PCB DISPOSAL BY THERMAL DESTRUCTION
U.S. Environmental Protection Agency-
Dallas, Texas
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
NTIS
J

-------
PB82-241860
EPA - 906/9-82-003
PCB DISPOSAL
BY
THERMAL DESTRUCTION
Compilation of Data and Information for the Approval of
Rollins Environmental Services, Deer Park, Texas
and
Energy Systems Company (ENSCO), El Dorado, Arkansas for
Disposal of Polychlorinated Biphenyls (PCBs)
February 1982
Final Report
Environmental Protection Agency
Region 6
Dallas, Texas

-------
50272 -10)			
REPORT DOCUMENTATION ;.i-_ report no. / g ;
	 F. P A -
PAGE
Title and Subtitle
PCB DISPOSAL BY THERMAL DESTRUCTION
7. Author(s)
3. Recipient's Accession No.
PB82 2 4186 0
5. Report Date
June 1981
6.
8. Performing Organization Rept. No.
Performing Organization Name and Address
Solid Waste Branch
Air & Hazardous Materials Division
U.S. Environmental Protection Agency, Region 6
Dallas, Texas 75270
10. Project/Task/Work Unit No.
11. Contract(C) or Grant(Q) No.
(C)
(G)
12. Sponsoring Organization Name and Address
U.S. Environmental Protection Agency-
Region 6
Dallas, Texas 75270
13. Type of Report & Period Covered
F inal 	 	
15. Supplementary Notes
14.
EPA/200/01
Abstract (Limit: 200 words)
A report on the sampling, analysis, and consideration of risks and
benefits associated with the incineration of polychlorinated biphenyls
(PCBs) at two commercial facilities in Deer Park, Texas and El Dorado,
Arkansas. Included are a summary, PCB incineration test reports,
polychlorinated dibenzo-p-dioxin (PCDD) and polychlorinated dibenzo-
furan (PCDF) emission sampling reports, a PCDD and PCDF analytical
chemistry report, air dispersion modeling results, an analysis of
risks and benefits, and letters and conditions of approval.
17. Document Analysis a. Descriptors
t>. Identifiers/Open-Ended Terms
c- COSATI Field/Group
18. Availability Statement
Release Unlimited
19. Security Class (This Report)
Unclassified	
20. Security Class (This Page)
	Vnqlas^iQd
21. No. of Pages
22. Price
(See ANSI—Z39.18)
See Instructions on Rm/»ree
OPTIONAL FORM 272 (4-77)
(Formerly NTIS-3S)
Deoartment of Commerce

-------
NOTICE
THIS DOCUMENT HAS BEEN REPRODUCED
FROM THE BEST COPY FURNISHED US BY
THE SPONSORING AGENCY. ALTHOUGH IT
IS RECOGNIZED THAT CERTAIN PORTIONS
ARE ILLEGIBLE, IT IS BEING RELEASED
IN THE INTEREST OF MAKING AVAILABLE
AS MUCH INFORMATION AS POSSIBLE.
I (*!

-------
Table of Contents
Page Number
1ii
1
16 (Division 1)
160 (Division 2)
293 (Division 3)
355 (Division 4)
416 (Division 5)
502 (Division 6)
512 (Division 7)
527 (Division 8)
594 (Division 9)
Title
Introduction
Summary
PCB Incineration Test
ENSCO, El Dorado, Arkansas
October 1979
PCB Incineration Test
Rollins, Deer Park, Texas
November 1979
Emission Tests for PCDDs & PCDFs
Rollins, Deer Park, Texas
August 1980
Emission Tests for PCDDs & PCDFs
ENSCO, El Dorado, Arkansas
August 1980
Determination of PCDDs and PCDFs
from Stack Effluent Streams at
Rollins and ENSCO
Air Quality Modeling Analysis
Rollins, Deer Park, Texas
Air Quality Modeling Analysis
ENSCO,El Dorado, Arkansas
Consideration of Risks and
Benefi ts/Alternatives
Concerning PCB Incineration
in Region 6
Approval Letters and Conditions
of Approval - ENSCO and Rollins

-------
INCINERATION OF PCBs - A SUMMARY OF EPA-REGION 6 APPROVAL ACTIONS
U.S. EPA, Reqion 6, Dallas, Texas
BACKGROUND - In response to mounting evidence that polychlorinated
biphenyls (PCBs) were a high risk chronic toxicant, Congress, in 1976,
banned the manufacture of these compounds. Final regulations*
published on May 31, 1979, prohibited the manufacture, distribution,
and use of PCBs after July 2, 1979, and detailed specific storage and
disposal requirements. All materials with PCB concentrations greater
than 500 ppm must be incinerated under EPA prescribed conditions (See
Table 1). A large portion of the electrical industry's PCBs fall into
this category. Other acceptable disposal methods for lower PCB con-
centrations (50-500 ppm) include landfillinq and incineration in high
efficiency boilers. In the regulations for PCB incinerators, there is
a provision which allows the Regional Administrator to require a trial
bum to ensure that the operational requirements of the incinerator are
met. This report summarizes the approval actions completed by EPA-
Region 6 for two PCB incinerators.
CHRONOLOGY OF FIRST TRIAL BURNS - EPA-Region 6 received applications
for PCB incineration from Energy Systems Company (ENSCO) at El Dorado,
Arkansas, on March 8, 1978, and from Rollins Environmental Services at

-------
- 2 -
Deer Park, Texas, on April 6fc 1978. The first PCB trial burns were
conducted at ENSCO in October 1979, and at Rollins in November 1979.
While both facilities demonstrated compliance with PCB regulations for
temperature, destruction efficiency, residence time, excess oxygen, and
acceptable carbon monoxide concentrations, laboratory results revealed
the possibility that very low levels of two toxic organic compounds were
present in some of the stack samples. These compounds were tetrachlorodibenzo-
p-dioxins (TCDD) and tetrachlorodibenzo-furans (TCDF). At this point
Region 6 determined that additional tests were warranted at ENSCO and
Rollins to determine whether dioxins and furans were actually being
emitted during PCB combustion and the quantities and sources of these
compounds, i.e. waste auxiliary fuels, PCBs. Both facilities used
waste fuels during PCB incineration to achieve their operating temperatures.
EPA paid for the sample collection, sample preparation, analytical
analyses, and the oil for these retests.

-------
- 3 -
TABLE 1
PC8	INCINERATOR REQUIREMENTS
Destruction Efficiency	99.9999%
Temperature	1150°C
Dwell Time	2 seconds
Excess Oxygen	3%
CO	< 100 ppm
HC1 Removal	99%
Particulates	State Permit
PC8 Feed'Rate	Test Bum

-------
- 4 -
CHRONOLOGY OF SECOND TRIAL BURNS - The second trial burn plan was
developed jointly by Region 6 and the Office of Pesticides and Toxic
Substances with agreement by the applicants. The plan included three
days of sampling during the incineration of a variety of fuels. Inciner-
ation on the first day was with the facility's normal waste fuel and
no PCBs. The second day of incineration was with normal waste fuel and
PCBs, while the third day was with PCBs and virgin fuel oil. The sampling
was conducted by an independent testing firm using methods and procedures
prescribed by EPA. EPA was present during all tests. The retests were
performed in August 1980.
SAMPLING METHODS ANO ANALYSIS FOR DIOXINS AND FURANS - The sampling of
stack emissions for dloxins and furans was with a modified EPA Method 5
2
Train . The train incorporated two resin cartridges, XAD-2 and florisll,
to adsorb the volatile organlcs (Figure 1). The resin cartridges were
located downstream from the heated probe and between the third and fourth
impingers. The sampling train was designed to collect both particulates
and organic compounds 1n the stack gas. Two resins were used to insure
adsorption of organlcs which may have passed through the first resin.
Resin cartridges and Impingers were water jacketed to maintain a temperature
below 70°F. The effect of ultraviolet light on photosensitive organic
compounds was minimized by covering the train with aluminum foil and
opaque plastic.
The stack gases were sampled Isokinetically for a period of four hours
during each of the three days of incineration. After being transported
to the laboratory, stack samples were extracted with an appropriate

-------
- 5 -
solvent and analyzed using capillary column gas chromotography and high
3
resolution mass spectrometry . For dioxins and furans, an added quality
control technique was used. The samples were spiked with a chlorina-37
labeled dioxin and furan which allowed the analyst to determine the
extraction efficiency of each sample.
RESULTS OF TRIAL BURNS - The results of both trial burns at ENSCO and
Rollins are provided in Tables 2 and 3. PCBs were not analyzed from
stack samples from the second trial burn since no PCBs were detected
during the first burn. The 1980 data are the results from the second
day when the auxiliary chemical waste fuel was being burned with PCBs.
When only chemical wastes were being incinerated, dioxins and furans
were also detected in the stack samples (Table 4). During day one,
both incinerators were operating below 1200°C.
PREDICTION OF AMBIENT AIR CONCENTRATIONS - The dioxin and furan emission
data from the second test were used to predict maximum ground level
concentrations. Calculation of the dioxlns/furans was by the use of the
RAM* air dispersion model which 1s applicable to level terrain. The
CRSTER^ model was used for the El Dorado facility, where the terrain 1s
not level. The model simulates plume dispersion behavior as 1t travels
downwind from the stack by utilizing parameters such as temperature and
velocity of stack gases, stack height, wind speed and direction, and
mixing height. Maximum concentrations for the dioxlns/furans, as predicted
-14 -IS
by the model, were in the order of 10 to 10 grams/cubic meter.

-------
- 6 -
Reproduced from
best available copy.
FIGURE 1 STACK SAMPLING TRAIN
7L0RISIL
7"
'ACUDM
LXSZ

-------
- 7 -
TABLE 2
ENSCO, El Dorado, Arkansas
Results of Trial 8ums
Oct. 1979	Aug. 1980
PC8 Feed Rate, lb/hr	636	722
Temperature °c (avg)	1254	1224
CO, ppm (max)	10	7
Excess 02» %	13.4	9.5
co2, %
Particulates, lb/hr	149	8
HCT Removal eff, %	98.9	99.8
PCB Destruction eff, %	99.99998	no PCB analysis
Combustion eff, %	99.98	99.99
*Dioxin, ng/dscf	not analyzed	0.005
*Furan, ng/dscf	not detected**	0,060
*ng/dscf » nanogram per dry standard cubic foot.
**detection limit of 1980 analysis was 1000 times greater than
1979 detection limit.

-------
- 8 -
TABLE 3
Ro11i ns Envi ronmental
Services, Deer Park,
Texas
Results
of Trial Burns


Nov. 1979
Auq. 1980
PCB Feed Rate, Ib/hr
2438
2502
Temperature, °c (avg)
1185
1175
CO, ppm (max)
11
38
Excess Og. % (min)
8.9
9.8
co2, %
9
8.1
Particulates, Ib/hr
39
24
HC1 Removal eff, %
98.4
99.9
PCB Destruction eff, %
99.99999
no PCB analysis
Combustion eff, %.
99.98
99.95
*01oxin, ng/dscf
Not Analyzed
0.018
*Furan, ng/dscf
Not Detected**
0.281
*ng/dscf * nanogram per dry standard cubic foot.
**detection limit of 1980 analysis was 1000 times greater than
1979 detection limit.

-------
- 9 -
TABLE 4
RESULTS OF TCDD AND TCDF ANALYSES FROM SECOND TRIAL BURN
Total TCDF
13.5 ng
23.0 ng
2.0 ng.
ROLLINS	Total TCDD
Day 1 (chemical waste)	6.94 ng
Day 2 (chemical	waste and	PCBs) 1.42 ng
Day 3 (fuel oil	and PCBs)	Not Detected
ENSCO
Day 1 (chemical waste)
Day 2 (chemical	waste and PCBs)
Day 3 (fuel oil	and PCBs)
0.26 ng*	2.0 ng
0.48 ng	6.0 ng
Not Detected Not Detected
~Excludes Impinger Water

-------
- 10 -
UPPER LIMIT RISK ASSESSMENT - With the verification that dioxins and
furans would be present in the stack gases during PCB incineration,
Region 6 decided to develop, with the assistance of EPA's Carcinogen
Assessment Group, an upper limit risk assessment to determine whether
these toxic compounds would pose a significant health risk to area
residents.
ASSUMPTIONS - Realizing that there was limited toxicologlcal evidence
for all but the 2, 3, 7, 8 isomer of TCDD and even less for the furans,
EPA-Region 6 developed the risk assessment as a "worst case" by making
very conservative assumptions to replace the missing toxicologlcal
information. The key "worst case" assumptions were as follows:
1.	All TCDD and TCDF were assumed to exhibit the same degree of
toxicity and carcinogenicity as the 2, 3, 7, 8 Isomer of
TCDO
2.	The toxicity and carcinogenicity of 2, 3, 7, 8 TCDF equals
that of 2, 3, 7, 8 TCDD
3.	The sensitivity of humans to the health effects of these
materials equals that of the most sensitive species observed
1n lab studies
4.	For 2, 3, 7, 8 TCDD, there Is no threshold level below which
carcinogenic risk is zero

-------
- 11 -
5.	The duration of exposure to this material will be a 70 year
average lifetime period
6.	The frequency of exposure is 24 hours per day, 7 days per
week for 70 years
7.	The intensity of exposure is calculated from ambient concentrations
of TCDD and TCDF which are predicted by a computerized air
dispersion model from the actually measured stack emissions
8.	The total population within the affected area 1s exposed to
the highest ambient concentration of TCDD and TCDF predicted
by the model
9.	All human exposure will be through inhaling tiny airborne
particles to which TCDD and TCDF adhere
10. The typical exposed person: a.
b.
c.
3
breathes 20 m /day
Ingests and retains 100% of the
particulates inhaled
absorbs 100% of the TCDD
and TCDF
has an average body weight
of 60 kg.

-------
- 12 -
EXPOSURE ESTIMATES - The intensity of exposure was predicted by the
previously discussed air computer modeling. The population exposed
to these concentrations were estimated in the following way:
Rollins: The population within a three-mile radius of the Rollins
stack included the total population of Deer Park, Texas, and an
estimated 1/3 of the population of the town of Deepwater. Also
included in the population estimate were an additional 300 employees
from nearby industries even though these people are only subject
to an eight-hour exposure. The 1980 population records were
furnished by the Houston-Galveston Area Council of Governments.
The computer modeling showed the maximum concentration of TCOD
and TCDF to occur at a point 1 1/2 miles from the stack with an
estimated 25,000 people within the 1 1/2 mile radius circle.
ENSC0: The total metropolitan population of El Dorado, plus
the suburban and rural population of most of the county, were assumed
to be exposed. These 1980 figures came from documents used by EPA for
water quality planning 1n the El Dorado area. The estimate 1s
50,000 people, with 24,000 residing 1n the area of maximum con-
centration (1.2 km).
The total estimated exposed population 1s assumed to be located and
subject to the highest predicted ambient concentrations of TC0D/TCDF.

-------
- 13 -
CALCULATION OF PREDICTED RISK - EPA's Carcinogen Assessment Group has
prepared for the Office of Air Quality Planning and Standards a Risk
Assessment Document® which summarizes data on TCDD. This document
estimates that the unit risk factor associated with breathing (70 year
3
exposure) air containing one nanogram of 2, 3, 7, 8-TCDD/m is 0.091,
i.e., 91 increased cases of cancer per 1000 people. This factor was
used with the predicted air concentrations at the two incinerators
to arrive at a site-specific upper limit of risk.
-7 *
The resulting individual risk for ENSC0 was found to be 3.9 x 10 ,
or less than one increase case of cancer per 2*500,000 exposed persons.
mC*
For the Rollins facility, the value was found to be 1.8 x 10 { or less
than one increased case of cancer in 50,000 exposed persons. These
numbers do not represent an estimate of the actual individual risk from
operation of the incinerators, but are an estimate of the upper level
of risk possible 1f all the worst case assumptions occurred. Such an
occurrence is virtually not possible.
PCB INCINERATION APPROVAL CONDITIONS - The PCB Regulations stipulate
the performance requirements which must be met"prior to the granting
of approval by the Regional Administrator (40 CFR Part 761). Additionally,
the regulations allow the Regional Administrator to include any other
requirements which are necessary to ensure that the operation of the
incinerator does not present an unreasonable risk of injury to health
or the environment from PC8s. The following PCB Incineration approval
conditions outline specific conditions which must be met to remain an
approved PCB disposal facility.
"Based on analyzed dicocin/furan levels where detected, plus levels of detection
where not detected. This is an additional conservative consideration.

-------
- 14 -
To ensure combustion efficiency greater than 99.9%, the approved facility
must operate above 1100°C, with excess oxygen greater than 3% and carbon
monoxide (CO) less than 10 times the CO2 level. If at any time during
PC3 incineration the combustion efficiency drops below 99.9%, or the
temperature drops below the minimum specified, an automatic cutoff mechanism
stops the flow of PCBs.to the incinerator.
The incineration of chlorinated compounds produces chlorine in the
combustion gases; 99% of this chlorine must be removed by the wet scrubber
to comply with the PCB approval conditions. Particulate emissions from
the stack of the facility must meet the State's particulate emission
permit limit. Annual retesting may be required by EPA to ensure
continued compliance with particulate and chlorine limits.
Other approval conditions include requirements for handling, storage,
marking, and record keeping. All PCBs received at the facility must
be recorded as to quantity, date received, and from whom the PCBs were
received. Also, the facility must record the date and quantity of PCBs
incinerated. These records and the incinerator operation will be
monitored regularly by EPA and by an independent contractor who will
inspect the facility at random times.
In addition to compliance with the PCB requirements, the incineration
facilities must meet all applicable State, local, and Federal regulations
during PCS incineration.

-------
- 15 -
References
1.	Federal Register, 44: 31514, May 31, 1979, (40 CFR Part 761).
2.	Method 5 - Determination of Particulate Emissions from
Stationaly Sources, 40 CFR 60,344, Appendix A.
3.	Tiernan, T.O. et aj_, Determination of Polychlorinated Dibenzo-
p-Diox1ns (PCDDs), Dlbenzofurans (PCDFs), and Biphenyls (PCBs)
in Stack Effluents and Other Samples Resulting from Assessments
of PCB Incineration Tests at Rollins Environmental Services,
Deer Park, Texas and Energy Systems Company (ENSCO), El Dorado,
Arkansas. U.S. EPA Cooperative Agreement #CR806846.
4.	User's Guide for RAM, November 1978, EPA-600/8-78-016A.
5.	Users Manual for Single-Source (CRSTER) Model, July 1977,
EPA-450/2-77-013.
6.	U.S. EPA Carcinogen Assessment Group, Office of Health and
Environmental Assessment, Office of Research and Development.
"Method for Determining the Unit Risk Estimates for Air Pollutants."
July 1980.

-------
Division 1

-------
/ 7
Disposal of Polychiorinated Biphenyls (PCBs)
and PCB-Contaminated Materials
Volume 4: Test Incineration of Electrical
Capacitors Containing PCBs
FP-1207, Volume 4
Research Project 1263-2
Final Report. September 1980
Work Completed, May 1980
Prepared by
ACUREX CORPORATION
Energy and Environmental Division
485 Clyde Avenue
Mountain View, California 94042
Project Manager
J. Steiner
Principal Investigators
N. W. Flynn
C. 0. Wolbach
Prepared for
Electric Power Research Institute
3412 Hill view Avenue
Palo Alto, California 94304
EPRI Project Manager
R. Y. Komai
Water Quality Control and Heat Rejection Program
Coal Combustion Systems Division

-------
J?
ORDERING INFORMATION
Requests for copies of this report should be directed to Research Reports Center
(RRC), Box 50490, Palo Alto, CA 94303. (415) 965-4081. There is no charge for reports
requested by EPRI member utilities and affiliates, contributing rtonmembers. U.S. utility
associations. U.S. government agencies (federal, state, and local), media, and foreign
organizations with which EPRI has an information exchange agreement. On request,
RRC will send a catalog of EPRI reports.
Copyrignt 1930 Electric Power Research institute, inc.
EPRi authorises me reproduction ana Oisiriounon oi all or any portion oi inn report and the preparation
of any aenvativt work based on tnis report in each cat* on the condition that any such reproduction,
distribution, and preparation shall acknowledge this report and EPRI if the source.
NOTICE
This report was prepared By the organicaiioryji named Beiow as an account of work sponsored by the
Electric Power Research institute, inc (EPRI) Neitner EPRI, members ol EPRI. the organizations) named
below, nor anv pe'Jon acting on tneir penait 
-------
/
-------
%0
EPRI PERSPECTIVE
PROJECT DESCRIPTION
This final report is the succeeding volume to the set of three volumes previously
published on disposal of polychlorinated biphenyls (PCBs) and PCB-contaminated
materials. This project (RP1263-2) is one of a series aimed at the safe disposal
of utility-generated wastes—specif1cally in this case, the sampling and analysis
during PCB incineration. Jointly funded under EPRI's Interdepartmental Solid-
Waste Disposal Program, the contractor presents the analytic considerations,
procedures, and results of the Energy Systems Company (ENSCO) PCB test burn in
El Dorado, Arkansas, in October 1979. The electric power-generating industry's
interest 1n this project stems from Its ownership of the majority of existing
PCBs in this country and its constraint by federal regulations to a limited
number of disposal options.
PROJECT OBJECTIVE
The overall objective of this study was to participate in and to document
sampling and analysis procedures utilized in the certification test burn of
PCBs. This is essentially a case study of an EPA-required test burn for a
conmercial PCB incinerator.
PROJECT RESULTS
The test found destruction of the PCBs to be highly efficient. Samples taken
from the stack, the scrubber, and the ash were analyzed for PCBs. PCBs were only
confirmed in the ash, but at levels below the federally regulated amount of
50 PPM.
An analytic problem, which arose and should be kept in mind during PCB analysis,
stemned from the perchlorlnation process. The principle is to convert whatever
chlorinated biphenyls are present Into decachlorobiphenyl (converting the 209
possible PCBs into a single compound). There are a number of problems arising
from this technique. Biphenyl (a non-PCB, nonregulated substance) is converted
v
-------
a i
to a PCB and counted as such. The mass of PCBs reported may be significantly
greater due to the higher molecular weight when chlorines are substituted onto
the molecule, i.e., the molecular weight may be doubled. False positives may be
obtained and, without confirmatory tests, one could erroneously report the
presence of PCB.
The analytic work by Acurex included the occurrence of false positives.
Because of the sensitivity of this incineration test, it was decided that
gas chromatography-mass spectrometry would be used for confirmation of PCBs.
None were found in the unperchlorinated stack gas samples or the scrubber water.
It is recommended that the perch1orination method of PCB analysis only be used as
a screening technique, and that positive results be confirmed on the unperchlori-
nated sample. There 1s a danger of reporting PCBs even if none are present,
unless a second technique 1s used to verify the PCB level obtained by the
perch1 orination analysis procedure.
The incineration process proved to be an effective and efficient way of
destroying PCBs. It will be a useful option for PCB disposal by utilities.
Ralph Y. Komai, Project Manager
Coal Combustion Systems Division
v1
-------
%x
CONTENTS
Section	Page
1	INTRODUCTION	1-1
2	PROCESS DESCRIPTION	2-1
2.1	Solid Waste Shredders	2-1
2.2	Rotary Kiln	2-9
2.3	Afterburner	2-13
2.4	Venturi Jet Scrubber	2-16
3	SAMPLING LOCATIONS, EQUIPMENT AND PROCEDURES	3-1
3.1	Location of Sampling Sites	3-1
3*1.1 Process Sampling Sites	3-1
3.1.2 Stack Sampling Site	3-9
3.2	Sampling Equipment	3-9
3.2.1	Process Stream Sampling Equipment	3-9
3.2.2	Stack Sampling Equipment	3-19
3.3	Sampling Procedures	3-23
3.3.1	Process Stream Sampling Procedures	3-23
3.3.2	Stack Sampling Procedures	3-24
4	ANALYTICAL PROCEDURES AND RESULTS	4-1
4.1	Introductory and General Approach	4-1
4.1.1	Pattern Recognition	4-2
4.1.2	Derivatization	4-3
4.1.3	Measurement of Individual PCB Components	4-4
4.2	Analysis of PCBs	4-7
4.2.1	Rotary K1ln Ash Analysis	4-7
4.2.2	Scrubber Liquor Effluent and Recycled Scrubber
Liquor Analysis	4.9
4.2.3	Stack Gas Analysis	4-12
vii
-------
a 3
Section	Page
4.3	Analysis for Total Organic Chlorine	4-18
4.4	Analysis of Particulate Matter	4-18
4.5	Analysis of Hydrochloric Acid	4-21
4.6	Analysis of N0X	4-21
4.7	Quality Control and Quality Assurance	4-21
4.8	Sunmary and Conclusion	4-22
5 RESULTS OF PCB TRIAL BURN	5-1
5.1	Retention Time	5-1
5.2	Combustion Efficiency	5-1
5.3	Feedrate of PCBs	5-1
5.4	Temperatures of Incineration	5-2
5.5	Automatic Shutoff for PCBs	5-2
5.6	Monitoring of Stack Emissions	5-2
5.7	Monitoring of Combustion Products	5-2
5.8	Water Scrubbers	5-2
5.9	Mass Air Emissions	5-4
5.10	Kiln Ash	5-4
REFERENCES	R-l
APPENDIX A	A-l
APPENDIX B	B-l
viii
-------
2-V
ILLUSTRATIONS
Elaure	Page
2-1	Solid/Liquid Waste Treatment Facilities of ENSCO	2-3
2-2	Process Flowchart of ENSCO Facilities	2-5
2-3	Capacitor Shredder Structure	2-7
2-4	Rotary Kiln	2-11
3-1	Liquid Organic Waste Sampling Site	3-3
3-2	Rotary Kiln Ash Sampling Site	3-5
3-3	Well Water Sampling Site	3-7
3-4	Recycled Scrubber Liquor Sampling Site	3-11
3-5	Lime Slurry Sampling Site	3-13
3-6	Caustic Solution Sampling Site	3-15
3-7	Scrubber Liquor Sampling Site	3-17
3-3	PCB/RCJl Impinger Train	3-21
3-9	Preparation Procedures for Particulate/HCH Sampling Train	3-25
3-10	Sampling Procedures for Particulate/HC2. Sampling Train	3-25
3-11	Sample Recovery Procedures for Particulate/HQ,	3-27
3-12	Preparation Procedures for PCB/RCS- Sampling Trains	3-29
3-13	Sailing Procedures for PCB/RCs, Sampling Train	3-31
3-14	Sample Recovery Procedures for PCB/RCi, Sampling Train	3-32
3-15	Preparation Procedures for N0x Sampling Train	3-33
3-16	Sanpling Procedures for N0X	3-34
3-17	Sample Recovery Procedures for NO	3-35
3-18	Sanpling Procedures for C02» 02, CO	3-36
4-1	fieneral Formula for Polychlorinated Siphenyls	4-1
4-2	General Analytical Scheme for PCB Samples	4-6
4-3	Analytical Scheme for Rotary K1ln Ash	4-8
4-4	Analytical Scheme for Scrubber Liquor and Recycled Scrubber Liquor 4-11
ix
-------
ar
Figure	Page
4-5	Scrubber Liquor Chromatogram	4-13
4-6	Analytical Scheme for PCB Stack Gas Sampling Train	4-15
4-7	Chromatogram of Perchlorinated PCB Train Sample	4-17
4-S	Method Blank Chromatogram	4-23
4-9	Decachlorobiphenyl Calibration Curve (SC/EC)	4-24
4-10	Biphenyl Standard Curve (GC/FID)	4-25
x
-------

TABLES
Table

Page
2-1
Feedrates of Solid Capacitors and Liquid PCBs During Trial Burn
2-9
2-2
Combustion Temperatures Versus Time for Trial Burn
2-14
2-3
Liquid Organic Waste Characteristics
2-25
2-4
Feedrates of Liquid Organic Waste, Well Water, and
Steam During Trial Burn
2-15
4-1
Rotary Kiln Ash Analysis as Decachlorobiphenyl
4-10
4-2
Results of Scrubber Liquor and Recycled Scrubber Liquor
Analysis as Decachlorobiphenyl
4-14
4-3
Results of Stack Gas Analysis
4-16
4-4
Biphenyl Analysis of PCB Stack Gas Sampling Train Samples by GC/MS
4-19
4-5
Analysis of Particulate Matter, HC2., N0X and RC Jl
4-20
5-1
Stack Emissions During PCB Trial Burn
5-3
5-2
Monitoring Results of Combustion Products
5-3
xi
-------
The first commercial PCB trial burn in the U.S. was conducted at ENSCO. This
trial burn was completed over a four-day period and consisted of a background burn
(no PCBs were destroyed) and three PCB burns (liquid PCBs and solid electronic
capacitors containing PCBs were destroyed). During the trial burn, an extensive
series of tests were performed in accordance with EPA's recently published rules
and regulations governing incineration of PCBs. Samples of all input streams and
all output streams to and from the ENSCO incineration facility were collected on a
daily basis and analyzed for PCBs.
Based on the results of these tests, it is the opinion of the contractor that the
ENSCO facility has successfully completed the first PCB trial burn. The
incineration facility effectively destroyed PCBs. PCBs were detected in only one
effluent stream from the facility (rotary kiln ash) but the PCB level of this
stream was below the lower limit of the regulations set by EPA (50 ppm).
Analytical problems arose in the use of the EPA perchlorlnation procedure for PCB
analysis. There were positive results of non-PCB origin (i.e., presence of
decachlorobiphenyl which was not formed from PC8s). This was confirmed by the use
of gas chromatography/mass spectrometry (GC/MS) on samples which were not
perchlorinated.
S-l
-------

