Dioxin Primer Prestn - EJnet.org · 2. Gas Phase Formation (Furnace to 500°C) From Related Chlorinated Precursors and from Simple Organics and Chlorine Donors (ca. 700°C) Formation

Chlorinated Dioxin and Furan Formation,

Control and Monitoring

Presented at

IIIICCCCCCCCRRRR MMMMeeeeeeeettttiiiinnnngggg

RRRReeeesssseeeeaaaarrrrcccchhhh TTTTrrrriiiiaaaannnngggglllleeee PPPPaaaarrrrkkkk

SSSSeeeepppptttteeeemmmmbbbbeeeerrrr 11117777,,,, 1111999999997777

Dr. Brian Gullett

EPA

(919) 541-1534

[email protected]

Dr. Randy Seeker

EER Corporation

(714) 859-8851

[email protected]

ICCR Meeting September 17,1997

AAAAcccckkkknnnnoooowwwwlllleeeeddddggggmmmmeeeennnnttttssss

❏ Jim Kilgroe, EPA Office of Research and Development

❏ Steve Lanier, EER Corporation, North Carolina

❏ Glenn England, EER Corporation, Irvine, CA


OOOOuuuuttttlllliiiinnnneeee ooooffff TTTTaaaallllkkkk

❏ Structures

❏ Concentrations of interest

❏ Measurement Techniques and Issues

❏ Fundamental Principles

❏ Dioxin and Furan Formation Chemistry

❏ Control Technologies

❏ Formation Conditions

❏ Monitoring Technologies

❏ Conclusions

❏ Recommendations


DDDDiiiiooooxxxxiiiinnnn aaaannnndddd FFFFuuuurrrraaaannnn SSSSttttrrrruuuuccccttttuuuurrrreeee

2,3,7,8 — Tetrachlorodibenzo(p)dioxin

OCl

Cl

Cl

Cl

5

2

3

98

7

6

10 1

4O

O

2,3,7,8 — Tetrachlorodibenzofuran

O

OCl

Cl

Cl

Cl

Dioxin

Furan

❏ There are 210 different congener configurations.

O1

2

346

7

89


CCCCoooonnnnggggeeeennnneeeerrrrssss TTTTooooxxxxiiiicccc EEEEqqqquuuuiiiivvvvaaaalllleeeennnntttt FFFFaaaaccccttttoooorrrr

Congener NATO Toxic Equivalent Factors (I-TEQ)

2378 TCDD 112378 PeCDD 0.5123478 HxCDD 0.1123678 HxCDD 0.1123789 HxCDD 0.11234678 HpCDD 0.01OCDD 0.0012378 TCDF 0.112378PeDF 0.0523478 PeCDF 0.5123478 HxCDF 0.1123789 HxCDF 0.1123678 HxCDF 0.1234678 Hx CDF 0.11234678 HpCDF 0.011234789 HpCDF 0.01OCDF 0.001

❏ 17 out of the 210 have toxic equivalency factors

World Health Organization (TEF)

“-” indicates no TEF because of lack of dataa) limited data setb) structural similarityc) QSAR modelling prediction from CYP1A induction (monkey, pig, chicken, or fish)d) no new data from 1993 reviewe) in vitro CYP1A inductionf) in vivo CYP1A induction after in ovo exposureg) QSAR modelling prediction from class specific TEFs


Relating TEQs to Total PCDD/PCDFT

EQ

(ng

/dsc

m @

7%

O2)

1 10 100 1000 10000

SD/FFSD/ESPESP

1000

100

10

1

0.11 10 100 1000 10000

FFESP

100

10

1

0.1

0.01

100

10

1

0.1

0.01

0.0010.1 1 10 100 1000 10000Total PCDD/PCDF (ng/dscm @

7% O2)

100

10

1

0.1

0.010.1 1 10 100 1000Total PCDD/PCDF (ng/dscm @

7% O2)

TE

Q (

ng/d

scm

@ 7

% O

2)

Hazardous Waste Incineratorsand Boilers

Medical Waste IncineratorsMunicipal Waste Incinerators

Cement Kilns

Data From Combustion Emissions Technical Resource Document

(CETRD) EPA 530-R-94-014, May 1994


OOOOuuuuttttlllliiiinnnneeee ooooffff TTTTaaaallllkkkk❏ Structures








❏ Conclusions

❏ Recommendations


LLLLeeeevvvveeeellllssss ooooffff DDDDiiiiooooxxxxiiiinnnn ooooffff IIIInnnntttteeeerrrreeeesssstttt❏ Currently levels in standards or proposed

standards➮ New Large Municipal Solid Waste Combustors

• 13 ng/dscm total or 0.2 TEQ (@7 % O2 )

➮ New Large (500lb/hr) Medical Waste Combustors

• 25 ng/dscm total or 0.6 TEQ

➮ New Large Hazardous Waste Incinerators (proposed)

• 0.2 ng/dscm TEQ

❏ What is a ng/dscm?➮ nanogram (10 -9 gram) per dry standard cubic meter

➮ 10 ng/dscm of PeCDD corresponds to .6 parts per trillion

• one person out of 100 world populations

➮ If room size of 100 x 100 x 10 ft

• if dioxin concentration is 1 ng/dscm there are approximately 0.000003 gm of dioxin in the room










❏ Conclusions

❏ Recommendations


PPPPCCCCDDDDDDDD////PPPPCCCCDDDDFFFF MMMMeeeeaaaassssuuuurrrreeeemmmmeeeennnnttttssss❏ Manual Sampling -> Laboratory Analysis

