EPA-6QQ/8-39-Ofc4a
July 1989
Municipal Waste Combustion
Multipollutant Study
Emission Test Report
Maine Energy Recovery Company
Refuse Derived Fuel Facility
Biddeford, Maine
Volume I: Summary of Results
By
6. Scheil, S. Klairan, M. Whitacre, and J. Surman
Midwest Research Institute
Kansas City, Missouri 64110
and
W. Kelly
Radian Corporation
Research Triangle Park, North Carolina 27709
EPA Contracts 68-02-4463, W.A. 2 and
68-02-4395, W.A. 27
Project Officer
James 0. Kilgroe
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
AIR AND ENERGY ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
TECHNICAL REPORT DATA
(Please read fxsiructwm on the reverse before comph
. 1. REPORT NO. 2,
EPA-600/8-89-064a
2
4. TITLE AMD SUBTITLE
Municipal Waste Combustion, Mirltip'ollutant Study,
Emission Test Report, 'Maine Energy Recovery Com-
pany Refure Derived Fuel Facility. Bidceford. Maine
5. REPORT DATE
July 1989
6. PERFORMING ORGANIZATION CODE
esft
?, authob(si Scheil, S. Klamm, M.Whitacre, and J.
Surman (MRI); and W.Kelly (Radian)*
8. performing organization report no.
9, PE RFORMING ORGANIZATION NAME AND ADDRESS
Midwest Research Institute
425 Volker Boulevard
Kansas City, Missouri 64110
10. PROGRAM ELEMENT NO,
11. CONTRACT/GRANT NO.
68-02-4463, Task 2;
68_02~4395, Task 27
12, SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Enerey Engineering Research Laboratory
Research Triangle Park, NC 27711
13. TYPE Of REPORT AND PERIOD COVERED
Task Final; 12/87 - 5/88
14. SPONSORING AGENCY CODE
EPA/600/13
15.supplementary notes AEERL project officer is James D, Kilgroe, Mail Drop 65, 919/
541-2854. (*) Radian Corporation, P.O. Box 13000, Research Triangle Park, NC
27709. (**) Volume I; Summary of Results.
is. abstract ij^e report gives results of an emission test of a new municipal solid waste
eombustor, in Biddeford, ME, that burns refuse-derived fuel and is equipped with a
lime spray dryer fabric-filter (SD;FF) emission control system. Control efficiency
of the SD/FF emission control system was measured for polychlorinated dibenzo-
dioxins (PCDD), polychlorinated dibenzofurans (PCDF), particulate matter (PM),
cadmium (Cd), chromium (Cr), arsenic (As), lead (Pb), mercury (Hg), sulfur diox-
ide (SC2), and hydrogen chloride (HC1). Additional continuous monitoring was con-
ducted for oxygen (02), carbon dioxide (CC2), carbon-monoxide (CO), nitrogen ox-
ides (NOx), and total hydrocarbons (THC). Average emissions of total PCDD were
290 ng/dscm (uncontrolled) and 1.3 ng/dscm (controlled),- Total PCDF emissions
were 590 ng/dscm (uncontrolled) and 2.9 ng/dscm (controlled). Control efficiency
was'about 99.5% for both dioxins and furans. Particulate emissions averaged 7400
mg/dscm (uncontrolled) and 33 mg/dsem (controlled), for an average particulate
control efficiency of 99.5%. Metals emissions varied from 500 micrograms/dscm
for As and Hg to 30,000 micrograms/dscm for Pb (uncontrolled), and from 6 micro-
grams/dscm for As and Cr to 160 micrograms/dscm for Pb (controlled). Metals
control efficiencies varied from 98.2% for Hg to 99.8% for Cr.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b. IDENTiPIE RS/CPEN ENDED TERMS
c. COSATI Field/Group
Pollution Fabrics
Emission ' Spray Drying
Refuse Calcium Oxide
Wastes
Combustion .
Filters
Pollution Control
Stationary Sources
Municipal Waste Com-
bustion
Refuse Derived Fuel
Fabric Filters
13B HE ,
14G 07A, 13 H
07B
21B
19. DISTRIBUTION! STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO.^^PAGES
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE, ^
Aofi
EPA Form 2220-1 (9-73)
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
-------
ACKNOWLEDGMENTS
This final report describes work performed by Midwest Research Institute
(MRI) under U.S. Environmental Protection Agency (EPA) Contract 68-02-4463,
Work Assignment No. 2, and Contract 68-02-4395, Work Assignment No. 27. The
EPA project officer is James D. Kilgroe, Air and Energy Engineering Research
Laboratory (0R0/AEERL). Clyde E. Riley, Emission Standards and Engineering
Division (OAQPS/ESED), Is the EPA task manager. Theodore G. Brna is the AEERL
program coordinator.
This work was supported with funds provided by the Municipal Waste Com-
bustion Program of the EPA Office of Environmental Engineering and Technology
Demonstration, and the National Incinerator Testing and Evaluation Program of
Environment Canada.
George Scheil, Senior Chemist, Field Measurements Section, of MRI s
Environmental Chemistry Department, was the project leader. The project was
conducted under the direction of Roy Neulicht, MRI1s Program Manager for
Incineration.
Process data were collected by Winton Kelly of Radian Corporation. HC1
measurements were conducted by a team from Entropy Environmentalists, Inc.,
led by Ran Jernigan. A list of the participants in the project is given in
Appendix N.
i f 'i
-------
ABSTRACT
This report describes the results of an emission test of a new municipal
solid waste eombustor which burns refuse-derived fuel and- which is equipped
with a lime spray dryer fabric' filter (SD/FF) emission control system. The
facility tested is operated by the Maine Energy Recovery Company and is
located in Bicdefcrd, Maine.
Control efficiency of the SD/FF emission control system was measured for
polychlor-'nated dibenzodloxins (PCDD), polychlorinated dibenzofurans (PCDF),
particulate matter (PM), cadmium (Cd), chromium (Cr), arsenic (As), lead (Pb),
mercury (Hg), sulfur dioxide (S02), and hydrogen chloride (HC1). Additional
continuous monitoring was conducted at various locations for oxygen (02),
carbon dioxide (C02), carbon monoxide (CO), nitrogen oxides (N0X), and total
hydrocarbons (THC). Process samples were also collected and analyzed for
metals and selected physical properties.
Average emissions of total PCOO were 290 ng/dscm (0.1 x 1Q"6 gr/dscf)
(uncontrolled) and 1.3 ng/dscm (5.7 x 10"10 gr/dscf) (controlled). Total PCDF
emissions were 590 ng/dscm (0.3 x 10~« gr/dscf) (uncontrolled) and 2.9 ng/dscm
(1.3 x 10"9 gr/dscf) (controlled). The control efficiency was about 99.5% for
both dioxins and furans. All of the above results are corrected to 12% C02.
The 17 specific PCDD/PCDF isomers, as well as 'the tetra through octa chlori-
nated total congeners, showed no significant change in distribution across the
control device.
Uncontrolled particulate emissions averaged 7,400 mg/cscm (3.23 gr/dscf),
and controlled particulates averaged 33 mg/dsctn (0.01 gr/dscf) (corrected to
12% CC2) for an average particulate control efficiency of 99.5%.
Metals emissions (uncontrolled) varied from 500 ug/dscm for arsenic and
mercury to 30,000 ug/dscm for lead. Controlled metals emissions varied from
6 ug/dscm for arsenic and chromium to 160 ug/dscm for lead. Metals control
efficiencies varied from 98.2% for mercury to 99.8% for chromium. The process
ash sample results were in general agreement with the concentrations measured
in the stack samples.
The continuous monitoring results and process data logging indicated that
the combustion process was never under optimum operating conditions. There
were frequent problems with feeder conveyors during all three test runs. ¦ CO
concentrations averaged 70 opm with , some short duration excursions above
200 ppm.
IV
-------
.The automatic SD/FF control system was not operating during these
tests. During the. first test, the stoichiometric lime-to-HCl + SQZ ratio was
1,7, which resulted in an S02 removal efficiency of 66% and an HC1 removal
efficiency of 98%, During the early stages of the' second test, the lime feed
rate was doubled to give a stoichiometric ratio of 3.4. During the third test
the stoichiometric ratio was 3.9 and removal efficiencies were improved to 90%
for SO2 and 99.4% for HC1.
This report was submitted in fulfillment of Contract 68-02-4463, Work
Assignment No. 2 and Contract 68-02-4395, Work Assignment No. 27 by Midwest
Research Institute under the sponsorship of the U.S. Environmental Protection
Agency. This report covers tests performed during the period from December 8
to 12,' 1987. Project work was completed as of May 1, 1988.
v
-------
CONTENTS
Page
Volume I
Acknowledgments. ill
Abstract "IV
Figures..... xi
Tables......xiii
Abbreviations ^
1.0 Introduction 1-1
1.1 Process Description . 1-1
1.2 Measurement Program....... 1-4
1.3 Quality Assurance/Quality Control (QA/QC) 1-7
1.4 Description of Report Sections 1-7
2.0 Summary of Emission Results..... 2-1
2.1 Process Data.. 2-1
2.2 PCOO/PCDF Emissions 2-3
2.3 Particulate Emissions. 2-17
2.4 Metals Emissions 2-17
2.5 Metals Content of'Process Samples. 2-17
• 2.6 Other Process Sample Analyses 2-22
2.7 Aci d Gases.... * .2—22
2.8 Other Gases. 2-22
2.9 Conclusions and Recommendations. 2-43
3.0 Process Description and Operation During Test Program 3-1
3.1 Facility Description 3-1
3.2 Summary of Operations by Test Run..., 3-7
3.3 Summary of Key Operating Parameters 3-8
4.0 Sampling Locations. 4-1
4.1 Spray Dryer Inlet... 4-1
4.2 Spray Dryer Outlet/Baghcuse Inlet 4-1
,4.3 Baghouse Outlet... 4-4
4.4 Cyclone Collector Ash 4-4
4.5 Baghouse Ash 4-4
4.6 Bottom Ash.... 4-6
4.7' Lime Slurry... 4-6
4.8 Refuse Derived Fuel 4-6
4.9 Samples Not Collected 4-6
5.0 Sampling and Analysis Procedures.... 5-1
5.1 Dioxin/Furan Sampling and Analysis 5-1
5.2 Combustion Sas—Particulate and Metals Sampling
and Analysis — 5-17
5.3 Ash, Lime Slurry, and RDF Sampling and Analysis.... 5-24
5.4 Continuous Gas Analyzers 5-27
5.5 Carbon Dioxide and Oxygen Sampling and Analysis
by EPA Method 3. 5-28
vi 1
Preceding Page Blank
-------
CONTENTS (continued)
Page
ۥ0 QA/QC. 6"" X
6.1 Introduction 6-1
6.3 Summary of QC Data*6*1
6.3 Audits 6-21
7.0 References . 7-1
8.0 Conversion Factors 8-1
viii
-------
CONTENTS
Page
Volume II
Appendices I ¦
A. Sample calculations.. A-l
B. Process data 6-1
Full-size plots'of key operating data ' B-2
Printouts of 4-min readings 8-21
C. Semlvolatile organics (dioxin/furan) MM5 field data........... C-l
Run Mo. 1 spray dryer inlet... C-2
¦ Run Mo. 1 baghouse outlet..... C-13
Run Mo. 2 spray dryer inlet C-24
Run No. 2 baghouse outlet..... C-36
Run No. 3 spray dryer inlet...., C-48
Run No. 3 baghouse outlet C-60
0. Particulate/metals MM5 field data and carbon
dioxide/oxygen M3 field data. D-l
Run No. 1 spray dryer inlet D-2
Run No. 1 baghouse outlet D—13
Run No. 2 spray dryer inlet... D-24
Run No. 2 baghouse outlet 0-36
Run No. 3 spray dryer inlet 0-48
Run No. 3 baghouse outlet D-59
M3 sampling and Qrsat analysis 0-70
Preliminary velocity traverse data D-83
Proof/blank trains set-up/recovery 0-86
E. Ash, lime slurry, and. RDF field data E-l
Baghouse ash................. E-2
Cyclone ash....... E-18
Bottom ash... ' E-28
Refuse-derived fuel........... E-36
Lime slurry.. E-40
F. CEM data.. F-l
Summary tables and concentration plots „ F-2
Run 1—spray dryer inlet, spray dryer outlet
(midpoint), baghouse outlet F-36
Run 2—spray dryer inlet, spray dryer outlet
(midpoint), baghouse outlet. F-91
Run 3—spray dryer inlet, spray dryer outlet
(midpoint), baghouse outlet. F-163
ix
-------
CONTENTS
Page
Volume III
Appendices
G. Laboratory analysis PCDDs/PCDFs G-l
Summary tables....... G—2
Quantitation report sheets...... G-9
Notebook pages - extraction G-34
Notebook pages - GC/MS G-58
H. Laboratory analytical results for particulate/metals H-l
Particulate data H-2
Metals data H-13
I. Laboratory analysis - other samples 1-1
J. Equipment and calibration data ' J-l
K. Field test logs.... K-l
Daily sampling log summary K-2
Field operations log K-4
Sampling location logs K-8
Field laboratory and sample identification log -
particulate/metals.. ' K-17
Field laboratory and sample identification log -
semi volatile organics K-36
Process sample identification log....... ............. K-52
Field sample traceability logs K-59
L. Quality assurance data. L-l
PCDD/PCDF blank trains L-2
PCDD/PCDF duplicate sample analyses.. L-10
PCDD/PCDF internal QA and instrument performance
results........L-16
PCDD/PCDF relative response factors L-45
EPA audit sample results............. L-70
Metals spikes and NBS standard recoveries t-82
Metals calibration data.... L-86
¦ Metals duplicate sample analysis L-91
CEM calibration data...... L-94
EPA audit reports. l-lll
M. Sampling and analysis procedures M-l
Sampling procedures M-2
Calibration protocol M-27
Radian draft method M-30
ASME draft methods M-57
N. Personnel N-l
x
-------
FIGURES
Number Page
1-1 MERC waste-to-energy process 1-2
1-2 Sampling and monitoring locations, Unit A 1-9
2-1 Mole fraction plot of PCDO (controlled) 2-9
2-2 Mole fraction plot of PCDO (uncontrolled).. 2-10
2-3 Mole fraction plot of PCDF (controlled) 2-11
2-4 Mole fraction plot of PCDF'(uncontrolled) 2-12
2-5 Resistivity plots of MERC baghouse ash 2-2.4
2-6 Run 1 S02 concentrations 2-26
2-7 Run 2 S0a concentrations 2-27
2-8 Run 3 S0a concentrations 2-28
2-9 Run 1 HC1 concentrations 2-29
2-10 Run 2 HC1 concentrations 2-30
2-11 Run 3 HC1 concentrations. 2-31
2-12 NQX concentrations at the baghouse outlet......... 2-32
2-13 Run 1 oxygen concentrations 2-35
2-14 Run 2 oxygen concentrations 2-36
2-15 Run 3 oxygen concentrations 2-37
2-16 .Run 1 C02 concentrations....... 2-38
2—17 Run 2 CC2 concentrations.2—39
2-18 Run 3 C0a concentrations 2-40
2-19 CO concentrations at the spray dryer outlet 2-41
2-20 THC concentrations at,the spray dryer inlet 2-42
3-1 Combustion process line at the MERC facility 3-2
3-2 Preparation of refuse-derived fuel at the MERC facility. 3-3
3-3 Combustion air scheme at the MERC faci 1 ity ... 3-5
3-4 Location of important temperature, pressure, and flow
sensors at the MERC facility.... 3-9
3-5 RDF heat release steam flow, steam temperature, and steam
pressure as a function of time during, the MERC test program.. 3-10
3-6 Combustion air pressure .as a function of time during the
MERC test program. 3-12
3-7. Overfire air flow pressure measured during the MERC test
program. 3-13
3-8 Flue gas temperature as a function of time during the MERC
test program. 3-14
3-9 Spray dryer operating parameters as a function of time during-
the MERC test progam 3-15
3-10 Differential pressures across the control devices during the
MERC test program 3-16
xi
-------
FIGURES (continued)
Number Page
4-1 Sampling and monitoring locations—Unit A.. 4-2
4-2 Spray dryer inlet—location of sampling ports and traverse
points.. 4-3
4-3 Baghouse outlet—location of sampling points and traverse
points. 4—5
5-1 Modified method 5 sampling train for semivolatile organic
compounds.. 5-3
5-2 Field lab set-up data sheet 5-4
5-3 Field lab recovery data , 5-7
5-4 Sample extraction scheme for inlet MM5 samples.... 5-9
5-5 Sample extraction scheme for outlet and blank MM5 samples 5-10
5-6 Sample cleanup and analysis scheme for PCDD/PCDF and other
organics in the MM5 train 5-11
5-7 Modified method 5 sampling train for particulates and metals... 5-18
5-8 MERC facility particulate/metals MM5 train field recovery
and analytical protocol 5-21
5-9 ' Schematic of spray dryer .inlet CEM equipment. 5-29
5-10 Schematic of spray dryer outlet CEM equipment........ 5-31
5-11 Schematic of baghouse outlet CEM equipment.. 5-33
xi i
-------
TABLES
Number Page
1-1 Summary of Sampling and Analysis Parameters and Methods 1-5
1-2 PCDO/PCDF AnaTytes 1-8
2-1 Summary of Key Operating Parameers During the MERC Test
Program in Biddeford, Maine..... 2-2
2-2 Summary of Selected Process Conditions and PCDD/PCDF
Emissions for MERC... 2-4
2-3 Uncontrolled PCDD/PCDF Emissions........ 2-5
2-4 Controlled PCDD/PCDF Emissions.. 2-6
2-5 Uncontrolled Flue Gas PCDD/PCDF Congener Distribution 2-7
2-6 Controlled Flue Gas PCDD/PCDF Congener Distribution 2-8
2-7 2,3,7,8-TCDD Toxic Equivalencies of Uncontrolled Emissions..... 2-13
2-8 2,3,7,8-TCDD Toxic Equivalencies of Controlled Emissions... 2-14
2-9 Ratio of Uncontrolled PCDD/PCDF to Particulate Emissions....... 2-15
2-10 Ratio of Controlled PCDD/PCDF to Particulate Emissions......... 2-16
2-11 Summary of Particulate Emissions for MERC Facility. 2-18
2-12 Specific Metals Mass Emissions Rates for MERC (Normalized to
12% CQ2 2-19
2-13 Ratio of Metals, to Particulate Mass for MERC................... 2-20
2-14 ' Metals Content of Process Samples....... 2-21
2-15 Process Samples—Bulk Characteristics........ 2-23
2-16 CEM Data Summary—Acid Gases.... 2-25
2-17 Molar Ratio of Actual Lime to Stoichiometric Lime for HC1 and
SO^a............. 2—33
2-18 CEM Data Summary—Other Gases.. 2-34
5-1 Summary of Sampling and Analytical Procedures.. 5-2
5-2 CEM Equipment Used at Spray Dryer Inlet 5-30
5-3 CEM Equipment Used at Spray Dryer Outlet 5-32
5-4 CEM Equipment Used at Baghouse Outlet.. 5-34
6-1 D1oxin/Furan Surrogate Recoveries for Field and QC Samples
for MERC MWC 6-3
6-2 Response Factor Comparison Table for Total Tetra-Octa
CDD/CDF, 6-5
6-3 Response Factor Comparison Table for Specific Tetra-Octa
CDD/CDF 6-6
6-4 Summary of Blank Train Data. 6-7
6-5 Precision Results for Duplicate Injection of MM5 Sample........ 6-8
6-6 Duplicate Metals Analysis Results............. 6-9
6-7 Ash and Lime Slurry Metals Spike and Recovery Data.. 6-10
6-8 Metals Instrument Check Standard Data and Percent Drift
Calculations. 6-12
6-9 Metals Blank Train Analyses........ 6-14
.Kill
-------
TABLES (continued)
Number Page
6-10 Calibration Results for Sampling Equipment 6-15
6-11 Semi volatiles Testing Isokinetics and Leak Check Summary 6-16
6-12 • Particulate/Metals Testing Isokinetics and Leak Check
Summary6—17
6-13 Calibration Gases. 6-18
6-14 Daily Zero Drift .in CEMS . 6-19
6-15 Daily Span Drift in CEMS 6-20
6-16 Span Cylinder Accuracy Checks. . 6-22
6-17 Precision for Duplicate Analyses of Process Samples.... 6-23'
6-18 Dioxin/Furan Results for the Instrument Performance Sample..... 6-24
6-19 Dioxin/Furan Results for Blank QA Performance Samples 6-24
6-20 Total Oioxin/Furan Results for Spiked QA' Performance Samples... 6-25
6-21 Isomer Specific Dioxin/Furan Results for Spiked QA
'Performance Samples 6-26
6-22 EPA Audit Samples PCDD/PCDF Results. 6-28
6-23 EPA Audit Results... 6-30
6-24 Results of EPA and Internal Metals Audit Sample Analysis 6-31
xiv
-------
LIST OF ABBREVIATIONS
AA Atonic absorption
AEERL A1r and Energy Engineering Research Laboratory
Btu British thermal unit
CDFE Chlorinated diphenyl ethers
CEH Continuous emission monitor
CVAAS Cold vapor atomic absorption spectroscopy •
DOPE Decachlcrlnated diphenyl ethers
dscf Dry standard' cubic foot
dscm Dry standard cubic meter
EICP Extracted ion current profile
EPA Environmental Protection Agency
ESED Emission Standards and Engineering Division
FD Forced draft
FSCC Fused silica capillary column
ft Foot
GC/MS Gas chrcmatography/mass spectroscopy
GFAAS Graphite furnace atomic absorption spectroscopy
gpi Gallon per minute
h Hour
HpCDD/HpCDF Heptachlorinated dibenzodioxin/furan
HRGC High resolution gas chromatography
HxCDD/HxCDF Hexachlorinated dibenzodioxin/furan
ICAP Inductively coupled argon plasma
10 Induced draft
in. Inch •
kg Kilogram
in3 Cubic meter
MERC Maine Energy Recovery Company
mg Milligram
MM5 .Modified Method 5 sampling train
MRI Midwest Research Institute
MS-SIM Mass spectrosocpy-selected ion monitoring
MSW Municipal solid waste
MW Megawatt
MWC Municipal waste conbustor
NDIR Nondisperslve infrared
NDPE Ncnachlorinated diphenyl ether
ng Nanogram
0CDD/0CDF , Octachlorinated dibenzodioxin/furan
PAS Performance audit sample
PCDO/PCDF Polychlorinated dibenzodioxin/furan
PeCDO/PeCOF Pentachlorinated dibenzodioxin/furan
PM Particulate matter
xv
-------
LIST OF ABBREVIATIONS (Continued)
ppm Parts per million
QAC Quality Assurance Coordinator
QAP Quality Assurance Plan
QA/QC Quality assurance/quality control
RDF Refuse-derived fuel
RPD Relative percent difference
RRF Relative response factor
SD/FF Spray dryer/fabric filter
TCDD/TCDF Tetrachlorinated dlbenzodioxins/furans
THC Total hydrocarbon
VOC Volatile organic compound
xv i
-------
SECTION LO
INTRODUCTION
As part of the U.S. Environmental Protection Agency's (EPA) effort to
develop and support regulations for municipal waste combustors under Sec-
tion 111 of the Clean .Air Act, EPA is sponsoring test projects at several new
municipal waste combustor (MWC) facilities. These test projects include mea-
surements to determine the emission levels of criteria pollutants, acid gases,
heavy metals, and semi volatile organic compounds as well as the collection
efficiencies of associated emission control systems.
Prior to EPA's decision to develop regulations for.MWCs,. Midwest Research
Institute (MRI), under contract to EPA, compiled the available data base for
the pollutants of interest {i.e., criteria pollutants, acid gases, metals, and
semivclatile organic compounds).1 Review of the omission data base was per-
formed to determine the information gaps regarding achievable emission
levels. Virtually no information was available for the pollutants of interest
from new MWC facilities that fired refuse-derived fuel (RDF). The Maine
Energy Recovery Company (MERC) located in Biddeford, Maine, was identified as
the first new RDF-firing MWC with wet-dry (spray dryer) scrubbing and a high
efficiency particulate collector to come on-line in the United States in the
1980s. To take advantage of the first opportunity to fill this information
gap, EPA made a cooperative arrangement with KTI Holdings, Inc. (the owner/
operator of MERC) to conduct emission measurements for the pollutants of
interest. This report describes the details of the measurement procedures and
the results of the MERC Test Project.
1.1 PROCESS DESCRIPTION
The Maine Energy Recovery Company (MERC) RDF power plant is located in
Biddeford, Maine (see Figure 1-1). The plant is currently operating at a
capacity of 454 Mg/d (500 tons/day) of municipal waste and wood chips.* The
wood chips, and to a lesser extent, fuel oil and natural gas, are supplements
to accommodate fluctuations in refuse volume and energy content of the RDF.
There are two combustors each consisting of a Detroit Stoker RDF spreader
stoker and a Babcock &.Wilcox controlled combustion zone boiler. Each unit is
rated at 158,200 MJ/h (150 x 108 Btu/h). In combination they can provide
steam for electric power generation up to 22 MW. For this test series, the
plant was fired with RDF only.
* A table for conversion of English units to Standard International (SI)
units is provided at the end of this volume.
1-1
-------
TIPPING
FLOOR
MSW
INFEED
MAGNETIC SEPARATOR
QTD
TROMMEL
SCREEN
SECONDARY SHREDDER
FLAIL
MILL
LOAD CELL I
U
i
o
FERROUS METAL
HOPPER
RDF
NON-PROCESSIBLES
TO LANDFILL
wooUips sTf hofWr turbine
SHEAR SHREDDER
FOR OVERSIZED WASTE
H P STEAM
POWER TO UTILITY
~bn
SPRAY DRYER/FABRIC FILTER
\AA7
C
BOILER
d1
ASH BIN
STACK
Figure 1-1. MERC Waste-to-Energy Process'(Adapted From KTI Holding Co. Brochure).
I
-------
Municipal solid waste (MSW) is received 1n packer trucks arid transfer
trailers and is off-loaded onto the floor of the tipping floor building. Non-
combustibles and potentially explosive or hazardous items are sorted and
removed by visual inspection and bucket loader. The MSW is processed using a
flail ¦mill, a magnetic separator, a trommel screen, and a secondary
shredder. The resultant RDF has a nominal 10-cm (4.0-1n.) top. size. - The
plant has the capability of burning small quantities of sewage sludge, but it
was not fired during the tests.
Dust control within the processing building is achieved through two sepa-
rate control systems. One system serves the tipping/processing area, while
the other serves the conveyors in the boiler building and RDF reclaim area.
Each system contains a bagrouse, fan dust hoods, and dust collection ducts at
key conveyor and transfer processing points. Dust-laden air is drawn through
one of two pulsed jet baghouses which exhaust in the vicinity of the boiler
forced draft fan intake. The baghouse exhaust thus becomes Incorporated into
the combustion air for the boilers. Dust contained by the baghouse is
returned to and becomes a part of RDF fuel.
Each boiler Is a balanced draft (employing both forced and induced draft
fans) water wall power boiler with a super heater, economizer, and air. pre-
heater. A traveling grate stoker 1s located at the bottom of the furnace.
Fuel from metered feeders is admitted above the grate at the front of the
boiler. A single auxiliary burner capable of firing natural gas or No. 2 fuel
oil is located on the right furnace sidewall directly above the primary fuel
combustion zone. It is used for startup, shutdown, and for periods where load
stabilization is- required. Medium-pressure superheated steam is delivered
from the boilers to a steam turbine which supplies power to an electric
generator.
Combustion air for the solid fuels is introduced into the furnace as
undergrate and overfire air. The boiler configuration and the location of
overfire air ports are designed to promote mixing and complete combustion of
organic material injected into the boiler as fuel.
A transmissometer in conjunction with oxygen (02) and carbon monoxide
(CO) monitor is used to optimize combustion efficiency and to control organic
compounds and CO. These monitors are equipped with continuous recording
devices.
Combustion gases from each boiler enter a spray dry adsorber followed by
a fabric filter. Exhaust from each baghouse vents through a common 74-m
(244-ft). stack. The spray dryer is a reaction vessel where a slurry of slaked
lime is sprayed into the flue gas which contains particulate matter, acid
gases, and other pollutants in gaseous and aerosol form. The slurry water is
evaporated by the flue gas heat, and the acid gases react with the slaked lime
(calcium hydroxide). Particulates, post-reaction compounds, and excess sor-
bent serve as nucleaticn points for adsorption-and agglomeration of volatile
metals and semivolatile organics. Some of the solids settle in the reaction
vessel and are removed. A baghouse then collects the remaining particulate
from-the gas stream. The excess lime in the bag filter cake provides a second
stage reaction site for further acid gas removal.
1-3
-------
The ash system removes ash from the grate discharge, generating .bank
hopper, air heater hopper, mechanical dust collector hopper, spray dryer, and
baghouse modules.
The boiler bottom ash discharges into a quench pit with a water seal and
is removed by a drag chain up an inclined dewatering conveyor. All of the
hopper discharges are through rotary seal valves. This ensures a positive
seal and prevents boiler gases from entering the ash conveyors and air from
entering the hoppers and boilers.
The ash from the baghouse modules discharges into six identical drag/
screw conveyors. Each set of these drag/screw conveyors then discharges into
one of two identical drag chain collecting conveyors.
The spray dryer and mechanical dust collector each discharge solids
directly onto the collecting drag chain conveyor. The generating hopper.and
air heater hopper discharge ash onto the transverse drag conveyor,' and it is
then deposited on the collecting conveyors into one of two identical ash con-
ditioning screw conveyors per system. The third screw conveyor in each system
is the ash conditioner. The water flow rate and screw speed are automatically
varied depending on the ash level within the conveyor. The ash conditioners
discharge onto the dewatering section of the bottom ash drag conveyor. It is
at this point that the fly and bottom ash streams mix. The bottom ash from
the stoker discharges into one of two submerged ash conveyors. The combined
ash streams are then dumped into a specially designed trailer for removal from
the site to a landfill.
1.2 MEASUREMENT PROGRAM
A total of three tests were conducted at the MERC facility during the
period of December 9 to 12, 1987.