SECTION 1
INTRODUCTION
This section of the report presents a brief summary of the technology used for the
disposal of PCBs by Incineration. A more detailed overview of this disposal
technology is available in an EPRI report (FP-1207, Volume 1) entitled "Disposal
of Polychlorinated Biphenyls and PCB-Contaminated Materials."
Since the early 1970s, when PCBs were first identified as a serious environmental
hazard, considerable research effort has gone into development of safe disposal
techniques. Incineration has received the most attention, since this disposal
technique actually destroys PCBs rather than simply containing them (eg. chemical
landfill) where the possibility still exists for the PCBs to enter the environment
at some later date.
There is no authoritative reference which can be cited which completely describes
the incineration conditions required for total destruction of PCBs. However, a
number of studies have been conducted using a variety of combustion devices to
thermally destroy PCBs. Tests have been conducted in cement kilns, utility
boilers, hazardous waste incinerators and sludge incinerators to determine their
PCB destruction efficiency. These kinds of combustion devices have been selected
since they can provide the retention time, turbulence and temperature
characteristics required to destroy PCBs, which are among the most thermally
stable hazardous compounds in use today. Based on these test results and other
PCB combustion research done on laboratory scale devices, the EPA has promulgated
a set of rules and regulations (40 CFR Part 761, Vol. 44, So. 106, May 31, 1979)
governing the destruction of PCBs by incineration. These rules and regulations
specify the combustion conditions which a combustion device must provide to
effectively destroy PCBs. EPA has also specified the types of PCBs which may be
destroyed in a given combustion device recognizing that not all combustion devices
can handle all types of PCBs. Finally, EPA has specified the type of pollution
control device required to control the pollutants generated by the incineration
process and has set stringent PCB emission and monitoring standards commensurate
with this effective combustion criteria and control technology.
1-1
-------
The PC8 trial burn at the ENSCO facility in El Dorado, Arkansas is the first
coimiercial attempt to destroy PCBs in accordance with these new rules and
regulations developed by EPA. The results from this trial burn will demonstrate
whether or not PCBs can be effectively destroyed using these criteria.
1-2
-------
3 o
SECTION 2
PROCESS DESCRIPTION
ENSCO owns and operates a solid and liquid waste treatment facility as illustrated
in Figure 2-1. The basic equipment arrangement used by ENSCO to conduct this PCB
trial burn is shown in Figure 2-2. This section of the report suirmarizes the
manner in which the process equipment was operated during this trial burn. The
descriptions contained in this section were obtained from several conversations
with ENSCO personnel and from observations made by Acurex during and after the
trial burn.
2.1 SOLID WASTE SHREDDERS
Figure 2-3 is a photograph of the shredder structure which houses two
hammermills. These two hammermills are manufactured by Gruendler Crusher and
Pulverizer Inc. and operate in series. The primary hammermill (approximately
30 hp) shreds the small electronic capacitors (2 in. by 2 in. by 4 in.) which then
drop into the secondary hanroermill (approximately 50 hp) and are reduced to
0.5 inch square pieces referred to as "fluff". These capacitors are contained in
weighed, 55-gal drums. The contents of these drums were manually dumped into the
primary shredder. The capacitors (analyzed by Northrop Services) contained 31.1 +
5.1 percent by weight of oil. This oil was 79.7 percent by weight
Aroclor^l016 (i.e., the capacitors contained 24.8 + 8.5 percent by weight of
Aroc lor ^1016). An ultimate analysis of these same type of capacitors done
three years ago (by TRW on EPA Contract 58-01-2966) yielded the following results
— 35.4 percent by weight carbon, 3.3 percent hydrogen, 12.4 percent chlorine,
16.9 percent oxygen, 32.0 percent ash and a higher heating value of 5,560 Btu/lb.
The average feedrate of capacitors for the three PCB tests was 980 Ib/hr. PCBs
accounted for 243 Ib/hr of this total amount. No large electrical utility
capacitors were incinerated during this trial burn.
Liquid PCBs (80.9 + 4 percent by weight of Aroclor® 1016(determined by Northrop
Services) are gravity-fed into the hopper under the secondary hammermill where
they mix with the "fluff" and actually adsorb onto the fluff (i.e., there is
2-1
-------
Figure 2-1. Solid/Liquid Waste Treatment Facilities of ENSCO
-------
SMALL
ELECTRONIC
CAPACITORS IN
55-GAL. DRUMS
LIQUID ORGANIC WASTES
NATURAL GAS
COMBUSTION AIR
LIQUID
PCBs
TOTALLY ENCLOSED
SHREDDERS
STACK-.
NATURAL GAS
STEAM
WELL WATER
LIQUID ORGANIC WASTES
3000 CU. FT.
PRIMARY
COMBUSTION
CHAMBER
0EM1STER
PAD
SMREDOER
FLUE
JASES
ROTARY KILN
MAKE-UP
WATER
ASH
DROP
WASTES
(OR WELL WATER)
COMBUSTION AIR
3000 CU. FT.
SECONDARY
AUGER
SCRUBBER
BLOWERS
COMBUSTION
CHAMBER
EXHAUST AIR
SIEVE TRAYS
TEMPORARY ASH
STORAGE ONSITE
LIME
STACK
DRAIN
SLUDGE
LAGOON
LIME
MIXING
TANK
CAUSTIC
CAUSTiC
MIXING
TANK
Figure 2-2. Process Flowchart of ENSCO Facilities
-------
Figure 2-3. Capacitor Shredder Structure
-------
3y
no excess liquid PCB). The PCB soaked fluff then falls into the auger for
transport to the rotary kiln. An ultimate analysis (by Northrop Services) of the
liquid PCBs indicated the following composition and heating values — 54.8 percent
by weight carbon, 2.6 percent hydrogen, 42.6 percent chlorine and 8,820 Btu/lb.
The average feedrate during this trial burn was 486 Ib/hr. Table 2-1 summarizes
the feedrate data obtained during the trial burn.
Table 2-1
FEEDRATES OF SOLID CAPACITORS AND LIQUID PCBs DURING TRIAL BURN
Component Feedrate3
10/15/80
10/16/80
10/17/80
10/18/80
Averaqec
Solid Capacitors
860b
1055
870
1015
980
PCB's in Solid Capacitors
0
268
224
238
243
Liquid PCBs
253b
447
478
533
486
PCBs in Liquid
0
362
387
431
393
aFeedrates in Ib/hr (determined by ENSCO and Northrop)
bNon-PCB materials during background burn
cAverage of 10/16, 10/17, 10/18
The hoppers of the hammermills are maintained under slight negative pressure
(approximately 0.1 1n. wg) to prevent the release of PCBs into the ambient air
during the shredding operation. Suction on the hoppers is provided by a blower
(approximately 3000 cfm at 1 psi) and this exhaust air is used as combustion air
for the rotary kiln.
2.2 ROTARY KILN
The PCB soaked fluff is discharged by the auger into the front of the rotary kiln
as illustrated in Figure 2-4. The rotary kiln is 34 feet long and has an internal
diameter of 7 feet. The kiln 1s refractory lined and rotates at approximately
0.5 rpm which results in a retention time of approximately 60 minutes for solids
2-9
-------
Figure 2-4. Rotary Kiln
-------
s £>
and approximately 3.7 seconds for gaseous material. There are lifters protruding
from the refractory to ensure solid material is "turned over" as it passes through
the kiln. The temperature inside the kiln (frontend) is measured by a thermocouple
and varied from 1200 to 1500°F (average 1400°F) during the tests. Table 2-2
illustrates the temperature variation with time. Liquid organic waste (chlorinated
aliphatic hydrocarbons), well water and natural gas are also injected into the
front of the rotary kiln. The characteristics of the liquid organic waste feed
stream is surnnarized in Table 2-3. Analysis of the natural gas (by ENSCO) yielded
75.0 percent by weight carbon, 25.0 percent hydrogen and a heating value of
24,190 8tu/1b. The average well water injection rate 1s 373 gal/hr. Only the
total flowrates of liquid organic waste, natural gas, and well water are measured.
There is no separate flowrate measurement for the rotary kiln or primary combustion
chamber. Approximately 75 percent of the total natural gas (averaged 2,260 scf/hr)
is consumed in the rotary kiln to ignite the waste material being fired. The
remaining 25 percent of the natural gas is used in the primary combustion chamber
through the two burners (in the large pilot lights). Ten percent of the total
liquid organic waste (averaged 590 gal/hr) is injected .In the rotary kiln compared
with 90 percent 1n the primary combustion chamber. Table 2-4 sunmarizes the dally
feedrates of these components.
A drag chain 1s used to remove the ash remaining 1n the kiln after destruction of
all the combustible matter. For these tests, the kiln ash was captured in steel
hoppers (foreground of Figure 2-4), covered with polyethylene and stored onslte.
The total volume of kiln ash produced during the trial burn was approximately
127.6 ft^ {no data for individual days). The effective density of this ash was
approximately 0.86 lb/ft3. Hence, the total amount of ash collected was
approximately 109.7 lb.
2.3 AFTERBURNER
The gases leaving the rotary kiln are injected with steam and fed into the primary
combustion chamber of the afterburner. This is a refractory lined chamber
(approximately 3000 ft^) into which most of the liquid organic waste 1s fired.
It has a residence time of approximately 2.19 seconds. The temperature of this
chamber 1s measured by a thermocouple 1n the crossover duet (between the primary
and secondary combustion chambers). The average temperatures measured during the
tests were 232Q°F, 2280°F, 2270°F and 2290°F. The liquid organic waste
has a low ash content (approximately 3.7 percent by weight) and the ash 1s molten
2-13
-------
37
Table 2-2
COMBUSTION TEMPERATURES VERSUS TIME FOR TRIAL BURN
Tl«	J0/lS/7»	10/16/79	10/17/79	10/18/79
1 Hji **t«rtniriwr Hat Ouct *11w Aftjjgjgjjjr Wot 3uet Kllw Afttrflurwr Hot 3uct lOln Afjjraurjjr Wot Ouet
10:00 m




2280





2350





2175





2275
11:00 m
u«o
2290
1890

2225

1440
2373
19S0
1310
2275

1480
2300
1900
1450
2290

1400
2360
1890
1400
2250
12:00
1410
2300
1875
1420
2400

1440
232S
1895
1320
2400

1480
2250
1875
1310
2225

1320
2350
1900
1320
2300
1:00 pa
1410
2280
1880
1380
2200

1450
2400
1980
1280
2300

1400
2350
1980
1300
2390

1480
2400
1980
1360
2200
2:00 d>
1480
2200
1940
1340
2300

1470
2200
1830
1350
2175

1450
2250
1850
1240
2390

1300
2240
1875
1300
2425
3:00 jm
U10
2280
1910
1400
2290

1400
2410
1925
1200
2260

1360
23a
1925
1300
2230

1400
2480
2000
1480
2300
*'0u a.
1400
23S0
19XS
1300
2300

1490
2300
190!
13S0
2310

1380
2400
1985
1360
222S

1400
2280
1910
13S0
2200
i=0O am
1400
2240
1860
1480
22SO

1400
2200
I860
1320
2250

1400
2280
1313
1430
2200




13 SO
2180
6:00 ga



1280
2190




1380
2300
7:00
3:00
187S

2300
1910
1340
2225
1890
1900

2200
1850
1470
2260
1890
1825

2250
1850
1500
2280
1900
1890

2260
1850
1280
2320
1925
1880
1130
2260
1860
1400
2290
1925
1850

2240
1875
1250
2380
1950
1890
1270
2240
1865
1370
2350
1975
1890
1300
2300
1940
1400
2280
1910
2000
1400
2190
1870
1450
2300
1940
1040
1460
2280
1900
1470
2300
1940
1900
1320
2300
1910
1500
2225
1900
1940
1450
2290
1900
1460
2280
1900
1860
1425
2240
1890
1480
2300
1950
1920
1480
228U
1920
1380
2300
1950
1940
1300
225(1
1900
1380
2350
1930
1825
1290
2350
2080
1360
2220
1905
1970
1300
2300
1950
1350
2250
1900
1890

225(1
1900
1380


1980

222!.
1375
1400


2010

2230
1890
1410
2300
1910
1950

2250
1900
1230
2225
1990
1900
1400
2240
1900
1280
2260
1890
1890
1380
2300
1905
1480
2300
1900
1940
1500
2410
1990
1320
2300
1890
192!
1370
2300
1940
1480
2280
1890
1980
1300
222
1880
1300
2250
1890
1925
1370
22S0
1890

1900
1280
227'..
1925



1900
1300
2300
1940



1900
1400
2260
1915



1880
1420
227!
1910



1820
1410.




1875
1300
2300
1950



1900
1400
2200
1875



1925
1280
2175
1830




1300
2300
1900




1400
2250
1840




1220
2180
1840




1400
2260
1900




1400
2288
1925




1450
2300
1950



J .tw?*p,t,,rM * («ttnrtntfl by OISCO)
;l'n * "otiry <1 In tmoaritur*
«I. » rn*r * :n3««»*r' au« ttKMriturt
not Duet • Rtfraetary Unto duet prior to scruMtr
2-14
-------
3 ?
Table 2-3
LIQUID ORGANIC WASTE CHARACTERISTICS
Property of
Wastea
Test lb
Test 2
Test 3
Test 4
Avera
-------
3 9
at these temperatures and collects in a pool on the combustion chamber floor. A
slag tap on one side of this chancer is used to remove excess molten slag.
The secondary combustion chamber (approximately 3000 ft ) 1s baffled to promote
mixing of the combustion gases. The temperatures measured by a thermocouple
located in the hot, refractory lined duct between this chamber and the scrubber
were 1920°F, 1900°F, 1910°F and 1920°F during the tests. The retention
time in this chamber is approximately 2.11 seconds. The total retention time of
both chambers and the crossover duct is approximately 4.53 seconds.
2.4 VENTURI JET SCRUBBER
The combustion gases leaving the secondary combustion chamber enter a
refractory lined duct which provides additional residence time (approximately
0.18 second) to complete combustion. The gases are quenched as they enter the
bottom of the scrubber. The quenched gases are drawn through the two sets of
sieve trays which are irrigated with recycled scrubber liquor, Hme and caustic
solution to remove particulates and neutralize acidic gases (HC2. and S02). This
spent scrubber liquqr falls to the bottom of the scrubber and drains to a tank
which overflows to the sludge lagoon for recycle. The gaseous and fine
Particulates not removed by the sieve trays are then passed through a venturl
section of ductwork (duct reduces from 48 in. to 30 in. and expands to 48 in.)
Into which recycled scrubber water is sprayed through a jet. Makeup well water is
also injected to compensate for the loss of moisture up the stack.
A sample of these cleaned gases are extracted for analysis by continuous stack gas
monitors (CO, 02) and by Orsat (C02) to determine combustion efficiency.
ENSCO has installed an Ecolyzer Model 3103 CO analyzer (electrolytic fuel cell
type) with dual ranges of 0 to 600 ppm and 0 to 100 ppm full scale. This monitor
is used to measure CO content of the gases on the 0 to 100 ppm scale and was
calibrated using certified gas supplied by EPA (7.0, 20.1 and 41.5 ppm CO). The
set point for the CO monitor alarm is 50 ppm. This alarm went off on several
occasions and shutoff the PCB feed to the rotary kiln but these shutoffs were not
sustained for very long periods of time. CO levels exceeded 50 ppm on those
occasions when fluctuations in the liquid organic waste feedrate created a fuel
rich combustion condition which, 1n turn, resulted 1n higher CO levels. The
average CO value measured during the tests was < 10 ppm. 02 is measured using a
2-16
-------

Taylor-Servomax Model QA-269 paramagnetic type analyzer with a range of 0 to
25 percent. The set point for this analyzer was 4.5 percent to ensure there is
always excess oxygen available for combustion. Oxygen levels averaged
13.4 percent. ENSCQ operations personnel used an Fyrite analyzer to measure the
CO2 content of the gases every 15 minutes. CD2 values averaged 5.5 percent
during the trial burn. The nondispersive infrared CO2 analyzer was not yet
operational during the tests.
The gases leaving the scrubber pass through a demister pad to remove entrained
water droplets and travel up the stack and enter the atmosphere. The stack is
equipped with a drain to collect the condensate which eventually returns to the
lagoon. According to ENSCO, the mesh demister pad shrank during the trial burn
leaving a gap, which varied from 0 to 6 in. in width around one half of its
circumference, between it and the stack wall. Subsequent to the trial burn this
demister has been replaced by a Chevron type demister supplied by Munters
Corporation. Fresh makeup water is also being used in place of recycled scrubber
liquor to rinse the demister to minimize carryover of salts from the spent
scrubber liquor.
2-17
-------
V/
SECTION 3
SAMPLING LOCATIONS, EQUIPMENT AND PROCEDURES
This section of the report summarizes the locations in the process where samples
were collected, the equipment used to collect the samples and the procedures used
to operate this equipment during the trial burn.
3,1 LOCATION OF SAMPLING SITES
Figure 2-2 illustrated the location of the sampling sites in the waste treatment
process. A description of each site is presented in the following subsections.
3.1.1 Process Sampling Sites
A number of samples (inputs and outputs) from various locations in the process
streams were collected during each test. These sites were identified 1n
Figure 2-2 and are described in further detail in the following paragraphs.
3.1.1.1	Shredded Electronic Capacitors — Sample Stream 1. Samples of the
shredded electronic capacitors (fluff) were collected from the hopper of the last
harnnermi11 just before the fluff fell into the auger system. These samples were
collected by Northrop Services Inc., of Little Rock, Arkansas under contract to
ENSCO. These fluff samples were collected in 1-qt steel cans and were composited
into cans of 1-gal capacity. All cans were new, triple rinsed with hexane
(pesticide grade), dried and sealed prior to field use. The cans were sealed at
all times except when fluff was being placed in the cans.
3.1.1.2	Liquid PCBs — Sample Stream 2. Samples of liquid PCBs which were gravity
fed into this same hopper were also collected by Northrop Services. These samples
were collected in the same type of steel cans used to collect the fluff samples.
3.1.1.3	Liquid Organic Waste — Sample Stream 3. Figure 3-1 illustrates the
sampling site at which samples of the chlorinated organic waste stream (used as
part of fuel for the rotary kiln) were collected. Samples were taken from a valve
in the pipeline just prior to introduction to the rotary kiln.
3.1.1.4	Rotary Kiln Ash -- Sample Stream 4. Samples of the ash from the rotary
kiln were taken directly from the steel ash hopper as illustrated 1n Figure 3-2.
3.1.1.5	Well Water -- Sample Stream 5. Figure 3-3 illustrates the sampling site
at which samples of the well water (makeup water to scrubber and water injected
3-1
-------
Figure 3-1. Liquid Organic Waste Sampling Site
-------
Figure 3-2. Rotary Kiln Ash Sampling Site
-------

Figure 3-3. Well Water Sampling Site
3-7
-------

into kiln) were collected. Samples Mere taken from a valve 1n the well water
pipeline leading to the scrubber.
3.1.1.6	Recycled Scrubber Liquor — Sample Stream 6. Samples of the' scrubber
liquor which was recycled to the scrubber from the sludge lagoon were taken
directly from the pipeline leading to the scrubber as shown in Figure 3-4.
3.1.1.7	Lime Slurry — Sample Stream 7. Figure 3-5 illustrates the sampling site
used to obtain samples of the lime slurry. These samples were collected from a
valve in the pipeline leading from the Hme mixing tank to the scrubb'er.
3.1.1.8	Caustic Solution — Sample Stream 8. Samples of the 50 percent caustic
solution injected into the scrubber were collected from the pipeline leading from
the caustic tank to the venturl scrubber as shown 1n Figure 3-6.
3.1.1.9	Scrubber Liquor — Sample Stream 9. Samples of the spent venturi
scrubber liquor were collected from an open portion of the pipeline transporting
the liquor to the sludge lagoon as shown in Figure 3-7.
3-1.2 Stack Sampling Site
Samples of the stack gases were collected at a platform which was located eight
stack diameters downstream of the breeching entering the base of the stack and
more than two stack diameters from the stack exit.
3.2 SAMPLING EQUIPMENT
This subsection of the report details the equipment used to collect the various
types of samples during the trial burn.
3*2.1 Process Stream Sampling Equipment
Each liquid process stream (liquid organic waste, well water, recycled scrubber
liquor, scrubber liquor, lime slurry and caustic solution) sample was collected 1n
a clean, amber glass jar (500 ml) with a teflon seal 1n the cap. The Individual
500 ml samples were transferred to a clean, amber glass jug (1 gal) to obtain a
composite sample for each test. Each jug was labelled, sealed and then
transported to Mountain View for analysis. The rotary kiln ash was sampled with a
clean steel can (1 qt) attached to clean steel handle. The individual quart
samples of ash were transferred to a clean, galvanized steel garbage can (20 gal)
to composite the ash sample for each test. After thoroughly mixing the ash 1n the
can, a clean, amber glass jar (250 ml) was used to collect and transport the
composite sanple of ash to Mountain View for analysis.
The individual shredded solid capacitors and the liquid PCBs samples were collected
in new, l-qt steel cans which were subsequently composited in new, 1-gal steel cans.
3-9
-------
Figure 3-4. Recycled Scrubber Liquor Sampling Site
-------
CO
Figure 3-5
\

-------
H7
Figure 3 6. Caustic Solution Sampling Site
3-15
-------
¦ »-W, si
Figure 3-7. Scrubber Liquor Sampling Site
- "rs-MT:
«*»»*-. V: 1	/rt	, 2
- * ***-^ *¦
-------
So
3.2.2 Stack Sampling Equipment
Various sampling trains were used to sample the stack gases during the trial
burn. Each type of train is described in detail in the following paragraphs.
3.2.2.1	Particulate Matter and Hydrochloric Acid Sampling Equipment. The
particulate matter and hydrochloric acid (HC£) present in the stack gases were
sampled simultaneously using an EPA Method 5 sampling train. This sampling train
consisted of the following components:
•	A 316 stainless steel nozzle with an inside diameter of
0.4978 in. to sample isokinetically
•	A heated, 316 stainless steel lined probe, 15 ft long, equipped
with thermocouples to measure the probe temperature and the stack
gas temperature and an S-type pitot tube to measure the stack gas
velocity pressure
•	A heated, circular oven containing a 142-nm, teflon coated,
stainless steel filter holder with a glass fiber filter
•	A teflon lined, braided stainless steel hose to connect the
outlet of the filter to the inlet of the implnger train
•	An impinger train containing 5 impingers (No. 1 — 100 ml
0.1 N NaOH, Mo. 2 — 100 ml 0.1 N NaOH, No. 3 — 100 ml
0.1 N NaOH, No. 4 -- dry, No. 5 — 200 gm silica gel)
•	A vacuum hose to connect the outlet of the Impinger train to a
vacuum pump
•	A 10-cfm carbon vane vacuum pump modified for low leakage around
the shaft
•	A vacuum hose to connect the pump outlet to the control module
•	A control module to measure the temperature, pressure, flowrate
and gas volume of the sampled gases
The material collected in the nozzle, in the probe, on the filter paper and on the
front half of the filter holder was considered to be particulate matter. The
material collected on the back half of the filter holder, in the teflon hose, and
trapped in the impinger train was considered to be HC&.
3.2.2.2	PCB and RC2. Sampling Equipment. The PC8 sampling train was identical to
the train specified in the EPA manual entitled "Sampling Methods and Analytical
Procedures Manual for PCB Disposal: Interim Report, Attachment E" with one
exception. A packed tube of XAD-2 resin was put in series with the Florisll ®
tube to trap RCi (total chlorinated organics). This modification was made with
the concurrence of EPA Region 6 since the EPA manual cited previously contains no
3-19
-------
S"i
methodology to measure RC£. Figure 3-8 illustrates this modification to the
impinger train. Hence, the PCB/RCx- sampling train consisted of the following
components:
•	A 316 stainless steel nozzle with an inside diameter of 0.371 in.
to sample isokinetically
•	A heated, 5-foot glass-lined probe equipped with thermocouples to
measure the probe temperature and stack gas temperature and an
S-type pitot tube to measure the stack gas velocity pressure
•	A 316 stainless steel connector to connect the outlet of the
probe to the inlet of the impinger train
•	An impinger train containing 3 impingers, 2 traps, and 1 impinger
in series (No. 1 — 200 ml distilled water; No. 2 — 200 ml
distilled water; No. 3 — dry; a 10-cm long glass trap, 2.2 cm
I.D. packed with 30/60 mesh Grade A Florisil1® , a 10 cm long
glass trap, 2.2 cm I.D. packed with 30/60 mesh XAD-2; No. 4 —
200 gm silica gel)
•	A vacuum hose to connect the outlet of the impinger train to a
vacuum pump
•	A 10-cfm carbon vane vacuum pump modified for low leakage around
the shaft
•	A vacuum hose to connect the pump outlet to the control module
•	A control module to measure the temperature, pressure, flowrate
and gas volume of the sampled gases
The material collected in this entire sampling train was analyzed for PCB and RCfc.
3-2.2.3 N0r Sampling Equipment, the N0X sampling train used during the trial
burn conformed to the equipment specified 1n EPA Method 7 and consisted of the
following components:
•	A heated, 5-ft glass-lined probe with a plug of glass wool to
remove particulates
•	A 2-liter glass flask with a thermowell to measure flask
temperature
•	A three-way flask valve to purge the probe, evacuate the flask
and draw a sample of gas
•	A 36-in. U-tube manometer to measure the vacuum in the flask
•	A vacuum pump to evacuate the flask to 3 1n. Hg absolute
Each flask contained the acidified peroxide absorption solution specified to
collect N0x>
3-20
-------
Figure 3-8. PCB/RC£ Impinger Train
-------
6" 3
* Th» fixed oases sampling train wss
3.2.2.4	and consisted of th, following
assembled in accordance with EP
components:	, ^ 	
» olua of glass wool to remove
• A stainless steel probe with a plug
particulates
.« ~««>ve moisture from the sampled g«es
. An air cooled condenser to remove moistu
. A teflon lined dUonr^ pump to draw a sample * «.
. A T^arSbag to contain th.	°< st"k 9"" @
- to analyze the contents of the Tedlar^bag for CO,
An Qrsat analyzer was used to ana yz
CO2 and Oj.
3.3 SAMPLING PROCEDURES	^ „d output streams from the EHSCO
The procedures used to sample th
facility are described in this section.
3.3.1 Process St——' Sanmlinn Procedures	,, water, recycled scrubber
All liquid process stream	^/solution) «re sampled in the
liquor, scrubber liquor, 1 me s * ^ ^ ^ ^ M 4!ttrllct , liquid
same manner. First, each line	^	^	j*. samp 1 a
sample was Hushed for several sec	^ ^	^ w1t„ th, „qu1d
container (a clean 500-ml amber g	^	Mediately
to be sampled. A 500-«l sample 0	Hquid stream had Its own sampling
transferred to a 1-gal cwositing jug.	ttflon.,1nrt Mps when not in use.
Jar and compositing Jug *>*'^ tactl test. At th.
Each Hpuld stream was	* d MS sealed, labelled and prepared
conclusion of the test, the compositing
for shipment to Mountain View for analysis.
, similar manner. Samples of the M1n ash were
The rotary Mln ash was sampled 1. a ^	^ e>n ^	„d
collected in a clean l-«t stMl an * ^ su|p1( Ms 1nwnrtely
"Ptied twice and then a sample was • ^ en(j of , Msti , cl«a„
transferred to the 20-9a' ccnposlt ns . ^ ^pojfte sample 1n the 20-gal
stainless steel rod was used to thdroug	^ uh ,u collected 1n an
compositing can. A 250-ml sample	^	fw tMplMnt.
amber glass jar which was sealed, lab*
tw	th® liquid PC8 stream were sampled directly with
The shredded capacitor stream and the mm
the l-qt cans every 15 to 30 minutes.
3-23
-------

3.3.2 Stack Sampling Procedures
The following paragraphs and flow charts summarize the procedures used to
sample the stack gases.
3.3.2.1	Particulate Matter and Hydrochloric Acid Sampling Procedures.
Figures 3-9, 3-10, and 3*11 Illustrate the procedures used to prepare the
particulate matter/HCZ sampling train prior to a test, the procedures used to
sample the stack gases and the procedures used to recover the sample from the
train. All procedures were 1n accordance with EPA Method 5. All tests were
1 hour in length and exceeded the minimum volume of gas sampled (30 ft^). All
tests were done Isokinetically.
3.3.2.2	PCS and RC2 Sampling Procedures. Figures 3-12, 3-13 and 3-14 depict the
procedures used for the PCB/RC2 sampling train. The sampling rate during each
test was isokinetic and never exceeded the maximum rate of 0.75 cfm specified by
the test method. The sampliny test time was 4 hours which was the maximum time
recommended. At no time during a test was any train disassembled (except for
condensate removal and leak check). Finally, for each series of tests a blank
train was set up, sealed for the duration of the test, and recovered as indicated
in Figure 3-14. The purpose of this blank train was to determine whether or not
the train was contaminated during setup and takedown on the stack.
3.3.2.3	NQJt Sampling Procedures. Figures 3-15, 3-16 and 3-17 sunmarize the
procedures used to conduct the N0X tests. All tests were conducted 1n
accordance with EPA Method 7.
3.3.2.4	CO. CO*. 0; Sampling Procedures. Figure 3-18 illustrates the
procedure used to measure the CO, CQg and 0j content of the stack gases. Ho
special preparations were required apart from making sure the Tedlar® bag was
leak free and the Qrsat was filled with fresh reagents.
3-24
-------
thoroughly clean sampling
TWIN COMPONENTS IN LABORATORY
NANOGRAOE ACETONE WASH OF
SAMPLING NOZZLE. PROBE. FILTER
HOLDER
PREPARE FRESH SOLUTION FOR HC*.
TESTS
RETAIN SAMPLE OF SOLUTION FOR
BLANK
RINSE IMPMGERS ANO CONNECTORS
I WINGER NO. 1. NO. 2, NO. 3
UITH Q.1 N NaOH
IMPINGER NO. 1: 100 ML. 0.1 N NiCH
IMPINGER NO. Z; 100 ML. 0.1 N NaOH
IMPINGER NO. 3: 100 ML. 0.1 N NaOH
BUBBLER NO. «: DRY
BUBBLER NO. S: SILICA GEL
INLET TO IMPINBER NO. 1 ANO
OUTLET TO BUBBLER NO. S
INLET ANO OUTLET OF SAMPLING
NOZZLE. PROBE
TRANSPORT SAMPLING TRAIN
COMPONENTS TO SAMPLING SITE
SEAL SAMPLING TRAIN COMPONENTS
WITH TIN FOIL TO PREVENT
CONTAMINATION
0.1 N NaOH
PUT PREUEIGHCO FILTER INTO
FILTER HOLDER
CHARGE IMPINGCR TRAIN WITH
PROPER SOLUTION
Figure 3-9. Preparation Procedures for Particulate/HC£ Sampling Train
3-25
-------
5" b
ATTACH N0Z2LE TO PROBE «NO PROBE
TO OVEN
MOUNT oven ANO PROBE ON MONORAIL
ATTACH IMPINGES TRAIN TO OVEN
ASSEMBLE SAMPLING TRAIN
COMPONENTS AT SAMPLING
SITE
LEAK CHECK ASSEMBLED SAMPLING
THAI ft AT 15' H6
TURN ON PROBE ANO OVEN HEATERS
ANO AOO ICE TO IMPINGER TRAIN
TEAM LEADER CHECK WITH PROCESS
OBSERVER FOR. START TIME
REMOVE SAMPLE PORT CAP
SCREV ON BCU REDUCER
INSERT PROBE THROUGH REDUCER
PROBE POSITIONED IN STACX
AT FIRST SAMPLING POINT
RECORD CLOCK Tit®
RECORD INITIAL DRY GAS WTER
READ if. T-, T„ FOR ISOKINETIC
CALCULATE AH
SET AH AT ORIFICE PETER
READ REMAINING SAUCES
START TEST; AT OUlSMTtt
START TIME
PROCESS OBSERVER TAKE QATA
THROUGHOUT THE TEST
RECORD QATA OK FIELO OATA
SHEET AT EACH POINT EVERT
i MINUTES
« POINTS PER TRAVERSE
z traverses at vr
5 MINUTES PE* POINT
SAMPLE EACH POINT ON TRAVERSE
STOP SAMPLING AFTER COMPLETING
TRAVERSE ANO REMOVE PROBE P*OH
STACX
RECORD FINAL DRY
SAS METER READING ANO
LEAK CHECK
TRANSFER SAMPLING TRAIN TO
NEXT SAMPLE PORT ANO REPEAT
PROCEDURE
PROBE AND OVEN HEATERS
AT ZSO°F
ZERO MAGflEHELIC GAUGES
RECORO LEAK RATE ON
FIELD DATA SHEET
PROCESS OBSERVER MAKES SURE
PROCESS 0PERATIN6 NORMALLY
CONNECT UMBILICAL TO CONTROL MOOULE
CONNECT FILTER HOLDER OUTLET TO
IMPINGER NO. I
CONNECT IMPINGER NO. 5 OUTLET
TO PUMP
CONNECT PUMP TO CONTROL MOOULC
Figure 3-10. Sampling Procedures for Particulate/HC£ Sampling Train
3-26
-------
TEFLON HOSE,
IMPIWGERS NO. 1
SILICA GEL
FILTER AND FILTER
HOLOER
NOZZLE, PROBE
IMSII AND BRUSH UITH
NANOGRAOE ACETONE
UASH WITH 01STILLED
MATER
TRANSFER WASHINGS
TO LABELLEO
POLYETHYLENE BOTTLE
SEAL WASHINGS IN
LABELLEO POLVETHY
LENE BOTTLE
TRANSFER SOLUTION
TO LABELLED
POLYETHYLENE BOTTLE
WASH FIITER HOLDER
WITH NANOGRAOE
ACETONE
WEIGH. RECORD
DISCARD
SEAL FILTER IN
LA8ELLED
PETRI DISH
MEASURE VOLUME
AND RECORD
Figure 3-11.
Sample Recovery Procedures for Partlculate/HCl
-------
IHM CIASSMM
M I Mi tlMISK tMtMM
wart
HUH MSMrW Mm
•ma miim Mniiia oiu
¦t«si mwr		
UK gini		
«rric M 4Mft IK(K
mum R.IIM.! MM
mm*, vn* MSflUM Mil
prtm m. nun
irwii m. « sat left an
t» HBO Men Iff
Mm
MM IV
mm mmuui wiiii mimm
mm mwmi act Mai. inm
MM WMOMOt MUM
•MSI
M MT MP MMCI
vim iismui huh
	*tM«
MMK
mmm mi • MSftttn
tttvci MM MMK
MM Its Sim MM
Mftl
KCIM IMK VMM ft
1» MWII m. 9
M fM MHII M
9MP TVR % Ml
1-tf