➮ Collect on filter & resin using Modified Method 5 train

• Glass probe and sample system components

➮ Complex & sophisticated laboratory analysis

❏ Cleanliness and QA/QC is essential

❏ Labor and time intensive = $$$$$➮ 3-5 days in the field to collect samples, 3-6 weeks

typical lab turnaround time

➮ Typically 3 months from start to finish

❏ Variations in the reference methods➮ EPA (23, 23A, 0010/8290) , California(428) , Canada

➮ EPA Methods can be found on Internet: http://134.67.104.12/html/emtic/cfr prom.htm


Overview Measurement TechniquesSampling onto Particle

Filter and Organic Resin (XAD-2) to concentration

from flue gas

Recovery from Probe, filter, and

resin

Concentrate field rinses and impingers

Solvent extract filter, resin, field rinses and

concentrate

Column cleanup and concentrate

Reporting

Analyze extract by high resolution gas

chromatography/high resolution mass spectrometry with selective ion monitoring

QA/QC spiking


FFFFiiiieeeelllldddd SSSSaaaammmmpppplllliiiinnnngggg

❏ QA/QC samples (field train blanks, field reagent blanks)

❏ Typically 4 to 5 complete samples, 4-6 field fractions each -> 16-30 samples shipped to the laboratory

Three test runs: typically three 10-12 hour days (4-6 hours sampling plus setup and sample recovery for a crew of 3-4)

Glass fiberfilter with

Teflon support

Filteroven

InclineManometer

Flow OrificeMeter

TTV

T

1 1 1

Condenser

XAD-2Sorbent Trap

RecirculationPump

Pump

DryGas

Meter

EPA Method 23

Thermocouple

Heated QuartzProbe Liner

SampleNozzle

T T T

PitotTube Incline

Manometer


RRRReeeeppppoooorrrrttttiiiinnnngggg ooooffff DDDDiiiiooooxxxxiiiinnnn DDDDaaaattttaaaa❏ Treatment of detection limits

➮ zero, half detection, or at detection limit

❏ TEQ treatment ➮ selection of TEF protocol

❏ Measurement technique➮ details of sampling and analysis procedures can

significantly influence the results

➮ can make data comparisons difficult

❏ Emissions Concentrations➮ Percent Oxygen (7% in U.S. vs. 11% in Europe)

• Rule of Thumb: Multiply European Data by 1.404 to compare with North American Data

➮ Different Rounding Procedures

❏ Reporting and treatment of blank and recovery results


MMMMeeeeaaaassssuuuurrrreeeemmmmeeeennnntttt LLLLiiiimmmmiiiittttssss❏ Field data often have large variability

➮ Rule of thumb: PCDD/PCDF data has ± factor of two variability

❏ Controlled emission concentrations data frequently below field blanks

❏ Sources of Variability➮ Process Variations

➮ Sampling Variations

• Sampling Conditions Variability

• Field sampling team

• Sample recovery

➮ Analytical Variations

• extraction and spike recovery

• matrix effects

• hysteresis effects


PPPPrrrraaaaccccttttiiiiccccaaaallll lllliiiimmmmiiiitttt ooooffff qqqquuuuaaaannnnttttiiiittttaaaattttiiiioooonnnn ((((PPPPLLLLQQQQ''''ssss))))

❏ Definitions➮ "PLQ is the lowest level above which quantitative

results may be obtained with an acceptable degree of confidence"

• Federal Register, V57, No. 250, December 29, 1992.

➮ EPA Method 301- " The PLQ is defined as 10 times the standard deviation, at the blank level. This corresponds to an uncertainty of ±30 percent at the 99-percent confidence level"

• analytical process variability

• does not address accuracy and precision from source variation, field sampling and site specific matrix effects.


PPPPrrrraaaaccccttttiiiiccccaaaallll lllliiiimmmmiiiitttt ooooffff qqqquuuuaaaannnnttttiiiittttaaaattttiiiioooonnnn ((((PPPPLLLLQQQQ''''ssss))))

❏ Significant controversy over how a PLQ should be defined for trace species

➮ What level can be routinely be measured with defined accuracy?

➮ What standard can be enforced with the currently available sampling and analysis procedures?

➮ Some studies have indicated that sampling and analytical variations are significant for PCDD/PCDF and that PLQ can be higher than established standards.

➮ Method 23 data are not corrected for spike recovery or blank levels










❏ Conclusions

❏ Recommendations


FFFFuuuunnnnddddaaaammmmeeeennnnttttaaaallll PPPPrrrriiiinnnncccciiiipppplllleeeessss

❏ Organics➮ Under combustion conditions (high temperatures

and with available oxygen),

• thermodynamic equilibrium levels of organics are negligible

• flame/oxidation chemistry of destruction is fast

➮ Formation results from poor mixing, quenches, cool temperature pathways, gas-particle interactions, lack of particle burnout, sooting.


TTTThhhheeeerrrrmmmmooooddddyyyynnnnaaaammmmiiiiccccssss

Fraction Theoretical Air

10-10

10-2

10-4

100

10-6

10-8

10-12

.01 .1 1

Mole Fraction

x x x xx

x

x

Methyl Chloride

Benzene

Biphenyl

Napthalene

CO

THC

Toluene

❏ Equilibrium levels very low except for very rich conditions


OOOOrrrrggggaaaannnniiiicccc KKKKiiiinnnneeeettttiiiiccccssss

Fraction Theoretical Air

10-10

10-2

10-4

100

10-6

10-8

10-12

.01 .1 1

Mole Fraction

❏ Rapid Destruction Kinetics ( modeling data 1227°C, 2240°F for 10 msec in stirred reactor, after W.J. Pitz et al, 1997)

Anthracene

Phenanthrene

Naphthalene

Benzene

Toluene


DDDDeeeessssttttrrrruuuuccccttttiiiioooonnnn KKKKiiiinnnneeeettttiiiiccccssss❏ Rapid Destruction Kinetics (Flow Reactor Data, After

