuum filter, where most of the water is removed (and sent to the process-water holding tank) and the sludge is sent to a settling pond. Table 9-2 lists operational data of lime FGD systems, showing the various absorbers used.


2.0-7/98 9-9

Lesson 9 ___________________________________________________________________________________

9-10 2.0-7/98

Table 9-2. (continued) Operational data for lime FGD systems on utility boilers

Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of modules

per

L/G ratio

Pres

plant name (gross) control coal boiler L/m3 gal/1000 ft3

kPa

Kansas City Power & Light

Hawthorn #3 90 Combustion Engineering

- 0.6 Mobile bed (marbles) 2 3.5 26.0 2.7


- 0.6 Mobile bed (marbles) 2 3.5 26.0 2.7

Monongahela Power

Pleasants #1 618 B&W ESP 3.7 Sieve tray 4 7.4 55.0 1.2

Pleasants #2 618 B&W ESP 4.5 Sieve tray 4 7.4 55.0 -

Utah Power & Light

Hunter #1 400 Chemico ESP 0.6 Countercurrent spray 4 5.7 43.0 0.6

Hunter #2 400 Chemico ESP 0.6 Countercurrent spray 4 5.7 43.0 0.6

Huntingdon #1 430 Chemico ESP 0.6 Countercurrent spray 4 5.7 43.0 0.6Note: A dash (-) indicates that no data are available.


2.0-7/98 9-11

Operating Experience Early lime FGD systems were plagued with many operational and maintenance problems. Scale buildup and plugging of absorber internals and associated equipment were prominent problems. However, scaling and plugging in lime FGD systems were not as severe as with other calcium-based FGD systems (EPA 1981). Scale buildup (CaSO4) on spray nozzles and entrainment separators was particularly troublesome. New spray nozzle designs and careful control of the recirculating slurry have reduced internal scrubber scaling (EPA 1975). Problems with the entrainment separators have also been reduced by careful separator design, installing adequate wash sprays, and monitoring the pressure drop across them. Additional techniques that reduce scale buildup are (Leivo 1978):

Control of pH - If a lime FGD system is operated above a pH of 8.0 to 9.0, there is a risk of sulfite scaling. Automatic control of the feed by on-line pH sensors has been successful.

Holding tank residence time - By providing retention time in the scrubber recirculation tank, the supersaturation of the liquor can be decreased before recycling to the scrubber. Typical residence times of 5 to 15 minutes have been used in some full-scale systems.

Control of suspended solids concentration - The degree of supersaturation can be minimized by keeping an adequate supply of seed crystals in the scrubber slurry. Typical levels in newer installations range from 5 to 15% suspended solids. Solids are generally controlled by regulating the slurry bleed rate.

Liquid-to-gas ratio - High liquid-to-gas ratios can reduce scaling problems because the absorber outlet slurry is more dilute, containing less calcium sulfates and calcium sulfites that cause scaling.

Another problem that has occurred concerns stack gas reheaters. Stack gas is reheated to avoid condensation on and corrosion of the ductwork and stack, and to enhance plume rise and pollutant dispersion. Reheating is accomplished by using steam coils in the stack, by using hot air supplied by auxiliary oil heaters in the stack, or by other methods previously mentioned. Some reheater failures were caused by acid attack to reheater components. Other reheaters vibrated too much, causing structural deterioration.

Corrosion of scrubber internals, fans and ductwork, and stack linings have been reduced by using special materials such as rubber- or plastic-coated steel and by carefully controlling slurry pH with monitors. Additional operation and maintenance problems and solutions are found in Lime FGD Systems Data Book, Second Edition (EPRI 1983).


Lesson 9 ___________________________________________________________________________________

9-12 2.0-7/98

Limestone Scrubbing

Process Chemistry Limestone scrubbers are very similar to lime scrubbers. The use of limestone (CaCO3) instead of lime requires different feed preparation equipment and higher liquid-to-gas ratios (since limestone is less reactive than lime). Even with these differences, the processes are so similar that an FGD system can be set up to use either lime or limestone in the scrubbing liquid (See Figure 9-1).

The basic chemical reactions occurring in the limestone process are very similar to those in the lime-scrubbing process. The only difference is in the dissolution reaction that generates the calcium ion. When limestone is mixed with water, the following reaction occurs:

CaCO O 3 (solid) 2++

3- - + H Ca + HCO + OH→

The other reactions are the same as those for lime scrubbing.

System Description The equipment necessary for SO2 absorption is the same as that for lime scrubbing, except in the slurry preparation. The limestone feed (rock) is reduced in size by crushing it in a ball mill. Limestone is sent to a size classifier. Pieces larger than 200 mesh are sent back to the ball mill for recrushing. Limestone is mixed with water in a slurry supply tank. Limestone is generally 2 to 4 times cheaper than lime, making it more popular for large FGD systems. Table 9-3 lists operations data for limestone FGD systems. Note the similarities in equipment and operating conditions to those of lime FGD systems.


2.0-7/98 9-13

See Table 9-3. Operational data for limestone FGD systems on utility boilers

Table 9-3. Operational data for limestone FGD systems on utility boilers

Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of modules

L/G ratio

Pres

plant name (gross) control coal per boiler

L/m3 gal/1000 ft3 kPa

Alabama Electric

Tombigbee #2 255 Peabody ESP 1.2 Countercurrent spray 2 9.4 70.0 1.0

Tombigbee #3 255 Peabody ESP 1.2 Countercurrent spray 2 9.4 70.0 1.0

Arizona Electric Power

Apache #2 195 Research-Cottrell ESP 0.5 Spray/packed bed 2 2.8 20.6 1.5

Apache #3 195 Research-Cottrell ESP 0.5 Spray/packed bed 2 2.8 20.6 1.5

Cholla #1 119 Research-Cottrell Cyclone/venturi 0.5 Spray/packed bed 1 6.5 48.9 0.1

Cholla #2 264 Research-Cottrell Cyclone/venturi 0.5 Spray/packed bed 4 6.5 48.9 0.1

Basin Electric Power

Laramie River #1 570 Research-Cottrell ESP 0.8 Spray/packed bed 5 8.0 60.0 -

Laramie River #2 570 Research-Cottrell ESP 0.8 Spray/packed bed 5 8.0 60.0 -

Central Illinois Light

Duck Creek #1 416 Environeering ESP 3.7 Rod deck packed tower

4 6.7 50.0 2.0

Colorado Ute Electrical

Craig #1 447 Peabody ESP 0.4 Countercurrent spray 4 6.7 50.0 1.6

Craig #2 455 Peabody ESP 0.4 Countercurrent spray 4 6.7 50.0 1.6

Commonwealth Edison

Powerton 450 Air Correction Division - UOP

ESP 3.5 Mobile bed (TCA) 3 8.0 60.0 3.0

Indianapolis Power & Light

Petersburg #3 532 Air Correction Division - UOP


Lesson 9 ___________________________________________________________________________________

9-14 2.0-7/98

Table 9-3. (continued) Operational data for limestone FGD systems on utility boilers

Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of modules

L/G ratio

Pres




La Cygne 820 B&W Variable venturi 5.4 Sieve tray 8 5.0 37.7 1.5

Jeffery #1 720 Combustion Engineering

ESP 0.3 Countercurrent spray 6 4.1 30.4 1.0


ESP 0.3 Countercurrent spray - 4.1 30.4 1.0

Lawrence #4 125 Combustion Engineering

Rod venturi 0.6 Countercurrent spray 2 4.0 30.0 0.6


Rod venturi 0.6 Countercurrent spray 2 2.5 19.0 0.6

Salt River Project

Coronado #1 350 Pullman Kellogg ESP 1.0 Weir crosscurrent spray

2 - - 0.4


2 - - 0.4

South Carolina Public Service

Winyah #2 280 B&W ESP 1.7 Venturi/sieve tray 2 6.3 47.5 1.1

Winyah #3 280 B&W ESP 1.7 Countercurrent spray 2 - - -

South Mississippi Electric

R. D. Morrow #1 200 Environeering ESP 1.3 Rod deck packed tower

1 6.6 49.0 2.0


1 6.6 49.0 2.0

Southern Illinois

Marion #4 173 B&W ESP 3.8 Countercurrent spray 2 9.9 74.0 1.5


2.0-7/98 9-15


Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of modules

L/G ratio

Pres



Springfield City

Southwest #1 194 Air Correction Division - UOP


Springfield Water, Light & Power

Dallman #3 205 Research-Cottrell Cyclone/ESP 3.3 Spray/packed tower 2 - - 0.2

TVA

Widows Creek #8 550 TVA ESP/venturi 3.7 Mobile packed bed and grid packing

1 3

8.0 60.0 0.5

Texas Power & Light

Sandow #4 545 Combustion Engineering

ESP 1.6 Countercurrent spray 3 - - -

Texas Utilities

Martin Lake #1 793 Research-Cottrell ESP 0.9 Spray/packed bed 6 - - 1.1



Monticello 800 Chemico ESP 1.5 Countercurrent spray 3 9.4 70.0 1.2Note: A dash (-) indicates that no data are available.

Lesson 9 ___________________________________________________________________________________

9-16 2.0-7/98

Operating Experience Early limestone FGD systems had scrubber operating problems similar to those of lime scrubbing systems. Plugged and clogged nozzles, scrubber internals, and mist eliminators (entrainment separators) resulted from inefficient SO2 absorption by limestone in the scrubber.

Increased absorption efficiency is achievable at high pH values since more alkali is available to dissolve the SO2 gas. However, scale buildup will occur if the scrubber is operated at very high pH values. The pH levels can be maintained by carefully controlling limestone and water feed rates. Low pH reduces removal efficiency; high pH causes scale buildup on scrubber internals.

As you can see from Tables 9-2 and 9-3, the SO2 removal efficiencies for various lime and limestone FGD installations range from 52% to 97%. These FGD systems were designed to meet existing air pollution regulations. Lime and limestone FGD systems are capable of removing SO2 with efficiencies in excess of 90%. The addition of small amounts of reagents (such as soluble magnesium) to the scrubber liquor can greatly increase SO2 removal efficiencies to as high as 99% (Devitt et al. 1978).

Another scrubber operating problem occurring in lime and limestone FGD systems is that calcium sulfite in the sludge settles and filters poorly. It can be removed from the scrubber slurry only in a semi-liquid or paste-like form. A process improvement called forced oxidation was developed by an EPA research laboratory to address this problem. In forced oxidation, air is blown into a designated section of the absorber module or into a separate reaction (oxidation) tank. The air oxidizes the calcium sulfite to calcium sulfate in the following reaction:

CaSO3 + H2O + 1/2 O2 → CaSO4 + H2O

Calcium sulfate formed by this reaction grows to a larger crystal size than calcium sulfite. As a result, calcium sulfate is easily filtered, forming a drier and more stable material that can be disposed of in a landfill or has the potential to be sold as a product to make cement, gypsum wallboard, or as a fertilizer additive.

Forced oxidation also helps control scale buildup problems on scrubber internals by removing the calcium sulfite from the slurry in the form of calcium sulfate, which is more easily filtered. This prevents calcium sulfites from oxidizing and precipitating out in the scrubber internal areas. Another method to prevent oxidation of calcium sulfite to calcium sulfate is by use of chemical inhibitors. Sulfur, magnesium and dibasic acid have all been tested and proven effective in inhibiting oxidation and thus reducing scaling in lime and limestone FGD systems.



2.0-7/98 9-17

Dual-Alkali Scrubbing

Dual- or double-alkali scrubbing is a third throwaway FGD process that uses a sodium-based alkali solution to remove SO2 from combustion exhaust gas. The sodium alkali solution absorbs SO2, and the spent absorbing liquor is regenerated with lime or limestone. Using both sodium- and calcium-based compounds is where the name dual or double-alkali comes from. Calcium sulfites and sulfates are precipitated and discarded as sludge. The regenerated sodium scrubbing solution is returned to the absorber loop. The dual-alkali process has reduced plugging and scaling problems in the absorber because sodium scrubbing compounds are very soluble. Dual-alkali systems are capable of 95% SO2 reduction.

Particulate matter is removed prior to SO2 scrubbing by an electrostatic precipitator or a venturi scrubber. This prevents the following: (1) fly ash erosion of the absorber internals and (2) any appreciable oxidation of the sodium solution in the absorber due to catalytic elements in the fly ash (EPA 1978).

Process Chemistry The sodium alkali solution is usually a mixture of the following compounds:

1. Sodium hydroxide (NaOH), also called caustic

2. Sodium carbonate (Na2CO3), also called soda ash

3. Sodium sulfite (Na2SO3)

The SO2 reacts with the alkaline components to primarily form sodium sulfite and sodium bisulfite (NaHSO3). The following are the main absorption reactions (EPA 1981):

2 NaOH + SO2 → Na2SO3 + H2O

NaOH + SO2 → NaHSO3

Na2CO3 + SO2 + H2O → 2NaHSO3

Na2CO3 + SO2 → Na2SO3 + CO2

Na2SO3 + SO2 + H2O → 2NaHSO3

In addition to the above reactions, some of the SO3 present may react with alkaline components to produce sodium sulfate. For example,

2NaOH + SO3 → Na2SO4 + H2O

Lesson 9 ___________________________________________________________________________________

9-18 2.0-7/98

Throughout the system, some sodium sulfite is oxidized to sulfate by:

2Na2SO3 + O2 → 2Na2SO4

After reaction in the absorber, spent scrubbing liquor is bled to a reactor tank for regeneration. Sodium bisulfite and sodium sulfate are inactive salts and do not absorb any SO2. Actually, it is the hydroxide ion ( OH− ), sulfite ion ( SO3

= ), and carbonate ion ( CO3

= ) that absorb SO2 gas. Sodium bisulfite and sodium sulfate are reacted with lime or limestone to produce a calcium sludge and a regenerated sodium solution.

2 3 3NaHSO SO O O + Ca(OH) Na + CaSO 1 / 2 H + 3 / 2 H (lime) (sludge)

2 2 3 2 2→ • ↓

Na + Ca(OH) + 1 / 2 H 2NaOH + CaSO 1 / 2 H (lime) (sludge)

2 2 2 3SO O O3 2→ • ↓

Na + Ca(OH) 2NaOH + CaSO (lime) (sludge)

2 2 4SO4 → ↓

At the present time, lime regeneration is the only process that has been used on commercial dual-alkali installations.

System Description The dual-alkali process uses two loops - absorption and regeneration. In the absorption loop, the sodium solution contacts the flue gas in the absorber to remove SO2. As shown in Figure 9-2, the scrubbing liquor from the bottom of the absorber is mixed with regenerated solution and sprayed in at the top of the absorber. A bleed stream from the recirculating liquid is sent to the reactor tank in the regeneration loop. The bleed stream is mixed with a lime slurry in a reactor tank, where insoluble calcium salts are formed and the absorbent is regenerated. The sludge from the reactor is then sent to a clarifier, or thickener, where the calcium sludge is drawn off the bottom, filtered, and washed with water. From the filter, the sodium solution is recycled to the clarifier, and the sludge is discarded. From the clarifier, the regenerated sodium solution is sent to a mixing tank where the sodium compounds and makeup water are added.


2.0-7/98 9-19

Figure 9-2. Typical process flow for a double-alkali FGD system

Some sodium sulfate solution is unreacted in the regeneration step. Additional sodium to makeup for that lost in the sludge is added to the regenerated solution in the form of soda ash or caustic soda. This regenerated absorbent is now ready to be used again.

Operating Experience The dual-alkali process has been installed and operating on both utility and industrial boilers for a number of years. Corrosion of, erosion of, and scale buildup on system equipment have not been major operating problems at dual-alkali FGD installations in the U.S. (EPA 1981). Operating data for the dual-alkali systems are presented in Table 9-4. Note the much lower L/G ratios of these systems compared to those of lime and limestone systems. The sodium solution is more efficient than both the lime and limestone slurries in absorbing SO2.

Lesson 9 ___________________________________________________________________________________

9-20 2.0-7/98

See Table 9-4. Operational data for double-alkali FGD systems on utility and industrial boilers

Table 9-4. Operational data for double-alkali FGD systems on utility and industrial boilers

Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of

modules

L/G ratio

Pres



Central Illinois Public Service

Newton #1 617 Buell ESP 2.5 Mobile bed 4 1.3 10.0 1.5

Louisville Gas & Electric

Cane Run #6 299 Combustion Equipment Association

ESP 4.8 Sieve plates 2 1.3 10.0 2.5

Southern Indiana Gas & Electric

A. B. Brown #1 265 FMC ESP 3.6 Variable-throat venturi

2 1.3 10.0 2.5

Caterpillar Tractor

East Peoria, IL 105 FMC Cyclone 3.2 Venturi 4 2.2 16.0 -

Joliet, IL 34 Zurn Cyclone 3.2 Dustraxtor 2 - - -

Morton, IL 19 Zurn Cyclone 3.2 Dustraxtor 2 - - -

Mossville, IL 70 FMC Cyclone - Venturi 4 1.2 8.6 -

Firestone Tire

Pottstown, PA 4 FMC Cyclone 3.0 Venturi 1 1.3 10.0 -

General Motors

Parma, OH 64 GM Environmental Cyclone - Bubble-cap plates 4 2.6 20.0 0.9Note: A dash (-) indicates that no data are available.


2.0-7/98 9-21

Some operating problems include regenerating scrubbing liquor and controlling the solids content of the sludge. Sodium sulfate, one of the compounds in the spent scrubbing liquor, is difficult to regenerate because it does not react efficiently with hydrated lime in the presence of sodium sulfite (Leivo 1978). Process conditions must be carefully controlled to adjust for the amounts of sodium sulfate and sodium sulfite that are formed in the spent scrubbing liquid. Another problem occurring in dual-alkali systems is that the solids content of the sludge can vary greatly, causing problems in handling and stabilizing the sludge for final disposal (Makansi 1982).


Sodium-Based Once-Through Scrubbing

Sodium-based once-through (throwaway) scrubbing systems are installed on a number of industrial boilers. These systems use a clear liquid absorbent of either sodium carbonate, sodium hydroxide, or sodium bicarbonate. According to Makansi (1982), sodium-based systems are favored for treating flue gas from industrial boilers for the following reasons:

• Sodium alkali is the most efficient of the commercial reagents in removing SO2, and the chemistry is relatively simple.

• They are soluble systems as opposed to slurry systems making for scale-free operation and fewer components.

• Such systems can handle the wider variations in flue-gas composition resulting from the burning of many different fuels by industry.

• The systems are often smaller, and operating costs are a small percentage of total plant costs.

• In some cases, these plants have a waste caustic stream or soda ash available for use as the absorbent.

These systems have been applied to only a few large utility boilers for these reasons:

• The process consumes a premium chemical (NaOH or Na2CO3) that is much more costly per pound than calcium-based reagents.

• The liquid wastes contain highly soluble sodium salt compounds. Therefore, the huge quantities of liquid wastes generated by large utilities would have to be sent to ponds to allow the water to evaporate.

Process Chemistry The process chemistry is very similar to that of the dual-alkali process, except the absorbent is not regenerated.

Lesson 9 ___________________________________________________________________________________

9-22 2.0-7/98

System Description A basic sodium-based throwaway FGD system is illustrated in Figure 9-3. Exhaust gas from the boiler may first pass through an ESP or baghouse to remove particulate matter. Sodium chemicals are mixed with water and sprayed into the absorber. The solution reacts with the SO2 in the flue gas to form sodium sulfite, sodium bisulfite, and a very small amount of sodium sulfate. A bleed stream is taken from the scrubbing liquor recirculation stream at a rate equal to the amount of SO2 that is being absorbed. The bleed stream is sent to a neutralization tank and aeration tower before being sent to a lined disposal pond.

Figure 9-3. Typical process flow for a sodium-based throwaway (single-alkali) FGD system

Some coal-fired units use ESPs or baghouses to remove fly ash before the gas enters the scrubber. In these cases, the absorber can be a plate tower or spray tower that provides good scrubbing efficiency at low pressure drops. For simultaneous SO2 and fly ash removal, venturi scrubbers can be used. In fact, many of the industrial


2.0-7/98 9-23

sodium-based throwaway systems are venturi scrubbers originally designed to remove particulate matter. These units were slightly modified to inject a sodium-based scrubbing liquor. Although removal of both particles and SO2 in one vessel can be economically attractive, the problems of high pressure drops and finding a scrubbing medium to remove heavy loadings of fly ash must be considered. However, in cases where the particle concentration is low, such as from oil-fired units, simultaneous particulate and SO2 emission reduction can be effective.

Operating Experience Presently a number of sodium-based throwaway FGD systems are in operation in the U.S., mainly on industrial boilers. Table 9-5 lists operating data for some of these systems. These systems are generally simpler to operate and maintain than lime or limestone systems. Therefore, reported operating problems have not been as severe or as frequent with the sodium-based system as with calcium-based systems. Control of pH, as with other FGD systems, is of prime concern to maximize absorption efficiency. Troubles with controlling pH can cause scale buildup and plugging of the sample lines. At high pH levels, the liquor absorbs CO2 and forms carbonate scale in systems where a high amount of calcium or magnesium is present (Makansi 1982). Other problems include ineffective entrainment separation, nozzle plugging, and failure of dampers, duct liners, and stack liners.

Lesson 9 ___________________________________________________________________________________

9-24 2.0-7/98

See Table 9-5. Operational data for sodium-based once-through FGD systems on utility and industrial boilers

Table 9-5. Operational data for sodium-based once-through FGD systems on utility and industrial bo

Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of

modules

L/G ratio

Pres



Nevada Power

Reid Gardner #1 125 Combustion Equipment Association

Cyclone/venturi 1.0 Sieve plate 1 0.2 1.6 0.7





Pacific Power & Light

Jim Bridger #4 550 Air Correction Division - UOP

ESP 0.6 Sieve plate 3 2.7 20.0 -

Alyeska Pipeline

Valdez, AK 25 FMC - 0.1 Disc-and-donut trays 1 1.6 12.0 -

Belridge Oil

McKittrick, CA 6 C-E NATCO - 1.1 Eductor venturi with variable disk

1 - - -

McKittrick, CA 6 Heater Technology

- 1.1 Eductor venturi with variable disk

1 5.4 40.0 -

McKittrick, CA 6 Thermotics - 1.1 Eductor venturi with variable disk

1 4.0 30.0 -

Chevron, USA

Bakersfield, CA 124 Koch Engineering - 1.1 Flexitrays 3 1.1 8.0 -

Double Barrel

Bakersfield, CA 6 C-E NATCO - 1.1 Spray tower/tray tower

1 3.3 25.0 -

FMC

Green River, WY 223 FMC ESP 1.0 Disc-and-donut trays 2 2.7 20.0 -


2.0-7/98 9-25

Table 9-5. (continued) Operational data for sodium-based once-through FGD systems on utility and industrial bo

Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of modules

L/G ratio

Pres



General Motors

St. Louis, MO 32 A. D. Little None 3.2 Impingement plate 1 - - -

Dayton, OH 18 Entoleter None 2.0 Vane cage 2 0.8 6.0 1.8

Tonowanda, NY 46 FMC Cyclone 1.2 Variable-throat venturi

4 2.7 20.0 -

Getty Oil

Bakersfield, CA 36 FMC None 1.1 Disc-and-donut tray/flexitray

1 1.1 8.4 -

Bakersfield, CA 445 In-house None 1.1 Flexitray 9 1.2 9.0 -

Orcutt, CA 2.5 In-house None 4.0 Packed tower 1 - - -

ITT Raynier

Fernandina Beach, FL

88 Neptune Airpol Cyclone 2.5 Variable-throat venturi

2 - - 5.5

Kerr-McGee

Trona, CA 245 Combustion Equipment Association

- 0.5-5

Plate tower 2 - - 1.5

Mead Paperboard

Stevenson, AL 50 Neptune Airpol Venturi 3.0 Bubble-cap plates 1 - - -

Northern Ohio Sugar

Freemont, OH 20 Great Western Sugar

None 1.0 Variable-throat venturi

2 - - -

Reichhold Chemicals

Pensacola, FL 40 Neptune Airpol None 2.0 Venturi 2 - - 6.0

Texasgulf

Granger, WY 70 Swemco Cyclone/ESP 0.8 Sieve plate 2 - - -

Note: A dash (-) indicates that no data are available.

Lesson 9 ___________________________________________________________________________________

9-26 2.0-7/98


Regenerable FGD Processes

Regenerable FGD processes remove SO2 from the flue gas and generate a saleable product. Regenerable products include elemental sulfur, sulfuric acid, or, in the case of lime or limestone scrubbing, gypsum (used for wallboard). Regenerable processes do not produce a sludge, thereby eliminating the sludge disposal problem. Most regenerable processes also achieve the following:

• Have the potential for consistently obtaining a high SO2 removal efficiency, usually exceeding 90%

• Utilize the scrubbing reagent more efficiently than nonregenerable processes

• Use scrubbing liquors that do not cause scaling and plugging problems in the scrubber

The major drawback of these processes is that systems using them are usually more complicated in design and are more expensive to install and operate.

