0071455914_ar018

CHAPTER 3.4STONE & WEBSTER–INSTITUTFRANÇAIS DU PÉTROLE FLUID

RFCC PROCESS

Warren S. LetzschFCC Program Manager

Stone & Webster Inc.Houston, Texas

HISTORY

Stone & Webster (S&W), in association with Institut Français du Pétrole (IFP), is thelicenser of the S&W-IFP residual fluid catalytic cracking (R2R) process. The originalS&W-IFP R2R (reactor 2 regenerators) process was developed during the early 1980s byTotal Petroleum Inc. at its Arkansas City, Kansas, and Ardmore, Oklahoma, refineries.Because the development of this process saw heavy input from an operating company, unitoperability and mechanical durability were incorporated into the design to ensure smoothoperation and long run lengths. To process the heavy, viscous residual feedstocks, whichcan contain metals in high concentrations and produce relatively high amounts of coke, thedesign incorporates an advanced feed injection system, a unique regeneration strategy, anda catalyst transfer system which produces extremely stable catalyst circulation. Recenttechnology advances have been made in the areas of riser termination, reactant vaporquench, mix temperature control (MTC), and stripping.

Today 26 full-technology S&W-IFP RFCC units have been licensed worldwide(revamp and grassroots), more than all other RFCC licensers combined. Within the PacificRim, S&W-IFP’s 19 licensed units outnumber the competition by more than 2 to 1. From1980 to 2001, there were 20 operating S&W-IFP FCC units totaling more than 190 yearsof commercial operation. A licensed R2R, located in Japan, is shown in Fig. 3.4.1. A list-ing of all S&W-IFP (full-technology) licensed R2R units is shown in Table 3.4.1.

While the conception of this technology was based on processing residual feed, thetechnology has been proved and is widely accepted for processing lighter gas oil feed-stocks. Stone & Webster and IFP have ample experience revamping gas oil FCC units toupgrade the feed injection system, combustion air distributor, riser termination device, etc.At present, more than 60 FCC units are processing over 2,400,000 barrels per day (BPD)of FCC feed employing the S&W-IFP feed injection technology. In fact, S&W-IFP sys-tems have replaced feed injection systems of virtually every competing licenser and havealways provided measurable benefits.

3.71

Source: HANDBOOK OF PETROLEUM REFINING PROCESSES

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)Copyright © 2004 The McGraw-Hill Companies. All rights reserved.

Any use is subject to the Terms of Use as given at the website.

PROCESS DESCRIPTION

R2R Converter

Two configurations of the grassroots R2R unit are offered. The first, and the most com-mon, is the stacked regenerator version shown in Fig. 3.4.2, which minimizes plot space.

3.72 CATALYTIC CRACKING

FIGURE 3.4.1 SWI-IFP RFCC unit located in Japan. Photograph shows second- and first-stageregenerators and main fractionator. Note the external cyclones on the second-stage regenerator.

STONE & WEBSTER–INSTITUT FRANÇAIS DU PÉTROLE FLUID RFCC PROCESS



STONE & WEBSTER-INSTITUT FRANÇAIS DU PÉTROLE FLUID RFCC PROCESS 3.73

The second is side-by-side regenerator design configuration, which is discussed in “FCCRevamp to RFCC” below and is more typical of FCC units which have been revamped toRFCC or for units larger than 100,000 B/D.

The process flow will be presented by using the stacked version shown in Fig. 3.4.2.The RFCC utilizes a riser-reactor, catalyst stripper, first-stage regeneration vessel, second-stage regeneration vessel, catalyst withdrawal well, and catalyst transfer lines. Processflow for the side-by-side configuration is identical except for the catalyst transfer betweenthe first- and second-stage regenerators.

Fresh feed is finely atomized with dispersion steam and injected into the riser throughthe feed injection nozzles over a dense catalyst phase. The small droplets of feed contactthe freshly regenerated catalyst and instantaneously vaporize. The oil molecules intimate-ly mix with the catalyst particles and crack into lighter, more valuable products.

Mix temperature control nozzles inject a selected recycle stream which quenches thecatalyst and feed vapor. This feature allows control of the critical feed-catalyst mix zonetemperature independent of the riser outlet temperature and provides some cooling ofthe regenerator. Riser outlet temperature (ROT) is controlled by the regenerated catalystslide valve.

As the reaction mixture travels up the riser, the catalyst, steam, and hydrocarbon prod-uct mixture pass through a riser termination device. S&W-IFP currently offers a number

TABLE 3.4.1 S&W-IFP Full RFCC Technology Units

Refinery Location Capacity, BPSD Start-up

A Kansas 20,000 1981B Oklahoma 25,000/40,000* 1982C Canada 19,000 1985D Japan 40,000 1987E Australia 25,000 1987F Canada 25,000 1987G China 23,000 1987H China 21,000 1989I China 28,000 1990J China 21,000 1990K China 21,000 1991L Japan 30,000 1992M Japan 31,600 1994N Uruguay 9,000 1994O Singapore 24,000 1995P Korea 50,000 1995Q Korea 30,000 1995R Thailand 37,000 1996S India 15,000 1997T Canada 65,000 2000U India 15,000 2001V India 26,000 2002W India 60,000 2003X India 65,000 2003Y Europe 30,000 2004Z Vietnam 65,000 2004

*Design capacity was 40,000 BPSD. Currently operating at 25,000BPSD.