Dellinger et al 1977)

100

.1

1

10

1000 160015001400130012001100

Per

cent

Rem

aini

ng

Temperature (°F)

Hexachlorobenzene

Dichloro-methane

Benzene

DioxinFurans

700Temperature (°C)

800600 900


SSSSoooouuuurrrrcccceeeessss aaaannnndddd CCCCoooonnnnddddiiiittttiiiioooonnnnssss ffffoooorrrr OOOOrrrrggggaaaannnniiiicccc HHHHAAAAPPPPssss ❏ From Unburned Fuel

➮ e.g., Benzene, Toluene, Ethylbenzene and Xylene (BTEX), hexane, naphthalene

❏ From Quenches➮ e.g., acetaldehydes, formaldehydes, acrolein

➮ also CO

❏ From By-product Chemistry in Rich Packets

➮ e.g., benzene and toluene

➮ e.g., polyaromatic hydrocarbons

❏ From Gas-Particle Interactions➮ e.g., chlorinated dioxin










❏ Conclusions

❏ Recommendations


DDDDiiiiooooxxxxiiiinnnn FFFFoooorrrrmmmmaaaattttiiiioooonnnn MMMMeeeecccchhhhaaaannnniiiissssmmmm

❏ 1. Undestroyed PCDD/PCDF originating with waste or fuel

❏ 2. Gas Phase Formation (Furnace to 500°C)

➮ From Related Chlorinated Precursors and from Simple Organics and Chlorine Donors (ca. 700°C)

➮ Formation of Dioxin or Precursors from Complex Organics and Chlorine Donors

❏ 3. Solid Phase Particle Mechanisms (< 500°C )

➮ de novo synthesis

➮ Catalyzed Precursor Mechanisms

➮ Condensation of Precursors and surface chlorination


DDDDiiiiooooxxxxiiiinnnn FFFFoooorrrrmmmmaaaattttiiiioooonnnn MMMMeeeecccchhhhaaaannnniiiissssmmmmssss

O

OCl

Cl

Cl

Cl

❏ 1. Undestroyed PCDD/PCDF originating with waste or fuel

Dioxin in Waste or Fuel or other feed materials

Failure Modes e.g.,flame bypasslow temperaturepoor mixingparticulate burnoutquenches

O

OCl

Cl

Cl

Cl

Dioxin remaining after combustion


DDDDiiiiooooxxxxiiiinnnn FFFFoooorrrrmmmmaaaattttiiiioooonnnn MMMMeeeecccchhhhaaaannnniiiissssmmmmssss❏ 2. Gas Phase Formation

➮ from Related Chlorinated Precursors and from Organics and Chlorine Donors (ca. 700°C)

O

OCl

Cl

Cl

Cl

OHClCl

Chlorophenols

Cl

Cl

Cl

Cl

PCBEvidence: PCDD/PCDF on soot from PCB fires

Lab and Bench Scale Studies of PCB, Chlorinated Benzene, Chlorinated Phenols, C

2 aliphatics have yielded PCDD/PCDF (e.g.,

Schaub and Tsang, Sidhu and Dellinger, Grotheer and Louw, Akki and Mulholland, Weber)

OHClCl

+

Precursors for downstream

formation


GGGGaaaassss PPPPhhhhaaaasssseeee MMMMeeeecccchhhhaaaannnniiiissssmmmmssss❏ An example of the current

thinking about the gas phase mechanisms (After Sidhu and Dellinger 1997)

R• +

OH

Cl

ClClOH

Cl

ClCl+

O•

Cl

ClCl

-OH-Cl

Cl

Cl

OHCl

ClO

Cl

Cl

Cl

ClO

Cl Cl

Cl

Cl

OCl

ClO

-Cl+H

Cl-HCl

RH

Cl

Cl

OHCl

ClO

•

Cl

Cl

O•Cl

ClO

Cl-Cl-Cl

Cl

Cl

OCl

ClO

Cl O

Cl

Cl

O

1,3,6,8 TCDD 1,3,7,9 TCDD

1,3,6,8 TCDD

Cl

-Cl2

Cl

Cl

Cl

ClO

2,4,6,8 TCDF

+R• -Cl

Cl

Cl

Cl

ClO

• Cl

Cl

Cl

Cl

ClO

2,4,6,8 TCDF

-Cl

1,3,6,8 TCDF

Cl

Cl

Cl

ClO

-OH -Cl


DDDDiiiiooooxxxxiiiinnnn FFFFoooorrrrmmmmaaaattttiiiioooonnnn MMMMeeeecccchhhhaaaannnniiiissssmmmmssss❏ 2. Gas Phase Mechanisms

➮ Formation of Dioxin or Precursors from Organics and Chlorine Donors

Complex Organic Structure

LigninWoodPaper

Plastics

Chlorine DonorPVCNaCl

HCl/Cl2

O

OCl

Cl

Cl

Cl

+

Evidence: Lab scale tests of vegetable matter, wood, lignin, coal with chlorine source have yielded PCDD/PCDF

Lab scale studies on fuel rich, simple hydrocarbon flames yielded PAH ( e.g., Marinov, et al and Senkan et al)

Precursors for

downstream formation


Fuel Rich Formation Pathways for Aromatics

H2CCCH+CH3

C3H4 (allene)

C3H5 (allyl)

C2H3+CH3

-H

-H2 +H

+H2CCCH

CH2(S)+C2H2C2H4+C2H3

CH2CHCHCH2

-H2

+H

CH2CHCCH2 H2CCCH3

Branched Aromatics

+H2CCCH or

+H2CCCH3

+H

-H

+H

-H

-H

+H-H

2

(C=C-C=C and C-C=C=C)

-H

H2C=C=CR H

2C-C=CR

R=H (propargyl)

R=CH3 (1-methylallenyl)