Two regenerable processes presently operating in the U.S. are the Wellman-Lord and the magnesium oxide. The Wellman-Lord process has been widely used in both sulfuric acid and petroleum refining industries but has only been installed on a limited number of industrial and utility boilers. The magnesium oxide process has been tested at a number of utility boilers, but the Philadelphia Electric Company's Eddystone and Cromby Stations are the only utility boilers presently operating this process. Because of the limited use of regenerable processes in the utility industry, these processes are not covered in this course. Information on these processes can be obtained from numerous EPA and EPRI publications specific to the demonstration projects.

Emerging Technologies

As shown in Table 9-1 the overwhelming choice for SO2 control by utilities has been the use of lime or limestone wet scrubbers. The Clean Air Act Amendments of 1990 require reductions in acid rain precursors both SO2 and nitrogen oxides (NOx). Utilities have options as to specifically how they will comply; however, a number of new and/or retrofit FGD technologies will have to be installed. Because of the regulatory requirements and efforts to provide more efficient and cost-effective FGD systems, a number of new technologies are being investigated and developed by vendors, utilities and governmental agencies (EPA and DOE).

Table 9-6 provides summary information on certain new technologies that EPA has evaluated as likely candidates for retrofit to meet acid rain control requirements (Princiotta and Sedman 1993). Table 9-6 provides a description of specific SO2 and combined SO2/NOx control technologies as well as estimates of the level of control and commercial availability. Table 9-1 is not intended to be an all inclusive listing of every emerging FGD technology, as there are a number of others that may be viable options pending pilot demonstration.


2.0-7/98 9-27

Table 9-6. SO2 and SO2/NOx control technologies for coal-fired boilers

Technology Description Control %1 Estimated commercial

SO2 NOx availability

Wet flue gas desulfurization (FGD)

Limestone or lime in water removes SO2 in a scrubber vessel. Additives may be used to enhance SO2 removal. A wet waste or gypsum is produced.

70-97 0 Current for new boilers and retrofit.

Dry FGD Lime in water removes SO2 in a spray dryer, which evaporates the water prior to the vessel exit. Produces a dry waste.

70-95 0 Current for low to moderate S coal for new boilers. High S coal retrofit, 5 yrs.

E-SOx/in-duct injection Lime and water are injected in a boiler duct and/or ESP (E-SOx) similar to a spray dryer.

50-70 0 Pilot scale only. Demonstrations required, 3-7 yrs.

Advanced silicate (ADVACATE)

Several variations. Most attractive: adding limestone to boiler, generating lime. Lime/fly ash collected in cyclone and reacted to generate highly reactive silicate sorbent. Moist sorbent added to downstream duct.

Up to 90 0 Pilot scale only. Demonstrations required, 3-7 yrs.

Limestone injection multistage burners (LIMB)

Low NOx burners and upper furnace sorbent injection. May use humidification to improve SO2 capture and ESP performance.

50-70 40-60 Wall-fired, current; T-fired3, 2 yrs

Natural gas reburning Boiler fired with 80-90% coal. Remaining fuel (natural gas) is injected higher in boiler to reduce NOx. Air added to complete burnout. Sorbent may be injected to capture SO2.

Without sorbent, 10-20; with sorbent 50-60

50-60 Demonstrations in progress

SNRB Ammonia (NH3) and lime/sodium injection upstream of catalyst-coated baghouse.

90 90 5 MWe pilot plant in operation.

NOxSO SO2/NOx absorption on alumina in fluid bed reactor.

90 90 5 MWe pilot plant in Clean Coal Technology (CCT) program.

WSA-SNOx Catalytic reduction of nitric oxide (NO) and oxidation of SO2 in two stages. Sulfuric acid recovery.

95 90 35 MWe pilot in CCT program; 1 unit in Denmark.

NONOx Ozone/NH3 promoted absorption of SO2/NOx in wet scrubber.

95 75-95 Commercial construction in Europe.

Lesson 9 ___________________________________________________________________________________

9-28 2.0-7/98


Technology Description Control %1 Estimated commercial


Activated char NH3 injection and absorption of SO2/SO3 on char; NO reduction.

90 70 Operational on 3 plants in Europe, 1 in Japan.

DESONOx One step variant of WSA-SNOx above. 85 80 20 MWe demo operating in Germany.

Amine absorption Amine absorption of SO2 and NOx followed by regeneration; acid production.

90+ 90+ Several vendors/processes; pilot-scale systems in operation.

Ferrous chelate additive Ferrous chelate added to magnesium/calcium FGD solubilizes NO.

90 30-70 3 MWe pilot plant in operation.

1. Control efficiency is % reduction from emission levels for uncontrolled coal-fired power plants. 2. Estimated commercialization for some technologies is strongly dependent on successful demonstrations. 3. T-fired = tangentially fired.

Source: Princiotta and Sedman 1993.


2.0-7/98 9-29

Summary

FGD systems have been installed and operated on many industrial and utility boilers and on some industrial processes for a number of years. These systems are capable of removing approximately 70 to 90% of the SO2 in the flue gas, depending on the operating conditions of the system. Some systems have achieved an SO2-removal efficiency of greater than 95%. The most popular FGD systems used on utility boilers are lime or limestone scrubbing. Approximately 75% of the FGD systems installed on utility boilers are either lime or limestone scrubbing. The use of dual-alkali systems on utility boilers is attractive because of their ability to remove SO2 very efficiently and to reduce scaling problems. The throwaway-sodium FGD systems have been used mostly on industrial boilers. These systems use a sodium scrubbing liquor that is very efficient in absorbing SO2 emissions, but they produce liquid wastes that can cause waste disposal problems. FGD systems used on utility boilers generate large quantities of liquid wastes. Therefore, throwaway-sodium systems have mainly been used on industrial boilers. Wellman-Lord FGD systems have been used to reduce SO2 emissions from utility and industrial boilers and from a number of industrial processes. These systems have the advantage of regenerating the scrubbing liquor and producing a saleable product instead of a sludge that can be a disposal problem. However, these systems are more expensive to install and operate than lime, limestone, or dual-alkali systems.

Over the past 25 years, a wealth of material has been written and documented concerning FGD control technology. The authors of this manual suggest that the readers utilize the many publications from EPA and the Electric Power Research Institute (EPRI) concerning this subject, particularly the proceedings from the FGD symposiums sponsored by the EPA.


Lesson 9 ___________________________________________________________________________________

9-30 2.0-7/98


2.0-7/98 9-31


1. True or False? Only wet FGD systems have been used on utility boilers.

2. ____________________-based slurries absorb SO2 better than ____________________; however, the former are much more expensive.

a. Sodium, lime or limestone b. Lime or limestone, sodium c. Gypsum, lime or limestone d. Limestone, lime

3. Solutions of sodium compounds are referred to as clear liquor solutions because the compounds are:

a. Blue b. Soluble c. Insoluble d. Transparent

4. True or False? Almost all FGD systems use a single wet scrubber for both SO2 and fly ash removal.

5. Which problem/problems must be considered when trying to remove both SO2 and fly ash in the same scrubber?

a. Pressure drops are higher b. The scrubbing liquid, if recirculated, can contain a high level of fly ash c. SO2 absorption efficiency is normally lower d. All of the above

6. Spray towers on most FGD systems require higher ____________________ (for equivalent SO2 removal) than other absorber designs.

a. Pressure drops b. Gas velocities c. Liquid-to-gas ratios d. All of the above

7. When the gas velocity is lowered, entrainment becomes ____________________; however, the scrubber system will be ____________________ costly.

a. More, more b. More, less c. Less, more d. Less, less

Lesson 9 ___________________________________________________________________________________

9-32 2.0-7/98

8. List five properties of the coal (or fuel) that will affect FGD operation.

________________________________________

________________________________________

________________________________________

________________________________________

________________________________________

9. Because flue gas contains some SO2 as it exits the absorber, FGD systems generally use ____________________ to prevent corrosion.

a. Additional absorbers b. Reheaters c. Special construction materials for downstream fans and ductwork d. Both b and c

Part 2

10. List three nonregenerable FGD processes.

________________________________________

________________________________________

________________________________________

11. Dissolving lime in water is referred to as:

a. Clarifying b. Slaking c. Raking d. Thickening

12. What is CaSO3 in the following reaction?

Ca SO 2H OH++3= + -+ + +2 →CaSO H O3 2+ 2

a. Sludge b. Liquid c. Gas

13. Lime FGD systems use a(an) ____________________ to remove fly ash from the flue gas before it enters the absorber.

a. Venturi scrubber b. Electrostatic precipitator c. Mechanical collector with precipitator or scrubber d. Any of the above


2.0-7/98 9-33

14. In early lime FGD systems, scale buildup and plugging of the ____________________ were particularly troublesome.

a. Spray nozzles b. Entrainment separator c. Scrubber internals d. All of the above

15. Operating a lime FGD system at a pH above 8.0 to 9.0:

a. Reduces scale buildup b. Increases the risk of scale buildup c. Is recommended d. Eliminates nozzle plugging

16. Most lime FGD systems on utility boilers operate at L/G ratios of:

a. 0.4 to 1.3 L/m3 (3 to 10 gal/1000 ft3) b. 3.0 to 8.0 L/m3 (25 to 60 gal/1000 ft3) c. 13 to 26 L/m3 (100 to 200 gal/1000 ft3) d. None of the above

17. ____________________ liquid-to-gas ratios reduce the potential for scale buildup.

a. High b. Low

18. Stack gas is reheated to:

a. Avoid condensation b. Enhance plume rise c. Give better pollutant dispersion d. All of the above

Part 3

19. Limestone FGD systems generally operate at ____________________ liquid-to-gas ratios than lime FGD systems because SO2 is ____________________ reactive with a limestone slurry.

a. Higher, more b. Higher, less c. Lower, more d. Lower, less

20. True or False? The chemistry for SO2 removal in a limestone slurry is very different from that for SO2 removal in a lime slurry.

21. The major difference in equipment for a limestone FGD system (compared to a lime FGD system) is in the:

a. Fly ash collection equipment b. Type of absorber c. Slurry feed preparation d. All of the above

Lesson 9 ___________________________________________________________________________________

9-34 2.0-7/98

22. True or False? Limestone is generally less expensive to purchase than lime.

23. In lime/limestone FGD systems, calcium sulfite formed as part of the sludge is difficult to remove from the slurry. One method used to eliminate this problem is to convert the calcium sulfite to calcium sulfate by the process called:

a. Forced oxidation b. Wellman-Lord c. Double-alkali d. Direct reduction

Part 4

24. Double-alkali processes generally use a ____________________ solution to absorb the SO2 from the flue gas and then react it with a ____________________ slurry to regenerate the absorbing solution.

a. Sodium, citrate b. Citrate, lime or limestone c. Sodium, lime or limestone d. Lime or limestone, sodium

25. In the double-alkali process, the sodium reagent is regenerated by reacting the sludge with lime. As part of this reaction, insoluble ____________________ are formed in the regeneration vessel.

a. Sodium salts b. Calcium salts c. Magnesium salts d. Citrate salts

26. Compared to lime and limestone scrubbing systems, double-alkali absorbers have a much lower:

a. Pressure drop b. Gas velocity c. Liquid-to-gas ratio d. All of the above

27. True or False? Using sodium-based scrubbing solutions (as compared to calcium-based) helps eliminate scale buildup.

Part 5

28. True or False? The two sodium compounds used most often in throwaway systems are sodium hydroxide (NaOH) and sodium carbonate (Na2CO3).

29. Sodium-based once-through FGD systems have been used on industrial boilers because:

a. Sodium is the most efficient of the commercial reagents b. They operate without scale buildup occurring c. They are often smaller and cheaper than other systems d. All of the above


2.0-7/98 9-35

30. Large utilities have not used sodium-based once-through systems because of the expense of the sodium reagent and the:

a. Limited efficiency b. Low fly ash removal c. Presence of soluble salts in the wastes (wastes cannot be discharged into rivers or lakes) d. All of the above

31. True or False? In a sodium-based once-through FGD system, the flue gas may first pass through a baghouse or ESP.

32. True or False? Sodium-based once-through systems are generally simpler to operate and maintain than lime or limestone FGD systems.

33. At high pH values, the scrubbing liquid in the sodium systems absorbs ____________________ and can form carbonate scale.

a. SO2

b. CO2

c. O2

d. CaCO3

Part 6

34. Regenerable FGD processes generate a saleable product such as:

a. Sulfur b. Sulfuric acid c. Gypsum d. All of the above

35. List at least three advantages that the regenerable process has over the nonregenerable FGD process.

________________________________________

________________________________________

________________________________________

Lesson 9 ___________________________________________________________________________________

9-36 2.0-7/98


2.0-7/98 9-37


1. False Dry FGD systems have been installed on some utility sized boilers (see Lesson 7).

2. a. Sodium, lime or limestone Sodium-based slurries absorb SO2 better than lime or limestone; however, the former are much more expensive.

3. b. Soluble Solutions of sodium compounds are referred to as clear liquor solutions because the compounds are soluble.

4. False Most FGD systems use two scrubbing stages: one for SO2 removal and another for fly ash removal.

5. d. All of the above Problems that must be considered when trying to remove both SO2 and fly ash in the same scrubber are:

• Pressure drops are higher

• The scrubbing liquid, if recirculated, can contain a high level of fly ash

• SO2 absorption efficiency is normally lower

6. c. Liquid-to-gas ratios Spray towers on most FGD systems require higher liquid-to-gas ratios (for equivalent SO2 removal) than other absorber designs. More liquid is used in spray towers because they have limited contact area available for absorption.

7. c. Less, more When the gas velocity is lowered, entrainment becomes less; however, the scrubber system will be more costly.

8. Five properties of coal (or fuel) that will affect FGD operation are:

• Heating value

• Sulfur content

• Chlorine content

• Ash content

• Moisture content

9. d. Both b and c Because flue gas contains some SO2 as it exits the absorber, FGD systems generally use reheaters and special construction materials for downstream fans and ductwork to prevent corrosion.

Lesson 9 ___________________________________________________________________________________

9-38 2.0-7/98

Part 2

10. Lime, Limestone, Double-alkali Three nonregenerable FGD processes are:

• Lime

• Limestone

• Double-alkali

11. b. Slaking Dissolving lime in water is referred to as slaking.

12. a. Sludge In the following reaction, CaSO3 is sludge.

Ca SO 2H OH++3= + -+ + +2 →CaSO H O3 2+ 2

13. d. Any of the above To remove fly ash from the flue gas before it enters the absorber, lime FGD systems can use any of the following:

• A venturi scrubber

• An electrostatic precipitator

• A mechanical collector with precipitator or scrubber

14. d. All of the above In early lime FGD systems, scale buildup and plugging of the spray nozzles, entrainment separator, and scrubber internals were particularly troublesome.

15. b. Increases the risk of scale buildup Operating a lime FGD system at a pH above 8.0 to 9.0 increases the risk of scale buildup.

16. b. 3.0 to 8.0 L/m3 (25 to 60 gal/1000 ft3) Most lime FGD systems on utility boilers operate at L/G ratios of 3.0 to 8.0 L/m3 (25 to 60 gal/1000 ft3).

17. a. High High liquid-to-gas ratios reduce the potential for scale buildup.

18. d. All of the above Stack gas is reheated to:

• Avoid condensation

• Enhance plume rise

• Give better pollutant dispersion


2.0-7/98 9-39

Part 3

19. b. Higher, less Limestone FGD systems generally operate at higher liquid-to-gas ratios than lime FGD systems because SO2 is less reactive with a limestone slurry.

20. False The chemistry for SO2 removal in a limestone slurry is very similar to that for SO2 removal in a lime slurry.

21. c. Slurry feed preparation The major difference in equipment for a limestone FGD system (compared to a lime FGD system) is in the slurry feed preparation.

22. True Limestone is generally less expensive to purchase than lime.

23. a. Forced oxidation In lime/limestone FGD systems, calcium sulfite formed as part of the sludge is difficult to remove from the slurry. One method used to eliminate this problem is to convert the calcium sulfite to calcium sulfate by the process called forced oxidation.

Part 4

24. c. Sodium, lime or limestone Double-alkali processes generally use a sodium solution to absorb the SO2 from the flue gas and then react it with a lime or limestone slurry to regenerate the absorbing solution.

25. b. Calcium salts In the double-alkali process, the sodium reagent is regenerated by reacting the sludge with lime. As part of this reaction, insoluble calcium salts are formed in the regeneration vessel.

26. c. Liquid-to-gas ratio Compared to lime and limestone scrubbing systems, double-alkali absorbers have a much lower liquid-to-gas ratio. Double-alkali systems use sodium which is more effective at acid gas absorption than lime and limestone per mole of compound used. Therefore less sodium and less scrubbing liquid are required.

27. True Using sodium-based scrubbing solutions (as compared to calcium-based) helps eliminate scale buildup. Sodium compounds do not form slake as readily as calcium compounds do.

Part 5

28. True The two sodium compounds used most often in throwaway systems are sodium hydroxide (NaOH) and sodium carbonate (Na2CO3).

Lesson 9 ___________________________________________________________________________________

9-40 2.0-7/98

29. d. All of the above Sodium-based once-through FGD systems have been used on industrial boilers because:

• Sodium is the most efficient of the commercial reagents

• They operate without scale buildup occurring

• They are often smaller and cheaper than other systems

30. c. Presence of soluble salts in the wastes (wastes cannot be discharged into rivers or lakes)

Large utilities have not used sodium-based once-through systems because of the expense of the sodium reagent and the presence of soluble salts in the wastes which means wastes cannot be discharged into rivers or lakes.

31. True In a sodium-based once-through FGD system, the flue gas may first pass through a baghouse or ESP.

32. True Sodium-based once-through systems are generally simpler to operate and maintain than lime or limestone FGD systems.

33. b. CO2 At high pH values, the scrubbing liquid in the sodium systems absorbs CO2 and can form carbonate scale.

Part 6

34. d. All of the above Regenerable FGD processes generate a saleable product such as:

• Sulfur

• Sulfuric acid

• Gypsum

35. Avoidance of sludge disposal problems Consistently higher SO2 removal Better utilization of reagent Use of clear liquid solutions (reduces scaling)

Four advantages that the regenerable process has over the nonregenerable FGD process include:

• Avoidance of sludge disposal problems

• Consistently higher SO2 removal

• Better utilization of reagent

• Use of clear liquid solutions (reduces scaling)


2.0-7/98 9-41

Bibliography Black & Veatch Consulting Engineers. 1983. Lime FGD Systems Data Book. 2nd ed. EPRI

Publication No. CS-2781.

Devitt, T., R. Gerstle, L. Gibbs, S. Hartman, and R. Klier. 1978, March. Flue Gas Desulfurization System Capabilities for Coal-Fired Steam Generators, Vols. I and II. EPA 600/7-78-032a and b. U.S. Environmental Protection Agency.

Hance, S. B., and J. L. Kelly. 1991. Status of flue gas desulfurization systems. Paper presented at the 84th Annual Meeting of the Air and Waste Management Association. Paper No. 91-157.3.

Leivo, C. C. 1978. Flue Gas Desulfurization Systems: Design and Operating Considerations. Vol. II. Technical Report. EPA 600/7-78-030b. U.S. Environmental Protection Agency.

Makansi, J. 1982, October. SO2 control: optimizing today’s processes for utility and industrial power plants. Power.

Mobley, J. D., and J. C. S. Chang. 1981. The adipic acid enhanced limestone flue gas desulfurization process - an assessment. Journal of the Air Pollution Control Association. 31:1249-1253.

Ponder, T. C., J. S. Hartman, H. M. Drake, R. P. Klier, J. S. Master, A. N. Patkar, R. D. Tems, and J. Tuttle. 1979. Lime FGD Systems Data Book. EPA 600/8-79-009. U.S. Environmental Protection Agency.


Smith, M., M. Melia, N. Gregory, and K. Scalf. 1981, January. EPA Utility FGD Survey: October - December 1980. Vols. I and II. EPA 600/7-81-012a and b. U.S. Environmental Protection Agency.

Tuttle, J., A. Paktar, S. Kothari, D. Osterhout, M. Heffling, and M. Eckstein. 1979. EPA Industrial Boiler FGD Survey: First Quarter 1979. EPA 600/7-79-067b. U.S. Environmental Protection Agency.

U.S. Environmental Protection Agency. 1976. Lime/Limestone Wet-Scrubbing Test Results at the EPA Alkali Scrubbing Test Facility. EPA Technology Transfer Capsule Report, Second Progress Report. EPA 625/2-75-008.

U.S. Environmental Protection Agency. 1976. Lime/Limestone Wet-Scrubbing Test Results at the EPA Alkali Scrubbing Test Facility. EPA Technology Transfer Capsule Report, Third Progress Report. EPA 625/2-76-010.

U.S. Environmental Protection Agency. 1978, March. Proceedings: Symposium on Flue Gas Desulfurization - Hollywood, Florida, November 1977. Vols. I and II. EPA 600/7-78-058a and b.

Lesson 9 ___________________________________________________________________________________

9-42 2.0-7/98

U.S. Environmental Protection Agency. 1979, May. Sulfur Emission: Control Technology and Waste Management. Decision Series. EPA 600/9-79-019.

U.S. Environmental Protection Agency. 1979, July. Proceedings: Symposium on Flue Gas Desulfurization - Las Vegas, Nevada, March 1979. Vols. I and II. EPA 600/7-79-167a and b.

U.S. Environmental Protection Agency. 1980, August. Controlling Sulfur Oxides. Research Summary. EPA 600/8-80-029.

U.S. Environmental Protection Agency. 1981, April. Control Techniques for Sulfur Oxide Emissions from Stationary Sources. EPA 450/3-81-004.

U.S. Environmental Protection Agency. 1981. Proceedings: Symposium on Flue Gas Desulfurization - Houston, Texas, October 1980. Vols. I and II. EPA 600/9-81-019a and b.

Table 9-2. (continued) Operational data for lime FGD systems on utility boilers

Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of

modules

L/G ratio

Pressure drop( ∆p)

Efficiency (%)


L/m3 gal/1000 ft3

kPa in. H2O Design Test



- 0.6 Mobile bed (marbles) 2 3.5 26.0 2.7 11.0 70.0 70.0


- 0.6 Mobile bed (marbles) 2 3.5 26.0 2.7 11.0 70.0 70.0

Monongahela Power

Pleasants #1 618 B&W ESP 3.7 Sieve tray 4 7.4 55.0 1.2 5.0 90.0 90.0

Pleasants #2 618 B&W ESP 4.5 Sieve tray 4 7.4 55.0 - - 90.0 90.0

Utah Power & Light

Hunter #1 400 Chemico ESP 0.6 Countercurrent spray 4 5.7 43.0 0.6 2.5 80.0 80.0

Hunter #2 400 Chemico ESP 0.6 Countercurrent spray 4 5.7 43.0 0.6 2.5 80.0 80.0

Huntingdon #1 430 Chemico ESP 0.6 Countercurrent spray 4 5.7 43.0 0.6 2.5 80.0 80.0 Note: A dash (-) indicates that no data are available.