Note: BPSD � barrels per stream-day.




FIGURE 3.4.2 S&W IFP RFCC unit process flow diagram.









of patented technologies for this service. These include rough-cut cyclones with extendedoutlet tubes, a linear disengaging device (LD2), a reactor separator stripper (RS2), andclose-coupled cyclones. These devices quickly disengage the catalyst from the steam andproduct vapors. Reactant vapors may be quenched after the initial catalyst-vapor separa-tion, minimizing thermal product degradation reactions. Reactant vapors are then ductedto the top of the reactor near the reactor cyclone inlets, while catalyst is discharged into thestripper through a pair of catalyst diplegs.

This ducting minimizes the vapor residence time and undesirable secondary thermalreactions in the vessel. The vapors and entrained catalyst pass through single-stage high-efficiency cyclones. Reactor products, inerts, steam, and a minute amount of catalyst flowinto the base of the main fractionator and are separated into various product streams.

Below each dipleg of the primary separator, a steam ring can be added to ensure smoothcatalyst flow out of the bottom. The stripper portion of this vessel can utilize four baffledstages or contain a proprietary packing material. Steam from the main steam ring fluidizesthe catalyst bed, displaces the entrained hydrocarbons, and strips the adsorbed hydrocar-bons from the catalyst before it enters the regeneration system. A steam fluffing ring, locat-ed in the bottom head of the stripper, keeps the catalyst properly fluidized and ensuressmooth catalyst flow through the spent catalyst transfer line.

Stripped catalyst leaves the stripper through the 45° slanted withdrawal nozzle and thenenters a vertical standpipe. The spent catalyst flows down through this standpipe and intoa second 45° lateral section that extends into the first-stage regenerator. The spent catalystslide valve is located near the top of this lower 45° transfer line and controls the catalystbed level in the stripper. Careful aeration of the catalyst standpipe ensures proper headbuildup and smooth catalyst flow. The flow rates from the aeration taps are adjustable tomaintain stable standpipe density for different catalyst circulation rates or different cata-lyst types. The catalyst enters the first-stage regenerator through a catalyst distributorwhich disperses the catalyst onto and across the bed surface.

Catalyst and combustion air flow countercurrently within first-stage regenerator vessel.Combustion air is distributed into the regenerator vessel by air rings. These air rings pro-vide even air distribution across the bed, resulting in proper fluidization and combustion.A pipe grid can be used as well. Partially regenerated catalyst exits near the bottom of thevessel through a hollow stem plug valve which controls the first-stage regenerator bed lev-el. A lift line conveys the partially regenerated catalyst from the first-stage regenerator tothe second stage, utilizing air injected into the line through the hollow stem of the plugvalve. Carbon monoxide–rich flue gases exit the regenerator through two-stage high-effi-ciency cyclones.

The operational severity of the first-stage regeneration is intentionally mild due to par-tial combustion. Low temperature results in the catalyst maintaining higher surface areaand activity levels. The coke burn percentage can be varied by shifting the burn to the sec-ond-stage regenerator, giving the RFCC the operating flexibility for residual as well as gasoil feedstocks. For residual feed, nearly 70 percent of the coke is burned in the first-stageregenerator while approximately 50 percent is burned during gas oil operation. Essentiallyall the hydrogen on the coke is burned off the coke in the first-stage regenerator; this step,coupled with low regenerator temperature, minimizes hydrothermal deactivation of thecatalyst.

As the catalyst enters the second-stage regeneration vessel, below the combustion airring, a mushroom grid distributes the catalyst evenly across the bottom head. This grid dis-tributor on the top of the lift line ensures proper distribution of air and catalyst. In the sec-ond-stage regenerator, the remaining carbon on the catalyst is completely burned off withexcess oxygen, resulting in a higher temperature compared to the first-stage regenerator.An air ring in this regenerator distributes a portion of the combustion air, while the lift airprovides the remainder of the air. With most of the hydrogen burned in the first stage,





moisture content in the gases in the second-stage regenerator is low. This allows highertemperatures in the second-stage regenerator without causing excessive hydrothermal cat-alyst deactivation.

The second-stage regenerator vessel has minimum internals, which increases the met-allurgical temperature limitations. Flue gas leaving the regenerator passes through two-stage external cyclones for catalyst removal. The recovered catalyst is returned to theregenerator via diplegs, and the flue gas flows to the energy recovery section.

If the feed Conradson carbon residue is greater than 7.0 wt %, a catalyst cooler will berequired for the second-stage regenerator (shown as optional in Fig. 3.4.2) to reduce thesecond-stage regenerator temperature to less than 760°C. A dense-phase catalyst coolerwill withdraw catalyst and return it, via an air lift riser, to just beneath the combustion airring. Heat is recovered from the catalyst by generating saturated high-pressure steam.Large adjustments in the catalyst cooler duty can be made by varying the catalyst circula-tion rate through the catalyst cooler. Fine catalyst cooler duty corrections can be made byadjusting the fluidization air rate in the cooler. Internal cyclones can be used in the secondregeneration, in this case due to the 760°C maximum temperature limit.