+H

CH3 CH

3

CH3

Benzene

After Marinov, Pitz, Westbrook, Castaldi and Senkan, 1996


Fuel Rich PAH Formation PathwaysPropargyl Radical Self-combine

H2C=C=CH

(phenyl)

(phenoxy)

(naphthalene)

(naphthyl)

(phenanthrene)(anthracene)

+ H

O+O

2

+ O

+ +H+H

+H+H+

O

-CO

-CO

+O2

-O

•

• •

+H

•

PAH+C2H

2

(naphthoxy)

After Marinov, Pitz, Westbrook, Castaldi and Senkan, 1996

Thermal Decomposition of Aliphatic Compounds

C2H6 C2H4 C2H2 •C2HCnH2n+2CnH2n

CnH2n-2

CH4 2•CH4 •C2H5 •C2H3

I II

1) Aromatic formation2) Chlorination

1) Aliphatic chlorination2) Cl-aromatic formation

Al2O CuO,

•C4H5C4H6C2H2

•C3H3•C4HxCly

•C3HxClyC4Cl6

(C4HxCly)

C2HClC2Cl2

Radical Propagation/Condensation

RadicalPropagation/Condensation

Diels-Alder Diels-

AlderCatalyticTrimerisation/

Fischer-Tropsch CatalyticTrimerisation/

Fischer-Tropsch

Clx

Clx

Clx

•Cl

•Cl

Clx

O

Clx

O

Clx

•Cl

PCDD/F

PAHbiphenyl

alkylbiphenyl

Gas Phase Mechanisms

Heterogeneous combustion reactions of C2 aliphatics (ca 600°C)(After Froese and Hutzinger, 1997)


DDDDiiiiooooxxxxiiiinnnn FFFFoooorrrrmmmmaaaattttiiiioooonnnn MMMMeeeecccchhhhaaaannnniiiissssmmmmssss❏ 3. Solid Phase Particle Mechanisms (< 500 °C )

Entrained or Captured Particulate

Matter

O

OCl

Cl

Cl

Cl

Low

Temperature

Precursors on Fly Ash Surface

Gas Phase Formation

of Precursors


DDDDiiiiooooxxxxiiiinnnn FFFFoooorrrrmmmmaaaattttiiiioooonnnn MMMMeeeecccchhhhaaaannnniiiissssmmmmssss

❏ 3a. de novo synthesis➮ fly ash can be source of carbon, catalysts and

chlorine as reagent for low temperature formation (ca. 300 °C)

➮ oxidative gasification reactions with carbon in particles to cleave dioxin/furan structures

O

OCl

Cl

Cl

Cl

Evidence: Laboratory scale tests of heated fly ash can generate PCDDPCDF

OxygenMoisture

ca. 300°C


DDDDeeee nnnnoooovvvvoooo FFFFoooorrrrmmmmaaaattttiiiioooonnnn❏ Effect of Temperature on de novo formation rates of

tetra- to octa- PCDD/PCDF (Schwartz 1990)


DDDDiiiiooooxxxxiiiinnnn FFFFoooorrrrmmmmaaaattttiiiioooonnnn MMMMeeeecccchhhhaaaannnniiiissssmmmmssss❏ 3b. Catalyzed Precursor Mechanisms (< 500 °C)

O

OCl

Cl

Cl

Cl

Chlorinated Precursors (e.g., Chlorophenols)

NonChlorinated Precursors (e.g., Aliphatics, Aromatices, Propene, BenzeneC2 hydrocarbons, Methane)

Evidence: Lab Scale data where PCDD/PCDF has been generated at low temperatures when different organics have been burned in the presence of chlorine sources and metals (e.g., Gullett, Froese, Addink)

Metal Catalyzed

Chlorine Source

+


CCCCaaaattttaaaallllyyyyzzzzeeeedddd PPPPrrrreeeeccccuuuurrrrssssoooorrrr MMMMeeeecccchhhhaaaannnniiiissssmmmmssss

TEMPERATURE (°C)

pg P

CD

D/2

µL IN

JEC

TIO

N2,500,000

2,000,000

1,500,000

1,000,000

500,000

0

150 200 250 300 350 400 450 500 550

Tetra-CDD

Penta-CDD

Hexa-CDD

Hepta-CDD

Octa-CDD

H

H

H

H

❏ Effect of Temperature on production with Cu(II) with chlorophenol


CCCCaaaattttaaaallllyyyyzzzzeeeedddd PPPPrrrreeeeccccuuuurrrrssssoooorrrr MMMMeeeecccchhhhaaaannnniiiissssmmmmssss

Tetra-CDD Penta-CDD Hexa-CDD Hepta-CDD Octa-CDD

CONGENER CLASS

pg P

CD

D/2

µL IN

JEC

TIO

N

10,000,000

1,000,000

100,000

10,000

1,000

100

CuO Fe2O3 ZnO NiO Al2O3 Blank

❏ Effect of various metals on dioxin formation from chlorinated phenol


CCCCaaaattttaaaallllyyyyzzzzeeeedddd PPPPrrrreeeeccccuuuurrrrssssoooorrrr MMMMeeeecccchhhhaaaannnniiiissssmmmmssss❏ Effect of iron oxidative state on PCDD with chlorophenol

10,000,000

1,000,000

100,000

10,000

1,000

100Tetra-CDD Penta-CDD Hexa-CDD Hepta-CDD Octa-CDD

Fe2O3Fe (III)

Fe2O4Fe (II, III)

FeFe (0)

BlankNo Fe

CONGENER CLASS

Pg

PC

DD

/2µL

INJE

CT

ION


DDDDiiiiooooxxxxiiiinnnn FFFFoooorrrrmmmmaaaattttiiiioooonnnn MMMMeeeecccchhhhaaaannnniiiissssmmmmssss❏ 3c. Condensation of Precursors and