2.0-7/98 9-10

Table 9-2. Operational data for lime FGD systems on utility boilers

Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of

modules

L/G ratio


Efficiency (%)


L/m3 gal/1000 ft3


Pennsylvania Power

Bruce Mansfield #1 917 Chemico 1st-stage venturi 3.0 Fixed-throat venturi 6 6.0 45.0 2.0 8.0 92.1 95.0

Bruce Mansfield #2 917 Chemico 1st-stage venturi 3.0 Fixed-throat venturi 6 6.0 45.0 2.0 8.0 92.1 95.0

Bruce Mansfield #3 917 Pullman Kellogg ESP 3.0 Weir crosscurrent spray

6 - - 0.7 2.8 92.0 95.0

Columbus & Southern Ohio Electric

Conesville #5 411 Air Correction Division

ESP 4.7 Mobile bed 1 6.7 50.0 2.0 8.0 89.5 89.7

Conesville #6 411 Air Correction Division

ESP 4.7 Mobile bed 2 6.7 50.0 2.0 8.0 89.5 89.5

Duquesne Light

Elrama 510 Chemico ESP 2.2 Variable-throat venturi

5 5.3 40.0 4.0 16.0 83.0 86.0

Phillips 408 Chemico Cyclone/ESP 1.9 Variable-throat venturi

4 5.3 40.0 4.0 16.0 83.0 90.0

Kentucky Utilities

Green River 64 American Air Filter Cyclone/ variable-throat venturi

4.0 Mobile bed 1 4.5 34.0 1.0 4.0 80.0 80.0


Cane Run #4 188 American Air Filter ESP 3.7 Mobile bed 2 8.0 60.0 1.0 40.0 85.0 87.5

Cane Run #5 200 Combustion Engineering

ESP 3.7 Countercurrent spray 2 7.4 55.0 0.1 0.5 85.0 91.0

Mill Creek #1 358 Combustion Engineering

- 3.7 - - 12.7 95.0 - - 85.0 86.6

Mill Creek #3 442 American Air Filter ESP 3.7 Mobile bed 4 8.7 65.0 1.6 6.5 85.0 85.7

Paddy’s Run #6 72 Combustion Engineering

ESP 2.5 Mobile bed (marbles) 2 2.2 16.5 2.9 11.5 90.0 90.0


2.0-7/98 9-9


Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of

modules

L/G ratio


Efficiency (%)


L/m3 gal/1000 ft3

kPa in. H2O

Design Test


La Cygne 820 B&W Variable venturi 5.4 Sieve tray 8 5.0 37.7 1.5 6.0 80.0 80.0


ESP 0.3 Countercurrent spray 6 4.1 30.4 1.0 6.0 80.0 60.0


ESP 0.3 Countercurrent spray - 4.1 30.4 1.0 6.0 80.0 60.0


Rod venturi 0.6 Countercurrent spray 2 4.0 30.0 0.6 2.5 73.0 73.0


Rod venturi 0.6 Countercurrent spray 2 2.5 19.0 0.6 2.5 52.0 52.0

Salt River Project


2 - - 0.4 1.5 66.0 82.0


2 - - 0.4 1.5 66.0 82.0

South Carolina Public Service

Winyah #2 280 B&W ESP 1.7 Venturi/sieve tray 2 6.3 47.5 1.1 4.5 45.0 90.0

Winyah #3 280 B&W ESP 1.7 Countercurrent spray 2 - - - - 90.0 90.0

South Mississippi Electric


1 6.6 49.0 2.0 8.0 52.7 85.0


1 6.6 49.0 2.0 8.0 52.7 85.0

Southern Illinois

Marion #4 173 B&W ESP 3.8 Countercurrent spray 2 9.9 74.0 1.5 6.0 89.4 89.4 Continued on next page

2.0-7/98 9-14


Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of

modules

L/G ratio

Pressure drop ( ∆p)

Efficiency (%)


L/m3 gal/1000 ft3


Springfield City

Southwest #1 194 Air Correction Division - UOP

ESP 3.5 Mobile bed (TCA) 2 5.5 41.0 1.5 6.0 80.0 87.0

Springfield Water, Light & Power

Dallman #3 205 Research-Cottrell Cyclone/ESP 3.3 Spray/packed tower 2 - - 0.2 0.7 95.0 95.0

TVA

Widows Creek #8 550 TVA ESP/venturi 3.7 Mobile packed bed

and grid packing

1

3

60.0 0.5 2.0 70.0 -

Texas Power & Light

Sandow #4 545 Combustion Engineering

ESP 1.6 Countercurrent spray 3 - - - - 75.0 -

Texas Utilities

Martin Lake #1 793 Research-Cottrell ESP 0.9 Spray/packed bed 6 - - 1.1 4.5 71.0 95.0



Monticello 800 Chemico ESP 1.5 Countercurrent spray 3 9.4 70.0 1.2 5.0 74.0 74.0

Note: A dash (-) indicates that no data are available.

9-15 2.0-7/98

Table 9-3. Operational data for limestone FGD systems on utility boilers

Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of

modules

L/G ratio


Efficiency (%)


L/m3 gal/1000 ft3


Alabama Electric

Tombigbee #2 255 Peabody ESP 1.2 Countercurrent spray 2 9.4 70.0 1.0 4.0 59.5 85.0

Tombigbee #3 255 Peabody ESP 1.2 Countercurrent spray 2 9.4 70.0 1.0 4.0 59.5 85.0

Arizona Electric Power

Apache #2 195 Research-Cottrell ESP 0.5 Spray/packed bed 2 2.8 20.6 1.5 6.0 42.5 97.0

Apache #3 195 Research-Cottrell ESP 0.5 Spray/packed bed 2 2.8 20.6 1.5 6.0 42.5 97.0

Cholla #1 119 Research-Cottrell Cyclone/venturi 0.5 Spray/packed bed 1 6.5 48.9 0.1 0.5 58.5 92.0

Cholla #2 264 Research-Cottrell Cyclone/venturi 0.5 Spray/packed bed 4 6.5 48.9 0.1 0.5 75.0 85.0

Basin Electric Power

Laramie River #1 570 Research-Cottrell ESP 0.8 Spray/packed bed 5 8.0 60.0 - - 90.0 90.0

Laramie River #2 570 Research-Cottrell ESP 0.8 Spray/packed bed 5 8.0 60.0 - - 90.0 90.0

Central Illinois Light

Duck Creek #1 416 Environeering ESP 3.7 Rod deck packed tower

4 6.7 50.0 2.0 8.0 85.0 85.0

Colorado Ute Electrical

Craig #1 447 Peabody ESP 0.4 Countercurrent spray 4 6.7 50.0 1.6 6.5 85.0 85.0

Craig #2 455 Peabody ESP 0.4 Countercurrent spray 4 6.7 50.0 1.6 6.5 85.0 85.0

Commonwealth Edison

Powerton 450 Air Correction Division - UOP


Indianapolis Power & Light

Petersburg #3 532 Air Correction Division - UOP



2.0-7/98 9-13

Table 9-4. Operational data for double-alkali FGD systems on utility and industrial boilers

Company and plant name

MW

FGD vendor

Fly ash control

%S in

SO2 absorber

No. of

modules per

L/G ratio

Pressure drop ( ∆p)

Efficiency (%)

(gross) coal boiler L/m3 gal/1000 ft3


Central Illinois Public Service

Newton #1 617 Buell ESP 2.5 Mobile bed 4 1.3 10.0 1.5 6.0 90.0 90.0


Cane Run #6 299 Combustion Equipment Association

ESP 4.8 Sieve plates 2 1.3 10.0 2.5 9.9 95.0 94.2

Southern Indiana Gas & Electric

A. B. Brown #1 265 FMC ESP 3.6 Variable-throat venturi

2 1.3 10.0 2.5 10.0 85.0 85.0

Caterpillar Tractor

East Peoria, IL 105 FMC Cyclone 3.2 Venturi 4 2.2 16.0 - - - 90.0

Joliet, IL 34 Zurn Cyclone 3.2 Dustraxtor 2 - - - - - 90.0

Morton, IL 19 Zurn Cyclone 3.2 Dustraxtor 2 - - - - - 90.0

Mossville, IL 70 FMC Cyclone - Venturi 4 1.2 8.6 - - - 90.0+

Firestone Tire

Pottstown, PA 4 FMC Cyclone 3.0 Venturi 1 1.3 10.0 - - - 90.5

General Motors

Parma, OH 64 GM Environmental Cyclone - Bubble-cap plates 4 2.6 20.0 0.9 8.0 - 90.0 Note: A dash (-) indicates that no data are available.

9-20 2.0-7/98

Table 9-5. Operational data for sodium-based once-through FGD systems on utility and industrial boilers

Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of

modules

L/G ratio


Efficiency (%)


L/m3 gal/1000 ft3 kPa in. H2O Design Test

Nevada Power


Cyclone/venturi 1.0 Sieve plate 1 0.2 1.6 0.7 3.0 90.0 -


Cyclone/venturi 1.0 Sieve plate 1 0.2 1.6 0.7 3.0 90.0 91.2


Cyclone/venturi 1.0 Sieve plate 1 0.2 1.6 0.7 3.0 85.0 91.2

Pacific Power & Light

Jim Bridger #4 550 Air Correction Division - UOP

ESP 0.6 Sieve plate 3 2.7 20.0 - - 91.0 91.0

Alyeska Pipeline

Valdez, AK 25 FMC - 0.1 Disc-and-donut trays 1 1.6 12.0 - - - 96.0

Belridge Oil

McKittrick, CA 6 C-E NATCO - 1.1 Eductor venturi with variable disk

1 - - - - - 90.0

McKittrick, CA 6 Heater Technology

- 1.1 Eductor venturi with variable disk

1 5.4 40.0 - - - 90.0

McKittrick, CA 6 Thermotics - 1.1 Eductor venturi with variable disk

1 4.0 30.0 - - - 90.0

Chevron, USA

Bakersfield, CA 124 Koch Engineering - 1.1 Flexitrays 3 1.1 8.0 - - - 90.0

Double Barrel

Bakersfield, CA 6 C-E NATCO - 1.1 Spray tower/tray tower

1 3.3 25.0 - - - 95.0

FMC

Green River, WY 223 FMC ESP 1.0 Disc-and-donut trays 2 2.7 20.0 - - - 95.0

Continued on next page 9-24 2.0-7/98

Table 9-6. (continued) SO2 and SO2/NOx control technologies for coal-fired boilers

Technology Description Control %1 Estimated commercial Comments


Activated char NH3 injection and absorption of SO2/SO3 on char; NO reduction.

90 70 Operational on 3 plants in Europe, 1 in Japan.

DESONOx One step variant of WSA-SNOx above. 85 80 20 MWe demo operating in Germany.

Amine absorption Amine absorption of SO2 and NOx followed by regeneration; acid production.

90+ 90+ Several vendors/processes; pilot-scale systems in operation.

Ferrous chelate additive Ferrous chelate added to magnesium/calcium FGD solubilizes NO.

90 30-70 3 MWe pilot plant in operation.

1. Control efficiency is % reduction from emission levels for uncontrolled coal-fired power plants. 2. Estimated commercialization for some technologies is strongly dependent on successful demonstrations. 3. T-fired = tangentially fired.

Source: Princiotta and Sedman 1993.

9-28 2.0-7/98

Table 9-5. (continued) Operational data for sodium-based once-through FGD systems on utility and industrial boilers

Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of

modules

L/G ratio


Efficiency (%)



General Motors

St. Louis, MO 32 A. D. Little None 3.2 Impingement plate 1 - - - - - 90.0

Dayton, OH 18 Entoleter None 2.0 Vane cage 2 0.8 6.0 1.8 7.0 - 86.0

Tonowanda, NY 46 FMC Cyclone 1.2 Variable-throat venturi

4 2.7 20.0 - - - 95.0

Getty Oil

Bakersfield, CA 36 FMC None 1.1 Disc-and-donut tray/flexitray

1 1.1 8.4 - - - 90.0

Bakersfield, CA 445 In-house None 1.1 Flexitray 9 1.2 9.0 - - - 96.0

Orcutt, CA 2.5 In-house None 4.0 Packed tower 1 - - - - - 94.0

ITT Raynier

Fernandina Beach, FL

88 Neptune Airpol Cyclone 2.5 Variable-throat venturi

2 - - 5.5 22.0 - 85.0

Kerr-McGee

Trona, CA 245 Combustion Equipment Association

- 0.5-5 Plate tower 2 - - 1.5 6.0 98.0

Mead Paperboard

Stevenson, AL 50 Neptune Airpol Venturi 3.0 Bubble-cap plates 1 - - - - - 95.0

Northern Ohio Sugar

Freemont, OH 20 Great Western Sugar

None 1.0 Variable-throat venturi

2 - - - - - -

Reichhold Chemicals

Pensacola, FL 40 Neptune Airpol None 2.0 Venturi 2 - - 6.0 24.0 - -

2.0-7/98 9-25

Table 9-5. (continued) Operational data for sodium-based once-through FGD systems on utility and industrial boilers

Company and

MW

FGD vendor

Fly ash

%S in

SO2 absorber

No. of

modules

L/G ratio


Efficiency (%)



Texasgulf

Granger, WY 70 Swemco Cyclone/ESP 0.8 Sieve plate 2 - - - - - 90.0

Note: A dash (-) indicates that no data are available. Continued on next page

2.0-7/98 9-25


Technology Description Control %1 Estimated commercial Comments


Wet flue gas desulfurization (FGD)

Limestone or lime in water removes SO2 in a scrubber vessel. Additives may be used to enhance SO2 removal. A wet waste or gypsum is produced.

70-97 0 Current for new boilers and retrofit.

State-of-the-art for higher S (sulfur) coal and FGD. Certain retrofits difficult.

Dry FGD Lime in water removes SO2 in a spray dryer, which evaporates the water prior to the vessel exit. Produces a dry waste.

70-95 0 Current for low to moderate S coal for new boilers. High S coal retrofit, 5 yrs.

Demonstration for high S coal retrofit is necessary, but may be limited to 90% SO2 removal.

E-SOx/in-duct injection Lime and water are injected in a boiler duct and/or ESP (E-SOx) similar to a spray dryer.

50-70 0 Pilot scale only. Demonstrations required, 3-7 yrs.

Potentially low cost retrofits. May be site-specific limits.

Advanced silicate (ADVACATE)

Several variations. Most attractive: adding limestone to boiler, generating lime. Lime/fly ash collected in cyclone and reacted to generate highly reactive silicate sorbent. Moist sorbent added to downstream duct.

Up to 90 0 Pilot scale only. Demonstrations required, 3-7 yrs.

Most promising emerging retrofit technology. Capable of 90% removal with costs 50% of wet scrubber.

Limestone injection multistage burners (LIMB)

Low NOx burners and upper furnace sorbent injection. May use humidification to improve SO2 capture and ESP performance.

50-70 40-60 Wall-fired, current; T-fired3, 2 yrs

T-fired wall-fired demonstration complete. Applicable to ≤ 3% S coal retrofits.

Natural gas reburning Boiler fired with 80-90% coal. Remaining fuel (natural gas) is injected higher in boiler to reduce NOx. Air added to complete burnout. Sorbent may be injected to capture SO2.

Without sorbent, 10-20; with sorbent 50-60

50-60 Demonstrations in progress May be only combustion NOx control for cyclones. Sensitive to natural gas price. New or retrofit.

SNRB Ammonia (NH3) and lime/sodium injection upstream of catalyst-coated baghouse.

90 90 5 MWe pilot plant in operation.

NOxSO SO2/NOx absorption on alumina in fluid bed reactor.

90 90 5 MWe pilot plant in Clean Coal Technology (CCT) program.

WSA-SNOx Catalytic reduction of nitric oxide (NO) and oxidation of SO2 in two stages. Sulfuric acid recovery.

95 90 35 MWe pilot in CCT program; 1 unit in Denmark.

NONOx Ozone/NH3 promoted absorption of SO2/NOx in wet scrubber.

95 75-95 Commercial construction in Europe.


2.0-7/98 9-27

2.0-7/98 10-1

Lesson 10 Design Evaluation of Particulate Wet Scrubbing Systems

Goal To familiarize you with the factors to be considered when evaluating particulate-pollutant scrubber design plans.


1. Explain the importance of the following factors in scrubber design:

• Dust properties

• Exhaust gas characteristics

• Static pressure drop

• Liquid flow rate

• Collection efficiency

• Removal of entrained droplets

2. Estimate the collection efficiency and pressure drop of a venturi scrubber using appropriate equations and graphs

3. Use the contact power method to estimate collection efficiency

4. Describe the strengths and limitations of the contact power method.

Introduction

In performing an evaluation of a new scrubbing system design, especially from a regulatory perspective, the major issue is whether the proposed design will achieve the required particle and/or gas removal efficiencies. In addition to addressing this basic issue, there is also the question of how effectively the proposed system will operate. For example, will the system be able to handle a sufficient range of expected operating conditions without requiring

Lesson 10 ___________________________________________________________________________________

10-2 2.0-7/98

excessive maintenance or downtime? Answering these questions is difficult since there is no one set of theoretical equations that will provide an absolute answer.

There are three basic approaches to evaluating the capability of a scrubbing system: (1) empirical relationships based on historical test data on similar scrubbers, (2) theoretical models based on basic engineering principles and (3) pilot scale test data. A scrubber vendor has access to all three (especially historical information) when designing a system. A person reviewing the design generally does not have easy access to this type of information. When conducting a review, first, start with the theoretical equations to verify the basic design then supplement this information with data on similar systems obtained from literature or the scrubber vendor.

In the previous lessons, you have become familiar with operating and maintenance data on a variety of scrubbing systems. This lesson will first present an overview of the general parameters that affect scrubber design and then cover the following:

• Theoretical models for estimating particle collection efficiency

• Estimating venturi static pressure drop

A reviewer can then use the equations in this lesson coupled with historical data to evaluate scrubbing systems. You will have the opportunity to practice using the equations presented in this lesson by working the three problems in the Review Exercise.

Particulate Scrubber Design Factors

In order to properly design a particulate wet scrubber, the vendor must obtain as much information as possible concerning the characteristics of the flue gas stream to be treated. This information must be obtained or estimated for both the average and maximum ranges that will occur. Scrubbing systems must be able to operate effectively at both the normal day-to-day conditions as well as to accommodate any maximum ranges.

Basically, the two most important site-specific parameters that must be evaluated by the designer are the particle and gas stream characteristics:

• Dust Properties - These include particle size distribution, concentration and chemical composition. The particle size distribution is the most important factor that affects scrubber design and operation. However, particle size distribution data is rarely available for most sources and generally must be estimated from similar type sources. The average and maximum particle concentrations (or grain loading) must be obtained to properly size the scrubber and the solids removal system. Chemical composition of the dust particle is important to determine if the material will cause any plugging problems or precipitate problems.

• Exhaust Gas Characteristics - These include the average and the maximum flow rates, moisture content, and chemical composition The flow rates determine the volume of gas to be treated and therefore, the size of the scrubbing system. The moisture content and chemical composition are important in determining the potential corrosiveness of any liquid streams, pH levels, saturation conditions and spent liquid treatment and disposal requirements.

Design Evaluation of Particulate Wet Scrubbing Systems ___________________________________________________________________________________

2.0-7/98 10-3

Vendors utilize the above information as the basis for their proposed design and provide estimates or guarantees for the following important scrubber operating parameters.

• Static Pressure Drop - This is dependent on the desired removal efficiency and mechanical design of the scrubber system. Table 10-1 presents typical ranges for various wet scrubbers.

• Liquid Flow Rate - This parameter is based on the evaporation rate and type of scrubbing system utilized. Values need to be identified for both normal and maximum operating conditions. Also, if applicable, the recirculation rate and permissible levels of suspended solids in the recirculated liquid need to be identified. Table 10-1 lists typical ranges for various wet scrubbers.

• Collection Efficiency - The particle removal rate at both normal and maximum levels should be identified.

• Removal of Entrained Droplets - The type and efficiency of the mist removal system should be clearly stated.

Table 10-1. Ranges of pressure drops and liquid-to-gas (L/G) ratios for various wet scrubbers

Pressure drop, ∆p Liquid-to-gas ratio1

Scrubber kPa in. H2O L/m3 gal/1000 ft3

Venturi 1.5-25.0 5.0-100.0

0.4-5.0 3.0-40.0

Spray tower 0.12-0.75

0.5-3.0 0.7-2.7 5.0-20.0

Cyclonic spray 0.4-4.0 1.5-10.0 0.3-1.3 2.0-10.0 Moving bed (good for removing particulate and gaseous pollutants)

0.5-6.0 2.0-24.0 0.4-8.0 3.0-60.0

Orifice (self-induced spray) 0.5-4.0 2.0-10.0 0.07-0.7 0.5-5.0 Mechanically aided (fan) 1.0-2.0 4.0-8.0 0.07-0.5 0.5-4.0 1. Higher L/G reflects those used for gas absorption.

All scrubbers are capable of removing particles from a gas stream. Because of their ability to achieve high particle removal efficiencies and handle heavy grain loadings without plugging, venturi scrubbers are the most popular scrubber used to remove particulate matter. Venturis produce high particle to liquid droplet velocities in order to achieve good particle removal and therefore are limited in their ability to remove gases. The remainder of this lesson will provide an overview of theoretical equations (mainly venturi type systems) to predict scrubber efficiency along with examples of their use.

Lesson 10 ___________________________________________________________________________________

10-4 2.0-7/98

Estimating Collection Efficiency and Pressure Drop

A number of theories have been developed from basic particle-movement principles to explain the action of wet scrubbing systems. Many of these start from firm scientific concepts, but yield only qualitative results when predicting collection efficiencies or pressure drops. The interaction of particulate matter having a given particle-size distribution with water droplets having another size distribution is not easy to express in quantitative terms. As a result of this complexity, experimentally determined parameters are usually needed to approach reality.

Collection Efficiency

Collection efficiency is frequently expressed in terms of penetration. Penetration is defined as the fraction of particles (in the exhaust stream) that passes through the scrubber uncollected. Penetration is the opposite of the fraction of particles collected (i.e. collection efficiency), and is expressed as:

Pt = 1 − η (10−1)

Where: Pt = penetration η = collection efficiency

Wet scrubbers usually have an efficiency curve that fits the relationship of

( )η = − −1 e f system (10-2)

Where: η = collection efficiency e = exponential function f (system) = some function of the scrubbing system variables

By substituting for efficiency, penetration can be expressed as:

Pt = 1 − η (10-3)

( )( )

( ) =

−− = −

−

systemf

systemf

e

e11

An equation for the scrubbing system variables, f (system), can be developed for a particular scrubber design. A vendor can measure the operating variables and the collection efficiency of an existing or pilot scale unit. This information can then be used to evaluate the efficiency of the system. Scrubber vendors and various consultants have developed equations and assembled data that can be used to design and evaluate their specific scrubbers. Unfortunately, much of this information is proprietary. In addition, an equation that has been designed for a venturi scrubber may not work well for


2.0-7/98 10-5

evaluating the design of an orifice or cyclonic scrubber. In other words, there is not one specific equation that can be used to estimate the collection efficiency of every scrubbing system. A summary of equations used for predicting collection efficiency can be found in the Wet Scrubber System Study, Volume 1, Scrubber Handbook (Calvert et al. 1972). Theoretical penetration models estimate the penetration value as a function of particle size. This correlation can be applied to the particle size distribution of a proposed system to estimate overall collection efficiency.

Limitations in using these correlations include the following: (1) there are often very complex mathematical relationships involved, and (2) all the data inputs are either not readily available or non-existent and must be estimated. Below is an example of one of the more refined models for the venturi scrubber.

The Infinite Throat Model for Estimating Venturi Scrubber Efficiency One method for predicting particle collection efficiency in a venturi scrubber is the infinite-throat model (Yung et al. 1977). This model is a refined version of the Calvert correlation given in the Wet Scrubber System Study (Calvert et al. 1972). The equations presented in the infinite-throat model assume that all particles are captured by the water in the throat section of the venturi. Two studies found that this method correlated very well with actual venturi scrubber operating data (Yung et al. 1977 and Calvert et al. 1978).

A summary of the infinite throat model using metric units is presented in equations 10-4 through 10-12. Equation 10-4 is the actual equation which predicts the penetration (Pt) for one particle size (diameter). Equations 10-5 through 10-12 provide calculations for parameters that are used for determining particle penetration. As discussed later, to obtain an overall penetration ( Pt ), you must integrate over the entire particle-size distribution.

( )0.7K

0.7K

tanK0.715.02K4.24K

BdlnPtpo

po1

po

0.5popo

p +

+−+

−=

−

(10-4)

Where: Pt(dp) = penetration for one particle size B = parameter characterizing the liquid-to-gas ratio, dimensionless Kpo = inertial parameter at throat entrance, dimensionless

Note: Equation 10-4 was developed assuming that the venturi has an infinite-sized throat length. This is valid only when l in the following equation is greater than 2.0.

Lesson 10 ___________________________________________________________________________________

10-6 2.0-7/98

ll

=3

2t D g

d l

Cd

ρρ

Where: l = throat length parameter, dimensionless l t = venturi throat length, cm CD = drag coefficient for the liquid at the throat entrance, dimensionless ρg = gas density, g/cm3

dd = droplet diameter, cm ρl = liquid density, g/cm3

As you can see from Equation 10-4, two parameters, Kpo and B, must be found before calculating the particle penetration. Kpo, the inertial parameter at the throat entrance, is calculated in Equation 10-5.

Kd v

dpop gt

g d=

2

9µ (10-5)

Where: Kpo = inertial parameter at the throat entrance, dimensionless dp = particle aerodynamic resistance diameter, cmA* vgt = gas velocity in the throat, cm/s µg = gas viscosity, g/cm•s dd = droplet diameter, cm

* The “A” signifies that the diameter is an aerodynamic diameter instead of a physical diameter.

All the variables in Equations 10-5 can be measured empirically except for the droplet diameter (dd) which is calculated in the following equation known as the Nukiyama Tanasawa equation.