The test matrix for this program is summarized in Table 1-1. MRI con-
ducted all sampling activities for the program, with the exception of the HC1
continuous emission monitoring system (CEM), which was operated by Entropy
•Environmentalists Inc.2 Radian Corporation provided support in monitoring the
process conditions during the test.
The basic,sampling program included:
1. , Sampling for polychlorinated dibenzodioxins (PCDD), polychlorinated
dibenzofurans (PCDF), particulate matter (PM), cadmium (Cd),
chromium (Cr), arsenic (As), lead (Pb), mercury (Hg), oxygen (02),
and carbon dioxide (CO2) at the spray dryer inlet.
2. Continuous emission monitoring of 02, CO, C02, sulfur dioxide (S02),
hydrogen chloride (HC1), and total hydrocarbons (THC} at the spray
dryer inlet.
3. Continuous monitoring of HC1, 02, and C02 at the fabric fiIter
inlet.
1-4
-------
TABLE 1-1. SUIWARY OP SAMPLING AND ANALYSIS PARAMETERS AND METHODS
Sampling location
Sample type
Sampling
fr equency/du ra t i on
for each run
Sanpl ing
nethod
Sample size
Analytical
parameters
Preparation
oethod^
Analytical method
L-Spray dryer
inlet
Combustion gas Continuous 4 h
2-Spray dryer
outlet
3-Fabric filter
outlet
Combustion gas
Combustion gas
Continuous 4 h
Continuous 4 h
MHS P/Hc
HH5-SV0f
M]
CEHS (HR1)
HM5-P/)tc
Run l--one shovelful
per grab. Three grabs
total.
Runs 2 and 3 —19 L pail
fuls per grab from the
ran feed. Three grabs
per run.
> 120 ftJ
> 140 Ft3
" 30 L
N/A
CEMS (Entropy)1
CFMS (MRI) N/A
CEHS (Entropy)1
120 ftJ
MM5-SV0 > 140 ftJ
M3 " 30 L
CEHS (MRI) N/A
CEMS (Entropy)*
Fiber drum
Particulate
Hetalsd
PCD0/PCDF9
"z- °2
CO , CO
i.
HC1
£
Particulate
Hetalsd
PCDO/PCOF^
co2. o2
so" •
HCI
Desiccation
Acid digestion
Solvent extraction
N/A
Gas conditioning
Gas conditioning
Des iccat ion
Acid digestion
Solvent extraction
N/A
N/A
N/A
Gravimetric (EPA M5)
ICAP/AASe
HRGC/HSh
Orsat
NDIR analyzer
Paramagnetic
Electrocheaical
Heated FID (Becknan 402)
1R gas filter correlation
HOIR analyzer
Polarographic analyzer
Specific ion electrode
Gravimetric (EPA MS)
ICAP/AASe
HRGC/HSh
Orsat
NDIR analyzer
Polarographic analyzer
Cheni luminescence
Electrocheaical
IR gas filter correlation
Archived
(cont inued)
-------
TABLE l-l (continued)
Sampling location3 Sample type
Sanpl ing
frequency/duration
for each run
Sanpling
nethod
Sanple size
Analytical
parameters
Preparation
method**
Analytical nethod**
A-Cyclone ash
discharge
Fly ash
Two grab samples
every 30 nin coo-
posited into two
separate sanples
each run
Scoop (S007)
1 kg
Metals
X Combustibles
X Carbon
Acid digestion
N/A
N/A
ICAP/AASe
AS IN Ei30
ASTM L 77 7
B-Fabric Filter
(bagtouse)
#Fly ash/spray
dryer residue
One grab sanple every
60 nin. cooposited
and split into four
jars for each run.
Scoop (S007)
1 kg
Metals
X Conbustibles
X Carbon
Resistivity
K factor
Acid digestion
N/A
N/A
N/A
N/A
ICAP/AASe
ASTM E830
ASTM E777
IEEE 548 1984
N/A
C-Botton ash
discharge
Bottom ash
i
CTi
D-Spray dryer inlet Lime slurry
Two grab sanples
every 30 nin con-
posited into two
separate sanples
each run
One grab sample every
hour composited into
one sanple for entire
test
Scoop (S007)
Tap (S004)
1 kg
LOO nL per grab
Metals
X Combustibles
X Carbon
Hetals
Acid digestion
N/A
N/A
Acid digestion
1CAP/AAS
ASTM E630
ASTM E777
ICAP/AASe
a Nunbers or letters refer to Figure 1-2.
b Sdraple preparation and analytical methods are described in detail in the Appendix of the Project Quality Assurance Work Plan referencing SW-B46 methods and draft netals protocols.
c Modified Method 5 train for particulates and netals.
^ Target metals are cadmium, total chroniun, mercury, lead, and arsenic.
e Inductively coupled argon plasma atomic emission spectroscopy and graphite furnace atomic absorption spectrometry.
f Modified Method 5 train for semivolatile organics.
9 PCDD/PCDF includes all tetra through octa dioxins and furans, all 2,3.7,8-substituted isomers, and to the extent possible, 1,2,3,4,8,-PeCDF and 1,2,3,4,7,9-HxCDF.
^ High resolution gas chromaLography/mass spectroscopy.
1 Entropy Environmentalists Inc., is conducting the HC1 monitoring under separate contract.
-------
4. Sampling for PCDD, PCDF, PM, Cd, Cr, As, Pbf Hg, 02s and C02 at the
fabric filter outlet.
5. Continuous monitoring of HC1, C02, S02, NO , and 02 at the fabric
filter outlet.
The specific PCDD/PCDF congeners of interest requested by EPA are listed
in Table 1-2.
In addition to the combustion gas sampling, RDF, ash, and lime slurry
samples were collected for analysis. Figure 1-2 identifies the sampling loca-
tions. All sampling for this project was conducted on the "A" side
combustor/boiler.
Operating parameters for the RDF combustion process were monitored by a
computer-controlled system. Seventy-two process parameters were selected
based on previous testing performed by Babcock and Wilcox, the combustor manu-
facturer. Data were logged using a portable computer linked to a printer port
of the facility's computer-controlled system. The selected operating param-
eters were scanned at intervals of 4 min, with data being transferred to the
portable computer and stored in Lotus 1-2-3"" files. The 4-min readings were
averaged for each test period. Intervals of 15 min or longer during which
flue gas sampling was suspended were not included in the test period averages.
1.3 QUALITY ASSURANCE/QUALITY CONTROL (QA/QC)
• An extensive QA/QC program was developed for this project. Audit samples
for metals and dioxln/furans were provided by both EPA and the MRI quality
assurance coordinator. Criteria for calibration accuracy and drift were
developed for both sampling equipment and the CEMs, as well as for the labora-
tory analyses. Selected samples were analyzed in duplicate. An external
audit was also conducted by Research Triangle Institute (RTI) in January
1988. The results of the QA/QC checks are discussed in Section 6.0.
1.4 DESCRIPTION OF REPORT SECTIONS
This report is presented in three volumes. Volume I includes the Summary
of Results (Section 2.0), Process Description and Operation (Section 3.0),
Sampling Locations (Section 4.0), Sampling and Analytical Procedures (Sec-
tion 5.0), Quality Assurance/Quality Control (Section 6.0), and References
(Section 7.0).
Volumes II and III contain the Appendices to this report with copies of
the field and laboratory data. Volume II includes sample calculations, pro-
cess data, field data sheets for the dioxin/furan and metals/particulate
trains, process sampling data sheets, and CEM data. Volume III contains
laboratory analysis for dicxin/furans and metals/particulates, miscellaneous
laboratory analyses, sampling train calibration data, field test logs, QA
information, sampling and analysis protocols, and the list of participants.
1-7
-------
TABLE 1-2. PCDD/PCDF ANALYTES
Isomer
Dioxins
Furans
,8-TCDD
TCDD
PeCDD
,4,7,8-HxCCD
2,3,7
Total
1,2,3,7,8-PeCOD
Total
1,2,3
1.2.3.6.7.8-HxCDD
1.2.3.7.8.9-HxCDD
HxCDD
,4,6,7,8-HpCDD
HpCDD
OCDD
Total
1,2,3
Total
Total
2,3,7,8-TCDF
Total TCDF
1,2,3,4,8-PeCDF
1.2.3,
2.3.4,
Total
1,2,3,
1,2,3,
1.2.3,
2.3.4,
1,2,3,
Total
1,2,3,
1,2,3,
Total
Total
7,8-PeCDF
7s8-PeCDF
PeCDF
4,7,8-HxCDF
6.7.8-HxCDF
4.7.9-HxCDF
8-HxCOF
9-HxCDF
HxCDF
4.6.7.8-HpCOF
4.7.8.9-HpCOF
HpCOF
OCDF
6.7
7.8
1-8
-------
o2,co, Cq,,S02,HCI, THC
hci,o2, co2,so2,nox
Opacily
M—F
Economizer
Spray
Dryer
Absorber
[Scrubber)
Cyclone
RDF
ID Fan
Bottom
Grate
Ash
Ash
Discharge
Stack
~ Preheater
Fabric
Filter
Baghouso
Combustion Gas
Ash Discharge
o Sample Locations
£~"j Ash Sample.Locations
A Plant CEMs
Off Line During Test
Identical
Boiler Unit B
Figure 1-2. Sampling and Monitoring Locations, Unit A.
-------
SECTION 2.0
SUMMARY OF EMISSION RESULTS
This section summarizes the results of the test program conducted at the
MERC facility in Biddeford, Maine, during the period from December 9 through
12, 1987. Standard international units (SIU) are used to present most of the
data. The main exceptions are data related to the process; these are shown in
the customary English units. A table providing conversion factors between SIU
and English units is provided at the end of this volume. In this section
trace organic and metal emissions are normalized to 1236 C02. Sample calcula-
tions are given in Appendix A. Uncorrected results and raw data are shown in
Appendices B through M.
Note that the nongaseous related results (PCDD/PCDF, particulate mass
loading, and metals) do not represent true furnace or uncontrolled emissions
because the spray dryer inlet sampling was conducted after a cyclone which
removes large particles. However, in this report the term uncontrolled emis-
sions will be used to identify emissions measured at the input to the spray
dryer/baghouse control system.
2.1 PROCESS DATA
The facility burned 1003£ RDF at full load conditions without auxiliary
fuel during all three test runs. Key operating parameters measured during
each test are summarized in Table 2-1. A complete process description is in
Section 3.0, Process Description and Operation, and detailed process logging
data are in Appendix B, Process Data. One major change in operation occurred
during the test series. The lime slurry feed rate was more than doubled about
45 min after the start of run 2 and held at that level for the remaining
tests. Facility-operators made this change because MRI's continuous, emission
monitor (OEM) .for SO, at the baghouse outlet was reporting emissions which
exceeded the facility operating permit.
Intermittent process problems occurred during all three tests and were
primarily related to RDF conveyor feed malfunctions. During test runs 1 and 3
the problems were severe enough to end the test early. In both cases the sam-
pling teams had completed at least two-thirds of the test period, and the par-
tial tests were judged to be acceptable by EPA personnel on site.
None of the process monitoring equipment was calibrated during the test
programs although Radian did perform a review of the data and compared read-
ings to the expected design values. With the exception of the plant S0t moni-
tor, which was not operational during the tests, there were no difficulties
encountered with the process monitoring data.
2-1
-------
TABLE 2-1. SUMMARY OF KEY OPERATING PARAMETERS DURING THE MERC TEST PROGRAM IN BIOCEFCRD, MAINE
Run 1
Run 2
Run 3
12-09-87
12-10-87
12-12-87
Average
Superheater steam
Flow rate (1,000 Ib/h)
106
109
108
108
Pressure (psig)
663
676
671
670
Outlet temperature (*F)
746
751
748
748
Combustion air
lotal airflow rate (1,000 Ib/h)
124
123
134
127
Undergrate airflow rate (1,000 1b/h)a
53
50
63
55
Overfire airflow rate (1,000 Jb/h)
71
73
70
71
Overfire air distribution (?)
57
59
52
56'
Undergrate air pressure (in. HzO)
-0.23
-0.86
-0.26
-0.45
Overfire air fan pressure (in. H20)
25.3
25.6
25.0
25.3
Air heater inlet air temperature (°F)
127
66
118
104
Air heater outlet air temperature (°F)
381
368
385
378
Excess oxygen {% by volume, wet)
Left side
5.59
5.77
5.78
5.71
Right side
7.91
8.13
8.02
8.02
Heat release (106 Btu/h>
Total (RDF + auxiliary fuel)
150
153
151
151
RDF only
150
153
150
151
Flue gas temperatures [°F)
Economizer inlet
779
788
801
789
Economizer outlet/air heater inlet
515
523
532
523
Air heater outlet
374
363
383
373
Spray dryer inlet
374
364
384
374
Spray dryer outlet/fabric filter inlet
277
278
279
278
Fabric fiIter outlet
268
268
268
268 •
Gas differential pressures (in. H?0)
Undergrate to furnace
0.46
0.34
0.44
0.41
Dust collector (cyclone)
3.02
3.07
3.37
3.15
Spray dryer
4.24
4.84
5.17
4.75
Fabric fiIter
7.16
7.89
8.22
7.75
Flue gas pressures (in. HjO)
Spray dryer inlet
-7.20
-7.25
-7.39
-7.28
Spray dryer outlet
-11.5 ,
-13.1
-13.4
-12.7
1.D. fan suction
-18.7
-21.0
-21.7
-20.5
Lime slurry feed rate (gpm)
2.91
6.70
7.80
5.80
Dilution water fee/3 rate (gpm)
6.95
3.39
4.89-
5.07
Total, lime slurry and water feed rate (gpm)
9.36
10.1
- 12.7
10.9
8 Undergrate airflow rate was calculated as the difference between the total airflow rate ana
overfire airflow rate,
^ Overfire air distribution was calculated as the overt ire airflow rate divided by the total
airflow rate.
2-2
-------
2,2 PCDD/PCDF EMISSIONS
Table 2-2 summarizes measured values of PCDD and PCDF at the inlet to the
spray dryer (uncontrolled) and the outlet of the baghouse (controlled). Rele-
vant data are presented in Appendix C (field data) and Appendix G (laboratory
data). The blank train analyses showed insignificant contamination. There-
fore, no blank correction was used. Blank train results are presented in
Section 6.2,1.
The PCDF fraction is about twice as large as the PCDO fraction for both
controlled and uncontrolled emissions. There is no significant variation
among the three runs. The average total PCDD/PCDF emission rate was 55 mg/h
(877 ng/dscm) uncontrolled and 0.3 mg/h (4.3 ng/dscm) controlled representing
an efficiency of about 99.5% for both the dioxins and furans.
The uncontrolled samples had separate analyses for the front and back
halves of the sampling train. The back half fractions showed concentrations
near detection limits indicating that the PCDD/PCDF are predominately asso-
ciated with the particulate matter at the control device inlet. Details of
the distribution are given in Appendix G including the GC/MS report sheets for
all samples.
Tables 2-3 (uncontrolled) and 2-4 (controlled) show the isomer specific
PCDD/PCDF concentrations measured. All 17 specific isomers listed in
Table 1-1 are well above detection limits in the uncontrolled samples for each
test. For the controlled samples, all 17 isomers were detected in the com-
pleted test (run 2), and most were also detected in the two partial runs.
Tables 2-5 (uncontrolled) and 2-6 (controlled) show the. PCDD/PCDF con-
gener distribution as mole fractions. Figures 2-1 (uncontrolled) and 2-2
(controlled) are plots of mole fraction vs. chlorine number of each PCDO con-
gener. Figures 2-3 and 2-4 are the mole fraction plots for each PCDF con-
gener. No significant change in the distributions occurred across the control
devices.
The EPA 2,3,7,8-PCDD toxic equivalencies3 are shown in Tables 2-7 (uncon-
trolled) and 2-8 (controlled). Due to lower equivalent toxicity for furans,
the total toxicity is similar for the dioxins and furans. 2,3,7,3-TCDD
accounts for -about 10% (uncontrolled) to 20% (controlled) of the total
toxicity.
Average controlled emissions of PCDD and PCDF are about 45 ng/g and
100 ng/g of particulate, respectively. Although the ratio of PC0D/PC0F to
particulate loading, is about 2Q% higher for controlled (Table 2-9) than for
uncontrolled (Table 2-10), this is not statistically significant considering
the mean and standard deviations of the two data sets.
2-3
-------
TABLE 2-2. SLMHARY OF SELECTED PROCESS CONDITIONS AND PCDD/PCDF EMISSIONS FOR HERC
Run 1 (12-09-87)
Uncontrolled Control led
Run 2 (12-10-67)
Control
efficiency
(*>
Uncontrolled Controlled
Run 3 (12-12-87)
Control
efficiency
«>
Uncontrolled Controlled
Average
Control
efficiency
<*)
Uncontrolled
Control
efficiency
Controlled (%)
Hue Gas Characteristics
Flow rate (dscfn)
Tenperature ("I)
Moisture (% by volute)
CO^ (% by volune)
02 (X by volune)
38.300
371
14.3
10.9
B.6
39.200
270
15.3
10.9
8.6
40.500
302
14.4
U.O
8.5
41.100
269
13.5
10.9
8.6
41,000
381
16.0
11.3
8.2
42.500
272
17.0
11.3
8.2
39.900
371
14.9
11.1
8.4
40,900
270
15.3
U.O
8.5
Process Operations
Stean load (10"* lb/h)
106
106
109
109
108
108
108
108
ro
-P*
Itlll) Results
Total PCDD (ng/dscn) 217
Total PCDU (corrected to 239
12% CO^, ng/dscn)
PCDF Results
Total PCDF (ng/dscn) 540
Total PCllf (corrected to 594
12% C0?, ng/dscn)
PCDO/PCDF Results
Total PCDD/PCOF (ng/dscn) 757
Total PCDD/PCDF (corrected to 833
12% CO^. ng/dscn)
2.3,7,8-TCDD toxic equivalent 11.4
(ng/dscn at 12% C02)a
1.25
1.38
2.67
2.94
3.93
4.32
0.066
99.42
99.42
99.50
99.50
99.48
99.48
99.42
259
282
514
560
772
842
12.9
1.45
1.59
3.14
3.45
4.58
5.04
0.087.
99.44
99.44
99.38
99.38
99.40
99.40
99.33
323
342
580
615
903
957
13.6
0.972
1.03
2.23
2.36
3.20
3.39
0.058
99.70
99.70
99.62
99.62
99.64
99.64
99.57
266
288
545
590
811
877
12.6
1.22
1.33
2.68
2.92
3.90
4.25
0.071
99.54
99.54
99.51
99.50
99.52
99.51
99.44
Enission Rales (uq/h)
Total PCDD
Total PCDF
Total PCDD/PCDF
14.000
35.200
49.200
83.3
17B
261
17.800
35.400
53.200
101
219
320
22.500
40,400
62.900
70.2
161
231
18.100
37.000
55,100
84.8
186
271
a USEPfl. Interim Procedures for Estimating Risks Associated with Exposures to Mixtures of PCDOs/PCDfs. EPA-625/3 87/012.3
I
I
-------
TABLE 2-3. UNCONTROLLED PCOD/PCDF EMISSIONS
. . , . , Corrected io 121 CO- (ng/dscm)a
Uncorrected (ng/dscm) ; £
Isomer Run I Run 2 Run 3 Average Run 1 Run 2 Run 3 Average
Diox i ns
2,3,7,8-TCDD
0.72
1 .5
1.5
1.2
0.80
1.6
1 ,
.6
1.3
Other TCDD
32
40
48
40
35
43
51
43
1,2,3,7,8-PeCDD
3.9
4.3
5.0
4.4
4.3
4.7
5.
.3
4.8
Other PeCDO
35
43
62
47
39
47
66
51
1,2,3,4,7,8-HxCDD
4.1
5.3
6.0
5.1
4.5
5.8
6,
,4
5.6
1,2,3,6,7,8-HxCDD
8.1
11
12
10.1
8.9
12
12
10.9
1,2,3,7,8,9-HxCDD
6.7
9.1
10.2
8.6
7.3
9.9
11
9.3
Other HxCDD
39
52
62
51
43
57
66
56
1,2,3,4,6,7,8-HpCDD
25
26
34
28
27
28
36
31
Other ElpCDD
21
24
32
26
23
26
34
28
OCDD
41
43
49
44
,45
47
52
48
Total PCDD
217
259
323
266
239
282
342
288
Eurans
2,3,7,8-TCDF
29
30
32
30
32
33
34
33
Other TCDF
199
182
192
191
219
198
204
207
1,2,3,4,8-PeCDF
2.9
3.7
3.5
3.4
3.2
4.0
3.
8
3.7
1,2,3,7,8-PeCDF
14
14
13
13.7
16
15
14
15
2,3,4,7,8-PeCDF
14
14
15
14.3
15
16
16
16
Other PeCDF
130
129
141
133
144
140
150
144
1,2,3,4 ,7,8-HxCDF
16
17
20
18
18
18
21
19
1,2,3,6,7,8-HxCDF
9.6
10
10
10.2
10.6
11
11
11
1,2,3,4,7 ,9-llxCDE
2.0
2.2
2.5
2.3
2.2
2.4
2.
,7
2.4
2,3,4,6,7,8-HxCDF
8.3
5.0
13
8.9
9.1
5.5
14
9.6
1,2,3,7,8,9-HxCDF
2.1
1 .0
1.2
1 .4
2.3
1. 1
1.
.3
1.(5
Other HxCDF
53
48
62
54
58
52
66
59
1,2,3,4,6,7,8-HpCDF
31
28
35
31
34
30
37
34
1,2,3,4,7,8,9-HpCDF
5.0
4.7
6.6
5.4
5.5
5.1
6.
,9
5.8
Other HpCDF
13
13
17
14
14
14
18
15
OCDF
12
12
16
13
13
13
17
14
Total PCDF
540
513
580
545
594
560
615
590
Tota1 PCDD + PCDF
758
772
903
81 1
833
842
957
877
COj correction factor3
1.1
1 .09
1.06
a CO^ correction factor x measured value = value corrected (normalized) to 12J CO^.
-------
TABLE 2-4. CONTROLLED PCDD/PCDF EMISSIONS
, b
...... Corrected to \2% CO- (ng/dscm)
Uncorrected (ng/dscm) . 2 J
Isomer Run I Run 2 Run 3 Average Run 1 Run 2 Run 3 Average
Diox i ns
2,3,7,8-TCDD
0,
.013
0,
.031
0,
,1la
0.015
0.014
0.034
0.01la
0.016
Other TCDD
0.
.17
0,
.17
0,
.11
0.15
0.19
0.19
0.11
0.16
1,2,3,7,8-PeCDD
0.
.020
0.
.021
0,
.023
0.021
0.022
0.023
0.024
0.023
Other PeCDD
0.
. 18
0,
.31
0,
.11
0.20
0.20
0.34
0.12
0.22
1 ,2,3,4,7,8-llxCDD
0.
.019
0.
.022
0.
.022
0.021
0.021
0.024
0.023
0.023
1,2,3,6,7,8-HxCDD
0,
.042
0.
.047
0,
.041
0.044
0.047
0.052
0.043
0.047
1 , 2,3,7,8,9-HxCDD
0.
.033
0,
.042
0,
.035
0.037
0.037
0.046
0.037
0.040
Other HxCDD
0.
.21
0.
.21
0,
.13
0.18
0.23
0.23
0.13
0.20
1 ,2,3,4,6,7,8-HpCDD
0,
,14
0.
. 16
0,
.13
0.14
0.16
0.18
0.13
0.15
Other HpCDD
0,
.13
0,
.14
0,
.11
0.13
0.14
0.15
0.12
0.14
OCDD
0.
.29
0.
.29
0.
.27
0.28
0.32
0.32
0.28
0.31
Total PCDD
1 ,
.3
1 .
.4
0,
.98
1.2
1 .4
1 .6
1 .0
1.3
Furans
2,3,7,8-TCDF
0.
.14
0.
,14
0.
.14
0.14
0.15
0.15
0.15
0.15
Other TCDF
0.
,90
0.
.85
0.
.75
0.83
0.99
0.94
0.80
0.91
1,2,3,4,8-PeCDF
0.
.016
0.
.016
0.
,011
0.014
0.017
0.018
0.012
0.016
1,2,3,7,8-PeCDF
0.
,063
0.
.065
0.
,044
0.057
0.069
0.071
0.046
0.062
2,3,4,7,8-PeCDF
0.
.064
0.
.059
0.
,055
0.059
0.070
0.065
0.058
0.064
Other PeCDF
0.
69
0.
,63
0.
,53
0.61
0.75
0.69
0.56
0.67
1,2,3,4,7,8-HxCDF
0,
.093
0,
,089
0.
.088
0.090
0.10
0.98
0.094
0.098
1,2,3,6,7,8-HxCDF
0.
,048
0.
,051
0.
,043
0.047
0.052
0.056
0.046
0.051
1,2,3,4,7,9-HxCDF
0,
,01 4a
0,
.0064
0,
,035a
0.0021
0.016a
0.0070
C.037a
0.0023
2,3,4,6,7,8,-HxCDF
0.
,055
0.
,038
0.
,042
0.045
0.060
0.042
0.045
0.049
1,2,3,7,8,9-HxCDF
0.
,0037
0.
,0055
0.
,0l0a
0.0031
0.0041
0.006
0.011a
0.0034
Other HxCDF
0.
,28
0.
,29
0.
,22
0.26
0.31
0.32
0.23
0.28
1,2,3,4,6,7,8,-HpCDF
0.
17
0.
,16
0.
.15
0.16
0.18
0.181
0.16
0.18
1 ,2,3,4,7,8,9-HpCDF
0.
,031
0.
035
0.
,033
0.033
0.034
0.038
0.035
0.036
Other HpCDF
0.
067
0.
066
0.
,059
0.064
0.074
0.072
0.062
0.069
OCDF
0.
070
0.
64
0.
,059
0.26
0.077
0.70
0.062
0.28
Total PCDF
2.
7
3.
1
2.
3
2.7
3.0
3.5
2.4
2.9
Total PCDD f PCDF
3.
9
4.
6
3.
3
3.9
4.3
5.0
3.5
4.3
CO- correclion
1 .
1
1 .
1
1 .
06
factor'3
a Denotes detection limits of undetected compounds which are considered zeroes in calculating averages.
b CO,, correction factor x measured value = value corrected (normalized) to 12I CO^.
-------
TABLE 2-5. UNCONTROLLED FLUE GAS PCDO/PCDF CONGENER DISTRIBUTION
Molecular
Mole
fraction3
Isomer
weight
Run 1
Run 2
Run 3
Averagf
Dioxins
2,3,7,8-TCDD
320
0.004
0.007
0.006
0.005
Other TCDO
320
0*180
0.183
0.180
0.181
1,2,3,7,8-PeCDD
¦ 354
0.020
0.018
0.017
0.018
Other PeCDD
354
0.177
0.182
0.209
0.189
1,2,3,4,7,8-HxCDD
388
0.019
0.020
0.018
0.019
1,2,3,6,7,3-HxCDD
388
0.037
0.040
0.035
0.038
1,2,3,7,8,9-HxCDD
388
0.031
0.035
0.031
0.032
Other HxCDO
388
0.180
0.200
0.191
0.191
1,2,3,4,6,7,8-HpCDD
422
0.103
0.091
0.096
0.097
Other HpCDD
422
0.090
0.085
0.089
0.088
OCDD
456
0.160
0.139
0.127
0.142
Furans
2,3,7,8-TCDF
304
0.059
0.065
0.062
0.062
Other TCDF
304
0.404
0.389
0.367
0.387
1,2,3,4,8-PeCDF
338
0.005
0.007
0.006
0.006
1,2,3,7,8-PeCDF
338
0.026
0.027
0.023
0.025
2,3,4,7,8-PeCOF
338
0.025
0.028
0.025
0.026
Other PeCOF
338
0.239
0.248
0.242
0.243
1,2,3,4,7,3-HxCCF
372
0.027
0.029
0.032
0.029
1,2,3,6,7,8-HxCDF
372
0.016
0.018
0.016
0.017
1,2,3,4,7,9-HxCDF
372
0.003
0.004
0.004
0.004
2,3,4,6,7,8-HxCDF
372
0.014
0.009
0.021
0.014
1,2,3,7,8,S-HxCDF
372
0.003
0.002
0.002
0.002
Other HxCDF
372
0.087¦
0.084
0.097 .
0.089
1,2,3,4,6,7,8-HpCDF
406
0.047
0.045
0.050
0.047
1,2,3,4,7,8,9-HpCDF
406
0.008
0.008
0.009
0.008
Other HpCOF
406
0.019
0.020
0.024
0.021
OCDF
440
0.016
0.017
0.021
0.018
a Mole fraction of each homo log is based on the tetra- thru octa- homo-
logs. PCDD mole fractions are based' on total PCDD, and PCDF mole frac-
tions are based on total PCDF.
2-7
-------
TABLE 2-6. CONTROLLED FLUE GAS PCOD/PCDF CONGENER DISTRIBUTION
Molecular
Run 1
Mole
fracti on®
Isomer
weight
Run 2
Run 3
Averag
Dioxins
2,3,7,8-TCDD
320
0*013
0.026
0.013b
0.013
Other TCDD
320
0*170
0.144
0.134
,0.149
1,2,3,7,8-PeCDD
354
0.018
0.016
0.026
0.020
Other PeCDD
354
0.158
0.234
0.130
0.174
1,2,3,4,7,8-HxCDD
388
0.015
0.015
0.023
0.018
1,2,3,6,7,8-HxCDD
388
0.034
0.033
0.042
0.036
1,2,3,7,8,9-HxCDD
388
0.027
0.029
0.036
0.030
Other HxCDD
388
0.166
0.145
0.131
0.148
1,2,3,4,6,7,8-HpCDD
422
0.105
0.101
0.120
, 0.108
Other HpCDD
422
0.093
0.088
0.106
0.095
OCOD
456
0.202
0.171
0.237
0.203
Furans-
2,3,7,8-TCDF
304
0.058
0.052
0.069
0.059
Other TCDF
304
0.369
0.315
0.367
0.350
1,2,3,4,8-PeCDF
338
0.006
0.005
0.005
0.005
1,2,3,7,8-PeCDF
338
0.023
0.021
0.019
0.021
2,3,4,7,8-PeCDF
338
0.023
0.020
0.024
0.022
Other PeCDF
338
0.254
0.208
0.232
0.231
1,2,3,4,7,8-Hx.CDF
372
0.031
0.027
0.035
0.031
1,2,3,6,7,8-HxCDF
372
0.016.