Surface Chlorination (< 500°C)➮ Surface condensation of precursors followed by

direct chlorination of adsorbed hydrocarbons

Entrained or Captured Particulate

Matter

+Surface

Chlorine Donor O

OCl

Cl

Cl

Cl

Evidence: Studies of passing combustion effluents with oxygen and HCl over fly ash (Altwicker) and entrained flow studies (Gullett et al)

Low

Temperature

Precursors condense on Fly Ash Surface


IIIInnnn FFFFlllliiiigggghhhhtttt FFFFoooorrrrmmmmaaaattttiiiioooonnnn800

0

200

400

600

1 1.5 3.532.520

200

500

400

300

100

Tem

pera

ture

(°C

)

Residence Time (sec)

D+F

Yie

ld (n

g/ds

cm)❏ Pilot Scale RDF

Combustion Tests (After Gullett, 1997)

❏ PCDD/PCDF has formed by 325 °C

❏ Levels decline with time

❏ Long residence times in APCD are not necessarily required for formation










❏ Conclusions

❏ Recommendations


Dioxin Formation Mechanism Regions

Catalytic formation of PCDD/PCDF on fly ash

PCDD/PCDF in waste stream

PCDD/PCDF formed from gas

precursors

PCDD/PCDF formed on particles

o

oCl

Cl

Cl

Cl

ESP


Parameter Observation Source

Entrained PM Increase Stack PCDD/F

Furnace formed PCDD/F associated with Particles

Quebec City,MWC

Montgomery Cty, MWC

CO Emissions Increase uncontrolled PCDD/F Mid-Conn, MWC

Low CO (<200 ppm) not sufficient to assure low stack PCDD/F

Mid-Conn & Mont. Cty MWCsVarious MWIs

High CO (>200 ppm) not sufficient to assure high PCDD/F

Cement kilnsUtility boilers

Carbon inFly Ash

Increase total PCDD/F formationAbsorbs PCDD/FDiverts PCDD/F from stack to fly ash

Borgess Med CtrQuebec CityDavis County

SIGNIFICANT PARAMETERS - FIELD STUDIES


SSSSIIIIGGGGNNNNIIIIFFFFIIIICCCCAAAANNNNTTTT PPPPAAAARRRRAAAAMMMMEEEETTTTEEEERRRRSSSS ---- FFFFIIIIEEEELLLLDDDD SSSSTTTTUUUUDDDDIIIIEEEESSSS

Parameter Observation Source

APCD Temp. PCDD/F Formation Increases Exponentially

Montg. Cty, MWCBorgess, MWIAsh Grove , Ck

Effects solid/vapor phase partitioning Montg. Cty, MWC

Quench Rate Rapid Quench decreases PCDD/Fformation rate

MWIs & HWIs withwet scrubbers compared to units with heat recovery & dry APCD

Cl in Fuel Negligible observed impact on PCDD/F stack emissions

Ash Grove CementPPG HAFWurtzberg, MWCHWIs with wet APCD

Catalytic Metals in fuel

Stack PCDD/F emissions impact not clearly defined

Quebec City, MWCAsh Grove Cement


PPPPiiiillllooootttt SSSSccccaaaalllleeee PPPPaaaarrrraaaammmmeeeetttteeeerrrr SSSSiiiiggggnnnniiiiffffiiiiccccaaaannnncccceeee

❏ Ability of specific parameters to account for observed PCDD/PCDF yields while holding all other parameters constant ( semi-partial correlations from two separate pilot scale programs)

Parameter RDF/Coal FlyAsh/Cl Injected intoCo-Combustion Natural Gas

RDF Feedrate 0.764 Not ConsideredQuench Rate 0.762 0.271Sorbent Injection 0.299 0.497SO

2 Conc. 0.234 Not Considered

HCl Conc. 0.132 0.380Sampling Temp. 0.016 Not Statistically SignificantOxygen Conc. Not Considered 0.092Cl

2 Conc. Not Considered 0.131

Duct Temp. Not Considered 0.614Residence Time Not Statistically sig. 0.271


IIIImmmmppppoooorrrrttttaaaannnntttt CCCCoooonnnnttttrrrroooolllllllliiiinnnngggg PPPPaaaarrrraaaammmmeeeetttteeeerrrrssss

❏ Combustion conditions as indicated by ➮ CO and Total Hydrocarbon

➮ Soot formation

➮ Particle entrainment and burnout

❏ Quench Rate and Air Pollution Control Temperature

❏ Fuel/waste parameters➮ Chlorine

➮ Sulfur

➮ Metals


RRRRoooolllleeee ooooffff CCCCoooommmmbbbbuuuussssttttiiiioooonnnn PPPPrrrraaaaccccttttiiiicccceeee

300 120 300 1200

5000

10000

15000

20000

25000

30000

PCDD/PCDF ( ng/dscm )

Furnace Formation

Downstream Formation

Furnace Formation

Downstream Formation

"Poor" Combustion Conditions

"Good" Combustion Conditions

Particulate Filter Temperature ( °C)

❏ Pilot Scale data after EER Corp., 1994)


Role of Combustion Practice

2000

0

500

1000

1500

0 200 800600400 1000

Good operationPoorPoor operation operation

R2=0.70

PC

DD

/PC

DF,

ng/

Sm 3

CO at 12% CO2, ppm


CCCCoooommmmbbbbuuuussssttttiiiioooonnnn CCCCoooonnnnddddiiiittttiiiioooonnnnssss

❏ Laboratory Data- CO, THC and HAP

❏ Good Combustion Conditions as indicated by low CO and Total hydrocarbons leads to low HAP emissions