( )dv

L Gdgt

= +50 918 1 5. / . (10-6)

Where: dd = droplet diameter, cm vgt = gas velocity in the throat, cm/s L/G = liquid-to-gas ratio, dimensionless

Once the droplet diameter is calculated using empirically derived values for the gas velocity (at the throat) and the L/G ratio, the value for Kpo can be determined (in Equation 10-5).


2.0-7/98 10-7

The second variable in Equation 10-4, the parameter characterizing the liquid-to-gas ratio (B), can be calculated using Equation 10-7.

( )B L GC

l

g D

= /ρ

ρ (10-7)

Where: B = parameter characterizing liquid-to-gas ratio, dimensionless L/G = liquid-to-gas ratio, dimensionless ρl = liquid density, kg/m3 ρg = gas density, kg/m3 CD = drag coefficient for the liquid at the throat entrance, dimensionless

Values for L/G, liquid density, and gas density can be measured. The value for CD is calculated using Equation 10-8.

( )CND

oo= + +0 22 24 1 015. .

ReRe N0.6 (10-8)

Where: CD = drag coefficient for the liquid at the throat entrance, dimensionless NReo = Reynolds Number for the liquid droplet at the throat inlet, dimensionless

The Reynolds Number is determined in Equation 10-9.

Nv d

ogt d

gRe =

v (10-9)

Where: NReo = Reynolds Number for the liquid at the throat entrance, dimensionless vgt = gas velocity in the throat, cm/s dd = droplet diameter, cm vg = gas kinematic viscosity, cm2/s

Equation 10-9 requires a value for the droplet diameter (dd) which was determined earlier (see Equation 10-6). The gas kinematic viscosity (vg) is a variable that can be measured. Once, you have solved for the parameters Kpo and B, you can calculate the particle penetration by using Equation 10-4.

Other equations that are included with the infinite throat model are presented below. Depending on data availability, Kpg, the inertial parameter for mass-median diameter is used instead of Kpo. A method for using the parameters Kpg and B to estimate particle penetration will be shown later.

Kd v

dpgpg gt

g d=

2

9µ (10-10)

Lesson 10 ___________________________________________________________________________________

10-8 2.0-7/98

Where: Kpg = inertial parameter for mass-median diameter, dimensionless dpg = particle aerodynamic geometric mean diameter, cmA vgt = gas velocity in the throat, cm/s µg = gas viscosity, g/cm•s dd = droplet diameter, cm

Equation 10-10 is identical to Equation 10-5 which calculates Kpo except for the particle aerodynamic diameter used. Equation 10-5 uses the particle aerodynamic resistance diameter (dp) and Equation 10-10 uses the particle aerodynamic geometric mean diameter (dpg). The parameter, dpg, is calculated in Equation 10-11.

( ) 5.0pcpspg Cdd ρ×= (10-11)

Where: dpg = particle aerodynamic geometric mean diameter,µmA dps = particle physical, or Stokes, diameter, µm Cc = Cunningham slip correction factor,dimensionless ρp = particle density, g/cm3

The Cunningham slip correction factor, Cc, which is required for Equation 10-11 can be found by solving Equation 10-12.

( )pg

-4

c dT106.211C ×+= (10-12)

Where: Cc = Cunningham slip correction factor, dimensionless T = absolute temperature, K dps = particle physical, or Stokes, diameter, µm

With values for Cc and dpg, you can solve Equation 10-10 for Kpg, the inertial parameter for mass-median diameter.


2.0-7/98 10-9

The infinite throat model becomes more useful for air pollution applications when the overall penetration ( Pt ) for a particle-size distribution is calculated. To obtain an overall penetration ( Pt ), you must integrate over the entire particle-size distribution.

As an aid in calculating the overall penetration, Equations 10-4 (penetration for one particle size) through 10-12 were solved for the overall penetration assuming a log-normal particle-size distribution at various values of Kpg and B. These results are plotted in Figures 10-1(a), (b), and (c) (Yung et al. 1977).

In Figure 10-1, Pt , the overall penetration, is plotted versus B, a dimensionless parameter characterizing the liquid-to-gas ratio, and versus Kpg, a dimensionless inertial parameter for mass-median diameter. Each figure has been plotted for a different geometric standard deviation for particle size, i.e., 2.5, 5.0, and 7.5. Figure 10-1(a) (with a geometric standard deviation of 2.5) represents particles with a narrower size range than Figures 10-1(b) and 10-1(c) (with geometric standard deviations of 5.0 and 7.5 respectively).

These figures show that collection increases (penetration decreases) as the values for both B and Kpg increase. From Equations 10-7 and 10-10, the value of B increases as the liquid-to-gas ratio increases and the value of Kpg increases as the particle geometric mean diameter increases, assuming other parameters in the equations remain constant.

Focusing on Figure 10-1(a), let’s compare the particle collection of two applications: one with a Kpg = 0.5 and another with a Kpg = 50. As you can see, where the value for Kpg is 0.5 (top line), particle collection starts off low and improves slightly as the value for B increases. This supports what we already know, namely, that small particles (Kpg is 0.5) are difficult to capture and increasing the liquid-to-gas ratio only slightly enhances collection. Whereas for larger particles, (Kpg is 50), particle collection starts at a higher level and improves dramatically as the liquid-to-gas ratio increases.

In summary, by knowing the particle-size distribution of the dust from an industrial source and the operating conditions of the scrubber, the terms B and Kpg can be calculated and the collection efficiency (penetration) can be estimated using the appropriate figure [Figure 10-1(a), (b), or (c)].

Lesson 10 ___________________________________________________________________________________

10-10 2.0-7/98

Figure 10-1. Overall penetration, Pt , versus B with Kpg as a parameter, with different geometric standard deviations σgm Source: Yung et al. 1977.


2.0-7/98 10-11

Example 10-1 illustrates how to use the infinite-throat model to predict the performance of a venturi scrubber. When using the equations given in the model, make sure that the units for each equation are consistent.

Example 10-1 Cheeps Disposal Inc. is planning to install a hazardous-waste incinerator that will burn both liquid and solid waste materials. The exhaust gas from the incinerator will pass through a quench spray and then into a venturi scrubber and finally through a packed bed scrubber. Caustic will be added to the scrubbing liquor to remove any HCl from the flue gas and to control the pH of the scrubbing liquor. The uncontrolled particulate emissions leaving the incinerator are estimated to be 1,100 kg/h (maximum average). The local air pollution regulation states that particulate emissions must not exceed 10 kg/h. Using the following data, estimate the particulate collection efficiency of the venturi scrubber.

dps, mass-median particle size (physical) 9.0 µm σgm, geometric standard deviation 2.5 ρp, particle density 1.9 g/cm3 µg, gas viscosity 2.0 × 10-4 g/cm•s vg, gas kinematic viscosity 0.2 cm2/s ρg, gas density 1.0 kg/m3 QG, gas flow rate 15 m3/s vgt, gas velocity in venturi throat 9,000 cm/s Tg, gas temperature (in venturi) 80°C Tl, water temperature 30°C ρl, liquid density 1,000 kg/m3 QL, liquid flow rate 0.014 m3/s L/G, liquid-to-gas ratio 0.0009 L/m3

Solution

Figure 10-2 gives an overview of the solution presented here. As you can see from the diagram, you must solve many equations and make many calculations to obtain the collection efficiency of the scrubbing system. Equations in the early steps serve as inputs to the later ones.

Lesson 10 ___________________________________________________________________________________

10-12 2.0-7/98

Figure 10-2. Overview of steps for completing Example 10-1

��

η = −� ��

��

��

��

��

��

� �

=�

µ

��

��

��µ�

( )�� = ×� �� ρ

��

��

��

��

��ρ�

( )��

��

�= + + ��

� ��

��

��

��

( )�

�

��

= +× −

��

��

��

��

( )�� = +� ��

�

��

� ! �

��

��

��!

( )"�#

� �

$ � ! ρ

ρ

��"��! ��

�� %�

��! ��ρ��ρ�

��

��"��

� �

��

�� =

�

��!� ��"��#��

��

��$


2.0-7/98 10-13

1. Calculate the Cunningham slip correction factor, Cc, using Equation 10-12.

( )ps

4

c dT1021.61C

−×+=

Given: dps = 9.0 µm, the mass-median particle size (physical) Tg = 80 °C, the gas temperature

( )( )

024.1m 9

802731021.61C4

c

=µ

+×+=−

2. Calculate the particle aerodynamic geometric mean diameter, dpg, using Equation 10-11.

( ) 5.0pcpspg Cd d ρ×=

Given: dps = 9.0 µm, mass-median particle size (physical) ρp = 1.9 g/cm3, particle density

From step 1: Cc = 1.024

( )

cmA 10 12.6 =

mA 12.6

g/cm 1.9 1.024 m9 = d

4-

5.03pg

×

µ=

×µ

Note: Steps 1 and 2 above would not have been required if the particle diameter had been given as the aerodynamic geometric mean diameter, dpg,, and expressed in units of µmA.

3. Calculate the droplet diameter, dd, from Equation 10-6 (Nukiyama Tanasawa equation).

( )dv

L Gdgt

= +50 918 1 5. / .

Lesson 10 ___________________________________________________________________________________

10-14 2.0-7/98

Given: vgt = 9,000 cm/s, gas velocity in venturi throat L/G = 0.0009 L/m3

( )dd = +

=

509 000

91 0 0009

0 0080

1 5

,. .

.

.

cm / s8

cm

4. Calculate the inertial parameter for the mass-median diameter, Kpg, using Equation 10-10.

dg

gt2

pgpg d 9

vdK

µ=

Given: vgt = 9,000 cm/s, gas velocity in venturi throat µg = 2.0 × 10-4 g/cm•s, gas viscosity

From step 2: dpg = 12.6 × 10-4 cmA From step 3: dd = 0.008 cm

( ) ( )( )( )

Kpg =×

× •

=

−

−

12 6 10 9 000

9 2 0 10 0 008

992

4 2

4

. ,

. .

cmA cm / s

g / cm s cm

5. Calculate the Reynolds Number, NReo, using Equation 10-9.

Nv d

ogt d

gRe =

v

Given: vgt = 9,000 cm/s, gas velocity in venturi throat vg = 0.2 cm2/s, gas kinematic viscosity

From step 3: dd = 0.008 cm

( )

360 /scm 2.0

cm) (0.008 cm/s 9,000N 2eo

=

=R


2.0-7/98 10-15

6. Calculate the drag coefficient for the liquid at the throat entrance, CD, using Equation 10-8.

( )CND

oo= + +0 22 24 1 015. .

ReRe N0.6

From step 5: NReo = 360

( )( )CD = + +

=

0 22 24360

1 015 360

0 628

0.6. .

.

7. Calculate the parameter characterizing the liquid-to-gas ratio, B, using Equation 10-7.

( )Dg

l

CG/LB

ρρ=

Given: L/G = 0.0009 L/m3 ρl = 1,000 kg/m3 ρg = 1.0 kg/m3

( ) ( )( )

B m=

=

0 0009 1 00010 0 628

143

3. / ,. .

.

L kg / m kg / m

3

3

Lesson 10 ___________________________________________________________________________________

10-16 2.0-7/98

8. Determine the overall penetration, Pt , from Figure 10-3. The geometric standard deviation, σgm , is 2.5.

σgm = 2.5 From step 4: Kpg = 992, use the line for 1,000 From step 7: B = 1.43

In figure 10-3, read Pt = 0.008.

Figure 10-3. Overall penetration, Pt , for Example 10-1, where the standard deviation, σgm, is equal to 2.5


2.0-7/98 10-17

9. Calculate the collection efficiency using equation below.

η = 1 − Pt

From step 8: Pt = 0.008

η = 1 − 0.008 = 0.992 = 99.2%

10. Determine whether the local regulations for particulate emissions are being met. The local regulations state that the particulate emissions cannot exceed 10 kg/h. The required collection efficiency can be calculated by using the equation below.

in

outinrequired dust

dustdust −=η

Given: dustin = 1,100 kg/h, the dust concentration leading into theventuri dustout = 10 kg/h, the dust concentration leaving the venturi

%1.99

991.0kg/h 1100

kg/h 10kg/h 1100required

==

−=η

Note: Figure 10-1 can also be used to determine some of the required operating variables. This can be done by solving the example problem in reverse. By entering the figures at the required efficiency (or Pt ), one can obtain various sets of Kpg and B values. These values for B and Kpg can be used to calculate the required L/G ratio or gas velocity in venturi throat (vgt) for a specific collection efficiency.

To test your knowledge of the preceding section, answer the questions in Part 1 of the Review Exercise and work problem 1.

Lesson 10 ___________________________________________________________________________________

10-18 2.0-7/98

Contact Power Theory A more general theory for estimating collection efficiency is the contact power theory. This theory is based on a series of experimental observations made by Lapple and Kamack (1955). The fundamental assumption of the theory is:

"When compared at the same power consumption, all scrubbers give substantially the same degree of collection of a given dispersed dust, regardless of the mechanism involved and regardless of whether the pressure drop is obtained by high gas flow rates or high water flow rates." (Lapple and Kamack 1955)

In other words, collection efficiency is a function of how much power the scrubber uses, and not of how the scrubber is designed. This has a number of implications in the evaluation and selection of wet collectors. Once you know the amount of power needed to attain a certain collection efficiency, the claims about specially located nozzles, baffles, etc. can be evaluated more objectively. The choice between two different scrubbers with the same power requirements may depend primarily on ease of maintenance.

Semrau (1959 and 1963) developed the contact power theory from the work of Lapple and Kamack (1955). The theory, as developed by Semrau, is empirical in approach and relates the total pressure loss, PT, of the system to the collection efficiency.

The total pressure loss is expressed in terms of the power expended to inject the liquid into the scrubber plus the power needed to move the process gas through the system.

P P PT G L= + (10-13)

Where: PT = total contacting power, kWh/1,000 m3 (hp/1,000 acfm) PG = power input from gas stream, kWh/1,000 m3 (hp/1,000 acfm) PL = power input from liquid injection, kWh/1,000 m3 (hp/1,000 acfm)

Note: The total contacting power (or pressure loss), PT, should not be confused with penetration, Pt, defined in the previous section. Penetration is the symbol used by Calvert to express the fraction of particulate matter escaping from a collector.


2.0-7/98 10-19

The power expended in moving the gas through the system, PG, is expressed in terms of the scrubber pressure drop.

PG = (2.724 × 10-4)*∆p, kWh/1,000 m3 (metric units)

or (10-14)

PG = 0.1575*∆p, hp/1,000 acfm (English units)

Where: ∆p = pressure drop, kPa (in. H2O)

*Note: These values are based on gas density at standard (70°F and 1 atm) conditions; see derivation of equation in Richards 1983.

The power expended in the liquid stream, PL, is expressed as:

PL = 0.28 pL(QL/QG), kWh/1,000 m3 (metric units)

or (10-15)

PL = 0.583 pL(QL/QG), hp/1,000 acfm (English units)

Where: pL = liquid inlet pressure, 100 kPa (lb/in.2) QL = liquid feed rate, m3/h (gal/min) QG = gas flow rate, m3/h (ft3/min)

The constants given in the expressions for PG and PL incorporate conversion factors to put the terms on a consistent basis.

The total power can therefore be expressed as:

P P PT G L= +

PT = 2.724 × 10-4∆p + 0.28 pL(QL/QG), kWh/1,000 m3

or (10-16)

PT = 0.1575∆p + 0.583 pL(QL/QG), hp/1,000 acfm

The problem now is to correlate this with scrubber efficiency.

Equation 10-2 in this lesson shows that efficiency is an exponential function of the system variables for most types of collectors.

( )η = − −1 e f system

Lesson 10 ___________________________________________________________________________________

10-20 2.0-7/98

Semrau defines the function of the system variables, f (system), as:

f (system) = Nt = αPTβ (10-17)

Where: Nt = number of transfer units PT = total contacting power α and β = empirical constants which are determined from experiment and depend on the characteristics of the particles

The number of transfer units (Nt) is a concept that originated with plate towers. Plate towers have discrete separation stages. A plate tower with three plates has three separation stages or transfer units. Transfer units apply as well to packed towers, even though they have continuous (rather than discrete) separations. The number of transfer units is higher in systems where the pollutants are difficult to capture. Transfer units will be discussed in greater detail in Lesson 11.

Combining Equations 10-2 and 10-17, efficiency then becomes:

βα−−=η TPe1 (10-18)

Table 10-2 gives values of α and β for different industries. The values of α and β can be used in either the metric or English units.

Table 10-2. Parameters α and β for the contact power theory

Scrubber design Aerosol α β

Venturi Talc dust Phosphoric acid mist Foundry cupola dust Open-hearth steel furnace fume

Odorous mist

2.97 1.33 1.35 1.26

0.363

0.362 0.647 0.621 0.569

1.41 Venturi evaporator Hot black liquor gas 0.522 0.861 Venturi and cyclonic spray Lime kiln dust (raw)

Black liquor furnace fume Ferrosilicon furnace fume Lime kiln dust (prewashed)

Black liquor fume

1.47 1.75 0.870 0.915

0.740

1.05 0.620 0.459 1.05

0.861 Continued on next page


2.0-7/98 10-21

Table 10-2. (continued) Parameters α and β for the contact power theory

Scrubber design Aerosol α β

Venturi condensation scrubber with: 1. Mechanical spray

generation 2. Hydraulic nozzles

Copper sulfate

Copper sulfate

0.390

0.562

1.14

1.06 Orifice Talc dust 2.70 0.362 Cyclone Talc dust 1.16 0.655 Source: Semrau 1960.

The contact power theory cannot predict efficiency from a given particle-size distribution. The contact power theory gives a relationship which is independent of the size of the scrubber. With this observation, a small pilot scrubber could first be used to determine the pressure drop needed for the required collection efficiency. The full-scale scrubber design could then be scaled up from the pilot information.

Example 10-2 A wet scrubber has been proposed to control particulate emissions from a foundry cupola. Stack test results reveal that the particulate emissions must be reduced by 85% to meet emission standards. If a 100-acfm pilot unit is operated with a water flow rate of 0.5 gal/min at a water pressure of 80 psi, what pressure drop (∆p) would be needed across a 10,000-acfm scrubber unit?

Solution 1. From Table 10-2, read the α and β parameters for foundry cupola dust.

α = 1.35 β = 0.621

2. Calculate the number of transfer units, Nt, substituting Equation 10-17 into Equation 10-2.

η = − −1 e Nt

η−

=1

1lnNt

Given: η = 85%, collection efficiency

Lesson 10 ___________________________________________________________________________________

10-22 2.0-7/98

Nt =−

ln.

11 0 85

= ln 6.66 = 1.896

3. Calculate the total contacting power, PT, using Equation 10-18.

βα Tt P = N

From step 1: α = 1.35 β = 0.621

From step 2: Nt = 1.896

1.896 = 1.35 PT0.621

1.404 = PT0.621

ln 1.404 = 0.621 ln PT 0.3393 = 0.621 ln PT 0.5464 = ln PT PT = 1.73 hp/1,000 acfm

4. Calculate the pressure drop, ∆p, using Equation 10-16.

P QQT

L

G=

01575. p + 0.583 pL∆

Given: PL = 80 psi, liquid inlet pressure QL = 0.5 gal/min, liquid feed rate QG = 100 acfm, gas flow rate

From step 3: PT = 1.73 hp/1,000 acfm

( )1 73 01575 0 5

100. . .=

p + 0.583 80

p = 9.5 in. H2

∆

∆ O


2.0-7/98 10-23

From the data in Table 10-2, you can see that the usefulness of Equation 10-18 is limited due to the lack of α and β values. Also, the contact power theory does not apply to a number of new wet collecting systems where a combination of collecting mechanisms are used, such as condensation scrubbers. The theory applies best when the power is applied in one scrubbing area (McIlvaine 1977), such as in a venturi scrubber. Multiple-stages devices and packed towers will have collection efficiencies varying from those of a venturi scrubber for a given power input. However, the concept of the contact power theory is still a very useful tool in evaluating scrubber design.

Pressure Drop

As discussed earlier, a number of factors affect particle capture in a scrubber. One of the most important for many scrubber types is pressure drop. Pressure drop is the difference in pressure between the inlet and the outlet of the scrubbing process. It is the sum of the energy required to accelerate and move the gas stream and the frictional losses as the gases move through the scrubbing system.

The following factors affect the pressure drop in a scrubber:

• Scrubber design and geometry

• Gas velocity

• Liquid-to-gas ratio

As with calculating collection efficiency, no one equation can predict the pressure drop for all scrubbing systems.

Many theoretical and empirical relationships are available for estimating the pressure drop across a scrubber. Generally, the most accurate are those developed by scrubber manufacturers for their particular scrubbing systems. Due to the lack of validated models, it is recommended that users consult the vendor's literature to estimate pressure drop for the particular scrubbing device of concern.

One expression was developed for venturis and is widely accepted. The correlation proposed by Calvert (Yung et al. 1977) is:

∆p = 8.24 × 10-4 (vgt)2 (L/G) (metric units)

or (10-19)

∆p = 4.0 × 10-5 (vgt)2 (L/G) (English units)

Where: ∆p = pressure drop, cm H2O (in. H2O) vgt = velocity of gas in the venturi throat, cm/s (ft/sec) L/G = liquid-to-gas ratio, dimensionless, L/m3 (gal/1000 ft3)

Lesson 10 ___________________________________________________________________________________

10-24 2.0-7/98

Using Equation 10-19 to calculate the pressure drop for the conditions given in Example 10-1, we get the following:

Given: vgt = 9,000 cm/s L/G = 0.0009 L/m3 ∆p = 8.24 × 10-4 (9,000)2 (0.0009) = 60 cm H2O (or 24 in. H2O)

Using Pilot Methods to Design Scrubbers

The semi-empirical theories previously discussed are useful for scrubber design and evaluation exercises because they can give qualitatively correct information. However, they have a number of practical limitations. It is not common practice to choose scrubber systems based only on this information. The uncertainties involved in particle-size determinations and the questions associated with using empirically determined parameters restrict the use of theoretical methods. Basically, too many variables are involved and accounting for them all in a simple theory is too difficult. The time and expense needed to obtain good input data for these methods may be better spent in developing pilot plant information.

Scrubbers that work primarily through impaction mechanisms have certain performance characteristics (such as efficiency and pressure drop) which are independent of scale. This consequence of the contact power principle provides the basis for using pilot systems. By using a small-scale scrubber (100 to 1,000 cfm) on the exhaust gas stream, the effectiveness of the equipment for removing the actual particles in the gas can be experimentally determined.

Pilot systems ranging from 170 m3/h (100 cfm) units to one-tenth the size of full-scale plants have been developed in the past. McIlvaine (1977) has compared the effectiveness of the various design methods. His work is summarized in Table 10-3.

Table 10-3. Methods for predicting venturi scrubber pressure requirements

Description

Expense (relative scale)

Time (months)

Most reliable

↓ ↓ ↓ ↓ ↓

Least reliable

1/10 size full-scale plants 2000-cfm pilot units 100-cfm pilot units Empirical curves based on similar processes

Impactor in situ particle sizing

100-1,000

30 5

0.2 2

12-24 3-6 2-3 0.2

1


2.0-7/98 10-25

The design of a wet collector system for a particulate-emission problem requires more than the application of a few design equations. The experience of scrubber manufacturers with specific industry installations, coupled with the use of pilot units, provides more reliable ways to determine the size of a system for a wide range of operating conditions. In many cases, theoretical models can complement such studies and provide qualitative data for wet collector evaluations.

Summary

When reviewing design plans for a proposed new wet scrubbing system, the most useful information is operating data from an installation on similar sources. There are theoretical relationships that can be used to estimate scrubber performance; however, they are specific to the physical design of one scrubbing system and often all the needed inputs are not available. Therefore, an evaluation of wet scrubber design plans should involve utilizing both theoretical relationships and operating information from similar sources to assure that the proposed system design can achieve the desired control efficiency and addresses potential operating problems.

There are a number of parameters that affect particle removal efficiency and must be considered in the design of a wet scrubbing system; they are the following:

• Dust properties (particle size distribution being most important)


• Static pressure drop

• Scrubber liquid flow rate

• Required particle removal efficiencies

• Removing entrained liquid droplets

The infinite throat model (for venturis only) and the contact power are two methods used to estimate scrubber performance that were discussed in this lesson. The infinite throat model correlates with operating data but is applicable only to venturi scrubbers. The contact power theory is applicable to various scrubber designs, but must have pilot plant data to predict efficiency.

To test your knowledge of the preceding section, answer the questions in Part 2 of the Review Exercise and work problems 2 and 3.