0.015
0.017,
0.016
1,2,3,4,7,9-HxCDF
372
0.005°
0.002
0.014°
0.001
2,3,4,6,7,8-HxCOF
372
0.018
0.011
0.017.
0.016
1,2,3,7,8,9-HxCDF
372
0.001
¦ 0.002
0.004°
0.001
Other HxCDF
372
0.093
0.086
0.088
0.089
l,2,3,4,6j7,8-HpCDF
¦ 406
0.052
0.045
0.056
0.051
1,2,3,4,7,8,9-HpCDF
406
0.01
0.009
0.012
0.010
Other HpCDF '
406
0.021
0.018
0.021
0.020
OCDF
440
0.020
0.163
0.020
0.068
Mole fraction of each homolog is based on the tetra- thru octa- homo-
logs. PCDD mole fractions are based on total PCDD, and PCDF mole frac
tions are based on total PCDF.
H
Compound not detected. Value is based upon detection limit and is
considered zero in calculating averages.
2-8
-------
Controlled Dioxins
0.5
0.4 -
0.3 -
0.2 -
0.1
4
5
6
7
8
Degree of Chlorination
~ Run 1 + Run 2 O Run 3
Figure 2-1. Mole Fraction Plot of PCDD (Controlled).
i
-------
Uncontrolled Dioxins
0.5
c
o
53
u
o
u.
o
o
2
0.4 -
0.3 -
0.2 -
0.1 -
Degree of Chlorinatlon
Run 1 + Run 2 O Run 3
Figure 2-2. Mole Fraction Plot of PCDD (Uncontrolled).
-------
Controlled Furans
0.5
0.4 -
0.3 -
0.2 -
0.1
5
6
7
8
Degree of Chlorinatlon
~ Run 1 + Run 2 O Run 3
Figure 2-3. Mole Fraction Plot of PCDF (Controlled).
-------
Uncontrolled Furans
0.5
0.4 -
0.3 -
0.2 -
0.1 -
4
5
6
7
8
Degree of Chlorinotion
~ Run 1 + Run 2 ^ Run 3
Figure 2-4. Mole Fraction Plot of PCDF (Uncontrolled).
i
-------
TABLE 2-7. 2,3,7,8-TCDD TOXIC EQUIVALENCIES OF UNCONTROLLED EMISSIONS
Toxic3 ng/dscm, corrected to 12% CC2
Isomer
equ i v.
Run 1
Run 2
Run 3
Average
Dioxins
2,3,7,8-TCDD
1
0.80
1.6
1.6
1.3
Other TCDD
0.01
0.36
0.43 ¦
0.51
0.43
1,2,3,7,3-PeCOD
0.5
2.1
2.3
2,65
2.4
Other PeCDD
0.005
0.19
0.24
0.33
0.25
1,2,3,4,7,8-HxCDD
0.04
• 0.18
0.23
0.26
0.22
1,2,3,6,7,3-HxCDD
0.04
0.36
0.46
0.49
0.44
1,2,3,7,8,9-HxCDO
0.04
0.29
0.39
0.43 -
0.37
Other HxCDD
0.0004
0.017
0.023
0.027
0.02
1,2,3,4,6,7,8-HpCOD
0.001
0.027
0.028
0.036
0.030
Other HpCDD
0.00001
0
0
0
0
OCOD
0
0
0
0
0
Total PCDD
4.4
5.7
6.3
5.5
Furans
2,3,7,8-TCDF
0.1
3.2
3.3
3.4
3.3
Other TCDF
0.001
0.22
0.20
0.20
0.21
1,2,3,4,8-PeCDF
0.001
0.0030
0.004
0.004
0
1,2,3,7,8-PeCDF
0.1
1.6
1.5
1.4
1.5
2,3,4,7,8-PeCDF
0,1
1.5
1.6
1.6
1.6
Other PeCDF
0.001
0.14
0.14
0.15
0.14
1,2,3,4,7,8-HxCDF
0.01
0.18
0.18
0.21
0.19
1,2,3,6,7,8-HxCDF
0.01
0.11
0.11
0.11
0.11
1,2,3,4,7,9-HxCDF
0.0001
0
0
0
0
2,3,4,6,7,8-HxCDF
0.01
0.091
0.055
0.14
0.10
1,2,3,7,8,9-HxCDF
. 0.01
0.023
0.011
0.013
0.02
Other HxCDF
0.0001
0.006
0.005
0.007
0.01
1,2,3,4,6,7,8-HpCDF
0.001
0.034
0.030
0.037
0.03
1,2,3,4,7,8,9-HpCDF
0.001
0.005
0.005
0.007
0.01
Other HpCDF
0.00001
0
0
0
0
OCDF
0
0
0
0
0
Total PCDF
7.1
7.1
7.3
7.2
Total PCDD + PCDF
11.4
12.9
13.6
12.6
a Toxic equivalent method used.
2-13
-------
TABLE 2-8. 2,3,7,8-TCDD TOXIC EQUIVALENCIES OF CONTROLLED EMISSIONS
yox1-ca ng/dscm, corrected to 12% C02
Isomer
equiv.
Run 1
Run 2
Run 3
Average
Dioxins
t.
2,3,7,8-TCDD
1
0.014
0.034
0.011°
0.016
Other TCDD
0.01
0.0019
0.0019
0.0011
0.0016
1,2,3,7,8-PeCDD
0.5
0.011
0.012
0.012
0.012
Other PeCDD
0.005
0.0010
0.0017
0.0006
0.0011
1,2,3,4,7,8-HxCDO
0.04
0.0008
0.0009
0.0009
0.0009
1,2,3,6,7,8-HxCDD
0.04
0.0019
0.0021
0.0017
0.0019
1,2,3,7,8,9-HxCDD
0.04
0.0015
0.0018
0.0015
0.0016
Other HxCDD
0.0004
0.0001
0.0001
0.0001
0.0001
1,2,3,4,6,7,8-HpCDD
0.001
0.0002
• 0.0002
0.0001
0.0002
Other HpCDD
0.00001
0
0
0
0
OCDO
0
0
0
0
0
Total PCOO
0.033
0.054
0.029
0.039
Furans
2,3,7,8-TCDF
0.1
0.015
0.015
0.015
0.015
Other TCDF
0.001
0.0010
0.0009
0.0008
0.0009
1,2,3,4,8-PeCQF
0.001
0
0
0
0
1,2,3,7,8-PeCDF
0.1
0.0069
0.0071
0.0046
0.0062
2,3,4,7,8-PeCDF
0.1
0.0070
0.0065
0.0058
0.0064
Other PeCDF
0.001
0.0008
0.0007
0.0006
0.0007
1,2,3,4,7,8-HxCDF
0.01
O.OOIO
0.0010
0.0009
0.0010
1,2,3,6,7,8-HxCDF
0.01
0.0005,
0.0006
0.0005.
0.0005
1,2,3,4,7,9-HxCDF
0.0001
0.0000°
0
0.0000°
0
2,3,4,6,7,8-HxCDF
0.01
0.0006
0.0004
0.0004,
0.0005
1,2,3,7,8,9-HxCDF
0.01
0'
0.0001
0.0001°
0.0000
Other HxCDF
0.0001
0
0
0
0
1,2,3,4,6,7,3-HpCCF
0.001
0.0002
0.0002
0.0002
0.0002
1,2,3,4,7,8,9-HpCDr
0.001
0
0
. 0
0
Other HpCDF
0.00001
0
0
0
0
OCDF
0
0
0
0
0
Total PCOF
0.034
0.033
0.029
0.032
Total PCDD + PCDF
0.066
0.087
0.058,
. 0.071
a See reference 3.
k Denotes detection limits of undetected compounds, which are considered
zeros in calculating averages.
2-14
-------
TABLE 2-9. RATIO OF UNCONTROLLED PCDD/PCDF TO PARTICULATE EMISSIONS
nq ana1,yte/q particulate
Isomer
Run 1
Run 2
Run 3
Average
D1ox1ns
2,3,7,8-TCOD
0.11
0.24
0.20
0.18
Other TCDD
4.8
6.6
6.4
5.9
1,2,3,7,8-PeCDD
0.58
0.71
0.66
0.65
Other PeCDO
5.2
' 7.2
8.2
6.9
1,2,3,4,7,8-HxCDD
0.60
0.88
0.79
0.76
1,2,3,6,7,8-HxCDO
1.2
1.8
1.5
1.5
1,2,3,7,8,9-HxCOD
0.99
1.5
1.3
1.3
Other HxCDD
5.8
8.7
8.2
7.6
1,2,3,4,6,7,8-HpCDD
3.6
4.3
4.5
4.1
Other HpCDD
3.2
4.0
4.2
3.8
QCDD
6.1
7.1
6.4
6.5
Total PCDD
32
43
42
39
Furans
2,3,7,8-TCCF
4.3
5.1
4.2
4.52
Other TCDF
29
30
25
28
1,2,3,4,8-PeCQF
0.44
0.61
0.46
0.50
1,2,3,7,8-PeCDF
2.1
2.3
1.7
2.04
2,3,4,7,8-PeCDF
2.0
2.4
1.9'
2.13
Other PeCDF
19
21
18
19.76
1,2,3,4,7,8-HxCDF
2.4
2.8
2.6
2.62
1,2,3,6,7,8-HxCDF
1.4
1.7
1.4
1.51
1,2,3,4,7,9-HxCDF
0.30
0.37
0.33
0.33
2,3,4,6,7,8-HxCDF
1.2
0.83
1.7
1.27
1,2,3,7,8,9-HxCOF
0.30
0.17
0.15
0.21
Other HxCDF
7.8
8.0
8.1
7.98
1,2,3,4,6,7,8-HpCOF
4.6
4.6
4.5
4.50
1,2,3,4,7 ,8,9-HpCOF
0.74
0.78
0.85
0.79
Other' HpCDF
1.9
2.1
2.2
2.06
OCDF
1.7
2.0
2.0
1.90
Total PCDF
80
86
76
81
Total PCDD + PCDF
112
129
' 118
120
Particulate loading
7.414
6.538 ¦
8'. 099
(g/dscm)
2-15
-------
TABLE 2-10. RATIO OF CONTROLLED PCDD/PCDF TO PARTICULATE EMISSIONS
Isomer
nq analyte/g particulate
Run 1
Run 2
Run 3
Average
Dioxins
2,3,7,8-TCOD
0*65
1,0
0.25a
0.56
Other TCDD
8.8
5.8
2.5
5.7
1,2,3,7,8-PeCDD
1.0
0.71
0.55
0.76
Other PeCDD
9.0
10
2.7
7.4
1,2,3,4,7,8-HxCDD
0.96
0.73
0.53
0.74
1,2,3,6,7,8-HxCDD
2.1
1.6
0.97
1.6
1,2,3,7,8,9-HxCDD
1.7
1.4
0.82
1.3
Other HxCDD
10
7.2
3.0
6.8
1,2,3,4,6,7,8-HpCDD
7.1
5.4
3.0
5.2
Other HpCDD
6.3
4,7
2,6
4.6
OCDD
15
9.8
6.4
10
Total PCDD
63
49
23
45
Furans
2,3,7,8-TCDF
7.0
4.7
3.4
5.0
Other TCDF
45
29
18
31
1,2,3,4,8-PeCDF
0.79
0.55
0.27
0.53
1,2,3,7,8-PeCDF
3.1
2.2
1.0
2.1
2,3,4,7,8-PeCDF
3.2
2.0
1.3
2.2
Other PeCDF
34
21
13
23
1,2,3,4,7,8-HxCDF
4.7
3.0
2.1
3.3
1,2,3,6,7,8-HxCDF
2.4
1.7
1.0 a
1.7
1,2,3,4,7,9-HxCDF
0.72a
0.21
0.83
0.07
2,3,4,6,7,8-HxCDF
2.7
1.3 ¦
1.0
1.7
1,2,3,7,8,9-HxCDF
0.19
0.18
0.25a
0.12
Other HxCDF ¦
14
9.6
5.2
9.6
l,2,3,4,6j7,8-HpCDF
8.4
5.6
3.6
5.9
1,2,3,4,7,8,9-HpCDF
1.5
1.2
0.78
1.2
Other HpCDF
3.4
2.2
1.4
2.3
OCOF
3.5
22
1.4
8.8
Total PCDF
135
106
54
98
Total PCDO + PCDF
198
155
78
143 .
Particulate loading
0.0219
0.0326
0.0445
(g/dscm)
a Based on detection limit of compound; considered as zero in calcu-
lating averages.
2-16
-------
2.3 PARTICULATE EMISSIONS
Particulate mass loading was 'determined by gravimetric analysis of the
filter, cyclone, and front half acetone rinses of the metals train. After
reaching constant weight, these fractions were digested for metals analysis.
Particulate emissions are summarized in Table 2-11. Particulate results were
blank corrected as specified in EPA Method 5. The PM control efficiency for
the SD/FF averaged about 99.5% for the three runs.
Field data are in Appendix D and the laboratory analyses are In Appen-
dix H. Further calibration data are in Section 6.0 and Appendix J.
2.4 METALS EMISSIONS
Table 2-12 summarizes the emission data for selected hazardous metals
(As, Cd, Cr, Pb, and Hg). Field data are presented in Appendix D, and rele-
vant laboratory data are presented in Appendix H. No blank corrections were
used, except for mercury. As and Cr were below detection limits in all
blanks. Cd and Pb were present at significant concentrations only in the
posttest blanks (about 10% of the controlled emission samples). Hg was
present in the front half blanks (probably the filter) at about 8 ug. Pb
emission rates dominate with about 30 mg/dscm (uncontrolled) and 0.1 mg/dscm
(controlled). The control efficiency varies from 98% for cadmium to 99.8% for
chromium, which is in general agreement with the relative volatilities of the
metals.
As shown in Table 2-13, the ratio of the selected metals to total partic-
ulate mass increased or remained constant across the control device. The
ratios increased by a factor of 2 for Cd and Hg, remained the same for As and
Pb, and decreased for Cr.
2.5 METALS CONTENT OF PROCESS SAMPLES
Table 2-14 shows the metals content of the process samples collected.
The selected metals are absent in the lime slurry with the exception of 4 pg/g
of As. Cd and Hg were not detected in the bottom ash, and the other metals
were present at levels of a few hundred micrograms per gram. Pb accounted for
about half of the total. No bottom ash sample was collected for run 3.
At the cyclone ash hopper (immediately upstream of the uncontrolled emis-
sions test location) and In the baghcuse ash hopper, Pb was the predominant
metal, as it was in the uncontrolled air emissions measurements. As, Cd, and
Pb are more concentrated in both cyclone and baghouse ash than in the uncon-
trolled air emissions samples. Cr concentration values in the uncontrolled
emissions fall between the Cr concentration values found in the cyclone and
baghouse ash samples.
2-17
-------
TABLE 2-11. SUMMARY OF PARTICULATE EMISSIONS FOR THE MERC FACILITY
Sampling location
Run no.
Date
Uncontrolled (spray dryer inlet)
1 2 3
12-09-87 12-10-87 12-12-87
Average
Controlled (fabric filter outlet)
1 2 3
12-09-87 12-10-87 12-12-87
Average
Sampling parameters
Flue gas characteristics
Gas volume sampled (dscf) 73.2 '112 88.8
Flue gas flow rate (dscfm) 41,500 42,100 42,500
Flue gas temperature (°F) 374 367 398
Moisture (percent by volume) 15.1 15.2 16.8
Sampling rate (percent isokinetic) 99.2 100 102
CO^ (percent by volume, dry) 10.9 11.0 11.3
O2 (percent by volume, dry) 8.6 8.5 8.2
42,000
380
15.7
11.1
8.4
61.0
39,800
270
16.8
106°
NDC
NDC
96.0
41,900
271
16.3
106
10.9
8.6
81.7
44,400
273
14.6
105
11.3
8.2
42,000
271
15.9
11 .0C
8.5C
ro Particulate results
00
Front half catch
(Probe, cyclone, filter)
mg-mass 14,000 19,000 19,000
gra i ns/dscf 2.94 2.61 3.33 2.96
grains/dscf (corrected to \2% CO^) 3.23 2.85 3.53 3.20
mg/dscm 6,730 5,990 7,630 6,780
mg/dscm (corrected to \2% 7,410 6,540 8,100 7,350
lb/h 1,040 942 1,210 1,070
kg/h 474 428 550 484
Removal efficiency (|)
34.3
0.009
0.009c
19.9
21 ,9C
2.96
1 .34
99.7
80.4
0.013
0.014
29.6
32.6
4.63
2.10
99.5
96.9
0.018
0.019
41 .9
44.5
6.94
3.15
99.4
0.013
0.014
30.5
33.0
4.84
2.20
99.5
Standard conditions are 68°F (20"C) and 1 atm (i.01325 x 10 Pa).
Results are adjusted for blanks.
Nol delermined from actual diluent data (method 3 sampling train leaked); inlet results used here for run 1.
I
-------
TABLE 2-12.
SPECIFIC METALS MASS EMISSION RATES FOR MERC (NORMALIZED TO \2% C02>
Meta 1
Run 1
Run 2
Run 3
Average
ug/dscm
g/h
ug/dscm
g/h
ug/dscm
g/h
ug/dscm
g/h
-
Uncontrolled
Arsen i c
462
33
513
36
511
37
495
35
Cadini um
990
70
1 ,030
73
1 ,230
89
1 ,080
77
Chromium
2,300
162
2,600
185
3,110
225
2,670
191
Lead
25,700
1 ,810
26,700
1 ,900
27,400
1 ,980
26,600
1,
,900
Mercury
461
34
315
22
341
24
379
27
Control led
Arsen i c
7.5 3b
0„509b
6.79b
0.483b
4.78b
0.361b
6.37
0.451
Cadmi um
II. 7b
0.79lb
10.2b
0.726b
14.8b
1.12b
12.2
0.879
Chrom i um
>
5.75b
0.389b
5.73b
0.408b
6.47b
0.488b
5.98
0.428
Lead
142
9.60
151
10.7
172
13.0
155
11.1
»Mercury
< 2.03a
< 0.137a
6.00b
0.427b
7.77b
0.586b
6.89
0.506
Removal Efficiency
Run 1
Run 2
Run 3
Average
Arsen i c
98.5
98.7
99.1
98.7
Cadini um
98.8
99.0
98.8
98.9
Chromi um
99.7
99.8
99.8
99.8
Lead
99.4
99.4
99.4
99.4
Mercury
< 99.6a
98.1
97.7
98.2
a
b
Some tractions below detection limit.
All fractions below detection limit.
Reported value is maximum possible.
Not included in averages.
-------
TABLE 2-13. RATIO OF METALS TO PARTICULATE MASS FOR MERC (ug/g)a
Metal
Run 1
Run 2
Run 3
Average
Uncontrolled
As
62.4
78.5
63.1
68.0
Cd
134
157
151
147
Cr
,310
398
384
364
Pb
3,460
4,080
3380
3,640
Hg
65
48
42
52
Control led
As
340^
2l0f|
lioj?
220
Cd
540°
320°
330°
400
Cr
26013
180
140
190
Pb
6510
4680.
3860,
5020
Hg
< 390c
190b
170
180
a Ratios are calculated with total train results by AA for the metals and
front half train results for the particulates. The ratio (ug/g) is calcu-
lated by dividing the concentration (yg) by the particulate loading (g).
k Some fractions below detection limit. Reported value is maximum possible.
c All fractions below detection limit. Not included in averages.
2-20
-------
TABLE 2-14. METALS CONTENT OF PROCESS SAMPLES (vg/g)
Run 1
Run 2
Run 3
Average
Cyclone ash
As ¦
32.2
32.7
38.9
34.6
Cd
30.7
35.2
38.1
34.7
Cr
383
438
424
415
Pb
2,100
2,090
2,070
2,087
Hg
15.3
5.00
13.4
11.2
Baghouse ash
As
47.6
47.2
53.6
49.5
Cd
129
80.1
77.4
95.5
Cr
152
159
167
159
Pb
2,770
1,290
2,130
2,063
Hg '
30.2
34.0
29.4
47.9
Bottom ash
As
7.24
14.0
b
10.6
Cd
< 1.50
2.73
b
2.73
Cr
¦ 169
312
- b
241
Pb
417
585
b
501
Hg
0.0640
1.22
b
0.64
Lime slurry
As •'
4.25
2.20
4.19
3.55
Cd
< 0.2264
< 0.231a
< 0.199a
< 0.219
Cr
< 0.943a
< 0.9603
< 0.8304
< 0.911
Pb
< 5.78a
< 5.88a
< 5.09a
< 5.58
h9
< 0.224a
< 0.216a
0.419
< 0*286
a Metal concentration below the indicated detection limit. Not included
in averages.
L f
Bottom ash was not collected during run 3.
2-21
-------
2.6 OTHER PROCESS SAMPLE ANALYSES
Table 2-15 shows the results of the analyses of general characteristics
measured in the ash and lime slurry samples. The baghouse ash resistivity was
also measured (Figure 2-5). Resistivity can be used to predict the perfor-
mance of an electrostatic precipitator. Resistivity was measured by Southern
Research Institute, Birmingham, Alabama, and the other analyses were performed
by Galbraith Laboratories, Inc., Knoxville, Tennessee. The analytical labora-
tory reports are in Appendix I.
2.7 ACID SASES
¦i
Table 2-16 summarizes the results of the dry basis CEM measurements for
SO2, HC1, and N0X which are not corrected to 12% C02. Figures 2-6 through 2-8
show the time plots for S02, Figures 2-9 through 2-11 show the time plots for
HC1, and Figure 2-12 shows the plots for N0X. Detailed CEM data are presented
in Appendix F. The acid gas results are difficult to assess because the pro-
cess was on manual control instead of the designed automatic system. The
sudden change in lime slurry feed during run 2 1s reflected prominently in the
S02 and HC1 plots.
The SOz analyzer used at the spray dryer inlet was subjected to poisoning
of the electrochemical cell with -consequent loss of sensitivity during each
run. The same type of analyzer located at the baghouse outlet was not
affected. The drift of the S02 analyzer at the spray dryer inlet was not con-
stant. Instead, it occurred in sudden steps, usually during the port
changes. Therefore, only the initial calibration was used to calculate the
results, and data are reported only until the first change in sensitivity.
Heavy alkaline dust loading at the spray dryer outlet affected HC1 mea-
surements during run 3. If dust builds up on the filter elements, the filter
effectively becomes an HC1 scrubber, which biases the readings. Although the
Entropy test crew could not find any apparent lime coating after the test, the.
analyzer was showing the character!stic low readings and slow response time2
which are associated with the sample system contamination of alkaline
material.
Table 2-17 shows the molar ratio of actual to stoichiometric lime for
each test. Stoichiometric lime is defined in Appendix A (Sample Calculations)
as the quantity of lime needed to exactly neutralize the average HC1 and S02
present. Since the peak concentrations of HC1 and S02 are about twice the
averages, the moderately low lime flow rates used for run 1 were too low to
neutralize the acid gases during maxima.
2.8 OTHER GASES
Table 2-18 summarizes the results of CEM measurements for 02, C02, CO,
and THC (corrected to 12% C02). Figures 2-13 through 2-15 show the oxygen
concentration plots for each run. Figures 2-16 through 2-18 show the C02
plots, Figure 2-19 shows CO plots, and Figure 2-20 shows the THC plot.
Detailed CEM data are presented in Appendix F.
2-22
-------
TABLE 2-15. PROCESS SAMPLES-BULK CHARACTERISTICS
Run L
Run 2
Run 3
Average
Bottom ash
Carbon, %
1.30
1.11
None collected
1,20
Ashs %
72.2
75.5
None collected
75.3
Cyclone ash
Carbon, %
1.06
0.89
1.60
1.18
Ash, %
98.3
98.3
9S.8
97.5
Baghouse ash
Carbon, %
5.48
2.62
3.67
3.92
Ash, %
89.1
96.0
90.2
91.8
Lime slurry
CaO, %
11.8
9.58
12.3
11.2
Solids, %
20.6
18.3
22.1
20.4
Specific gravity
1.13
1.12
1.14
1.13
2-23
-------
t r
Project: 6506
Date: 1/26/88
%H20- 14.8
T r
T ^ 1 r , 1 p
O 9DEC87
~ 10DEC87
O12DEC87
A UNLABELED
2.8
t i j
2.6
i 1
2.4
! I
2.2
i t
2.0
L .L
1.8
i i
1.6
i t
1.4
84 ¦
112
144
182
227
283
352
441
183
233
291
359
441
541
666
826
Temperature
Figure 2-5. Resistivity Plots of MERC Baghouse Ash,
oc
°F
2-24
-------
TABLE 2-16, CEM DATA SUMMARY—ACID GASES
Average RSOa
(ppm> (I)
Run I
30? Dryer inlet 83.2 16,1
Baghouse outlet 28.4 43.3
Removal efficiency 66.0
HC1 Dryer inlet 478 NAd
Dryer out Iet 63 NA
Baghouse outlet 9 NA
Removal efficiency 98.2
NO Baghouse outlet 203 8,72
Run 2b
SO, Dryer inlet 76,9 11.3
^ Baghouse outlet 21.2 45.6
Removal efficiency 72.3
HCI Dryer inlet 566 NA
Dryer outlet 8 NA
Baghouse outlet 4 NA
Removal efficiency 99.3
N0x Baghouse outlet 206 8.25
Run 3
SO- Orysr inlet 115 20.3
Baghouse outlet 12,0 18.6
Removal efficiency 89.6
HCi Oryer inlet 540 NA
Dryer out Ietc 1 NA
Baghouse outlet 3 NA
Removal efficiency 99.4
NO^ Baghouse out Iet 210 9.76
Average of 5 runs
SO, Dryer inlet 91.8
Baghouse outlet 20.5
Removal efficiency 76.0
HCI Dryer inlet 528
Dryer outlet 24
Baghouse outlet 5
Removal efficiency 99
NQ^ Baghouse outlet 206
3
b
RSD (relative standard deviation) = C100 x standard deviation) f mean.
Spray dryer lime slurry flow rate was increased t>y "002 at aBouf 13:45
during run 2.
c HCI results at the dryer outlet are questionable for run 3,
W'
NA - Data is. not availahle for HCI, wdich is contained in a separate
report by Entropy Environmentalists Inc. (Ref, 2)
2-25
-------
Spray Dryer
Inlet
200
150 -
100 -»
16
1«
17
18
19
"dm* (24-hr)
Baghouse
Outlet
120 -i
100 -
40 -
20-
13.5
1?
17.5
18
•15J
IS
15
Thr» (24—ht)
Figure 2-6. Run 1 S02 Concentrations,
2-26
-------
200
150
100 ¦
S3
Spray Dryer
Inlet
—I 1 ! 1 1 1 !
12 !3 .14 15 16 17 18 ' 19
Tim# (24-hr)
120
1
' I
100
80
60 -
40 -t
20
1
Baghouse
Outlet
_| , , , , ! , !
12 13 14 IS 16 17 18 19
Tims (24-hr)
Figure 2-7, Run 2 S02 Concentrations.
2-27
-------
250
Spray Dryer
Inlet
a
a.
21X3 -
150
100
50
0-«-
11
-1
19
12
13
14
15
I
16
Time (24—hr)
17
ia
120
Baghouse
Outlet
E
a
a
u
c
s
O
100
80
60
40
20
A
A
1 1 1 i 1—
11 12 13 14 15 16
Tim# (24—hr)
"*V^
17
18
19
Figure 2-8. Run 3 S02 Concentrations.
2-28
-------
320
300
280
260
240
220
200
180
160
140
120
100
80
60
40
20
0
I i
! Midpoint
16:30
Inlet + 10
Outlet
17:30
CLOCK TIME
Figure 2-9. Run 1 HC1 Concentrations.
-------
CO
o
XJ
E
CL
Q.
<
OH
\~
2:
UJ
o
z
o
o
o
X
70
60 -
50 -
40 -
30 -
20
10 -
0
13:00
14:00
Inlet -r 1 0
Midpoint
Outlet
T
15:00 16:00
CLOCK TIME
I
17:00
18:00
Figure 2-10. Run 2 HC1 Concentrations.
1
-------
X)
E
O-
Cl.
<
cm
i—
z
LU
O
z
o
o
o
X
Inlet *- 20
ju Outlet
r wV*v,^vv>v
Mjdj^oint
11:16 12:16 13:16 14:16 15:16
CLOCK TIME
Figure 2—11. Run 3 HC1 Concentrations
h
1 1 1
16:16 17:16 18:16
-------
Run No. 1
Run No. 2
Run No. 3
350 ¦
300 -
150 ¦
200 ¦
150 -
100 -
50 -
0
12
350 -i
300 ¦
250 •
200 -
150
100
50
0
13
14
IS
1?
1 1 r 1 1 r~
12 13 14 15 16 17
Hm* (24-hr)
13
IB
—I.
19
-1
19
Figure 2-12. NO Concentrations at the Baghouse Outlet.
2-32
-------
TA8i_E 2-17. MOLAR RATIO OF ACTUAL LIME TO STOICHIOMETRIC' LIME
FOR HC1 AND S02
Run
Actual lime
Actual lime
Stoichiometric lime
for HC1
Stoichiometric lime
for HC1 and.SO2
1
2.27
1.68
2
3.49
2.74
3
5,52
3.87
Average
3.75
2.76
2-33
-------
TABLE 2-18. CEM DATA SUMMARY—OTHER GASES
RSDa
Average (%)
Run 1b
0,, * Dryer in let 7.9 14,8
1 Dryer outlet 7.4 14.9
Baghouse outlet 8.0 14.3
CO-, t Dryer inlet 11.6 8.45
• Dryer outlet 11,5 6.78
Baghouse outlet 11.6 8.71
CO, ppm Dryer inlet 62.8 36.1
Run 2b
0„ % Dryer inlet 8.5 10.6
1 Dryer outlet 7.9 13.7
Baghouse outlet 8,4 !2,4
C0_, % Dryer inlet 11.2 7.29
Oryer outlet 11,7 2,82
Baghouse outlet 11.1 8.47
CO, ppm Dryer inlet 68.5 17.3
Run 3
Q_, % Dryer inlet 8.4 13.3
Dryer outlet 7.8 15.5
Baghouse outlet 8.6 15.4
CO,, % Dryer inlet 11.2 12.0
Dryer outlet 11.7 3.42
Bagtiouse outlet 11.3 10.8
CO, ppm Dryer inlet 89.9 55.9
THCa, ppm Dryer inlet 1.14 86.0
a RSD (relative standard deviation) = (100 x standard deviation) f mean,
b The total hydrocarbon analy2er was not functioning properly for runs 1
and 2 and had span drift during run 3.