❏ After PERF, 1997


EEEEffffffffeeeecccctttt ooooffff SSSSooooooootttt DDDDeeeeppppoooossssiiiittttssss

❏ Boiler tube simulation with #2 fuel oil, copper napthenate, dichlorohexane

❏ Significant PCDD/PCDF formation

❏ Required all 3

➮ Soot

➮ Copper

➮ Chlorine

❏ Formation even at low Cl (10ppm)

❏ Sulfur Dioxide suppresses PCDD/PCDF

1,000,000

100,000

10,000

1,000

100

10

1

Cl=500

Cl=10Cl=500

SO2=0SO2-600

PC

DD

+P

CD

F (ng

/dsc

m)

Data after Raghunathan, Lee & Kilgroe, 1997


After Soot Deposits

Before Sooting

Combustion AirNatural gas or fuel oil

dichloro-hexane

Dioxin Sampling

end plate

furnace

Duct

to CEM

Sampling Temperature 240 °C, [HCl]= 500 ppm

Impacts of Soot Formation

Data after Raghunathan et al, 1997


Role of ParticlesQuebec City MB/WW MWC

CD

D/C

DF

Em

itted

/Ref

use

Fed

(µm

ole/

tonn

e)

Uncontrolled Ash/Refuse Fed(kg/tonne)

10

0

2

4

6

8

4 86 16141210

Data after Barton et al, 1990


RRRRoooolllleeee ooooffff PPPPaaaarrrrttttiiiiccccuuuullllaaaatttteeee CCCCaaaappppttttuuuurrrreeee TTTTeeeemmmmppppeeeerrrraaaattttuuuurrrreeee

1

10

100

1000

0.0018 0.002 0.00221/[ESP Temperature (K)]

Stack - Inlet Total PCDD+PCDF(ng/dscm @ 7% O2)

Electrostatic Precipitator Temperature (°C)

250 225 200 175

❏ Dioxin formation very sensitive to the temperature of the Particulate Collection Device


LLLLoooowwww TTTTeeeemmmmppppeeeerrrraaaattttuuuurrrreeee FFFFoooorrrrmmmmaaaattttiiiioooonnnn❏ Data after Swartz et al (1990)


Role of Particulate Capture Temperature

100%

-100%

-50%

0%

50%

-250%

-200%

-150%

420 600580560540520500480460440

Con

trol

Effi

cien

cy %

ESP Inlet Temperature (°F)Pinellas County-MB Peekskill-MB Peekskill-MB Oswego-

Stack air emission =Inlet emission

300250 275225ESP Inlet Temperature (°C)


Mechanisms in Particle Collection Devices

PARTICULATE MATTER -WITH ABSORBED ORGANICS INCLUDING CDD/CDF

GAS PHASE ORGANICSCDD/CDF & PRECURSORS

APCD

SURFACE CATALYZED REACTIONS

ORGANICS DESORBED

FROM SURFACES

PARTICLES ESCAPING

PARTICLES COLLECTED

GAS PHASE ORGANICS

AND CDD/CDF

PARTICLES MAY

CONTAIN CDD/CDF

PARTICLES CONTAINING CDD/CDF


CCCChhhhlllloooorrrriiiinnnneeee IIIImmmmppppaaaaccccttttssss❏ ASME Field

Correlations

➮ Lack of Relationship between Percent Chloride Feed and Mass of PCDD/F Emitted.

➮ Other parameters are controlling

215219220224227230231234235238239246251252253258270271272273274281282283284382388389396398399400

H

C

S

F

A

D

✣

✕

J

L

L

J

E

B

G

P

S

F

A

D

✣

✕

L

L

E

G

C

H

P

C

✕EC

✕

S

G

B

C

E

L

K

D

A

P

J

H

F

G

GG

G

G GG

D

DDD

DDD D

FF

FF

FFFFFFB

BBB

BBB

✕✕✕✕✕

✕✕

✕✕✕ ✕

✕✕✕✕✕

✕✕✕✕✕ H

H

HH

H H

HH

H

H

HHH

PPHPPP

P

FFF

AAAA

AAA

A A✕

AAA A

C A

C

S

S

SS

S

SS

S

S

C

CCCC

✣✣

EEEE

E

EEE

EEE

EE

L

DDEE

✣

JJJJ

806040200-2

-1

0

1

4

3

2Lo

g ng

PC

DD

/PC

DF

per

dscm

7%

02

Percent Chlorine in the Feed100


EEEEffffffffeeeecccctttt ooooffff HHHHCCCCllll CCCCoooonnnncccceeeennnnttttrrrraaaattttiiiioooonnnn❏ Pilot scale RDF data (after Gullett, 1997) indicate

that HCl concentration effect is dependent on quench rate

3500

500

1000

1500

2000

2500

3000

0 654321HCl, g/min

Tota

l, ng

/min

RDF,kg/min

HCl,g/min

SO2,g/min

QU, °C/s TS, °C Ca,g/min

1.12 X-Axis 338 00.011 Legend

426426°°C/sC/s 304304°°C/s C/s 531 531°°C/sC/s

Total PCDD/PCDF


RRRRoooolllleeee ooooffff AAAAmmmmbbbbiiiieeeennnntttt CCCChhhhlllloooorrrriiiinnnneeee❏ Ambient levels of chlorine in atmospheric

particulate ➮ from World Meteorological Organization

➮ Total global tropospheric Cl concentration

• organic bound 3.8 ppb (e.g., long lifetime CFCs, carbon tetrachloride, methylchloroform, )

• HCl 1.4 ppb

• particulate bound <0.23 ppb

➮ elevated near coast

• some measurements indicate levels of 35 ppb

• acidification of salt spray

❏ Chlorine levels in ambient air (ppb) can be orders of magnitude greater than required to generate ppt levels of PCDD/PCDF