Lesson 10 ___________________________________________________________________________________

10-26 2.0-7/98


2.0-7/98 10-27

Review Exercise Questions

Part 1

1. Which approach(es) can be used to evaluate the capabilities of scrubbing systems?

a. Empirical relationships b. Theoretical models c. Pilot scale test data d. a and b, only e. a, b, and c

2. Two important parameters in the design and operation of wet scrubbing systems that are a function of the process being controlled are:

a. Static pressure drop and collection efficiency b. L/G ratio and pressure drop c. Dust properties and exhaust gas characteristics d. Liquid flow rate and L/G ratio

3. True or False? Particle size distribution is the most critical parameter in choosing the most effective scrubber design and determining the overall collection efficiency.

4. Static pressure drop of a system is dependent on the:

a. Mechanical design of the system b. Collection efficiency required c. Size of the system d. a and b, only

5. The scrubber used most often to remove particulate matter from exhaust streams is a ____________________ scrubber.

6. The term penetration is defined as:

a. The fraction of particles collected in a scrubber b. The amount of gaseous pollutants absorbed in the scrubbing liquor c. The fraction of particles that passes through a scrubber uncollected

7. True or False? There is no one simple equation that can be used to estimate scrubber collection efficiency for all scrubber types.

8. True or False? Efficient particle removal requires low gas-to-liquid (relative) velocities.

9. A model used to estimate particle collection in venturi scrubbers is:

a. The infinite-throat model b. The penetration model c. The short-stack model d. The impaction model

Lesson 10 ___________________________________________________________________________________

10-28 2.0-7/98

Part 2

10. The contact power theory is dependent on ____________________ data to determine required collection efficiency:

a. Process b. Pilot test c. Theoretical d. Fan curve

11. In the equation used in the contact power theory, PT = PG + PL, the symbol PT represents:

a. The penetration of the system b. The collection efficiency c. The total pressure loss, or contacting power, of the scrubbing system

12. According to the contact power theory, the ____________________ the pressure drop is across the scrubbing system, the higher the collection efficiency will be.

a. Lower b. Higher

13. Which of the following factors affect the pressure drop of a scrubbing system?

a. Scrubber design and geometry b. Gas velocity c. Liquid-to-gas ratio d. All of the above


2.0-7/98 10-29

Problem 1

A company has submitted an application to increase production of its lime kiln by 20%. The kiln is currently controlled by a venturi scrubber that the applicant feels is capable of handling the added gas volume and dust loading. The system has a quench chamber that is capable of cooling the extra gas volume. Given the information below, calculate the new collection efficiency.

Existing Proposed

dps , mass-median particle size (physical)

8.0 µm 8.0 µm

σgm, geometric standard deviation of particle distribution

2.5 2.5

ρp, particle density 1.7 g/cm3 1.7 g/cm3

µg, gas viscosity 2.0×10-4 g/cm·s 2.0×10-4 g/cm·s vg, gas kinematic viscosity 0.2 cm2/s 0.2 cm2/s ρg, gas density 1.0 kg/m3 1.0 kg/m3 QG, gas flow rate 18 m3/s 22.5 m3/s vgt, gas velocity in venturi throat 85,000 cm/s 10,625 cm/s Tg, gas temperature (in venturi) 80oC 80oC Tl, water temperature 30oC 30oC ρl, liquid density 1,000 kg/m3 1,000 kg/m3 QL, liquid flow rate 0.016 L/s 0.016 L/s L/G, liquid-to-gas ratio 0.00089 0.00071 dust loading 455 kg/hr 545 kg/hr efficiency 98.8% calculate

Problem 1: Student Worksheet (This space is provided for you to work problem 1)

Lesson 10 ___________________________________________________________________________________

10-30 2.0-7/98

Problem 1: Student Worksheet (cont’d) (This space is provided for you to work problem 1)


2.0-7/98 10-31


Lesson 10 ___________________________________________________________________________________

10-32 2.0-7/98



2.0-7/98 10-33

Problem 2

A wet scrubber is used to control dust emissions from a foundry. The system design and test data is summarized below. Due to new air quality requirements, the source will be required to control particulate to a removal efficiency of 95%. What would be the new pressure drop to attain 95% removal if no other operational changes were made?

Operating test data

∆p, pressure drop 9.0 in. of H2O QL, liquid feed rate 150 gal/min QG, gas flow rate 22,000 acfm pL, water pressure 90 psi


Lesson 10 ___________________________________________________________________________________

10-34 2.0-7/98



2.0-7/98 10-35

Problem 3

A company proposes to increase the production rate of its lime kiln by 20%. Calculate the increase in pressure drop that will result from the new operating conditions.

Existing Proposed

vgt, gas velocity at the throat 8,500 cm/s 10,625 cm/s L/G (dimensionless in metric units) 0.00089 0.00071


Lesson 10 ___________________________________________________________________________________

10-36 2.0-7/98



2.0-7/98 10-37

Review Exercise Answers Answers to Questions

Part 1

1. e. a, b, and c The following approaches can be used to evaluate the capabilities of scrubbing systems: • Empirical relationships • Theoretical models • Pilot scale test data

2. c. Dust properties and exhaust gas characteristics Two important parameters in the design and operation of wet scrubbing systems that are a function of the process being controlled are dust properties and exhaust gas characteristics.

3. True Particle size distribution is the most critical parameter in choosing the most effective scrubber design and determining the overall collection efficiency.

4. d. a and b, only Static pressure drop of a system is dependent on the:

• Mechanical design of the system • Collection efficiency required

5. Venturi The scrubber used most often to remove particulate matter from exhaust streams is a venturi scrubber.

6. c. The fraction of particles that passes through a scrubber uncollected The term penetration is defined as the fraction of particles that passes through a scrubber uncollected.

7. True There is no one simple equation that can be used to estimate scrubber collection efficiency for all scrubber types.

8. False Efficient particle removal requires high gas-to-liquid (relative) velocities.

9. a. The infinite-throat model The infinite-throat model is used to estimate particle collection in venturi scrubbers.

Part 2

10. b. Pilot test The contact power theory is dependent on pilot test data to determine required collection efficiency.

Lesson 10 ___________________________________________________________________________________

10-38 2.0-7/98

11. c. The total pressure loss, or contacting power, of the scrubbing system In the equation used in the contact power theory, PT = PG + PL, the symbol PT represents the total pressure loss, or contacting power, of the scrubbing system.

12. b. Higher According to the contact power theory, the higher the pressure drop is across the scrubbing system, the higher the collection efficiency will be.

13. d. All of the above The following factors affect the pressure drop of a scrubbing system: • Scrubber design and geometry • Gas velocity • Liquid-to-gas ratio


2.0-7/98 10-39

Solution to Problem 1

Answer: The collection efficiency of the venturi scrubber under the new scenario is 97.5%.

Solution:

1. Calculate the particle aerodynamic geometric mean diameter, dpg. Since the mass-median particle size, dps, is given, first calculate the Cunningham slip correction factor, Cc, using Equation 10-12.

ps

-4

c d)T10(6.21+1=C ×

Given: dps = 8.0 µm, mass-median particle size (physical) T = 80°C, gas temperature

880)+(273 )10(6.21+1 = C

-4

c×

= 1.027

Now, calculate dpg using Equation 10-11.

dpg = dps (Cc × ρp)0.5

Given: ρp = 1.7g/cm3, particle density

dpg = 8.0 µm (1.027 × 1.7 g/cm3)0.5 = 10.57 µmA = 10.57 × 10-4 cmA

2. Calculate the droplet diameter, dd, from Equation 10-6 (Nukiyama Tanasawa equation).

dd = 50/vgt + 91.8 (L/G)1.5

Given: vgt = 10,625 cm/s, gas velocity in venturi throat L/G = 0.00071

1.5d (0.00071) 91.8 +

cm/s 625,1050d =

= 0.00644 cm

Lesson 10 ___________________________________________________________________________________

10-40 2.0-7/98

3. Calculate the inertial parameter for the mass-median diameter, Kpg, using Equation 10-10.

Kd v

dpgpg gt

g d=

2

9µ

Given: vgt = 10,625 cm/s, gas velocity in venturi throat µg = 2.0 × 10-4 g/cm•s, gas viscosity

From step 1: dpg = 10.57 × 10-4 cmA From step 2: dd = 0.00644 cm

Kpg = ×× •

10.57 10 cmA) (10,625 cm / s)9 (2.0 10 g / cm s) (0.00644 cm)

-4 2

-4

=1,024

4. Calculate the Reynolds Number, NReo, using Equation 10-9.

Nv d

ogt d

gRe =

v

Given: vgt = 10,625 cm/s, gas velocity in venturi throat vg = 0.2 cm2/s, gas kinematic viscisity

From step 2: dd = 0.00644 cm

N oRe = (10,625 cm / s)(0.00644 cm)0.2 cm / s2

= 342


2.0-7/98 10-41

5. Calculate the drag coefficient for the liquid at the throat entrance, CD, using Equation 10-8.

( )CND

o= + +0 22 24 1 015. .

Re NReo

0.6

From step 4: NReo = 342

( )( )CND

o= + +0 22 24 1 015. .

Re 342 0.6

= 0.639

6. Calculate the parameter characterizing the liquid-to-gas, B, using Equation 10-7.

( )B L GC

l

g D

= /ρ

ρ

Given: L/G = 0.00071 ρl = 1,000 kg/m3, liquid density ρg = 1.0 kg/m3, gas density

From step 5: CD = 0.639

( ) ( )( )B

mm

= 0 000711 000

10 0 639

3

3.

, /. / .

kg kg

= 1.11

7. Find the overall penetration, Pt , using Figure 10-1(a). The geometric standard deviation, σgm, is 2.5.

From step 3: Kpg = 1,024 From step 6: B = 1.11

Read Pt = 0.025 (Note: you have to estimate where the 1,024 line would be.)

Lesson 10 ___________________________________________________________________________________

10-42 2.0-7/98

8. Calculate the collection efficiency using the equation below.

η = 1 − Pt

From step 7: Pt = 0.025

η = 1.0 − 0.025 = 0.975 = 97.5%


2.0-7/98 10-43


Answer: The new pressure drop to attain 95% particle removal is 21 in. of water.

Solution:

1. Obtain values for α and β for foundry cupola dust from Table 10-2.

α = 1.35

β = 0.621

2. Calculate the number of transfer units, Nt, using Equation 10-18.

η = − −1 e Nt

η−

=1

1lnNt

Given: η = 95%, collection efficiency

Nt =−

ln.

11 0 95

= ln 20

= 3.0 transfer units

3. Calculate the total contacting power (PT) required.

Nt = α PTβ

From step 1: α = 1.35 β = 0.621

From step 2: Nt = 3.0

3.0 = 1.35 PT0.621

PT0.621 = 3.0/1.35

PT0.621 = 2.22

Lesson 10 ___________________________________________________________________________________

10-44 2.0-7/98

0.621 ln PT = ln 2.22

ln PT = 1.28

PT = 3.61 hp/1,000 acfm

4. Calculate the pressure drop for the given operating conditions using Equation 10-16.

PT = 0.1575∆p + 0.583 pL(QL/QG), hp/1,000 acfm

Given: pL = 90 psi, water pressure QL = 150 gal/min, liquid feed rate QG = 22,000 acfm, gas flow rate

From step 3: PT = 3.61 hp/1,000 acfm

3.61 = 0.1575 ∆p + 0.583 (90) (150/22,000) 3.61 = 0.1575 ∆p + 0.358 ∆p = 21 in. of water


2.0-7/98 10-45


Answer: At the new operating conditions, the pressure drop will increase 13 cm (5 in.) of water.

Solution:

1. Solve for the existing pressure drop using Equation 10-19.

∆p = 8.24 × 10-4 (vgt)2 (L/G) (for metric units)

Given: vgt = 8,500 cm/s, existing gas velocity at throat L/G = 0.00089, existing liquid-to-gas ratio

∆p = 8.24 × 10-4 (8,500)2 (0.00089) ∆p = 53 cm (or 21 in.) of water

2. Solve for new pressure drop using Equation 10-19.

∆p = 8.24 × 10-4 (vgt)2 (L/G)

Given: vgt = 10,625 cm/s, proposed gas velocity at throat L/G = 0.00071, proposed liquid-to-gas ratio

∆p = 8.24 × 10-4 (10,625)2 (0.00071) ∆p = 66 cm (or 26 in.) of water

3. Solve for the increase in pressure drop at the new operating conditions.

new ∆p − old ∆p = increase in ∆p In metric units: 66 cm − 53 cm = 13 cm of water In English units: 26 in. − 21 in. = 5 in. of water

Lesson 10 ___________________________________________________________________________________

10-46 2.0-7/98


2.0-7/98 10-47

Bibliography

Brady, J. D., and L. K. Legatski. 1977. Venturi scrubbers. In P. N. Cheremisinoff and R. A. Young (Eds.), Air Pollution Control and Design Handbook. Part 2. New York: Marcel Dekker.

Calvert, S. 1977. How to choose a particulate scrubber. Chemical Engineering. 84:133-140.


Kashdan, E. R., and M. B. Ranada. 1979. Design Guidelines for an Optimum Scrubber System. EPA 600/7-79-018. U.S. Environmental Protection Agency.

Lapple, C. E., and H. J. Kamack. 1955. Performance of wet dust scrubbers. Chemical Engineering Progress. 51:110-121.

McIlvaine, R. W. 1977, June. When to pilot and when to use theoretical predictions of required venturi pressure drop. Paper presented at the meeting of the Air Pollution Control Association. Toronto, Canada.

Nukiyama, S., and Y. Tanasawa. 1983. An experiment on atomization of liquid by means of air stream (in Japanese). Transactions. Society of Mechanical Engineers. Japan. 4:86.


Richards, J. R. 1983, September. Wet Scrubber Inspection and Evaluation Manual, EPA 340/1-83-022. U.S. Environmental Protection Agency.


Rimberg, D. B. 1979, March. Tips and techniques on air pollution control equipment O & M. Pollution Engineering. (pp. 32-35).

Rimberg, D., and Y. M. Peng. 1977. Aerosol collection by falling droplets. In P. N. Cheremisinoff and R. A. Young (Eds.), Air Pollution Control and Design Handbook. (pp. 747-777). New York: Marcel Dekker.

Semrau, K. T. 1960. Correlation of dust scrubber efficiency. Journal of the Air Pollution Control Association. 10:200-207.

Semrau, K. T. 1963. Dust scrubber design - a critique on the state of the art. Journal of the Air Pollution Control Association. 13:587-593.


Lesson 10 ___________________________________________________________________________________

10-48 2.0-7/98

Sparks, L. E. 1978. SR-52 Programmable Calculator Programs for Venturi Scrubbers and Electrostatic Precipitators. EPA 600/7-78-026. U.S. Environmental Protection Agency. Research Triangle Park, NC.

Stairmand, C. J. 1956. The design and performance of modern gas-cleaning equipment. Journal. Institute of Fuel. 29:58-81.

Strauss, W. 1975. Industrial Gas Cleaning. Oxford: Pergamon Press.


Yung, S., S. Calvert, and J. F. Barbarika. 1977. Venturi Scrubber Performance Model. EPA 600/2-77-172. U.S. Environmental Protection Agency. Cincinnati, OH.

2.0-7/98 11-1

Lesson 11 Design Review of Absorbers Used for Gaseous Pollutants

Goal To familiarize you with the factors to be considered when reviewing absorber design plans for the permit process.


1. Explain the importance of the following factors in absorber design:


• Liquid flow

• Pressure drop

• pH

• Removal of entrained liquids

2. Estimate the liquid flow rate, the diameter, and the packing height of a packed tower using appropriate tables and equations

3. Estimate the number of plates and the height of a plate tower using appropriate tables and equations

Introduction

Gas absorbers are most often used to remove soluble inorganic contaminants from an air stream. The design of an absorber used to reduce gaseous pollutants from process exhaust streams involves many factors including the pollutant collection efficiency, pollutant solubility in the absorbing liquid, liquid-to-gas ratio, exhaust flow rate, pressure drop, and many construction details of the absorbers such as packing, plates, liquid distributors, entrainment separators, and corrosion-resistant materials. These have been discussed in detail in the previous lessons.

Lesson 11 ___________________________________________________________________________________

11-2 2.0-7/98

The same three basic review approaches discussed for particle removal are applicable for gas absorber evaluation:

1. Empirical relationships based on historical data

2. Theoretical principles based on gas chemistry and physics

3. Pilot scale test data

The theoretical relationships for gas absorption have been well defined over the many years that gas absorption has been studied; however, they can be very complex and are dependent on the mechanical design of the scrubber. As with particulate scrubbers, empirical relationships and general rules of thumb are often used to evaluate absorber designs and there is no one easy set of equations to evaluate the design of all absorbers.

All wet scrubbing systems are able to collect both particulate and gaseous pollutants emitted from process exhaust streams. However, spray towers, plate towers, packed towers, and moving-bed scrubbers are most often used for gaseous pollutant removal. This lesson will focus on equations used to estimate liquid flow rate, the diameter and the height of a packed tower, and the diameter and number of plates used in a plate tower to achieve a specified pollutant removal efficiency.

In evaluating an absorption system, the reviewer can use the equations in this lesson to estimate critical operating parameters or component sizes, then supplement this information with operating information on the particular scrubber type from previous lessons to complete the review process.

Review of Design Criteria

The principal design criteria are the exhaust flow rate to the absorber, measured in units of m3/min (ft3/min, or acfm), and the gaseous pollutant concentration, measured in units of parts per million (ppm). The exhaust volume and pollutant concentration are set by the process exhaust conditions. Once these criteria are known, the vendor can begin to design the absorber for the specific application. A thorough review of the design plans should consider the factors presented below.

Exhaust gas characteristics - average and maximum flow rates to the absorber, and chemical properties such as dew point, corrosiveness, pH, and solubility of the pollutant to be removed should be measured or accurately estimated.

Liquid flow - the type of scrubbing liquid and the rate at which the liquid is supplied to the absorber. If the scrubbing liquid is to be recirculated, the pH and amount of suspended solids (if any) should be monitored to ensure continuous reliability of the absorbing system.

Pressure drop - the pressure drop (gas-side) at which the absorber will operate; the absorber design should also include a means for monitoring the pressure drop across the system, usually by manometers.

pH - the pH at which the absorber will operate; the pH of the absorber should be monitored so that the acidity or alkalinity of the absorbing liquor can be properly adjusted.

Design Review of Absorbers Used for Gaseous Pollutants ___________________________________________________________________________________

2.0-7/98 11-3

Removal of entrained liquid - mists and liquid droplets that become entrained in the "scrubbed" exhaust stream should be removed before exiting the stack. Some type of entrainment separator, or mist eliminator, should be included in the design.

Emission requirements - collection efficiency in terms of parts per million to meet the air pollution regulations; collection efficiency can be high (90 to 99%) if the absorber is properly designed. The agency review engineer can use the equations listed in this lesson to estimate the absorber removal efficiency, liquid flow rate, tower diameter, and packing height. However, these equations can only estimate these values, and they should not be used as the basis to either accept or reject the design plans submitted for the permit process.

Absorption

Absorption is a process that refers to the transfer of a gaseous pollutant from a gas phase to a liquid phase. More specifically, in air pollution control, absorption involves the removal of objectionable gaseous pollutants from a process stream by dissolving them in a liquid.

The absorption process can be categorized as physical or chemical. Physical absorption occurs when the absorbed compound dissolves in the liquid; chemical absorption occurs when the absorbed compound and the liquid (or a reagent in the liquid) react. Liquids commonly used as solvents include water, mineral oils, nonvolatile hydrocarbon oils, and aqueous solutions.

Some common terms used when discussing the absorption process follow:

Absorbent - the liquid, usually water, into which the pollutant is absorbed.

Solute, or absorbate - the gaseous pollutant being absorbed, such as SO2, H2S, etc.

Carrier gas - the inert portion of the gas stream, usually air, from which the pollutant is being removed.

Interface - the area where the gas phase and the absorbent contact each other.

Solubility - the capability of a particular gas to be dissolved in a given liquid.

Absorption is a mass-transfer operation. In absorption, mass transfer of the gaseous pollutant into the liquid occurs as a result of a concentration difference (of the pollutant) between the liquid and gas phases. Absorption continues as long as a concentration difference exists where the gaseous pollutant and liquid are not in equilibrium with each other. The concentration difference depends on the solubility of the gaseous pollutant in the liquid.

Lesson 11 ___________________________________________________________________________________

11-4 2.0-7/98

Absorbers remove gaseous pollutants by dissolving them into a liquid called the absorbent. In designing absorbers, optimum absorption efficiency can be achieved by doing the following:

• Providing a large interfacial contact area

• Providing for good mixing between the gas and liquid phases

• Allowing sufficient residence, or contact, time between the phases

• Choosing a liquid in which the gaseous pollutant is very soluble

Solubility

Solubility is a very important factor affecting the amount of a pollutant, or solute, that can be absorbed. Solubility is a function of both the temperature and, to a lesser extent, the pressure of the system. As temperature increases, the amount of gas that can be absorbed by a liquid decreases. From the ideal gas law: as temperature increases, the volume of a gas also increases; therefore, at the higher temperatures, less gas is absorbed due its larger volume. Pressure affects the solubility of a gas in the opposite manner. By increasing the pressure of a system, the amount of gas absorbed generally increases.

The solubility of a specific gas in a given liquid is defined at a designated temperature and pressure. Table 11-1 presents data on the solubility of SO2 gas in water at 101 kPa, or 1 atm, and various temperatures. In determining solubility data, the partial pressure (in mm Hg) is measured with the concentration (in grams of solute per 100 grams of liquid) of the solute in the liquid. The data in Table 11-1 were taken from The International Critical Tables, a good source of information concerning gas-liquid systems.

Table 11-1. Partial pressure of SO2 in aqueous solution, mm Hg

Grams of SO2 per

100g H2O

10°C

20°C

30°C

40°C

50°C

60°C

70°C

0.0 - - - - - - - 0.5 21 29 42 60 83 111 144 1.0 42 59 85 120 164 217 281 1.5 64 90 129 181 247 328 426 2.0 86 123 176 245 333 444 581 2.5 108 157 224 311 421 562 739 3.0 130 191 273 378 511 682 897 3.5 153 227 324 447 603 804 - 4.0 176 264 376 518 698 - - 4.5 199 300 428 588 793 - - 5.0 223 338 482 661 - - -


2.0-7/98 11-5

Solubility data are obtained at equilibrium conditions. This involves putting measured amounts of a gas and a liquid into a closed vessel and allowing it to sit for a period of time. Eventually, the amount of gas absorbed into the liquid will equal the amount coming out of the solution. At this point, there is no net transfer of mass to either phase, and the concentration of the gas in both the gaseous and liquid phases remains constant. The gas-liquid system is at equilibrium.

Equilibrium conditions are important in operating an absorption tower. If equilibrium were to be reached in the actual operation of an absorption tower, the collection efficiency would fall to zero at that point since no net mass transfer could occur. The equilibrium concentration, therefore, limits the amount of solute that can be removed by absorption. The most common method of analyzing solubility data is to use an equilibrium diagram. An equilibrium diagram is a plot of the mole fraction of solute in the liquid phase, denoted as x, versus the mole fraction of solute in the gas phase, denoted as y. (See Appendix A for a brief refresher on mole fractions.) Equilibrium lines for the SO2 and water system given in Table 11-1 are plotted in Figure 11-1. Figure 11-1 also illustrates the temperature dependence of the absorption process. At a constant mole fraction of solute in the gas (y), the mole fraction of SO2 that can be absorbed in the liquid (x) increases as the temperature decreases.

Figure 11-1. Equilibrium lines for SO2 - H2O systems at various temperatures

Under certain conditions, Henry's law may also be used to express equilibrium solubility of gas-liquid systems. Henry's law is expressed as:

p = Hx (11-1)

Where: p = partial pressure of solute at equilibrium, Pa x = mole fraction of solute in the liquid H = Henry's law constant, Pa/mole fraction

Lesson 11 ___________________________________________________________________________________

11-6 2.0-7/98

From Equation 11-1, you can see that H has the units of pressure per concentration. Henry's law can be written in a more useful form by dividing both sides of Equation 11-1 by the total pressure, PT, of the system. The left side of the equation becomes the partial pressure divided by the total pressure, which equals the mole fraction in the gas phase, y. Equation 11-1 now becomes:

y = H'x (11-2)

Where: y = mole fraction of gas in equilibrium with liquid H' = Henry's law constant, mole fraction in vapor per mole fraction in liquid x = mole fraction of the solute in equilibrium

Note: H' now depends on the total pressure.