OtM M SMMlll MM MtM
IM H MM!
MV M MOM! Mil * Ml (
MM IM t MS M

CUM M SM.lt MMMM IM 14 MMtl
OtM M jnt(l M1N Ml MM IM H MMIS
OIMI M SBHtIV MM MS MIM/M. MMM IM M tMMlS
Ml M MOMM MtM



u\
Preparation frtcHml f«r
rca/DCt Infl Inf Tr*lm
-------

attach nozzle to probe and probe
TO 1OT1NSER HO. 1
MOUNT IMPINGER CASE AND PROBE
(Iff MONORAIL
ASSEMBLE SAMPLING TRAIN
COMPONENTS AT SAMPLING
SITE
CONNECT UMBILICAL TO CONTROL MODULE
CONNECT PROBE OUTLET Tft JMPINSER NO. 1
3SB PES
IMPINGER MO. 3
CONNECT TUBE TO OUTLET OF JMPINSER
NO. A
CONNECT FLORISIL i JCAO-2 TUBES WITH
•IT CONNECTOR
CONNECT OUTLET OF IMPINGER NO. 4 TO
PUMP
CONNECT PUMP TO CONTROL MODULE
ZERO MAGMEHELIC SAUGES
PROBE HEATER AT 250°F
LEAK CHECK ASSEMBLES SAMPLING
TRAIN AT 13" HG AT FRONT OF NOZZLE
RECORD LEAK RATE ON
FIELD OATA SHEET
TURN ON PROBE AND OVEN HEATERS
ANO ADO ICE TO IMPINGER TRAIN
TEAM LEADER CHECK WITH PROCESS
OBSERVER FOR START TIME
PROCESS OBSERVER MAKES SURE
PROCESS OPERATING NORMALLY
RECORO CLOCK TIME
RECORO INITIAL DRY SAS METER
READ 4P. T., T„ FOR ISOKINETIC
CALCULATE AH
SET «.H AT ORIFICE METER
*EAO REMAINING GAUGES
PROBE POSITIONED IN STACK
AT FIRST SAMPLING POINT
REMOVE SAMPLE PORT CAP
SCREH ON ULL REDUCER
INSERT PROSE THROUGH REDUCER
START TEST AT DESJWATBJ
START TIME
PROCESS OBSERVER TAKE DATA
THROUGHOUT THE TEST
RECORO DATA ON FIELD OATA
SHEET AT EACH POINT EVERY
20 MINUTES
SAMPLE EACH POINT ON TRAVERSE
3	POINTS PER TRAVERSE
4	TRAVERSES AT S0°
20 MINUTES PIR POINT
STOP SAMPLING AFTER COMPLETINS
TRAVERSE AW) REMOVE PROBE FROM
STACK
RECORO FINAL ORY
SAS METER READING ANO
LIAS CHECK
TRANSFER SAMPLING TRAIN TO
NEXT SAMPLE PORT ANO REPEAT
PROCEDURE
figure 3-13. Sampling Procedures for PCB/RCA Sampling Train
3-31
-------
WASH WITH
NANOGRADE lit KANE
AND ACETONE
TRANSFER UASHINGS
TO LABELLED GLASS
BOTTLE AND S£AL
SILICA GEL
SEAL MASillNGS IN
LABELLED GLASS BOTILE
SEAL Willi GLASS
CAPS AND LABEL
UEIGII, RECORD
DISCARD
FLORISIL AND XAD-2
MASH Ml III
NANOGRADE IIEXANE
AND ACETONE
IHPlNGiKS NO. I
TRANSFER SOLUTION
TO LABELLEO
GLASS BOTTLE
REMOVE FROM
IMPINGERS
MEASURE VOLUME
AND RECORD
Figure 3-14. Sample Recovery Procedures for PCB/RCl Sampling Train
-------
THOROUGH* CI (AN
FLASKS AND VMVFS
KIAII SAHPLi OF SOLUTION

PREPARE FRESH A8SOARING

ACIDIFIEO H20?
FOR HANK

SOLUTION

PIPETTE 25 HI. Of
ABSWUNG SOiUIIOM
INIO EACH FLASK
U
•
u
u
SEAL FLASK *AL¥E
TRANSPORT FLASKS
TO SAMPLING SITE
Figure 3-15. Preparation Procedures for NOx Sampling Train
-------
(o %
ASSEMBLE THE SAMPLING
TRAIN COMPONENTS AT
THE SAMPLING SITE
CONNECT PROBE
TO FLASK
RECORO FLASK ANO VALVE VOLUME
RECORO FLASK TEMPERATURE
RECORD BAROMETRIC PRESSURE
RECORO FLASK PRESSURE
PURGE THE PROBE
WITH STACK GAS
TURN FLASK VALVE TO
SAMPLE POSITION ANO
SAMPLE CASES
TURN FLASK VALVE TO
PURSE POSITION
SHAKE FLASK FOR
5 MINUTES ANO
PREPARE FOR NEXT TEST
EVACUATE FLASK TO LESS THAN
3" nq ABSOLUTE PRESSURE
TURN ON PROBE HEATER
CHECK FOR LEAKS
CONNECT flask to MANOMETER
ANO VACUUM
Figure 3-16. Sampling Procadures for NOx
3-34
-------
L 3
RECORD BAROMETRIC
PRESSURE
RECORD FLASK TEMPERATURE
,£T FLASK SIT FOR
MINIMUM 16 HRS
TRANSFER FLASK CONTENTS
TO SAMPLE BOTTLE
RINSE FLASK TWICE WITH
01STILLED WATER.
ADD TO SAMPLE BOTTLE
SHAKE FLASK
FOR 2 MIN
SEAL ANO LABEL BOTTLE
FOR SHIPMENT
CONNECT FLASK TO MANOMETER!
RECORD PRESSURE '
^gure 3-17. Sample Recovery Procedures for NOx
3-35
-------
£ V
COWKCT »«08t TO
Q.ASS CONOENSATE TUP
ASSEmLE SAWUIM rut*
COWONENTS AT SAWLJNO SITE
EVACUATt TIBUAP
MS AM SEAL Off WITH OW
mm ON PROSE HCATtX AMI
M» ICC TO C0N0EHSATE TRAP
POSITION PNOBE AT POINT NO. J
IN STAC*
UWU UNTIL MS IS ruu.
sim. qff us ano man
KUASE CUW ON TtOUIX MS
CONNECT	OUTLET or CONDENSATE
TW TO INLET Of PUMP
CONNECT POTP TO KOTAKCTTIt
CONNECT	ROTAMETER TO TEOLAR M6
Figure 3-18. Sampling Procedures for C02, 02t CO
3-36
-------

SECTION 4
ANALYTICAL PROCEDURES AND RESULTS
*	the procedures used to analyze the saaples
This section of the report deta	s the ^ults of these analyses in
collected during the PCS trial burn and discus
detail.
4.1 INTRODUCTORY AND GENERAL APPROACH
'•1 INTRODUCTORY ANU tacnow*. 		
pC8s are a complex family of compounds and Isomers (Figure 4-1) (£). Aroclors are
mixtures of PCSs defined by the average number of chlorine atoms par molecule and
isolated by boiling point range. A comnon analytical procedure utilizes
recognition of unique gas chromatographic patterns for given Aroclors. Industrial
treatment such as heating, mixing with other materials, or environmental aging
(weathering) can change the Isomeric distribution of a given Aroclor. Using
Pattern recognition to identify given Aroclors can be a formidable task. To
overcome this problem, the compounds can be perch!ormated to decachloroblphenyl
CDC8) (x and y in Figure 4-1 both equal 5). Blphenyl (the carbon skeleton of PCBs
without any chlorines) will also be chlorinated and must be determined
independently. These considerations determined the analytical approach to the
samples analyzed for this project.
F^gure 4-1. seneral Formula for Polychlorlnated Blphenyls
4-T
-------
C b
quantitative analysis of PCSs presents a unique challenge for the analytical
chemist. There are 209 potential PCS compounds and isomers, each one having Its
own chromatographic properties. The formidable problem of quantitation has been
limited to three discrete methods. This section describes these methods and
discusses some of the advantages and disadvantages of each.
The three methods used for quantitation are:
•	Pattern recognition
•	Derivatization
•	Measurement of individual PC8 components
4.1.1 Pattern Recognition
PC8 mixtures (known as Aroclors in the United States) are formulated for different
industrial purposes. A fortunate coincidence of this production phenomena is the
gas chromatographic profile of the different formulations. Each of the Aroclor
mixtures, when analyzed hy gas chromatography, has a unique elution pattern. This
observation has led to an analytical technique known as pattern recognition.
Pattern recognition, described 1n detail by Hutzlnger (JL) affords the following
advantages:
•	Identification of the Aroclor mixture which may 1n turn help
localize the source of environmental contamination
•	May be accomplished by the use of relatively simple, nonexpensive
gas chromatographic equipment
•	Has the potential of low artifact production, which enables more
accurate quantitation 1f the samples are relatively "clean"
Pattern recognition has become an established technique. The analytical chemist,
however, must be aware that this method has definite disadvantages when applied to
certain types of samples. Samples derived from combustion products and
environmental samples that have undergone "weathering" may not be amenable to
pattern recognition techniques. Sanples containing chlorinated pesticides or
mixtures of Aroclors may also pose a problem when accurate quantitation 1s
necessary. These compounds have the same extraction characteristics as PCBs, and
may co-chromatograph with PC8s, altering the Aroclor pattern and frustrating
pattern recognition and quantitation.
4-2
-------
Q 7
The disadvantages may be summarized as follows:
•	Samples from combustion products and "weathered" samples may have
undergone chemical changes and pattern alterations, thus
obscuring efforts to use patterns for Aroclor number assignments
•	Because of Interferences from other sources the quantitation of
PCBs is subject to significant error
4.1.2 Derivatlzatlon
Derivatization theoretically simplifies the analysis of PC8s by converting all 209
homologs to a single chromatographable species. The compound formed, DC8, is a
fully chlorine-substituted blphenyl that is prepared by a reaction involving
antimony pentachloride. The perchlorination technique, originally proposed by
Berg and later by Armour (£), has been used with success and has inherent
advantages.
•	Lowers the detection limit of the analysis
•	Gives a "cleaner" analysis by conversion of PCB homologs to one
species
•	Yields unequlvlcal quantitation of OCB
As the title of the Armour paper suggests and the text reinforces, this procedure
should be used only as a confirmation technique and the quantitative accuracy of
the procedure 1s highly dependent on the test sample PCB mixture. Correction
factors for OCB analysis, to convert back to the original weight of PCB present.,
may range from approximately 0.4 for Aroclor 1221 to 0.80 for Aroclor 1260.
The factor cannot be known precisely without accurate knowledge of the actual
homolog distribution within the original perch!orinated sample. There is a large
weight change 1n the conversion of PCBs to DCS. This large weight change 1s
caused by the addition of chlorine atoms to the precursor PCB homolog molecule.
The number of chlorine atoms added 1s* of course, dependent upon which homolog 1s
being derivatized and the original number of chlorines present 1n the molecule.
Thus a reported DCB number could be twice as high as the actual PC8 level present.
Samples obtained from combustion sources will definitely have homolog distribution
changes when compared with the profile of the original feed material. These
distribution changes make 1t Impossible to estimate the actual contribution from
real chlorinated biphenyls by a factor of any less than two.
4-3
-------
C f
Additional errors that may be incorporated into the derivatization process include
artifact DCS production. Haile (3) has documented DCS methodology applied to
combustion sources that resulted in false positives. Confirmation of PCS in a
duplicate, nonperchlorinated aliquot suggested that the DCS measured in the sample
was an artifact of non-PCS combustion products. To date, no confirmed explanation
of the origin of this artifact has been presented.
The DCS methodology will, in all cases, overestimate the sample concentration of
PCS. This fact should be recognized and experimental design should be approached
with regards to correction mechanisms to reflect real PCS levels.
4.1.3 Measurement of Individual PCB Components
A third approach used to quantltate PC8 materials avoids the errors intrinsic to
both derivatization and pattern recognition. This method, proposed by Webb and
McCall (4) was partially incorporated into the Federal Register and was recoimiended
for use when pattern recognition- seemed to be nonapp11 cable. Additional work
published by Eichelberger, Harris and Sudde (1) laid the groundwork for an
alternate approach using GC/MS. A detailed explanation of this methodology has
been published in EPA Report 600/7-79-047 entitled "Measurenent of PCS Emissions
from Combustion Sources."
Briefly, the method measures each PCS peak, determines the number of chlorines 1n
the compound representing the peak, then applies a response factor derived from a
PCS having the same number of .chlorine substituents to that peak. This procedure
is applied consecutively to each PCS peak 1n a sample chromatogram and the
resultant weights are summed yielding the total weight of PC8 1n the sample
Injection. The use of GC/MS analysis with this methodology results in the most
accurate determination of PCB levels. The current availability of capillary gas
chromatographic systems that are Interfaced directly Into the mass spectrometer
promises to update the state-of-the-art for PC8 analysis.
Capillary gas chromatographic systems are generally characterized as more
sensitive than packed column systems and possess far better resolution. The
combination of these factors, when used with mass spectrometry should, 1n theory,
be the most accurate quantitation procedure available.
4-4
-------

Advantages of the measurement of Individual PC8 components method are as follows:
•	Accuracy does not depend on correct identification of Aroclor by
pattern recognition; therefore, weathered samples and samples
derived from combustion sources may be quantltated as accurately
as pure standards
•	Eliminates the potential for false positive Identifications that
would occur as a result of perchlorination
Disadvantages
0 Ultimate accuracy will be achieved only by use of GC/MS which is
expensive and requires highly trained staff
•	Overall detection limits are higher than 0C8 and GC methodologies
•	Requires numerous standards for determining response factors
•	Requires extensive data reduction
The quantitation procedure for PC8s by perch!orination, as recomnended by EPA (6),
was developed by Armour for detection of PCS residues 1n biological materials. To
achieve accurate numbers, the actual percent chlorine 1n the particular PCS
mixture must be known. A close approximation can be obtained if a particular
commercial mixture pattern can be identified. Prior to the development of the
perchlorination procedure, pattern recognition and Integration of multiple GC
peaks by various formulas was the preferred method to quantltate PCBs. These
methods have the conmon analytical problems associated with determining response
factors, integration of peaks, potential for overlapping peaks from compounds of
no interest, and the problem of changing Isomer ratios eaused by environmental
factors.
It was then decided to use the DC8 methodology to determine the potential presence
and approximate auantity of PCBs, and to verify the presence of PCBs by GC/MS
identification of specific isomers.
The general analytical approach (Figure 4-2) was to:
•	Extract the organic fraction from a given sample
•	Perch!oHnate a portion of the raw extract
•	Analyze for KB
4-5
-------
10
DCS PRESENT
STOP
STOP
CONCENTRATION
SPLIT
SAMPLE
EXTRACTION
WITH HEXANE
QUANTITATE
8IPHENYL 8C/FID
PERCHLORINATION
RAW EXTRACT
SC/EC ANALYSIS
FOR OCB
GC/MS VERIFY
PC8 & BIPHENYL
QUANTITATE DCS
Figure 4-2. General Analytical Scheme for PCB Samples
4-6
-------
71
If XB is detected then:
•	Analyze for and quantitate blphenyl in the raw extract
•	Verify the presence of PCS compounds in the raw extract by SC/MS
•	Quantitate the DC8 level in the perch 1orinated fraction
4.2 ANALYSIS OF PC8s
This section describes the analytical schemes used to analyze for PCBs in the
samples of kiln ash, scrubber liquor, recycled scrubber liquor from the sludge
lagoon and stack gases. In all cases the samples collected on test day No. 1
represent a "background" test burn condition during which capacitors containing no
PCBs were incinerated. The incineration conditions were essentially the same for
the background test on day No. 1 as for days No. 2, No. 3, and No. 4 during which
time PCBs were incinerated.
4.2.1 Rotary Kiln Ash Analysis
Figure 4-3 illustrates the analytical scheme for the k'Hn ash. Kiln ash was
received in the laboratory as bulk solid material in a 250-ml antoer glass jar. By
visual inspection, approximately 90 percent of the bulk was small black flakes
ranging in size from approximately 2 nm to a fine powder. A few large pieces (up
to 75 rmi by approximately 20 urn) of metallic-like material were also present.
A representative sample of kiln ash (approximately 50 gm) was transferred directly
from the sample container to a preweighed soxhlet thimble. The extraction was
carried out in accordance with the soxhlet extraction technique '1n attachment E of
the EPA manual. Following extraction and concentration to 2 ml, the extract was
divided into 1-ml portions. One ml of the extract (approximately 50 percent of a
50-gm sample) was submitted to perch 1or1nation 1n accordance with the EPA report
entitled "A Method for Sampling and Analysis of PolychlorInated Biphenyls (PCB's)
1n Ambient Air." Following the derlvatization, the kiln ash samples and method
blank were analyzed by gas chromatography/electron capture (SC/EC). The first
GC/EC runs showed peaks throughout the chromatogram that were out of the normal
working range of the detector. The original perch!orlnated fractions were far too
concentrated and were subsequently diluted (1000 to 1). This simple dilution
enabled the analyst to resolve the 0C8 peak from other peaks in the chromatogram
and to quantitate the DCB peak.
4-7
-------
7 X
SPLIT
SAMPLE
CONCENTRATE TO
2 MLS
SOXHLET EXTRACTION
w/HEXANE
50g
KILN ASH
Positive
1 dcs r
SC/EC ANALYSIS
FOR DCS
PERCHLORINATION
GC/EC ANALYSIS
FOR PCS
O.S ML
FLORISIL CLEANUP
GC/MS ANALYSIS
FOR PCB
0.5 ML RAW
CONCENTRATE
GC/FIO.QUANTITATION
OF BIPHENYL
Figure 4-3. Analytical Scheme for Rotary Kiln Ash
4-8
-------
73
Due to the presence of the DC8 peak in the kiln ash samples, underivatized portions
of the raw extracts were submitted for biphenyl analysis by gas chromatography/
flame ionization detection (GC/FID) and confirmation by GC/MS. Mass spectra are
shown in Appendix A. A second portion of the raw extract was submitted for
Florisi t® cleanup and GC/EC analysis. A PCB pattern was tentatively identified
and later confirmed by capillary GC/MS analysis.
Analysis of the 0.5-ml portion of raw extracts (underivatized, no Florisil®
cleanup) by capillary GC/MS verified the presence of PCBs ranging from the mono
isomer to the penta isomer. Higher chlorinated isomers may have been present, but
were not detected. Table 4-1 sumnarizes the results of all the analyses performed
on the rotary kiln ash sanples.
4.2.2 Scrubber Liouor Effluent and Recycled Scrubber Liquor Analysis
Figure 4-4 illustrates the analytical scheme for the scrubber liquor. Liquor
samples were received in the laboratory in 1-gal amber glass jugs. The jugs were
thoroughly shaken (due to the fact that approximately half of the volume of the
jug was a murky-grey sediment) prior to removing a 1-liter sample for extraction.
The extraction procedure was conducted according to Attachment B entitled "A
Tentative Method of Testing for Polychlorinated Biphenyls 1n Water" found in the
EPA manual. After completion of the extraction and concentration to
2 milliliters, a visual evaluation was performed to determine the necessity for
further chromatographic cleanup procedures. Every sample (8) in the group was
judged appropriate for Florisi 1® cleanup (the sample was colored) and were thus
treated in accordance with Section 10.3 of Attachment E of the EPA manual.
In the case of scrubber liquor, the concentrated extract was divided into two
equal fractions (1.0 ml each), one of which was submitted to Florisil®
chromatography. For the recycled scrubber liquor, the entire concentrated
fraction^s submitted to Floris 11 ^chromatography. Following the cleanup with
Florisi nS', the fractions were concentrated on a Kuderna-Danish evaporator and
divided into two equal samples. These samples (representing one-half of the
recycled scrubber liquor, and one-fourth of the scrubber liquor) were
perchlorinated using the same procedure as for the perchlorination of kiln ash
samples. In most samples a dilution of 1 to 100 was required to achieve a stable
baseline on the recorder of the electron capture gas chromatograph. The samples
required dilution prior to analysis to adjust for the unavoidable matrix resulting
4-9
-------
7V
Table 4-1
ROTARY KILN ASH ANALYSIS AS DECACHLOROBIPHENYL
Corrected DC8 Value (ppm)a
DC3 Value (ppm)b
Biphenyl Value (ppm)c
Wet Weight Kiln Ash Extracted
(gm)
Percent dry weightd
Day #1 Day #2 Day #3 Day #4
0.52 41.94 20.15 27.21
1.25	44.63 24.09 28.98
0.73
2.73
4.00
1.78
50.0744 49.7722 50.2641 50.1284
98.7
98.0
95.4
99.0
Corrected DCB value using Section 23.7 from "A Method of Sampling and
Analysis of PCBs in Ambient Air," EPA Report 600/4-78-048 for presence of
biphenyl
^Uncorrected DCB value including biphenyl
cCorrected to DCB and for biphenyl conversion efficiency of 43.3 percent
4-10
-------
76"
11 SCRUBBER
EFFLUENT
SEPARATORY FUNNEL
EXTRACTION WITH
HEXANE
CONCENTRATE
TO 10 ML
AND SPLIT
5 ML FL0RI5IL
CLEANUP ANU
SPLIT
RAH EXTRACT
POSITIVE
PERCHLORINATION
RAW EXTRACT
DCS
SC/FID SIPHENYL
QUANTITATION
GC/EC ANALYSIS
FOR 0C8
GC/FIO BIPHCNYL
QUANTITATION
GC/M5 ANALYSIS
FOR 0C8
PERCHLORINATION
SC/MS
VERIFICATION
S ML
RAU EXTRACT
CONCENTRATE TO
2 ML ANO
SPLIT
1L RECYCLE
SCRUBBER
EFFLUENT
CONCENTRATE TO
2 ML ANO SPLIT
CONCENTRATE TO
10 ML ANO
FLORISIL CLEANUP
SEPARATORY FUNNEL
EXTRACTION WITH
HEXANE
LIQUID SAMPLES
Figure 4-4. Analytical Scheme for Scrubber Liquor and Recycled Scrubber Liquor
4-11
-------

(jfr
after extraction and FT cri si I ^-/cleanup. The levels of DCS found were at the
limit of detection for the electron capture detector and in some cases were
outside the boundary of responses from the usual 20-pg standard of DCS used as the
lower limit.
(R)
Unperchlorinated, F1orisi 1 ^-treated fractions were submitted for GC/FID
analysis. Blphenyl was undetectable within the sample matrix. The low levels of
DCS detected with the electron capture detector were out of the sensitivity range
of the flame ionization detector. A chromatogram of the scrubber liquor is shown
1n Figure 4-5. The results of the GC/EC analysis for DCB are shown in Table 4-2.
4.2.3 Stack Sas Analysis
Figure 4-6 illustrates the analytical scheme for the PCB train used to sample the
stack gases. The PCS train samples were received from the field 1n three separate
containers. Florlsil^and XAD-2 tubas were sealed outside the ground glass
joints with stretched parafilm and were contained within 2-l1ter wide mouth nalgene
bottles. Impinger catches were received 1n 2-Hter amber glass bottles. Probe
rinses and impinger washes (1 to 1 hexane/acetone) were collected in 500-ml amber
glass bottles. 81ank trains were prepared, were recovered each day, and were
shipped in identical containers. Blank Florlsil^and XAD-2 tubes were shipped
in the same wide mouth nalgene bottle as the corresponding test day sample tubes.
Sample extraction procedures were followed as prescribed in Attachment E of the
EPA manual. The extraction procedure included soxhlet extraction of combined
XAD-2 and Florisil®tubes, separatory funnel extraction of the Impinger catches
with hexane and a 20-ml portion of acetone used as a rinse, and the hexane portion
of a water wash of the hexane-acetone impinger and probe rinse fraction in a
1-liter separatory funnel. These extracts were combined, cleaned and concentrated
as outlined in Section 7 of Attachment E. The final volume was adjusted to
40 "ml. The 40-ml concentrate was then divided to two portions — a 35-ml portion
that was used'for PCB analysis and a 5-ml portion used for total organic chloride
(RCi.) analysis. The 35-ml portion was further concentrated to 5-ml. A 3-ml
portion was submitted for perchlorlnation as outlined in Section 3.1 of
Attachment E. The final 2-ml portion was stored at 4°C for further analysis.
Analysis of the perch1 or1nated PCB train samples by GC/EC required dilution by a
factor of 100. Results of the DCB analysis is shown in Table 4-3. Representative
4-12
-------
7 7
3 mP 74-If*/ ?/*»'*:/ S.ff.

is Pi?
20. 35
28:
*§
§§:
H
i§:
U
31.
74
4-13
-------
7 2
Table 4-2
RESULTS OF SCRUBBER LIQUOR AND RECYCLED SCRUBBER LIQUOR ANALYSIS AS
0ECACHL0R08IPHENYL
Scrubber Effluent Recycle Scrubber Effluent
Dav #1 Dav #2 Dav #3 Dav #4 Oav #1 Dav #2 Dav #3 Day #4
GC/MS	UUUU	-	UUU
Confirmation
of PC3
DCS (ppb)	£4 ^4 <7 ~11	ND	ND <2 NO
Dilution	1:100 1:100 1:100 1:100	1:10 1:10 1:100 1:10
Sample Size	1L 1L 1L 1L	1L	1L 1L 1L
Portion of Sample 1/4 1/4 1/4 1/4 1/2 1/2 1/2 1/2
Perch 1 orinated
U » unconfirmed
4-14
-------
79
SODIUM BIPHENYL
REDUCTION
CHLORIDE ION
ELECTRODE ANALYSIS
3 ML PERCHLOR1NATION
GC/ECD DCS
QUANTITATION
POSITIVE
DCS
SC/MS VERIFICATION
OF OCB
ORGANIC CHIORIOE
COMBINED EXTRACTS
ORGANIC
LAYER
ORGANIC
LAYER
35 ML
CONCENTRATE TO
S ML AND SPLIT
XAD-2 RESIN
SEPARATORY FUNNEL
HEXANE EXTRACTION
IMPINGER SOLUTIONS
WATER
SEPARATORY FUNNEL
HEXANE EXTRACTION
SC/MS VERIFICATION
OF PCS
CONCENTRATE TO 40 ML
AND SPLIT
IMPINGER ANO
PROBE RINSES
(HEXANE/ACETONE)
FLORISIL RESIN
GC/FID BIPHENYL
QUANTITATION
PCB
TRAIN
SOXHLET EXTRACTION
(HEXANE)
2 ML
HOLD PENDING DCB
ANALYSIS
COMBINED
SORBENTS
Figure 4-6. Analytical" Scheme for PCB Stack Gas Sampling Train
4-15
-------

Table 4-3
RESULTS OF STACK GAS ANALYSIS

Oay #1
79-3472
(ug/train)
Day #2
79-3474
(uq/tra1n)
Oay #3
79-3476
(uq/train)
Day #4
79-3478
(uo/train)
Capillary 6C/MS for
PCSs
NO
NO
NO
NO
0C84— GC/EC
35
120
63
167
DCS - GC/MS
Confirmed
DCS
Confirmed
OCS
Confirmed
DCS
Confirmed
DC8
Biphenyl by GC/FIC&
~
~
*
*
31phenyl by GC/MS
NO
3
1
NO
Corrected DC3 levelsc
35
116
61
167
ajha equation used to obtain the above GC/EC results 1s
AT:ClnjKtMRF * Aoount s*,# Volu"» *7*1
^Detected at or below the level of quantitation (detection limits: 50 pg
GC/EC, 100 pg GC/MS)
C81phenyl levels subtracted when detected and quantified by GC/MS
4-16
-------

Chroroatograra of Perchlorinated PCB Train Sample
Figure 4-7.
-------
fx
chromatograms are shown in Figure 4-7. The DCB levels observed in these analyses,
after subtraction of mean PCB blank train values of DC8, indicated a need for
biphenyl correction. GC/FID analysis for biphenyl were performed on an
underivatized portion of the PCB train samples. Chromatograms obtained for these
analyses were inconclusive and suggested that further analysis was in order.
GC/MS analysis of the perch 1 orinated fractions confirmed the presence of DCB (see
Figure 4-7) in approximately the concentration range as measured with GC/EC
analysis. Capillary GC/MS chromatograms of underivatized samples were evaluated
for the presence of biphenyl and for PCBs. This evaluation suggested the biphenyl
was present at very low levels, but no PCBs could be detected in the underivatized
samples. A GC/MS procedure was used for quantitation of biphenyl. The results of
this analysis is shown in Table 4-4. Mass spectra are shown in Appendix A.
4.3	ANALYSIS FOR TOTAL ORGANIC CHLORINE
A portion of the PCB train concentrate (5-ml of 40-ml total) was submitted to
total organic chlorine analysis. Ligget's (7J sodium biphenyl reduction method
was used.. The sample was solvent exchanged from hexane to halide free toluene,
then reduced with sodium biphenyl. After cleanup and decomposition, the sample
was subjected to chloride analysis with a chloride specific electrode. A method
blank and a standard Aroclor 1260 sample were run for QA/QC.
The detection limit for the method was determined to be 0.5 mg total organic
chloride in the train. There was no organic chloride detected in samples from
days 2 through 4. A slight positive was found for day 1 (approximately 0.9 mg).
This is believed to be a contamination problem. The chloride spike and recovery
sample gave 100-percent recovery. The standard Arclor sample gave 96 percent
recovery based on a literature value of 60 percent wt/wt chlorine in Aroclor 1260.
It is concluded that no organic chlorine was collected in the PCB train at the
detection limits of the methodology.
4.4	ANALYSIS OF PARTICULATE MATTER
Analysis of'the particulate matter collected in the stack gas sampling train was
in accordance with EPA Method 5. All filters and acetone residues were desiccated
and weighed to the nearest 0.05 mg on a Mettler Model H51AR analytical balance-
Blanks were treated in an identical manner. Table 4-5 sunmarlzes the results of
these analyses.
4-18
-------
$2>
Table 4-4
BIPHENYL ANALYSIS OF PCS STACK SAS SAMPLING TRAIN SAMPLES BY SC/MS
Acurex
I.D.
79-3479
Identification
Area	yg/Train Biphenyl4
Counts uq Biphenvl/train Corrected to DCS
79-3472
PCB
Train Day #1
ND
ND
NO
79-3474
PCB.
m
Train Day #2
10315
2.68
3.87
79-3476
PCB
Train Day #3
4440
1.15
1.166
79-3478
PCS
Train Day #4
ND
ND
ND
79-3477b
PCB
Train Blank
526
0.14
0.20
Day #3
PCB Train Blank
Day #4
470
0.12
0.17
Note: Biphenyl conversion efficiency to DCB is 42 to 48 percent
aAmount Blphenyl x x 0.45
bTwo of the four blank trains were selected at random for analysis.
4-19
-------

Table 4-5
ANALYSIS OF PARTICULATE MATTER, HCi, NOx and RCia


Oav #1
Day #2
Oav #3
Oav #4
Total
particulate (mg)
420.21
1351.31
820.63
1198.05
Total
chloride (ragj
97.0
170.0
450.0
320.0
Total
NOx (ug fOz)b
1.5
5.2
40.0
20.0

52.0
78.0
60.0
27.0


71.0
6.3
9.4
91.0


27.0
72.0
37.0
is.a
Total
RC ft
NO
NO
NO
NO
(rag organic chlorine)0
Results are corrected for blanks
^Micrograms of nitrogen dioxide
cLower detectable limit 1s 0.5 mg
4-20
-------