SSSSuuuullllffffuuuurrrr IIIImmmmppppaaaaccccttttssss❏ Pilot scale RDF data indicate that SO2 lowers dioxin

formation

8000

0

2000

4000

6000

0 87654321

Tota

l, n

g/m

in

SO2, g/min

1.12 kg/min1.12 kg/min 1.93 kg/min1.93 kg/min 0.5 kg/min0.5 kg/min

RDF,kg/min

HClg/min

SO2,g/min

QU, °C/s TS, °C CA,g/min

Legend 1.9581 X-Axis 428 338 0

Total PCDD/PCDF


EEEEffffffffeeeeccccttttssss ooooffff CCCCooooaaaallll CCCCooooffffiiiirrrriiiinnnngggg wwwwiiiitttthhhh RRRRDDDDFFFF ((((EEEEffffffffeeeeccccttttssss ooooffff SSSSuuuullllffffuuuurrrr))))

❏ Cofiring of high sulfur coal suppresses RDF's PCDD/PCDF formation (>73%)

❏ PCDD/PCDF decreases with cofiring cycles, suggesting build up of sulfur effect

❏ Return to PCDD/PCDF RDF baseline is slow

120

0

20

40

60

80

100 RDF

+Coal

RDF

RDF

RDF

+Coal

Avg

. D/F

Yie

ld (

ngT

EQ

/dsc

m)

Full Scale RDF Data prior to APCD ( After Gullett, 1997)










❏ Conclusions

❏ Recommendations


DDDDiiiiooooxxxxiiiinnnn ccccoooonnnnttttrrrroooollll❏ Good Combustion Practice

➮ Uniform high temperature

➮ Good mixing with sufficient air

➮ Minimize entrained, unburned particulate matter

➮ Feed rate uniformity

➮ CO and Total Hydrocarbons Emissions as indicators

➮ Temperature at particulate control device

❏ Air Pollution Control➮ Spray Dryer/fabric filter

➮ Wet Scrubbers

➮ Carbon injection

➮ Carbon beds

➮ Catalysts poisoning agents and inhibitors

➮ Catalytic Oxidations


CCCCoooonnnnddddiiiittttiiiioooonnnnssss MMMMoooorrrreeee LLLLiiiikkkkeeeellllyyyy ttttoooo LLLLeeeeaaaadddd ttttoooo PPPPCCCCDDDDDDDD////PPPPCCCCDDDDFFFF FFFFoooorrrrmmmmaaaattttiiiioooonnnn

❏ Poor Combustion Conditions ➮ Mixing, Temperature, Quenches, Transients

➮ Sooting conditions

➮ High CO and Total hydrocarbons

❏ High particulate entrained from combustion process with poor burnout (high carbon)

❏ Particulate holdup in critical temperature window (150-450°C)

❏ Particulate matter which contains metal that can catalyze formation of dioxin

❏ Waste or fuel with complex organics and/or lignin like structure

❏ Sufficient Chlorine

Listed in Order of Relative Importance


CCCCoooonnnnddddiiiittttiiiioooonnnnssss ttttyyyyppppiiiiccccaaaallllllllyyyy aaaassssssssoooocccciiiiaaaatttteeeedddd wwwwiiiitttthhhh lllloooowwww PPPPCCCCDDDDDDDD////PPPPCCCCDDDDFFFF FFFFoooorrrrmmmmaaaattttiiiioooonnnn

❏ Low Particulate Matter ❏ Low Temperature or no Particulate Control

Device❏ Rapid Quenches through dioxin

reformation window❏ Good Combustion Practice❏ Low "catalysts" Metals❏ High Sulfur❏ Insufficient Chlorine

Listed in Order of Relative Importance










❏ Conclusions

❏ Recommendations


MMMMoooonnnniiiittttoooorrrriiiinnnngggg aaaannnndddd CCCCoooonnnnttttiiiinnnnuuuuoooouuuussss PPPPeeeerrrrffffoooorrrrmmmmaaaannnncccceeee

AAAAssssssssuuuurrrraaaannnncccceeee

❏ No direct real time monitors exist➮ Emerging technologies (REMPI) hold some promise

for more direct measurement

❏ Currently must rely on periodic manual sampling and analysis and monitoring of parameters that are known to impact formation and emissions e.g.,

➮ combustion parameters

➮ CO/THC indicators of combustion conditions

➮ PM control device temperature


TTTThhhheeee uuuusssseeee ooooffff CCCCOOOO aaaassss iiiinnnnddddiiiiccccaaaattttoooorrrr ffffoooorrrr PPPPCCCCDDDDDDDD////PPPPCCCCDDDDFFFF

❏ Carbon monoxide is a stable combustion intermediate

❏ It can be continuously and reliably monitored

❏ It can be used as an indicator ➮ that conditions are same as compliance tests

➮ that conditions are good combustion conditions

❏ But not as a direct indicator of trace organic emissions


Laboratory Data- CO,

THC and HAP


CO and PCDD/PCDF in MWCs

2000

0

500

1000

1500

0 200 800600400 1000

Good operationPoorPoor operation operation

R2=0.70

PC

DD

/PC

DF,

ng/

Sm 3

CO at 12% CO2, ppm


CO and PCDD/PCDF in Hazardous Waste Incinerators


THC and PCDD/PCDF in Hazardous Waste Incinerators


AAAAddddvvvvaaaannnncccceeeedddd MMMMoooonnnniiiittttoooorrrriiiinnnngggg TTTTeeeecccchhhhnnnniiiiqqqquuuueeeessss

❏ Resonance Enhanced Multi-Photon Ionization "REMPI"

❏ Principles➮ Molecules have characteristic energies of their

intermediate vibrational states

➮ Selection of laser wavelength in resonance with that vibrational state causes photoionization