Equation 11-2 is the equation of a straight line, where the slope (m) is equal to H'. Henry's law can be used to predict solubility only when the equilibrium line is straight. Equilibrium lines are usually straight when the solute concentrations are very dilute. In air pollution control applications, this is usually the case. For example, an exhaust stream that contains a 1,000-ppm SO2 concentration corresponds to a mole fraction of SO2 in the gas phase of only 0.001. Figure 11-2 demonstrates that the equilibrium lines are still straight at this low concentration of SO2.

Figure 11-2. Equilibrium diagram for SO2 - H2O system for the data given in Example 11-1

Another restriction on using Henry's law is that it does not hold true for gases that react or dissociate upon dissolution. If this happens, the gas no longer exists as a simple molecule. For example, scrubbing HF or HCl gases with water causes both compounds to dissociate in solution. In these cases, the equilibrium lines are curved rather than straight. Data on systems that exhibit curved equilibrium lines must be obtained from experiments.


2.0-7/98 11-7

Henry's law constants for the solubility of several gases in water are listed in Table 11-2. The units of Henry's law constants are atmospheres per mole fraction. The smaller the constant, the more soluble the gas. Table 11-2 demonstrates that SO2 is approximately 100 times more soluble in water than CO2 is.

Table 11-2. Henry's law constants for gases in H2O1

Gas 20°C 30°C

N2 80.4 92.4

CO 53.6 62.0 H2S 48.3 60.9

O2 40.1 47.5

NO 26.4 31.0 CO2 1.42 1.86

SO2 0.014 0.016

1. Expressed in H × 10-5, atm/mole fraction.

The following example illustrates how to develop an equilibrium diagram from solubility data.

Example 11-1

Given the data in Table 11-3 for the solubility of SO2 in pure water at 303°K (30°C) and 101.3 kPa (760 mm Hg), calculate y and x, plot the equilibrium diagram, and determine if Henry's law applies.

Table 11-3. Equilibrium data

cso2

(g of SO2 per 100 g of H2O)

pso2

(partial pressure of SO2)

y (mole fraction of

SO2 in gas phase)

x (mole fraction of SO2 in

liquid phase)

0.5 6 kPa (42 mm Hg) 1.0 11.6 kPa (85 mm

Hg)

1.5 18.3 kPa (129 mm Hg)

2.0 24.3 kPa (176 mm Hg)

2.5 30.0 kPa (224 mm Hg)

3.0 36.4 kPa (273 mm Hg)

Lesson 11 ___________________________________________________________________________________

11-8 2.0-7/98

Solution In steps 1 and 2, convert the data for the concentration of SO2 in water and the partial pressure of SO2 in air into mole fraction units.

1. Calculate the mole fraction of SO2 in the gas phase, y, by dividing the partial pressure of SO2 by the total pressure of the system.

ypP

y

so

T=

=

=

2

6

0 06

kPa101.3 kPa.

The mole fractions of SO2 in the gas phase (y) are tabulated in Table 11-4.

2. Calculate the mole fraction of the solute (SO2) in the liquid phase, x, by dividing the moles of SO2 dissolved into the solution by the total moles of liquid.

xO

= moles of SO in solutionmoles of SO in solution + moles of H

2

2 2

Where: moles of SO2 in solution = cSO2/64 g SO2 per mole

moles of H2O = 100 g of H2O/18 g H2O per mole = 5.55 moles

x =cso2

// .

.

. .

.

6464 555

0 564

0 564

555

0 0014

2cso +

=+

=

The mole fractions of the solute in the liquid phase are tabulated in Table 11-4.


2.0-7/98 11-9

Table 11-4. Equilibrium data for Example 11-1

OH g 100

SO of g

2

2

SO = c

2

2SOp

(kPa)

y = p/101.3 x =

c / 64c

so

so

2

2/ .64 5 55+

0.5 6.0 0.060 0.0014 1.0 11.6 0.115 0.0028 1.5 18.3 0.180 0.0042 2.0 24.3 0.239 0.0056 2.5 30.0 0.298 0.0070 3.0 36.4 0.359 0.0084

3. Plot the mole fraction of SO2 in air, (y), against the mole fraction of SO2 dissolved in water, (x).

Figure 11-2. (repeated) Equilibrium diagram for SO2 - H2O system for the data given in Example 11-1

The plot in Figure 11-2 is a straight line; therefore, Henry's law applies.

7.420042.00056.0180.0239.0

xySlope ≈

−−==

The slope of the line (∆y/∆x), Henry's law constant (H'), is approximately equal to 42.7.


Lesson 11 ___________________________________________________________________________________

11-10 2.0-7/98

Absorber Design

Theory

The first step in designing an air pollution control device is to develop a mathematical expression describing the observed phenomenon. A valid mathematical expression describing absorber performance makes it possible to determine the proper absorber size for a given set of conditions, and predict how a change in operating conditions affects absorber performance. A number of theories, or models, attempt to analytically describe the absorption mechanism. However, in practice, none of these analytical expressions can solely be used for design calculations. Experimental or empirical data must also be used to obtain reliable results.

The most widely used model for describing the absorption process is the two-film, or double-resistance, theory, which was first proposed by Whitman in 1923. The model starts with the three-step mechanism of absorption previously discussed in Lesson 2. From this mechanism, the rate of mass transfer was shown to depend on the rate of migration of a molecule in either the gas or liquid phase. The two-film model starts by assuming that the gas and liquid phases are in turbulent contact with each other, separated by an interface area where they meet. This assumption may be correct, but no mathematical expressions adequately describe the transport of a molecule through both phases in turbulent motion. Therefore, the model proposes that a mass-transfer zone exists to include a small portion (film) of the gas and liquid phases on either side of the interface. The mass-transfer zone is comprised of two films, a gas film and a liquid film on their respective sides of the interface. These films are assumed to flow in a laminar, or streamline, motion. In laminar flow, molecular motion occurs by diffusion, and can be categorized by mathematical expressions. This concept of the two-film theory is illustrated in Figure 11-3.

Figure 11-3. Visualization of two-film theory


2.0-7/98 11-11

According to the two-film theory, for a molecule of substance A to be absorbed, it must proceed through a series of five steps. The molecule must:

1. Migrate from the bulk-gas phase to the gas film

2. Diffuse through the gas film

3. Diffuse across the interface

4. Diffuse through the liquid film

5. Mix into the bulk liquid

The theory assumes that complete mixing takes place in both gas and liquid bulk phases and that the interface is at equilibrium with respect to pollutant molecules transferring in or out of the interface. This implies that all resistance to movement occurs when the molecule is diffusing through the gas and liquid films to get to the interface area, hence the name double-resistance theory. The partial pressure (concentration) in the gas phase changes from pAG in the bulk gas to pAI at the interface.

A gas concentration is expressed by its partial pressure. Similarly, the concentration in the liquid changes from cAI at the interface to cAL in the bulk liquid phase as mass transfer occurs. The rate of mass transfer from one phase to the other then equals the amount of molecule A transferred multiplied by the resistance molecule A encounters in diffusing through the films.

NA = kg(pAG − pAI) (11-3)

NA= kl(cAI − cAL) (11-4)

Where: NA = rate of transfer of component A, g-mol/h•m2 (lb-mole/hr•ft2) kg = mass-transfer coefficient for gas film, g-mol/h•m2•Pa (lb-mole/hr•ft2•atm) kl = mass-transfer coefficient for liquid film, g-mol/h•m2•Pa (lb-mole/hr•ft2•atm) pAG = partial pressure of solute A in the gas pAI = partial pressure of solute A at the interface cAI = concentration of solute A at the interface cAL = concentration of solute A in the liquid

Lesson 11 ___________________________________________________________________________________

11-12 2.0-7/98

The mass-transfer coefficients, kg and kl, represent the flow resistance the solute encounters in diffusing through each film respectively (Figure 11-4). As you can see from the above equations, as the value for a mass transfer coefficient increases, the amount of pollutant transferred (per unit of time) from the gas to the liquid increases. An analogy is the resistance electricity encounters as it flows through a circuit.

Figure 11-4. Resistance to motion encountered by a molecule being absorbed

Equations 11-3 and 11-4 define the general case of absorption and are applicable to both curved and straight equilibrium lines. In practice, Equations 11-3 and 11-4 are difficult to use, since it is impossible to measure the interface concentrations, pAI and cAI. The interface is a fictitious state used in the model to represent an observed phenomenon. Using the interface concentrations in calculations can be avoided by defining the mass-transfer system at equilibrium conditions and combining the individual film resistances into an overall resistance from gas to liquid and vice versa. If the equilibrium line is straight, the rate of absorption is given by the equations below:

( )N K p pA OG AG A= − * (11-5)

( )N K c cA OL A AL= −* (11-6)

Where: NA = rate of transfer of component, A, g-mol/h•m2 (lb-mole/hr•ft2) pA

* = equilibrium partial pressure of solute A at operating conditions


2.0-7/98 11-13

cA

* = equilibrium concentration of solute A at operating conditions KOG = overall mass-transfer coefficient based on gas phase, g-mol/h•m2•Pa (lb-mole/hr•ft2•atm) KOL = overall mass-transfer coefficient based on liquid phase, g-mol/h•m2•Pa (lb-mole/hr•ft2•atm) pAG = partial pressure of solute A in the gas cAL = concentration of solute A in the liquid

An important fact concerning Equations 11-5 and 11-6 is that they impose an upper limit on the amount of solute that can be absorbed. The rate of mass transfer depends on the concentration departure from equilibrium in either the gas (pAG - pA

* ) or liquid ( cA

* - cAL) phase. The larger these concentration differences are, the greater the rate of mass transfer becomes. If equilibrium is ever reached (pAG = pa

* and cAL = cA* ) absorption

stops and no net transfer occurs. Thus, the equilibrium concentrations determine the maximum amount of solute that is absorbed.

At equilibrium, the overall mass-transfer coefficients are related to the individual mass-transfer coefficients by the equations below.

1 1

K kHkOG g l

= + ′ (11-7)

1 1 1

K k HOL l g= +

′ k (11-8)

H' is Henry's law constant (the slope of the equilibrium). Equations 11-7 and 11-8 are useful in determining which phase controls the rate of absorption. From Equation 11-7, if H' is very small (which means the gas is very soluble in the liquid), then KOG ≈ kg, and absorption is said to be gas-film controlled. The major resistance to mass transfer is in the gas phase. Conversely, if a gas has limited solubility, H' is large, and Equation 11-8 reduces to KOL ≈ kl. The mass-transfer rate is liquid-film controlled and depends on the solute's dispersion rate in the liquid phase. Most systems in the air pollution control field are gas-phase controlled since the liquid is chosen so that the solute will have a high degree of solubility.

The discussion so far has been based on the two-film theory of absorption. Other theories offer different descriptions of gas molecule movement from the gas to the liquid phase. Some of the significant mass-transfer models follow. For these theories, the mass-transfer rate equation does not differ from that of the two-film method. The difference lies in the way they predict the mass-transfer coefficient. It has been shown that the rate of mass

Lesson 11 ___________________________________________________________________________________

11-14 2.0-7/98

transfer depends on a concentration difference multiplied by a resistance factor. Like most theories describing how something functions, absorption theories provide a basic understanding of the process, but due to the complexities of "real life" operations, it is difficult to apply them directly. Concentrations can easily be determined from operating (c and p) and equilibrium ( cA

* and pa* ) data of the system. Mass-transfer coefficients are

very difficult to determine from theory. Theoretically predicted values of the individual mass-transfer coefficients (kg and kl) based on the two-film theory, do not correlate well with observed values. Overall mass-transfer coefficients are more easily determined from experimental or operational data. However, the overall coefficients apply only when the equilibrium line is straight.

Mass-Transfer Models The following discussion on mass-transfer models is taken from Diab and Maddox (1982).

Film Theory (Whitman 1923) - First, and probably the simplest theory proposed for mass transfer across a fluid. Details of this model are discussed in the text because it is the most widely used.

Penetration Theory (Higbie 1935) - Assumes that the liquid surface in contact with the gas consists of small fluid elements. After contact with the gas phase, the fluid elements return to the bulk of the liquid and are replaced by another element from the bulk-liquid phase. The time each element spends at the surface is assumed to be the same.

Surface-Renewal Theory (Danckwerts 1951) - Improves on the penetration theory by suggesting that the constant exposure time be replaced by an assumed time distribution.

Film-Penetration Theory (Toor and Marchello 1958) - Combination of the film and penetration theories. Assumes that a laminar film exists at the fluid interface (as in the film theory), but further assumes that mass transfer is a nonsteady-state process.

Mass-transfer coefficients are often expressed by the symbols KOGa, kla, etc., where "a" represents the surface area available for absorption per unit volume of the column. This allows for easy determination of the column area required to accomplish the desired separation. These mass-transfer coefficients are developed from experimental data and are usually reported in one of two ways: as an empirical relationship based on a function of the liquid flow, gas flow, or slope of the equilibrium line; or correlated in terms of a dimensionless number, usually either the Reynolds or Schmidt Number.

Figure 11-5 provides an example comparing the effect of two types of packing materials on the mass-transfer coefficient for SO2 in water (Perry 1973). Packing A consists of one-inch rings and packing B consists of three-inch spiral tiles. As can be seen from this example, packing A has the higher transfer coefficient and would provide a better service in this application. Note that G' is the gas mass flow rate per cross-sectional area of tower (i.e. ft2). Similar figures are used extensively to compare


2.0-7/98 11-15

different absorbers or similar absorbers with varying operating conditions. It should be noted that these estimated mass-transfer coefficients are system and packing-type dependent and, therefore, do not have widespread applicability. The Chemical Engineers' Handbook gives a comprehensive listing of empirically derived coefficients. In addition, manufacturers of packed and plate towers have graphs in their literature similar to the one in Figure 11-5.

Figure 11-5. Comparison of overall absorption coefficient for SO2 in water Source: Perry 1973.

Although the science of absorption is considerably developed, much of the work in practical design situations is empirical in nature. The following sections will apply the principles discussed to the design of gas absorption equipment. Emphasis has been placed on presenting information that can be used to estimate absorber size and liquid flow rate.


Procedures

The effectiveness of an absorption system depends on the solubility of the gaseous contaminant. For very soluble gases, almost any type of absorber will give adequate removal. However, for most gases, only absorbers that provide a high degree of turbulent contact and a long residence time are capable of achieving high absorption efficiencies. The two most common high-efficiency absorbers are plate and packed towers. Both of these devices are used extensively to control gaseous pollutants. Absorber design calculations presented in this lesson will focus on these two devices.

Numerous procedures are used to design an absorption system. These procedures range in difficulty and cost from short-cut "rules of thumb" equations to in-depth design procedures based on pilot plant data. Procedures presented here will be based on the

Lesson 11 ___________________________________________________________________________________

11-16 2.0-7/98

short-cut "rules of thumb." The approaches discussed in this lesson are for single component systems (i.e., only one gaseous pollutant).

When an absorption system is designed, certain parameters are set by either operating conditions or regulations. The gas stream to be treated is usually the exhaust from a process in the plant. Therefore, the volume, temperature, and composition of the gas stream are given parameters. The outlet composition of the contaminant is set by the emission standard which must be met. The temperature and inlet composition of the absorbing liquid are also usually known. The main unknowns in designing the absorption system are the following:

• The flow rate of liquid required

• The diameter of the vessel needed to accommodate the gas and liquid flow

• The height of absorber required to achieve the needed removal

Procedures for estimating these three unknowns will be discussed in the following sections.

Material Balance In designing or reviewing the design of an absorption control system, the first task is to determine the flow rates and composition of each stream entering the system. From the law of conservation of mass, the material entering a process must either accumulate or exit. In other words, "what comes in must go out." A material balance helps determine flow rates and compositions of individual streams. Figure 11-6 illustrates the material balance for a typical countercurrent-flow absorber. The solute is the "material" in the material balance.

Figure 11-6. Material balance for countercurrent- flow absorber


2.0-7/98 11-17

The following procedure to set up a material balance and determine the liquid flow rate will focus on a countercurrent gas-liquid flow pattern. This is the most common flow pattern used to achieve high-efficiency gas absorption. For concurrent flow, only a slight modification of this procedure is required. Equations for crosscurrent flows are very complicated since they involve a gradient pattern that changes in two directions. They will not be presented here.

X = mole fraction of solute in pure liquid Y = mole fraction of solute in inert gas Lm = liquid molar flow rate, g-mol/h (lb-mole/hr) Gm = gas molar flow rate, g-mol/h (lb-mole/hr)

Engineering design work is usually done on a solute-free basis (X, Y) which means we ignore the amount of pollutant being transferred from the gas to the liquid. This makes the material balance calculations easier because we do not have to continually account for the change in mass of the flue gas as it is losing pollutant, or of the liquid as it is gaining pollutant. The solute-free basis is defined in Equations 11-9 and 11-10.

Y yy

=−1

(11-9)

X xx

=−1

(11-10)

In air pollution control systems, the percent of pollutant transferred from the gas to the liquid, y and x, is generally small compared to the flow of gas or liquid. Therefore, from Equations 11-9 and 11-10, Y ≈ y and X ≈ x. In this lesson, it is assumed that X and Y are always equal to x and y respectively. If y (inlet gas concentration) ever becomes larger than a few percent by volume, this assumption is invalid and will cause errors in the material balance calculations.

An overall mass balance across the absorber in Figure 11-7 yields Equation 11-11.

lb-mole in = lb-mole out (11-11)

Gm(in) + Lm(in) = Gm(out) + Lm(out)

For convenience, the top of the absorber is labeled as point 2 and the bottom as point 1. This changes Equation 11-11 to Equation 11-12.

Gm1 + Lm2 = Gm2 + Lm1 (11-12)

Lesson 11 ___________________________________________________________________________________

11-18 2.0-7/98

In this same manner, a material balance for the contaminant to be removed is obtained as expressed in Equation 11-13.

Gm1 Y1 + Lm2X2 = Gm2Y2 + Lm1X1 (11-13)

Equation 11-13 can be simplified by assuming that as the gas and liquid streams flow through the absorber, their total mass does not change appreciably (i.e., Gm1 = Gm2 and Lm1 = Lm2). This is justifiable for most air pollution control systems since the mass flow rate of pollutant is very small compared to the liquid and gas mass flow rates. For example, a 10,000-cfm exhaust stream containing 1,000 ppm SO2 would be only 0.1% SO2 by volume, or 1.0 cfm. If the scrubber were 100% efficient, the gas mass flow rate would change from 10,000 cfm at Gm1 to 9999 cfm at Gm2. The transfer of a quantity this small is negligible in an overall material balance. Therefore, Equation 11-13 can be reduced to Equation 11-14.

Gm(Y1 - Y2) = Lm(X1 − X2) (11-14)

By rearranging terms, Equation 11-14 becomes Equation 11-15.

( )Y Y LG

X Xm

m1 2 1 2− = − (11-15)

Equation 11-15 is the equation of a straight line. When this line is plotted on an equilibrium diagram, it is referred to as an operating line. This line defines operating conditions within the absorber: what is going in and what is coming out. An equilibrium diagram with a typical operating line plotted on it is shown in Figure 11-7. The slope of the operating line is the liquid mass flow rate divided by the gas mass flow rate, which is the liquid-to-gas ratio, or Lm/Gm. The liquid-to-gas ratio is used extensively when describing or comparing absorption systems. Determining the liquid-to-gas ratio is discussed in the next section.


2.0-7/98 11-19

Figure 11-7. Typical operating line diagram

Determining the Liquid Requirement In the design of most absorption columns, the quantity of exhaust gas to be treated (Gm) and the inlet solute (pollutant) concentration (Y1) are set by process conditions. Minimum acceptable standards specify the outlet pollutant concentration (Y2). The composition of the liquid flowing into the absorber (X2) is also generally known or can be assumed to be zero if it is not recycled. By plotting this data on an equilibrium diagram, the minimum liquid flow rate required to achieve the required outlet pollutant concentration (Y2) can be determined.

Lesson 11 ___________________________________________________________________________________

11-20 2.0-7/98

Figure 11-8(a) is a typical equilibrium diagram with operating points plotted for a countercurrent-flow absorber. Point A (X2, Y2) represents the concentration of pollutants in the liquid inlet and the gas outlet at the top of the tower. At the minimum liquid rate, the inlet gas concentration of solute (Y1) is in equilibrium with the outlet liquid concentration of solute (Xmax). The liquid leaving the absorber is saturated with solute and can no longer dissolve any more solute unless additional liquid is added. This condition is represented by point B on the equilibrium curve.

In Figure 11-8(b), the slope of the line drawn between point A and point B represents the operating conditions at the minimum flow rate. Note how the driving force decreases to zero at point B. The slope of line AB is (Lm/Gm)min, and may be determined graphically or from the equation for a straight line. By knowing the slope of the minimum operating line, the minimum liquid rate can easily be determined by substituting in the known gas flow rate. This procedure is illustrated in Example 11-2.

Determining the minimum liquid flow rate, (Lm/Gm)min, is important since absorber operation is usually specified as some factor of it. Generally, liquid flow rates are specified at 25 to 100% greater than the required minimum. Typical absorber operation would be 50% greater than the minimum liquid flow rate (i.e., 1.5 times the minimum liquid-to-gas ratio). Setting the liquid rate in this way assumes that the gas flow rate set by the process does not change appreciably. Line AC in Figure 11-8(c) is drawn at a slope of 1.5 times the minimum Lm/Gm. Line AC is referred to as the actual operating line since it describes absorber operating conditions.


2.0-7/98 11-21

Figure 11-8. Graphic determination of liquid flow rate

The following example problem illustrates how to compute the minimum liquid rate required to achieve a desired removal efficiency.

Lesson 11 ___________________________________________________________________________________

11-22 2.0-7/98

Example 11-2 Using the data and results from Example 11-1, compute the minimum liquid rate of pure water required to remove 90% of the SO2 from a gas stream of 84.9 m3/min (3,000 acfm) containing 3% SO2 by volume. The temperature is 293°K and the pressure is 101.3 kPa.

Solution 1. Determine the mole fractions of the pollutants in the gas phase, Y1 and Y2.

Then, sketch and label the drawing of the system as shown in Figure 11-9.

Y1 = 3% SO2 by volume = 0.03 mole fraction of SO2

Y2 = 90% reduction of SO2 from inlet concentration = (10%) (Y1) = (0.10) (0.03) = 0.003 mole fraction of SO2

Figure 11-9. Material balance for Example 11-2

2. Determine the mole fraction of SO2 in the liquid leaving the absorber to achieve the required removal efficiency. At the minimum liquid flow rate, the gas mole fraction of pollutants going into the absorber, Y1, will be in equilibrium with the liquid mole fraction of pollutants leaving the absorber, X1, (the liquid will be saturated with SO2). At equilibrium:

Y H1 1= ′ X


2.0-7/98 11-23

and Henry’s law constant from Example 11-1 is

′ =H 42 7. mole fraction of SO in airmole fraction of SO in water

2

2

X YH1

1

0 0342 70 000703

=′

=

=

..

.

3. Calculate the minimum liquid-to-gas ratio using Equation 11-15.

( )Y Y LG

X Xm

m1 2 1 2− = −

Therefore,

air of mol-g waterof mol-g4.38

0000703.0003.003.0

GL

XXYY

GL

m

m

21

21

m

m

=

−−=

−−=

4. Convert the exhaust stream flow rate, QG, to the exhaust gas molar flow rate, Gm (from units of m3/min to units of g-mole/min). At 0°C and 101.3 kPa, there are 0.0224 m3/g-mole for an ideal gas.

First, convert the volume of gas from 0 to 20°C (from 273 to 293°K). At 20°C:

0 0224 0 024. / . / m g - mol 293273

m g - mol of air3 3

=

Therefore,

G Qm G=

g - mol of air0.024 m3

1

Given: QG = 89.4 m3/min

Lesson 11 ___________________________________________________________________________________

11-24 2.0-7/98

Gm =

=

89 4

3 538

.

,

m / min 1 g - mol0.024 m

g - mol of air / min

33

5. Calculate the minimum liquid flow rate, Lmin. The minimum liquid-to-gas ratio was calculated in step 3.

LG

m

m

=min .38 4 g - mol of water

g - mol of air

Therefore, ( ) ( )L Gm mmin .= 38 4

From step 4: Gm = 3,538 g-mol of air/min

( )

kg/min 448,2

mol-kgkg 18

minmol-kg136.0=

:units mass toConvertingmin

waterof mol-kg0.136

min waterof mol-g000,136

air of mol-g waterof mol-g4.38

minair of mol-g538,3L minm

=

=

=

=

6. Figure 11-10 illustrates the graphical solution for this problem. To obtain the actual operating line, multiply the minimum operating line by 1.5.