2-34
-------
Spray Dryer
Inlet
10
A
15,5
HL5 17 17i 13 115 IS
Spray Dryer
Outlet
Baghouse
Outlet
10 -
5 -
15.5
r*"
16
—j—
1tL5
17 17.5
TTrna (24-hr)
—r~
18.5
Figure 2-13. Run 1 Oxygen Concentrations.
2-35
-------
Spray Dryer
Inlet
15
10
12
l
13
—r~
14
1 1 1
IS 16 17 18 19
15 •
Spray Dryer
Outlet
10
12
15 -|
i r
13 14 15
18 IS
Baghouse
Outlet
i
13
j
H
-I 1-
15 16
fiir* (24- Hr)
i r
17
19
Figure 2-14. Run 2 Oxygen Concentrations.
2-36
-------
Spray Dryer
Inlet
15-1
10
N
11 12
-i 1 i 1 1 r~
13 H 15 16 17 18
1
19
Spray Dryer
Outlet
10
T"~
16
~"1
19
11 12 *3 14 • 15
17 18
Baghouse
Outlet
15
»2 13
i i
14 15 18
T,im (24-hr)
17 It 19
Figure 2-15. Run 3 Oxygen Concentrations.
2-37
-------
Spray Dryer
Inlet
IS 1
10'
15.5
i
16
I
115
_t—
17
-T—
175
18
_j—
165
Spray Dryer
Outlet
15
10
IS J
15
i
16.5
17
I
175
18
185
1
19
Baghouse
Outlet
155
15
165
17 175
Timo (24-hr)
Figure 2-16. Run 1 C02 Concentrations.
165
2-38
-------
Spray Dryer
Inlet
10
1 [—
12 13 14
I 1 1 1 1
15 16 17 18 IS
Spray Dryer
Outlet
15 -i
-r~
13
AiW\
T ' I 1 I I
•5 ' 18 1? 18 19
Baghouse
Outlet
I
13
14
i
15
Tim« (24-fcr)
Figure 2-17. Run 2 C02 Concentrations.
2-39
-------
Spray Dryer
Inlet
15
—r~
13
"T-
14
15
I
16
i
17
_r_
18
15 -]
Spray Dryer
Outlet
|fff \/s^WvV'\/V
10 ¦
5 ->
11 12
~1
13
I
14
15 18
-T-
17
-T-
13
Baghouse
Outlet
15 i
10 ¦
1 r :—! 1 —i-
11 12 13 14 15 15
Urn* {24— hr)
17 18
Figure' 2-18. Run 3 CO, Concentrations.
2-40
-------
Run No. 1
250
200
150
100 ¦
50
u
\ ^WjYV ^
15.5 15 16.5
~~1
17
—J—
1TJ5
18
"I—
15.5
—I
19
Run No. 2
Run No. 3
250
200
ISO
100 -
50 -i
vu
Turns (24—hr)
Figure 2-19. CO Concentrations at the Spray Dryer Outlet.
2-41
-------
Run #3
10 -
8 -
4-
2-
t
11 12 13 14 15 16 17 18 19
Time (24-hr)
Figure 2-20. THC Concentrations at the Spray Dryer Inlet.
2-42
-------
The oxygen and C02 results Indicate that no significant leakage, dilu-
tion, or reaction of. C02 occurred across the- control devices. The C02
analyzer at the spray dryer outlet shows fewer excursions than the other ana-
lyzers because of slower response time for this Instrument.
The CO measurements indicated that the unit was experiencing periodic
combustion disturbances as evidenced by a number of CO transients which
exceeded 200 ppm. It is expected that the CO transients were related to tem-
porary combustion upsets caused by nonuniform refuse feed conditions. This
could account for the moderately elevated emissions of PCDO/PCDF measured
upstream of the spray dryer.
2.9 CONCLUSIONS AND RECOMMENDATIONS
Based on the results obtained during this project, the following conclu-
sions are made:
» Removal efficiency across the spray dryer/fabric filter was about
99.5% for dioxins, furans, and particulates.
• The PCDF levels were about twice as large at the PCOD levels for
both controlled and uncontrolled emission values.
• No significant change occurred across the control devices for the
molar distributions of the tetra- through octa-COF and CDD.
Metals (As, Cd, Cr, Pb, and Hg) removal efficiencies varied from
98.2% for Hg to 99.9* for Cr.
Concentrations of the five selected metals measured in the ash sam-
ples were in general agreement with the concentrations found in the
stack samples.
• Removal efficiency for S02 varied from 66* In run 1 to 90« in run 3,
in direct relation to the amount of slaked lime fed to spray dry
adsorber.
• Control efficiency for HC1 varied from 9B% in run 1 to 99% in run 3.
• No air dilution or absorption of C02 occurred across the control
devices.
The following recommendations are suggested for further study of this
type of facility:
• Data should be obtained far, the performance of the automatic Time
control system.
• The effect of combustion conditions on the emission of PCDO/PCDF,.
CO, N0X, and other pollutants should be investigated.
2-43
-------
SECTION 3.0
PROCESS DESCRIPTION AND OPERATION DURING TEST PROGRAM
This section contains a description of the MERC Waste-to-Energy facility
located in Biddefcrd, Maine. This section also summarizes the operation of
the facility and the key operating parameters that were measured during the
test program.
3.1 FACILITY DESCRIPTION
The MERC facility consists of two identical combustion process lines with
emission control devices that exhaust to a common stack. The combustion pro-
cess line is Illustrated in Figure 3-1. Refuse-derived fuel (RDF) enters the
combustor and is fired with preheated combustion air. Auxiliary fuel (natural
gas or fuel oil) can be used during startup, shutdown, or for load stabiliza-
tion. The combustion gases pass through the superheater, economizer, and
combustion air preheater heat recovery sections. The combustion gases then
pass through a cyclone to remove large particulate, an alkaline spray dryer to
control acid gas emissions and lower flue gas temperature, and a fabric filter
to reduce particulate emissions. The flue gas then exhausts to the atmosphere
through a 244-ft high stack which is common to both units.
The MERC facility is rated at 23 mg/h (600 tons/day) of RDF. The facil-
ity is owned by KTI Holdings, Inc., and was designed and built by General
Electric Environmental Services Company. Each unit can generate 47,628 kg/h
(105,000 lb/h) of steam at a temperature of 760°F and pressure of 675 psig
(superheated). The steam from the boilers- is supplied to a steam turbine
which generates up to 22 MW of electricity. The electricity is sold to
Central Maine Power.
3.1.1. Preparation of Refuse-Derived Fuel
At the MERC facility, preparation of RDF follows the scheme shown in
Figure 3-2. Solid waste from local municipalities is received in packer
trucks and transfer trailers and unloaded on the tipping floor, which is
enclosed. The waste is visually inspected, and potentially explosive or haz-
ardous items are removed. Oversized waste is removed and sent to a shear
shredder. The sorted waste is reduced in size by a flail mill and is combined
with the end product from the shear shredder. Then a magnetic separator
removes ferrous metal for reclamation. A trommel screen separates
nonorocessible wastes, and the remaining refuse is shredded to a nominal top
size of 10 cm (4 in.) by the secondary shredder. At this point, the waste has
become RDF. MERC estimates that 23 Mg/h (607 tons/day) of solid waste is
processed to produce 19 Mg/h (500 tons/day) of RDF.
3-1
-------
Dilution Lime
Slurry
Water
Combustion Air
Preheater
Spray
Dryer
Fabric / T.
Filter
Wv IDFan
Cyclone
Economizer
RDF
Auxiliary
Fuel
Combustor/
Boiler
Grate
Sittings
Stack
Bottom
Ash
Ash
Discharge
Figure 3-1. Combustion Process Line at the MERC Facility.
-------
Wood Chips
Sewage
Sludge
Municipal
Refuse
(MSW)
MSW
RDF
To Feed
Conveyor
Ferrous Metal
(to be reclaimed)
Nonprocessibles
to Landfill
Oversized
Waste
Flail
Mill
Secondary
Shredder
Tipping
Floor
Trommel
Screen
Shear
Shredder
Sludge
Hopper
Magnetic
Separator
Feed Hopper
for Combustor
Figure 3-2. Preparation of Refuse-Derived Fuel at the MERC Facility.
-------
As the RDF enters the combustor feed hopper, wood chips or sewage sludge
may be added, if desired. To date, only wood chips have been used. Sewage
sludge can be received into a separate hopper which is enclosed by a hydraul-
ically operated steel cover. The sewage sludge has a design moisture content
between 12% and 21% and a design feed rate of 0.833 yd3/h. This amount of
sludge, as a percentage of the total fuel volume, has an insignificant effect
on the boiler's firing rate. The fuel, whether RDF or RDF mixed with wood
chips and/or sewage sludge, is metered from the hopper by dual feeders to the
stoker.
3.1.2 Combustion Air
Air from the tipping floor area and boiler penthouse is withdrawn by a
forced-draft (FD) fan to supply the air heater section of the heat recovery
system. The preheated combustion air is split to supply the natural gas
burners, overfire air ports, and undergrate air. The combustion air scheme is
shown in Figure 3-3. The slightly negative pressure in the tipping floor area
ensures a continuous movement of air through the processing building prevent-
ing excessive accumulation of odors from the solid waste.
3.1.3 Combustor and Boiler
The combustion system consists of a Detroit stoker RDF spreader stoker
and a Babcock and Wilcox controlled combustion zone boiler. The combustion
zone boiler is rated at 158,200 MJ/h (150 x 106 Btu/h) of steam.
The stoker is a traveling grate located at the bottom of the boiler. The
fuel from the feeders enters the front of the boiler. A single auxiliary
burner capable of using natural gas and No. 2 fuel oil is located on the right
sidewall above the primary combustion zone. It. can be used for startup,
shutdown, or to maintain stable combustion conditions. The sulfur content of
the natural gas and fuel oil is limited by the air permit to a maximum of
0.7%.
The boiler is balanced draft. One fan (forced-draft) is used to feed
combustion air, and the second fan (induced-draft) located ahead of the stack
is used to draw out the combustion gases. A control system based on 02 and CO
concentrations is used to optimize combustion efficiency. The target level of
excess oxygen is in the range of 7% to 8% on a dry basis.
In addition to the waterwal1s in the combustion zone, the heat recovery
system includes superheater, economizer, and combustion air heater sections.
At the exit to the air heater, the flue gas temperature is approximately 204°C
(400°F).
3.1.4 Cyclone, Spray Dryer, and Fabric Filter
The combustion gases from the air heater enter a multicyclone dust col-
lector which removes large particulate. Next, an alkaline spray dryer is used
to control acid gas emissions. The spray dryer is a reaction vessel where
lime slurry is sprayed into the flue gas that contains PM, acid gases, and
other pollutants in gaseous and aerosol form. The slurry water is evaporated
3-4
-------
Tipping
Floor
Combustion
Air
F.D. Fan
Boiler
Penthouse
Undergrate
Air
Natural
Gas Burner
Overfire Air
Air
Heater
Total Air
Flow Meter
Secondary Air
Flow Meter
Overtired Air
Flow Meter
Figure 3-3. Combustion Air Scheme at the MERC Facility.
-------
by the flue gas heat and the acid gases react with the lime. Particulate and
excess lime serve as nucleation points for adsorption and agglomeration of
volatile trace metals and semivolatile organics.
The rate of lime addition and the flue gas temperature at the exit to the
spray dryer can be controlled separately. Slaked lime, which is introduced as
a slurry, is diluted with water before it enters the reaction vessel at rates
appropriate to achieve the desired S02 concentration at the inlet to the
fabric filter. The rate of slurry addition is varied based on the continu-
ously monitored S02 concentration at the outlet of the fabric filter.
The facility is required by its operating permit to maintain an outlet
S02 concentration of < 30 ppm. At no time during the test program, however,
were the facility's S02 monitors providing accurate readings. The spray dryer
outlet temperature is directly controlled by the amount of -diluting water
added and is typically 138° to 149°C (280° to 300°F).
The fabric filter collects the particulate from the gas. stream. The
excess lime in the bag filter cake provides a second-stage reaction site for
further acid gas removal. The fabric filter, unit has six modules. In a con-
tinuous cycle, five modules fiIter flue gas while' one module is being
cleaned. The total time to complete a fabric filter cleaning cycle is about
18 minutes.
3.1.5 Ash Handling
The ash system removes ash from the stoker discharge quench pit, generat-
ing bank hopper, air heater hopper, mechanical dust collector hopper, spray
dryer hopper, and fabric filter module hoppers. All of the hopper discharges
are through rotary seal valves. This ensures a positive seal to prevent
boiler gases from entering the ash conveyors and air from entering the hoppers
and boilers.
The ash from the fabric filter modules discharges into six identical
drag/screw conveyors. Each set of these conveyors discharges into one of two
identical drag chain collecting conveyors. The spray dryer and mechanical
dust collector discharge ash directly onto these drag chain collecting con-
veyors. The generating hopper and air heater hopper discharge "ash. onto a
transverse drag conveyor which feeds to the drag chain collecting conveyors.
The combined fly ash from,each collecting conveyor is fed to one of two ident-
ical ash conditioning screw conveyors. The ash is conditioned by water added
at a controlled rate.
The bottom ash from each stoker discharges into one of two submerged drag
chain ash conveyors. The discharge of the ash conditioners is deposited into
the dewatering section of the bottom ash drag conveyor. At this point, the
fly and bottom ash streams combine. The combined ash streams are then dumped
into a specially designed trailer for removal'from the site.
Dust control within the processing building is achieved through two sepa-
rate control systems. One system serves the tipping/processing area, while
the other serves the conveyors in the boiler building and RDF reclaim area.
3-6
-------
Each system contains a baghouse, fan duct hoods, and dust collection ducts at
key conveyor and transfer processing points. Oust-laden air is drawn through
one of two pulsed jet baghouses which exhaust in the vicinity of the boiler
forced-draft fan intake. The baghouse air exhaust is thus incorporated into
the combustion air for the boilers. Oust captured by the baghouses is
returned to and becomes a part of the RDF fuel.
3.2 SUMMARY OF OPERATIONS BY TEST RUN
Three test runs were conducted on Unit A from December 9-12, 1987. Dur-
ing each test run ROF only was fired.
3.2.1 Operations During Run 1
Run 1 was originally scheduled for December 3, but power problems in the
afternoon delayed it until December 9. Both units were down overnight and the
facility was still experiencing operational problems on the morning of
December 9. The units were started up in the morning and were preheated on
natural gas. However, problems with the feeder conveyors delayed bringing the
boilers up to full load until 1400. At 1500, CEM data indicated that the
boilers were stabilized.
Test 1 began at 1530 and continued until approximately 1840 when the
Unit A forced-draft fan failed. Two of three traverses (160 min) had been
completed at the time of the shutdown. Since replacement of the fan motor
required overnight work, Test 1 was considered to be completed by EPA
personnel on site.
3.2.2 Operations During Run 2
Run 2 was conducted on December 10, 1987. The fan was repaired at
approximately 0100 that morning and both units were back on-line. However, at
1030 there was a feeder conveyor failure and a unit shutdown occurred. The
units were brought back on-line at 1200, and Test 2 began at 1245. Testing
continued and was completed at 1800. All three traverse ports were sampled
for a complete run (240 min).
The facility operators decided to increase the lime slurry feed rate at
1330. Minor excursions of S02 were'being experienced and the facility did not
want to exceed its permit limit of 30 ppm. Therefore, the lime slurry feed
rate was increased from approximately 3 gpm to values ranging from 7 to
8 gpm. The test average was 7.8 gpm. This increase substantially reduced the
SO2 concentration at the midpoint and outlet locations.
3.2.3 Operations During Run 3
Run 3 was conducted on December 12, 1987. Originally scheduled for
December 11, problems had continued with feeder conveyors throughout the day
on the 11th, so testing was postponed until the 12th. Test 3 began at 1115.
A brief test interruption occurred between 1138 and 1200 due to a feeder mal-
function. Testing continued until 1525, restarted at 1815, but was stopped at
1330 due to recurring feeder problems. Throughout run 3, the lime slurry rate
3-7
-------
was maintained between 7 and 8 gpm. Oue to the late hour and the fact that
the facility estimated that the delay time would be 4 to 8 h, the test was
considered complete at the end of two complete port traverses plus three of
eight points for the third (190 min).
3.3 SUMMARY OF KEY OPERATING PARAMETERS
This section summarizes the values of key operating parameters during the
test program. The purpose of evaluating the operating parameters was to
determine: (a) if the system was operating at normal conditions, and (b) if
the system was operating at similar conditions during each of the three test
runs. Only selected key parameters are discussed in this section.
The operating data were recorded by computer every 4 min. The complete
set of operating data showing each 4-min value is included in Appendix B.
Plots of the 4-min data versus time are presented in this section. The plots
have been reduced so that all three runs can be shown on one page. Full-sized
plots for each run are included 1n Appendix B if more detail is required by
the reader. The locations of temperature, pressure, and flow sensors are
indicated in Figure 3-4.
Average values for selected operating parameters over the actuaT testing
intervals are summarized in Table 2-1. On average, the combustor operating
conditions appear to be the same for all three runs. The only variation of
consequence is the higher airflow and economizer inlet flue gas temperature
during run 3. Although the combustion operating conditions appear similar,
there is no way to judge if the entire combustor system reached the same
degree of thermal equilibrium for each run.
The emission control system was operated differently during each run.
The average lime slurry feed rate was increased during each test, with run 2
being higher than run 1, and run 3 being higher than run 2. This increase in
slurry flow, combined with the higher spray dryer inlet temperature and
airflow during run 3, is consistent with the increase in pressure drop across
the spray dryer and fabric filter during each test.
3.3.1 Steam Load and Heat Release
In Figure 3-5, RDF heat release, superheater steam flow, superheater
steam pressure, and steam temperature at the superheater outlet are plotted
against time. The ROF heat release is calculated from the steam flow minus
the heat content supplied by any auxiliary fuel (natural gas or fuel oil).
During this test program, only RDF was fired, and sampling was discontinued
during periods when auxiliary natural gas firing was necessary. Thus, for
this test program, the RDF heat release is equivalent to the total heat
release.
For all three runs, these combustion parameters were similar and normal
during the time periods in which the manual sampling trains were operating.
The relative standard deviation of the steam load averaged 4% during the
sampling periods.
3-8
-------
Dilution Lime
Water Slurry
Combustion Air
Preheater
Fabric J / T-
Filter
Wv IDFan
Spray
Dryer
Cyclone
Economizer
RDF
Auxiliary
Fuel
Combustor/
Boiler
Grate
Sittings
Stack
Bottom
Ash
Ash
Discharge
1 - Superheater steam Nowrate, pressure, temperature and economizer inlet flue gas temperature
2 ¦ Economizer outlet Hue gas temperature and excess oxygen
3 - Air healer outlet flue gas temperature and pressure
4 - Spray dryer inlet Hue gas temperature and pressure
5 - Spray dryer outlet due gas temperature and pressure
6 - Fabric (ilter outlet temperature
7 - Dilution water teedrate
8 - Lime slurry (eedrate
Figure 3-4. Location of Important Temperature, Pressure, and Flow Sensors at the MERC Facility.
-------
Run 1
Run 2
Run 3
BOO
600
400
200
A.
Steam Temp.
SH. Pressure
Start
Test
Port
Change
RDF Heat
''"SK Fiow ''
V"'1
ti u<4
lH-i n»i
15:25 16:00
17:00
Time
(12/9)
Start stop/Start Stop/
Test// stop/Start Start
Start
Test Stop/Start
/
Port
Change
S'*r >T*1 r*-** j ? tf (, "h VVv
Stop/Start
18:00 12:45 14:00 15:00 16:00 17:00
Time
(12/10)
11:15 13:00 14:00 15:0016:0017:00 18:00 19:00
Time
(12/12)
h-H Indicate periods in which manual sampling
trains were not operating
KEY
~ RDF heat release (10° Btu/hr)
+ Superheater steam flow (1000 Ib/hr)
0 Superheater steam pressure (psig)
A Steam temperature at the superheater outlet (°F)
Figure 3-5. RDF Heat Release Steam Flow, Steam Temperature, and Steam Pressure
as a Function of Time During the MERC Test Program.
-------
3.3.2 Combustion Air
Overfire air distribution, undergrate-to~furnace differential pressure,
arid excess oxygen are plotted against time in Figure 3-6. The overflre air
distribution was calculated by dividing the overfire air mass flow rate by the
total mass flow rate.
The variation in excess oxygen was greater during run 3 than in runs 1
and 2. During run 3, the relative standard deviation was 22%, as compared to
16% and 12% for runs 1 and 2. However, the average concentrations were not
significantly different.
The overfire air (OF) distribution was lower and undergrate-furnace dif-
ferential pressure was higher during run 3. The average OF air distribution
was 60^ during runs 1 and 2, but decreased to 50% during run 3. The
undergrate-furnace differential pressure increased to 0.4 in H20 during run 3
from 0.3 in H2Q during run 2. It was 0.5 In run 1.
The overfire airflow pressures were measured in the combustor. The
pressures measured during the MERC test program are presented in Figure 3-7.
As the figure shows, once the combustor is optimized, the pressures do not
vary. Pressurized air from two airswept spouts is also used to spray the RDF
across the grate as it enters the combustor. The airswept pressure is varied
in a set range so that the RDF is sprayed evenly across the grate.
3.3.3 Temperature Profile
The inlet and outlet flue gas temperatures of the economizer, air heater,
spray dryer, and fabric filter are plotted against time in Figure 3-8. The
economizer inlet, economizer outlet, and air heater outlet temperatures were
10 to 20 degrees hotter during run 3. However, after the spray dryer, the
flue gas temperature during run 3 was the same as during runs 1 and 2. The
spray dryer outlet temperature was very consistent during all three runs.
3.3.4 Spray Dryer and Fabric Filter
The operation of the spray dryer and fabric filter was evaluated using
two plots. The first plot (Figure 3-9) included the spray dryer inlet and
outlet temperatures, the lime slurry and dilution water feed rates, and. the
fabric filter differential pressure.- The second plot (Figure 3-10) includes
the flue gas differential pressures across the cyclone, spray dryer and fabric
filter.
The differences in spray dryer operation during the runs are shown
clearly in Figure 3-9. Quring run 2, the lime slurry feed rate was increased
significantly because of the high S07 concentration monitored at the fabric
filter outlet by the test conductor (more than double the permit level of
30 ppm). The lime slurry feed rate was increased from 3 gpm to over 7 gpm and
remained at this level through run 3. A corresponding decrease in the dilu-
tion water feed rate was observed such that the total feed rate of lime slurry
and dilution water increased only slightly. The spray dryer outlet tempera-
ture remained constant throughout all three test runs.
3-11
-------
Run 1
Run 2
Run 3
20-
co
I
15
10
5 -
Star!
Test
Port
Change
Stop
Test
Left Side
v T-'-V
Right Side
rTi-n-r'"' >->
15:25 16:00
17:00
Time
(12/9)
Start
Test s<°P'S|arl
W'
V/i/^
Port
Change
i\TW
Slop/ Stop/ Slop/Slarl
st?rt Start Start
Stop/Start
1.0
0.8
0.6
0.4
0.2
18:00" " "12:45" " 14:00 -15.00 16:00 ' 17:00 " n:i5 13:00 " 15.00 " 17:00 19:00
Time „ Time
(12/10) KEY (12/12)
~ Excess Oxygen (left side, % by volume, wet)
_ + Excess Oxygen (right side, % by volume, wet)
Start
Test
Port
Change
Overfire
'«
"V
Undergate
Stop
Test.
"'"'.Vf'
s,arl Stop/Start Port Stop Start Stop/STarl Stop/Start End Test
"Test change Test \| W ft tf stop/slafl >¦
>f"l
•V'
A
Y
!\
A
' Vvs' ^
\j\r
run
u
15:25 16:00 17:00 1 "" 18:00' "" 12:45' "14:00 "15:00 16:00 - 17:00 "" 11:15 13:00" 15:00 • ' 17:6o ::' '19:00
Time Time Time
(12/9) (12/10) (12/12)
VC CV
[-—^ Indicate periods in which manual sampling
trains were not operating Q Overfire air distribution (fraction)
+ Undergrate-furnace differential pressure (in H20) period
Figure 3-6. Combustion Air Pressure as a Function of Time During the MERC Test Program.
I
-------
9
E
Combustion
Zone Boiler
F
Grate
Pressure (in H20)
A - 23"
B -24"
C ¦ 23"
D - 24"
E - 23" '
F - 24"
G -13"
H - 24"
1 - Airswept Spout - Range of 9" to 24"
J - Airswept Spoul - Range of 9" lo 24"
Figure 3-7. Qverfire Air Flow Pressure Measured During
the MERC Test Program.
3-13
-------
Run 1
Run 2
Run 3
800
600-
400
200
Start
Test
V"
Port
Change
Stop Start
Test Test Stop/Start Change Test Test I Start
T~17 w —'—. *snVr
Economizer Inlet
LOau tld O^a
t,a l1 u'
Economizer Outlet
I
Air Heater Outlet/Spray Dryer Inlet
\
V
;9tfuOHeSBfiSg8wO00O»4O«o
Spray Dryer Outlet/Fabric Filter Outlet
15:25 16:00 17:00 18:00
Time
(12/9)
|wr'W^y
I'nt^un^y
Port Stop Start S'OP
st
V;
l{
3 e ewttw'
,»w
/• Vs."."
ueriMM* i
End
Test
Slop I Stop/
Start Start
Wv*--\?
ta to
Stop/Start
pofl
12:45 14:00 15:00 16:00 17:00 11:15 13:00 14:0015:00 16:0017:00 18:0019:00
Time
(12/10)
Time
(12/12)
KEY
Indicate periods in which manual sampling
trains were not operating
~ Economizer inlet gas temperature (°F)
+ Economizer outlet/air heater inlet gas temperature (°F)
0 Air heater outlet gas temperature (°F)
A Spray dryer inlet gas temperature (°F)
x Spray dryer outlet/fabric filter inlet temperature (°F)
V Fabric filter outlet gas temperature (°F)
Figure 3-8. Flue Gas Temperature as a Function of Time During the MERC Test Program.
-------
Run 1
Run 2
Run 3
400
300
200
100-
Slafl
Test
Port
Change
"^,l" ,|]' HI, M I'll Mil IH.l,."'
SD Inlet
Stop start Port
Test Test Slop/Start Change
A/w\/v
SD Outlet
Dilution Water
Lime Slurry
V\K/v/\a/\
TT
Stop/Start
Stop/Start
Slop Start / / Stop/ End
Test TestII Start SlopfSlarl Stop/Start -|-eS|
•Mm
15:25 16.00
17:00
Time
(12/9)
18:00
i ttt[T 11 n 11 n n n (i
12.45 14:00 15:00 16:00 17:00
w
'/ jVv** ** */wvv
WwM
v\/V
AA\
Time
(12/10)
11:15 13.00 14.00 15:0016:0017:0016.00 19:00
Time
(12/12)
KEY
i. .1 , .. . . . . . ~ Spray dryer inlet gas temperature (°F)
r~"1 Indicate periods in which manual sampling . „ . .. . . ,.u
trains were not operating + Spray dryer outlet/fabric filter inlet gas temperature ("F)
0 Lime slurry feedrate (gpm x 10)
A Dilution water feedrate (gpm x 10)
Figure 3-9. Spray Dryer Operating Parameters as a Function of Time
During the MERC Test Program.
-------
10
Starl
Test
/
Run 1
Port
Change
Fabric Filter
•VVV"
, Spray Dryer
Dust Collector
V
Run 2
Stop Start Port
Test Test Stop/Start Change
\
¦A
V\A
^*\/W
"\
/""'•J
-Axn
V^A-a^
i/v\a
VW^M,"w"
VA. v'
Run 3
Stop/Start
/.Stop/Start
?2 st°p's-
A
Av
*41
Aa v"
End
Stop/Start jest
15:25 16:00
17:00
Time
(12/9)
18:00
12:45 14:00 15:00 16:00 17:00
Time
(12/10)
11:15 13:00 14:00 15:00'16:00' 17:00 18:00 19:00
Time
(12/12)
|- H Indicate periods in which manual sampling
trains were not operating
KEY
~ Dust collector differential pressure (in HaO)
+ Spray dryer differential pressure (in H20)
0 Fabric filter differential pressure (in H20)
Figure 3-10. Differential Pressures Across the Control Devices
During the MERC Test Program.
-------
The' differential pressures across all three control devices (cyclone,
spray dryer, and fabric filter) increased during runs 2 and 3, with run 3
beinq the highest. For run 2, the Increase in lime slurry way have caused the
increase, since the pressure drop across the cyclone did not change signifi-
cantly. However, for run 3, a combination of increased airflow rate and lime
slurry may have caused the increased pressure drop.
3-17
-------
SECTION 4.0
SAMPLING LOCATIONS
Sampling locations for unit A are shown in the process line schematic in
Figure 4-1, Each sampling location is discussed in the following sections.
4.1 SPRAY DRYER INLET
Parameters that were measured at the spray dryer inlet (location L in
Figure 4-1) included PCDO/PCDF, HC1, S0a» PM, Pb, Hg, Cd, Cr, As, THC, C0?,
CO, and 02» A side view and top view of the sampling location are shown in
Figure 4-2. The sampling location had four lC-cm (4-in.) inside diameter
flanged ports on the side of a horizontal rectangular duct with a cross-
sectional area of 1.544 (16.62 ft2). Three of these ports gave access to
the cross-sectional plane used for the full traverse MM5 and Hethod 3 sam-
pling. The fourth port, located 1.22 m (4 ft) downstream, was used for the
CEMS probes at a fixed sampling point near the center of the duct. All of the
ports had 23-cm (9-in.) nipples and were accessible from the floor at the top
of the spray dryer.