Hot regenerated catalyst flows into a withdrawal well from the second-stage regenera-tor. The withdrawal well allows the catalyst to deaerate properly to standpipe densitybefore entering the vertical regenerated catalyst standpipe. This design ensures smooth andeven catalyst flow down the standpipe. Aeration taps, located stepwise down the standpipe,serve to reaerate the catalyst and replace gas volume lost by compression. Flow rates forthe aeration taps are adjustable to maintain desirable standpipe density, allowing for dif-ferences in catalyst circulation rates or catalyst types. The catalyst passes through theregenerated catalyst slide valve, which controls the reactor temperature by regulating theamount of hot regenerated catalyst to the reactor. The catalyst then flows down the 45°slanted wye section to the riser base. Fluidization in the wye section ensures stable andsmooth dense-phase catalyst flow to the feed injection zone. A straight vertical sectionbelow the feed nozzles stabilizes the catalyst flow before feed injection and serves as areverse seal, preventing oil flow reversal.

Flue Gas Handling

Each RFCC flue gas system is generally unique from one unit to the next because of localenvironmental requirements and refiner preference. An example of a basic flue gas han-dling system is shown in Fig. 3.4.3. The flue gas line from the second regenerator will havea flue gas slide valve and orifice chamber. The first-stage regenerator flue gas slide valve(FGSV) controls the pressure differential between the two regenerator vessels, while thesecond-stage regenerator FGSV directly controls the pressure of the second-stage regen-erator.

Large-capacity RFCC units may employ a power recovery train and tertiary cyclonesystem on the first-stage regenerator flue gas stream to drive the air blower. Depending onlocal particulate emission requirements, an electrostatic precipitator (ESP) or other partic-ulate recovery device such as a third-stage cyclone system or flue gas scrubber may beused to recover entrained particulates. Here a flue gas scrubber is included. More stringentSOx and NOx emission requirements may necessitate a flue gas scrubber, SOx capturingcatalyst additive, or similar process for SOx recovery and/or a selective catalytic reduction(SCR) unit for NOx mitigation.

A CO incinerator is located just downstream of the first-stage regenerator power recov-ery equipment and oxidizes all CO gases to CO2, utilizing fuel gas and combustion air. Exittemperature is typically 980°C with 1 percent excess O2. Gases from the CO incineratorcombine with second-stage regenerator flue gases and enter a flue gas cooler, where heat






FIGURE 3.4.3 Flue gas handling process flow diagram.








is recovered as high-pressure superheated steam. After going through the wet scrubber, theflue gases are finally dispersed into the atmosphere through a stack.

Catalyst Handling

The RFCC catalyst handling system has three separate and unique functions:

● Spent catalyst storage and withdrawal● Fresh catalyst storage and addition● Equilibrium catalyst storage and addition

The spent hopper receives hot catalyst intermittently from the second-stage regeneratorto maintain proper catalyst inventory during operation. In addition, the spent catalyst hopperis used to unload, store, and then refill the entire catalyst inventory during R2R shutdowns.

The fresh catalyst hopper provides storage of catalyst for daily makeup. A loader, locat-ed just beneath the hopper, loads fresh catalyst from the hopper to the first-stage regener-ator. Fresh catalyst makeup is based on maintaining optimal unit catalyst activity andshould be on a continuous basis.

Unique to R2R designs is a third hopper which is used for equilibrium catalyst. Likethe fresh catalyst hopper, the equilibrium catalyst hopper provides storage of catalyst fordaily makeup. Equilibrium catalyst serves to flush metals from the unit equilibrium cata-lyst in processing of residual feeds with high metal content. However, equilibrium catalystusually does not contribute much to cracking activity.1 As a result, the equilibrium catalystaddition rate is based on targeted metal content on unit equilibrium catalyst, while thefresh catalyst makeup rate is based on maintaining unit catalyst activity. An equilibriumcatalyst loader is located just beneath the hopper which supplies equilibrium catalyst to thefirst-stage regenerator. It is critical that the equilibrium catalyst be compatible with resid-ual operations and usually should not be more than one-third of the total catalyst makeup.

RFCC FEEDSTOCKS

The most significant advantage of the S&W-IFP R2R process is the flexibility to processa wide range of feedstocks. Table 3.4.2 lists the range of feedstock properties which havebeen successfully processed in the S&W-IFP R2R.

Feedstock to the R2R can take a variety of forms, from a hydrotreated vacuum gas oil(VGO) to a virgin highly aromatic atmospheric tower bottoms (ATB) such as ArabianLight ATB. The R2R feedstock can also be a blend of various unit streams such as a VGOplus coker vacuum gas oil, vacuum tower bottoms (VTB), deasphalted oil (DAO), slopwax, or lube extract. In fact, the number of possible feed constituents to the R2R is quitelarge since almost any hydrocarbon stream can be considered as a potential R2R feed.