4.5	ANALYSIS OF HYDROCHLORIC ACID
Analysis of the HC2 collected in the irapingers of the particulate sampling train
was done according to the argentometric method described in the 14th Ed'ition of
"Standard Methods for Examination of Water and Wastewater" published by the
American Public Health Association in 1975. The total volume of impinger liquid
was measured and a representative aliquot of the liquid was titrated with silver
nitrate to determine the chloride content of the liquid. A blank was also run on
the distilled water used in the Impinger trains. Table 4-5 summarizes the results.
4.6	ANALYSIS OF N0X
Analysis of the N0X samples collected during the trial burn were done in
accordance with EPA Method 7 — the phenoldisulfonic acid procedure. A blank of
the absorbing solution was also analyzed. The results are summarized in Table 4-5.
4.7	QUALITY CONTROL AND QUALITY ASSURANCE
Quality control/quality assurance measures necessary to insure the integrity of the
ppb level analysis of PCBs require strict attention to the linearity of the
analysis (i.e., calibration curves) and to the quality of solvents and reagents
used in the sample preparation stage.
In addition to Acurex laboratory quality control measures, PCS audit samples were
supplied by William D. Langley, Ph.D. of the Environmental Protection Agency
Region 6, Houston, Texas (see Appendix B). Acurex performed and passed this audit
check which consisted of determining Aroclor mixtures (qualitative) and
establishing the PCB concentrations (quantitative) in each of these liquid samples
prior to conducting any analysis of the field samples.
Solvent purities, based on analysis of a representative sample from identifiable
lot numbers were quality checked by GC/MS. Samples, representing 250.ml of
solvent concentrated to 1 ml on a Kuderna-Danlsh evaporator, were submitted for
analysis. Electron Impact spectra were collected during the entire programmed
chromatographic run. Diagnostic PCB 1ons were plotted as a function of time to
determine if any PCB compounds were present. This procedure allowed evaluation of
solvent purity to the low ppb level. All solvents used 1n the analysis were
judged "PCB free."
4-21
-------
?(>.
Method blanks were used to further ensure solvent and reagent purity. These
samples were prepared 1n an identical manner to the perchlorinated field samples
and showed no contribution at the retention time of KB. Method blank
chromatograms are shown in Figure 4-8.
Perchlorination efficiency was monitored to insure the ability to quantitatively
derlvatize known standards. Results of these perch!or1nation experiments
established the average efficiency of the reaction as 100 percent. The Aroclor
preparation used in these experiments was Aroclor 1260 prepared in 1 mg/ml stock
solution s supplied by the EPA. Literature values indicate Aroclor 1260 is
composed of approximately 60 percent C£, with an average of 6.3 Ci/molecule and an
average molecular weight of 372. Therefore 74.5 percent of the weight of DCS
detected represents the weight of Arochlor 1260 derivatized (342/499). Using th«
formula:
/Area counts of\/ ¦» nca\ .
{ derlvitized )(	je Std) * (°*745)
\Aroclor 1260 /\ 0.137 ££ "7
P"'	x (100) * Perchlorination
/Wt inject tQa)\	efficiency
\ Arochlor 1260
Perch lori nation efficiency may vary for different Aroclor mixtures. Two levels of
Aroclor 1260 were perchlorinated. Samples were derivatized using the same
procedure s outlined 1n Section 4.2.1. The area counts observed from the GC/EC
analysis were 25.619 and 56.781 respectively for 100 pg and 250 pg of Aroclor 126Q%
As a standard for these experiments, a 50-pg Injection of DCS was used which
resulted in 0.137 ac/pg of KB. A portion of Aroclor 1260 was subjected to total
organic chlorine analysis. The actual weight percent chlorine for this standard
material was found to be 57.3 percent.
Standard calibration curves were established by Injecting dilutions of authentic
standards (8) for decachlorobiphenyl by GC/EC and for blphenyl by GC/FID.
Calibration curves (see Figures 4-9 and 4-10) for these compounds show a linear
response that w« plotted by the use of the linear regression equation.
4.3 SUMMARY /WO CONCLUSION
As can readily be seen frun the data 1n Tables 4—1 through 4—3» the kiln ash
contained by far the highest levels of 0C8 observed. The difference between the
background bum (day 1) and the PC8 burn (days 2 through 4) is quite evident. Th%
4-22
-------
V

?. r»5
3:??
i-.S*
4.9*
Hi
**sz
At 36
!v.fc3
3. 41
*.7?
¥%
il:»
U:«
1 ^ 22
1S.*«


iV:&
ii:||
; j|;||
1 V*.**
PMD
Figure 4-8. Method Blank Chromatogram
4-23
-------
60
50
g 40
=3
O
u
< 30
25
20
10
5
25
0
Pg a
LINEAR REGRESSION CORRELATION 0.999
J	L_
20 50 100
1
200
_L_
300
_L_
400
J	
500
±
600
700
800
900
1000
Pg STO OCB INJECTED
4-9. Decachloroblphenyl Calibration Curve (GC/EC)
-------
IBOKi
I50K
C 100K
o
50K
10K
500
200
300
400
100
Ng BIPHENYL INJECTED
Figure 4-10. Blphenyl Standard Curve (GC/FID)
-------
9 0
presence of PCBs In unperchlorlnated fractions of the kiln ash extract was
confirmed by GC/MS identification. In addition, numerous chlorinated and
unchlorinated compounds were identified and shown to be predominately polynuclear
aromatics.
Results of analyses of the two liquid streams (the scrubber effluent and the
recycle scrubber effluent) are also clear and specific for the presence of DCS.
However, the levels were barely above the quantitation limits by GC/EC and the
presence of PCBs was not verifiable by GC/MS.
The stack gas samples also showed the presence of DCB and at sufficient levels for
good quantitation. The presence of DC8 1n the perch1 orinated fractions was
confirmed by GC/MS. However, PCBs could not be detected in the unperchlorlnated
portions of the saaple. These results support the work of Halle (as reported by
Levins) which conclude? that the perch 1or1nation procedure can result in false
positive values for PCBs when applied to combustion source emissions. Furthermore,
this validates the necessity of GC/MS verification for PCBs when the DCB procedure
1s used for the quantitation of PCBs in combustion sources.
In conclusion, PC8s are present 1n the kiln ash at the parts per million level;
PCBs are not present at the detectable limits of the methodology in the scrubber
effluent streams, and are nbt present in the stack gas emissions above 50 ug/m3--
(less than 50 ppb weight PCBs in stack gas per unit weight PCB burned).
Based upon the available literature and the experience that Acurex has gained from
participation 1n this test incineration, the following recommendations are made:
• Utilize the DC8-perch1orination methodology with strict
reservations
Divide all samples to be perchlorInated into two or more
fractions. If any of the derivatized samples show positive
results, confirm with a different method (i.e., GC/MS)
e Adopt a method similar to that outlined in the article by Webb &
McCall and presented 1n "Measurement of PCB Emissions from
Combustion Sources," EPA-600/7-79-047, February 1979 for the
definitive evaluation of suspected.PC8-contain1ng samples from
combustion sources
4-26
-------
<71
SECTION 5
RESULTS OF PCS TRIAL BURN
This section of the report s una arizes the test results from the PCS trial bum
conducted at the Energy Systems Company in El Dorado, Arkansas.
5.1	RETENTION TIME
The EPA regulations specify that retention time in the Incinerator must be
2 seconds at 2192°F with 3 percent excess oxygen 1n the stack gas or 1.3 seconds
at 2912°F with 2 percent excess oxygen 1n the stack gas.
The rotary kiln had a retention time of "3.7 seconds at 1400°F (average) during
the trial burn. The afterburner, which consisted of a primary and secondary
combustion chamber, and a passageway had retention times of -2.19 seconds at
2290°F (average primary), -2.11 seconds at 1915°F (average secondary), and
-0.23 second at 2280°F (passageway) during the trial burn. The average excess
oxygen level (measured at the scrubber outlet) was 13.35 percent.- Hence, the
ENSCO facility exceeded the minimum EPA requirements for retention time,
temperature and excess oxygen required for efficient destruction of PCBs.
5.2	COMBUSTION EFFICIENCY
EPA requires that the combustion efficiency of the Incinerator be at least
99.9 percent. Based on the results of the gas analysis at the scrubber outlet
using the continuous stack gas monitors (for CO and 02) and the Orsat (for
COg), the concentration of CO averaged <10 ppm and the average concentration of
CO2 was 5.45 percent. Hence, the combustion efficiency of the Incinerator
(greater than 99.98 percent) exceeded the minimum criteria of 99.9 percent.
5.3	FEEDRATE OF PCBs
The feedrate of liquid PCSs and the feedrate of shredded electronic capacitors fed
to the incinerator were measured and recorded during the trial bum. The average
feedrate for the 3 PC8 test bums was 393 lb/hr of liquid PC8s and 243 l*>/hr of
PCBs 1n the shredded capacitors for a total average PCB Input of 636 lb/hr
5-1
-------

5.4	TEMPERATURES OF INCINERATION
The temperatures of the incineration process were continuously measured during the
trial burn using thermocouples in the walls of the chambers. As indicated
previously, the average temperatures in the rotary kiln, primary combustion
chamber and secondary combustion chamber were 1400°F, 2290°F and 1915°F
respectively.
5.5	AUTOMATIC SHUTOFF FOR PCSs
The EPA regulations specify that the flow of PCSs to the incinerator must be
automatically shutoff when the combustion temperatures drop below the specified
limits. The EN SCO Incinerator is equipped with automatic shutoff capability. The
automatic shutoff device can be triggered by low-temperature (<2050°F), high CO
(>50 ppm) and low excess oxygen (<4.5 percent) and, in fact, was triggered on
several occasions for short periods of time due to fluctuating CO levels. The
fluctuating CO levels resulted from variations 1n the liquid organic waste feed
rate which caused momentary fuel rich conditions occasionally, resulting in CO
levels >50 ppm.
5.6	MONITORING OF STACK EMISSIONS
The stack emissions were monitored during the PCB trial burn and the results are
presented in Table 5-1.
5.7	MONITORING OF COMBUSTION PROOUCTS
EPA requires continuous monitoring for Og and CO and periodic monitoring for
COg during the incineration of PCBs. Table 5-2 Illustrates the results for the
trial bum. These results were obtained from measurements made at the outlet of
the venturi scrubber.
5.8	WATER SCRUBBERS
EPA requires that water scrubbers be used to control emissions of HCl during PCB
incineration. Since the scrubber liquor drains to the sludge lagoon and 1s
recycled back to the scrubber, no effluent leaves the ENSCO property. These two
streams were monitored during the trial bum for PCBs and PC8s were not detected
in either stream.
5-2
-------
3
Table 5-1
ST** EMISSIONS OURIHS PCS TRIAL BUR"
Parameter	g	37.2
Oxygen (X)	^	ND
Carbon Monoxid# (<)	2.8
i»o
Carbon Dioxide (%)	3,5
.	3» 5
Oxides of Nitrogen (lb/hr)	22-3
12.S
12.8
Hydrochloric Acid (lb/hr)	^
Total Chlorinated Organics (lb/hr)
ND
pCBs (rag/kg feed)	^ g 176.S
Total Particulate Matter (lb/hr)
^Background test bum; no PCBs 1ncin
DNot detected
Day #3
•10/17/71
Day #4
10/18/79
16.6
15.6
NO
ND
2.8
3.0
3.3
3.4
58.3
41.2
NO
ND
ND
ND
107.3
1S6.7
Table 5-2
<«	Day #3 inriR/?9
n *1	0#y #io in/17/79 10/18/79
,!R«/t9	isSg/zs. *¦-' „.
Parameter iX-^	^ 13.8
°*y9«n (*) 1A*2	6" 4	4
Carbon Monoxide (PP"0 ^	g g 4.9
c*bon Dioxide (%) 4'5	99.991 »-992
Combustion Efficiency
5-3
-------
q *j
5.9	MASS AIR EMISSIONS
When burning nonliquid PC8s (e.g., solid electronic capacitors) EPA requires that
the mass air emissions from the incinerator shall be no greater than 0.001 gm PCS
per leg of PCS fed to the Incinerator. Since PCBs were not detected at considerably
below this specified concentration, this standard was not exceeded during this PCS
trial burn.
5.10	KILN ASH
PC8s were detected 1n the rotary kiln ash at levels <50 ppm which 1s lowest limit
of PCSs currently regulated by EPA. This ash can be landfllled at an approved
chemical landfill.
In sumnary, the ENSCO incineration facility has satisfied the requirements listed
in Annex I of the EPA rules and regulations governing the disposal of PCSs by
incineration as published in the Federal Register, Volume 44, No. 106 on Thursday,
May 31, 1979 and has successfully completed the PCS trial burn.
5-4
-------

REFERENCES
!• Hutzlnger, 0., et al. The Chemistry of PC3's CLeveland: CSC Press, 1974.
2. Armour, J. "Quantitative Perchlorlnation of Polychlorlnated Slphenyls as a
Method for Confirmatory Residue Measurement and Identification." JAQAC.
56:986, 1973.
3.
Halle, C. PCB Interlaboratory Verification Analysis. Midwest Research
Institute, Final Report for ErA Contract 68-02-1399, December 27, 1976.
4.	«,t*.	A. ttC.lt.	,0r ElKtr°" CWtlJr" G"
Chromatography.11 J. Chroma. Sc1.. 11.3oo,
5.	£td»H>«-3.r, J.. « al.	Polycnlor 1na«d Sibyls
Problem." Anal. Chen.. 46:227, 1974.
c -	, ,	< am. u Atmosohere. Environmental Protection
*• Environmental Assessme"* PCBi m r,w
Agency. SPA-450/3-77-441,, November l^.
7.	Llgget, L. M. -Determination of Organic Halogen with Sodium 81Phenyl
Reagent." JAM Chen, 24, April 1954.
8.	Baker, J. T. and Supelco
R-l
-------
lb
APPENDIX A
-------
*»
I
w
RIC
01/82/80 I0t50s00
SAMPLEt 79-3472 P
-------
I
100.0"
RIC
RIC	OATAi PCB3474 It
81/02/80 10118(80	CALIt C1288 13
SAHPLEt 79-3474 PC8 TRAIN DAY? 4UL1NJ
RAMCEl C 1,1693 LABELi H 1,20.8 QUANi A	0, 1.0 BASEi U 20| 3
SCANS 1 10 600
/A

T
m
U4»
200
3tM
500
liM
139776.
600 SCAN
IAiCA UMf

-------
KIC
BiWOB 3t24t00
SAMPLE! 79-3476 KB TRAIN 0AYI3 4UL
RfiHBEl Q 1,2009 LfiBELt H 1,20.0
MTtii PC83476 01
CfiLlt CI200 §3
SCONS I TO 1600

215296.

868
13s 20
I860
16«40
1280
28(80
1400
23i28
1600 SCAN
26H0 TINE
-------
>
I
o»
66.7-1
RIC
RIC	OATAs PC83478 II
81/02/80 0H4I00	CALIs C1280 13
SAMPLES 79-3478 PCS TRAIN DAYI4 2UL INJECTED
RANGES G 1.1641 LABELS H 1.28.0 OUANi A 8. 1.8 BAOs U 20. 3
1892
SCANS I TO 1641
447
-------
I
1
RIC
RiC	DAT(H KILNi »1525
i2/29/79 2Btl7»99	(MLIt CI229F3C #3
SAMPLE! 3.0UL KILN ASH BAYM BACKGROUND BURN
RfiHGEt G I, t IMBLl H 9. 4.9 QUANt A 9. 1.9 BtGEi U 29, 3
SCANS I TO J400
280
3i20
TOTAL ION CURRENT
QROMATOGRttl
^..ouJL
114016.
10100
000
13s 20
I
1000
16(40
1200
20:00
—¦I
1400 SCAN
23j20 TIME
-------
RIC
12/23/79 I8H3sO0
ShHPLE: 3.6UL KILN ASH 0AYI2 PCB BURII
RAHCEl G 1,2800 LABEL« H 1.20.0 QUAHi A
-i	1	1	1	1	r
500
fetO
ieM.
16*40
QATAi KILN2 II	SCANS I TO 2000
CALIi CI223790 13
1.0 BASE: U 20. 3
433808.
2000 SCAN
33il» TIME
-------
RIC	DATA: KILM3 »I956	SCANS I TO 1387
12/29/79 I7i45:00	CALIi C122979C 13
SAMPLE: 3UL KILN ASH 0AVI3 PCB 6URH
RAIIGEt G 1.1987 LABELi N 1.20.0	flUAHi A 0, 1.6 BASEt (J 20, 3
560128
i	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	n	1	1	
500	1000	1500	SCAN
8> 20	IbHO	25:06	TINE
-------
T
160.6-1
R1C	OATA: KILH4 «79?
12/63/73 I5:3bi06	ChLI: CI2973 166
SAHPLEi 2.0UL KIUI ASH 14 AFTER FLORISIL 6?. FRACTION
RAHGEi C 1.1408 LABEL* H 1.50.0 QUAHl A 0, 1.0 BASE* U 26. 3
856
SCAHS 106 TO 1260
136
172
RIC
366
545
_Lol

466
6:4ft
SuO
10:00
IbS 40
600
IJjW
17440.
1200 som
20:60 IIHE
C
-------
IOS'
APPENDIX 3
-------
I Ob
t	1 UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
\*S2ZJ	KCSION VI
6608 Hornwood Drive
Houston, Texas 77074
Mr. Norman M. Hynn	j " '3'3
Staff Chemist
Accurex Corporation
485 Clyde Avenue
Mountain View, CA 94042
Dear Mr. Flynn:
I have received your report of analysis of quality assurance samples
for PC8 content which I supplied to you August 7, 1979. The following
are the results reported by you and the expected true values of the
samples when prepared according to the directions in ny letter to
" ^Ported Composition:
Aroclor 1254; 1.0 ug/1 iter (ppb)
Expected Actual Composition:
Aroclor 1254; 1.44 ug/liter (ppb)
-	Reported Composition:
Aroclor 1016; 0.75 ug/Mter (ppb)
Expected Actual Composition:
Aroclor 1016; 1.12 ug/liter (ppb)
-	Reported Composition:
Aroclor 1254; 0.95 ug/1 (ppb)
Aroclor 1016; 0.64 ug/1 (ppb)
Expected Actual Compos I Llunr-
Aroclor 1254 ; 0.87 ug/1 (ppb)
Aroelor 1016; 0.39 ug/1 (ppb)
Your Aroclor specie identifications are correct in each Instance and all
quantitative values would be within the range of acceptability based on
similar type samples analyzed annually by State and Federal laboratories
for proficiency evaluation.
Sincerely,
William 0. Langley, Ph.D.
Chief, Laboratory Section
cc:
Oscar Ramirez, Jr., 6ASA
Charles Gazda, 6ASASC
B-3
Mr. Steiner.
Sample No. 1
Sample No. 2
Sample No. 3
-------
I o 7
PARTICULATE CALCULATIONS	/©-/S--7?
1. Volume at dry gas sampled at standard conditions. 68" F, 29.92 inert Hg (set)
or Vm stack ' (-y * Vwstd) (r^'^vg + ^>)(p?") (acf at suck conditi°M>
2.	Stack gas moisture condensed at standard conditions (set)
Vv*std » 0.04707 Vlc	f/c m 3SLC+
. /5*. 83?
3.	Stack gas proportion of water vapor, by voiume
«	"*«"	e<" a.avS~	oactt' **C
° V«SM ~ Vm»f	^ '?/'<*
» SLSS
4.	Stack gas dry molecular weigM (Ib/tb-mote)
MWd » 0.44 (*C02> ¦" 0.32 (%02) + 0-28 (%N2 * %CO)
-
5.	Stack gas molecular weight (lb/lb-mo»e)
MW, • MWd (1 - Bw0) * 18 (Bwo)
• ^
-------
/ 0 ?
7. Stack gas velocity at stack conditions (ft/sac)
V,. 85.49 (Cp) K/^>„g
» /5i/
8. Stack gas volume at standard conditions (scfm|
Cp * 0,7/
Q, « 60 (1 - Bw0) Vsavg A, ^xs avg2° 46o)
3SQ
or Q# ¦ 60 Vsavg Aj (actm) »
9. Test percent isokinetic
... 17.33 (Ts>ava ~ 460) [0.04707 (W,c) * Vm^]
• vs Ps 0„'
?7. 9
10. Particulate matter concentration, gr/acf
C, « 16.432	" C,/^0¥
or C- * 15.432 u ' ¦ (gr/acf) ¦
*	vmstack
11.	Emission rata of particulate matter, tb/hr
EH « 0.00857 (Qs) Ct
• S"*-. 8
12.	Percent excess air at sampling point
#CL
(*.*)
100 [%Oz - 0.5 (%CO)]	^ _
% ^ " 0.264 [%N2 - (%Oz - 0.5 x %CO)] "	7
13. Emission rate of particulate matter, lb/10* Btu
B-5
-------
<*. 	
l«x iocaww 3i Oitk. ..
¦& I &4 1(/mc
Hiimntm tauun ^9* (f /
Sialic haul	4) '.**/	
it P'cuuM	 .. 			
NlMMMt	/X.' 	
PARTICULATE TEST FIELD DATA SHEET
.1177.
Hon* Sin • Mumbat
WO.
FILTER DAIA

Quel flinnnnm dim	
M l~ JLtQ	
JA&mArt..
. Mai NumbM	.~7I ft~t	
. Um Baa Nunfew —Q
ytsk
iHO
T
m
I*,?
jl|0
S\tt
(31f
	
NUMBER
TMK
FINAL WT
1


*¦- \.0\H






•MPtNGER
vmyuts
lic
SILICA
PEL
Eg
TIME
CO?
y.9
o>
TO
CO
SAMPLE
KMNI
CLOCK
1ME
VELOCIIV
•IE AO
* »|
orifice
MEIER
M« *|
GAS
MEIER
VOLUME ffl*
TEMPERATURES *F |
PUMP
VACUUM
m MB
/IF

STACK
mow
MPtNOCH
ORGANIC
MODULE
OVEN
OAS MEIER |
IN
OUT
1
o
.ObS
/.7«/
f^.S37
i4h
767


26 f
P*
?>
b
.2^-



-OJV
2)-?
ho7.

2d 3


2 tf
is
it?
IP


J
16.0
OAiT
(7f

Mr
2£2_


a.7*
yf
r*
&


¥
tC.O
./y
2.0

W*
2AT


?*n

.4
XfT-o
MS

&H.QQ
Ml
&2f


no

*&
A3
as.


itf'O



Hi**




jfeiL
dQt



i
3O-0

Ml.
tw ^os-
141
%1 a



7*
fio
#
it
r»/»
¦X
ic.o j
•7r
/.99
'?
tSl
xi, i


24 J
r?_
/frt
/o
r/7
/ 03 / ^j>aru^



2Jt
We.tl
tVi- u
If*
iSZ
2fe2_



k«>
Rl i
1/

j
i j
/
S&Q
. /O -
-0?,
PJflJ
Hfc]



3d
ll \
/g
.3^-

Z.>/



171
1®_
1
ft /
/*>
J

t.
fjf-P
,o*
2,'*
fcVl 23
/V7
Z70


a.7 o

tf/
<3
.1#

&H-
fcfl- o



























£W> LEMi














*o*a














(pin a



	




















1"
«• I-



AMMOIU


4,2<

xS









O
-£>
('••kIidio* • *9
-------
I 1 o
7z*rh.
PARTICULATE CALCULATIONS	/©-/* - ??
1. Volume of dry gas sampled at standard conditions, 88* F, 29.32 inch Hg (set)
vmM • 17.84 ^. *y. WO	-1- /.a/1
or Vmstack • ("T1 * Vwstd) (l^vg^0) fr) (>Cf #t **** C0ndit,0n,)
2.	Stack gas moisture condensed at standard conditions (set)
VwM • 0.04707 VIC	y/c .
/S'.fSf
3.	Stack gas proportion of water vapor, by volume
a	Vwstd
s<
WO Vwstd + vmrtd
O.ZS'S
4.	Stack gas dry molecular weight (Ib/lb-mole)
MWd » 0.44 (%C02) ~ 0.32 (%02) ~ 0.28 (%N£ ~ %CO)
. *9. /3
5.	Stack gas molecular weight (Ib/lb-mole)
MW, » MWd (1 - Bwo) «¦ 18 (Bw0)
• +L&. 2*
8. Pressure stack, in. Hg
p. ¦ p»
• St*. TC
B-7
-------
11
7. Stack gas velocity at stack conditions (ft/sac)
V, » 85.49 lCp) tv®
Cp S «.?/
• /f, S
*¦ Stack gaa voiuma at standard condition# (sefm)
» \Ta^vg ' x
Qs * 60 (1 - Bw0) Vaavg A,
or Qa « 60 Vsavg A, (acfm) *
'• T«at parcant isokinetic
17.33 (Tg avg > 4«ft [0.04707 (W|e> * VmStdl
?£. 8
^ Particulate matter concentration, gr/acf
* 15.432	$2-
5 ^ Vmatd
« C. » 15.432 tt=~— (Qr/aef) »
"	Vmstack
^3, S03
B « 2.679 x 10*

(a.oS9¥~)

^nwaion rata of partlculata mattar, lb/lv
5* » 0.00857 (Qt) C%
^ ^•TCam axcaaa air at sampling point
1QO f*07 - 0.5 (*CO)i . . /C3.7
* 0.264 [%N2 - {*02 " U,S x u,i
^ Bmiaaion rata of particulate mattar, Ma/lCP Btu
B-3
-------
Plant (Mi(0
IM. loj/t/ll_
Inl tucakon >/oCj
BaionMliic Pitiuil
SlMlC PlMlUfl 			
Slack PftwNf	 yt.ih
LATE 1
,7W
PAHTICIJLATE TEST FIELD DATA SHEET
HelilB &l> I NwJLm

BWO
Slack Pwii>i Mdm
Oucl Ouniioiu		
Slart Im		
(lyilni

FMfR OAIA
Nol CoaMacianl _
- PUd Huwtli		
. Malar Baa Numlw	Of J
. Of*ca CoaKicianl		 . ilr:!
1X1'
CO
«o
£NJ»
NUMBER
TARE
flNAL WI










IMPtNGER
yft-VMES

3?
saicA
Q£l
TIME
0*3
7.f
COj
IJ
y 1 **
^JS2gt
o»

CO
SAMF1E
POINT
CLOCK
l«K
VELOCITY
HEAD
AP «ft wq
omricE
MEIER
lHa wg
OAS
METER
VOLUME FT*
TEMPERATURES *F
PUUP
vacuum
m Hg
>/5p
leak fa-
.08 &'SA
p,i^ *K
SI AC*
f*nOBE
IMPtNOEn
ORGANIC
MOOUtE
OVEN
OAS MEIER
M
OUT
/
6
il

7W
/ST?
271


J2&?
tr




%
<
ill
Mfr


Zm. ?!
/sj>
27°


17i


?,&


}

.bls
LlL
671 .fe7
/rfe
ztl



^7
•Vfa .
rf


0

, afi>
i.yi
Lib 2!

IB a


2« 3
rf.r

,f
2y


3d




























H £
«¦ »C

ft}
kfu.2-1

iao



£7
if£
7
•13


fO'if
B2.G>3

-2*7.





LSil> IMJ














.ax














If A V














*-n7 ¦
nviow


A->d

/•*?




^•5"



'	

0
-------
I IS
PARTICULATE CALCULATIONS	/0 ^
1. Volume of dry gat sampled at standard conditions, 68" F, 29.92 inert Hg (scf)
vm^ • ,7.6.^1	^	Of.O,f
\'m,avg
°r Vm stack

stack conditions) #
2.	Stack gas moisture condansad at standard conditions (scf)
Vw^ ¦ 0.047X37 Vle	~ -3/f V-/7./
¦ /r. wo
3.	stack gu propoiHon of vatot vapor. By volunw
B ,	^sa	
BuiA *
*° "	* Vmstd
£* 3
4.	Stack gas dry molecular weight (Ib/lb-mole)
MWd ¦ 0.44 (%C02) * 0-32 (%02) ~ 0-28 (%N2 * %CO)
¦ ^.//
5.	Stack gas molecular weight (lb/lb-mole)
MW, - MWd (1 - Bw0) + 18 (Bw0)
• SX./9
®- Pressure stack, in. Hg
p« ¦ Ph +-^-
* b 13.6
¦
B-10
-------
" y
7. Stack gas velocity at stack conditions (ft/sec)
M « avn ~ 460
v..,cp, ^	^ = o
• /¥.&
8.	Stack gas volume at standard conditions (scfm)
Qs » 60 (1 - Bwo) Vsavg A, (t, avg2 + 46o)	*
or Qt ¦ 60 Vsavg Aj (acfm) ¦
9.	Test percent isokinetic
*3, 7CC
17.33 (Ts,avg + 460) [0.04707 (W,e) + Vm^]
• vs % °na
• f s. 3
10. Particulate matter concentration, gr/scf
c» " 15 432 " 0-3-863.	/S5?)
M.
or C. ¦ 15.432 p ¦' (Qr/acf) «
vmstack
11.	Emission rate of particulate matter, Ib/hr	//CC
ER » 0.00857 (Q$) C$
* /0?. 3
12.	Percent exeats air at sampling point
(**. O
100 [%02 - 0.5 (%CO)]	^
* ** * 0.264 [«N2 - (%02 -Oix %CO)] " "• A
13. Emission rate of particulate matter, lb/10* Stu
E « 2.679 x 10«
fe)F (acrrfesj)
B-n
-------
DM f/l/jC-O
om. lain/it	
s*s&4 .....	
l«M IacMon
PARTICULATE TEST FIELD DATA SHEET
flMMMC fV«U|J» .3/. 7j	tA«fO .	
, '	IOMH.StmtNumbt, tXICW	fTDHRoBT
-.XZ,		I wnjjufi
—Op. 76
Jf _.....
SUM ftWIM
Slack ftMiM
Mmw4 III
COj
Of
CO
0*
i
N
I*1
swA
l<#»°
SAMPLE
PCMNI
CLOCK
(ME
VELOCITY
ME AO
dP w lag
ORIFICE
MEIER
411 at- «B
OAS
TEMPERATURES 'f
rut*
M&JUM
M llf
¦JEr
LIaK
. 001Q. '*M
Pltot nff
MEIER
VOLUME n*
STACK
PROBE
MPtNOER
ORGANIC
MOOULE

OAS MEIEfl

IN
OUT
>c
0
.09
J.73
7a3.12-1
/rr
2$*/


2C6,


A?
3


f
.11
2.1*
707. Sk
is*7
Z7r


ISO

10
/7
,*y


<0
.HI
2
112. XT

272.