EPA Method 1 was used to select the number and location of the sampling
points for MM5 and Method 3 .sampling. Those ports were located approximately
one equivalent duct diameter 1.22 m (4 ft, 1 in.) downstream of an expansion
joint and 1.4 duct diameters upstream of the spray dryer inlet. Twenty-four
traverse points were required, and the point location diagram is presented in
Figure 4-2.
A cyclonic flow check was conducted prior to sampling. The average
degree of rotation was two degrees. This met the criterion in Method 1 which
specifies that the average degree of rotation must be no greater than
20 degrees.
The average volumetric gas flow rate and temperature in the duct during
the three test runs was 1,170 dscm/min (41,000 dscfm) and 191"C (376°F). The
average static pressure was -215 mm H2Q (-8.5 in. H2Q).
4.2 SPRAY DRYER OUTLET/BAGHQUSE INLET
Parameters that were measured at the spray dryer outlet (location 2 in
Figure 4-1) were HC1, carbon dioxide, and oxygen. The sampling location had
three 10-cm (4-in.) ports on top of a horizontal rectangular duct with a
cross-sectional area of 1.839 m* (19.79 ft2). Two of these ports were used
for the CEMS probes at fixed sampling points near the center of the duct. The
ports were accessible from the top of the duct. These ports were located
4-1
-------
t n^.
Boiler
Baghouse
Spray
Dryer
(ENy RDF Feed
V
Cyclone
vw
I.D. Fan Stack
F.D. Fan
H
Ash Discharge
~^Ash Handling System
OCombustion Gas Sampling Locations
v-1 Process Sampling Locations
Figure 4-1. Sampling and Monitoring Locations—Unit A.
-------
4' —*|
Gas Flow
TOP VIEW
CEMs Port
MM5 Ports
SIDE VIEW
Traverse
Point
8
7
6
5
4
3
2
1
Figure 4-2. Spray ..Dryer Inlet—Location of Sampling Ports
and Traverse Points.
Distance from
Inside Wall
at Port (Inches)
3^8
«3
15'
21-
28-
341
401
46 I
r~-
50"
1 2 3 4 5 6 7 8
«•••••••
• • • «
Port 1
Port 2
Port 3
Flow Toward
Observer
4-3
-------
approximately 1.4 duct diameters (1 duct diameter = 1.35 m = 4 ft, 5 in.)
downstream from a bend in the duct and 1.1 duct diameters upstream from the
first pair of baghouse module inlets.
4.3 BAGHOUSE OUTLET
Parameters that were measured at the baghouse outlet (location 3 in
Figure 4-1) included PCDO/PCDF, HC1, S02, NO , carbon dioxide, oxygen, partic-
ulate, lead, mercury, cadmium, chromium, and arsenic. Two side views of the
sampling location are shown in Figure 4-3. There were two sampling locations
on the duct. Each had three 10-cm (4-in.) inside diameter flanged ports
located on one side of the vertical rectangular duct. The three ports at each
location gave access to a cross-sectional plane of 1.839 m2 (19.79 ft2) in the
duct having an equivalent diameter of 1.35 m (4 ft, 5 in.).
Two of the ports at the upstream location were used for the CEMS probes
at fixed sampling points near the center of the duct. These ports were
accessible from a permanent platform located approximately 1.52 m (5 ft) below
the ports. The ports were located 1.8 duct diameters downstream of a
90-degree duct bend and 3.6 duct diameters upstream of the camperec bypass
duct.
The ports at the downstream location were used for the MM5 and M3 sam-
pling and were accessible from the top of the bypass duct. These ports were
located approximately 4.3 duct diameters downstream of a 90-degree duct bend
and 1.1 duct diameters upstream of the dampered bypass duct. Twenty-four
traverse points were required according to Method 1, and the point location
diagram is presented in Figure 4-3. A cyclonic flow check yielded an average
rotation of four degrees.
The average volumetric gas flow rate and temperature measured at the
downstream location during the three test runs was 1,170 dscm/min
(41,200 dscfm) and 133°C (271°F). The average static pressure was -512 mm H20
(-20.2 in. H20).
4.4 CYCLONE COLLECTOR ASH
Cyclone collector ash was sampled from a 5-cm (2-in.) gate-valved dis-
charge pipe (location A in Figure 4-1) connected to the bottom of. the larger
inclined pipe that transferred, the ash from the rotary seal valve on the
bottom of the hopper to the drag chain conveyor.
The original scope for mass ash collection sampling at this location was
not accomplished. Ash could not be mass sampled due to the expense and man-
power that would have been required to reroute the inclined pipe before and
after each test run.
4.5 BAGHOUSE ASH
Baghouse ash was sampled from a 5-cm (2-in.) gate-valved discharge pipe
connected to the bottom of the horizontal drag/screw conveyor approximately
3.05 m (10 ft) downstream from the rotary seal valve discharge on the bottom
of the hopper on unit A baghouse compartment 5.
4-4
-------
C5
v—
i
in
i
Gas Flow from
Fabric Filter
=0 CEMs Ports
rzO MM5 Ports
i By-Pass
K
To I.D. Fan
Damper
o o o
o o o
reverse
Point
8
7
5
4
3
2
1
Distance from
Inside Wall
at Port (Inches)
3^8
„ 3
15:
' 21 -
28;
34 ^
40 S
46 ^
~
Port 1
Flew Away
from Observer
Port 2 Port 3
Figure 4-3. Baghouse Outlet—Location of Sampling Points
and Traverse Points.
4-5
-------
This was the only location where uncontaminated baghouse ash could be
sampled, because the expense of installing pipes on the other five conveyors
and the extra time required to collect those samples were not warranted. The
samples taken from one compartment were representative of the total baghouse
discharge because of the gas flow distribution and the overlapping of compart-
ment cleaning cycles.
4.6 BOTTOM ASH
Furnace bottom ash was sampled from the drag chain conveyor leaving the
quench pit beneath the furnace bottom and ash chute (location C in Fig-
ure 4-1). Samples could not be collected before the quench without making
costly modifications to the bottom ash chute so as to keep ambient air from
entering the furnace at the sampling point. Also, samples taken from such a
point probably would not have been representative of the bottom ash.
The bottom ash samples taken after the quench pit could have been contam-
inated occasionally by overflow from the conveyor systems that were moving ash
from the baghouses and all collection points in units A and B. When screw
conveyors moving ash from the drag chain conveyors to the final ash discharge
were overloaded or obstructed, the ash from these conveyors overflowed into
the quench pit. Also,- the Unit A bottom ash discharge was downstream of the
Unit B discharge. Therefore, when Unit 8 was operating, the bottom ash sam-
ples represented ash from both units. However, both units are identical and
the RDF feed to both are from a common source.
4.7 LIME SLURRY
Lime slurry samples were taken from the main feed line at a tap
(location D in Figure 4-1) located within 6.10 m (20 ft) of the injection
point at the top of the spray dryer. This sampling location was ideal for
obtaining representative sample increments immediately upstream of the slurry
injection.
4.8 REFUSE DERIVED FUEL
During run 1, RDF samples were taken from a point along the process con-
veyor belt, A—16 (location E in Figure 4-1). During runs 2 and 3, RDF samples
were taken from the A-17 feed line immediately before injection into the
furnace.
4.9 SAMPLES NOT COLLECTED
Samples were not collected directly from the spray dryer ash discharge
hopper, because it had been assumed that very little ash, if any, accumulated
in the hopper. A gate-valved pipe attached to the hopper was opened and no
ash flow resulted. However, this could have been due to the negative pressure
at that point. Physical constraints made it impossible to collect samples
after the rotary seal valve.
Preheater, economizer, and grate siftings ashes were not part of the test
protocol and were not sampled.
4-6
-------
SECTION 5.0
SAMPLING AND ANALYSES PROCEDURES
The descriptions of the sampling and analysis methods provided in this
section are brief. Detailed descriptions are included in Appendix M. The
sampling and analysis methods used for each parameter are summarized in
Table 5-1.
5.1 DIGXIN/FURAN SAMPLING AND ANALYSIS
5.1.1 Sampling Equipment Preparation
Combustion gas samples for dioxin/furan (PCDD/PCDF) analysis were col-
lected using a Modified Method 5 (MM5) sampling train. Figure 5-1 shows
details of this train. Space restrictions at the spray dryer inlet location
additionally required the probe connection at the sample box to be connected
with a flexible heated 1.22-m (4-ft) Teflon line and a glass union. Remaining
assembly of the sampling train followed the procedure outlined in S-l, Stan-
dard Operating Prccedure for MM5 train, located in Appendix M-l. Figure 5-2
shows a data sheet for the field lab setup.
5.1.1.1 XAD-2 Cleanup and Trap Preparation-
Adsorbent resin .used for the testing was XAD-2 (Alltech Associates/
Applied Science, 20/50 mesh, 90 1 pore size), Soxhlet extracted with methylene
chloride. The resin is then dried overnight with a gentle stream of prepuri-
fied nitrogen.
5.1.1.2 Storage of Extracted XAD-2—
Any precleaned XAD-2 resin not used immediately (within 2 weeks) was
stored under high purity methanol until needed. Traps were packed with dry
XAD and sealed .with glass plugs. XAD cartridges must then be used within
2 weeks of packing.
5.1.1.3 Glassware Cleaning—
Prior to testing, all sampling train glassware was cleaned and individual
pieces were marked. A data sheet for each assembled train was filled out,
detailing precisely which pieces of glassware were used. All glassware used
for collection or storage of organic semi volatile compounds was cleaned
according to the following procedure:
1. Wash in hot soapy water (Alconox or equivalent),
2. Rinse five times with tap water.
5-1
-------
TABLE 5-1. SUMMARY OF SAMPLING AND ANALYTICAL PROCEDURES
Parameters
Sampling method
Analytical method
<_n
i
ro
Particulate
Metals (Cd, Cr, As, Pb, Hg)
Molecular weight
PCDD/PCDF
Cyclone ash
Fabric filter ash
Bottom ash
Lime slurry
CEMS
MM5 combined train with metals
MM5 combined train with
particulate
M3
MM5
Composite grab sample
Composite grab sample
Composite grab sample
Single grab sample
C02
02
CO
NO
S02
HC1
Gravimetric
ICAP/AAS
Orsat
HRGC/MS
ICAP/AAS (metals)
*ASTM E777 (% combustibles)
ASTM D3174 (% carbon)
ICAP/AAS (metals)
*ASTM E777 (% combustible)
ASTM D3174 (% carbon)
IEEE 548-1984 (resistivity)
ICAP/AAS (metals)
*ASTM E777 (% combustible)
ASTM D3174 {% carbon)
ICAP/AAS (metals)
% Lime
Specific gravity
NDIR analyzer
Paramagnetic and
polarographic analyzer
NDIR analyzer
Chemiluminescence
Electrochemical
Various (refer to Entropy
report)
* Actually Galbraith Laboratories Method ME-6 based upon the ASTM method, but utilizes a
Leco CR-12 IR detector for analysis.
-------
Quartz/Glass Liner Jv
Cyclone (Optional)
Potentiometer \ Filter
T/C Check
Valve
Reverse - Type
Pitot Tube
cn
CO
Stack
Wall
Thermocouple
I
^ Manometer
Sample Box
Silica Gel
Probe
Heater
© © © © © ©
vacuum Line
Ice Bath
Thermometers By-pass
Valve
Vacuum
Gauge
vaivt! oaugt
rrnv-i^H^
Orifice
Main Valve
%
Manometer
Dry Test
Meter
Airtight
Pump
© Condenser with Ice Water Jacket
© XAD Resin Cartridge with Ice Water Jacket
(5) Modified Greenburg-Smith lmpinger,100mL of Double Distilled in Glass H2Q
© Modified Greenburg-Smith Impinger, 100mL of Double Distilled in Glass H20
© Modified Greenburg-Smith Impinger, Empty
© Modified Greenburg-Smith Impinger, SiC>2
Figure 5-1. Modified Method 5 Sampling Train for Semivolatile Organic Compounds.
-------
MODIFIED METHOD 5 TRAIN
SEMIVOLATILE ORGANICS SAMPLING
FIELD LAB SET UP DATA
Project No. 8910 L (02-12)
Plant: Maine Energy Recovery Co.
Run No.
Biddeford, ME
Sampling Location:
Sampling Train No.
Sample Box No.
Unit A,
Train to be used for:
Source Sample
Pre-Run Proof Blank
Post-Run Proof Blank
Location Blank
COMPONENT NO.
Sample Box Set Up Date/Person;
Sample Box Leak Check;
TRAIN COMPONENT
Nozzle
Probe •
$ Probe Blank-off
cfm @
in.
Hg vacuum
LOADING DATA
a Glass Union Blank-off_
Glass Union
Sample Transfer Line
? S-T Line Blank-off
Cyclone
Flask
Long 90° Adapter
Filter Holder Front
Filter Holder Back
Short 90° Adapter
Condenser
T/C-well U-Adapter
XAD Cartridge
U-Adapter (A)
3rd Impinger Mod-GBS
U-Adapter (B)
4th Impinger GBS
U-Adapter (C)
5th Impinger GBS
U-Adapter (D)
6th Impinger Mcd-GBS
Gooseneck
FiIter No.
Initial
Weight
(grams)
100 mis D.I. H,0
100 mis D.I,
Empty
H20
'200 g Silica gel
Component changes during run:
Figure 5-2.
Field Lab Set-up Data Sheet.
5-4
-------
3.
Rinse three times with deionized distilled water.
4. Rinse three times with acetone.
5. Rinse three times with hexane.
6. Air dry and seal to prevent contamination.
The blanks for all assembled trains were proof-rinsed prior to field use
to verify cleanliness. See below for details of glassware proofing.
5.1.1.4 Pretest Calibration of Equipment-
Prior to field testing, the Modified Method 5 (MM5) sampling equipment
was calibrated. This equipment includes the dry gas meter, gas meter tempera-
ture instruments, nickle-plated nozzles, p-'tot tubes (S-type), barometer, and
thermocouples used in conjunction with sampling. Pretest and posttest cali-
bration data sheets are Included in Appendix J. Procedures are noted in
Appendix M-3.
5.1.1.5 Proof Rinsing of Glassware—
Before any MM5 sampling took place, blank proof-rinsing of glassware for
each fully assembled train was performed. The trains were assembled as if
ready for field use, then were broken down as if a test were complete. The
trains were then cleaned and rinsed by the appropriate methods (see Sec-
tion 5-. 1.3, Sample Recovery). Sample components were set aside 'as samples for
blank analysis.
5.1.2 Sample Train Operation
Sampling time for each run was intended to be a total of 4 h (240 min).
However, plant malfunctions restricted sampling time for runs 1 and 3 to 160
and 190 min, respectively.- PCDD/PCDF sampling was performed concurrently with
particulate/metals sampling with train start/stop times between them varying
by no more than a 5-min interval.
Leak checks on the probe/train. assembly were performed prior to beginning
each run, after the traverse of one port, prior to traversing a new port, and
following the run. All leak checks conducted prior to a traverse were done at
> 15 in. HjO vacuum. Leak checks performed following a traverse were done at
the highest vacuum encountered during the traverse. Acceptable limits for
leak rates were established by EPA Method 5.
Both sampling locations required the use of three separate sampling ports
(see Section 4.0, Sampling Locations, for schematic). Traverse lengths at
both locations were 50 in. A total of 24 points were sampled, at both loca-
tions, using 8-point traverses in each of the three ports. Sampling time was
10 min/point, with readings taken every 5 min. In all cases the inlet sam-
pling train was operated with a cyclone and catch flask in place for removal
of large particulates. The outlet train had a much lower particulate loading,
allowing use of the cyclone bypass without a catch flask. Joints within all
trains were greaseless; threaded glassware was used with Teflon ferrules for
seal. Nickel nozzles were used at all times.
5-5
I
-------
One set of semivolatile blank train samples was taken from both the inlet
and outlet locations. The blank trains were assembled, taken to the appropri-
ate location, leak-checked, and left idle (no heat or flow) at the sampling
site for the same length of time as the normal train. The blank train was
leak-checked again whenever the normal trains were leak-checked, thereby
approximating the volume of ambient air taken in by the normal trains. Blank
trains were recovered exactly as if they contained a sample.
5.1.3 Sample Recovery
Following each test, the sampling train and probe assembly were discon-
nected, and acetone-rinsed aluminum foil was placed over connections, prevent-
ing contamination or sample loss during transport to the field lab. A copy of
the sample recovery data sheet is shown in Figure 5-3. Recovery of the probe
was performed as follows:
1. Nozzle was removed and flask attached to probe outlet.
2. Nozzle was rinsed and brushed with acetone until clean,
3. Nozzle was rinsed and brushed with hexane.
4. Probe was rinsed and brushed with acetone into flask twice.
5. Clean flask was attached.
6. Probe was rinsed again and brushed with acetone. Probe inlet was
sealed with Teflon-wrapped stopper or thumb. Probe was then tipped
and acetone allowed to pass through its length several times. If
acetone was dirty, flask was replaced and process repeated until
clean. All rinses were saved as one sample.
7. Probe was rinsed and brushed into flask three times with hexane.
Probe was tipped to allow hexane to pass through probe several
times. An empty flask was used at the start of each rinse. All
rinses were saved as one sample.
The heated flexible Teflon sample transfer lines were recovered in a man-
ner identical to that of the probes.
MM5 semivolatile sample trains were recovered in the field lab according
to the following procedure:
1. Each impirger was weighed and its weight recorded. The amount of
condensate collected was calculated and given to personnel perform-
ing data reduction.
• 2. Amount of silica gel exhausted was documented.
3. Empty sample containers with labels and lids were weighed.
5-6
-------
MODIFIED METHOD 5 TRAIN
SEMI VOLATILE ORGAN ICS SAMPLING
FIELD LAB RECOVERY DATA
Project No. 8910 L (02-12)
Plant; Maine Energy Recovery Co.
Run No.
Siddeford, ME
Train was used for:
Sampling Location: Unit A,
Sampling Train No.
Sample Box No.
Sample Box Recovery Oate/Person
Nozzle/Probe/Sample Transfer Line Recovery Date/Persons
BACK HALF RECOVERY
Impinger Condenser XAD 3rd 4th
Final Wt.(g)
Initial Wt.(g)
Net Wt, (g) '
Description/
Color
Sample Bottle
Tare Wt, (g)
Sample Number
5th
6th
Condensate
Collected
(grams)
Pour condensate in condenser outlet into XAD resin cartridge
Rinse Solvents
Components
Rinsed:
Sample Bottle
Final Wt. (g)
Net Wt. (g)
Acetone/
Hexane*
Filter Support
Filter Holder Back
Short 90° Adapter
Condenser
T/C-well U-Adapter
Acetone/
Hexane*
3rd,4th,5th Impingers
U-Adapters (A,B,C)
Component
Sample Bottle
Tare Wt. (g)
Sample Number
Rinse Solvents
Sample Bottle
Final Wt. (a)
Net Wt. (g)'
FRONT HALF RECOVERY
Nozzle/Probe Cyclone/Flask
Glass Union (loose contents!
Sample Transfer Line
F i1ter
Long 90° Adapter
Cyclone/Flask
.Filter Holder Front
Acetone/Hexane*
Acetone/Hexane*
Acetone rinses {with Brushing) precede Hexane rinses.
Figure 5-3, Field Lab Recovery Data.
5-7,
-------
4. Sample numbers were entered on data sheets.
5. Sample train components were rinsed into appropriate sample con-
tainers according to format on data sheet.
6. pH was checked and adjusted with concentrated Baker Instra Analyzed
nitric acid as necessary,
7. Full sample container with lid was weighed, allowing calculation of
the net weight of sample. The reweighing of samples after shipment
verified integrity of the sample.
8. Liquid levels were marked on each sample container, allowing a quick
determination of leakage.
All samples (probe rinses and recovered train) were sealed, double-
bagged, and stored in coolers with ice. The ice was checked periodically and
during shipment to ensure proper storage conditions were maintained.
5.1.4 PCDO/PCDF Analysis
The analysis of all MM5 samples was performed in accordance with the
draft ASME procedures,with the exception of some minor changes normally made
by MR I and used in previous projects. These are listed below, and no other
deviations from the referenced protocol were necessary. Figures 5-4, 5-5, and
5-6 show sample extraction analysis schemes for PCDO/PCOF compounds.
1. All glassware was thoroughly detergent-washed, 5X rinsed in warm tap
water, 3X rinsed with distilled water, rinsed with bulk acetone, air-dried,
and stored until use. Immediately before use, all glassware was rinsed with
high purity acetone followed with a 2X rinse with the solvent used in the
method. Glassware blanks were collected and analyzed as appropriate.
2. Condensate and rinse samples from MM5 trains were extracted according
to EPA Method 8280* but were analyzed according to the draft ASME procedure
like the rest of the MM5 train samples.
3. The sample cleanup columns used were different from those specified
in the ASME Method.The columns were 1 x 10 cm columns packed with 1.0 g of
silica gel and 4.0 g of 40» w/w sulfuric acid-modified silica gel; and 1 x
30 cm columns packed with 6.0 g of alumina covered with 1 cm of anhydrous
sodium sulfate. The cleanup procedure and sorbent cleaning and preparation
procedure are provided in Reference 6. In accordance with best professional
judgment and good analytical techniques, a third and/or fourth column, accord-
ing to the ASME protocol, or a performance equivalent, was utilized as
necessary.
4. HRGC columns used by MRI for determination of all 2,3,7,8-substituted
and total PCDD/PCDFs are 60-m DB-5 columns.
The above changes in the analysis method are considered to be minor. A
brief summary of the referenced analysis method is presented below.1*
5-8
-------
o
Probe Rinse
©
Cyclone/Filter
©
XAD Resin
© ©
Back Half Rinse Condensate
Apparatus Rinse —*¦
Add 1JC Method
Internal Standards
(Surrogates)
Add C Method
Internal Standards
(Surrogates)
Combine Extracts
Combine Extracts
Combine Extracts
Apparatus Rinse
Apparatus Rinse
Rotoevaporation
Rotoevaporation
Separatory
Funnel
Extract
Soxhlet
Extract
Soxhlet
Extract
Separatory
Funnel
Extract
Figure 5-4. Sample Extraction Scheme for Inlet MM5 Samples.
-------
Filter/XAD
Probe Rinse/Back Half Rinse/Condensate
Apparatus Rinse ~
4 Add 13C Method Internal ~
Standards (Surrogates)
Combine Extracts
Rotcevaporation
Soxhlet
Extract
Separatory
Funnel
Extract
Figure 5-5. Sample Extraction Scheme for Outlet and Blank MM5 Samples.
5-10
-------
Concentrate
by N2
r
Column
* Acidifis
* Acidifiec
Clean-up
3d Silica
Alumina
Identify
PCDD/PCDF
Analyze by
HRGC/MS
Quantitate
PCDD/PCDF
* = Add Recovery Internal Standards
Figure 5-6. Sample Cleanup and Analysis Scheme for PCDD/PCDF and
Other Organies in the MM5 Train.
5-n
-------
Each MM5 train composite sample is analyzed for the specified PCDD/
PCDF. Prior to extraction, one component of each MM 5- front-half and back-half
sample fraction is spiked with the method internal standards, The MM5 com-
ponents are extracted, composited, and concentrated by rotoevaporation. Each
sample extract is cleaned up using the column cleanup techniques described
below.
A 1 x 10 cm chromatography column is packed with 1.0 g of silica gel and
4.0 g of 40% w/w sulfuric acid-modified silica gel. A second chromatography
column (1 x 30 cm) is packed with 6.0 g of alumina covered with 1 cm of
anhydrous sodium sulfate. The silica gel (type 60, EM reagent, 100/120 mesh)
and alumina (acid alumina, AG 4, Bio-RAO) are Soxnlet extracted with methylene
chloride and activated at 130° and 190°C, respectively, before use. Each
sample extract was added -to the top of separate silica gel columns along with
two 0.5-mL washes of the sample container. The columns are eluted with 45 mL
of hexane directed onto the alumina columns. The alumina columns are eluted
with an additional" 20 mL of hexane, and the hexane is archived. The alumina
columns are then eluted with 20 mL of 20% v/v methylene chloride/hexane.
These samples are concentrated under nitrogen to about 1 mL, transferred to a
1-mL conical vial, and further evaporated just to dryness. Immediately prior
to analysis, the sample residue is taken up in 23 pL of tridecane and the
recovery internal standard is added.
The extracts were analyzed by high resolution gas chromatography/mass
spectrometry with selected ion monitoring (HRGC/MS-SIM) using a 60 m x 0.25 mm
DB--5 fused silica capillary column (FSCC). For analysis of all 2,3,7,3-
substituted PCDO/PCOF extracts were analyzed using a 60 m x 0.25 mm DB-5
column. The mass spectrometer operated at 3000 resolution units.
The levels of dioxins and furans were calculated by comparison of the
response of the' samples to calibration standards. The response of the
recovery internal standards was monitored from run to- run for conformance to a
50% criterion envelope. Concentrations of each PCD0/PCDF congener were deter-
mined by comparison to the appropriate response factor determined from the
calibration curve. Final concentrations are reported in units of ng/dscm.
5-1.5 GC/HS Data Reduction
The quantification of the samples was based on the internal standard
method in which a constant amount of the recovery internal standard is added
to all samples, blanks, and calibration standards just prior to analysis. The
raw data for the quantification of the sample components consists of the com-
puter-measured peak areas of the characteristic mass fragmentation ions of the
method internal and recovery internal standards. The raw data are converted
to concentrations by using mass spectrometric response factors relative to the
internal standard.
If none of a target analyte was found, and thus could not be quantitated,
the lower limit of detection is reported. This lower limit of detection is
determined using the following steps:
5-12
-------
1. A background area is determined by inspection of the GC/MS data.
2. Background area x 2.5 = lowest individual ion area that can be used
to confirm the presence of the compound of interest.
3. Using the area in (2) and theoretical ion ratios, the minimum area
that can be quantitated is determined.
4. The area from (3) is applied to the calculations to obtain the min-
imum concentration of the compound of interest that can be detected.
If a target analyte was found to be saturated, the sample was diluted,
recovery internal standard added, and reanalyzed. The method internal stan-
dard recoveries and relative response factors were determined on the undiluted
sample. The total amount of the analyte in the sample was determined using
the following steps:
. I. The unsaturated peaks from the original sample were entered onto a
LOTUS 1-2-3 spreadsheet along with the method internal and recovery internal
standards.
2. The original. saturated peaks, which were on scale in the diluted sam-
ples were entered onto a new LOTUS 1-2-3 spreadsheet with the recovery
internal standard.
3. The amounts (rig) in both samples, undiluted and diluted, were then
added together for the amount of analyte.
5.1.5,1 Determination of the Relative Response Factors (RRF)--
For initial calibration and certification of the GC/MS method, a minimum
of three compound levels covering a significant portion of the linear range of
the instrument was used to determine instrument sensitivity and linearity.
Ideally, the response factors are constant over the entire concentration
range of interest. However, the response factors may vary with concentra-
tion. The relative response factor (RRF) is plotted against the area or peak
height of the analytes in the calibration standards, using a minimum of three
concentrations over the range of interest. The relative response factor is
calculated according to the equation:
where:
As
Cis
*RF" Mfy E(<-
Area or peak height of the primary characteristic ion of the
compound being quantified;
Concentration or amount of the recovery internal standard;
5-13
-------
Ais = Area or peak height of the primary characteristic ion of the
recovery internal standard; and
C$ = Concentration or amount of the compound in the calibration
standard.
5.1.5.2 Peak Identification-
Selected ion monitoring (SIM) was used to analyze the MM5 extracts for
both qualitative and quantitative peak identification.
Qualitative peak identification refers to the peak eluted within the
retention time windows set for that analyte. The sample spectrum is compared
to that of a calibration standard. The intensity of the two largest ions in
the molecular cluster must match the ratio observed for a standard within
±20%. System noise at low concentration or interferences may skew the ion
ratio beyond the ±20% criteria. If the analyst's best judgment is "that a peak
that does not meet the qualitative criteria, i.e., is a match, the peak may be
included in the calculation, with a footnote explaining the data and the
reason for relaxing the criteria.
After a chromatographic peak is identified as a positive match, the com-
pound is quantitated based either on the integrated abundance of the extracted
ion current profile (EICP) or the SIM data for the primary characteristic ion
in the appropriate tables listed in the analytical protocol. If interferences
are observed for the primary ion, the secondary and tertiary ion are used for
quantitation. For the ion used, the RRF is determined using that ion. The
same criteria are applied to the internal standard compounds.
Primary, secondary, and tertiary ions are extracted from the recon-
structed ion. chromatograms. Ratio criteria for'the ions must be met before
the analyte is quantified.
In this study, quantification was based on the primary ion.
Following sample analysis, the appropriate response factor was taken from
the calibration curve generated from Eq. (1) and the concentration of the com-
pound in the sample calculated according to the equation:
Sample Concentration = [ais)(RRF) (2)
It is very important to determine the correct value for Cis, the concen-
tration of the internal standard relative to the original sample matrix.
5-14
-------
All of the peaks were summed for each analyte, and then those were summed
to yield the total mass in the sample.* For a concentration-per-peak or eon-
centration-per-analyte reporting format, each value is carried through the
calculations appropriate manner.
5.1.6 Data Entry.
Data transfer and reduction are essential functions in summarizing infor-
mation to support conclusions. It is essential that these processes be per-
formed accurately and, in the case of data reduction, accepted statistical
techniques be used.
The entry of input data was a HP 110 portable computer which can utilize
the LOTUS 1-2-3 spreadsheet software package.
At a minimum,¦example calculations must be included,with the summarized
data to facilitate review. The entry of input data and calculations should be
checked and the signature/initials of the data technician and reviewer(s)-
accompany all data transfers with and without reduction.
5.1.7 GC/MS Data Validation
The principal criteria was used to validate the integrity of the GC/MS
data acquired and reported during this program were the following;
1. Verification on a frequent basis by the analytical task leader that
all raw data generated in the preceding week had been stored on magnetic tape
and/or in hard copy and that storage locations were documented in the project
records.
2. Examination of at least 5% of the raw data (e.g., chromatograms) on a
daily basis by the organic analytical task leader to verify adequacy of docu-
mentation, confirm peak shape and resolution, assure that the computer was
sensing peaks appropriately, etc.
3. Confirmation that raw areas for internal standards and calibration
standards and raw and relative areas for surrogate compounds were within 50%
of the expected value.