What gives the R2R unit the flexibility to process this wide range of feedstocks is pri-marily the two-stage regenerator design and the minimization of delta coke inherent in thefeed injection, catalyst-vapor separator, and stripper designs. A common index which indi-cates a feedstock’s tendency to produce feed-derived coke is the Conradson carbonresidue (CCR). As the residual content of a feedstock increases, so does the CCR amount.Table 3.4.3 compares the maximum CCR levels that can be processed in a two-stage regen-erator and in a single-stage regenerator.

Recently the increasing need to convert the bottom of the barrel into clean transporta-tion fuels (low-sulfur) coupled with the decreasing availability of sweet crudes has ignit-





ed an interest in hydrodesulfurization and residual hydrodesulfurization (RDS). Reynolds,Brown, and Silverman showed that it is economically feasible to upgrade VTB using a vac-uum RDS (VRDS) process into feedstock for the S&W-IFP R2R unit.2 Processing 100 per-cent VTB in the RFCC is considerably more attractive than processing it in traditionalthermal processors such as delayed and fluid cokers since catalyst yields are superior tothermally derived products.

Operating Conditions

Like traditional FCC units, the S&W-IFP R2R unit can be operated in maximum distillate,maximum gasoline, or maximum olefin operational modes. Conversion is decreased formaximum distillate operations and increased for the maximum olefin operations by adjust-ing the riser outlet temperature and catalyst activity. Typical range of ROTs required for thethree operation modes are as follows: maximum distillate, 510°C ROT minimum; maxi-mum gasoline, 510 to 530°C ROT; and maximum olefins, 530 to 560°C ROT. For maxi-mum distillate operation, MTC, discussed in “Mix Temperature Control” below, is criticalin order to maintain the required mix temperature to ensure vaporization of the heavy resid-ual feed at lower riser outlet temperatures. Likewise, reactant vapor quench technology,


TABLE 3.4.2 Commercial RFCC FeedstockOperation Experience

Property Range

Gravity, °API 18–29Conradson carbon residue, wt % 0–9Sulfur, wt % 0.1–2.4Nitrogen, wt % 0.05–0.35Metals (Ni � V), wt ppm 0–50540°C � components, LV % 0–58

Note: °API � degrees on the American PetroleumInstitute scale; LV � liquid volume.

TABLE 3.4.3 Heavy-Feed Processing Capabilities of VariousHeat Rejection Systems

System Conradson carbon residue, wt %

Single-stage regenerator

Full combustion 2.5Partial combustion 3.5Partial combustion � MTC 4.0Catalyst cooler* 10.0

Two-stage regenerator

Alone 6.0With MTC 7.0Catalyst cooler* 10.0

*Economic rather than technical limit.




discussed in “BP Product Vapor Quench” below, is especially critical during maximumolefin operations to reduce postriser thermal cracking at the elevated reactor temperatures.

Other typical operating conditions of the R2R unit are shown in Table 3.4.4. Examples ofobserved commercial product yields from S&W-IFP R2R units are shown in Table 3.4.5.

RFCC CATALYST

Catalyst Type

A successful residual cracking operation depends not only on the mechanical design of theconverter but also on the catalyst selection. To maximize the amount of residual content inthe RFCC feed, a low-delta-coke catalyst must be employed. Delta coke is defined as

Delta coke � wt % carbon on spent catalyst � wt % CRC

where CRC � carbon on regenerated catalyst, or as

Delta coke �

Delta coke is a very popular index and, when increased, can cause significant rises inregenerator temperature, ultimately reducing the amount of residual feed that can beprocessed. Commercial delta coke consists of the following components:

coke wt % feed��catalyst/oil ratio


TABLE 3.4.4 Typical RFCC OperatingConditions

Reactor

Pressure, kg/cm2 gage 1.1–2.1Temperature, °C 510–550MTC recycle, vol % feed 10–25Feed dispersion steam, wt % feed 2.5–7.0Stripping steam, kg/1000 kg 2.0–5.0

First-stage regenerator

Pressure, kg/cm2 gage 1.4–2.5Temperature, °C 620–690CO/CO2 0.3–1.0O2, vol % 0.2Coke, burn, wt % 50–70

Second-stage regenerator

Pressure, kg/cm2 gage 0.7–1.4*Temperature, °C 675–760O2, vol % 2.0Coke burn, wt % 30–50

*Second-stage regenerator pressures reflect a stackedregenerator configuration. For side-by-side regenerator con-figurations, the second-stage regenerator pressure would besimilar to the first-stage regenerator pressure.




● Catalytic coke (deposited slowly as a result of the catalytic reaction)● Feed-derived coke (deposited quickly and dependent on feed CCR)● Occluded coke (entrained hydrocarbons)● Contaminant coke (coke produced as a result of metal contaminants)

Because the feed-derived coke becomes a large contributor to the overall delta coke inprocessing residual feeds, it is crucial that the overall delta coke be minimized in a resid-ual FCC operation.

Stone & Webster-IFP typically recommends a catalyst with the following properties,which characterize it as a low-delta-coke catalyst:

● Low rare-earth ultrastable Y (USY) zeolite● Equilibrium microactivity test (MAT) activity 60 to 65● Low-delta-coke matrix

At high metal loadings the operator may also consider catalyst with vanadium trapsand/or nickel passivators.