AC = 1.5 AB AC = 1.5 (38.4) = 57.6


2.0-7/98 11-25

Figure 11-10. Graphical solution to Example 11-2


Sizing a Packed Tower

Packed Tower Diameter

The main parameter affecting the size of a packed column is the gas velocity at which liquid droplets become entrained in the exiting gas stream. Consider a packed column operating at set gas and liquid flow rates. By decreasing the diameter of the column, the gas flow rate (m/s or ft/sec) through the column will increase. If the gas flow rate through the column is gradually increased (by using smaller and smaller diameter columns), a point will be reached where the liquid flowing down over the packing begins to be held in the void spaces between the packing. This gas-to-liquid ratio is termed the loading point. The pressure drop of the column begins to increase and the degree of mixing between the phases decreases. A further increase in gas velocity will cause the liquid to completely fill the void spaces in the packing. The liquid forms a layer over the top of the packing and no more liquid can flow down through the tower. The pressure drop increases substantially, and mixing between the phases is minimal. This condition is referred to as flooding, and the gas velocity at which it occurs is the flooding velocity. Using an extremely large-diameter tower would eliminate this problem. However, as the diameter increases, the cost of the tower increases.

Normal practice is to size a packed column diameter to operate at a certain percent of the flooding velocity. A typical operating range for the gas velocity through the columns is 50 to 75% of the flooding velocity. It is assumed that, by operating in this range, the gas velocity will also be below the loading point.

Lesson 11 ___________________________________________________________________________________

11-26 2.0-7/98

A common and relatively simple procedure for estimating flooding velocity (thus, setting a minimum column diameter) is to use a generalized flooding and pressure drop correlation. One version of the flooding and pressure drop relationship for a packed tower is in the Sherwood correlation, shown in Figure 11-11 (Calvert et al. 1972). Figure 11-11, was developed from experimental data, knowing the physical properties of the gas and liquid streams and tower packing characteristics. In Figure 11-11, the terms and units must be used as presented since the relationships are based on empirical data. The “X” axis (or abscissa) is a function of the physical properties of the gas and liquid streams. The “Y” axis (ordinate), is also a function of the gas and liquid properties as well as the packing material utilized. The graph is used to predict what conditions will cause flooding to occur. Since flooding is an unacceptable operating condition, this sets a minimum tower diameter for a given set of gas/liquid conditions. Knowing minimum unacceptable diameter, a larger, operating diameter can be specified.

Figure 11-11. Generalized flooding and pressure drop correlation Source: Calvert et al. 1972.


2.0-7/98 11-27

The procedure to determine the tower diameter is given below.

1. Calculate the value of the abscissa.

Abscissa = LG

ρρ

g

l

0.5

(11-16)

Where: L and G = mass flow rates (any consistent set of units may be used as long as the term is dimensionless) ρg = density of the gas stream ρl = density of the absorbing liquid

2. From the point calculated in Equation 11-16, proceed up the graph to the flooding line and read the ordinate, ε.

3. Rearrange the equation of the ordinate and solve for G'.

( )( )( )( )

′ =

Gg

Fg c

l

ε ρ ρ

φµ10.2

0.5

(11-17)

Where: G' = mass flow rate of gas per unit cross-sectional area of column, g/s•m2 (lb/sec•ft2) ρg = density of the gas stream, kg/m3 (lb/ft3) ρl = density of the absorbing liquid, kg/m3 (lb/ft3) gc = gravitational constant, 9.82 m/s2 (32.2 ft/sec2) F = packing factor given in Table 11-5 for different types of packing (Bhatia 1977) φ = ratio of specific gravity of the scrubbing liquid to that of water µl = viscosity of liquid

4. Calculate G' at operating conditions. G' at operating conditions is a fraction of G' at flooding conditions.

( )( )′ = ′G f Goperating flooding (11-18)

Where: f = the percent of flooding velocity, usually 50 to 75%

Lesson 11 ___________________________________________________________________________________

11-28 2.0-7/98

5. Calculate the cross-sectional area of column A from Equation 11-19.

A GG operating

=′

(11-19)

6. Calculate the diameter of the column from Equation 11-20.

dt =

4Aπ

0.5

(11-20)

= 1.13 A0.5

Table 11-5. Packing data1

Packing Size (in.)

Weight(lb/ft2)

Surface area, a (ft2/ft3

packing volume)

Void fraction

(%)

Packing factor, F

(ft2/ft3)

Raschig rings

(ceramic and porcelain)

1/2 1

1 1/2 2 3

52 44 42 38 34

114 58 36 28 19

65 70 72 75 77

580 155 95 65 37

Raschig rings

(steel)

1/2 × 1/32 1 × 1/32 2 × 1/16

77 40 38

128 63 31

84 92 92

300 115 57

Berl saddles

(ceramic and porcelain)

1/4 1/2 1 2

55 54 48 38

274 155 79 32

63 64 68 75

900 240 110 45

Intalox saddles

(ceramic)

1/4 1/2 1 2

54 45 44 42

300 190 78 36

75 78 77 79

725 200 98 40

Intalox saddles

(plastic)

1 2 3

6.00 3.75 3.25

63 33 27

91 93 94

30 20 15

Pall rings (plastic)

5/8 1 2

7.0 5.5 4.5

104 63 31

87 90 92

97 52 25



2.0-7/98 11-29

Table 11-5. (continued) Packing data1

Pall rings (metal)

5/8 × 0.018 thick

1 1/2 × .03 thick

38 24

104 39

93 95

73 28

Tellerettes 1 2 3

7.5 3.9 5.0

55 38 30

87 93 92

40 20 15

1. Note: Data for guide purposes only. Source: Bhatia 1977.

Example 11-3 This example illustrates the use of Figure 11-11 for computing the minimum allowable diameter for a packed tower. For the scrubber in Example 11-2, determine the column diameter if the operating liquid rate is 1.5 times the minimum. The gas velocity should be no greater than 75% of the flooding velocity, and the packing material is two-inch ceramic Intalox saddles.

Solution 1. Determine the actual gas and liquid flow rates for the system. For Example

11-2, the gas molar flow rate in the absorber, Gm, was 3,538 g-mol/min and the minimum liquid flow rate, Lmin, was 2,448 kg/min. The actual liquid flow rate in the absorber should be 1.5 times the minimum flow rate:

L = Lmin × 1.5 = (2,448 kg/min) (1.5) = 3,672 kg/min

Assuming the molecular weight of the exhaust gas is 29 kg/mol, convert the gas molar flow rate (Gm) to mass flow rate (G).

G = Gm × (29 kg/kg-mol) G = (3,538 g-mol/min)(29 kg/kg-mol) = (3.538 kg-mol/min)(29 kg/kg-mol) = 102.6 kg/min

2. Using Equation 11-16, calculate the abscissa for Figure 11-11.

Abscissa LG

g

l=

ρρ

0.5

Lesson 11 ___________________________________________________________________________________

11-30 2.0-7/98

The densities of air and water at 30°C are:

ρg = 1.17 kg/m3 ρl = 1,000 kg/m3

Abscissa = 3,672102.6

1171 000

0.5.,

3. Using Figure 11-12, with the abscissa of 1.22, move up to the flooding line and read the value of ε on the ordinate.

ε = 0.019

Figure 11-12. Generalized flooding and pressure drop correlation for Example 11-3


2.0-7/98 11-31

4. Calculate the superficial flooding velocity, G' using Equation 11-17. The superficial flooding velocity is the flow rate per unit of cross-sectional area of the tower.

( )( )( )( )

′ =

Gg

Fg l c

l

ε ρ ρ

φµ 0.2

0.5

Given: ρg = 1.17 kg/m3, density of air at 30°C ρl = 1,000 kg/m3, density of water at 30°C gc = 9.82 m/s2, the gravitational constant F = 40 ft2/ft3 (131 m2/m3), the packing factor for two- inch ceramic Intalox saddles (see Table 11-5) φ = 1.0, the ratio of specific gravity of the scrubbing liquid(water) to that of water µl = 0.0008 Pa•s, the viscosity of liquid

From step 3: ε = 0.019

( )( )( )( )

( )( )( )′ =

= •

G

s

0 019 117 1000 9 821 131 0 0008

2 63

0.2

0.5. . .

.

. kg / m at flooding2

5. Calculate the superficial gas velocity at operating conditions (G'operating) using Equation 11-18.

G′operating = (f)(G′ flooding)

Where: f = 75%

From step 4: G'flooding = 2.63 kg/s•m2

G'operating = (0.75)(2.63 kg/s •m2) = 1.97 kg/s•m2

Lesson 11 ___________________________________________________________________________________

11-32 2.0-7/98

6. Calculate the cross-sectional area of the packed tower using Equation 11-19.

A GGoperating

=′

From step 1: G = 102.6 kg/min From step 5: G'operating = 1.97 kg/s•m2

( )( )A =

•

=

102 6 11 97

0 87

..

.

kg / min min / 60 sec kg / s m

m

2

2

7. Calculate the tower diameter using Equation 11-20.

d At =

4 0 5

π

.

Where: π = 3.14

From step 6: A = 0.87 m2

( )dt =

=≈

4 0 87314

1 05

0 5..

.

.

m1.1 m

8. Use Figure 11-11 to estimate the pressure drop across the absorber, ∆p, once the superficial gas velocity for operating conditions has been set. First, plug G'operating back into Equation 11-17 and rearrange the equation to get the ordinate, ε.


2.0-7/98 11-33

( ) ( )( )( )( )( )( )

ε φ µρ ρ

ε

= ′

=• •

=

G Fg

m

l

g l c

2 0.2

2 3 0.21 97 1 131 0 0008

117 1 000 9 82

0 0106

. / .

. , .

.

kg / s m m Pa s

kg / m kg / m m / s

2 2

3 3 2

The ordinate equals 0.0106 and the abscissa equals 1.22. Then from Figure 11-13, read ∆p. The pressure drop equals 0.0416 m of water/m of packing.

Figure 11-13. Generalized flooding and pressure drop correlation for Example 11-3

Packed Tower Height

The height of a packed column refers to the depth of packing material needed to accomplish the required removal efficiency. The more difficult the separation, the larger the packing height required. For example, a much larger packing height would be required to remove SO2 than to remove chlorine (Cl) from an exhaust stream using water as the absorbent because Cl is more soluble in water than SO2 is. Determining the proper height of packing is important since it affects both the rate and efficiency of absorption.

Lesson 11 ___________________________________________________________________________________

11-34 2.0-7/98

A number of theoretical equations are used to predict the required packing height. These equations are based on diffusion principles. Depending on which phase is controlling the absorption process, either Equation 11-5 or 11-6 is used as the starting point to derive an equation to predict column height. A material balance is then set up over a small differential section (height) of the column.

The general form of the design equation for a gas-phase controlled resistance is given in Equation 11-21.

( )( )

Z GK aP

dYY Y YOG Y

Y

= ′− −∫

2

1

1 * (11-21)

Where: Z = height of packing, m G' = mass flow rate of gas per unit cross-sectional area of column, g/s•m2 KOG = overall mass-transfer coefficient based on the gas phase, g-mol/h•m2•Pa a = interfacial contact area, m2

P = pressure of the system, kPa Y1 = inlet gas pollutant concentration Y2 = outlet gas pollutant concentration Y* = pollutant concentration in gas at equilibrium

In analyzing Equation 11-21, the term G'/KOGaP has the dimension of meters and is defined as the height of a transfer unit. The term inside the integral is dimensionless and represents the number of transfer units needed to make up the total packing height. Using the concept of transfer units, Equation 11-21 can be simplified to:

Z = HTU × NTU (11-22)

Where: Z = height of packing, m HTU = height of a transfer unit, m NTU = number of transfer units

The concept of a transfer unit comes from the assumptions used in deriving Equation 11-21. These assumptions are: (1) that the absorption process is carried out in a series of contacts, or stages, and (2) that the streams leaving these stages are in equilibrium with each other. The stages can be visualized as the height of an individual transfer unit and the total tower height is equal to the number of transfer units times the height of each unit. Plate towers operate in this manner where they have discrete contact sections. Although a packed column operates as one continuous separation (differential contactor) process, in design terminology it is treated as discrete sections (transfer units) in order to perform a mass balance around a small subsection of the tower. The number and the height of a transfer unit are based on either the gas or the liquid phase. Equation 11-22 now becomes:


2.0-7/98 11-35

Z = NOGHOG = NOLHOL (11-23)

Where: Z = height of packing, m NOG = number of transfer units based on an overall gas-film coefficient, KOG NOL = number of transfer units based on an overall liquid- film coefficient, KOL HOG = height of a transfer unit based on an overall gas-film coefficient, m HOL = height of a transfer unit based on an overall liquid-film coefficient, m

The number of transfer units, NTU, can be obtained experimentally or calculated from a variety of methods. For the case where the solute concentration is very low and the equilibrium line is straight, Equation 11-24 can be used to determine the number of transfer units (NOG) based on the gas-phase resistance. Equation 11-24 can be derived from the integral portion of Equation 11-21.

N

Y mXY mX

mGL

mGL

mGL

OG

m

m

m

m

m

m

=

−−

−

+

−

ln 1 2

2 21

1 (11-24)

Where: NOG = number of transfer units based on an overall gas-film coefficient, KOG Y1 = mole fraction of solute in entering gas Y2 = mole fraction of solute in exiting gas m = slope of equilibrium line X2 = mole fraction of solute entering the column Gm = molar flow rate of gas, kg-mol/h Lm = molar flow rate of liquid, kg-mol/h

Equation 11-24 may be solved directly or graphically by using the Colburn diagram, which is presented in Figure 11-13. The Colburn diagram is a plot of the NOG versus ln[Y1 − mX2/Y2 − mX2] at various values of (mGm/Lm). The term (mGm/Lm) is referred to as the absorption factor. In using Figure 11-14, first compute the value of [Y1 − mX2/Y2 − mX2]; next read up the graph to the line corresponding to (mGm/Lm), and then read across to obtain the NOG.

Lesson 11 ___________________________________________________________________________________

11-36 2.0-7/98

Figure 11-14. Colburn diagram Source: Perry 1973.

Equation 11-24 can be further simplified for situations where a chemical reaction occurs or if the solute is extremely soluble. In these cases, the solute exhibits almost no partial pressure; therefore, the slope of the equilibrium line approaches zero (m → 0). For either of these cases, Equation 11-24 reduces to Equation 11-25.

N YYOG = ln 1

2 (11-25)

The number of transfer units depends only on the inlet and outlet concentration of the solute. For example, if the conditions of Equation 11-25 are met, achieving 90% removal of any pollutant requires 2.3 transfer units. Equation 11-25 applies only when the equilibrium line is straight and the slope approaches zero (for very soluble or reactive gases).


2.0-7/98 11-37

Values for the height of a transfer unit used in designing absorption systems are usually obtained from experimental data. To ensure greatest accuracy, vendors of absorption equipment normally perform pilot plant studies to determine the HTU. For common absorption systems, such as NH3 and water, manufacturers have developed graphs for estimating HTU. These graphs do not provide the accuracy of pilot plant data, but are less expensive and easier to use. Figure 11-15 gives a typical example of these graphs for an ammonia and water system. In this figure, the superficial liquid flow rate is plotted versus the HOG with the superficial gas rate as a parameter. For a given liquid flow rate, the height of a transfer unit for the 1-inch Tellerettes is less than that for the 1-1/2 inch Raschig rings. Therefore, a system would need less Tellerette packing to accomplish the same removal. For this example the Tellerettes would be more efficient. It is also common to plot gas rate versus the HOG and have the liquid rate as a parameter. Additional information on other gas-liquid systems can be found in Chemical Engineers' Handbook (Perry 1973). In applying these data, process conditions must be similar to conditions at which the HTU was measured.

Figure 11-15. Column packing comparison for ammonia and water system

Lesson 11 ___________________________________________________________________________________

11-38 2.0-7/98

When no experimental data are available, or if only a preliminary estimate of absorber efficiency is needed, generalized correlations are available to predict the height of a transfer unit. The correlations for predicting the HOG or the HOL are empirical in nature and are a function of:

1. Type of packing 2. Liquid and gas flow rates 3. Concentration and solubility of the pollutant 4. Liquid properties 5. System temperature

These correlations can be found in engineering texts such as Chemical Engineers' Handbook (Perry 1973), Wet Scrubber System Study, Volume I (Calvert et al. 1972), or Mass Transfer Operations (Treybal 1968). For most applications, the height of a transfer unit ranges between 0.3 and 1.2 m (1 to 4 ft) (Calvert 1977). As a rough estimate, 0.6 m (2.0 ft) can be used.

Example 11-4 From pilot plant studies of the absorption system in Example 11-2 it was determined that the HOG for the SO2-water system is 0.829 m (2.72 ft). Calculate the total height of packing required to achieve 90% removal. The following data were taken from the previous examples.

m, Henry’s law constant for the equilibrium diagram for SO2 and water system (see Example 11-1).

42 7. kg - mol of waterkg - mol of air

Gm, molar flow rate of gas 3.5 kg-mol/min

Lm, molar flow rate of liquid 3,672 kg/min × kg-mol/18 kg= 204 kg-mol/min

X2, mole fraction of solute in entering liquid

0 (no recycle liquid)

Y1, mole fraction of solute in entering gas

0.03

Y2, mole fraction of solute in existing gas 0.003


2.0-7/98 11-39

Solution 1. Calculate the number of transfer units, NOG, using Equation

11-24.

( )( ) ( )( )

( )( )

N

Y mXY mX

mGL

mGL

mGL

N

OG

m

m

m

m

m

m

OG

=

−−

−

+

−

=

−

+

−

=

ln

ln ..

. . . .

. .

.

1 2

2 21

1

0 030 003

142 7 3 5

20442 7 3 5

204

142 7 3 5

204

504

2. Calculate the total packing height, Z, using Equation 11-23.

Z = HOG × NOG

Given: HOG = 0.829 m, height of a transfer unit

From step 1: NOG = 5.04 m

Z = (0.829 m)(5.04) = 4.18 m of packing height


Sizing a Plate Tower

Another scrubber used extensively for gas absorption is a plate tower. Here, absorption occurs on each plate, or stage. These are commonly referred to as discrete stages, or steps. The following discussion presents a simplified method for sizing or reviewing the design plans of a plate tower. The method for determining the liquid flow rate in the plate tower is the same as previously discussed. Methods for estimating the diameter of a plate tower and the theoretical number of plates follow.

Plate Tower Diameter

The minimum diameter of a single-pass plate tower is determined by using the gas velocity through the tower. If the gas velocity is too fast, liquid droplets are entrained,

Lesson 11 ___________________________________________________________________________________

11-40 2.0-7/98

causing a condition known as priming. Priming occurs when the gas velocity through the tower is so fast that it causes liquid on one tray to foam and then rise to the tray above. Priming reduces absorber efficiency by inhibiting gas and liquid contact. For the purpose of determining tower diameter, priming in a plate tower is analogous to the flooding point in a packed tower. It determines the minimum acceptable diameter. The actual diameter should be larger.

The smallest allowable diameter for a plate tower is expressed in Equation 11-26.

( )d Qt G g= ψ ρ0.5

(11-26)

Where: QG = volumetric gas flow, m3/h ψ = empirical correlation, m0.25h0.5/kg0.25

ρg = gas density, kg/m3

The term ψ is an empirical correlation and is a function of both the tray spacing and the densities of the gas and liquid streams. Values for ψ in Table 11-6 are for a tray spacing of 61 cm (24 in.) and a liquid specific gravity of 1.05 (Calvert et al. 1972). If the specific gravity of a liquid varies significantly from 1.05, the values for ψ in Table 11-6 cannot be used.

Table 11-6. Empirical constants for Equation 11-26

Tray Metric Ψa English Ψb

Bubble cap

0.0162 0.1386

Sieve 0.0140 0.1198 Valve 0.0125 0.1069 a. Metric Ψ is expressed in m0.25 h0.5/kg0.25, for use with QG expressed in m3/h, and ρg expressed in kg/m3.

b. English Ψ is expressed in ft0.25 min0.5/lb0.25, for use with QG in cfm, and ρg expressed in lb/ft3.

Source: Calvert et al. 1972.

Depending on operating conditions, trays are spaced with a minimum distance between plates to allow the gas and liquid phases to separate before reaching the plate above. Trays should be spaced to allow for easy maintenance and cleaning. Trays are normally spaced 45 to 70 cm (18 to 28 in.) apart. In using Table 11-6 for a tray spacing different from 61 cm, a correction factor must be used. Figure 11-16 is used to determine the correction factor, which is multiplied by the estimated diameter. Example 11-5 illustrates how to estimate the minimum diameter of a plate tower.


2.0-7/98 11-41

Figure 11-16. Tray spacing correction factor Source: Calvert et al. 1972.

Example 11-5 For the conditions described in Example 11-2, determine the minimum acceptable diameter if the scrubber is a bubble-cap tray tower. The trays are spaced 0.53 m (21 in.) apart.

Solution To determine the minimum acceptable diameter of the plate tower, we will use Equation 11-26:

( )d Qt G g= ψ ρ0.5

From Example 11-2, the following information is obtained:

QG, gas flow rate = 84.9 m3/min ρg, gas density = 1.17 kg/m3

Lesson 11 ___________________________________________________________________________________

11-42 2.0-7/98

1. Convert the gas flow rate, QG, to units of m3/h.

QG = (84.9 m3/min) (60 min/hr) = 5,094 m3/h

2. Determine the empirical constant, ψ. From Table 11-6, the value for ψ is 0.0162 m0.25 h0.5/kg0.25.

3. Calculate the minimum diameter, dt, of the plate tower using Equation 11-26.

( )d = Qt Gψ ρg

0.5

Given: ρg = 1.17 kg/m3

From step 1: QG = 5,094 m3/h From step 2: ψ = 0.0162 m0.25 h0.5/kg0.25

( ) ( )[ ]dt =

=

0 0162 5 094 117

12

0.5. , .

.

m

4. Correct the diameter using Figure 11-16. The tray spacing for each tray is 0.53 m but the values in Table 11-6 are for a tray spacing of 0.61 m. Read a correction factor of 1.05.

Figure 11-17. Tray spacing correction factor for Example 11-5


2.0-7/98 11-43

5. Adjust the minimum plate tower diameter value by using the correction factor.

( )

( )Adjusted d from step 3 factor

d m 1.05 m

t

t

= ×

==

d correctiont

1 21 26..

Note: The value of 1.26 m is the minimum estimated tower diameter based on priming conditions. In practice, a larger diameter based on economic conditions is usually chosen.

Number of Theoretical Plates

Several methods are used to determine the number of ideal plates, or trays, required for a given removal efficiency. These methods, however, can become quite complicated. One method used is a graphical technique. The number of ideal plates is obtained by drawing "steps" on an operating diagram. This procedure is illustrated in Figure 11-18. This method can be rather time consuming, and inaccuracies can result at both ends of the graph.

Figure 11-18. Graphic determination of the number of theoretical plates

Lesson 11 ___________________________________________________________________________________

11-44 2.0-7/98

Equation 11-27 is a simplified method used to estimate the number of plates. This equation can only be used if both the equilibrium and operating lines for the system are straight. This is a valid assumption for most air pollution control systems. This equation, taken from Sherwood and Pigford (1952), is derived in the same manner as Equation 11-24 for computing the NOG of a packed tower. The difference is that Equation 11-27 is based on a stepwise solution instead of a continuous contactor, as is the packed tower. (Note: This derivation is referred to as the height equivalent to a theoretical plate, or HETP instead of HTU.)

N

Y mXY mX

mGL

mGL

LmG

p

m

m

m

m

m

m

=

−−

−

+

ln

ln

1 2

2 21

(11-27)

This equation is used to predict the number of theoretical plates required to achieve a given removal efficiency. The operating conditions for a theoretical plate assume that the gas and liquid streams leaving the plate are in equilibrium with each other. This ideal condition is never achieved in practice. A larger number of actual trays are required to compensate for this decreased tray efficiency.

Three types of efficiencies are used to describe absorption efficiency for a plate tower:

1. An overall efficiency, which is concerned with the entire column

2. Murphree efficiency, which is applicable with a single plate

3. Local efficiency, which pertains to a specific location on a plate

A number of methods are available to predict these plate efficiencies. These methods are complex, and values predicted by two different methods for a given system can vary by as much as 80% (Zenz 1972).

The simplest of tray efficiency concepts, the overall efficiency, is the ratio of the number of theoretical plates to the number of actual plates. Since overall tray efficiency is an over-simplification of the process, reliable values are difficult to obtain. For a rough estimate, overall tray efficiencies for absorbers operating with low-viscosity liquid normally fall in a 65 to 80% range (Zenz 1972).

Example 11-6 Calculate the number of theoretical plates required for the scrubber in Example 11-5 using the same conditions as those in Example 11-4. Estimate the total height of the column if the trays are spaced at 0.53-m intervals, and assume an overall tray efficiency of 70%.