5.1.8 SC/MS Analytical Data
The quantification of the sample components was based on the internal
standard method in which a constant amount of the internal standard 1s added
to all samples, blanks, and calibration standards.
The raw data for the quantification of the sample components consisted of
the computer-measured peak areas of the characteristic mass fragmentation ions
of the internal standards and the analytes of interest. The raw data were
* A check must be made to ensure that the number of peaks measured does not
exceed the theoretical maximum number of congeners for each isomeric group.
5-15
-------
converted to concentrations by using mass spectrometry response factors rela-
tive to the internal standard.
Concentrations of each PCOD/PCOF homolog were calculated by first calcu-
lating a response factor and then calculating a final, concentration in nano-
grams per sample using the following equations (with TCDO as an example):
Relative Response Factor (R.F.) = is^ x Fa (31
A(IS) C(std)
where: A(std) = Area of 10ns mf2 320 and 322 for the unlabeled 2,3,7,8-TCQD
in the standard;
A-/is) = Area of ions m/z 332 and 334 for the 13Cl2-2s3,7,8-TCDD in
the standard;
C/tS\ = Concentration of 13C12-2,3,7,8-TC00 in the standard
(ng/niL); and
C/Std) = Concentration of unlabeled 2,3,7,8-TCDD in the standard
(ng/ml).
c = A(sample) C(IS)
(sample) A(IS) x RF "I- (4)
where: *-(samDle1 = Total concentration of all TCDO isomers in the sample
(ng/sample);
A/_a Dl^\ = Total area of ions m/z 320 and 322 for all TCDD isomers
in the sample;
A(IS> = Area of ions m/z 332 and 334 for the *3Cl2-2,3,7,8-TCD0
in the sample; and
C/T<-\ = Concentration of l3C12-2,3,7,8-TCDD in the sample
{ ' (ng/ml).
The concentration of total TCDF is calculated with the above equations
using the response of ions m/z 304 and 306 to measure the concentration of
unlabeled TCDF and the response of ions m/z 316 and 318 for the 13Cl2-2,3,7,3-
TCOF. Similar procedures were used for each of the PCDD/PCDF homologs.
All data were qualified as "estimated" concentrations or tentative
identifications, except where pure isomer standards were used to verify the
results.
5-16
-------
•Recovery of method internal standards was monitored during analysis of
samples. Target recovery values were 50% to 150X as outlined in the QA plan.
Recovery values were consistently low for all the method internal stan-
dards, ranging from 39% to 61% for all samples and compounds. Approximately
70^ of all the method internal standard recoveries were within objectives
(I.e., > 50%). When recovery limits for internal standards and precision
limits were exceeded, the following actions were initiated:
1. The spiking procedure was checked, including the solution concentra-
tion, preparation, techniques, and calculations.
2. The archive and 50$ methylene chloride/hexane fraction of the column
cleanup procedure were concentrated and-analyzed,
3. The data and spiking procedure was checked by a third person.
Monitoring for ion masses of possible furan interferents (chlorinated
diphenyl ethers, CDPE) was completed for all samples (except duplicates) where
a positive furan response was obtained. , The CDPE compounds were monitored
simultaneously with the PCDFs. Specifically, m/z 374 (HxDPE) was monitored
vs. m/z 304 and 306 (TCDF); m/z 408 (HpDPE) vs. m/z 338 and 334 (PeCDF);
m/z 444 (CDPE) vs. m/z 374 and 376 (HxCDF); m/z 478 (NDPE) vs. m/z 408 and 410
(HpCDF); and mfz 512 (DOPE) vs. m/z 442 and 444 (OCDF). These monitoring pro-
cedures are provided in Table 1 of the ASME analytical procedure.
By monitoring the PCDF Ions and corresponding CDPE ions within a reten-
tion window, the presence of the possible interferents could be verified.
Simultaneous responses for the specific PCOF homologs at the appropriate ratio
and retention time and no response to the corresponding CDPE provided positive
identification of the PCDF.
5.2 COMBUSTION GAS—PARTICULATE AND METALS SAMPLING AND ANALYSIS
The combustion gas sampling for particulate and metals analysis at the
spray dryer inlet and the baghouse outlet was performed according to proce-
dures specified in EPA Methods 1 through 5 and in Reference 7. This EMB
metals protocol can be found in Appendix M. These procedures were followed
except as discussed below where deviations from these methods and selected
options in these methods as performed during the test program are described.
5.2.1 Equipment and Sampling Preparation
The combustion gas samples for metals and particulates were collected in
the sampling train shown in Figure 5-7. Due to space restrictions at the
spray dryer inlet, a flexible heat-traced Teflon sample transfer line con-
nected the glass probe liner to the cyclone of the sampling train. The sample
transfer line was not used, and a bypass replaced the cyclone in the train
used at the baghouse outlet location. The sodium hydroxide Impinger was used
as an acid trap to protect downstream train components.
5-17
-------
Quartz/Glass Liner I
\
Thermocouple
Cyclone (Optional)
Potentiometer V Filter
Reverse - Type
Pitot Tube
U~l
I
00
T/C Check
Valve
Stack
Wall
Manometer
Sample Box
Probe
Heater
Silica Gel
©©©©©©
Vacuum Line
Ice Bath
Vacuum
Gauge
Thermometers By-pass
Valve
Orifice
o
Main Valve
Manometer
Dry Test
Meter
Airtight
Pump
© Modified Greenburg-Smith Impinger, Empty
0 Greenburg-Smith Impinger.lOOmL Nitric Acid/Hydrogen Peroxide
© Greenburg-Smith Impinger.lOOmL Nitric Acid/Hydrogen Peroxide
(I) Greenburg-Smith Impinger.lOOmL Acid/Permaganate
© Greenburg-Smith Impinger.lOOmL 0.5 N NaOH
© Modified Greenburg-Smith Impinger, SiC>2
Figure 5-7. Modified Method 5 Sampling Train for Particulates and Metals,
-------
All sampling nozzles were made of nickel. These nozzles were used
because nickel was not one of the target metals. The nozzles were riot rinsed
with acid during sample recovery. A list of other equipment used specifically
for this test program may be found in Table J-2 in Appendix J.
Each train component and piece of glassware was individually identified
with a component number recorded on the field lab set-up data sheet before
each use. This prevented the accidental exchange of components among the
trains. Three trains were set up to be used for one purpose throughout the
test program: inlet train, the outlet train, and a blank train used at both
the inlet and outlet during two of the sampling runs.
5.2.2 Sampling Train Operation
Samples were extracted for 10 min at each of the 24 sampling points dur-
ing a complete run (run 2) or until the run had to be stopped because of pro-
cess problems (runs 1 and 3), Sampling and combustion gas data were recorded
every 5 min during each run. Sampling was conducted according to U.S. EPA
Method 5 with appropriate modifications as necessary to accommodate the two
sampling locations. Leak checks were performed at the beginning and end of
each run and before and after each port and train component change.
Static pressure determinations were made several times during each run,
and the results were averaged and used in the final calculations for combus-
tion gas volumetric flow rates and the isokinetic sampling rates for both
trains used at each location. A concurrent velocity head reading and total
pressure reading (impact tip of pi tot) were recorded. A I'-tube manometer con-
nected to the impact pi tot line was used to measure total pressure. The
velocity pressure was obtained by multiplying the velocity head, by the squared
pitot tube coefficient. The static pressure was obtained by subtracting the
velocity pressure from the total pressure.
5.2.3 Sample Recovery
After a sampling run, each train was disassembled into sections (probe,
sample transfer line, sample box) before being transferred to the field lab.
The nozzles were sealed with plastic cap p'ugs and the probe outlets were
sealed with glass blank-offs. The ends of each sample transfer line were con-
nected with the glass union that had connected the line to the probe* liner.
The sample box inlet was covered with aluminum foil that did not contact any
sample surface.
Certain apparatuses, i.e., probe flasks and brushes, used for recovery
were designated for use on only one train to prevent cross-contamination of
samples. All sample containers had preprinted, computer-generated labels.
Replicates of each label were used to (a) identify the sample container,
(b) identify sample container over-wrap, and (c) verify recovery by entry Into
the field laboratory log book. The sample identification logs are in
Appendix K.
5-19
-------
The nozzles were removed from the probes and brushed and rinsed with ace-
tone until clean. Each probe was brushed and rinsed with acetone at least
three times or until clean into a flask attached to one end. The flask was *
replaced, and the probe was rinsed again by passing acetone back and forth
through the probe several times to remove any residual particulate. The sam-
ple transfer lines were recovered in the same manner, and all probe/nozzle/
sample transfer line acetone rinses from each train were combined as one
sample.
The probe liners and sample transfer lines were then rinsed three times
with 0.1 N nitric acid. A brush was not used, and the nozzles were not rinsed
with acid. The acid rinses for each train were combined and saved as one sam-
ple from each train.
The probe liners and sample transfer lines were rinsed again with ace-
tone, which was discarded.
After removal of the filter and the loose cyclone/flask catch, the front
half of the filter holders and the remaining front half components in the sam-
ple box were brushed and rinsed with acetone until particulate recovery was
complete. The same components were then rinsed with 0.1 N nitric acid at
least three times. The acetone and acid rinses were saved as separate
samples.
The amount of condensate collected in the impingers was determined by
weight change of the impi ngers after each run and was used to determine the
moisture content of the gas samples. The impinge** solutions were recovered as
specified in the EMB metals protocol.7 The impinger containing sodium
hydroxide was rinsed with deionized water.
All samples were stored in coolers containing bagged ice before and dur-
ing shipment to the analytical laboratory. The samples were then placed in a
refrigerated room to await analysis.
5.2.4 Particulate Analysis
All samples for particulate analysis were checked for leakage and loss
during shipment by examining the liquid level marks on the containers. No
losses occurred. The acetone samples were weighed and transferred to tared
glass beakers. The sample containers were rinsed into the beakers with
weighed amounts of acetone. The acetone was evaporated at about 70"F (room
temperature) and ambient pressure. The acetone residues and the combined
filter/cyclone catches were desiccated at ambient temperature and pressure,
and they were weighed to a constant weight as defined in Method 5. Acetone
and filter blanks were treated in the same manner. Figure 5-8 illustrates the
sequence of sample fraction analyses.
5-20
-------
Fronl hall recovery fractions (store on ice)
Field Recovery
Moisture Determination
Melals Analysis
Combine
Analyzo by cold vapor AA
for Hg.
Analyze by cold vapor AA
for Hg.
Analyze by ICAP and/or
appropriate AA procedure
for Cd, total Cr, As and Pb.
Analyze by ICAP and/or
appropriate AA procedure
for Cd, tolal Cr, As and Pb.
Cyclone/llask
or cyclone bypass,
Front hall Hilar holder
(brush/rinse)
Nozzle, probo.
sample transfer line
(brush/rinse)
Probe, sample
transfer line
(rinse)
Cyclone/llask
or cyclone bypass,
front half filter holder
(rinse)
Cyclone/flask
catch, filler
0.1 N nitric acid rinse:
To recover residual
trace melals subsequent
to acetone rinse,
0.1 N nitric acid rinse:
To recover residual
trace melals subsequent
to acetone rinse.
Digest each 0.5 g portion
with HF and nllric acid
using pressure relief vessels
In a rrilciowave oven.
Digest each 0.5 g portion
with HF and nitric acid
using pressure relief vessels
In a microwave oven,
Dilute to volume wlih
0.5 M boric acid in distilled
deionlzed water maintaining
a matrix ol 12% nitric acid
and 8% HF. Take aliquol
for Hg analysis.
Dilute lo volume with
0.5 M boric acid In distilled
delonized water maintaining
a matrix of 12% nilric acid
and 8% HF, Take aliquol
for Hg analysis.
Combine. As per Method 5,
desiccate at ambient
temperature and pressure,
weigh lo constant
weight for particulate
determination.
Acetone rinse: iniiial
rinse lo recover particulates.
As per Method 5, evaporate
to dryness at ambient
temperature and pressure,
desiccate, weigh to consianl
weight for particulate
determination.
Acetone rinse: Iniiial
rinse to recover particulates.
As per Method 5, evaporate
lo dryness ai ambient
temperature and pressure,
desiccate, weigh to constant
weight for paniculate
determination.
Add potassium permanganate,
sulfuric acid, nilric acid,
potassium persulfale. Digest
at 95°C In a water bath for
2 hours. Cool. Add
hydroxylamine hydrochloride
and stannous sulfate.
Add potassium permanganate,
sulfuric acid, nitric acid,
potassium persulfale Digest
at 95°C In a water bath for
2 hours. Cool. Add
hydroxylamine hydrochloride
and stannous sulfate.
Figure 5-8. MERC Facility Particulate/Metals MM5 Train Field Recovery and Analytical Protocol.
-------
Field Recovery
Moisture Deierminaiion
Welals Analysis
Combine.
Adjust pH lo < 2.
Take aliquot
(or Hg analysis,
Archive until other
analyses complete.
Discard.
Analyze by ICAP and/or
appropriate AA procedure
for Cd, total Or, As, and Pb.
Analyze by cold vapor AA
for Hg.
Analyze by cold vapor AA
for Hg.
Weight as per Method 5
for moisture determination.
Recover. Rinse with
0.1 N nitric acid.
Weigh as per Method 5 for
moisture determination.
Recover. Rinse with
delonlzed distilled water.
Impingers 1.2,3
Condensate and nitric
acid/hydrogen peroxide
Rinse as per Method 5 lor
moisture determination.
Recover. Rinse with
potassium permanganate.
Rinse with
0.1 N nitric acid
Impinger 5
Sodium Hydroxide
Back half filler holder,
filler support, connecting
glassware Ihrough
ihird Impinger.
Impinger 4
Acidified potassium
permanganate.
Weigh as per Method 5 for
moisture determination.
Recover.
Impinger 6
Silica Gel
Reduce volume to near
dryness (20 mL). Digest
with niiric acid and
hydrogen peroxide. Add
hot water and heat.
Filler if necessary and
dilute to volume wlih
distilled deionized water.
Add sulfuric acid, nitric
acid, potassium persulfate.
Digest at 95°G In a water bath
for 2 hours. Cool. Add
hydroxylamine hydrochloride
and stannous chloride.
Add potassium permanganate,
sulfuric acid, niiric acid,
potassium persulfate. Digest
at 95°C in a waler balh for
2 hours. Cool. Add
hydroxylamine hydrochloride
and stannous sulfate.
Back half recovery fractions (store on ice)
Figure 5-8. (Continued)
-------
5,2.5 Metals Analysis
All front half sample fractions underwent a microwave digestion procedure
employing a CEM Corporation microwave digestion system with hydrofluoric acid
and nitric acid as the digestion matrix. The digestates were diluted to vol-
ume with.0.5 M boric acid to reduce the destructive capabilities of the hydro-
fluoric acid on analytical instrument components.
An aliquot of each of the digestates was taken and analyzed for mercury
by cold vapor atomic absorption spectroscopy according to SW-846,5
Method 7470. The balance of the digestates was analyzed for cadmium, total
chromium, lead, and arsenic by inductively coupled argon plasma atonic emis-
sion spectroscopy (I CAP) according to SW-846, Method 6010. Where necessary,
graphite furnace atomic absorption spectroscopy was used to analyze for lead
and arsenic.according to SW-846, Methods 7421 and 7060, respectively.
The combined condensate and nitric acid impinger fraction was digested
according to SW-846, Method 3050. An aliquot of the digestate was removed for
nfercury analysis by Method 7470. The balance of the digestate was analyzed
for cadmium, total chromium, lead, and arsenic by Method 6010. Where neces-
sary, Methods 7421 and 7060 were used to analyze for lead and arsenic,
respectively.
The acidified potassium permanganate impinger fraction was analyzed for
mercury according to Method 7470. The sodium hydroxide impinger fraction was
not analyzed. Figure 5-8 illustrates the sequence of the sample fraction
analyses.
5,2.6 Data Reduction
The particulate loading was calculated with the following equations.
Mass of particulate collected:
mn ^mp ~ mb^filter/cyclone + ^mp ~ ""b^ acetone rinses
mn = Mass of particulate collected from source (g)
rrip = Gross mass of particulate (g)
nijj = Mass of particulate in blanks (g)
Particulate concentration (actual):
C = (0.001 g/mg) (mr/Vm(std))
C = particulate loading (mg/dscm)
i = volume of dry gas sampled corrected to standard conditions
• (dscffl § 1 atm and 68°F)
Eq. (5)
Eq. (6)
5-23
-------
Particulate concentration (corrected to 12% CO2):
CN = C(12/CO2) Eq. (7)
= particulate concentration (mg/dscm), corrected to 12% C02
C02 = percent by volume of C02 in combustion gas, dry-basis
Particulate mass emission rate:
M = C Qstd (60 min./h) (10"« kg/mg) Eq. (8)
M = particulate mass emission rate (kg/h)
Qstcj = combustion gas volumetric flow rate on dry basis at standard
conditions 68°F, 1 atm (m3/min.)
Metal analyte concentration:
cmetal = Cco " cb)/vm(std) (9)
^metal = ITie^al concentration in combustion gas (yg/dscm)
CQ = mass of metal detected in sample (ug)
Cb = mass of metal detected in blank (pg)
vm(std) = volume of dry gas sampled corrected to standard conditions
(dscm @ 1 atm and 68°F)
5.3 ASH, LIME SLURRY, AND ROF SAMPLING AND ANALYSIS
Five different types of grab samples were collected at the test site:
cyclone ash, baghouse discharge ash, bottom ash, RDF, and lime slurry. All
samples requiring a separate analysis (i.e., organics vs. metals) were split
in the field and placed in' an appropriately prepared and marked sample
'container.
Grab sample collection typically began 40 to 60 min after start of the
test, with grabs being taken about every hour thereafter. Final samples were
collected about 1 h after the completion of a test. The exception to this was
the lime slurry, which was sampled only once each test period, 2 h into the
run.
5.3.1 Cyclone Ash
5.3.1.1 Sampling—
Cyclone ash was sampled through a 5-cm (2-in.) gate valve located on the
bottom side of an inclined cylindrical chute connecting the cyclone to the ash
conveyor system. The sampling procedure consisted of opening the sample port,
clearing the sample port with a steel rod to ensure against bridging, and col-
lecting sample in a 5-gal. steel pail. Following the test, the ash was mixed
for homogenization and placed in the appropriate sample jars.
5-24
-------
On several occasions very little cyclone ash was available for sam-
pling. In these cases, any sample obtained was still mixed in and split
evenly among the appropriate containers.
5.3.1.2 Analysis-
Analysis for percent combustibles and percent carbon was performed by
Galbraith Laboratories. Percent combustibles analysis followed ASTM
Method E777 using a Leco CR-12 IR detector. Percent carbon analysis followed
ASTM Method D 3174.
5.3.1.3 PCOD/PCDF—
Samples for organics analysis were extracted with benzene, with the
extract currently archived at the MERC plant location.
5.3.1.4 Metals—
The sample was prepared for analysis by digestion with microwave diges-
tion procedure using a CEM Corporation microwave digestion system with hydro-
fluoric acid and nitric acid as the digestion matrix. The digestates were
diluted to volume with 0.05 M boric acid to reduce the destructive capabili-
ties of the hydrofluoric acid on analytical instrument components.
The analyses performed followed SW-846* Methods 6010 (ICAP), 7421 (lead
furnace), 7060 (arsenic furnace), and 7470 (mercury cold vapor), with method
modifications as detailed in Appendix H-2.
5.3.2 Baghouse Discharge Ash
5.3.2.1 Sampling—
Baghouse discharge ash was collected from a 5-cm (2-.in.) gate valve
placed on the bottom side of a conveyor roughly 3.5 m (10 ft) downstream from
the baghouse discharge. A 19-L (5-gal.) steel pail was placed beneath the
port for sample collection. Following each run, the sample was homogenized
(mixed) and split into appropriate containers for each analysis.
In addition to the samples collected above, two 19-L (5-gal.) pails
double-lined with polyethylene bags were filled with baghouse discharge ash.
These samples were turned over to EPA personnel for further analysis.
5.3.2.2 Analysis—
Analysis for percent combustibles and percent carbon was performed by
Galbraith Laboratories. The percent combustibles analysis performed followed
ASTM Method- E777 using a Leco CR-12 IR detector. Percent carbon analysis
followed.ASTM Method D 3174.
5.3.2.3 PCDD/PCCF—
Baghouse discharge ash samples were extracted with benzene. The extracts
are currently archived at the MERC plant location.
5.3.2.4 Metals—
The sample was prepared for analysis by digestion with microwave diges-
tion procedure using a CEM Corporation microwave digestion system with hydro-
fluoric acid and nitric acid as the digestion matrix. The digestates were
5-25
-------
diluted to volume with 0.05 M boric acid to reduce the destructive capabili-
ties of the hydrofluoric acid on analytical instrument components.
The analyses performed followed SW-8465 Methods 5010 (ICAP), 7421 (lead
furnace), 7060 (arsenic furnace), and 7470 (mercury cold vapor), with method
modifications as detailed in Appendix H-2«
5,3.2.5 Resistivity-
Resistivity of baghouse discharge ash was determined by Southern Research
Institute using Method IEEE 548-1984. The samples were tested at a constant
humidity of 14.8% water vapor by volume. Results, included in Appendix I,
show maximum ash resistivity over the 325° to 390°f temperature range.
5.3.3 Bottom Ash
5.3.3.1 Sampling-
Bottom ash samples were taken from a conveyor that carries the ash from
the quench tank to a disposal area. As with other ash samples, individual
aliquots were mixed for homogenization and then split into the appropriate
sample containers. Due to problems with the process, EPA requested that no
run 3 bottom ash sample be collected.
5.3.3.2 Analysi s—
Analysis for percent combustibles and percent carbon was performed by
Galbraith Laboratories. Percent combustibles analysis followed ASTM
Method E777 using a Leco CR-12 IR detector. Percent carbon analysis followed
ASTM Method 0 3174.
5.3.3.3 PC00/PCDF—
Bottom ash samples collected for organics analysis were extracted with -
benzene. The extracts are currently archived at the MERC plant location.
5.3.3.4 Metals—
The sample was prepared for analysis by digestion with microwave diges-
tion procedure using a CEM Corporation microwave digestion system with hydro-
fluoric acid and nitric acid as the digestion matrix. The digestates were
diluted to volume with 0.05 M boric acid to reduce the destructive capabili-
ties of the hydrofluoric acid on analytical instrument components.
The analyses performed followed SW-846S Methods 6010 (ICAP), 7421 (lead
furnace), 7060 (arsenic furnace), and 7470 (mercury cold vapor), with method
modifications as detailed in Appendix H-2.
5.3.4 Lime Slurry
5.3.4.1 Sampling—
Lime slurry samples were taken from the feed line at the top of the spray
dryer. Prior to collection of each sample, the line was purged into a
bucket. Each grab consisted of three samples: one each for organics analy-
sis, metals analysis, and physical properties analysis. One grab was col-
lected for each run, approximately 2 h into the test.
5-26
-------
5.3.4.2 Analysis—
Percent lime, percent solids, and specific gravity analyses of the lime
slurry samples were performed by Galbraith Laboratories.
5.3.4.3 Metals—
The sample was prepared for analysis by digestion with microwave diges-
tion procedure using a CEM Corporation microwave digestion system with
hydrofluoric acid and nitric acid as the digestion matrix. The digestates
were diluted to volume with 0.05 M boric acid to reduce the destructive
capabilities of the hydrofluoric acid on analytical instrument components.
The analyses performed followed SW-846* Methods 6010 (ICAP), 7421 (lead
furnace), 7060 (arsenic furnace), and 7470 (mercury cold vapor), with method
modifications as detailed in Appendix H-2.
5.3.5 Refuse Derived Fuel
5.3.5.1 Sampling—.
RDF composite samples for run 1 were collected at the beginning, of the
long transfer conveyor belt which feeds into the plant's ram loading system.
A grain shovel was used for sampling a chute that feeds RDF onto the conveyor
belt; one shovelful was collected each grab. Samples were stored in a fiber
drum lined with double plastic bags.
Runs 2 and 3 RDF composite samples were obtained at the ram feed on
boiler A. A 19-L (5-gal.) pail was lowered by rope into the ram feed and
allowed to fill. The full pail was emptied into a fiber drum lined with
double plastic bags. s
5.3.5.2 Analysis-
Following completion of the test program, all RDF samples were turned
over to EPA personnel for storage and analysis.
5.4 CONTINUOUS GAS ANALYZERS
Three sets of MR I CEMs were used to monitor gaseous emissions from the
incinerator. Oxygen and carbon dioxide were measured at all three loca-
tions. SO2 was monitored at the spray dryer inlet and baghouse outlet. THC
and CO were measured at the spray dryer inlet, and N0X was measured at the
baghouse outlet. Each analyzer system was leak checked from the probe before
and after each test. All analyzers were also zeroed and spanned before and
after each test. The average of both calibrations was used to calculate the
final concentrations. All calibration gases were introduced at the inlet to
the conditioning manifold. The data loggers corrected the CO and C02 analyzer
outputs for non-linearity. The current C02 concentration was used to correct
the CO readings for the loss of C02 through the Ascarite scrubber. All data
were logged continuously with 1-min averages.
The accuracy of the working standard calibration gases was checked by
comparing them against EPA protocol no. 1 cylinders at the end of one test
day. Due to time constraints, the protocol cylinders could not be recertified
before the test. Instead, the protocol cylinders were returned to the vendor
5-27
-------
for recertification after the test. Since some of the span cylinders were
changed during the test, the various span cylinders were also included in the
comparative checks. Cylinder accuracy results and analyzer calibration data
are presented in Section 6.0.
A schematic of the equipment used at the spray dryer inlet is shown in
Figure 5-9. Table 5-2 is the detailed equipment list. Figure,5-10 and
Table 5-3 provide similar details for the spray dryer outlet. Figure 5-11 and
Table 5-4 describe the equipment used at the baghouse outlet. HC1 was also
measured by Entropy Environmentalists under a separate EPA contract. Details
of their equipment can be found in Reference 2.
5.5 CARBON DIOXIDE AND OXYGEN SAMPLING AND ANALYSIS BY EPA METHOD 3
Carbon dioxide and oxygen determinations were made during the three sam-
pling runs to obtain data for calculating the molecular weight of the combus-
tion gas and for adjusting pollutant concentration and emission results to a
standard excess air volume, i.e., 12% C02 or 7% 02. The procedures used dur-
ing this test program are contained in MR 11s SOP S-4 (see Appendix M) and are
consistent with EPA Method 3.
Sample extraction was performed at a constant rate during the course of
the particulate/metals MM5 train sampling at the spray dryer inlet and the
baghouse outlet locations. Integrated multipoint sampling was accomplished
with a stainless steel tube attached to the MM5 probe so that samples would be
extracted near the same points used for particulate/metals samples. Flue gas
analysis was done with an Orsat analyzer. The sampling data and analysis
results are in Appendix 0.
For each traverse at a port (10 min at each of eight points for the MM5
train), the first sampling point was the point farthest from the port. The
integrated gas sampling train was purged with combustion gas for approximately
5 min before sampling commenced. Sampling was discontinued as a precaution at
the point nearest the port to avoid dilution of the sample from any in-leakage
through the port seal. Although in-leakage was minimized by the seals, a
slight amount of in-leakage. could have occurred from time to time because of
the very negative pressure in the ducts. Concurrent temperature readings did
indicate, however, that any in-leakage would have been minimal. As a result,
the integrated gas sampling concurred with 85% to 95% of the MM5 sampling.
The inlet tip of the stainless steel tube was positioned approximately
3 in. behind the components (nozzle and pitot tube) of the MM5 probe tip to
prevent flow interference. The tube tip was in the area that would have been
affected by in-leakage during sampling at the points •closest to the ports.
More than the minimum number of points required by EPA Method 3 were sam-
pled. The sampling point matrix approximated that required by EPA Method 1
for 21 points and was sufficient to obtain reliable results for carbon dioxide
and oxygen concentrations during the three sampling runs.
5-28
-------
Heated Sample Line
Span
Zero
Probe
Rec
Rec
Rec
Rec
Rec
Logger
Printer
SO
CO
CO.
Oxygen
THC
PP N
Span
Gases
A to D
Conv.
CONDITIONER
Permapure
Dryer
Filter
Pump
Manifold
Figure 5-9. Schematic of Spray Dryer Inlet CEM Equipment.
5-29
-------
TABLE 5-2. CEM EQUIPMENT USED AT SPRAY DRYER INLET
Probe
Sampling line
Conditioner
Zero gas
Span gases
THC analyzer
Oxygen analyzer
Carbon dioxide analyzer
Carbon monoxide analyzer
Sulfur dioxide analyzer
Recorders
A to 0 "converter
Data logger
Printer
Sintered Inconel, SO mm by 500 mm, 3-ym pore size.
Technical Heaters model LP212-5, electrically
heated Teflon tubing operated at 120"C (248°F).
MRI built with Permapure filter/coalescer and
extractive dryer, Teflon diaphragm pump and
capillary flow splitters to each analyzer. All'
parts before dryer are heated to > 120°C
(248°F). Flow controls, sample line blowback and
pressure/vacuum gauges are included. Zero and
span gases are introduced at conditioner inlet.
Prepurified nitrogen from high pressure cylinder,
the nitrogen supply is also used for the dryer
module as the drying gas and to blow back the
sample line.
Detailed list is in Appendix J.
Beckman model 402, heated oven flame ionization
detector. Has built-in zero and span controls.
Horiba model PMA 200, paramagnetic sensor.
Horiba model PIR 2000S, nondispersive infrared.
Horiba model PIR 2000L, nondispersive infrared.
The analyzer inlet has an ascarite/si1ica gel
cartridge to prevent the C02 interference.
Whitacre model P310, electrochemical sensor.
Heath model SR 204 strip chart recorders. Used
for backup to the data logger.
MRI built.system based on Wintek MCS control
modules with 12 bit analog to digital resolution
and RS-232 host computer interface.
Zenith model Z-181 portable PC with GWBASIC
logging program.
Okidata model 182 printer to record all
operations.
logger
5-30
-------
Heated Sample Line
Probe
Logger
Oxygen
Span
Gases
PP N
A to D
Conv.
CONDITIONER
Condensation
Dryer
Filter
Pump
Manifold
Figure 5-10. Schematic of Spray Dryer Outlet CEM Equipment.