Catalyst Addition

Virgin residual feeds may contain large amounts of metals, which ultimately are deposit-ed on the catalyst. Because of the mild two-stage regenerators, the catalyst metal contentcan be allowed to approach 10,000 wt ppm (Ni � V) before product yields are significantlyaffected. For an RFCC operation, catalyst addition is based on maintaining catalyst activ-ity as well as metals on catalyst as opposed to maintaining only activity for typical FCCgas oil operations. The most economical way to maintain both activity and metals is to addboth fresh catalyst and purchased equilibrium catalyst. Equilibrium catalyst is an effectivemetal-flushing agent; however, equilibrium catalyst does not contribute much crackingactivity.1 As a result, equilibrium catalyst is added with fresh catalyst in order to econom-ically control both the unit catalyst activity and metal content. Care must be taken that theequilibrium catalyst chosen is compatible with residual operations and should not be morethan one-third of the total catalyst additions.


TABLE 3.4.5 Commercial RFCC Product Yields

Unit (year)

A (1987) B (1993)

Feed properties:540°C � components, LV % 36 58

CCR, wt % 5.9 4.9Gravity, °API 22.3 25.1Yield (LV %):

Dry gas, wt % 4.3 3.2C3-C4 24.9 30.5

Gasoline 60.2 61.5Light cycle oil 17.5 14.0Slurry 6.6 4.9Coke, wt % 7.8 8.0Conversion 75.9 81.1




TWO-STAGE REGENERATION

In the S&W-IFP two-stage regeneration process, the catalyst is regenerated in two steps:50 to 70 percent in the first-stage regenerator and the balance in the second-stage regener-ator. The first-stage regeneration is controlled by operating the first stage in an oxygen-deficient environment, producing significant amounts of carbon monoxide. Since the heatof combustion of carbon to carbon monoxide is less than one-third of that for combustionto carbon dioxide, much less heat is transferred to the catalyst than in a single-stage full-combustion regenerator. For example, a 30,000 BPSD R2R unit with a feed gravity of22.5° API and a coke yield of 7.5 wt % at 66 percent coke burn in the first-stage regener-ator has reduced the heat transferred to the catalyst by approximately 25 � 106 kcal/h overa full-burn single-stage regenerator.

The remaining carbon on the catalyst is burned in the second-stage regenerator in full-combustion mode. Because of the possible elevated temperature, external cyclones areemployed to minimize regenerator internals and allow carbon-steel construction.

Comparison of Two-Stage and Single-Stage Regeneration with a CatalystCooler

Although both systems operate to control regenerator temperatures, the principles of oper-ation are significantly different. The advantages of the two-stage regeneration systembecome apparent as the feed becomes heavier and/or its metal content increases. The ben-efits of a two-stage regeneration system over a single-stage system with a catalyst coolerare briefly described as follows.

Lower Catalyst Particle Temperature. A catalyst cooler removes heat after it isproduced inside the regenerator, while less heat is produced in the regenerator with atwo-stage regenerator design. This results in a lower catalyst particle temperatureduring combustion, reducing overall catalyst deactivation. Since the combustion isoccurring in two steps, the combustion severity of each step is low. In the first-stageregenerator, the catalyst enters the bed from the top through the spent catalystdistributor while the combustion air enters the bed at the bottom of the vessel. Thiscountercurrent movement of catalyst and air prevents the contacting of spent catalyst(high carbon) with fresh air containing 21 percent oxygen. All these factors result inlower catalyst thermal deactivation for the two-stage regeneration system.

Lower Hydrothermal Deactivation. While the catalyst is only partially regenerated inthe first stage, most of the water formed by the combustion of the hydrogen in thecoke is removed in this vessel. Figure 3.4.4 shows the percentage of hydrogen on cokeburn as a function of carbon burn. Since the temperature of the first-stage regeneratoris low, catalyst hydrothermal deactivation is significantly reduced. In the second-stageregenerator, where the bed temperature is high, moisture is minimal and does not posea significant hydrothermal deactivation risk for the catalyst.

Better Metal Resistance. When refiners run high-metal feeds, it is very advantageousto be able to run with high metal levels on the equilibrium catalyst. Studies haveclearly shown that high metal levels (particularly vanadium) lead to excessive catalystdeactivation in the presence of steam and oxygen. Since most of the steam in aregenerator comes from the hydrogen in the coke, the moisture content can becalculated in a straightforward manner. For a single-stage regenerator this will usuallybe more than 10 percent moisture. When steam and vanadium react in the presence of





oxygen, vanadic acid is formed, which attacks the alumina in the catalyst zeolitestructure. Massive dealumination causes the collapse of the zeolite structure, and theresulting catalyst is left with little activity. The equations

2V � O2 → V2O5

andV2O5 � 3H2O → 2VO (OH) 3

Vanadic acid

describe the generation of vanadic acid. As a result, catalyst in a single-stage regeneratoroperating in the presence of excess oxygen and steam is prone to vanadic acid attack. AlsoV2O5 has a very low melting temperature and can be liquid at typical regenerator condi-tions.