2.0-7/98 11-45

Solution 1. Estimate the number of theoretical plates by using Equation 11-27.

N

Y mXY mX

mGL

mGL

LmG

p

m

m

m

m

m

m

=

−−

−

+

ln

ln

1 2

2 2

1

From Example 11-5 and the previous examples, the following data are obtained:

m = 42.7, Henry’s law equilibrium constant Y1, (inlet gas) = 0.03 mole fraction Y2, (outlet gas) = 0.003 mole fraction X2, (inlet liquid) = 0.0 mole fraction Lm = 204 kg-mol/min, the molar flow rate of liquid Gm = 3.5 kg-mol/min, the molar flow rate of gas

( )( ) ( )( )

( )( )

Np =

−−

−

+

=

ln ..

. . . .

ln. .

.

0 03 00 003 0

142 7 3 5

20442 7 35

204

20442 7 35

3 94 theoretical plates

2. Estimate the actual number of plates assuming that the overall efficiency of each plate is 70%.

Actual plates = estimated plates

70%

tes = 3.940.70

Actual pla

= 5.6 or 6 plates (since you can’t have a fraction of a plate)

3. Estimate the height of the tower, Z, by using the following equation:

Z = (Np × tray spacing) + top height of tower

Lesson 11 ___________________________________________________________________________________

11-46 2.0-7/98

The top height of the tower is the distance that allows the gas-vapor mixture to separate. This distance is usually the same distance as the tray spacing.

Z = (6 plates) (0.53 m) + 0.53 m = 3.18 + 0.53 = 3.71 m

Note: This height is approximately the same as that predicted for the packed tower in Example 11-4. This seems logical since both packed and plate towers are efficient gas-absorption devices. However, due to the many assumptions, no concrete generalization can be made.

Summary

For gas absorption, the two devices most often used are the packed tower and the plate tower. Both of these devices, if designed and operated properly, can achieve high collection efficiencies for a wide variety of gases. Other scrubbing systems can be used for absorption, but are limited to cases where the gases are highly soluble. For example, spray towers, venturis, and cyclonic scrubbers are designed assuming the performance is equivalent to one single equilibrium stage (i.e., NOG = 1) (Perry 1973).

The equations and procedures used in designing packed and plate towers are very similar. Both are based on solubility, the mass-transfer model, and the geometry of the tower. The main difference is that the equations for a plate tower are based on a stepwise process, whereas those for a packed tower are based on a continuous-contacting process. Care must be taken when applying any of the equations presented in this lesson (or in other texts). Some of the equations are empirical and are applicable only under a similar set of conditions. Used correctly, these procedures can be a useful tool in checking absorber designs or in determining the effect of a process change on absorber operation.

When checking the design plans for the permit process, the agency engineer should check its files or another agency's files for similar applications for absorber installations. A review of these data will help determine if the absorber design specifications submitted by the industrial source's officials are adequate to achieve pollutant removal efficiency for compliance with the regulations. The agency engineer should require the source owner/operator to conduct stack tests (once the source is operating) to determine if the source is in compliance with local, state, and federal regulations. The agency engineer should also require that the source owner/operator submit an operation and maintenance schedule that will help keep the scrubber system on line.



2.0-7/98 11-47

Review Exercise Questions

Part 1

1. Of the wet collectors listed below, which is/are the best device(s) for removing gaseous pollutants from process exhaust streams?

a. Packed tower b. Plate tower c. Venturi scrubber d. Centrifugal scrubber e. a and b

2. In the absorption process, the solute is the:

a. Inert portion of the gas stream b. Area where the gas phase and liquid phase come into contact with each other c. Gaseous pollutant that is absorbed d. Capability of a gas to be dissolved in a liquid

3. A very important factor affecting the amount of a pollutant that can be absorbed is its ____________________.

4. In an absorber, as the temperature of the system increases, the amount of pollutant that can be absorbed ____________________.

a. Increases b. Decreases

5. A plot of the mole fraction of the solute in the liquid phase versus the mole fraction of the solute in the gas phase is called:

a. The partial pressure b. An equilibrium diagram c. A concentration gradient

6. What is one form of the equation for Henry's law?

a. x = Hp b. H = xp c. H = x/y d. y = H'x

7. In describing the solubility of various gases in water, the ____________________ Henry's law constant is, the more soluble the gas is.

a. Smaller b. Larger

Lesson 11 ___________________________________________________________________________________

11-48 2.0-7/98

Part 2

8. In the double-resistance, or two-film theory, a ____________________ zone exists that includes a gas and liquid phase on either side of the interface.

a. Soluble b. Mass-transfer c. Droplet

9. True or False? The two-film theory implies that all resistance to movement occurs when the molecule (gaseous pollutant) is diffusing through the gas and liquid films.

10. In absorption equations, the concentration of a gaseous pollutant is usually expressed by its:

a. Diffusion rate b. Total pressure c. Partial pressure

11. In calculating the rate of mass transfer of pollutant A, NA, using the equation

( )N K p pA OG AG A*= − , the term KOG is the:

a. Equilibrium concentration of pollutant A b. Mass-transfer coefficient for the gas film c. Mass-transfer coefficient for the liquid film d. Overall mass-transfer coefficient based on the gas phase

12. True or False? Overall mass-transfer coefficients are only valid when a plot of the equilibrium data yields an equilibrium line that is straight.

Part 3

13. In absorption calculations, a(an) ____________________ equates the gas and liquid concentrations coming into the absorber with the gas and liquid concentrations going out of the absorber.

a. Material balance b. Energy balance c. Transfer unit

14. In air pollution calculations, the mass of the pollutant is usually very ____________________ compared to the mass of exhaust gas being treated and the mass of the liquid used in the absorber.

a. Small b. Large


2.0-7/98 11-49

15. In the graph below, the line AB is the:

a. Equilibrium line b. Actual operating line c. Minimum operating line

16. The slope of the actual operating line is:

a. Minimum liquid-to-gas ratio b. Gm/Lm (actual) c. Lm/Gm (actual)

17. True or False? In the following figure, point B represents absorber conditions where the liquid leaving the absorber is saturated with the pollutant and can no longer absorb any additional pollutant, unless more liquid is added.

Lesson 11 ___________________________________________________________________________________

11-50 2.0-7/98

Part 4

18. In designing a packed tower, the normal practice is to make the tower diameter so that the unit will operate at ____________________ of the flooding velocity rate.

a. 50 to 75% b. 100% c. 150%

19. True or False? The Sherwood correlation can be used to calculate the tower diameter of a packed tower, if the minimum liquid rate, Lm, and the gas flow rate, G, through the absorber are known.

20. In estimating packing height in a packed tower, the packing sections are broken up into discrete sections called:

a. Transfer units b. Gas-film coefficients c. Liquid-film coefficients

21. The packing height, Z, can be estimated from the following equation:

Z = HTU × NTU

What are the terms HTU and NTU?

HTU: ________________________________________________________________

NTU: ________________________________________________________________

22. True or False? The Colburn diagram can be used to estimate the number of transfer units based on an overall gas-film coefficient, NOG, if the absorption factor (mGm/Lm), the inlet and outlet pollutant concentrations, and the liquid recycle concentrations are known.

23. The height of a transfer unit is a function of:

a. Type of packing b. Liquid and gas flow rates c. Pollutant concentration and solubility d. Liquid properties and system temperature e. All of the above

24. For most packed tower applications, the height of a transfer unit can be estimated to be:

a. 3 to 4.6 m (10 to 15 ft) b. 0.3 to 1.2 m (1 to 4 ft) c. 1.82 to 3 m (6 to 10 ft)


2.0-7/98 11-51

Part 5

25. In a plate tower, if the gas velocity through the tower is too fast, liquid droplets become entrained in the gas stream, causing a condition called:

a. Pumping b. Streaking c. Priming

26. True or False? For the purpose of determining a plate-tower diameter, priming in a plate tower is the same as the flooding point in a packed tower.

27. In a plate tower, the following equation

N

Y mXY mX

mGL

mGL

LmG

p

m

m

m

m

m

m

=

−−

−

+

ln

ln

1 2

2 21

is used to calculate the:

a. Number of transfer units based on an overall gas-film coefficient b. Number of transfer units based on Henry's law constant c. Number of theoretical plates

28. In plate towers, the efficiency of each plate, or tray, is usually ____________________.

a. 20 to 30% b. 65 to 80% c. 90 to 100%

Lesson 11 ___________________________________________________________________________________

11-52 2.0-7/98


2.0-7/98 11-53

Problem 1

A medical waste incinerator utilizes a packed scrubber to remove HCl and other soluble gases. Given the operating conditions below, estimate the scrubbing liquid volumetric flow rate, QL, (essentially water with some caustic added to control pH) to achieve the required removal efficiency.

QG, gas flow 15,000 acfm at 500oF Y1, concentration of HCI in inlet gas 1,000 ppm or 47 lb/hr

Y2, concentration of HCI in outlet gas 30 ppm or 1.4 lb/hr

X2, concentration of HCI in inlet liquid 0

m, Henry’s Law equilibrium constant 1.1

actual flow rate 1.5 times minimum

ρl, density of water 8.35 lb/gal

R, ideal gas constant at 70°F 380 scf/lb-mole

molecular weight of water 18 lb/mole

molecular weight of HCI 36 lb/mole

Lesson 11 ___________________________________________________________________________________

11-54 2.0-7/98



2.0-7/98 11-55


Lesson 11 ___________________________________________________________________________________

11-56 2.0-7/98


2.0-7/98 11-57

Problem 2

A sewage treatment plant utilizes a countercurrent flow, packed bed scrubber to control odor emissions. The scrubbing liquid uses potassium permanganate solution in water and the packing material is 1 inch Berl ceramic saddles. Because of development in the area, the treatment plant needs to increase capacity by 25%. Given the data below, can the present tower accommodate the added flows?

Existing Proposed QG, volumetric flow rate of gas 10,000 acfm 12,500 acfm QL, volumetric flow rate of liquid 100 gal/min

125 gal/min

T, temperature of gas 70oF 70oF µl, viscosity of liquid 0.018

centipoise 0.018 centipoise

ρl, density of liquid 64 lb/ft3 64 lb/ft3

ρg, density of gas 0.075 lb/ft3 0.075 lb/ft3

φ, ratio of specific gravity of scrubbing liquid to that of water

1.01 1.01

F, packing factor 45 ft3 45 ft3 tower diameter 4 ft 4 ft

Constants and assumptions:

R, ideal gas constant (at 70°F) 380 scf/lb-mole gc, gravitational constant 32.2 lb/sec2 molecular weight of flue gas (assume

it is essentially air) 29 lb/lb-mole

1 gal 0.134 ft3


Lesson 11 ___________________________________________________________________________________

11-58 2.0-7/98



2.0-7/98 11-59


Lesson 11 ___________________________________________________________________________________

11-60 2.0-7/98


2.0-7/98 11-61

Problem 3

Calculate the number of theoretical plates required by a scrubber given the data below. Also, estimate the total height of a column if tray spacing is 2 ft and the overall tray efficiency is 70%. Assume that the top spacing is also 2 ft.

QG, volumetric gas flow rate 15,000 acfm at 500oF Y1, concentration of HCI in entering gas 1,000 ppm or 0.01 mole fraction Y2, concentration of HCI in exiting gas 30 ppm or 0.00003 mole fraction m, Henry’s law equilibrium constant 1.1 X2, concentration of HCI in inlet liquid 0 QL, volumetric liquid flow rate 123 gal/min ρl, density of water 8.35 lb/gal R, the ideal gas constant (at 70°F) 380 scf/lb-mole molecular weight of water 18 lb/mole


Lesson 11 ___________________________________________________________________________________

11-62 2.0-7/98



2.0-7/98 11-63

Review Exercise Answers Answers to Questions

Part 1

1. e. a and b Packed towers and plate towers are better than venturi scrubbers and centrifugal scrubbers for removing gaseous pollutants from process exhaust streams.

2. c. Gaseous pollutant that is absorbed In the absorption process, the solute is the gaseous pollutant that is absorbed.

3. Solubility A very important factor affecting the amount of a pollutant that can be absorbed is its solubility.

4. b. Decreases In an absorber, as the temperature of the system increases, the amount of pollutant that can be absorbed decreases.

5. b. An equilibrium diagram A plot of the mole fraction of the solute in the liquid phase versus the mole fraction of the solute in the gas phase is called an equilibrium diagram.

6. d. y = H'x One form of the equation for Henry's law is:

y = H'x

7. a. Smaller In describing the solubility of various gases in water, the smaller Henry's law constant is, the more soluble the gas is.

Part 2

8. b. Mass-transfer In the double-resistance, or two-film theory, a mass transfer zone exists that includes a gas and liquid phase on either side of the interface.

9. True The two-film theory implies that all resistance to movement occurs when the molecule (gaseous pollutant) is diffusing through the gas and liquid films.

10. c. Partial pressure In absorption equations, the concentration of a gaseous pollutant is usually expressed by its partial pressure.

11. d. Overall mass-transfer coefficient based on the gas phase In calculating the rate of mass transfer of pollutant A (i.e. NA) using the equation

( )N K p pA OG AG A*= − , the term KOG is the overall mass-transfer coefficient based on the

gas phase.

Lesson 11 ___________________________________________________________________________________

11-64 2.0-7/98

12. True Overall mass-transfer coefficients are only valid when a plot of the equilibrium data yields an equilibrium line that is straight.

Part 3

13. a. Material balance In absorption calculations, a material balance equates the gas and liquid concentrations coming into the absorber with the gas and liquid concentrations going out of the absorber.

14. a. Small In air pollution calculations, the mass of the pollutant is usually very small compared to the mass of exhaust gas being treated and the mass of the liquid used in the absorber.

15. c. Minimum operating line In the graph below, the line AB is the minimum operating line.

16. c. Lm/Gm (actual) The slope of the actual operating line is Lm/Gm (actual).


2.0-7/98 11-65

17. True In the following figure, point B represents absorber conditions where the liquid leaving the absorber is saturated with the pollutant and can no longer absorb any additional pollutant, unless more liquid is added.

Part 4

18. a. 50 to 75% In designing a packed tower, the normal practice is to make the tower diameter so that the unit will operate at 50 to 75% of the flooding velocity rate.

19. True The Sherwood correlation can be used to calculate the tower diameter of a packed tower, if the minimum liquid rate, Lm, and the gas flow rate, G, through the absorber are known.

20. a. Transfer units In estimating packing height in a packed tower, the packing sections are broken up into discrete sections called transfer units.

21. HTU = height of a transfer unit NTU = number of transfer units In the equation, Z = HTU × NTU, which estimates the packing height, Z:

HTU = height of a transfer unit NTU = number of transfer units

Lesson 11 ___________________________________________________________________________________

11-66 2.0-7/98

22. True The Colburn diagram can be used to estimate the number of transfer units based on an overall gas-film coefficient, NOG, if the absorption factor (mGm/Lm), the inlet and outlet pollutant concentrations, and the liquid recycle concentrations are known. See Figure 11-14.

23. e. All of the above The height of a transfer unit is a function of the following: • Type of packing • Liquid and gas flow rates • Pollutant concentration and solubility • Liquid properties and system temperature

24. b. 0.3 to 1.2 m (1 to 4 ft) For most packed tower applications, the height of a transfer unit can be estimated to be 0.3 to 1.2 m (1 to 4 ft).

Part 5

25. c. Priming In a plate tower, if the gas velocity through the tower is too fast, liquid droplets become entrained in the gas stream, causing a condition called priming.

26. True For the purpose of determining a plate-tower diameter, priming in a plate tower is the same as the flooding point in a packed tower.

27. c. Number of theoretical plates In a plate tower, the following equation:

N

Y mXY mX

mGL

mGL

LmG

p

m

m

m

m

m

m

=

−−

−

+

ln

ln

1 2

2 21

is used to calculate the number of theoretical plates.

28. b. 65 to 80% In plate towers, the efficiency of each plate, or tray, is usually 65 to 80%.


2.0-7/98 11-67


Answer: To achieve the required removal efficiency, the operating liquid flow rate, QL, should equal 123 gal/min.

Solution:

1. Determine mole fraction of HCl in gas and liquid phases. (See Appendix A for help converting from ppm to mole fractions.)

Given: Y1 (gas in) = 1,000 ppm or 0.001 mole fraction Y2 (gas out) = 30 ppm or 0.00003 mole fraction X2 (liquid in) = 0 ppm X1 (liquid out) = unknown

2. Convert gas flow, QG, from acfm to molar units (Gm).

Gm = QG × temperature correction × R to standard (in absolute units)

mole/min-lb 8.21

scf 380mole-lb

500+ 46070 +460

minacf 15,000 = Gm

=

××

Lesson 11 ___________________________________________________________________________________

11-68 2.0-7/98

3. Calculate the concentration of HCl in the existing liquid (X1) at the minimum flow rate. At the minimum liquid flow rate, the gas mole fraction of HCl entering the absorber, Y1, will be in equilibrium with the liquid mole fraction leaving the absorber, X1. At equilibrium:

Y1 = m X1 X1 = Y1/m

Given: m = 1.1, Henry’s law equilibrium constant

X1 = (0.001)/1.1 X1 = 0.00091

4. Compute the minimum Lm/Gm using the following equation:

( )

minm

m

minm

m

21m

m21

GL=1.07

)000091.0(GL=0.00003001.0

XXGLYY

−

−

−=−

5. Calculate the minimum liquid flow rate, in gallons per minute.

(Lm/Gm)min = 1.07 (Lm)min = Gm (1.07)

From step 2: Gm = 21.8 lb-mole/min

(Lm)min = (21.8 lb-mole/min) 1.07 = 37 lb-mole/min

Convert from molar flow rate

minmole-lb to volumetric flow rate (gal/min).

min

.imum Q gal

L= × ×

=

37 lb - molemin

18 lblb - mole lb

gal / min8 35

82


2.0-7/98 11-69

6. Calculate operating liquid flow rate, QL.

Operating Q = 1.5 minimum liquid flow rate

= 1.5 82 gal / min= 123 gal / min

L ××

Lesson 11 ___________________________________________________________________________________

11-70 2.0-7/98


2.0-7/98 11-71


Answer: No, the existing tower (as is) cannot accommodate additional flows.

Solution:

1. Convert the proposed gas and liquid volumetric flow rates to mass units.

Given: Q = 12,500 acfm at 70 F

G = 12,500 ftmin

- mole380 ft

29 lb- mole

954 lb / min

Given: Q = 125 gal / min

L = 125 galmin

0.134 ft gal

lbft

= 1,072 lb / min

G

3

3

L

3

3

°

× ×

=

× ×

lblb

64

2. Calculate the abscissa in the flooding curve.

Use equation 11-16:

Abscissa = LG

×

ρρ

g

l

0.5

Given: ρg = 0.075 lb/ft3, the density of gas ρl = 64 lb/ft3, the density of liquid

From step 1: L = 1.072 lb/min G = 954 lb/min

0.038lb/ft 64

lb/ft 0.075 lb/min 954lb/min 1,072Abscissa

5.0

3

3

=

=

3. Calculate the area of the tower using one of the following two equations.

A =

A = 0.7854 d2

πr2

×

Lesson 11 ___________________________________________________________________________________

11-72 2.0-7/98

Where: A = area of cross-section of tower, m2 (ft2) r = radius of tower, m (ft) d = diameter of tower, m (ft)

Given: d = 4 ft

A = (0.7854) (4 ft)

12.56 ft

2

2=

4. Next, calculate the superficial gas velocity (G').

′G = G /A

From step 1: G = 954 lb/min From step 3: A = 12.56 ft2

′ × ×

= •

G = lbmin sec

112.56 ft

1.27 lb / sec ft

2

2

95460

min

5. Calculate ε, the ordinate in the flooding curve.

ε φ µρ ρ

= G 2′ Fg

l

g l c

0.2

Given: F = 45 ft3, packing factor φ = 1.01, ratio of the specific gravity of scrubbing liquid to that of water µl = 0.018 centipoise, viscosity of liquid ρg = 0.075 lb/ft3, the density of gas ρl = 64 lb/ft3, the density of liquid gc = 32.2 lb/sec2, the gravitational constant

From step 4: G' = 1.27 lb/sec•ft2

( ) ( )( )( )( )( )( )ε =

lb / sec ft ft 1.01 0.018 centipoise

lb / ft lb / ft lb / sec

2 3

3 3 2

127 45

0 075 64 32 2

0 21

2 0.2.

. .

.

•

=


2.0-7/98 11-73

From the coordinates on the graph in Figure 11-11, the new operating point (x = 0.038 and y = 0.21) would be above the flooding line and this is unacceptable. Note that the facility could still increase throughput and utilize the same tower by switching to a different packing material with a lower packing factor. For example, by using 2 inch plastic Tellerettes with a factor (F) of 20, the new ε would be 0.09 which would be well within acceptable limits.

Lesson 11 ___________________________________________________________________________________

11-74 2.0-7/98


2.0-7/98 11-75


Answer: Number of theoretical plates = 6.08 Total height of column = 20 ft

Solution:

1. Convert gas and liquid volumetric flow rates (QG and QL) to molar units (Gm and Lm).

Gm = QG × temperature correction × R to standard (in absolute units)

G

L

m

m

= × ×

=

× ×

=

15,000 acfmin

460 + 70460 + 500

lb - mole380 scf

21.8 lb - mole / min

= galmin

8.35 lbgal

lb - mole18 lb

57 lb - mole / min

123

2. Calculate number of theoretical plates, Np.

N

Y mXY mX

mGL

mGL

LmG

p

m

m

m

m

m

m

=

−−

−

+

ln

ln

1 2

2 21

Given: Y1 (inlet gas) = 0.01 mole fraction Y2 (outlet gas) = 0.00003 mole fraction X2 = 0, concentration of HCI in inlet liquid m = 1.1, Henry’s law equilibrium constant

From step 1: Gm = 21.8 lb-mole/min Lm = 57 lb-mole/min

Lesson 11 ___________________________________________________________________________________

11-76 2.0-7/98

( )( ) ( )( )

( )( )

N

n 193.52

p =

−−

−

+

=

=

ln ..

. . . .

ln. .

.

0 01 00 00003 0

111 218

5711 218

57

5711 218

1

6081n 2.37

theoretical plates

3. Calculate the number of actual plates.

Actual plates = estimated theoretical plates

overall efficiency

Given: overall tray efficiency = 70%

From step 2: estimated number of theoretical plates = 6.08

Actual plates = 6.08/0.70 = 8.7 plates = 9 plates (since you can’t have a fractional plate)

4. Estimate the height of the tower, Z.

Z = (Number plates) × (tray spacing) + top spacing

Given: tray spacing = 2 ft top spacing = 2 ft

From step 3: number of actual plates = 9

Z = (9 × 2 ft) + 2 ft Z = 20 ft

Note this is a rather tall tower for this separation. By increasing the liquid flow the tower height could be reduced. For example, by doubling the liquid flow rate the tower height could be reduced to half the size.


2.0-7/98 11-77

Bibliography


Bhatia, M. V. 1977. Packed tower and absorption design. In P. N. Cheremisinoff and R. A. Young (Eds.), Air Pollution Control and Design Handbook. New York: Marcel Dekker.


Danckwerts, P. V. 1951. Industrial and Engineering Chemistry. 43:1460.

Diab, Y. S., and R. N. Maddox. 1982. Absorption. Chemical Engineering. 89:38-56.

Higbie, R. 1935. Transactions of AIChE. 31:365.

MacDonald, J. W. 1977. Packed wet scrubbers. In P. N. Cheremisinoff and R. A. Young (Eds.), Air Pollution Control and Design Handbook. Part 2. New York: Marcel Dekker.

Marchello, J. M. 1976. Control of Air Pollution Sources. New York: Marcel Dekker.

McCabe, W. L., and C. J. Smith. 1967. Unit Operations of Chemical Engineering. New York: McGraw-Hill.



Sherwood, K. T. and R. L. Pigford. 1952. Absorption and Extraction. New York: McGraw-Hill.

Theodore, L., and A. J. Buonicore. 1975. Industrial Control Equipment for Gaseous Pollutants. Vol. I. Cleveland: CRC Press.

Toor, H. L., and J. M. Marchello. 1958. Journal of AIChE. 4:97.

Treybal, R. E. 1968. Mass Transfer Operations. 2nd ed. New York: McGraw-Hill.

Whitman, W. G. 1923. Chemical and Metallurgical Engineering. 29:147.

Zenz, F. A. 1972. Designing gas absorption towers. Chemical Engineering. 79:120-138.

Lesson 11 ___________________________________________________________________________________

11-78 2.0-7/98

Documents

ePA-AIR POLLUTION-LESSON