5-31
-------
TABLE 5-3. CEM EQUIPMENT USED AT SPRAY DRYER OUTLET
Probe
Sampling line
Conditioner
Zero gas
Span gases
Oxygen analyzer
Carbon dioxide analyzer
A to 0 converter
Permapure model F-500-6Z5-12 sintered stainless
steel, self-cleaning bypass type with 5-um pore
S126 «
Technical Heaters model LP212-5, electrically
heated Teflon tubing operated at 120*C (248°F).
MRI built with Permapure filter, water condensate
trap, Teflon diaphragm pump, and flow controls for
each analyzer. Zero and span gases are introduced
at conditioner inlet.
Prepurified nitrogen from high pressure cylinder.
Detailed list is in Appendix J.
Beckman model 7003, polarographic sensor.
Horiba model PIR 200GS, nondisoersive infrared. -
MRI built system based on Wintek MCS control
modules with 12 bit analog to digital resolution
and RS-232 host computer interface.
Data logger
Computer has built-in printer
Epson model HX-20 portable computer with GWBASIC
logging program.
5-32
-------
Heated Sample Line
Probe
CO,
Rec
Rec
Rec
Rec
Printer
Oxygen
SO.
NOx
Logger
Span
Gases
pp N
A to D
Com/.
CONDITIONER
Permapure
Dryer
Filter
Pump
Manifold
Figure 5-11. Schematic of Baghouse Outlet OEM Equipment.
5-33
-------
TABLE 5-4. CEM EQUIPMENT USED AT BAGHOUSE OUTLET
Probe
Sintered stainless steel, 50 mm by 500 mm, 3-ym
pore size.
Sampling line
Technical Heaters model LP212-5, electrically
heated Teflon tubing operated at 120"C (248°F).
Conditioner
MRI built with Permapure fiIter/coalescer and
extractive dryer, Teflon diaphragm pump, and flow
splitters to each analyzer. All parts before
dryer are heated to > 120°C (248°C). Flow con-
trols, sample line blowback, and pressure/vacuum
gauges are included. Zero and span gases are
introduced at conditioner inlet.
Zero gas
Prepurifled nitrogen from high pressure cylinder,
the nitrogen supply is also used for the dryer
module as the drying gas and to blow back the
-
sample line.
Span gases
Detailed list is in Appendix J.
Oxygen analyzer
Beckman model 742, polarographic sensor.
Carbon dioxide analyzer
Horiba model PIR 2000S, nordispersive infrared.
Nitrogen oxides analyzer
Bendix model 8101-B, chemiluminescence detector.
The conditioning manifold includes a nitrogen
dilution system to maintain a constant dilution
factor for both calibration and sample gases.
Sulfur dioxide analyzer
Whitacre model P310, electrochemical sensor.
Recorders
Heath model SR 204 and Sol tec model 1243 strip
chart recorders. Used for backup to the data
logger.
A to D converter
MRI built system based on Wintek MCS control
modules with 12 bit analog to digital resolution
and RS-232 host computer-interface.
Oata logger
Zenith model Z-181 portable PC with GWBASIC
logging program.
Printer
Okidata model 182 printer to record all logger
operations.
5-34
-------
During- the first sampling run at the baghouse outlet, the diaphragm pump
in the integrated gas sampling train leaked. This leakage was not obvious
during the leak checks before and after the, sampling run, but it was discov-
ered after the Orsat analysis from both the'1 inlet and outlet locations were
compared. The leak was corrected and an additional leak check procedure was
employed. With the probe tip plugged and the train operating at a vacuum
greater than the sampling vacuum, the end of the tube that was connected to
the gas sample bag was submerged in water and.observed for at least 1 mln. If
no bubbling was observed, there was no leak.
5-35
-------
SECTION 6.0
QA/QC
6.1 INTRODUCTION
This section describes activities performed by project personnel as part
of internal quality control (QC) functions, as well as the quality assurance
(QA) audits and reviews that were conducted independently of the project team
by MR I Quality Assurance Coordinators. This section also includes a summary
of field and laboratory technical systems audits conducted by EPA and Research
Triangle Institute (RTI). Copies of the audit reports are included in
Appendix L.
Tests performed at the MERC facility and the subsequent analysis and
reporting of results were performed under the direction of the Project Leader,
Dr. George Scheil. Field tests and' sampling were coordinated by the Field
Sampling Task Leader, Mr. James Surman. Metals and organic analyses were per-
formed under the supervision of Ms. Eileen McClendon and Dr. John Coates,
respectively.
QA activities were performed under the direction of Mr. Dennis Hooton,
Quality Assurance Coordinator (QAC) for the Environmental Systems Depart-
ment. All QA reports and corrective actions were reported to department and
project management and to Ms.'Carol Green, Quality Assurance Manager for MRI.
6.2 SUMMARY OF QC DATA
Summaries and discussions of QC data for the various analyses are
presented below.
6-2.1 Dioxin/Furan Analyses
6.2.1.1 Method Internal Standard Surrogate Recoveries-
Each sample analyzed for dioxins and furans was spiked' with lJC-PCCD/PCDF
method internal standards to determine surrogate recoveries and to quantitate
native PCDD/PCDF compounds.
Recovery values for the method internal standards were consistently low,
ranging from 30% to 70% for-all surrogates in the field samples and audit
samples (spiked XAD and water). Approximately 70^ of all the method internal
standard recoveries for the field samples were within the QAPP objective
(i.e., > 50% recovered). Although these generally low surrogate recovery
values were investigated by the technical staff, no conclusive explanation was
identified other than possibly a procedural loss of compounds (e.g.,' in the
6-1
-------
column cleanup step). This theory is supported by the consistency of the low
recoveries from the extracted samples, the high recoveries in the (nonex-
tracted) instrument performance sample, and. the highly accurate results for
native PCDD/PCDF in the audit samples.
The impact of the low recoveries on sample data appears .minimal because
the calculation of native PCOD/PCDF is not dependent on absolute recovery of
the surrogates. The accuracy of the quantitation method is supported by the
results of the spiked performance audit samples (PAS) which had similar sur-
rogate recovery values {see Appendix L for PAS quantitation reports), but were
well within the QA objective of 50% to 150% accuracy for spiked native PCOD/
PCDF. Performance audit sample results are discussed'in detail later in this
section.
Complete results of the surrogate recovery values are in the quantitation
summary tables presented in Appendix G of this report. A summary of the
reported surrogate recoveries is presented in Table 6-1.
6.2.1.2 GC/MS Calibration Checks for PCDD/PCDF Analyses—
The response factor comparison tables for total and specific tetra-octa
COO/CDF are found in Tables 6-2 and 6-3,.respectively. These tables include
the response factors of the daily calibration standards compared with the
average response factors of the calibration standard curve. This is used to
check the calibration drift of the mass spectrometer. The computer-generated
spreadsheets for the calibration curve and daily standards along with the Mass
Spectrometry Notebook are found in Appendix L.
Variability of response factors were less than +20% RSD for all PCDD/PCDF
compounds during initial calibration; 99% for all continuing calibration
checks were also within the ±20% RSD criteria during analyses of samples.
6.2.1.3 PCDO/PCOF Blank Data-
Three types of blanks were collected from the uncontrolled and controlled
locations: proof blanks, field blanks, and post blanks. Only the controlled
blanks were analyzed.
Trace levels of ccta-CDD, relatively close to the detection limit, were
found in all three of the- blanks. Trace levels of TCDD, also relatively close
to the detection limit, were reported for the proof and post blanks, and were
attributed to possible laboratory contamination. Based on technical review by
project analysts, it was concluded that the amounts of analytes found in the
blanks are relatively low and would not have a significant impact on sample
results.
Complete results for the blank samples are found in Appendix L of this
report. A summary of the blank results is presented in Table 6-4.
6-2
-------
TABLE 6-1. DIOXIN/FURAN SURROGATE RECOVERIES FOR FIELD AND QC SAMPLES FOR MERC MWC
13C-TCDD
l3C-PeCDD
1 3C-HxCDD
13C-HpCDD
'3C-0CDD
13C-TCDF
13C-PeCDF
' 3C-HxCDF
13C-HpC(
Blanks
Proof
46.5
50.3
55.4
48.9
50.3
47.1
50.2
57.8
54.2
Field
46.4
50.3
60.0
55.9
56.9
50.1
55.4
62.5
59.0
Post
52.3
54.0
59.1
55.6
52.9
54.2
56.0
65.0
57.6
EPA audits
130
51.0
58.9
57.6
63.7
67.8
55.6
63.4
60.0
60.8
145
54.8
59.7
59.8
61.2
63.7
53.9
61.5
64.4
59.4
176
44.9
49.8
61.8
63.6
65.5
47.7
52.2
65.8
64.4
QA samples
Performance sample
83.3
97.4
84.7
86.7
90.0
89.7
100.9
87.9
86.6
Spiked filter
48.7
53.7
58.0
60.6
64.1
46.1
42.7
46.9
49.8
Blank filter
46.4
57.6
58.1
61.7
60.1
46.8
56.9
56.8
57.4
Spiked XAD
59.3
63.1
69.2
72.0
72.7
60.3
66.8
70.7
68.1
Blank XAD
55.9
58.4
69.2
66.4
65.1
54.5
65.5
67.7
68.2
Spiked XAD/filter
44.6
54.5
62.2
59.7
58.9
50.7
54.8
64.1
60.0
Blank XAD/filter
42.2
53.1
59.9
62.4
64.0
48.3
56.6
60.1
60.9
Spiked water
30.2
37.0
57.7
50.7
44.3
41.8
44.6
59.2
61.1
Blank water
57.6
59.5
59.6
62.3
66.3
48.5
61.2
61.4
60.8
Uncontrolled
Run 1
FH
46.9
58.9
54.5
61.0
60.6
50.7
58.1
60.3
56.0
BH
43.5
45.5
56.8
52.9
48.4
46.0
49.2
58.1
54.1
Run 2
FH
48.1
57.9
54.7
59.4
52.3
54.3
48.3
57.3
55.3
BH
56.9
53.7
58.7
56.1
52.9
52.2
54.3
61.0
58.1
Run 3
FH
46.1
54.9
51.3
52.2
56.2
46.5
54.7
52.0
51.1
BH
59.4
55.3
56.0
56.1
53.1
48.2
55.9
60.3
57.2
(continued)
-------
TABLE 6-1 (continued)
13C-TCDD i3C-PeCDD i3C-HxCDD isC-HpCDD 13C-0CDD 13C-TCDF i3C-PeCDF i3C-HxCDF i3C-HpCDF
Control led
Run 1 45.8 47.2 52.6 50.1 48.3 47.7 49.9 55.2 51.4
Run 1 dup. 45.1 48.0 50.8 50.1 50.0 44.5 48.2 53.5 50.4
Run 2 39.4 43.7 49.5 47.4 44.8 39.2 47.0 55.7 48.0
Run 3 58.8 56.1 57.9 53.3 52.9 51.2 55.0 59.0 55.0
^C-TCDD is used in calculating total TCDD and 2,3,7,8-TCDD.
13C-PeCDD is used in calculating total PeCDD and 1,2,3,7,8-PeCDD.
i3C-HxCDD is used in calculating total HxCDD, 1,2,3,4,7,8-HxCDD, 1,2,3,6,7,8-HxCDD, and 1,2,3,7,8,9-HxCDD.
•3C-HpCDD is used in calculating total HpCDD and 1,2,3,4,6,7,8-HpCDD.
13C-0CD0 is used in calculating total 0CDD and total 0CDF.
!3C-TCDF is used in calculating total TCDF and 2,3,7,8-TCDF.
13C-PeCDF is used in calculating total PeCDF, 1,2,3,4,8-PeCDF, 1,2,3,7,8-PeCDF, and 2,3,4,7,8-PeCDF.
i3C-HxCDP is used in calculating total HxCDF, 1,2,3,4,7,8-HxCDF, 1,2,3,6,7,8-HxCDF, 1,2,3,4,7,9,-HxCDF,
2,3,4,6,7,8-HxCDF, and 1,2,3,7,8,9-HxCDF.
^C-HpCDF is used in calculating total HpCDF, 1,2,3,4,6,7,8-HpCDF, and 1,2,3,4,7,8,9-HpCDF.
-------
TABLE 6-2. RESPONSE FACTOR COMPARISON "ABLE fQR TOTAL TETRA-OCTA COD/CDF
Response Factor (RRF) a [Area(S)/Area(lS)J * ICone.(I$5/Conc.(S)J
Date analyzed;
1/29/88
L/30/88
1/30/88
1/31/88
1/31/83
3/i8/aa
3/10/88
Standard cone,:
1/23/88
DF100
DF10O
OFlOO
DF1O0
OFlOO
OF 10-3
PFlOq
Data file no,:
Average
Range
Range
A29KQ6
A30XQ2
A3QXQ3
A31XQ2
A3LXq3
C18XQ6
ciaxq?
Area R.I.5 I;
response
% RSD
(8^)
tm)
30476a
508768
955800
S00192
841816
1016270
1084380
Ar§a ft.I.5 2*
factor
-------
TABIE 6-3. RESPONSE FACTOR COMPARISON TABLE FOR SPECIFIC TETRA-OCTA CtJD/COF
Response Factor
-------
TABLE 6-4. SUMMARY OF BLANK TRAIN DATA
Sample
Analytes identified
Amount found (ng)
13012-13016 proof blank
15012-15016 field blank
18012-18016 post blank
Total,TCDD
Octa-CDO
Octa-CDO
Total TCDD
0.027
0.14
O.Ofi
0.024
0.073
6.2.1.4 Precision Results for Duplicate Injections—
Duplicate analysis by replicate injection was performed for the con-
trolled location run 1 MM5 sample. Table 6-5 summarizes-the results including
the average and relative percent difference {RPD). Results are reported in
total nanograms per sample with no blank correction. Precision is calculated
by taking the difference (range) between the analysis results divided by the
average times 100. The computer-generated spreadsheets for these samples are
in Appendix L.
Precision for PCDD/PCDF, reported as totals, was 5% (RPD) or less. Pre-
cision values reported for specific isomers ranged from less than IX (RPD) to
12% (RPD) for compounds detected'above 0.08 ng and 15% (RPD) to 56% (RPD) for
compounds detected below 0.08. ng.
Although no data quality objective was specified for replicate injection
precision, results are quite consistent for analytes detected above 0.08 ng
per compound.
6.2.2 Metals
The.QC checks for metals included the analysis of selected samples in
duplicate, samples spikes and SRM results, monitoring instrument calibration
drift, and the'analysis of the blank train samples.
6.2.2.1 Duplicate Sample Analyses-
Nine of the metals samples (alT from run 1) were selected for duplicate
analysis. The results of the duplicate analyses are shown in Table 6-6.
Detailed data are in Appendix L. The percent difference for all duplicate
analyses were less than 17%, except for the arsenic analysis of. the lime
slurry which was slightly higher at 33% difference.
6.2.2.2 Spiked Sample Analysis and NBS SRM Results—
Two feed samples were selected for spiked sample analysis and. NBS
SRM 1633a, Trace Elements in Coal Fly Ash, was also analyzed. The results are
presented in Table 6-7. Detailed data are in Appendix L.
6-7
-------
TABLE 6-5. PRECISION RESULTS FOR DUPLICATE INJECTION OF MM5 SAMPLE
Analyte
Amount
found
(ng)
Amount
found
(ng)
Average
(ng)
Precision
(RPD)a (%)
Total tetra-COF
3.72
3.65
3.69
2
Total penta-CDF
2.97
3.11
3.04
5
Total hexa-CDF
1.71
1.75
1.73
2
Total hepta-CDF
0.953
0.979
0.97
3
Octa-CDF
0.251
0.259
0.26
3
Total tetra-CDD
0.674
0.683
0.68
1
Total penta-CDD
0.716
0.716
0.72
0
Total hexa-CDO
1.08
1.14
1.11
5
Total hepta-CDQ
0.963
0.968
0.97
1
Octa-CDO
1.06
1.019
1.04
4
2,3,7,8-Tetra-CDF
0.504
0.528
0.52
5
1,2,3,4,8-Penta-CDF
0.0561
0.0662
0.061
17
1,2,3,7,8-Penta-CDF
0.225
0.252
0.24
11
2,3,4,7,8-Penta-CDF
0.228
0.245
0.24
7
1,2,3,4,7,8-Hexa-CDF
0.332
0.317 ¦
0.32
5
1,2,3,6,7,8-Hexa-CDf
0.171
0.176
0.17
3
1,2,3,4,7,9-Hexa-CBF
NO5
ND
-
-
<0.0511
< 0.0498
2,3,4,6,7,8-Hexa-CDF
0.196
0.194
0.19
1
1,2,3,7,8,9-Hexa-CDF
0.0133
0.0233
0.018
56
1,2,3,4,6,7,8-Hepta-CDF
0.604
0.603
0.60
< 1
1,2,3,4,7,8,9-Hepta-CDF
0.111
0.124
0.12
11
Octa-CDF
0,251
0.259
0.26
3
2,3,7,8-TCD0
0.0465
0.0464
0.046
< 1
1,2,3,7,8-Penta-CDD
0.0725
0.0683
0.070
6
1,2,3,4,7,8-Hexa-CDD
0.0690
0.0806
0.075
15
1,2,3,6,7,8-Hexa-CDO
0.152
0.135
0.14
12
1,2,3,7,8,9-Hexa-CDD
0.119
0.129
0.12
8
1,2,3,4,6,7,8-Hepta-CDO
0.510
0.508
0.51
< 1
Octa-CDO
1.056
1.02
1.04
3
GC/MS data file:
8910A31X4
8910A31X5
Sample ID: ,1028-1032
(MM5 outlet)
1028-1032
(MM5 outlet Rep.
inj.)
Analyis date:
1/31/88
1/31/88
a RPD (relative percent difference) = 100(range t mean),
k ND = Not detected.
6-8
-------
TABLE 6-6. DUPLICATE METALS ANALYSIS RESULTS
Arsenic Cadmium Chromlura Mercury Lead
final final final final final
Sample Units . results results results results ' results
Run I
Cyclone ash
Dup
% di f ference
Run I
11me SIurry
Dup
i di fference
ug^a
ug/g
ug/g
yg/g
30.3
34.1
12.0
4.95
3.55
33,1
50.5
JO.8
1,00
< 0.229
< 0.224
m
590
37?
3.44
< 0.955
c 0.931
NA
< 7.96
< 22.6
m
< 0.225
< 0.223
NA
2,168
2,028
6.69
< 5.85
< 5.70
NA
Run 1
inlet acetone yg total
r i nse
Duo Wg total
% difference
Run 1
Outlet ace+one yg total
r inse
Dup yg total
I difference
Run
Inlet front half yg total
Dup yg'total
I difference
Run 1
Outlet front half yg total
Dup . yg tataI
$ di f f erence
Run 1
Inlet back half yg total
Dup yg total
.* d" f ference
Run 1
Inlet permanganate yg/g
Dup yg/g
i difference
94.5
94.5
0.00
< 2.58
< 2.58
NA
799
751
6.29
< 9.03'
< 9.03
NA
I 2 I
158
• 205
3,943
.52
NA
NA
NA
NA
NA
122
469
203
3,961
0.46
2.45
I..03
0.45
< 0.750
4.29
< 3.16
15.4
< 0.750
3.69
< 3.19
15.9
NA
15.0
NA
2.94
770
3,824
726
44,417
716
3,897
681
44,472
3.12
1 .39
6.29
0.12
12.7
< 4.47
< 9.55
204
11 .7
< 4.47
< 10.3
203
8.37
NA
NA
0.57
4.27
2.04
NA
14.7
4.20
1 .72
NA
14.8
1 .69
16,93
NA
0.47
NA
NA
0.00121
NA
NA
NA
0.00134
NA
NA
NA
10.13
NA
Run 1
Outlet permanganate •yg/g
Dup yg/g
J difference
NA
NA
NA
NA
NA
NA
NA
NA
NA
0.000769
0.000758
1 .37
NA
NA
NA
NA « riot applicable or not analyzed.
-------
TABLE 6-7. ASH AND LIME SLURRY METALS SPIKE AND RECOVERY DATA
Arsenic
Cadmium
Metal
Chromium
Mercury
Lead
Baghouse ash (yg/g)
With spike (yg/g)
Spike level (yg/g)
Recovery (%)
Lime slurry (yg/g)
With spike (pg/g)
Spike level (yg/g)
Recovery (%)
NBS SRM 1633a (yg/g)
Certified value (yg/g)
Recovery (%)
Low reference spike (yg total)
Spike level (yg total)
Recovery (%)
High reference spike (yg total)
Spike level (yg total)
Recovery (%)
47.6
134
98.6
88.0
4.95
3.55
16.1
9.48
125
127
145
87.5
5.83
4.90
119
59.1
50.0
118
129
227
98.
99.
a
a
8.98
9.48
94.8
a
1.00
NA
5.07
4.90
104
52.4
50.0
105
152
185
98.6
33.4
a
a
5.97
9.48
63.0
223
196
114
4.61
4.90
94.0
54.5
50.0
109
80.2
244
197
82.9
a
a
19.0
19.0
100
a
0.16
NA
9.60
9.80
98.0
97.3
100
97.3
2,767
6,011
3,254
99.7
a
a
19.8
28.4
69.5
124
72.4
172
17.0
14.7
116
1,836
1,650
111
a = Below detection limits,
b = NA - Not applicable.
-------
6.2.2.3 Instrument Calibrations-
Calibrations for the three analyzers are summarized in Table 6-8.
Detailed data are 1n Appendix L. AT1 calibration checks met the requirements
given in the project QA Plan.
6.2.2.4 Blank Train Analyses—
The blank train sample analyses are shown in Table 6-9. The posttest
blank showed that some sample carryover occurred. Mercury was the only ele-
ment consistently measured in the blanks. A blank correction of 8 u9 was
applied for mercury. Detailed data are in Appendix L.
6.2.3 Sampling Equipment
A summary of equipment calibration results is presented in Table 6-10.
Acceptable ranges for the calibrations are included. All of the equipment
fell within acceptable limits.
Isokinetic performance and leak check results are presented for senivcla-
tile data (Table 6-11) and particulate/metals data (Table 6-12). All of the
test data fell within isokinetic limits of 100 ± 10%. 'Leak checks were
acceptable in all cases except one. The final leak check, the run 3 inlet
semivolatiles train, showed such a high leak rate that no vacuum could be
drawn. It is believed that this large leak was caused when the probe nozzle
was severely jarred upon being removed from the stack. No indications of a
leak of such magnitude were observed during 'test operations.
6.2.4 Continuous Emission Monitoring
The QA/QC checks for the MR I CEMs included dally leak checks, zero, and
span drift measurements, and comparison of the working standards against EPA
protocol no. 1 gas cylinders.
6.2.4.1 Leak Checks—
The MR I gas analyzers were operated as three separate CEM systems. Leak
checks of each of the three monitoring systems were done before and after
every run. The leak checks consisted of sealing the sampling probe, producing
a vacuum equal to the highest, observed vacuum on the system between the probe
and the sampling pump, sealing off the pump, and checking for a leakage rate
of < 4% of normal flow. All leak checks were completed successfully. Leak
check results are noted on the data logger printouts shown in Appendix F.
6.2.4.2 Calibration Drift—
The working standards are shown in Table 6-13. The span gases were
certified to at least ±2% by the manufacturers. Each analyzer was calibrated
twice daily with zero and span gases. Table 6-14 shows the dally zero drift,
and Table 6-15 shows the daily span drift for each analyzer. Only the S02
analyzer used at the inlet failed to meet the unofficial daily drift criterion
of < 10%. A careful review of the data indicated that the span drift was
sudden rather than gradual, that it usually occurred during a port change, and
it was clearly identifiable by a sudden drop in the measured concentration.
The probable cause was poisoning of the electrochemical sensor. Therefore,
only data before the first sensitivity loss were reported, and only the ini-
tial calibration values were used in calculating concentrations for this
analyzer.
6-11
-------
TABLE 6-8. METALS INSTRUMENT CHECK STANDARO DATA AND PERCENT DRIFT CALCULATIONS
.•CAP run
Element $ Element
Cd drift Cr
drift
Element
Hi'
drift
Element %
Pb dri ft
3/23/88
initial ICS
ICS1
ICS2
ICS3
3/23/88
initial ICS
ICS1
ICS2
ICS3
4/12/88
i ni T i a I 1CS
ICS1
ICS2
6/7/88
5,1069
5.106
5.1484
5.1555
5.1287
5.2551
5.3029
5.3222
5.0635
5.0869
5.1201
0.02
0.81
0.95
2.46
3.340
3,77
0.46
1.12
5.0202
4.9996
5.0565
5.0662
5.0635
5.2280
5.2559
5.2256
50.424
5.0437
5.1225
0.41
0.72
0.92
3.25
3.80
3.20
0.03
1 .59
5.102
5.1285
5.3784
5.3894
5.1273
5.3269
5.402!
5.4371
5.0489
5.0892
5.0146
0.52
5.42
5.63
3.89
5.36
6.04
0.80
0.68
5.1591
5.1897
5.3095
5.3963
5.0773
5.2546
5.2510
5.2624
5.0269
5.0338
5.08!1
0.59
2.92
4.60
3.49
3,42
3.65
0.14
1.08
Initial ICS
• ICS1
ICS2
5.0223
4.9558
5.0530
1 .32
0.61
5.0368
4.9838
5.0637
NA
1 .05 NA
0.53 NA
NA
NA
5.0824
5.2699
5.4326
3.69
6.89
GFAA run
Absorbance
dr i f t
Hg
CVAAS run
Absorbance
¦%
dri ft
3/31/88 Pb
Initial ICS
ICS1
ICS2
1CS3
ICS4
4/1/88 As
Initial ICS
ICS!
1CS2
1CS3
ICS4
J CSS
4/4/88 As
Initial ICS
1CS1
CS2
CS3
CS4
CS5
CS6
0.119
0.119
0.120
0.121
0.123
0.197
0.208
0.215
0.205
0.211
0.212
0.211
0.214
0.231
0.233
0.235
0.225
0.234
0,00
0.84
1 .68
3.36
5.58
9.14
4.06
7.11
7.61
1.42
9.48
10.43
11.37
6.64
10.90
1/7/88
Initial ICS
1CS1
ICS2
ICS3
ICS4
4/9/88
Iniital ICS
ICSl
iCS2
4/13/88
In i i taI I CS
ICSl
ICS2
4/14/B8
Initial i CS
ICSl
ICS2
0. 144
0.142
0.141
0.144
0.144
0.244
0.262
0.269
0.236
0.236
0.239
0.223
0.226
0.228
1.39
2.08
0.00
0.00
7.38
10.25
0.00
1.27
1.35
2.24
(Continued)
6-12
-------
TABLE 6-8 (Concluded)
GFAA run
Absorbanca
drift
Hg
IVAAS run
Absorbance
¦it
dr i f t
4/8/88 Pb
6/7/88
Initial ICS
ICS)
ICS2
0.147
0,135
0,' 25
8.16
14.97
Initial iCS
ICS!
ICS2
I C$3
0.256
0.256
0.271
0.273
0.00
5,06
5.64
6/6/88 As
Initial ICS
ICSt
ICS2
ICS3
0.199
0.204
0.215
0.22?
2.51
8.04
14.07
6/6/08 Pb
Initial lCS
ICS I
1CS2
0.229
0.231
0.232
0.87
1.31
Mote: These analyses were conducted on a Jarrell-Ash Model 1 155 A iCP-AES, a Perkin-Elmer
Model 5000 Zeewan atomic absorption spectrometer and a Perkiri-Elmer 30309 atomic absorption
spectrometer, The atonic abosrption units were equipped with a hollow cathode lamp for PB or
el'ectrodeless discharge (amps for Hg or As.
6-13
-------
TABLE 6-9. METALS BLANK TRAIN ANALYSES
As Cd Cr Pb Hg
(ug) (vg) (wg) (vg)
Proof blank
Acetone rinse
< 1.29a
, < 0.375
< 0.638
0.79
< 1.52
Front half
< 3.87
1.14
< 1.91
1.68
5.76
Back half
< 0.14
.0.439
< 1.11
0.671
0.0063
Permanganate < 0.00049
Total < 5.3b < 2„Qb < 3.7b 3.1 5.8
Stack blank
Acetone rinse < 1.29 < 0.375 < 0.638 < 0.44 < 1.55
Front half < 7.74 < 2.25 < 3.83 < 2.64 9.23
Back half < 0.141 0.329 < 1.12 3.82 < 0.0052
Permanganate, < 0.00049
Total < 9.2b < 3.0b < 5.Sb - < 6.9b 10.8b
Posttest blank
Acetone- rinse <1.29 <0.375 <0.538 2.64 1.78
Front half < 6.45 3.42 < 3.19 17.8 8.64
Back-half < 0.143 < 0.399 < 1.13 4.32 < 0.0066
Permanganate ' < 0.0005
Total < 7.9b < 4.2b < 5.0b 24.8 10.4
The metal.values shown to right of < denote detection limit of the
analysis.
Totals include detection limit.
6-14
-------
TABLE 6-10. CALIBRATION RESULTS FOR SAMPLING EQUIPMENT
Parameter
Acceptance limit
Pass/fai1
Probe nozzle
3 measurements within
0.1 mm
All pass
Gas meter volume4 (Y-factor)
Post-test ±5% of pretest
All pass
Gas meter temperature
±5°F
All pass
Stack temperature sensor
±1.5%
All pass
Final impinger temperature sensor
±5°F
All pass
Filter temperature sensor
±5"F
All pass
Aneroid barometer
±2.5 mm Hg
Pass
S-type pi tot tube
Method 2 criteria
All pass
a Actual values as follows: 1
Pretest
Posttest
Console 10
Console 3
Console 9
Console 6
1.127
1.0158
1.0229
0.9870
1.1381
1.0182
1.0153
0.9813
6-15
-------
TABLE 6-11. SEMIV0LAT1LES TESTING ISOKINETICS3 AND LEAK CHECKb SUMMARY
% samp I i ng Lssk rs*!"© Pr ss su r©
Date Run No. isokinetic Leak check port cm^/min (ft"Vitifn) mm NjO (in. HjO)
12/9/37 1-1nlet 99.2
1-Out let 100.1
12/10/87 2-lnlet 99.4
2-Outlet 100.3
52/12/87 3-Inlet 100.9
3-Out I e-t 104.8
Initial
2
84.9
Port change
2
28.3
Continue
3
141
F i na 1
3
28.3
Initial
2
340
Port change
2
141
Continue
3
141
Final
3
28.3
Initial¦
2
226
Port change
2
113
Con"i nue
3
34.9
Port change
3
26.3
Con* i nue
I
141
F i na 1
1
56.6
Initial
2
84.9
Rep 1 ace XAD
2
84.9
Conti nue
3
54.9
Port change
2
141
Continue
2
84.9
Port change
3
84.9
Conti nue
1
84.9
Final
1
28.3
Initial
2
141
Port change
2
170
Conti nue
3
198
Port change
3
28.3
Conti nue
1
84.9
Final
1
141
Initial
2
28.3
Port change
2
28.3
Continue
3
28.3
Port change
3
28.3
Conti nue
1
28.3
Final
1
(0.003)
381'
(15)
(0.001)
127
(5)
(0.005)
381
(15)
(0.001)
178
(7)
(0.012)
381
(15)
(0.005)
457
(18)
(0.005)
381
(15)
(0.001)
457
(18!