Staging the regeneration can be particularly effective in this situation. In the first-stageregenerator, most of the hydrogen (and subsequent water vapor) is removed at low tem-perature without the presence of oxygen. This is followed by a full-burn second-stageregenerator where there is excess oxygen but very little moisture. Vanadium destruction ofthe catalyst structure is minimized, since very little V2O5 is present in the first-stage regen-erator because of the lack of oxygen and lower temperature, while vanadic acid is mini-mized in the second-stage regenerator by lack of water. In other words, the reaction

2V � O2 → V2O5

proceeds very slowly in the first-stage regenerator because of a lack of oxygen while thereaction

V2O5 � 3H2O → 2VO (OH)3

proceeds slowly in the second-stage regenerator because of low steam content.The two-stage regeneration is clearly less severe with regard to catalyst deactivation;

and this, coupled with the newer generation of catalyst with vanadium traps, will allowrefiners to run heavier crudes more efficiently and economically than ever before.

5�2

5�2


FIGURE 3.4.4 Hydrogen and hydrocarbon burnrates.




Catalyst Cooler System

The S&W-IFP heavy residual R2R units (feed CCR greater than 6.0 wt % or 7.0 wt % withMTC) design includes well-proven catalyst cooler technologies. The same catalyst coolerdesigns can also be provided for an existing FCC regenerator. These designs are operatingin more than 20 crackers, with several more in the design and construction stages. A fewfeatures of the Stone & Webster catalyst cooler systems are

● Dense-phase, downward catalyst flow● Slide-valve-controlled catalyst circulation● Turndown capability from 0 to 100 percent● No tube sheet required● High mechanical reliability● Cold wall design● All carbon-steel construction● High heat-transfer and low tube wall temperature● 100 percent on-stream factor

Catalyst cooler duties can range from as low as 2 � 106 kcal/h up to 35 � 106 kcal/h.In the event that more than a 35 � 106 kcal/h cooler is required, multiple catalyst coolerscan be employed on a regenerator.

A schematic diagram for a catalyst cooler coupled to a regenerator (second-stage regen-erator in a two-stage regeneration system) is shown in Fig. 3.4.5. Catalyst level inside thecooler is controlled by the inlet catalyst slide valve. Gross temperature control of theregenerator is achieved by the bottom catalyst slide valve, and fine temperature control isachieved by the cooler fluidization air. An optional design eliminates the inlet slide valveand operates with the cooler full of catalyst.

S&W-IFP Technology Features

S&W-IFP offers many technology features which improve the product selectivity, unitcapacity, and operability of our R2R designs. These same features are available to refinerswho wish to upgrade existing FCC units. In fact, various aspects of the S&W-IFP FCCprocess have been applied to more than 100 FCC revamps.

Feed Injection System. The feedstock injection system and lower portion of the feedriser are the most critical parts of the R2R/FCC. The earlier pioneering and patenteddevelopments of Total Petroleum Inc. have convinced the refining industry of thevalue and benefits of advanced feed injection. Basic elements of the S&W-IFP feedinjection system are as follows:

● Dense-phase flow of catalyst up to the feed injection point, employing small quantitiesof steam to stabilize catalyst flow and maintain a uniform catalyst flux across the riser

● Atomization of the feed external to the riser using steam in a simple but efficient two-fluid nozzle not involving complex internals subject to plugging and erosion

● Introduction of feed into an upward-flowing dense phase of catalyst in a manner whichachieves the penetration and turbulence necessary to accomplish rapid heat transfer fromthe hot catalyst to the fine oil droplets, ensuring rapid vaporization





Table 3.4.6 lists actual commercial product yield improvements observed after replacingolder feed injection systems with the S&W-IFP design.

Basic elements of the S&W-IFP feed injection nozzles are shown in Fig. 3.4.6. Thistwo-fluid nozzle works by injecting oil under pressure against a target plate to break theoil into thin sheets that the steam shears as it moves across and through the oil. The oil mistis injected into the riser through a specially designed tip which ensures maximum risercoverage without impinging and damaging the riser wall.

This feed injection system was developed for residual FCC operations where the resid-ual feed is highly viscous and difficult to atomize. To provide adequate atomization of theresidual feedstock, this nozzle design uses oil pressure, steam pressure, and steam rate. Forvacuum gas oil feedstocks which are considerably easier to atomize, oil pressure and steamrates can be significantly reduced below those of residual operations.

Mix Temperature Control

An important concern in processing heavy feedstocks with substantial amounts of residualoil is to ensure rapid feed vaporization. This is critical to minimize unnecessary coke dep-osition due to incomplete vaporization. Unfortunately, in conventional designs, the mixtemperature is essentially dependent on the riser outlet temperature. Typically the mix tem-perature is about 20 to 40°C higher than the riser outlet temperature and can be changedonly marginally by the catalyst/oil ratio.