>2*1
fr
>?
/2-
,3

*3
IS"
<0(a
/.C2
7/r«.fir/
/«*
174


is-?.
ZH

7
.2/


lO
^Q7
/.**
1#1 »?
/r7
27^



*2-
ir?
y
.26

>'

1 %
1*
«T>

7i*n
/«-/
7.SO


PS3
7?
to
p,
'2-V

ftP^
ao


727. /6a*








































* *
ao
n
i.tt 7

/ar*
m 1


2.CI
11
VI
7
.2*


K

i.->y
730 *0
«c.o
S.7I


^£»
!> 4£
"-C At/












. <3
*0 Q




•M3
V
*?8fc





w1


•••«+ • -
WVWIM









RM



^Jtto
pm»<

-------
11 i
7Z*/¥
PARTICULATE CALCULATIONS	SO-/3- 7f
1. Volume of dry gas sampled at standard conditions. 68* F, 29.92 inch Hg (set)
Vmstd ¦ 17.64^2.(tX?Io) » ¥3. 676	<*-=/• otf
or staC)(
(t - Vw»a) fa * (77) (,c< " """iuon,,
2.	Stack gas moisture condensed at standard conditions (scf)
VwM . 0.04707 «c	^ „ 375" * /S. R.
" /S. 5»S
3.	Stack gas proportion of water vapor, by volume
B VW«*
w0 Vwstd * Vmstd
0. SL9&
4.	Stack gas dry molecular weight (Ib/lb-mole)
MWd • 0.44 (%C02) ~ 0-32 (%02) ~ 0.28 (%N2 «¦ SCO)
. St. so
5.	Stack gas molecular weight (Ib/lb-mole)
MW, - MWd (1 - Bw0) ~ 18 (B^,)
• 2&. 71
6.	Pressure stack, in. Hg
p. * % -m
- *9.
-------
I 17
7. Stack gas velocity at stack conditions (ft/sac)
' T3.avg * 460
Vj " 35.49 (Op) ,V^.lvg J	O.
• /S.S
8.	Stack gas voiuma at standard conditions (scfm)
Q, ¦ 60 (1 - Bw0) VSjyg A, (tJ—TSS?) ^§2) *	^OQ
or Qa ¦ 60 Vsavg A, (acfrn) »
9.	Tast parcant isokinetic
17.33 (T, avg * 480) [0.04707 (WIc) * Vm^)
"	» p. 6n*
r&. s
10. Particulate matter concentration, gr/scf
Mce
Cs » 15.432 JJJ2- »	33	Co • /// 3^
std
Mn
or C- » 15.432 -n-p— (gr/aef) «
*	vmstack
11.	Emiaaion rata of particulate matter, Ib/hr	#OL
EH » 0.00857 (Qs) C8
/«. 7
12.	Percent axcasa air at sampling point

100 [%02 - 0.5 (%CO)l	0<9 0
% ^ " 0.284 [%N2 - (%02 - 0-3 *	"
13. Emiaaion rata of particulate matter, lb/10* Btu
e • ^x 1°* fe) f U"'ii2)
B-H
-------
Hiii Bj/SCiO	t		 ftiiinn	^
101 It fit. Stake Pkum .. .
iMlouM	--..v i., Stack PintM	
* ../VVf •„ mmnm** . ,/tf.' ...
	 P4o4 CovNicml _ |^/_	
	Wd HuXm	*0
Start lax .JQfTJk			M«l»i Dm NlaUi ^ tST	i
PARTICULATE TEST FIELD DATA SHEET
Nouia Sua I Nunta	
nw.
ftLtfH OA (A
Opaiata AUtlAiljJ-.
. OAca CeaWicianl.

*5*

U)

>r1
6*>
ll*
NUMBED
TAHC
FINAL Wl










tMPMOEII
.yoi|Mts
TME
SILICA
OEt


CO/
» d
o»
wz
CO
SAMPLE
PCMNT
CLOCK
IME
VELOCIIV
HE AO
AP m «g
OMfKX
MEIER
Mm «|
OAS
MEIER
VOLUME Fl'
lEMPERAIURES *F
ruur
MCUUM
« H>
y/E*
**!*r
'.0C&%tS
ott.
SIACK
pnoec
MPMdEtt
onoANic
MOOULE
CMEll
OAS MEIER
IN
our
b

.if)
1.0 C,
Vf.72*

2.7^


X77
?7


.S2.

S-
5*
.II
*.12.

/&c?
17/


2"7*

•r?
'6


4
to
.09
2 3;
O 3C
/*?
XIL.


2.7?

rt?
'y
-.1

3
IS
.dC

62.1*7
/S-0
ZG>?


2-7>
?9

.r
-It

Z
XJO
.OSS'
7.^1
(jr. it

lc*y


*77
u


??

I
2.S
.U(f
t.,T?


2d* ?


27*
fit

f
-2/

off
ao


ix.Mt













aa^kX

























b
3o
.//
2JJ
12..137

Jlo^f


217 7
?2
v
11
II?

«•

.US
3Cb
76. J 3
lioO
zo7


2.77
?/
q?
f?
• 3V


HO
./O
1*1

i&±_
t€r t
1-71



?/
r<
, /
• 3»

3


K, *)/
2,70


,2*\r

rt
4T
.

z
<9
~t»s-
z.o
)&K /9
KmO




03
9*
t
17

/
£Z
.or
2./v?
n.ft?
'*9
n
-------
PARTICULATE CALCULATIONS
PC3/ZCL B*s*U»€
so//s/7f
1. Vdlume of dry gas sampled at standard conditions. 88* F, 29.92 inch Hg (scf)
2. Stack gas moisture condensed at standard conditions {scf)
•	3 9. r* z
3.	Stack gas proportion of water vapor, by volume
a	Vw*d
wo ^std^std
O• 5/3
4.	Stack gas dry molecular weight (lb/lb-mole)
MWd « 0.44 (%C02) ~ 0.32 (%02) * 0.2S (%Ng * %CO)
5.	Stack gas molecular weight (lb/lb-mole)
MW, - MWd (1 - Bwo) -¦ 18 (Sw0)
•	^6. 6/
8. Pressure stack, in. Hg
m,avg
(acf at stack conditions)
OC. sm /.O0 7
VwKd » 0.04707 V»c
-
8SC + AO. S*

B—16
-------
I % O
7. Stack gas velocity at stack conditions (ft/sac)
v. ¦ «v ^
» /Q • S>
8.	Stack gas volume at standard conditions (scfrn)
Q, » 00 <1 - Bwe) Vs^g A,	46o) (2O2)
or Qa » 60 Vaavg A, (actm) •
9.	Test parcant isokinetic
17.33 (T» avo + ^ [0.04707 (Wlc) * Vmst(J]
*•	»v,l»,6n'
s
10.	Particulate matter concentration, gr/sef
Mn
C. ¦ 15.432 -n-2- ¦
»	vm#td
M_
or C- • 15.432 ttt-2— (gr/acf) »
*	Vmstack
11.	Smiaaion rata of particulate matter, Ib/hr
EH * 0.00857 (Qf) C,
*
12 Percent excess air at sampling point
100 [%02 - 0.5 (SCO)]
% ^ * 0.264 [%n2 " (*02 - 04 x %CO)l *
13. Emisaion rata of particulate matter, lb/10* Btu
B-17
-------
GO
I
Ot>
riMI fchf lUlUffiMllC PfClMHa	. i f
*m....se.r.!£zz±		 smcpmmim	r.#.2H	
1«M Lotalio* - $Tl(^	 			black Pihwm		 	
L		
«*« - 		Mot Cortfccum	1 (ij£	
Omummom m. a m
tfiJL	
i?.M/
PARTICULATE TEST FIELD DATA SHEET
Nauli Sim I Nuntaf f ft>7 -
Uol«cuUi Wstgltf
BWO	
mien oaia
sun i«
opwMoi	vr^
Main Bm Hiilm	! V !• 007
Onkc* CbiMicmih .
Mr
fWrT
5-*;


NUMBER
TAKE
FINAL wt










IMPINGED
YQtUMtS-
loo
IQO


skica
oft

TIME
COj
/.f
o»

CO
SAMPLE
Man
CLOCK
»•*

OAS
MEIER
VOLUME Fl1
TEMPERATURES *F
niMP
VACUUM
« Hi
>/SF
ita<( ».04» ^

JO

1.77
2*to.i 7
1*7
261



7(r
77
i r





h 3V
lot 12

H~y



H

IX

O.ovT * 1
f«lj
ti?












!S"H*















1 •"
/o
.itl
/.#
M.W1
/Vf
in



77

it
.jv/
0.011 ft It'l
2


lr?X

• a
2i
•14'



,* ,lfrr
• lit
lfc»T
fMM|C||l?l I .*9
-------
I % 2-
FCB #3-
PARTICULATE CALCULATIONS
i. Volume ot dry gas sampled at standard conditions. 68* F, 29.92 inch Hg (sef)
vmstd	. /of. *07	+<¦ « / eel
or Vmstack *
(•T1 + Vw«d) (l^7g^) (^f) (8Cf 8t C0nditi0n$l
2.	Stack gas moisture condensed at standard conditions (set)
Vwstd * 0.04707 Vlc	m QQq + /Q.o + //.A
* va • 7+6
3.	Stack gas proportion of water vapor, by volume
-	Vwstd
O-
wo Vwwd - ym3td
» 6.397
4.	Stack gas dry molecular weight (Ib/lbwnole)
MWd « 0.44 f%C02) ~ 0.32 (%02) ~ 0.28 {%N2 * %CO)
•
5.	Stack gas molecular weight (Ib/lb-mote)
MW, - MWd (1 - B^) + 18 (Bw0)
• as- SSL
6.	Pressure stack, in. Hg
p» * pd +?ib
' fif. 70
B-19
-------
/ 0-3
7. Stack gas vetocity St stack conditions (ft/sac)
rr5,avg * 460
V, . 8S.49 (C|j)
* /^" • s
3. Stack gas voluma at standard conditions (sctm)
q5 »so (i - vs^g a,	46b) (ada)
or Qa » 80 VsaVg A, (acfm) •
9- Teat percent isokinetic
17.33 (T, avg ~ 460) [0.04707 (Wle) ~ Vm^l
W-	« V9 Ps 6n»
10. Particulate matter concentration, gr/aef
J1e_
/mstd

Mo
or C- » 15.432 «=-=— (gr/acf) «
¦	vm^tack
11. Einiaaion rata of particulate matter, lb/hr
ER * 0.00897 (Qs) C,
a
12.	Percent excess air at sampling point
100 t*Og " 0-3 (*CO)l
* * 0.284 - (%©2 - 0.S x TW-WJ
13.	Emission rata of particulate matter, lb/19 8tu
& ¦ 2.979 X 104
M0
-------
w
I
M
tf
n«rt
W





* 1
7
,?//

2
2f
.11
O.K 1
*U. Yr
iff




71
go

,??2-

3 •J
to
.(0
07 7
m.i*
If 7




%o
*1
2


tnP
So












i.t'fc fH'Cj















1
to
If
1 .oi
w.7n
It*
Hi



to
10
<£
^7/
»fll0 ~/
2
$£>
.tsr
l'7
UU7
i $7
27f



4*
*3
1
•?f7

J •*
1"

W
tllAO
HO
29J



4~?
*7


-»I7
£nl



ni*t>3














Hill*










1 »il»
|£0
•®7

•HfrrrZ
Ift
}€

ll

n
<>V



2
140
*67
0 .JS
IriS.tT
uo
L$*



ze
2?

U*


$04. &2


*
ir





















i *«•
l|*>
•Pi
{
'/
P
-t-
-------
} 7-J
'°/'7 /??
PARTICULATE CALCULATIONS
1. Volume of dry gas sampled at standard conditions. 68" F, 29.92 inch Hg (set)
Vm»td ~(TP" ' ft) •	^ 007
\ m.avg /
or Vm stack
(^T~ Vwstd) (t^J0 + 46o)(pj^ (acf at stacK condltions)
2.	Stack gas moistura condensed at standard conditions (set)
Vwstd * °-04707 vic	^tc * ?9a + /6> f + /S"- 6
' 38.
3.	Stack gas proportion of water vapor, by volume
a	Vwstd
wo Vwatd ~ Vmst
-------
I 2- b
7. Stack gas velocity at stack conditions (ft/sec)
V, . 
-------
CO
IV)
iw /;Atffo
Om IP~~! 7~?f_
Ym« location
flltQimllC hMMM 21.7f
Suae Pimmm* 		
Slack hMM			
particulate test field data sheet
No/fla Sua < NuanMt
3	V M in**
mn.
FILTER OAT A

—. ... Pdol c
'OtfklMl * £ f
NUMBER
TAKE
final wt
P'tTf n ¦ f>
hM h
f" 1 J



SUrt Tm 		

M.IM. i f0^' i J>Oj
4 a 5 '
rinMirum ,*99



QpMitM -fir -CklL?	
		 QiAf



IMPINGED
w*5-
ii#

SKICA
act
*•» llU S
t/f.
JS
*Z
V*
|«A*
I"
S^-
%
«*»
SAMPLE
POINT
I «"
2
J73
M!L

1-+
tup
11-
l
jZ
-tMi
o'tl
oO
35
<
7h
4

1WWM
CLOCK
TIME
1.0
H*
M.
VELOCITY
HEAD
AP M «a
7ST
.<*7
£2L
Jfi.
_iL

JifiL
ia*
"iP"
OiL
0**.
2±0


OAW ICE
METER
tH« a|
Jit
£££
OAS
METER
VOLUME ft*
r*EW
£fcM£.
.err sso.iz.
10
it
41
.W

it
&L
i77f
till
ili
h£l
I,JX
L*£
A-rv)
1®
lI£1
ill

frr*-7'/
5r^«r7
fr/i"*7
r77,72,
\u.\%7
unp
H7.I*
ui.i1
»*•»»»
temperatures *f
stack
±£2>
i£?
li£

1£1
Isx

liil.
i£t
til
i2
8
m
IZlIIL
t.rlg

s
i££.
PROBE
2"?^,
2Jf
?72-
121
%
2*Z
27^
ME
m-
tins'
MPNMEft

fpr i Lm&jLi
OROANIC
MODULE
OVEN
OAS METER
IN
*

z^t
II

iL

??
i
22
goy
OUT
11
12
J1
Tt.
n
to
zz
%L
&0£
el57

£

IE

A.
t
y/Sf
-m.


tit
£2=Ik_
J*

Jtl
,*6/


iiO
iiO

!¦*«/< ftp/1-
¦gi^ i
I'umilD IBI i:i
P
-i
-------
) X2
FCS m 3
rt/ti/y?
PARTICULATE CALCULATIONS
1. Volume of dry gas sampled at standard conditions. 68* F, 29.92 inch Hg (set)
(' P ~ —-
T " ' ^I •	• /.OOf
m.avg '
or Vmstack*
(t* Vw„„) (;r:;,l)(^j <»ct it stack conditions)
2.	Stack gas moisture condensed at standard conditions (scf)
Vwn„ . 0.0470T Vlc	W ^
' ¥3.9)7
3.	Stack gas proportion of water vapor, by volume
Vwstd
swo » vw^ - Vmftd
O. 36/
4.	Stack gas dry molecular weight (Ib/lb-mole)
MWd « 0.44 (%C©2H' 0.32 (%02) ~ 0.28 (%N2 * *CO)
'O
5.	Stack gas molecular weight (Ib/lb-mole)
MW, - MWd (1 - Bwo) ~ 18 (Swo)
" if.
8. Pressure stack, -in. Hg
pst
p. • p«,
.*9. ¥7
B-25
-------
)^°!
7. stack gas velocity at stack condition* (tt/sac)
— /Tt.ava *
V,- 85.49 (Cp) ^av9v/ Fa MWs
» /5". 7
a. Stack gas voiurrt* at standard conditions (sefmj
/ 528 \ / P»_>
Q, - SO (1 - S^q) Vsavg A,	TSSo) \^S33,
or Qa » 60 Vs8vg Aj (acfm) •
9- Test parcant isokinetic
17.23 (T, avg ~ 460) [0.04707 (W1e) * Vm^d*
%l 		«v,ps6n»'
m
10.	Particulate matter concentration, gr/scf
Mp
C. « 15.432 n ¦' ¦ *
S	VmStd
Mn	_
or Ca ¦ 15.432	(gr/scf) •
11.	Emission rats of particuiats mattar, ib/hr
ER > Q.00857 (Qs) C,
¦
12.	Pareant excass air at sampling point
100 I*0>> - P.S (*CO)l
* * QJ264 %N2 - {*
-------
o>
«
N
ktoicO
D«M
imikinm . StgC
1
xK
iJ trt i«»-
¦woawNic PWtMM*
Sialic Pranun _
Stack fwnoil	
3*».J'7
¥1.11
PARTICULATE TEST FIELD DATA SHEET
J -f i
Hal IIm Sin ft Numtiw I 1
Mutotulw WfigM.
OWO . . 		
11* ''
rn.iEH oa i a
Start la

Op**" -fl Jiufj.
/*1
	
__?io	


< siA
V
Fti"

SAMPIE
POMT
lh

.M

£1$
|i«
i

M
( .*

m.
••MUM
CI OCX
IK
.22.


1»
%o
i«0
llL
0£
it*
CO
IM.
1SL
%oo
J*
AO
VCtOCJIY
IKAO
tf ¦ *|
j£.
.01
<41
.il
Jl
ill
¦IA
JL
A
onricc
MEiai
SHta »g
<121
7 Oo
'ML
Ui
Lol
3£±
2fT
M
m.
sir
W
WL

OAS
METER
VOLUME fl'
BEM
iyv.gr
Ui.tr
17 ?«
-------
/ 3 I
PLANT	OPERATOR
LOCATION
DATE
TIME
AMBIENT PRESSURE (IN. HG.) 7.
AMBIENT TEMPERATURE C°F)^ /
SAMPLING POINT
TEST NUMBER f j
N0X SAMPLING DATA	^



Initial Conditions
Final Conditions

Time
Flask
Number
Flask
Volume
(ml.)
Flask
Vacuum
(in. Hg.)
Flask
Temp
en
Flask
Vacuum
(1n. Hg.)
Flask
Temp
CF)

/«'!>
ft
W
2X.L
sor
)A


1 £0*
n

17- 9
7iV
j


jbi5
ii
^ 2o3»
11-Z
?2F



Pi)
i±

It 3

.?


-------
13:z-
PLANT	OPERATOR
LOCATION
DAT5 /0*tt-77
TIME
AMBIENT PRESSURE (IN. HS.) J % 7X
AMBIENT TEMPERATURE («F)	,
SWUM
TEST NUMBER
NO„ SAMPLING DATA
Time
Flask
Number
Flask
Volume
(ml.)
Initial Conditions
Final Conditions
Flask
Vacuum
(in. Hg.)
Flask
Temp
(#F)
Flask
Vacuum
(In. Hg.)
Flask
Temp
CF)
tSYO

•2D10
J&.V
fo
\A


$9
2o^\
. >f-S

*
•\V
nis

•2O&0
77. C
gr
1.2.4
<\Z

*2-
2JQ"%b>

tf
\.\x
V*
*	bccSflLKx
oJb*-v^
B-29
-------
133

PLANT 0
LOCATION
DATE /<3-/7-77
TIME
AMBIENT PRESSURE (IN. HG.) Jf,7^
AMBIENT TEMPERATURE (*F)
SAMPLING POINT
OPERATOR
TEST NUMBER
3
NO„ SAMPLIN6 DATA
^viStf$3
4-
Tint®
riuk
Number
nufc
Vol urn
(ml.)
Initial Conditions
Final Conditions
Flask
Vacuum
(1n. Hg.)
ntsk
Flask
Vacuum
(1n. Hg.}
Flask
Temp
CF)
IMS
*IS

a$.7
£3
V?

ma
'n

22U-
X5
/V

ftS)
'i
X*
¦2*6
£3
Ao
fcf




7*
~ 4

B-30
-------
I Si
PLANT C
LOCATION
DATE lo
TIME
AMBIENT PRESSURE (IN. HG.) ZL%
AMBIENT TEMPERATURE (*F)
SAMPLING POINT
OPERATOR
TEST NUMBER
y,
NOx SAMPLING DATA
M 1 ——	¦	^
£p&*2>
Vbt*2>
V
Tin
Flask
Number
Flask
Vol una
(ml.)
Initial Conditions
Final Conditions
Flask
Vacuum
(1n. Hg.)
n«ik
(3
Flask
Vacuum
(1n. Hg.)
Flask
Tamp
CFJ

» ^
20 5 G>
27<1

A 5

)<&$
*//
?OJO
2&> I
*7
./3

ifti


J£T, 1
rj

67
tSfo
*3
20^
?a. /
27
/.?5

B-31
-------

ANALYTTCAL REPORT FORM
SAMPLE NUMBER
79- 3¥S"T
7?- ?¥*&'*'
7f - 3V8 7
TP- 3?SS
~7?-
7f- By 9/
7?-
"7?-3V?3
SAMPLE I. 0.
Of//
7?
7?-5¥fV
7?-3Vf*r
7?-3yf7
7? '3Vf&
yf-yyff
7P-33&0
&ty/.
0ff/
IM/Z
asyJ-
m<2
P/tr3
&&.//
fo&z/f
y
SPECIES
/&*&¦ 3
3
v/ty-3 /£*!&¦/
&f/3
7M/3
M097
prf/x^st/S
psy?/%s&S7
/*//£&&
3MZS- 7r-?<> 9f
,SiM*S SZteoD Jttp 3&0&W
/»&* PMSd* &4f&0T77&t/.
lev '•
RESUp *4$ j
jC.3
as
7/0
&>7
7£/
A"2
7?.i2 !
I
37/
/?!&
77-2.
<&.£
/$/
//.£%
y_2ueB 4rr&* /
-------
/ 34
3o St /t 'ne to/t$/71
A/Ox Cl/cu/a^s	f/aiM "
~^Yn.
ty-Of"*-. ft ^
V	T, TI J
I7.M (v1 --5.)
/7.A¥. (»Q90- ze) / afl.aa /.aA
V	*** 5*? /
is^s.r
6. 2H3X»0
V",c
• 5
6. 20 *io
'(t)
(_S£i. \
.7
S. 1 2 x i o
fro Q» C
to (*#4, 3SS )(8.lax,o7 )
a. rt»/hr
B-33
-------
/ s ?
//O* Ca/cu /a //bns
3as< **/ts/&o
F/asM n
Ts+i (V - v.
^ Tf Ti /
I?. ( V. - asV^l . T? ^
1 \r4 TcJ
\5*9 *** '
/fi5*7. 8
m
4.243X10
'(*)
IL «3tf1 ^ t/C5 f V'O^ \
1 s*. 8 y
a. 38 xio
V /8«^.
.6
(•o Q» C
<•0 ( 3S*6 ")^2.38x vo"4 )
&.6 Ife/hr
B-34
-------
12?
3as* I'ne 'off/79
Cl/cU/&//6„s	""*«
T
1 5-haC
^3+tL
ty-ofji- 1}
V Tf Tl /
»7.c,<* ( Vj - as) . 7^
/7.^y (ao3o-as)( ^a.sa _ /.*/ \
39^ £- 8 J
-a
3"- »4 X 10
60 Qs C
fco C MC..3S8 Xs- 'X Xlo*«)
0.1 (b/hr
8-35
-------
/ 3*?
la*
Oy Ca./cu/a.li'or'S
- iO
V T, Ti /
&
F/a*M fS
17. ( V, - as) ^5. - 2.^
11.(0*/ (ao*s- »g )/^a8.qa _ /~ 5/
\	SVC-/
/S3 t. 2.
'(*)
6.a4^xvo
s/ J-/.«
&.2M3 *10
(S/-& \
V '03* 2-y
/.?£" X /0
CrO Q» C
to ( *6, 35*3 X/.75". X/o"*0
4.3 lfe/hf
8-36
-------
/ V o
Isa
'Vt*
A/Ox Ca./cu/ail'07'S
(^-V^(l. 1)
K T, Ti /
PC 3 + ' 'o/tb/v9
Wash /f
17. c**1 ( Vf - a s
/?.&y (ao^o-as}f	- />sa
V ssa. 5.^0
/QSS' 7
£.243*10
5
6. SH3 X IO
— 6
2.V2.X/0
'(t)
(— 1
V '*ss.7 J
(so C
wo ( y*,£&3
6». *f Ib/hr*
B-37
-------
Hi
TV*
rf/Oy C*./eu/&6'or>S
ty-ofit _ iO
V Tt Ti J
PC 3*' /o//*/^9
/z/asA ¥
f7.Cr*| ( v4 - as)
n»6# («tevy- as) /ag.sa* /.aa \
\S'33L	y
/a 33.7
* assumed
- r / N
6. a«43x»<3
6.ao xio
f 78* ft \
V /03S.7 y
• 4»
a.44 x/«
fro Qs C
4o ( *3,833	x/0~6)
7.0 ll*/hr
8-38
-------

A/Ca./iu/&l*/or*S
Ik Of-iO(l
?»<	v rf Tl J
17. Cr<4 ( - as) f	"\
Tt- '
.* ? (aoso - as} / ag.sv _ s./s. *\
\ sz* sve J
PC8«/ /o/zb/yf
/^Z&sJ* //
n
/7 60.^
£. 2<4Z*\0
-5
6.ao *
r(t)
\QS( .s )
-7
ji.a3 x io
fro Q, C
to ( 43,833 )(a.*3 Xio"7 ^
0.6 )4»/hr
B—39
-------
j
4/0X Ql/cu/aAo"*
4")
?
ra+a.
PC 3 * / *o//<>
^/a$k a
'7. Cr<# ( Vj -as
Kl-i)
/'?•&/ (zozt -•>£¦} a*?* _ //a ^
\ S3a 5V7 J
/So 5"* 6
6. a *43xiO
6.a«*3 xio*S
'ft)
Lsj± \
\ /SoS.* J
/. Bo x toT7
40 Qs C
to ( *3,S*3 )( /.aox/o"7)
O.S" lt/ir
B-40
-------
/ i-i
T
1 S+dL
?
3+4.
17. <*H
4/COl/cu /& AonS
. i')
V T, Ti /
PC S + a- 'o/n/-n
F Ms <4 /5"
r?. 6 ? ^03^-ag.sq _ /oa \
V	^3 y
/fi¥S\ 7
4. 241 x to
V~,e
5
&. 20 *»o
• 7
3-/7 x /«
'(t)
(_H-)
V tans.-?J
(so Q* C
to ( *3,700 X*-'7x/o"7)
0-8 ffc/f»r
B-41
-------
H-r
A/0* Ca/cu /S
PC 3#-sl j a/nfa
F/asJ* n
*sc ~ ls±i (V - V.
? 1
3+4.
)(!.!)
V T# Ti /
l7.Cr*| ( Vf - as) ^5- - 2.^
n.c,tf (*c7f- 3*) (*e.o9 _
v SA9 S*5* y
C
/S33. a.
4.a^3xiO
V",C
6. ao xid 5
/. a&
'(t)
(-12-)
\ /B32.3.J
£R •* bO Qs C
» tmO ( V3 ,7«©
s 3*3 Ife/lir
B-42
-------
y ^
T
P
//or>S
'•«	V T* Ti /
I?.M Cvf - as)^l . 2.^
/7.^^ (?*¥*9 _ ///g
V »-* s«J y
/=£•3* a '0/47/79
/8«0.~?
6. 2*43 XIO
5
6.2
-------
/ i i
3 •+ 2. /c//t
F/aak 3
T
p
a+4.
A/Ox Ca./eu./a.Ab"S
0,-0(1 - ^
V T, Tt /
17. <#q ( V. - as) . 7J ^
W 7fV
^ Jtg.s*? _ /&8 \
\	538/
/?.
/373.3
6. 243X10
-S
6. SO *10
/~33 * '0
'fe)
(. /S7S. 33
fro Q* C
Co (	AJ3 X/974)
3« 5* /*/4r-
B-44
-------

A/Ox Ca,/cu/A^er»s
Is* ty-vS)(T, _ Pi °)
Kt* tl J
I7.G,<( (Vf - S!)^l.
17. iy (s.OSA-Xs') (£®i££_ /S9 \
\ £-9-9 svy J
r£3*-*> 'o//e/
/8/0. e
fr. 243X10*
V",<
-5
fr.ao %\o
"(t)
p*M
V /s/o. ©y
- 7
&. 75" x /e
fro Qs C
fro C ¥*, X (S.75-X/©-7])
/• 8 !t*/hr
B-45
-------
ii
-------
/So
A/Ox /& //ar>s
PCS -#3 /c//e/"?9
/C/tfS^ 'f
ia* ty-OfJi . p: ^
?>*<	^ Tf Ti J
I7.fr* (V1 - ss)^|.
/7. (ao9Q	«3 . /.3f ^
\	5V7 /
/«s.3
6. a**3xJ0 r | >H |
k
\ /?60.3 y
-7
5". */©
c»o Q* C
fro ( *3 aes )(j.7V x/o"' )
A 5* /t/hr
B-47
-------
) si
A/Ox Ca./cu/a^O^S
la* ty-Of* . 1 >
^ K% 'J
'7.c**t (Vf - as)(l. 7* ^
^ Tc '
/7.C,? (aovf-as}/'sa^o - ££? \
\	5V7 )
7*CS #3 toMfa
^/asA 3
/5r
8-48
-------
/ 5"2-
NOZZLE MEASUREMENT
DIAMETER
DIMENSION
a Ml
B
C.
D.
E.
F.
mo
JLC&-
JL2JL
r? 7 I
DATE
RECORDED BY
M.Cht'PS
AVG.