➮ Ionization yields are very high

➮ Ionization is selective, molecule specific

➮ Coupled with Time-of-Flight Mass Spectrometer yields 2D analytical tool


REMPI ApparatusGas

Probe Gated Valve

ISTriggerGate

Reflectron-TOF-MS

SHG + PS

LambdaOPPO

COHERENTInfinityNd-YAG

1 mJZ ns0.15 cm-1

Trigger

Controland Data

Acquisition

DSO108/s

Preamplifier

Signal


Correlations of mono, di, tri dioxins with TEQ

No. of R2 SS Variables in ModelIsomers1 0.729 66.55 2,3,7-TrCDD1 0.537 112.501 2,6-DiCDD1 0.501 122.515 2-MCDF2 0.736 66.040 2,3,7-TrCDD 2,4-DiCDD2 0.733 65.377 2-MCDD 2,3,7-TrCDD2 0.733 65.467 2,3,7-TrCDD 2-MCDF3 0.767 57.136 2-MCDD 2,3,7-TrCDD 2,4-DiCDF3 0.760 68.068 2,8-DiCDD 2,3,7-TrCDD 2,4-DiCDF3 0.757 69.038 2-MCDD 2,3,7-TrCDD 2-MCDF4 0.770 56.535 2-MCDD 2,3,7-TrCDD 2,4-DiCDF4 0.769 56.639 2-MCDD 2,8-DiCDD 2,3,7- TrCDD 2,4-DiCDF4 0.760 58.807 2,8-DiCDD 2,3,7-TrCDD 2-MCDF 2,4-DiCDF5 0.771 58.189 2-MCDD 2,8-DiCDD 2,3,7-TrCDD 2-MCDF

2,4-DiCDF

Predictive Models of TEQ with Low Chlorinated Congeners


CCCCoooonnnncccclllluuuussssiiiioooonnnnssss❏ Dioxin levels of interest

➮ very low concentrations

➮ difficult to measure

➮ care must be taken when comparing measurements due to differences in measurement and reporting techniques

➮ levels near analytical detection limits and practical quantitation limits

❏ Monitoring➮ No direct real time monitors exist

➮ Must rely on Parameter monitoring e.g.,

• CO/THC indicators of combustion conditions

• PM control device temperature

➮ Emerging technologies (REMPI) hold some promise for more direct measurement


CCCCoooonnnncccclllluuuussssiiiioooonnnnssss❏ Formation Mechanisms

➮ Major pathways are defined

➮ Different pathways dominant under different conditions involving both combustion and post combustion regions

• Gas Phase mechanisms

• Solid Phase mechanisms

➮ Kinetics of mechanisms are not established

• Cannot a priori predict PCDD/PCDF formation levels

➮ Key importance of

• Good Combustion Practice

• Particulate matter

• PM control Device Temperature


CCCCoooonnnnddddiiiittttiiiioooonnnnssss MMMMoooorrrreeee LLLLiiiikkkkeeeellllyyyy ttttoooo LLLLeeeeaaaadddd ttttoooo PPPPCCCCDDDDDDDD////PPPPCCCCDDDDFFFF FFFFoooorrrrmmmmaaaattttiiiioooonnnn

❏ Poor Combustion Conditions ➮ Mixing, Temperature, Quenches, Transients

➮ Sooting conditions

➮ High CO and Total hydrocarbons

❏ High particulate entrained from combustion process with poor burnout (high carbon)

❏ Particulate holdup in critical temperature window (150-450°C)

❏ Particulate matter which contains metal that can catalyze formation of dioxin

❏ Waste or fuel with complex organics and/or lignin like structure

❏ Sufficient Chlorine


RRRReeeeccccoooommmmmmmmeeeennnnddddaaaattttiiiioooonnnnssss❏ Examine source design and operation for conditions

potentially favoring dioxin formation

❏ Examine existing dioxin emissions data bases for similar sources

❏ Support research leading to screening and predictive models for PCDD/PCDF from practical combustion systems

❏ Field test on sources

➮ If funds limited focus on more likely sources based upon current understanding of PCDD/PCDF mechanisms

➮ Data collection should include parameters known to influence PCDD/PCDF formation and control

➮ Data engineering analysis to interpret and generalize to other configurations

➮ Address practical quantitation limit

❏ Emphasis on continuous performance assurance and advanced monitoring techniques

Testing RecommendationsSource Fuel Potential Rationale/parameters

Incinerators Waste High Complex organic, Entrained PM, PM Control, metals & Cl, PM Control Temp

Boilers Wood/biomass Mod-High Complex Organics, entrained PM, PM control, Cl, Metals, PM Control Temp

Boilers Landfill gas Moderate Transients, Complex HC, Cl, Low PM

Boilers Residual Oil Mod-Low Sooting, metals, S/Cl, PM control TempIC Engines Landfill gas Moderate Transients, Complex HC, Cl,

Low PMBoilers Coal Low-mod S/Cl ratio, GCP, Entrained PM, PM

Control, C in ash, metalsHeaters Process gas Low GCP, transients, potential catalysts, low

PMFlares Waste Gas Low No PM control, rapid quenchBoilers Distillate Oil Low Sulfur, GCP, low ClIC Engines Diesel Low PM present,Low Cl, Temp ProfileTurbines Distillate Oil Low GCP, no PM control, thermal

quenchBoilers Natural Gas Very Low GCP, very low PM, Low metals andClIC Engines Natural Gas Very Low Very low PM, GCP, Low metals and ClTurbines Natural Gas Very Low GCP, very low PM, Rapid

Quench, Low metals and Cl

Documents

Dioxin Primer Prestn - EJnet.org · 2. Gas Phase Formation (Furnace to 500°C) From Related Chlorinated Precursors and from Simple Organics and Chlorine Donors (ca. 700°C) Formation