In many cases, raising the riser outlet temperature to adjust the mix temperature is notdesirable since this may result in undesirable nonselective cracking reactions with high


HP Steam

BFW

Blowdown

CatCooler

FluidizationAir

Lift Air

FIGURE 3.4.5 General catalyst cooler arrangement.




production of dry gas. The problem becomes even more critical with less severe operatingconditions for maximum distillate production. To address this problem and make the aboveobjectives compatible with each other, the riser outlet temperature must be independentlyadjusted. This is achieved with MTC developed and patented by IFP-Total.

MTC is performed by recycling a selected liquid cut downstream of the fresh feedinjection zone. It roughly separates the riser into two reaction zones:

● An upstream zone, characterized by high temperature, high catalyst/oil ratio, and veryshort contact time

● A downstream zone, where the reaction proceeds under more conventional and mildercatalytic cracking conditions

Creating two separate cracking zones in the riser permits fine tuning of the feed vapor-ization and cracking to desired products. With MTC, it is possible to raise the mix tem-perature while maintaining or even lowering the riser outlet temperature. Figure 3.4.7illustrates the MTC nozzle arrangement and the three temperature zones.


FIGURE 3.4.6 S&W IFP feed injection nozzle.

TABLE 3.4.6 Incremental S&W-IFP FeedInjection System Product Yields

Delta yields

Product Unit A Unit B

Dry gas, wt % �0.0 �1.3C3/C4, LV % �1.5 �1.5Gasoline, LV % �3.4 �6.2Light cycle oil, LV % �1.6 �4.5Slurry, LV % �6.5 �0.3Coke, wt % �0.0 �0.1Conversion, LV % �4.9 �4.8




The primary objective of the MTC system is to provide an independent control of themix temperature. However, as a heat sink device similar to a catalyst cooler, MTC can beused to increase the amount of residual feed processed in the unit.

Riser Termination Device

Numerous studies have shown that postriser vapor residence time leads to thermal crack-ing and continued catalyst cracking in the reactor vessel. Unfortunately, these postriservapor-phase reactions are extremely nonselective and lead to degradation of valuable liq-uid products, high dry-gas make, and high hydrogen transfer in liquefied petroleum gas(LPG) olefins (low olefin selectivity). The factors that contribute to these phenomena aretemperature, time, and surface area. S&W-IFP’s riser termination technology is designedto control all three factors.

S&W-IFP offer a variety of termination technologies to effectively control postrisercracking. Rough-cut cyclones with modified outlet tubes, a linear disengaging device(LD2), and a close-coupled version referred to as a reactor separator-stripper (RS2) have allbeen successfully used. A close-coupled system that includes a dilute phase stripper hasalso given state-of-the-art performance. Two of these separators are shown in Fig. 3.4.8.

These technologies offer the refiner options that are easy to operate, give low catalystcarryover, and can provide dilute phase stripping.

BP Product Vapor Quench

This technology was developed and patented by Amoco (now BP) and is offered to theindustry by Stone & Webster and IFP under an exclusive arrangement. Reactant vapors are


FIGURE 3.4.7 Mix zone temperature control.




quenched after leaving the riser termination system substantially free of catalyst, by inject-ing a light cycle oil quench. By employing quench technology, nonselective thermal reac-tions are arrested, resulting in higher gasoline yields and lower dry gas production. Inaddition, use of the quench technology further preserves the LPG olefins and gasolineoctane, minimizes the formation of diolefins, and enhances gasoline stability.

The effectiveness of vapor quench is shown in Table 3.4.7. The data indicate that areduction in dry gas production is observed even at low riser outlet temperatures. Asexpected, the impact of quench in terms of dry gas reduction and gasoline yield improve-ment is more marked at higher temperatures.

The combination of the S&W-IFP riser termination devices and Amoco’s vapor quenchvirtually eliminates undesirable postriser reactions.

Stripper Design

The traditional disk and doughnut stripping technology has been successfully used ingrassroots and revamp designs. However, these designs lose efficiency when the catalystflux rates approach 1100 to 1200 lb/ft2 � min. Structured packing can be used in the placeof the disk and doughnuts or shed decks, with the result being

1. More stages of stripping

2. Use of the entire cross-sectional area of the stripper for catalyst flow

3. Less catalyst entrained to the Rx cyclones

4. Reuse of the existing stripper shell

For a new FCC unit, a stripper can be designed that will operate satisfactorily at 2 to 3times the design catalyst flux. The improved contacting is due to the lower catalyst veloc-ity going down the stripper which allows smaller steam bubbles to rise rather than havingthem either coalesce into larger bubbles to go up the stripper or be simply swept along withthe catalyst to the bottom exit of the vessel.


Linear Disengager Reactor Separator Design with Integral Stripper

FIGURE 3.4.8 Linear disengager and reactor separator and stripper.




The two types of strippers are shown in Fig. 3.4.9. Since the stripper is a multistagecontacting tower, putting in more efficient contactors improves the overall performance.This is completely analogous to replacing trays with packing in a distillation tower.

MECHANICAL DESIGN FEATURES

The S&W-IFP R2R mechanical design philosophy is based on multiple concepts to pro-vide high reliability and maintainability with longer run lengths. Mechanical design effortshave focused on areas of an FCC unit that have historically caused high maintenance costsand increased downtime. These efforts have resulted in an overall mechanical design capa-ble of providing up to 5 years of operation between turnarounds. Some of the features arediscussed here.