NOZZLE SERIAL
B-49
-------
I S3
NOZZLE MEASUREMENT
RECORDED BY
diameter
D'MENSION
N°2a-S SERIAL
B-50
-------
NOZZLE MEASUREMENT
DIAMETER
DIMENSION
A '
IZMZZ
p mqH
avci .H0fT3
NOZZLE SERIAL
°A^
RECORDED BY
B-51
-------
Pitt 4/1B/79	Otitic* Neitr 520-084 opantors Rautnann
time 1500	Orlfic* lUsaclielic 80615BU28 "•* iMpenturc
imwlric hrwun	30.26 frkurf Cellbratloa Meter 259453
Mint Itf ere lure	70®F Caatral Nodule 520 - 084
METER CALIBRATION DATA	Orfice Plate 084
Oriflc*
Meaawter
"ref
(U.MS.t
•rifle*
Ihfwlirtlt
Prlaery
Meter
1
Ory Test
Hater
'<9
«ll
bt Voliat
frtaary Heter
»r
Ift.1!
bi VeltiM
Icepcritari
Tlae
t
(¦la. |
«

try hit Meter
u
itt.*\
Frtaury Meter
Dry lest Meter
Ulet,
Milt.
V ™
**«•..
Ulet.
Outlet.
Av«.,
lde '"f>
0.77
0.8


191.702
208.476
586.812
603.572
66
66
66
61
67
68
35
.991
.690
66
' 66
68
70




16.774
16.760













1.76
1.8


208.476
22A.467
mm
55
§§
66
§2

70
28
1.
004
.684
»
fin

72




19.991
20.134













2.75
2.8


226.467
249.748
623.706
645.245
§Z
6§
67

72
71
24
1.
nil
.681
bo
66
/o
U




21.281
21,539













•3.78
3.8


249.748
269.224
645.245
ftM 441
68
66
68
70
72
72
19
I*
015
.675
71
. 67
72
74




19.481
19.748




























































































K 0.683
0
Aver
*9*

-------
7-25-79
o*u 7-27-79
n* 10:15
Sarwetric Pressor* 29.88
Mini lartritm 30.02
75°. 72°
Orifice Meter 570 085
Orifice Hagaekeltc 806291BW89
Prlaury Calttretlen Meter 259453
Control Nodule 570 _ 035
Op€r,torl Seegnlller
Net talb fenperature
METER CALIBRATION DATA
Orifice Plate 085
Orifice
Mawneter
"U
(Ml
Orifice
Mafadietlc
"V.
frhury
Meter
Dry lest
He Ur
<'ta. «i.|
(is Mim
Prlaary Mettr
'f
6m Koliw
Dry Test Meter
,»«
leaperature |
l(ae
t
lain.)
a
*0
Primary Meter |
Dry Test Meter |
'•let,
%• I*f>
OMtHt,
v <•*>
'pa «*f>
Inlet a
Ut if\
Outlet.
<*r>
*»« .
"da «*r»
0.80
0.8
•

249.622
239.562
108.450
098.450
76
77
76.75
75
75
75
19.9
17
019
.719
77
77
75
75




10.060
10.000












,
1.80
1.8




Zfi
ft
76.75
76
77
76t>25
13.6
I.
nifi
.699
M'Z ¦TiTiWT?TH
;8
17
ZZ'SZ





10.040
10.000













2.82
2.8


2W.742
759.
128.450
lift 4UI
-if-
77
77
78.5
76
W~
76
~W~
76
11.0

.696




10.080
10.000













3.84
3.8


279.778
269.742
138.450
128.450
75
76
75.75
75
75
75
75
9.5
1.
m
.689
76
76




10.036
10.000















!





































































i






a 1.019
K 0.701
1 Average
I" ft»
-------
/J-7
Test Section
Velocity
(fps).









a.
u
W
i
e
u>
•
e
to
VO
•
e
VO
VO
•
o
ft
u>
S
s
•
o
«•»
r».
•
o
*0
•
e
a
in
•
a
«r>
•
e
WO
VO
•
a






at
1
e
10
•
e
a
m
M
«
•
e
in
(0
•
e
N
U»
8
s
•
o
s
•
o
¦
LD
S
1!
•
a
»
•
o









s»
•1
s
•
e
>
»•>.
e
*
e

s
d

w
•¦I
•

U9
«•«
«
e









8-S4
-------
DATE 8-17-80
TIME 0830
BAROMETRIC PRESSURE 30.06
AMBIENT TEMP &9°f
TEST PITOT TUBE	P-12
STANDARD PITOT TUBE	UNITED SENSOR
TEST SECTION LOCATION	UINDTUNNEL
OPERATORS	JARMAN
NOZZLE SIZE
NOZZLE-PITOT SPACING
PITOT-TC SPACING
PITOT TUBE CALIBRATION DATA
-------
IS1
EPRl FP-1207 Volume 4
Jill—
tlililil
"ir
nstUi
\r
c if —
J 2,3
!l»
7
Below are five index cards that allow for filing according to the
four cross-references in addition to the title of the report. A brief
abstract describing the major subject area covered in the report
is included on each card. For information regarding index card
subscriptions to past and future EPRl publications contact the
Research Heoorts Center. P.O. Box 50*90, Paio Alto. California
94303. Telephone (415) 965-4081.
EPRl FP-1207
VotUlM 4
R PI263-2
Final Report
Saptamfiar 1M0
IPM FP-1207
VotaOTM *
API283*2
Ptaai Report
September 19W
«PW FP-12DT
VokuM4
RPiaW-2
Pfeul Report
' 1M
EPRl
Oisposai of Poiychlorinated Blphenyls (PCSs) and
PCB-Contaminated Materials volume 4: Test
Incineration of Electrical Capacitors Containing PCSs
Contractor Acurax Corporation
voiuma 4 sraaanta a study of a trial Bum eonouctad to datarmina wnatnar
liduid poiycnlonnatad Oienanvis iPCSa) and snraooad aiactronic caoacitora can
ba incinaratad according to Snvironmantai PfCtaetion Aganey 'aguiationa. Tha
raoort somaina (1) a bnaf ovarviaw of ousiianao litaratura on ®C3 Burning, <» a
dMertotlon of tna Snargy Syatama Company facility, (3) a dlacuaaion of
sampling and analyaia orocaouraa. (4) a luminary of raauita ootainad from tna
trial bum. and (9 a eoilaction of raw data and sampia calculation# uaad to
oetain tna raauita.
SPRI Preiaet Managar fl. Y. Kornai
Cwaa-rafartneaa:
1. F»-1S07, V0X«n«4 2. WSSJW 1 Watar Quality Comrot ana Haa» Bawation P-oqram
*.	 	
iU«CT«IS »OW««	INITItun
»•« omt •« 1«41> »«• AIM. CA 94203 ¦••1*3000
EPRl FP-1207, VOLUME 4	EPRl
Disposal of Poiychlorinated Blphenyls {PCSs) and
PCs-Contaminated Materials volume 4: Test
Incineration of Electrical Capacitors Containing PC8e
Contractor Acurax Corporation
Votuma 4 graaarrta a study of a trial bum conducted to datarmina wftatnar
liquid Bolyeftlortnatad bipnanyla (PCSai andinraddad aiactronic saoaciton can
ba ineMaraxao according to Swronmamal Protection Agancy regulations. Tha
raoort contama ft) a Brief overview of pueilanad litaratura on PCI burning, ffl a
daacrtptlon of tna Snargy Syatama Company facility, 13) a aiacuaaion of
sampling im analyaia oroeaduraa. (4t a summary of raauita oBtainaa from tna
trial bum. and (8) a coiiactton of raw data ana sampia caicuiationa uaad to
o«aln tna raauita.
KPW Protect Managar R. Y. Komal
Craafrrafaraneaa:
i. IPniWMJBr.valume* 2. WISH 1 waiar Quality Cental ana Haai Wanetlaw Program
4.	 		
K»fiTuf|
I1MII-UM
iiiermc vows* ais«*«CH .
•mi ornot an mil. mm mi*, s* msos
RP1263-2		
EPRl
Disposal of Poiychlorinated Blphenyls (PCSs) and
PC8-Contaminated Materials volume 4: Test
Incineration of Electrical Capacitors Containing PCSs
Contractor Aeurw Corporation
Votuma * prtaama a atuay ol a trial Bum conduetad to determine wnetnar
iWuM pefyoniennatae wpnanyie (PCIal and snraddad jieotronic caoacitora can
Ba inelnaratad aoooraing to frwironinentai Protection Aganoy reguiaflona. Tha
raoaR aornama (1) • brtaf overview of puoilafted litaratura on PCI burning, (2 ¦
jaacrtotion of tna Snargy Syatama Comoany faeiilty, (3) a diaeuaaten of
sampling and analyaia eraoadwaa. (4) a summary of raauita ootainad fmm ma
trial bum, ana (A a collection of raw data and sampia caieuiattona uaad to
retain tna raauita.
IFRl Protect Managar R. Y. Komal
Cwawafaranaaa!
i.vmiw-iaBr.vaMiM4 i**ma
4.
3. WMarQuaMvCannel ana MaameweMn Program
SkSCraiC lows* Mlltaca NSTiruTS
•«« OlttM Ml 10411. tilt. ZA 14303 
-------
160
Division 2
-------
/CI
THE PCB INCINERATION TEST
MADE BY ROLLINS ENVIRONMENTAL SERVICES (TX)
AT DEER PARK, TEXAS
	NOVEMBER 12-16. 1979	
A Report to the United States
Environmental Protection Agency, Region 6
	Dallas. Texas	
-------
U2
RECIPIENTS OF COPIES OF THIS REPORT
Mr. James S. Sales	3 copies
U.S. EPA Region 6
Solid Waste Branch (6AAHSW)
1201 Elm Street
Dallas, IX 7S270
Mr. E. B. Stewart	1 copy-
Texas Air Control Board
3520 Shoal Creek Blvd.
Austin, TX 787S8
Mr. Jay Snow	1 copy-
Texas Dept. of Water Resources
Box 13246
Capitol Station
Austin, TX 78711
Mr. W. B. Philipbar, Jr.	i copy
Rollins Environmental Services, Inc.
Box 2349
Wilmington, DE 19899
RES (TX), Inc.
Box 609
Deer Park, IX 77S36
Original
retained
on file
i
-------
}b 3
Observers and Participants in PC3 Incineration Test
November 12*15, 1979
U.S. EPA, Region 6
Charles Gasdo
Frank Hall
Jim Sales
Phil Schwindt
Texas Department of Water Resources
Clarence Johnson
Texas Air Gantrol Board
G. McArthur
Son Vo
Harris County Pollution Control Department
G. M. Combs
TRW (EPA contractor)
R. J. Johnson
City of Houston, Board of Air Quality Control
Virginia Khab
Rollins Environmental Services
R. P. Ellis Wilmington, DE
RES(TX), Inc. Operating Personnel
Environment One Corporation
Mike Kawahata
Stack Sampling team
ii
-------
/
-------
lie
LIST OF TABLES
I	Physical and Chemical Properties of Feed Streams
II	Fuel Feed Rates During PCS Incineration Test
III	Operating Temperatures During PC3 Incineration Test
IV	Combustion Efficiencies During PCB Incineration Test
V	Operating Temperatures, Flue Gas Concentrations and Air flows
VI	Analyses of Water Streams
VII	Concentrations of Certain Components in the Flue Gas During
the PCB Incineration Test
VIII Metal Concentration in the Incinerator Feed During the
PCB Incineration Test
IX PCB Destruction Efficiencies During PCB Incineration Test
X Scrubber Efficiency Based on Chlori de-balance During the
PCB Incineration Test
XI Incineration Residence Tim During the PCB Incineration Test
XII Quality of Outfall 001 Effluent During the PCB Incineration Test
nil PCB Concentration in the Final Equipment Flush Using Auxiliary Fuel
iv
-------
/£ b
LIST OF FIGURES
1.	Plant Layout
2.	Schematic of Incinerator Complex
3.	Schematic of Incinerator Stack
4.	Schematic of Hot Duct
5.	Incinerator Complex-Plan View
6.	PCB Feed Piping Layout
7.	Schematic of Carbon Monoxide and Lew Temperature Automatic Cut-offs
8.	Control Circuitry for PCS Incineration Test
4
9.	Schematic of the Carbon Monoxide and Oxygen Continuous Analyzers
10.	Volume of Afterburner and Loddby Burner
11.	Volume of Hot Duct
12.	Fan Curves
13.	Loddby and Hot Duct Temperatures During PCB Incineration Test
(strip charts)
14.	Kiln Temperatures During PCB Incineration Test (strip charts)
15.	Carbon Monoxide and Oxygen Concentrations During the PCB
Incineration Test (strip charts)
16.	Water Flows to Scrubber System During PCB Incineration Test
(strip charts)
17.	&P Across Venturi Throat During the PCB Incineration Test
(Circular Chart)
18.	Scrub Water Retention Tank: pH During the PCB Incineration
Test (strip charts)
19.	The pH of Outfall 001 affluent During the PCB Incinerate
Test (strip charts)	lon
v
-------
\i*7
LIST OF .APPENDICES
I. PC3 Incineration Test: Operating Data Logs
II. Letter: Ms. A. Harrison, U.S. EPA to J.D. Neel, RES(TX), Inc.
Letter: J.D. Neel, RES(TX), Inc. to Ms. A. Harrison, U.S. EPA
Letter: ¦ W.D. Langley, U.S. EPA to Ms. D. Cass, NUS
Letter: J.W. Bright, NUS to W.D. Cooper, RES(TX), Inc.
III. Mass Flow Rate of Flue Gas (Calculation)
IV. Southern Petroleum Laboratories, Inc Certificate of Analysis
No #78196-78199
V. NUS Corporation Data Sheets:
Composite Sauple of PCB material in T-Z7
PCB Analyses for Fuel Composites
PCB Analyses for Test No 1 Water Samples
PCB Analyses for Test No. 2 Water Samples
PCB Analyses for Test No 3 Water Sauries
PCB .Analyses for Test No 4 Water Samples
PCB Analyses for Flushes
VI. Environment One Corporation
Report on Stack Gas Sampling and Analyses
vi
-------
Oft
SUMMARY
A liquid waste containing 35 wt% PC3's was incinerated
at Rollins Environmental Services (TX), Inc in Deer Park,
Texas with a destruction efficiency >99.99997% and a combustion
efficiency of 99.993%. A chloride balance showed that the
scrubber was 98.4% efficient. The minimum incineration control
temperature was 11464 C; and, the dwell time was calculated as
2.68 seconds.
Continuous monitoring of the flue gas showed the carbon
monoxide concentration was <12 ppm and the oxygen concentration,
was >7%.
All automatic cut-off systems were operational during the
incineration.
No PCB's, chlorinated aibenzofurans nor, chlorocarbons
were detected in the flue gas. Water quality of Outfall OQl
water was well below the specifications in TDWR Permit No OI429
and NPDES Permit No TXOOOS941 and contained no detectable PCB»s
1
-------
/
INTRODUCTION
In accord with. C?R 40- 761.40 RES (TX) , Inc submitted to
the Regional Administrator of U.S. EPA Region 6 an application
for permission to incinerate liquid and solid (mainly capacitors)
wastes containing PCB's at its Deer Park, Texas facility. In
this application of April 5, 1978 RES (TX), Inc. asked, in the
main, for relief from a trial incineration of PCB's because it
had incinerated PCB's under subcontract to TRW which was operat-
ing as a U.S. EPA contractor.^ When this relief was denied
RES (TX) , Inc submitted a revised application for liquid and
solid PCB wastes on February 28, 1979: having submitted on
February 14, 19 79 a proposed plan for conducting and monitoring
a trial incineration.
Subsequently, RES (TX), Inc encountered mechanical problems
with the capacitor shredding machinery and elected to temporarily
delay that part of the test incineration relating to solid wastes.
In accord with this decision RES (TX), Inc submitted a second
revised application on October 4, 19 79 which pertained solely to
the incineration of liquid PCB's.
After several communications amongst Region 6, RES (TX), Inc,
and, RES (TX), Inc sub-contractors personnel, agreement to the
proposed plan was reached and the incineration test was made
during the week of November 11, 19 79 at Deer Park, Texas.
As sub-contractors to RES (TX) , Inc, Environment One^^ took
(1)	Destroying Chemical Wastes in Commercial Scale Incinerators -
Facility Report No.6 Rollins Environmental Services, Inc., Ackerman,
D.G. , et al, U.S. Environmental Protection Agency, Office of Solid
Waste Management Programs, Washington, D.C. 20460, June 19 77.
(2)	Environment One Corporation, 2773 Balltown Road, Schenectady,
NY 12309.
-------
/70
the stack gas samples and analysed them. NUS ^ anal/zed the
water samples for PC3's and the incinerator feed streams for
metals content. The elemental composition of the incinerator
feed streams was determined by SPL. Certain other parameters ,
such as pH and Total Organic Carbon (TOC) , of the water samples
were measured in the RES (TX), Inc laboratory.
THE PC3 TEST EQUIPMENT
These incineration tests were made at RES (TX) , Inc plant
site at 2027 Battleground Road, Deer Park, Texas (Fig l).'-5)
The incinerator complex consists of a Bartlett-Snow rotary
kiln and a modified Loddby liquid burner both discharging into
a common afterburner chamber. Flue gas from the afterburner
goes to a Flexi-Venturi scrubber (dilute lime slurry) and a
bubble tray absorber (water) to remove particulates and acidic
gases. The scrubbed flue gas leaves through a stack. This
complex is shown in Figures 2,3,4,5, and 6.^
The temperatures throughout the complex were measured with,
certified, calibrated thermocouples (Figs 3 5 I) and the results
continuously recorded (Figs 13 5 14). A minimum temperature
profile was maintained by using the hot duct thermocouple
(Fig 4) to activate the automatic feed cut-off (Figs 7 5 8)
TT1 NUS Corporation, 900 Gemini Avenue, Houston, TX 77053. — —
(4)	Southern Petroleum-Laboratories, Inc , P.O. Box 20807 ,
Houston, TX 77025.
(5)	A U.S. Dept of Interior, Geologic Survey topographic map
the La Port'e, Texas quadrangle and an aerial map of RES(TX), Inc
plant site are shown in Appendices 1 and 2 to Rollins Environment
Services Report and Application for Approval of Proposed PCB
fill submitted to U.S. EPA Region 6 on" December 1. !*''«.	-L
(6)	A detailed discussion of the principal parts of this complev-
is given in Application for Approval to Incinerate Polvchloriwar* .
3ishenvl Compounds and Wastes at Rollins Environmental' servic-a^-rSS
Inc. Jeer Park, Texas, suomitted to U.S. EPA Region 6. reariia^r^sLfi
1379.
-------
17/
whenever that temperature fell below 1113®C.
The pressure drop across the Venturi throat was continuously
measured and recorded (Fig 17) to showAp was maintained above
40 inches (water) .
A Beckman Model 865 NDIR continuous analyser and controller
for carbon monoxide (CO) was used with the instrument set to
automatically shut off the PCB feed whenever the CO concentration
exceeded 80 ppm (Figs 7 3 8). A Thermox WDG oxygen (03) analyzer
was used to measure and record the O2 concentration in the flue
gas. Additionally, the Thermox WDG was set to automatically
shut off the PCB feed whenever the O2 concentration was less
than 3% (5igs 85 3),
The operation of the temperature, carbon monoxide and
oxygen cut-off mechanisms was demonstrated prior to the test.
By mutual agreement between Region 6 and RES(TX) the
carbon dioxide concentration was measured by the Orsat Method
in grab samples in lieu of continuous analysis with recording
as specified in CFR 761.40-7, and, the required combustion
efficiency in CFR 761.40-2 was increased from 99$ to 99.9%
(Appendix II).
THE PC3 INCINERATION TEST PROCEDURE
During the period October 25th through November 8th, 1979
RES(TX), Inc obtained from several local companies S2 ,900 kgs
of PCB liquid wastes which were blended in a baffled and mech-
anically stirred tank (T-27). Analysis of a sample of this
blend showed 470,000 mg/kg of PCB's and a specific gravity of
1.340: individual isomers were not identified but Arochlors
1254 and 1260 appeared to be predominant (Appendix V).
4
-------
17 a.
A liquid waste material (in storage tank T-6G) was used as
th.e auxiliary fuel for these tests as well as for the background
test (Test 1), and its properties are shown in Table I, Test
No 1. The PC3 fuel used for the three PC3 tests was prepared
by blending 22,700 kgs o£ auxiliary fuel with the 52 ,900 kgs of
PCB wastes in the blend-feed tank:' the properties are shown in
Table I.
During each test, grab samples were taken from the feed
line (Fig 6) every IS minutes and composited by consecutive tvo
hour periods. Each composite sample was analyzed for PCB con-
centration, see Table I and Appendix V. Those composite saaplas
were in turn composited as one sample for each test and analyse
for elemental composition, heat of combustion, ash content,
chloride scrub and specific gravity (Table I).
During all tests auxiliary fuel was fed to the kiln at • an
average rate of 840 kg/hour. For Test No 1, the background
auxiliary fuel was fed to the Loddby liquid burner at an average
rate of 2,151 kg/hour. For the PCB tests, Numbers 2,3, and 4, ^
PCB fuel was fed at average rates of 3,253; 2,996; and 3,253 kg/
hour, respectively (Table II).
While the tests were being made no other fuels were fed to
either the kiln or the Loddby liquid burner. Moreover,, each t®st
was preceeded and followed by the incineration of only auxiliaj-y
fuel for periods of at least 30 minutes. Once the testing and
auxiliary fuel flushing were completed, RES(TX), Inc reverted
the normal incineration of wastes.
During each test the following sampling programs pertained.
Grab samples of the Loddby fuel were taken every fifteen minut®,
5
-------
i73
(Figs 2 and 6). Grab samples of water streams were taken every
thirty minutes: well water, lime slurry, touch-up tank (TUT)
effluent (Fig 2) and Outfall 001 (Fig 1). Stack samples were
taken during a 2 40 minute time period (Fig 3). Grab samples
were taken from the hot duct approximately every two hours for
carbon dioxide (C02) analysis by the Orsat Method (Fig 2).
THE PCB INCINERATION TEST RESULTS
Only auxiliary fuel was used in the background test made on
November 12, 19 79. The average feed rate to the Loddby burner
was 2,151 kg/hr for the 448 minute test period (Table II, Test
No 1). The average temperatures in the afterburner and the hot
duct were 1262*and 1183° C, respectively (Table III, Test No 1).
With an average CO2 concentration in the hot duct of 8.3 vol!
and an average CO concentration in the stack of 0.3 ppm the
combustion efficiency (?) was 99 .999% (Tables IV and V).
No PCB's were detected in the auxiliary fuel (Table I,
Test No 1), nor in the water streams (Table VI, Test No 1), nor
in the stack gas (Table VII, Test No 1).
The replicate PCB-fuel incineration tests, Numbers 2,3, and
4, were made on November 13,14, and IS, respectively, with the
following (average) conditions. The fuel had a PCB content of
34.8 wt% and a Heat of Combustion of 5,770 kcal/kg (Table I).
The feed rate to the Loddby liquid burner was 3,170 kg/hr for
391 minutes (Table II). The temperatures in the afterburner and
the hot duct were 1304*and 1183*C, respectively (Table III). A
combustion efficiency of 99.993% was acheived with a CO2 concen-
tration of 9.1 vol* and a CO concentration of 6.9 ppm (Table IV).
(7) Defined as 100 x {(C02) - (C0)}/(C02).
6
-------
I7H
The stack gas samples showed no detectable amounts of PC3' s
(Table VII). (The limit of detectability of PC3's is 1 ppb and
no detectable amount of PCS's is reported as "less than 1 ppb'*
or, "<1 ppb") The input water samples showed no detectable PCB's
in the well water and 1 ppb in the lime slurry (Table VI, Test
Nos 2,3, and 4). The validity of the 4 ppb reported for the
lime slurry composite in Test No 2 is questioned. The output
water samples showed 1 ppb in the TUT effluent and no detectable
PCB's in Outfall 001 water (Table VI, Test Nos 2,3, and 4).
Effluent water from Outfall 001 is regulated by Texas
Department of Water Resources (TDWR) Permit No 01429 and b>;
National Pollution Discharge Elimination System (NPDES) Permit
No TX0005941. During the test period, November 12 through 17
1979, the quality of Outfall 001 effluent water was in compile
ance with the limitations specified in both permits (Table XIX)
Samples of stack gas were taken and analyzed according to
the applicable U.S. EPA methods discussed in their Interim Rep0rt
of February 10, 1978^8^. As mentioned above, there were no PC3»S
detected nor were there any RC1 or chlorinated dibenzofurans.
the backbround test with only auxiliary fuel used, the stack
emission rate was IS.34 kg/hr for particulate, 8.8 kg/hr for
total chloride, 0.S8 kg/hr for NOX and 3.8 kg/hr for C02. During
the three tests with PC3-fuel being fed to the Loddby burner the
average emission rates were 17.65 kg/hr for particulates, 27.7
kg/hr for total chloride, 0.64 kg/hr for NOX and 4.1 kg/hr for
COj. (The data for each test is shown in Table VII).
(8) Sampling Methods and Analytical Procedures Manual for PC.n
DispoJT1: Interim report, aeard. J.ri. III. ana Schaum, J.,
of solid Waste, U.S. Environmental Protection Agency, Wa siiiagto^,®
D.C., February 10, 1978.	»
7
-------
I7f
At the request of U.S. EPA Region 6 the stack gas sampling
time period was set at 240 minutes to be able to demonstrate a
PC3 destruction efficiency of 99.99999%. As shown in Table IX
the average destruction efficiency was >99.99997%.
The scrubber efficiency, calculated from total chloride, is
shown in Table X for each test. For the background test, Test
Mo 1, the efficiency was 98.8% and for the PCB tests the average
efficiency was 9 8.4%.
The average residence time in the hot zone (Loddby chamber,
afterburner and hot duct) was calculated as 2.68 seconds based
on fan performance curves (Fig 12) and standard methods (Appendix
III) .
During all tests, the following data was continuously re-
corded: afterburner and hot duct temperatures (Fig 12); kiln
temperature (Fig I4); carbon monoxide and oxygen concentrations
(Fig IS); water flows to scrubber system (Fig 16);AP across
Venturi throat (Fig 17); the pH o£ scrubber water retention tank
(Fig 18); and, the pH of Outfall 001 outfall water (Fig 19)
EQUIPMENT FLUSHING
PC3 incineration Test No 4 was completed at 0108 hrs on
November 16, 1979 and the remaining PCB fuel was incinerated
under the prevailing conditions.
The tank trailer, blend-feed tank and all the auxiliary
equipment used to transport, transfer, blend and feed the PCB
tuel were flushed three times with auxiliary fuel and each flush
was incinerated separately.
For the first flush 7500 kg of auxiliary fuel was pumped
into the tank trailer which was then driven stop-and-go-wise for
3
-------
17b
at'least 30 minutes to slash the fuel in the tank trailer. The
flush fuel was pumped into the blend-feed tank and was- fed to the
Loddby liquid burner. This sequence was repeated once [second
flush).
After the third tank trailer load of auxiliary fuel was pumped
to the blend-feed tank an additional 48,900 kg of auxiliary fuel
was pumped into the blend-feed tank for a final tank flush of
56,400 kg. During the incineration, three grab samples were
taken from the feed line, each of which showed <1 mg/kg PCB's.
The data for the incineration of the excess PCB fuel and the
three equipment flushes are shown in Table XIII and Figs 13, 14.
IS, 16, 17, 18, and 19.
ANALYTICAL PROCEDURES
Both the NUS and Environment One laboratories demonstrated
their capability to analyze for PCB's using PCB test samples
submitted by the U.S. EPA (Appendices II and VI).
NUS laboratories used the method in Attachment 3 in the
f 81
U.S. EPA Interim Report. The analyses for metals in the fuel
feeds were made by Atomic Adsorption Spectroscopy using an IR-
Model 2S1 AA Spectrophometer (Table VIII).
SPL laboratories sub-contracted the carbon-hydrogen analyses
to Schwarzkopf Microanalytical Laboratory in Woodside, New York.
The sulfur analyses were made using a Dohrman Microcoulombmeter:
the results (Table I and Appendix IV) could be high because of
positive chloride interference. The nitrogen content was deter-
mined by the Kjeldahl method. After the sample was digested in,
a Parr bomb, the chloride was determined by titration with silve^
nitrate solution.
9
-------
n 7
NUS laboratories after extensive investigation concluded
that no accepted procedure existed for the detection of trace
amounts of polychlorinated dibenzofurans in PC3's (see letter
NUS to RES, Appendix II).
CONCLUSIONS
Liquid wastes containing 351 PCS's were incinerated with
a destruction efficiency >99.99997% and a combustion efficiency
of 99.99 31, The Flexi-Venturi/bubble-cap tray absorber scrubbed
the flue gas with a 9 8.4% efficiency, based on the chloride
balance.
Continuous monitoring of the flue gas showed that the CO
concentration was less than 12 ppm, the O2 concentration was
o
greater than 7% , and the minimum hot duct temperature was 1146 C.
The incineration dwell time was 2.68 seconds.
The effluent water quality at Outfall 001 was well below
the specifications in the TDWR Permit No 01429 and the NPDES
Permit No TX0005941. The quality of the flue gas emitted from
the incinerator stack was below the specifications in TACB Permit
No R-679.
All controllers, recorders, and automatic shut-offs re-
quired by CFR-761-4Q and the Regional Administrator for U.S.
EPA Region 6 were functional during the incineration tests.
10
-------
TAUI.E I
PHYSICAL AND QUMICAL PKOPERTIES OF FEED STREAMS
PCB Incineration Test Novenfoer 12-16, 1979
Ifaui
No.
Elemental Cowposit ion^wtl
Concentration of PCB's^in 3
Composited grab samples , X10
II
CI
N
1
JSfi

lieat
offcl
foiiib lCJ
kcal/kg
(hloride
Ash , ,¦» content S|>ec
Content, of .ami >. grav
wtl t;as wt *» . hn/I
I
63.65
8.86
24.2
0.072
0.072
<0.001
<0.001
<0.001
10,409
0.2
13
0.996
2
48.58
4.66
43.8
0.027
0.036
330
380
380
6,700
nd
38
1.261
3
50.94
4.73
54.3
0.015
0.035
360
380
360
5,700
0.7
54
1.189
4
50.44
4.80
56.3
0.016
0.034
325
340
310
310 4,900
0.5
50
1.261
(a).	Analyses were uiacie by SPL with the C/ll subcontracted to Schwartzkopf Micro-analytical Lab.
(b).	Analyses were made by NIJS Laboratories.
(c).	Analyses wew made by RES (TX) Laboratory.
nd = none detected
-------
TABLE II
R1EL 1EED RATES HIRING PCB INCINERATION TESTS
November 12-15, 1979
Total	Average feed rate
Duration >	feed,..
Test	of test/ '	used;
No.	mins	liters	liters/iuiu kg/lu
1	448	16,065	36	2151
2	368	15,687	43	3253
3	390	16,519	42	2996
4	415	17,766	43	3253
(a).	Duration of incineration of fuel was calculated from data log sheets reproduced
in Appendix 1.
(b).	Calculated from recorded gallons multiplied by 3.78 liters/gallon.
-------
TABU- III
Oi'liRATING TEMPIiRAHiRES WRING PCB HOMiHATION TESTS*8*
Novender 12-IS, 1979
Teiii|>erature,


llot Duct


Afterburner

Kiln


Ambient
Test
No.
Mill
Max
Ave
Min
Max
Ave
Min
Max
Ave
Min
Max
Ave
]
1161
1210
1183
1197
1301
1262
709
875
737
14.4
17.5
16.4
2
1113
1218
1178
1211
1338
1280
737
883
822
13.3
20.0
15.3
3
1147
1212
1186
1269
1360
1312
762
925
871
15.6
17.2
16 6
4
1148
1220
1184
1252
1396
1321
709
988
854
12.8
15.0
13.6
(a). Values shown were calculated from data in operating sheets reproduced in Appendix I.
(h). Conversion from *l: to *C was made with the use of the temperature conversion table in
Handbook of Chemistry, I.ange, N.A., Ed., llandliook Publishers, Inc., Sandusky, Ohio, 1941.
-------
/*'
TABLE IV
CCMBUSTICN EFFICIENCIES DURING PC3 INCINERATION TESTS
Clock
time
hrs
November 12-15 , 1979
Carbon,,
dioxide,
vol%
Carbon
monoxiae,
Dtnn
Cb)
Combustion
efficiency,
%
Cc)
1455
1640
1845
1950
Average
8.0
8.0
8.6
8.4
0.50
0.00
0.50
1.75
99.999375
ion.000000
99.999419
99.997917
mrm—
1515
1700
1905
Z015
Average
9.3
10.0
9.2
9.0
6.5
6.0
4.5
5.5
99.993367
99.994000
99.995109
99.993889
99.994
1130
1245
1430
1630
Average
9.0
9.2
8.4
8.2
11.0
10.5
9.75
9.0
99.987778
99.988587
99.988393
99.989024
	
1845
2010
2150
2440
Average
9.6
8.4
9.2
8.8
6.0
3.0
5.5
5.0
99.993750
99.996429
99.994022
99.994318
¦goss—
(a).	Data from Environmental Ote, Inc. report reproduced in Appendix IV.
Samples of hot duct gas were analyzed for carbon dioxide by the
Orsat method.
(b).	Data from Operations data sheets reproduced in Appendix I.
(c).	Defined in 40CFR 761.40 a2 as
SSzl^JSSL x ioo
(COO
14
-------
TABLE V
OPERATING TEMPFRAIHIiFS. FllJE )
r;>Tc)~ is-
I 10(1
1301
1174
875
17.0
1.75
8. 15

67
32
1115
1266
1190
767
17.0
1.50
8.15

67
32
1 130
1206
1161
712
17.0
1.25
7.73

67
32
14-15
1270
1178
730
17.0
0.50
7.94

67
32
1500
1255
1194
754
17.2
0.50
8.57
8.0
65
32
ISIS
1253
1196
754
17.2
1.00
8.57

65
32
153ft
1254
1196
759
17.5
0.00
8.IK

65
32
1545
1246
1188
727
17.5
0.00
8.57

65
32
1600
1237
1196
747
17.5
0.00
8.99

65
32
1615
1242
1195
748
17.2
0.75
8.57

65
32
1630
1242
1195
746
17.0
0.00
9.41

65
32
1645
1242
1195
749
16.7
0.00
9.41
8.0
65
34
1700
1253
1191
719
16.4
0.00
9.82

65
34
1715
1290
1203
709
16.1
0.00
9.82

65
33
1730
1229
1183
746
16.1
0.00
10.24

65
31
1745(d)
1197
942
437
16.1
6.0
10.24

65
32
1800
1292
1210
816
16.1
0.00
10.24

65
34
1815
1296
1175
710
16.1
0.50
9.82

65
33
l.S 30
1262
1201
741
16.1
1.00
10.87

65
33
1845
1272
1192
738
16.1
0.50
10.45
8. b
65
33
1900
1278
1199
738
16.1
1.25
10.87

65
31
I'M 5
1288
1201
752
16.1
1.00
10.45

65
33
1 930
1266
1203
746
15.8
1.00
11.08

65
32
1«J45
1280
1196
753
15.6
0.50
10.24
8.4
65
33
2000
1278
1195
750
15.6
1.75
10.15

65
30
2015
1276
1196
746
15.0
1.50
10.03

65
30
2030
1282
1196
740
14.7
1.75
10.03

65
30
2045
1280
1198
737
14.4
2.25
10.21

65
31
(•V	Smiinarizod from original data sheets reproduced in Appendix I.
(ft)	llot diict gas analysis for carim dioxide by Orsat uvtiiod.
(c)	H) - forced draft fan; II) = induced draft fan.
(d)	At 1738 lirs. a line plug s'.oppod feed ami decrease . . , ..turn*. ,.^h:i.»s»uu	anlv.it«m» .» »»*->
-------
TABUF. V (cont.)
OPERATING TLMP1KA11IRFS, FIJUE GAS CONCENTRATIONS AMP AIR FLOWS: TEST SO. 3 (a)
:'U>ck