(0.008)
381
(15)
(0.004)
152
(6)
(0.003)
381
(15)
(0.001)
178
(7)
(0.005)
381
(15)
(0.002)
228
(9)
(0.0035
431
(17)
(0.003)
609
(24)
(0.003)
381
(15)
(0.005)
381
(15)
(0.003)
533
(21) ¦
(0.003)
508
(20)
(0,003)
381
(15)
(0.001)
533
(21)
(0.005)
381
(15)
(0.006)
i 27
(5)
(0.007)
381
(15)
(0.001)
152
(6)
(0.003)
381
(15)
(0.005)
178
(7)
(0.001)
381
(15)
(0.001)
178
(7)
(0.001)
381
(15)
(0.001)
178
(7)
£0.001)
381
(15)
c
Unatta I li-
able
a The QC objective for isokinetics was 100 + 101.
b The QC objective for leak checks was a leak-free train or a leakage rate less than or equal to 0,02
cfm, or iess than 4l of the average sampling rate (whichever is ess).
c Nozzie tip severely jarred upon removal from stack. Unable to draw vacuum for final leak check.
Sampie line believed to be leak free until withdraw) from stack.
6-16
-------
TABLE 6-12. PARTICULATE/METALS TESTING ISOKINETICS3 AND LEAK CHECK0 SUMMARY
I
Samp Ii ng
Leak
rate
Pressure
Date Run No. isokinetic
Leak check
port
cm^/m i n
S f tVmi n)
mm HjO (
i n. i-i^O*
12/9/87 1-lnlet 99.2
Initial
1
113
(0.004)
381
(15)
Port change
1
56.6
(0.002)
152
(6)
Cqnt i nue
2
84.9
(0.003)
381
(15)
Final
2
84.9
(0.003)
178
(7)
i-Outlet 106.1
Initial
1
56.6
(0.002)
381
(15)
Port change
1
0.000
(0.000)
381
(15!
ContInue
2
84.9
£0.003)
381
(15)
Final
2
28.3-
(0.001)
381
(15)
12/10/8? 2-lnlet 100.!
Initial
1
170
(0.006)
381
(15)
Port change
1
56.6
(0.002)
101
(4)
Cont i nue
2
170
(0.006)
331
<15)
Port change
2
56.6
(0.002)
127
(5)
Cont i nue
3
1 41
(0.005)
381
(15)
F ina 1
3
56.6
(0.002)
152
. (6)
2-Outlet 105.8
Initial
l
113
(0.004)
381
(15)
Port change
1
28.3
(0.001)
127
(5)
Cont in lie
2
84.9
(0.003)
381
(15)
Port change
2
0.000.
(0.000)
152
(6)
Cont1nue
3
113
(0.004)
381
(15)
Final
3
28.3
(0.00!s
127
(5)
12/12/87 3-1nlet 101,7
Initial
1
84.9
SO.003)
381
(15>
Port change
1
56.6
(0.002)
152
(6)
Cont i nue
2
84.9
(0.003)
381
(15)
Port change
2
56.6
(0.002)
178
(7)
ContInue
3
84,9 .
(0.Q03)
381
(15)
Final
3
56.6
(0.002)
178
(7)
3-Out let 104.7
initial
1
170
(0.006)
381
(15)
Port change
1
28.3
(0.001)
127
(5)
Continue
2
198
(0.007)
361
(15)
Port change
2
28.3
(0.001)
132
(6)
Continue
3
I 13
<0.004)
381
(15)
Final
3
198
(0.007)
152
(6)
3 The OC objective for isokinetics
was 100 + .108.
k The QC ob ject i ve for leak checks
was a leak-?
ree train or
a leakage
rate less
than or equal
to 0.02
cfm, or less than 4f of the average sampling
rats (whichever is less
) -
6-17
-------
TABLE 6-13. CALIBRATION GASES
Gas mixture
Supplier
Grade
Analyzer
Zero gases
Span gases
Nitrogen
Aira
106.1 ppm NO In nitrogen
93.1 ppm SO2 in nitrogen
413 ppm SO2 In nitrogen
10.2 ppm propane in HC1 air
14.01% oxygen, 12.00% C02,
296 ppm CO .in nitrogen
12.09% oxygen, 12.00% C02,
2,924 ppm CO
14.01% 02 in nitrogen
Airco
Matheson
Scott
Matheson
Scott
Scott
Scott
Scott
Scott
Prepurified
Zero
1%,
protocol 1
2%
1%,
protocol 1
1%, ¦
protocol 1
All but THC
THC
N0X
SO 2
SO 2
THC
C02, CO (R1+R2)
02, C02 (R3)b
CO ,
02 (R1+R2)
System is purged with zero air, calibration gas flow is then turned off for
zero reading. This zero point agrees with the zero measured from ultra high
purity gases.
b R1+R2 = calibration during run 1 and run 2.
R3 = calibration during run 3.
6-18
-------
TABLE 6-14. DAILY ZERO DRIFT IN CEMS
Run 1
Run 2
Run 3
Initial Final % drift
Initial Final i drift
Initial Final
% drift
Oxygen:
Dryer inlet
Dryer outlet
Baghouse outlet
2.79
0.42
0.40
0.0
0.59
0.32
4.9
0.3
0.1
1.99
0.58
0.28
2.51
0.31
0.22
0.8
0.6
0.1
1.64
0.64
0.92
1.95
0.53
0.32
0.7
0.2
1.3
Carbon dioxide:
Dryer inlet
Dryer outlet
Baghouse outlet
1.63
1.21
1.09
1.53
1.70
1.58
0.2
0.7
0.7
1.93
1.60
0.91
1.92
1.64
1.56
0.0
0.1
0.8
1.97
1.14
0.98
1.86
1.60
1.08
0.2
0.7
0.2
Carbon monoxide:
Dryer inlet
2.30
0.28
4.0
2.30
2.11
0.4
2.54
1.35
, 2.4
Total hydrocarbons:
Dryer inlet
-
-
-
_
-
_
10.29
11.14
1.0
Sulfur dioxide:
Dryer inlet
Baghouse outlet
3.10
2.46
3.71
2.38
1.1
0.1
4.35
2.68
5.29
2.57
3.1
0.2
8.10
2.44
8.37
2.31
0.5
0.2
Nitrogen oxides:
Baghouse outlet
1.53
1.15
0.8
1.61
1.27
0.7
1.94
1.20
1.6
-------
TABLE 6-15. DAILY SPAN DRIFT IN CEMS
Run 1
Run 2
Run 3
Initial
Final
%
drift
Span
conc.
Initial
Final
%
drift
Span
conc.
Initial
Final
%
drift
Span
conc.
Oxygen:
Dryer inlet
Dryer outlet
Baghouse outlet
58.79
54.79
54.91
57.52
54.48
52.49
2.7
0.9
4.4
14.0
14.0
14.0
59.21
55.31
54.70
59.23
54.11
52.02
0.9
1.7
4.9
14.0
14.0
14.0
51.05
48.27
46.08
50.82
46.63
40.72
1.1
3.3
11.2
12.1
12.1
12.1
Carbon dioxide:
Dryer inlet
Dryer outlet
Baghouse outlet
82.05
77.92
82.70
80.99
77.93
82.75
1.2
0.6
0.5
12.0
12.0
12.0
81.63
77.85
82.95
81.73
77.95
82.94
0.1
0.1
0.8
12.0
12.0
12.0
80.87
77.74
82.00
80.68
77.72
81.98
0.1
0.6
0.1
12.0
12.0
12.0
Carbon monoxide:
Dryer inlet
52.09
50.19
0.2
336
51.92
48.62
6.5
336
49.90
52.55
7.8
338
Total hydrocarbon:
Dryer inlet
-
-
-
-
-
- ,
-
-
86.71
100.61
15.8
10.2
Sulfur dioxide:
Dryer inlet
Baghouse outlet
68.90
86.32
46.83a
82.54
41.6
4.5
413
93.1
50.10
83.26
21.35a
82.62
96.1
0.6
413
93.1
86.75
81.61
24.33a
78.15
132.5
4.3
413
93.1
Nitrogen oxides:
Baghouse outlet
48.25
48.90
2.2
106.1
49.44
49.57
1.0
106.1
48.55
49.04
2.6
106.1
a Final SO2 calibration not used to calculate sample concentrations. See text.
-------
6.2.4.3 Calibration Cylinder Check for Continuous Monitoring of Combustion
Gases-
Span cylinder accuracy checks were performed for CO, C02, S0a, 02l and
THC. Results are presented in Table 6-16.
Accuracy values ranged from 91% to 106% for all gas measurements, well
within the objective of BS% to 115% recovery.
6-2.5 Process Samples
Duplicate sample analysis results (performed by Galbraith Laboratories)
are presented in Table 6-17. Samples from run 1 were used for duplicate
analysis, which followed the same methodology as normal sample analysis. Pre-
cision results for the duplicate analyses were all 3% (RPD) or better, except
for percent carbon analysis of the cyclone ash and the bottom ash, which
varied by 12% (RPD) and 42% (RPD), respectively.
6.3 AUDITS
Several independent audits were conducted during this project. These
included analysis of performance audit samples, systems audits, and audits for
data quality. These independent audits are summarized below.
6.3.1 Performance Audit Samples
Performance audit samples (PAS) were prepared by the QAC using standard
solutions independent of and separate from project calibration standards-
Actual amounts or concentrations of the PAS were not disclosed to the analysts
until the results of analysis were reported in writing to the QAC. The QAC
calculated accuracy results and reported these to project and department man-
agement and to the QAM.
Audit samples for PCDD/PCDF and metals were also provided by EPA and
analyzed together with the field samples.
Two types of PCDD/PCDF audit samples were processed along with the MERC
samples; (1) spikes and blanks, and (2) instrument performance. Clean XAD
and water matrices were spiked by the QAC with known amounts of PCDD/PCDF and
extracted by the analyst in the same manner as the samples. Blanks were run
to check for contamination of sampling materials. Eight spikes and blanks
were extracted and analyzed: (a) XAD, (b) filter, (c) water, and (d) XAD/
filter.
The instrument performance sample,is a known amount of a spiked solution
with method and recovery internal standards added. It is given to the mass
spectrometer operator by the QAC. This sample is used to independently verify
that the GC/MS is operating properly while the spikes and blanks verify that
the extraction procedure is adequate. A summary of the spikes and instrument
performance results -for this project are listed in Tables 6-18 through 6-21.
6-21
-------
TABLE 6-16. SPAN CYLINDER ACCURACY CHECKS
Test mixture
Analyzer
Measured
value
Accuracy
(*)
None
NQX
No second
cylinder
available
9.5 ppm propane
THC
9.7
102
HBS 1666b
93.1 ppm SO2
SO2 inlet
87.5
94
Standard for outlet
14.01% 0 2
02 inlet
Dryer outlet
Baghouse
outlet
14.00
12.75
14.17
100
91
101
BAL 936 EPA P 1
12.4% CO2
CO2 dryer
outlet
Dryer outlet
Baghouse
cutlet
11.9
12.06
12.38
96
97
100
BAL 3172 EPA P 1
409.7 ppm CO
CO
434.6
106
BAL 102 EPA P 1
412.7 ppm S02
SO2 dryer
outlet
Baghouse
outlet
381
400.8
92
97
SAL 1907 EPA P 1
4.01% 02
02 dryer
outlet
Dryer outlet
Baghouse
outlet
3.99
3.70
4.16
99
92
104
Standard used for
runs 1 and 2
12.0% C02
CO2 dryer
outlet
Dryer outlet
Baghouse
outlet
12.11
11.99
11.98
101
100
100
-
295 ppm CO
CO
287.4
97
Span gases used (same as used for Run 3}
02 12.09% 02
CO2 12.00% CO2
CO 297.4 ppm CO (338 ppm corrected for C02)
THC 10.2 ppm propane
SO2 93.1 ppm S02 at baghouse outlet
413 ppm SO2 at dryer inlet
N0X 106.1 ppm NO
6-22
-------
TABLE 5-17. PRECISION FOR DUPLICATE ANALYSES OF PROCESS SAMPLES4
_____
% Carbon % Ash % CaO % Solids gravity
Fabric filter ash
Cyclone ash
Bottom ash
5.57
88.96
5.40
89.26
X =
5,49
x =
89.11
RPO
3%
RPO
1*
1.12
98.30
0.99
98.38
x =
1.06
x =
98.34
RPO
12%
RPD
0.1%
1.58
74.48
1.03
75.86
x =
1.30
x =
75.17
RPO
42%
RPD
2%
Lime slurry - - 11.76 20.60 1.131
11.91 _ 20.65 _ 1.130
x = 11.84 x = 20.53 x = 1.13
RPD 1% RPD 0.2% RPO 0.1%
a Run 1 samples used for duplicate analysis.
b Precision is expressed as range percent deviation (RPD): -x 100
6-23
-------
TABLE 6-18. DIOXIN/FURAN RESULTS FOR THE INSTRUMENT PERFORMANCE SAMPLE
Analyte Found (ng)
Theoretical (ng)
Accuracy (%)
2,3 s 7,8-TCD0 68.7
67.8
101
TABLE 6-19. DIOXIN/FURAN
RESULTS FOR BLANK QA PERFORMANCE SAMPLES®
Sample
Analytes detected
Amount (ng)
XAD (04512)
Octa-CDD
0.032
Filter (04517)
none
-
XAD and filter (04519)
Octa-CDD
0.027
Water (04515)
Octa-CDD.
0.010
a Detection limits for dioxins and furans ranged from 0.002 to 0.04 ng.
6-24
-------
14)
cun
(SO
91
82
77
72
95
91
87
70
71
83
TABLE 6-20. TOTAL DIOXIN/FURAN RESULTS FOR SPIKED QA PERFORMANCE SAMPLES3
Spiked XAD + f iIter
Spiked XAD (045.13) Spiked filter (04516) (04518) Spiked water
Found Theor. Accuracy Found Theor. Accuracy Found Theor. Accuracy Found Theor.
(ng) (ng) (2) (ng) (ng) (|) (ng) (ng) (|) (ng) (ng)
1 .043
1 .0
104
0.391
0.40
98
0.362
0.40
91
0.362
0.40
1.959
2.0
90
0.74
0.80
93
0.663
O.flO
83
0.655
0.80
9.420
10.0
94
3.98
4.0
100
3.287
4.0
82
3.069
4.0
5.0071
5.0
100
1.83
2.0
92
1.676
2.0
84
1.435
2.0
4.554
5.0
91
1.53
2.0
77
1.688
2.0
84
1 .901
2.0
0.9790
1 .0
98
0.361
0.40
90
0.377
0.46
94
0.365
0.40
0.965
1.0
97
0.352
0.40
88
0.344
0.46
86
0.347
0.40
7.016
7.5
94
2.66
3.0
89
2.507
3.0
84
2.104
3.0
2.010
2.5
80
0.740
1.0
74
0.754
1.0
75
0.705
1.0
4.694
5.0
94
1.69
2.0
85
1.735
2.0
87
1.661
2.0
the data quality objective for accuracy of 50-i50!(.
-------
TABLE 6-21. ISOMER SPECIFIC DlOXIN/FURAN RESULTS FOR SPIKED QA PERFORMANCE SAMPLES
Ana 1yte
Spiked XAD (04513)
Sp i ked f i1ter
(04516)
Spiked XAD +
(04518)
f i1ter
Spiked water (04514)
Found
(ng)
Theor.
(ng)
Accuracy
(?)
Found
(ng)
Theor.
(ng)
Accuracy
(1)
Found
(ng)
Theor.
(ng)
Accuracy
(1)
Found
(ng)
Theor.
(ng)
Accuracy
(1)
2,3,7,8-Tetra-CDF
1 .
.0331
1 .0
103
0.391
0.40
98
0.363
0.40
91
0.362
0,
.40
91
1,2,3,7,8-Penta-CDF
1 .
.131
1.0
113
0.430
0.40
108
0.358
0.40
90
0.362
0.
.40
91
2,3,4,7,8-Penta-CDF
0,
.938
1.0
94
0.396
0.40
99
0.3503
0.40
88
0.339
0,
.40
85
1,2,3,4,7,8-Hexa-CDF
2,
.530
2.5
101
0.929
1 .0
93
0.869
1 .0
87
0.961
1,
.0
96
1,2 ,3,6,7,8-Hexa-CDF
2.
.159
2.5
86
1 .001
1.0
100
0.8097
1 .0
81
0.797
1,
.0
80
2,3,4,6,7,8-Hexa-CDF
2,
.356
2.5
94
1 .024
1.0
102
0.812
1 .0
81
0.778.
1,
.0
78
1,2,3,7,8,9-Hexa-CDF
2.
.174
2.5
87
0.984
1.0
98
0.714
1 .0
71
0.425
1.
.0
43a
1,2,3,4,6,7,8-Hepta-CDF
2,
.908
2.5
116
1 .120
1 .0
112
0.971
1 .0
97
0.900
1,
.0
90
1,2,3,4,7,8,9-Hepfa-CDF
2,
.387
2.5
95
0.803
1 .0
80
0.801
1.0
80
0.608
1,
.0
61
Octa-CDF
4.
.554
5.0
91
1 .53
2.0
77
1 .701
2.0
85
1.901
2,
.0
95
2,3,7,8-TCDD
0,
,979
1 .0
98
0.361
0.40
90
0.377
0.40
94
0.364
0.
.40
91
1,2,3,7,8-Pen fa-CDD
0.
.965
1 .0
97
0.352
0.40
88
0.344
0.40
86
0.347
0,
.40
87
1 ,2,3,4 ,7,8-Hexa-CDD
3,
.276
2.5
131
1 .088
1.0
109
0.972
1 .0
97
1.073
1 ,
.0
107
1 ,2,3,6,7,8-Hexa-CDD
3,
.368
2.5
135
1 .37
1.0
137
1 .464
1 .0
146
1.163
1 ,
.0
116
1 ,2,3,7,8,9-Hexa-CDD
2
.425
2.5
97
0.988
1 .0
99
0.8705
1 .0
87
0.565
1,
.0
57
1,2,3,4,6,7,8-Hepta-CDD
2.
.010
2.5
80
0.723
1 .0
72
0.754
1 .0
75
0.705
1 ,
.0
71
Oc1a-CDD
4,
.694
5.0
94
1 .69
2.0
85
1 .735
2.0
87
1.661
2,
.0
83
a This value did not meet the data quality objective for accuracy of 50-1501.
-------
All but one of the target analytes were within the data quality objective
of 50% to 150% accuracy. Therefore, the overall completeness of the accuracy
determinations was essentially 100%.
6.3.2 EPA Audit Samples ,
6.3.2.1 PCOD/PCDF Audit Samples--
Three samples were submitted by EPA as XAD resin in wide-mouth jars. The
entire contents of each sample were extracted and analyzed for PCDO/PCDF.
Results of the analyses and summaries of EPA evaluations are shown in
Table 6-22. The EPA results are provided in Table 6-23. The complete EPA
report of the audit results and calculations is Included in Appendix L.
5.3.2.2 Metals Audit Samples-
Four audit samples for metals were supplied by EPA. The analysis results
of these samples are shown in Table 6-24. Only sample 102 for chromium and
lead showed results beyond the desired 90% to 110% accuracy limits. The lead
result improved to within the acceptance limits after realigning the instru-
ment. The chromium result did not change when the standard addition method
was used. After further discussion with,the QAC, a separate EPA reference
sample, WP283, was added as a further check for chromium. Results for the EPA
WP283 sample were consistently within criteria, indicating that accuracy of
the results are acceptable.
6.3.3 Systems Audits
6.3.3.1 Laboratory Systems Audit —
A laboratory systems audit on January 25 and 27, 19B8, was performed
concurrently by Mr. Joseph Evans and Or. S'nri Kulkarni of Research Triangle
Institute, and MRI's project QAC. The technical systems audit consisted of
reviews of procedural documents,- discussions with laboratory personnel, and
inspection of laboratory facilities and equipment maintenance records. In
addition, sample handling procedures and custody records were inspected by the
auditors to assure sample integrity.
No major problems were identified during this audit. The RTI auditors,
inconcurrence with MR I, did make two specific recommendations for documenting
actual laboratory practices and to improve QC records. These recommendations
were: (1) that MRI's final report reflect that the metals analyses of samples
from the MM5 sampling train were digested according to the "Radian Draft
Method" and analyzed by the SW-846 Method 6010, and (2) that criteria for QC
checks of analytical balances be established. RTI's laboratory audit report
is included in Appendix L.
6.3.3.2 Field' Technical Systems Audit-
In addition to the technical systems audit of the laboratory performed by
RTI, a check list for field sampling and facility operations was also devel-
oped by RTI to assist Dr. Brna (AEERL Project Officer) and Mr. Riley {0AQPS
Project Officer) in the evaluation of project procedures during sampling
activities.
6-2?
-------
TABLE 6-22. EPA AUDIT SAMPLES PCDD/PCOF RESULTS (ng/sample)
Isomer
F-I76a
F-145
. F-130
2,3,7,8-TCDO
0.008
0.849
0.547
Other TCDD
0.008
8.71
16.4
1,2,3,7,8-PeCDD
0.010
1.85
3.36
Other PeCDO
0.010
.15.3
27.4
1,2,3,4,7,8-HxCDD
0.023
3.97
4.57
1,2,3,6,7,8-HxCDD
0.031
3.54
6.45
1,2,3,7,8,9-HxCDD
0.015
4.88
8.46
Other HxCDO
0,009
• 16.5
28.0
1,2,3,4,6,7,8-HpCDD
0.012
13.2
21.3
Other HpCOD
0.012
11.3
19.3
Octa-CDD
0.023
22.4
33.6
2,3,7,8-TCDF
0.004
4.31
6.28
Other TCDF
0.004
20.8
34.8
1,2,3,4,8-PeCDF
0.009
0.789
1.30
1,2,3,7,8-PeCDF
0.011
1.96
3.26
2,3,4,7,8-PeCDF
0.009
2.24
3.91
Other PeCDF
0.009
25.0
39.6
1,2,3,4,7,8-HxCDF
0.005
5.83
11.0
1,2,3,6,7,8-HxCDF
0.005
2.39
4.48
1,2,3,4,7,9-HxCDF
0.007
0.212
0.446
2,3,4,6,7,8-HxCDF
0.007
, 2.34
4.6
1,2,3,7,8,9-HxCDF
0.008
0.123
0.281
Other HxCDF
0.005
14.5
25.1
1,2,3,4,6,7,8-HpCDF
0.008
16.5
• 27.8
1,2,3,4,7,8,9-HpCDF
0.016
1.17
1.50
Other HpCDF
0.009
3.83
5.10
Octa-COF
0.022
10.0
5.91
(continued) -
6-28
-------
TABLE 6-22 (continued)
Isomer
F-176a
F -145
F-130
EPA Evaluation:
X of PCDD within 90%
100*
100*
100*
confidence level
% of PCDF within 90*
100*
86*
86%
confidence level
Average % error outside
0*
5.5%
6.3%
the 9Q% confidence
level
a Analyte not detected; value denotes detection limit.
6-29
-------
TABLE 6-23. EPA AUDIT"RESULTS
Results of performance audit sample Ho. 130
Results of PCDD:
0 of the 11 different PCDO are riot within the 90% confidence intervals.
0 of the 11 different PCDO are not within 50% of the 90% confidence intervals.
Results of the PCDF:
2 of the 14 different PCDF are not within the 90% confidence intervals,
2 of the 14 different PCDF are not within 50% of the 90% confidence intervals.
Results based on 2,3,7,8-TCQD equivalency factors:
Based on the 2,3,7,8-TCDD toxic equivalency factors, the average percent error
outside the 90% confidence limits was 6.3%, with an average bias of +5.5%.
Results of performance audit sample No. 145
Results of PCDD:
0 of the 11 different PCDO are not within the 90% confidence intervals.
0 of the 11 different PCDD are not within 50% of the 90% confidence intervals.
Results of the PCDF:
2 of the 14 different PCDF are not within the 90% confidence intervals.
2 of the 14 different PCDF are not within 50% of the 90% confidence intervals.
Results based on 2,3,7,8-TCDD equivalency factors:
Based on the 2,3,7,8-TCDD toxic equivalency factors, the average percent error
outside the 90% confidence limits was 5.5%, with an average bias of +5.2%.
Results of performance audit sample Ho. 176
Results of PCDD:
0 of the 11 different PCDD are not within the 90% confidence intervals.
0 of the 11 different PCDO are not within 50% of the 90% confidence intervals.
Results of the PCDF:
0 of the 14 different PCDF are not within the 90% confidence intervals.
0 of the 14 different PCDF are not within 50% of the 90% confidence intervals.
Results based on 2,3,7,8-TCDD equivalency factors:
Based on the 2,3,7,8-TCDD toxic equivalency factors, the average percent error
outside the 90% confidence limits was 0.0%, with an average bias of 0.0%.
Note: For those PCDD and PCDF reported as not detected by the Audi tee, three
times the reported detection limit (in parentheses) was used for the
calculations. The identical procedure was used for calculating the
confidence intervals.
6-30
-------
TABLE 6-24. RESULTS OF EPA AND INTERNAL .METALS AUDIT'SAMPLE ANALYSIS
Audit
As
Accuracy
Cd Accuracy
Cr
Accuracy
Pb
Accuracy
no.
(ug)
m .
(vg/mL) (J£)
(ug)
(%)
(ug)
m ¦
101 2270
91
NA
NA
253
101
999
100
102 1120
97
NA
NA
169
135
585
117
103 NA
NA
494
103
NA
NA
NA
NA
104 NA
NA
161
101
NA
NA
NA
NA
Reanalysis 3/23/88
102 NA
NA
NA
NA
167
134
¦ NA
NA
WP 283 1 NA
NA
NA
NA
1.34
107
NA
NA
Reanalysis 3/28/88
102 NA
NA
NA
NA
155
124
NA
NA
WP 283 1 NA
NA
NA
NA
1.28
102
NA
NA
NA = Not analyzed or riot applicable.
The certified value of the EPA WP 283 conc. 1 standard is 1.25 yg/nl. The audit
sample true values are:,
101 2500 ye As, 250 ug Cr, 1000 vq Pb
102 1150 u? As, 125 ug Cr, 500 ug Pb
103 480 ug/mL Cd
104 160 ug/mL Cd
6-31
-------
Based on the completed audit check list for the field sampling and analy-
sis, the audit rating was "acceptable With recommendations" and a concensus
that MRI was following the protocols as stated in the QAPP. In general, all
recommendations referred to refinement of investigative studies for future
tests. The audit did identify temperature reading mismatches between the ID
fan and fabric filter and between the air heater outlet and dry scrubber
inlet, RTI's field audit report is included in Appendix L.
6.3.4 Data Audit
An audit of data quality was conducted for data generated from the PCDO/
PCDF analyses. This audit report and follow-up actions/comments are presented
in Appendix L.
Project records included information regarding sample analyses, qualita-
tive observations, laboratory procedures, and calculations, plus evidence of
technical review. The data were found to be traceable, and the documentation
indicated that sample preparation and analysis were performed in accordance
with the test plan, QAP, and referenced methods.
5-32
-------
SECTION 7.0
REFERENCES
1. U.S. Environmental Protection Agency. Municipal Waste Combustion Study:
Emission Data Base for Municipal Waste Combustors. EPA-53Q/SW87-02lb,
June 1987.
2. U.S. Environmental Protection Agency. Emission Test Report, HC1 Continuous
Monitoring for Municipal Waste Combustion Study, Maine Energy Recovery
Company, Solid Waste-to-Energy Facility Refuse-Derived Fuel Process,
Biddeford, Maine, April 8, 1988 (EHB Report Mo. 88-MIN-06A).
3. U.S. Environmental Protection Agency, Interim Procedures for Estimating
Risks Associated with Exposures to Mixtures of Chlorinated Dibenzo-p-
dioxins and Dibenzofurans. EPA-625/3-87/012, Risk Assessment Forum,
October 1986.
4. ASME, Analytical Procedures to Assay Stack Effluent Samples and Residual
Combustion Products for Polychlorinated Dibenzo-p-dioxins (PCDD) and Poly-
chlorinated Dibenzofurans (PCOF). Draft analytical protocol, September 18,
1984.
5. U.S. Environmental Protection Agency. Test Methods for Evaluating Solid
Waste. SW-846, 3rd ed., November 1986,
6. U.S. Environmental Protection Agency. Determination of 2,3,7,8,-TCDD in
Soil and Sediment. Region 7, September 1933.
7. U.S. Environmental Protection Agency. Emission Measurement Branch (EMB)
iMetals Protocol. Draft report, August 1, 1987.
7-1
-------
SECTION 8.0
CONVERSION FACTORS
Multiply
English units
By
To obtain
SI units
lb/h
4.536 x 10"i
. kg/h
gr/dscf
2.288 x 109
ng/dscm
gr/dscf
2.288 x 103
mg/dscm
foot
0.3048
meter
inch
2.54
centimeter
tons/day
0.907
mg/day
gallon
3.785
11 ter
Btu/h
1.05488
MJ/'n
ft3/min
28316.8
cm3/mi n
inches H,0
0.0394
mm Hz0
8-1
------- |