Cold Wall Design

The cold wall design concept is emphasized throughout the unit in the riser, reactor, regen-erators, catalyst cooler, external transfer lines, slide valves, and external cyclone. Internalrefractory insulation of vessel pressure parts sufficiently reduces the skin temperatures topermit use of less expensive and easier-to-maintain carbon-steel materials. Lower metaltemperatures result in less thermal expansion of the components, minimizing the need forexpansion joints to compensate for differential thermal expansion between interconnectedcomponents and transfer lines.

External surface areas of the pressure parts are exposed for on-line inspection, therebyreducing inspection and maintenance costs. The internal refractory protects the pressureshell from catalyst erosion, while metal hot spots can be readily detected before theyprogress to a potentially dangerous level.

Feed Nozzle Fabrication

The S&W-IFP proprietary feed injection nozzles are installed through sleeves in the riserwall. Erosion of the riser wall is avoided by careful selection of the entrance angle of thesleeve and the design of the nozzle spray angle. The nozzle tip and atomizing chamber aremade from erosion-resistant material to virtually eliminate wear. In the unlikely event oferosion, those surfaces exposed to erosive conditions are easily replaced and are designedso that normal maintenance can be performed during a scheduled turnaround with removalof the nozzle from the vessel sleeve. Typically, it is only necessary to inspect the nozzlesat turnaround, and only rarely is any maintenance required.


TABLE 3.4.7 Impact of Reactor Vapor Quenchon FCC Yields

Unit A Unit B Unit C

Temperature, °CRiser outlet 513 549 532After quench 484 519 494

Yield shifts, wt %Dry gas �0.23 �0.80 �0.66Gasoline �0.43 �1.80 �2.89




External Cyclones

Cold wall external cyclones are used on the second-stage regenerator to remove them fromthe internal, hot environment. The cyclones are attached directly to the cold wall regener-ator and the minimal differential thermal expansion is easily accommodated. The size andlength/diameter ratio of the external cyclones are not limited by the internal dimensions ofthe regenerator; therefore, more efficient cyclones can be designed with a shorter, lessexpensive regenerator. In addition, the external cyclones offer longer turnaround cycles,are insensitive to thermal excursions, and are subject to direct inspections while in opera-tion. The cyclones can be easily monitored for mechanical reliability by using infraredcameras and for process performance by monitoring the dipleg levels with level indicators.Internal cyclones could be used where second-stage temperatures are not expected toexceed 1400°F (760°C).

Combustion Air Rings

The S&W-IFP design utilizes proprietary combustion air rings instead of dome or pipegrids. The design provides optimum air distribution and mixing, both vertically and later-ally, and overcomes problems of material cracking, distributor erosion, and nozzle erosionexperienced with other designs. The use of properly designed nozzles and high-densityrefractory material on the rings eliminates all damage due to erosion. A combustion airring is shown in Fig. 3.4.10.

FCC REVAMP TO R2R (SECOND-STAGEREGENERATION ADDITION)

Adding a second-stage regenerator is an effective means of converting an existing FCCunit to residual service without losing throughput. To date, three FCC units have been


Disk & Doughnut Trays Structured Packing

FIGURE 3.4.9 Stripper geometric considerations.




revamped to include a second-stage regenerator and allow the processing of heavy resid-ual feedstocks. These designs retain the existing regenerator as the first-stage regeneratorand the reactor/stripper. A new second-stage regenerator, catalyst transfer lines, and a COincinerator; a new or supplemental air blower; and a revamp of the flue gas handling facil-ities are required. By operating the first-stage regenerator in partial combustion mode, asexplained earlier, no additional heat removal facilities will be required up to a feed CCRof 6.0 wt %. Shown in Fig. 3.4.11 is an FCC unit revamped to include a second-stageregenerator; the figure indicates both new and existing equipment.

REFERENCES

1. Raymond Mott, “FCC Catalyst Management for Resid Processing,” First FCC Forum, Stone &Webster Engineering Corporation, The Woodlands, Tex., May 11–13, 1994.

2. B. E. Reynolds, E. C. Brown, and M. A. Silverman, “Clean Gasoline via VRDS/RFCC,”Hydrocarbon Processing, April 1992, pp. 43–51.

3. Warren S. Letzsch, “Catalytic Cracking Update,” 4th Stone & Webster/IFP Licensors Forum,Houston, Tex., May 2–5, 2000.

4. Warren S. Letzsch, “Stone & Webster/IFP Your Catalytic Cracking Supermarket,” 12th AnnualStone & Webster Refining Seminar, San Francisco, Oct. 10, 2000.

5. Warren S. Letzsch, “Advanced Fluid Cracking Technologies,” NPRA Annual Meeting 2001, PaperAM-01-65.

6. Warren S. Letzsch, “Fluid Catalytic Cracking Meets Multiple Challenges,” NPRA AnnualMeeting 2002, Paper AM-02-26.


FIGURE 3.4.10 Combustion air ring.





FIGURE 3.4.11 Side-by-side regenerator RFCC revamp design.




Documents

0071455914_ar018