•





Fan
louvres,
t ;

Temperature, C


Has
Moni tor


open
f'l)~
Afterburner
llot ftict
Kiln
Ambient
CO. ppm
°2, 1
C02.
Fjjfc)" in
10 IS
1342
1147
762
15.6
14.0
7.11

70
18
1100
1350
1188
881
15.6
14.0
7.32

70
18
Ills
1360
1203
882
15.6
12.0
7.32

70
l'J
1130
1342
1198
878
16.1
11.0
7.73
9.0
70
19
11 IS
1348
1198
876
16.4
11.0
7.32

70
19
1200
1343
1203
881
16.4
11.5
7.73

70
l'J
1215
1338
1203
867
16.4
11.0
7.84

70
19
1230
1339
1201
885
16.4
11.0
7.84

70
18
12-15
1335
1198
856
16.1
10.5
8.15
9.2
70
18
1500
1338
1203
870
16.4
11.0
8.15

70
19
1315
1335
1204
865
16.7
10.0
7.94

70
19
1330
1325
1194
860
16.7
9.0
8.15

70
19
1345
1329
1212
877
16.7
9.0
8.36

70
20
1 100
1298
1195
873
17.0
10.5
9.41

70
22
1415
1271
1178
859
16.7
11.0
9.82

70
22
1 130
1273
1174
878
16.4
9.75
9.A1
8. 1
70
21
1145
1275
1173
881
15.6
9.5
9.82

70
22
1500
1277
1179
884
16.7
9.5
9.41

70
22
1515
1284
1180
881
17.0
9.0
9.82

70
22
1530
1277
1173
850
17.2
7.0
10.66

70
23
1515
1272
1171
838
17.2
8.0
10.45

70
22
1600
1283
1177
856
17.2
9.0
10.66

70
22
I (.IS
1278
1175
843
17.2
9.0
11.91

70
22
1630
1278
1175
893
17.2
9.0
10.87
8.2
70
22
1645
1269
1172
925
17; 2
9.0
10.24

70
22
1700
1304
1187
887
16.7
9.0
9.82

70
22
1715(e)
1321
1157
878
16.7
8.0
8.36

70
23
1730
1341
1199
872
16.4
7.0
8.98

70
23
(a)	Summarized from original data sheets reproduced in Appendix I.
(It)	llot duct gas analysis for carbon dioxide by Orsat method.
(c)	I'D = forced draft fan; II) = induced draft fan.
(d)	Ou PCB feed at 1030 Ins.; low tem|>e rat lire shut-off at 1034 hrs. then on at 1042 hrs.
lei	Off fuel at 1708 Ins: remainder of time on hacKgrotud fuel purge.
-------
TAIUJE V (cont.)


fif'HRATINfi TBff'f laiURfs
. Flirt
GAS CONCEYIKATinNS AND AIR
FI/JVS:
TfcST NO. 2


1 vk







Fan
louvres,
t i,i:e,

Temperature, C

Gas Monitor

9.
»
open
:n -.
Atteilmmer
lk)t ftict
Kiln
Ambient
CO, ppm
-» %
C02.t(l>)
FD
fcT ID
1130
1243
1154
772
20.0
4.5
7.32

72
20
14 IS
1243
1204
830
19.4
6.0
7.32

70
23
isoo
1246
1218
773
19.2
4.0
7.32

70
20
ISIS
1227
1216
760
18.3
6.5
7. 73
9.8
70
20
1S30
1211
1195
747
18.1
6.0
8.15

70
22
ISIS
1215
1184
771
17.2
5.5
8.15

70
20
l(.00)
(c)
00
(e)
I lot duct gas analysis for airbon dioxide by Orsat method.
H> = force draft fan; II) = induced draft fan.
JO sauftle collect Jon train plug/jed: /O flou off at 1605 Jirs. then on at 1025 hrs.
Off IKS fuel at 20SO Ins.: reminder of time on background fuel purge.
-------
TABLE V (cont.)
PP!:KVI I\r. IIMI'lianiUFS, FLUE c;as cokcENTRAT1lA'S .and air FLO'S: TEST \0. 4 (a)
• . '-a
1 ! .".Hi ,

Temperature,
'C

f-as
Von itor

® open
F

Afterburner
Hot ftict
Kiln
Ambient
CO, ppm
~)
lc)
JAl
Hot duct yas analysis for carbon dioxide by Orsat method.
11) = forced draft fan: Il> =	araf<;	.
Off lfB fuel at 2212 his., on fuel at 2227 Ins.
-------
TABl£ VI
ANALYSIS 01- WATER SITU-AMS: TEST MO. 1
Well water
Lime slurrv feed
I
HIT effluent
I
Oil fall 001
I
Composite
N6.(a)
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
6
PCB coiic, (b)
PPb
Chloride conc,(c)
"¦g/1	
<	1
<	1
<	1
<	1
<	1
<	1
40
40
SO
Sulfate conc,(c)
	mg/1
nd
nd
25
Pn(c)(d)
8.1
7.9
7.3
a.
b.
c.
d.
e.
C.rab samples collected every 30 minutes were composited by consecutive 2 hour periods.
\nalyses were made by MJS l.al)oratories.
Analyses were made by Rl:S(Tx) laboratories.
Standard units via glass electrode.
IhS = total dissolved solids.
f. ItC - total organic carbon.
T0S(c)(e)
hh;/1
300
310
350
m ,(c in i
mi!11
34
15
53
< 1
4499

105
3.8
8200
4
< 1
4300

140
2.9
7550
3
< 1
4300

140
7.9
84 50
19
< 1
3800

85
6.9
6900
4
< 1
3900

125
7.0
7200
8
< 1
3699

102
7.4
7000
4
--
--

--
--
--

oC
-------
TABU: Vi (cont'd)
ANALYSUS OF WATliR STREAMS: TEST NO. 2
i Well water
l ime sliiirv feed
i
o i
lUT effluent
(litfall 001
Composite
No.(a)	
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
PCB conc,(b)
		
<	1
<	1
<	1
1
< 1
4
Chloride cone, (c) Sulfate cone,(c) p"(c)(d)
mg/1		rog/l	
200
50
5
40
8.5
8.2
Tl>S(c)(e)
	mg/1
790
392
a.	Oral) samples collected every 30 minutes were composited by consecutive 2 hair periods.
b.	Analyses were made by NMS Laboratories.
c.	Analyses were made by RliS(Tx) laboralories.
<1. Standard units via glass electrode.
e.	il)S = total dissolved solids.
f.	T(Xr - total organic carbon.
Tu:,(c)(fi
42
36
1
7189
90
1.9
12200
83
< 1
74 9 8
95
2.3
12550
61
1
7800
85
11.3
15350
7
2
4200
95
7.4
6900
129
< 1
4200
120
7.1
7550
92
< I
4300
110
--
7200
46
-------
i Well water
I.line slurry feed
l
I
^ i
- i
HIT. effluent
. anfall 001
i
TABU VI (cont'd)
ANALYSIS OF WATER STREAMS: TEST NO. 3
Canposite
No.(a)
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
IXZB cone, (b)
	l^1'	
<	1
I
<	1
Chloride conc,(c)
		
SO
S000
Sulfate cone,(c) j>*'(c)(«J)
.	mil				
30
95
8.S
8.7
Tl»S(c)(e)
	
480
7824
IU , (c jll )
118
82
1
1
( )
< 1
2
1
<	1
1
<	1
7898
82
3.4
14400
88
9497
105
8.1
13650
88
5498
95
8.7
7800
82
4697
4700
4750
88
125
50
7.2
8.3
8.4
8200
13400
7650
60
101
207
caO
a.
b.
c.
d.
e.
f.
f.rab samples collected every 30 minutes were composited by consecutive 2 liour periods.
Analyses were made by NLIS laboratories.
Analyses were made by RLS(Tx) laboratories.
Standard units via glass electrode.
TliS = total dissolved solids.
TIC- - total organic carbon.
-------
TABLE vi (cont'd)
ANALYSES OF WATER STREAMS: TEST NO. 4
Composite
No.(a)
PCB cone, (b)
PHb	
Chloride cone,(c)
'"H/1	
Sulfate conc,(c)
	i»g/l
i>ii(c)(a)
v>
M • III Ml' iJUIMI '•	— •»»»•• J »»"	nv| *
b.	Analyses were made by WIS laboratories.
c.	Analyses were made by RliS(Tx) laboratories.
d.	Standard units via glass electrode.
e.	TUS = total dissolved solids.
f.	TCI' - total organic carbon.
TDS(c)(e)
•i^/1	
TOC,(c)(f)
Well water
1
< 1
400
nd
8.5
360
122
•
2
2
300
nd
7.7
390
107
1
3
1
SS
40
7.9
352
61
1
X
1
4
5
1
40
30
8.2
308
6
Lime slurry feed
1
1
< 1





i
i
L
3






j
4
1





i
i
»
5
--





'1UT effluent
1
< l
8597
95
9.9
11800
S4

2
1
7998
85
10.2
13800
57

3
1
8997
62
10.8
13840
44

4
5
2
2
8400
105
8.5
8400
114
'Outfall 001
1
< 1
5400
65
6.1
9250
130
|
2
1
5300
75
6.8
10100
38
j
3
1
5198
75
7.0
9500
12
•
1
4
< 1
5800
85
6.3
8250
29
1 .
S

— ~
~ -
- -
- -
	
-------
TABLE VII
CONCENTRATION OF CERTAIN COMPONENTS IN HIE FLUE GAS
WIRING TIIE PCB INCINERATION TESTS (a)
Cotn>onent
PCB's
ItCI's
Benzofurans
Particulates
Total chloride
NOXU)
Carbon dioxide
Test l"
ad
nd
nd
IS. 34
8.B
0.58
3.8
Test
Concentration, kg/hr , ^
FTtn	 Test 3^"
nd
nd
nd
19.55
28.5
0.63
4.3
ltd
ltd
nd
18.11
28.6
0.58
4.0
Test 4^
nd
nd
nd
15.28
26.1
0.71
4.1
(a).	Sampling and analyses of flue gas were made by Environment One Corp.
((>).	[hiring Test 1 only backgromd fuel was incinerated.
(c).	During this Test the PCB feed was as described in Table I.
(d).	Polychlorinated dibenzofuratis.
(e).	Analyzed by EPA Method 5.
(f).	Analyzed according to "Sampling Methods Analytical Procedures Manual for PCU Disposal," Feb.10,1U78
(g).	Analyzed by EPA Method 7.
(It).	Values are average of 4 grab sauries taken from the hot duct and analyzed by the Orsat method.
-------
TABUS VIII
Metal
(b)
METAL CONCENTRATION IN HIE INCINERATOR FEED
IUR1NG 11IE PCB INCINERATION TESTS
Test 1
November 12-15, 1979
Concentration, mg/1
(a)
Test 2
Test 3
Barium
Cadiniiw
~iroiuiun
Copper
Lead
Manganese
Nickel
Silver
Zinc
<4
<2
84
4.4
<2
<2
<2
<2
5.8
<4
<2
8
<2
24
<2
<2
<2
<2
<4
<2
7
<2
34
<2
<2
<2
<2
(a). Analyses were made by NtIS laboratories.
Test 4
<4
<2
6.8
<2
34
<2
<2
<2
<2
(b). Standards were prepqred from organo-metallie conipoiaids and xylene. For the
standards 5 litis were dissolved in 100 mis of xylene.
-------
TABU: IX
u
in
Run
No.
\
2
3
4
IO DESTRUCTION EFFICIENCIES HIRING KB INCINERATION TEST
(a)
lO input.
Kilograms
(b)
<)»
4723
4398
4177
12
In flue
gas,
	cck_
<1
<1
<1
<1
(c)
PCB ouput
Iir™FdT
water;
PPb
<1
1
1.3
1.4
Total
kilograms
0.0011
0.0014
0.0015
AVERAGE
Destruct iou >
efficiency,
I
>99.9991)77
>99.999968
>99.999064
>99.999970
(a).	Based on the 240 niin. PCB stack sampling tine period.
(b).	Calculated from the flow rates in Table II and the concentrations in Table 1.
(c).	Values reported in Environment Clie report (Ari|iendix IV) and sliown in TableVII.
(d).	For the piir]>ose of this calculation all values in TableVI, except Run no.II, shown as
<1 (lielow detectable limit) were taken as 1.
(e).	Instruction efficiency is defined as {(input - output) / (input))100. Since the condition
in (d) was used the efficiencies are reported as greater than (>) the calculated value.
L
-------
TABIJ3 X
Run
No.
1
2
3
4
SCRUBBER EFFICIENCY BASED ON aILORIDO-BALANCE
DURING THE PCB INCINERATION TESTS
November 12-IS, 1979
Chloride input,^ kg/hr
I'CB fuel
to Ixtddhy
519
1413
1641
1824
(d)
Auxiliary
fuel to
kiln
183
183
183
183
Chloride
output, kg/hr
Flue gas
(b)
8.8
28.5
28.6
26.1
Scrubber
efficiency,^)
wti
98.8
98.2
98.4
98.7
(a).	Chloride analyses by SPL and recorded in Table 1.
(b).	Chloride analyses by Environment One and recorded in Table VII.
(c).	Define as {(Input-Output)/Input)100.
(d).	Auxiliary fuel was fed to the Loddby during this background test.
-------
TABU; XI
INCINERATION RESIDENCE TIKE DURING PCB
INCINERATION TESTS
Novender 12-15, 1979
Average Temp Density	Weiglit of	Residence
Rim	llot Zone,	flue gas,*3'	flue gas,*	* time,*0'
no. 	"R .	Ibs/cu ft	.lbs		sec
1	2763	0.0144	122.1	2.74
2	2796	0.0142	120.4	2.71
3	2853	0.0139	117.8	2.65
4	2870	0.0138	117.0	2.63
(a)	Calculated from the formula in#4.2 in SajnplinR Methods and Analytical Procedures Manual
for PCB Disposal: Interim Report, Beard, J.II., Office of Solid Waste, U.S. Environmental
Protection Agency, Washington, D.C., February 10, 1978.
(b)	Calculated froiu Density of flue gas and the volume of the hot zone (Loddby burner, after-
burner and hot duct) at 8476 cu. ft.
(c)	Calculated from the weight of flue gas in the hot done and the mass flow rate of the flue
gas (See Aj>|»endix III).
-------
TAB1E XII
QUALITY OF OUITALL 001 UFFLUONT UJIUNG
THE PCB INCINERATION TEST
Week of November 12t 1979
Limits, kg/day	
NPDfiS'aJ	TH®Cn	November, 1979*c)
V" — " ¦ k* I	H u ¦¦ ¦	^ ¦ .. I ¦	- ¦¦ ¦ — ... ¦	¦¦ I ¦ .if	¦ I-. I
Parameter
Daily
Av.
Daily
Max
Daily
Av.
Daily
Max
12 13
14
15
16
17
nons(d)

	
70
140
nd
_ _
nd
	
—
TSs(e)
132
263
103
205
9.6
--
12.7

5.5
T0C(f)
184
368
184
369
" (g)
--
25.5
--
10.5
Oil/Grease
30
60
30
60
3.6

0.3
—
0.4
ilienol ics
0.09
0.18
0.09
0.18
0.03
—
nd
--
nd
HI
6 to 9
, continuous record00
7.1/8.0
7.4/8.1
7.4/7.5
7.6/8.3
6.8/8.3
(a)	According to NPDBS Pemit No. TX0005941.
(b)	According to TDWR Permit No. 01429
(c)	Values fomd by RIIS(TX) laboratory using accepted analytical procedures.
(d)	Biochemical Oxygen Demand-5 day.
(e)	Total Suspended Solids.
(f)	Total Organic Carbon.
(g)	TOC instrument instable, no reliable result available.
00	liach day the minimum and maximum values shown on the strip chart are recorded.
-------
146
TABLE HI!
PC3 CONCENTRATION IN THE FINAL
EQUIPMENT FLUSH USING AUXILIARY FUEL
November 17, 1979 from 1650 to 1315 hrs.
NUS
Samgl^	sam^^	Total PCB concentration
nolai
f
no
-------
i<\1
FIG'JRE 1
PLANT LAYOU
anx i-arm
OOOOOO oooooo
3 0O00oo ooooooo
o 30
Scruboe
Water
Basin
5!ijl-Fire
I—'House
PCB Warehouse
Pressure tanks
Kiln
Loddby liquid burner
Contr-o-1 room
Water storage
Compressor building
Lime slurry facilities
After-burner
Stack
Touch-up tank (TUT)
Canal
Outfall 001
Outfall 201
Drainage ditch
Tucker Bayou
-------
11UIR1-: 2
SQUiHATlC OF INClNliRAlOR (DM'lJ-X
Afterburner
Feed Chute
Door
&
Kiln
Gas Flow
Direction
Liquid
burner
Feed
Liquid Feed
Sample Point
ORSAT Sampling
Point
Fresh
r Feed
T/C
Stack Gas ~
Sampling
Points\ I
Hot
uct
Stack
Scrubtter
indue el
Draft
Fans
Limb
SIurry
Feed
Di sc liarge
W« ter
Toucl/ up Tank
"^crubber
Water
Sampling
Point
-------
l
-------
thermocouple 151
—y
in
o
Q
14- iter 5
Ca< e
-------
FIGURE 5
INCINfiRATION SYSTEM - PI AN Vll-W
18ft.
Kiln
Exit
Duct
Kiln
ID 10ft. 6in.
Conveyor
11 ft.
Afterburner
16 ft.
/ I	,;xit
/ ' liiside Width 14ft. Duct
. Inside Height 13ft. 6in.
ID
5ft
i
-------
FIGURE 6
PCB FEED PI1MNG LAYOUT DURING PCB INCINERATION TEST
November 12-15, 1979
Tank *27
Control Valve
Background
ll'G Trailer
Bleeder
Valve
Background Fuel
Trip Valve
o
1.5 in.
line
PCB Trip
Valve
I;1 ow
Meter
Waste Liquid
Feed Giui
fx}
1.0 in. line
-------
103
FIGURE ?
SCHEMATIC OF CARBON MONOXIDE AND LOW
TEMPERATURE AUTOMATIC CUT- OFFS
Fuses
3-1
Power Supply
5x1 I	J	
rrr\~\
Low Temperature breaker C203S*F)
contact relay
/ High CO breaker (300 ppnO
contact relay
3-15 0n/0f£ Switch
3-16 On/Of£ Switch
D-o
Solenoid
Solenoid

3b
-------
20*4
FIGuPE 3

CTNTHDL dRCLTTSy FDR PCS INdMEPATICN TEST
Novenoer 12-15, 1379
normal PC3
Recorder
CR-3
Sectarian
Co
ncrnaL
Thermox
02
normal
..PCS
PCS
TD-263
i i R-l
q?


H»rmox O2
Bedanan CO
y\—A'/fj/f
CR-230 CR-1 CP
off
CR-2
OR
U
fuse
^T"
"As
OwT
"O^T
-O	
\	S L-l
o
R-2
O
TD-1
¦o
o^n
stop
"^-LP

Tenp below 2035*F
Terp relay operates
above 2035aF
Tenp permissive light
High CO
Low O2
WI2T7 by-pass
Shutdown kiln door
and conveyor
O2 permissive light
CO permissive light
WLTV
Alarm relay
Kiln door
Kiln conveyor
WLTV
waste liquid trip valve
TD
time delay relay
CS
control relay
M
motor ccntacter
3
relay
S
solenoid
r
light
XLAFM CONTACTS
nogral\ PCS
R-4
-------
zos
FIGURE 9
SCnEMATIC CF THE CARBON MONOXIDE .AND OXYGEN CONTINUOUS ANALYZERS
119 Volt
Cn/0££ Switc
2 Per
CONTROL
PANEL
UNIT |	
~
Rectordgrj _
• •
110 Volts
~l
Emergency Cn/Of£
"I
Capacitance alarm and
override circuits
1 jjZD Signal to Recorder^
1
PCB sqlid
feed
PCB lilquid feed
, Signal to Record**
s^tch
CUBICAL
FOR
INSTRUMENT
SIGNALS
Signal^
1 i
s td Cubical
i
I I
CO PROBE
I	
i:o volt
close kiln door and
stop kiln conveyor belt
Activate trip
valves

02 ANALYZER
N2 Gas Purge Line

-------
Z0(o
FIGURE 10
VOLUME OF AFTERBURNER /\ND LODDBY 3URNER
IM THE INCINERATOR COMPLEX
18ft
Kiln
exit
duct
lift
14ft
Afte/t/timer
Hot
Due
16f//6.2Sft
/ / insid^
/ Avdiam./
13 . Sf t
Height
Loddby
liquid
burner
	Af terburner	 	Loddby		^
18 x 14 x 13.5"	3402 V5- ir x 9.77 x 16" 491 cu. f t
V2» {(3.5 x 14)/2}x 13.S- 331
V3- {(S x 14)/2}x 13.S- 473
V4- 11 x 14 x 13.S-	2079
6285 cu.ft.
NOTE: All dimensions are inside.
I
3-1
-------
•207
FIGURE 11
VOLUME OF HOT DUCT IN THE INCINERATOR COMPLEX
5.25ft
diam.
17.7ft.
7.875ft
radius
6.5ft
h	*
JL.
5.25ft
1 _
<-
6.5ft
* 5.25ft
7.16ft diam.
90
elbow
4ft.
Venturi
Scrubber
3.00ft diam.
V3
V4
V*
7rC2 .625) 2C6.5) - 140.7
2,
141
267.8	536
rr (2. 625) " C17 . 7) - 383.2	383
tt/3{ C2. 625) " * C3. 25) 2 + (72. 78) **} (5 .25) - 142.9 143
ir(2 .625) (7. 875) Cir/2)
i
ir(3.25) (7.875) Cir/2) - 410.5
tt/3{ (3. 58) 2 * CI - SO) 2 + C28. 84)1l}C4.00)- 85.6
411
86
1700 cu.ft.
NOTE: All dimensions are internal.
40
-------
Sii
performance corves
FFgwmv—		J	^cltlcl
3o«j«u. h> i r>\? -±|^2i2il
| Cm w»7?P- >4 7 7*
•il Xe	
j 2. - ?U ^
i foO rpn, IjZO	*F	.	Hg .fl#72lt> c«i(t 6jr v/A!(^--
WL
i :•.
_nj-—1_
• .-,..
• • t .
~r~"
- i..
. : »
:: ' I- : j
^trr

1 • l -i-: :;
-—u —-j—
•¦I ¦ i i. : 1—-
ri-ihrri
I j' i^
. '0. & 2&2TI
—r-4-
I ¦ ' ''V r'"t'*'; **| '*' yf-*~	~ -• '• ¦.; j r::
'' ' "i ¦ : •!.»' ' -»-* ''1 • » '»<»»- '¦	['*-«	: t
:j J • .• . .1.:. ,	i;:;"T.r
52....
. t
_ :.j—
fo.


r I

"T.::
J.-:
•	I	«...
• « •! • I • '
y-y^7-'-~



.. » .
:: :-V.
•J...,
;:*i::




r. \ .r:
- — • • . ,4 «

•Ml? t —
hMs
H -v:t:::2
.	i.I-.. I..
•'In:.!*. 1.*? i•
-_L~J—" •¦ • ^ -..1^4
-l • V,

I...
:i.H
J.-•*• 4
» • . « - - • • r * I • ¦	•ll'tlil
h.^h fr
• . I «>J. 'MfMt.l	• • •« I Hi
~••I 
-------
10*1
FIGURE 13
AFTERBURNER -AND HOT DUCT TEMPERATURES CURING PC3
	INCINERATION TEST	
Run 1 incineration of background fuel
Noveniber 12, 1979 from 1400 to 2045 hrs.
ot du
ot duct
¦Afterburner
Afterburner
(a). Feed line plue?ed.causing temperature decrease which activated
shut-down a-t
42
-------
-2)0
FIGURE 13 (cont'd)
AFTERBURNER -AND HOT IXICT 7"3ff)EHA7LTRES CURING
	PC3 INCINERATION TEST	
Run 2 incineration o£ PC3 fuel
Novender 13, 1979 from 1430 to 2113 hTS.
I i i o.,.
terbumer
or* srack sanolina train slugged at 160 S hrs. and PQ feed was
shut-off? Problem was corrected and PC3 feed was started at 162S hrs.
Shut off PC3 feed at 2050-hrs. and purged system for 0.3 hr. with
background fuel.
43
-------
a"
FIGURE 15 (cont'd)
AFTERBURNER AND HOT DUCT TEMPERATURES DURING
	PCB INCINERATION TEST	
Run 3 incineration of PQ3 fuel
November 14, 1979 from 1045 to 1750 hrs.
Ho'Duct
fterburner
(a). Shut off PC3 feed at 1708 hrs. and purged system for 0.5 hr with
background fuel.
-------
7)7
FIGURE 15 (cont'd)
AFTERBURNER AND HOT DUCT TEMPERATURES CURING
	PCB INCINERATION TEST	
Pun 4 incineration of PC3 fuel
Novejnber IS - 16, 1979 from 1300 to 0100 hrs.
Hot' Duct*
'terbuxner
(a).	Plug in kiln feed line: shut off PCS feed at 2212 hrs.
(b).	Problem corrected and PCB feed resumed at 222? hrs.

-------
w
FIGURE 15 (cont'd)
.AFTERBURNER ; .AND HOT DUCT TEMPERATURES DURING
PC3 INCINERATION TEST
Burn-out of excess PC3 fuel
November 16, 1979 from 0108 to 0355 hrs.
Afterburner
46
-------
an
i
i
FIGURE 13 f cont'd)	I
' !
AFTERBURNER- -AND HOT OICT TEMPERATURES DURING
PC3 INCINERATION TEST
Incineration of First Flush
November 16, 1979 from 0450 to 0750 hrs.
77
47
-------

FIGURE 15 fcont' 
-------
Afterburner
-------
HI
FIGURE 15 [cont'd)
AFTERBURNER -AND HOT DUCT TEMPERATURES DURING
?C3 INCINERATION TEST
Incineration of Third Flush
November 16-17, 1979 from 1630 to 1315 hrs.
(concluded)
SO
-------
3/*
FIGURE 14
KILN TBPESATURES DURING ?C3 INCINERATICN TEST
Rin 1 incineration of background fuel
November 12, 1979 from 1400 to 2045 hrs.
51
-------

FIC^URE 14 (cont'd)
KILN TEMPERATURE DURING PCB INCINERATION TEST
Run 2 incineration of PCB fuel
November 13, 1979 from 1430 to 2115 hrs.
-------
2iv
FIGURE 14 (cont'd)
KILN TEMPERATURE DURING PCB INCINERATION TEST
Run 3 incineration of PCB fuel
November 14, 1979 from 1045 to 1730 hrs.
-------
w)
FIGURE 14 (cont'd)
KILN TEMPERATURE DURING PCB INCINERATION TEST.
Run 4 incineration of PCB fuel
November 15-16, 1979 from 1800 to 0100 hrs.
-------

I
i
FIGURE 14 (cont'd")
KILN TEMPERATURE DURING PG INCINERATION TEST
Burn-out o£ excess PCB fuel
Novcsnber 16. 1979 from 0108 to 0355 hTs.
-------

FIGURE 14 (cont'd)
KILN TEMPERATURE DURING PCB INCINERATION TEST
Incineration of First Flush
November 16, 1979 from 0430 to 0730 hrs.

&
Ltili11
56
-------
cm
FIGURE 14 (cont'd)
KILN TEMPERATURE DURING PCB INCINERATION TEST''
Incineration of Second Flush
November 16, 1979 from 0930 to 1215 hrs.
-------
7V:
FIGURE 14 (cont'd)
KILN TEMPERATURE DURING PCB INCINERATION TEST
Incineration of Third Flush
November 16-17, 1979 from 1630 to 1315 hrs.
*1-	r., i Mil ii'i .! mi r 1111 n 11 tn. i.: i.] i. i	«

LUi.i
ll
1
fct
i
; 'o!1'
nrhMiii

58
-------
FIGURE 14 (cont'd)
KILN TEMPERATURE DURING PCB INCINERATION TEST
Incineration of Third Flush
November 16-17 1979 from 1630 to 1315 hrs.
(concluded)
uiUf-trfrrffl
rrrffrhfi
I ¦ i

w
r1
-------
Ill
FIGURE 15
CARBON MONOXIDE AND OXYGEN CONCENTRATIONS
DURING THE PCB INCINERATION TEST
t ¦* M' / •" nm
. «l J ». ' if!.
Run 1 incineration of background fuel
November 12, 1979 from 1400 to 2045 hrs

jk** |j*n ;.!ii
I ! !
-rH-S
i-1.1 -

n

if
trizsti
:! I
! i1 ijj
_ujxj !,y
8;

60
-------

FIGURE 15 (cont'd)
CARBON MONOXIDE AND OXYGEN CONCENTRATIONS
DURING THE PCB INCINERATION TEST
Run 2 incineration of PCB fuel
November 13, 1979 from 1430 to 2115 hrs.
WttM' ¦£
>sc/	^ "'y *'	Li.
yj Kiwi z
II	I	i a	a	b M TS
1ktuAauu^uuJu s
lit	i ^
J_!.L:g ......
8:-
:i.n.! n.
: ,s
QcbV-"
, ir -/ {JoSo^/Asr
1!! ! iL.-J
61
-------
224
/
FIGURE 15 (cont'd)
CARBON MONOXIDE .AND OXYGEN CONCENTRATION
DURING nffi PCB INCINERATION TEST
Run 3 incineration of PCB fuel
November 14, 1979 from 1045 to 1730 hrs.
8 -Jr i
'"**? r.T-
J/*
US
%.'
I*
*1?. 3
fA
	°2
'W-. -	.*•.
\
CO - ^	«'
3
*•
1
1.
"//¦///prf
jj°	8	8	,
§	, . ; x b(%uUbw4
g				—
1
62
-------
3S
\ o
FIGURE 15 (cont'd)
CARBON MONOXIDE .AND OXYGEN CONCENTRATIONS
DURING THE PCB INCINERATION TEST
CO-

a
Pun 4 incineration of PCS fuel
	 November 15-16, 1979 from 1800 to 0100 hrs.
W7f


8 ^	
* At,
u
U
S
J-
S
3
s
<2?
¦w
at.
&.
3k .

\
v
6!
-------
23/
FIGURE IS (cont'd)
CARBON MCNOXIIE .WD OXYGEH G^fCENTRATICNS
CURING THE PCB INCINERATION TEST
Bum-out of excess PCS fuel
November 16. 1979 from 0108 to 0335 hrs.
64
-------

FIGURE 15 (cont'd)
CARBON MONOXIDE AND OXYGEN CONCENTRATIONS
CURING THE PQ INCINERATION TEST
Incineration of First Flush
November 16, 1979 from 0430 to 0730 hrs.
CO
v-*$~rr

?>-
>r.

-f~ tf-
3
•c"
3
65
-------
233
FIGURE 15 ("cont'd)
CARBON MQNOXIEE AND OXYGEN CONCENTRATIONS
CURING THE PCB INCINERATION TEST
Incineration of Second Flush
November 16, 1979 from 0950 to 1215 hrs.
66
-------

rIGJRB 15 f cont'd!
C.AE3CN NDNOXIDE ANT) OXYGEN CONCENTRATIONS
CURING THE ?C3 INCINERATION TEST
Incineration of Third Flush
a-	Movember 16-17. 1979 from 1650 to 1515 hrs.
»	¦	.*7 .
?Q	3	a	3
.«¦ -¦ *
CO



>
S ^^=-8
* «»»>•
"V
2
-5	.
S
I	_ "S.
.! >
3 —

¦«
a	i
Ho	S	a	8	8 ) §

r cont' dl
67
-------
US
FIGURE 15 [cont'd)
CARBON MONOXIDE AND OXYGEN CONCENTRATIONS
DURING THE PCS INCINERATION TEST
Incineration of Third Flush
November 16-17, 1979 from 1630 to 1315 hrs.
mi -1 rnr>
.Li.}J. UjTTWS
LUirrm
r."i1
.iiiHu i

:-:-rr8
^Tii i
! I i i i i ' •
ill I ! i I I
| H Him
I.J.I.
L'*h~rn\ i1
ijr: hM]!11!
Mill i U
!:! ! I 1 M _
! .LI.!...'..' '
-r x. 8
M M : 1 i j
_ 11
!~n I ; j II
.! U i\i

iu; ui_
j i
J 1.1.
¦~rr" 8

ilLi-Milld
58
-------
7U
FIGURE 16
WATER FLCWS TO SCRUBBER SYSTEM XRING
?C3 INCINERATION TEST
Run 1 incineration of background fuel
November 12, 1979 from 1400 to 2045 hrs.
2PM (a)
4m (a)
3PM (a)
?Tow Co ,
*ull Nozzle
Total
^ ^ ^ ^ ^ co us ^	-	r	.T!
—a--o—o—o—<«-o-o—o—o §	o	oT^c	o~cl
lilt' 1 ! i i •¦! -i -i	-i	- i
i	I	I W	: I
6PM (a) I a.
: r, i ; ¦ * , ; , J i JgL i J
o-s^-{s-s-®s-s-s--s--§-o-asv-s-§
i I	1; '	f , i m*. i ¦ 1

-l.

t.
i



.Alii
![! '



:t|!i
CTi
jLJi.
2»
!Hii;
s:
P;

id
_],& |
45* '
;£»-* I
-srsf -
>J ">¦ 1
^vu-»
¦Sri'
1
S.->:
W.: ;
Tal Corrected time scale added.
.3.

69
-------
r>7
- .
FIGURE 16 fcont' d")
WATER FLCWS TO SCRUBBER SYSTEM DURING
	?C3 INCINERATION TEST	
Run 2 incineration of PC3 fuel
November 13, 1979 from 1430 to 2115 hrs.
4PM (a)
6PM (a)
3PM (a)
ro i> 111
S
Total
i^FloW1 to
Bull.Nbzz

—oj— A-tni
a a o -
(a). Corrected time scale added.
70
-------
^52
FIGURE 16 [cont'd)
10AM (a)
12N (a)
2PM (a)
4PM fa)
WATER FLOWS TO SCRUBBER SYSTEM DURING
	PC3 INCINERATION TEST	
Run 3 incineration of PC3 fuel
November 14, 1979 from 1045 to 1730 Hts.
o oi
r !
HlOWj tri
Bull Nozzle
Ml''' I ( *71	I I ' i «	mt	M
— 1 1 M ili A. JS#	-J 03 .	^	-AJifta,.¦ fij - fj f_n
°,r? ? ? It® ? ° ? ? o
'otal Flow
	i " ! ' i i B, X I J,	— gujft J. j. U
, N U)	gft _0	Vj	50	O	*+	
'-o-c-o-o-S-o-o-o c 5 oTpfo cTo!
.. LiJ—		j -. j	- j |
|Hil
[fii
«-i- :
¦"j"*?
iy-i
(a).
1 1	, «
Corrected time scale added.
71
-------
Ill
FIGURE 16 (cont'd]
WATER FLOW TO SCRUBBER SYSTEM DURING
	PC3 INCINERATION TEST	
Run 4 incineration of PC3 fuel
November 15-16, 1979 from 1800 to 0100 hrs
I 01 I
! 00
	OJ
o o
W* fi'	Q— #-» — .•
O'
^ — Nj		jo o o o:
t 1 - 'fee
I
Total
S—flow
- — £ —c~^
0 3'.*b C O
?•?
-------
24°
FTGUSE 16 (cont'd]
WATER FLOWS TO SCRUBBER SYSTEM DURING
PQ INCINERATION TEST
Burn-out o£ excess PC3 fuel
November 16, 1979 from 0108 to 0335 hrs.
-------
¦w
FIGURE 16 (cont'd)
WATER FLOWS TO SCRUBBER SYSTEM DURING
PC3 INCINERATICN TEST
Incineration of First Flush
November 16, 1979 from 0450 to 0730 hrs.
Flow to
Bull Nozzle
-—-33.
O O	O si
Total
Flow
7 i
-------
2h2
FIGURE 16 (cont'd')
water flows to scrubber system during
PCS INCINERATION TEST
Incineration of Second Flush
November 16, 1979 from 0930 to 1215 hrs.
Flow to
Bull Nozzle
Total
Flow
75
-------

FTGUPS 16 fcsnt'dl
riATBR FLOWS TC SCRUBBER SYSTEM X'RING PC3 INCINERATION TEST
Incineration of Third Flush
fCont' d)
75
-------

FIGURE 16 (cont'd!
WATER FLOWS TO SCRUBBER SYSTEM DURING
PC3 INCINERATION TEST
Incineration of Third Flush
November 16-17, 1979 from 1915 to 1515 hrs.
(cont' d]
77
-------
ll1f
FIGUKE 16 (cont'd)
WATER PLOWS TO SCRUBBER SYSTEM DURING
?C3 INCINERATOR TEST
Incineration o£ Third Flush
November 16-17, 1979 from 1915 to 1515 hrs.
(concluded)
78
-------
ZHb
FIGURE 17
A ? ACROSS '."ENTURI THROAT CURING
THE ?C3 INCINERATION TESTS
November 12-16, 1979
— SUNDAY
M
£ a ' _
j
- Seeded
Test 2
es>.' o

iHowaiw-** ^
79
-------
aw 7
FIGURE 18
SCRUB WATER RETENTION TANK: pH DURING
THE PC3 INCINERATION TEST
Run 1 incineration of background fuel
November 12, 1979 from 1400 to 2045 hrs


hi umi
jiiwp"
i.	l—CH» |
TfSr^nnr'
SO
-------
m y
PluJPr 13 {cont'd)
SCRUB WATER SHTENTICN TANK: pH DURING
THE PC3 INCINERATION TEST
Run 2 incineration o£ PCS fuel
November 15, 1979 from 1450 to 2115 tvrs.
IMMSUfc
81
-------
2^
FIGURE _13_ (cont'd)
SCRUB WATER RETENTION TANK: pH HIRING
THE PC3 INCINERATION TEST
Run 3 incineration of PC3 fuel
November 14, 1979 from 1045 to 1750 hrs.


82
-------
7SV
.	' ** 3URT.VG
	^a^s@a22Lssr
4 »=inwttiaTnT	
Xovevber 15.1* .	* ° ^
h
2**Mh
-------
3 
FIGURE 13 (cont'd)
SaUB WATER RETENTION TANK: pH DURING
THE PCS INCINERATION TEST
Burn-out of excess PC3 fuel
November 16, 1979 from Q108 to 0555 hrs.
339
"TMllTTl
34
-------
2S-2
?T(7J3E 13 Ccsnt' d)
SCRUB WATER RETENTION TANK: pH HIRING
THE PCS INCINERATION TEST
Incineration of First Flush
Mnvgmber 16. 1979 from 0430 to 0750 hrs.
Ui "¦
if niiirrrr ""*!•*"
¦iiMMimBflgjai
3 S
-------
7S3
FIGURE 13 (cont'd]
———	j
SCRIJB WATER RETENTION TANK: pH CURING
THE PC3 INCINERATION TEST
Incineration o£ Second Flush
November 16, 1979 from 0950 to 1215 hrs.
36
-------
FIGURE 18 (cont'd)
SCRUB WATER RETENTION TANK: ?H CURING
THE PG INCINERATION TEST
Incineration of Third Flush
November 16-17, 1979 from 1650 to 1515 hrs.
imam muLT«j**' —<
J."'l1IHlilB|TITTTHr
WII*!-" ¦

(cant'd)
-------
iSb"
FIGURE 13 (cont'd)
SGO '.VATER RETENTION LANK: ?H DURING
THE PC3 IMCINERATICN TEST
Incineration of Third Flush
November 16-17, 1979 from 1650 to 1315 hrs.
	"5
IIWH 11 i — ¦! i !¦- — '' • T'JU.
"FTH
SaiSSg
!i!:


S3
-------

FIGURE 19
THE pH OUTFALL 001 DURING THE
PCB INCINERATION TEST
Test 1 incineration of background fuel
November 12, 1979 from 1400 to 2045 hrs.
itr
|8flM
oql'
Outfall
39
-------
25?
FIGURE 19 ("cont'd")
THE pH AT OUTFALL 001 DURING THE
PCB INCINERATION TEST _
Test 3 incineration of PCB fuel
November 14, 1979 from 1045 to 1730 hrs.
o
90
-------
25V
FIGURE 19 'cont'd)
THE ^ AT CUTrALL 001 HIRING THE
PC3 INCINERATION TEST
Test Z incineration of PCB fuel
November 13, 1979 from 1430 to 2115 hrs.
91
-------

FIGURE 19 (cont'd")
THE p*4 AT OUTFALL 001 CURING THE
?C3 INCINERATION TEST
Test 4 incineration of PC3 fuel
November 15-16, 1979 from 1800 to 0100 hrs.
92
-------
zoo
FIGURE 19 (cont'dp
THE pH CUTFALL 001 DURING THE
PC3 INCINERATION TEST
Burn-out of excess PC3 fuel
November 16, 1979 from 0108 to 0555 hrs.
93
i
-------
W\
FIGURE 19 (cont'd]
THE ?H AT CUTFALL 001 DURING THE
PG INCINERATION TEST
Incineration of First Flush
November 16, 1979 from 0430 to 0730 hrs.
-------

FIGURE 19 Tcont'd)
THE pH AT OUTFALL 001 DURING THE
PC3 INCINERATION TE5T
Incineration of Second Flush
November 16, 1979 from 0950 to 1215 hrs.
-------
I'bl
FIGURE 19 fcont'dl
THE ?H AT OUTFALL 001 DURING THE
PC3 INCINERATION TEST
Incineration of Third Flush
November 16-17, 1979 from 1650 to 1315 hrs.
Outfall C01
i
fcont' dl
96
-------
2bH
FIGURE 19 Cccnt'd)
THE pH OF OUTFALL 001 DURING THE
PQ INCINERATION TEST
Incineration of Third Flush
November 16-17, 1979 from 1630 to 1515 hrs.
(cent' d)
^itaii! ooi
97
-------
2<>Z
FIGURE 19 (cont'd)
THE pH OF OUTFALL 001 DURING THE
PC3 INCINERATION TEST
Incineration of Third Flush
November 16-17. 1979 from 1650 to 1515 hrs.
fconcluded)

98
-------
2U
APPENDIX I

-------
TRfAI. TEST OF LMHII) J'CB fNCINFPATION
BY: ROU.INS ENVIKOM«ENTAL SFWVICES (TX), INC.
hate "/ri/rfr)	rti&s: /
Rl«N »i / k,A	)
timi:
IKWS
/tfco
TIWl'ERATURKS (°F)
MONITORS
FANS (I) OI'IMM)
I kit Juct
Afterburner
Kiln
Stack
Ambient
cfrnror1
^2 (t) J" Cty (*)
i
Forced air
in
J/^3
J37£>£?

frJ.jT
/ 73
1
3-/3 1
i
6,7
3Z
l<7/3
J/7^
J3//
/^/Z


7 3e
3/3 j
&7
JZ
JZ

J/2/
JJpjL
/J23

6>ZS
/ Z3 \ 7. 73 J
6,7
1^3
J/33
J3/7
/3J3

6,2-3
0-3c>
W '}
67
3z
/ 3vo
/3/J
J/3/
JJ9/
7339

&S.O
<93o
3-37 j
£3
3Z
J/33
JJSB
/39o

L3.fi
/ oo
0.*1 j
t,S

/33&
J/S73
J/7o
JJ7>f
i3*/o

&>3.3
0. oo
&-S1 j
&3T
Jz
/6vo
j/39-
J233
/377

C3.3
O.po
8, ft !
(?3
3Z
/£>&
J/33
JM
W

l*3o
p73
8-S1 i
63
3Z
/&>3o
J/03
JJ&3
/37-Z

t>2.5
0.OO
f-ti j
<£iT
JZ
&
/le*73
Jt83
3Jt,B
l3S?o

&Z.o
0.OO
W !
6?3~~
/'/oo
2/73
J133
/32£

6>/.3~
0.O0
9.3i i
7,3
3+
/7/3
J/9?
J3S3
/3oj

C/.o
0. oo
i-et !
&>3
33
17Jo
J/3i

/37{

Uo
Pop
1
/O l^ j
63
V
rUJ>*
*
/727
//%?- lOl
J/8b
UAvitC
3/3
uu<
iS
MO
i- & Jet
o.o
j-u, b-ntj
1
/o z^ !
* j
tuJ-
t,3
31
-------
TRIAL IliST OF LIOUIU 1»CB INCINERATION
BY: ROLLINS BJVIROW'ENTAL SERVICES (TX), INC.
PATE	/)/&£ z
W*N ' 1 (wU/y/Uae^)
TIME
lift IPS
/3oo
TIMPERATURES (°F)
MONITORS
Gnwrr_^irrT~a^T
• i 	
FANS (t) 01
.NED
in
Hot duct
Afterburner
Kiln
Stack
Ambient
Forced air
JJ-'O
J35&
/5'0/

6/<0
0OO 1 /O. Z5
/J/0

&/¦ O
0.3o I ?.3Z j
6,5
33
/33o
2/9J
J3/>^
/36>fc>

6 b £>
1 1
/.oo } /o.37 {
45
53
/0*5
J'77
3J.2X
/3&/

6>/o
1 1
0*<£o I iO'tfj* i
i i
£.5
33
/r/oo
J/9o
JJJJ
/3&Q

£>/o
/.2.S i /e.37 j
65
! ^ 1^
1 I
<9/5
J/9*/
JJSo
S3&3-

6,/.0
t.OO J j

{&£>
J/97
J3U
/375~

La.3'
/.oo ! //,os !
1 1
bZ
31
33
/9*¦& j
{,5
Slooo
3/83
J J33
/3$l

&& »o
US i /!>¦& |
&5
3o
3o/5
J/84-
J3J3
/375

59. o
1 1
/>5o i /0.03 '
¦ t

30
Ja.%f
J/35
JJt/o
A563

58-5
/-7S | /o.o3 J

Jo
-5/

<2/83
JJ36,
/353

53. o
J.Jj j /OV |
65






i i
i i
i i
i i

	






i i
i i
I I







I 1
i i
I I
( 1


i
i
1
i
1
1


1 1
1 1
1 I
1 1
1 1
1 A
i

-------
/7.0
£.o
+
7-3Z
7o
13
JjJfT
JJ7-/-
"/2Ji
JU-3/
Jj
C/5
'/.o
97-0
7o
2#
JJOoL-
J30Cr
<&3
6>0.5
SS
7. 7J
7#
Jo
23

/SS/
J/73
J3S7
/*2J
6>O.0
So
3.30
59.5
6>.o
3-3/,
7o
7o
&
/33S
53.0
So
.5
9. Jo
7o
U /fk/L$Jt> .14S ^	kd-di^My -jkoJ
Jo
-------
THIA1. TEST OF LIQUID FCB INCINERATION
BY: ROLLINS ENVIROH'IiNTAL SERVICES (TX), INC.
HATE
fiif

HI'N »	2iuui^ )
Itot duct
•=Z/S^7
MJ
J/37
Ml
J jJ/
6
JJ05~
MU
»L37Z
J3&0
JJ&7
J&7
J*/3
762/
HtZJ
/&3
'VMLduL-f
Stack
Ambient
S7-0
S7 &
S7o
S1Q
S7o
SJ.Q
SC,£~
56>o
Sc.o
J>C o
SC-o
cn~nw
c 0
So
Jf£
So
£3
S3
La
£3_
£5
3.0
3.3
Ufy'AUh
MONITORS
T^ITTWTIT
+
6.67
#.£7 !
V0 I S.73
3*76
B- 73
M/
W
V/
8.73
837_
Ml.
9-Jo
f
W WW VHJHUfiU
FANS C%X C>PliNI-.D
Forced air [ II)
7 o
7o
7o
7o
7o
7o
70
7o
To
7o
7v
HL
7^
Jo
20
Zo
20
Li
_Io_
Jo
2o
ZD
2o
Zo
2o
2o
>!^(-)Ua££z\u}
-------
IHIAI. TEST (IP LinUIll IHJB INCINERATION
t03OH>
/o3<7 A
(0<+ 2. /
ilMli
IK"* IRS
TRWEHAIURES (°F)
? a r
MONITORS
FANS (I) OI'LMil)
llot duct
Afterburner
Kiln
Stack
Ambient
Gft (PPM)
cur
C02 err
Forced air
ID
/&
7020

&£>o
/lo
7.JZ

7o
'9
!/3o
jm
eW?
/6/J

6>/*o
// O
7.7J

7o
H

J/88
JV-5J
!i>08

(0/ S
t/o
131

lo
!
I
/Zoo
J/?7
J'fco
/&>/7

&>/.£
// S'
7- 73

7o
/2/s
jm
JZ3~

US
// 0
Ify

7o
/3

o2/S8
J^3S
/373

6/.0
/0.5
3./S

7o

/3oo
J/97
J'A/o
&?7

US
//O
8-tS

7o
>9
/3/S
J/99
JVJtf
039

&ZD
Joo
7.W

7o

/3jo
J/3Z
J^/6
08o

6>2.0
9 o


7c
/f
/3&
JJ/3

/£?/&

bZ£>
9o
8.31,

7£>
2o
Jj/co
J/&3
J36>d
/tc<7

&2'.£
/D 3


7o
J/JL
J/33
J3/j
/*?)

bZD
//O
q.si

7o
22-






1



-------
trial test of liouid pea incineration
BY: ROLLINS ENVIRfWENTAL SERVICES (TX) , INC.
PATE
RI'N * //fW+L, +

Hot-3nct
TEMPERATURES (°F)

l\fterburner
J323
Kiln
7^/3
"Stack
Ambient
6>/ S
GfTfTPM)
9.7J>
MONITORS
-nrWTWTIT
?.£>/
FANS (l> OPENED
Forced air T~ It)
~~ \T,
7#
Ji<77
J23JC
76/7

Jjjy
/6>lV~

u
9.0i
6>2 0
9.3
9<77
7v
7o
1 4?
! <2J.
J/S&
33 >73
tt>/3
J"7<7
<23V/
J6bZ
^2^
9.o
Ml
L,30
1.0
70.£>i>
7o
7t>
i
JiTo
J3JJ-
7^7

-13*7/

tJ.O
Bo
Jo.fe
6>$0
9o

7o
7e>
! ^LL
i J2JL

J33L

C*3o
9-o
1/.9/
7o
» JX
J"77
J333
/&<£>
(*3o
9.e
70.57
7o !
c2/<77
JM
/6>97
6>3o
9.0
/0.2l o
1.0
3-98
70
! &
«	
i
i
.1	
*
V	9
I	I
	,	1—
I -

-------
/7S(jlks " ^
THIAI. TEST 01- I.IOUII) I'^lt INCINEPATION
BY: KOI.I,INS ENVHtfYtWVAl SERVICES (TX), INC.
HATE

TIME
in ips
/Boo
/3(3
/BJcf
/S4S
tfao
o
o*
/9¥73
Ji4P3

Ji/33
c2.S
S£.S
56. o
S~hc>
SLo
SUo
56. o
COW
J.s
MONITORS

O
"OTTir
7 73
h
772.
6> o J 356
6.o
S.o
6. O
6' 0
s.o
t/.o
J.o
/.0
s.o
6-o
SS
S.o
$.5
656
818
7 31
3 SO
3'/3
07B
813
9*//
3 56
9.oo
94/
9.00
3 73
WTFJ
FANS (%> Ol'lNlil*
Forced"air 7 T'l)
63

63
6£
63
6S
6>3
63
63
63
63
63
6/S
6&
63
Jo
3c
3o
So
3o
So
3c>
So
30
3o
So
3D
3o
3o
So
-4
-------
*Us/J = 0//	)
7 4i/jU a-liO
TRIAL TEST OF LIQUID PCB INCINEPATION
BY: ROLLINS FNVIROM'ENTAL SERVICES (TX), INC.
f-f
y
Z
&33J'h. f ,
Itot duct
TIME
llttIRS
ofcJcO
O
C2J/S
Jk&o
JJJo
jjVs
¦V'
00
J^/S
343o
O/oo
Jib
JiJ5~
J3#j
J/3#
Mo/
Jyj.
jjj?
JUS
&/¦/&/
*>V7
9
I
7&7J
tJ&r
/3&o
/jfaf_
A^Z
JJ37
«/3j
/. £)
SS.S'
SS-O
3SO
ss,o
SLO
356
35.Q
SS.O
5C*o
££,o
S^.o
SSo
So I 70-03
A
+ 0 i /D%1
/DO+
So
£¦o
£>s
7o
h. O
£5-
SO
SO
£.5"
7.Sl
H#
&.1&
Jo. 00
/0-6,6,
0-37
£BjL
/o-l*/
/o <75
fO.lJ-
FANS (l> OPIMil)
Forced air ~T In
£>3
68
63
UZ
7/
7/
7/
71
72-
72
77-
72
11-
J3
J7
26
2.3~
2C
jjr
zr
2cT
2S
2d
zs
-------
37jr
APPENDIX II
i cS
-------
Ill*
United State!
Srwironmemal Protection
Agency
Region 5
' 20! Sim Street
Dallas TX 75270
Arkansas. Louisiana,
Okianoma. Tenas,
New Mexico
seEPA
RECEIVED HOV 3 Q 1S78
November 22, 1978
Mr. Jerry Neel, Vice President
Rollins Environmental Services, Inc.
P.0. Sox 609
Deer Park, Texas 77536
Dear Mr. Neel:
This letter is in response	to your general intentions letter of September 1
1978 in which you proposed	not to purchase and install a continuous carbon
dioxide (CO2) analyzer for use during PCB incineration.
My staff has reviewed the relative merits of your request based on the
incinerator1s design and prior combustion experience. This alternative
proposal is submitted for your consideration. Periodic monitoring of
COj and a higher combustion efficiency of 99.9* during PCB incineration
would be substituted for continuous CO2 monitoring. EPA Region 5 would
retain the right to require continuous C02 monitoring in the future if con-
ditions warranted this.
A formal reply from Rollins is necessary if you accept this alternative.
Sincerely,
Aaiene Harrison
Regional Administrator
cc: John Rollins, Jr.
Rollins Environmental Services, Inc.
W. S. Philipbar
Rollins Environmental Services, Inc.
Matt Straus
Office of Solid Waste (WH-562)
Jack Cannichael
Texas Department of Health
Jay- Snow
Texas Department of Water Resources

109
-------
%-n
C©PY
i iLdJ ¦ !X1 hlU.
R A
iottns
February 5, 1979
Ms. Adlene Harrison
Regional Administrator
Region VI
U.S. Environmental Protection Agency
1201 Elm Street
Dallas, Texas 75270
Dear Ms. Harrison:
This letter is in response to your letter of November 22, 197S,
in which you submitted an alternate proposal for continuous CO2
monitoring of our incinerator during PCB disposal.
Having reviewed your proposal, we acc.ept your alternative and
will perform periodic monitoring of CO2 and maintain a com-
"b'unsTian efficiency of 99.9% during PCB incineration.
Thank you for your consideration in eliminating the requirement
for continuous COj monitoring. Our PCB test proposal will be
mailed this week, and we look forward to your staff's reply so
that we may proceed with an incinerator test as soon as possible*
Sincerely,
Jerry D.	Neel
Vice President
JDN/rw
cc: Mr.	Matt Strauss, E.P.A.
Mr.	Bill Stewart, T.A.C.B.
Mr.	Jay Snow, T.D.W.R.
Mr.	W. B. Philipbar, Jr., R.E.S.
ROLLINS ENVIRONMENTAL SERVICES (TX) INC.
110
-------
Z7S
tto
l^yr^

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION VI
5608 Hornwood Drive
Houston, Texas 77074
Ms. Dianne Cass	0 -iuV ''373
Chemist
NUS Corporation
900 Gemini Avenue
Houston, Texas 77053
Dear Ms. Cass:
I have received your report of analysis of quality assurance samples
for PC3 content which I supplied to you in October 1979. The follow-
ing are* the results reported by you and the expected true values of
the samples supplied to you.
Sample No. 1 - Reported Composition:
Aroclor 1254; 2 ug/1iter (ppb)
Expected Actual Composition:
Aroclor 1254; 4.8ug/1iter (ppb)
Sample No. 2 - Reported Composition:
Aroclor 1016; 2 ug/1iter (ppb)
Expected True Composition:
Aroclor 1016; 3.2 ug/1iter (ppb)
Your PCB specie identifications were correct in each instance, and I
have determined that you quantitative values would be within the range
of acceptability based on similar type samples analyzed annually by
State and Federal laboratories for proficiency evaluation.
Sincerely,
William D. Langley, Ph.D.
Chief, Laboratory Section
cc:
Oscar Ramirez, Jr., 6ASA
Charles Gazda, 6ASASC
111
-------
274
=ENUS
3ca serviiisii AvsNiue
HOUSTON. TSXA3 77GS8
7ia-*aa-iaio
south csnthal G^e»ATtaivjs
CCFIPCRATiaN
gyrus wm. aica division
C-30-00-02/80-010
February 13, 1980
Mr. Doug Cooper
Rollins Environmental Services
P.O. Box 609 _
Deer Park, Texas 77536
Dear Mr. Cooper:
After reviewing the chemical literature and talking to several
people engaged in research in the determination of chlorinated di-
benzofurans in environmental samples, we have regretfully come to
the conclusion that no reliable widely accepted procedure for de-
termining trace amounts of "total" chlorinated diben2ofurans exists
at the present time. This conclusion was reached for the follow-
ing reasons:
1.	Of the numerous polychlorinated bizenzofuran derivatives
theoretically existing only three are commercially avail-
able.
2.	No procedure for definitely isolating trace amounts of
chlorinated dibenzofurans from polychlorinated biphenyls
was found.
3.	The aqueous samples analyzed typically contained less
than <0.001 mg/liter total PCB. If the original
polychlorinated biphenyl used in the test contained
chlorinated dibenzofurans in only trace quantities as is
reported in the chemical literature, there is little
likelihood that chlorinated dibenzofurans would be pre-
sent in the solutions in measureable amounts.
If additional information is required, please call.
Sincerely,
Jerry W. Bright
Analytical Laboratory Manager
/ss
112
3«NW*u 3*»>CS9 viano« OAK rwa • ia*5 cscw-an koao . -rrraau-OH. ,9aao . i-iai 3*3-9300
-------
2 %o
APPENDIX III
liS
-------
28/
.Appendix III
MASS FLOW RATE OF FLUE GAS
The maximum combined horsepower of the two induced draft fans,
operating in series, is 900. From the composite fan performance curve,
Figurel2, the rated capacity is 31,000 cu ft/min o£ gas at 170#F and a
density of 0.0472 lb/cu ft.
The fan capacity ¦ 81,000 cu ft/min x 0.0472 lbs/cu ft * 63.72 lbs/sec.
60 sec/min
The scrubber cools the flue gas by adiabatic huraidfication and the
exhaust gas at 170 *F is saturated with 0.4327 lb water for each pound of
dry gas. Therefore, the fans move
63.72 lbs/sec x 1 ¦ 44.48 lbs/sec dry gas
I7732T
Since very small quantities of water are produced by combustion of
highly chlorinated hydrocarbons, the quantity of gas moving thru the hot
zone is 44.5 lbs/sec.
The density of the gas calculated by the formula
520 °R . * lbs/cu ft
(15.1)ix 'ft) —
where x is the temperature of the gas, is 0.01S1 lbs/cu ft at a *^mperature
of 1223 °C.
114
-------
2*2
AVERAGE RESIDENCE- TIME
Volume of Loddby chamber
Volume of afterburner
Volume of hot duct
491 cu	ft
6235 cu	ft
1700 cu	ft
8476 cu	ft
Test No 1 (background) at 2763 °R:
320 °R - 0.0144 Ib/cu ft
U^.IJ (2765 "RJ
(0.0144 Ib/cu ft)(3476 cu ft)- 2.74 sees.
(.44.48 lb/secj
Test No 2 at 2796 °R:
320 °R - 0.0142 lb/cu ft
(13.1) (2796 °R)
(0.0142 lb/cu ft) (8476 cu ft) » 2.71 sees.
(44,48 lb/sec)
Test No 3 at 2853 °R:
(0.0139 lb/cu ft)(8476 cu ft) » 2.65 sees.
(44.48 lb/sec)
Test No 4 at 2870 #R:
520°R - 0.0138 lb/cu ft
(13.1) (.2^70 -Rj
(0.0138 lb/cu ft)(8476 cu ft) « 2.63 sees.
(44.48 lb/sec)
S20°R - 0.0139 lb/cu ft
113
-------
2 S3
APPENDIX IV
lib
-------
Z W
SOUTHERN PETROLEUM LABORATORIES, INC.
Certificate of Analvsis Mo. "3I96-"3199
" O. 3CX :0907
HOUSTON "SXAS 7T035
'713! 568.44«a	39
* 0. 30X 12788
LArAY = T-H. .OUISIANA res>
:319) 984-2374
Company:
Sample of:
For:
Attn:
Tes t 1,2,3,4
Rollins Environmental Service
P 0 Box 609
Deer Park, Texas 77336
Douglas Cooper
Invoice Mo. S9621
December 31 , 1975
Rollin 4
7 8196
test 3
73197
test Z
73 i 93
test 1
78199
tCSt 4
Carbon
Hydrogen
S
Sulfur
Chloride
Nitrogen
50 .94';
4.73$
0.0351
43 . S8's
4.66%
0 .036'
6 3.6 5's
8 . 861
0 .072'
50 .44'.
4.80*
0.034%
- - - - too much CI interference 	
54.3	43.3	24.2	56.3
0.015wt* 0 . 0 2 7 w t» 0.072wt« 0.016wt'
Southern Petroleum Laboratories, Inc.
117
-------
zs.r
APPENDIX V
US
-------
ItL
W 11 ICS	ANALYTICAL SERVICES LABORATORY
SOUTH CcNT^AL OPERATIONS
¦—S^ST-r-ir'rvi	900 GEMINI AV6NU6 • -OUSTON. TgxAS 770S8
~QfqpCjRAi it—iNi	713^88-1310
"¦-ollins Environmental Services, Inc.
Industrial Waste Disposal
P.O. 3ox 609
Deer Park, Texas 77536
Attn: Doug Cooper
Client No. .
Otte Sampled .
Date Received .
Date Reported «
11/9/7?
t 1 / q / 7 Q
II / 14
m.
Composite sample of PC3 material in T-27
I N US
I Sample
!	No.
!29110127
Total ?C3* 470,000 mg/kg
Specific Gravity 1.340
jp»ci«i instructions This sample was received in a plastic bottle with a waxed paper liner ~"j
*Alchough individual isomers were not identified, it appeared that a predominance of j
PCS 125* and ?C3 1260 existed.
Tail tMulO rapenad in mq/lii«r unlaw oin«rnn« netad
119
-------
ISIUS
co=*=graticn
IS 7
ANALYTICAL SERVICES LABORATORY
SOUTH CENTRAL OPERATIONS
300 GEMINI AVENUE . HOUSTON, TEXAS 77053
713-188-1810
Rollins Environmental Services,
Industrial Waste Disposal
P.O. 3ox 609
Deer Park, Tx 77536
Attn: Doug Cooper
Inc.
Client So. ___
Data Samplad _
Data Raeaivad.
Data Raportad.
11/12, 13,
15/79
11/13.14,15.16/79
11/20/79
N US
Samp la
No.
29110165
Test
1
Fuel
Composite
I


Total
PCB
<1 mg/kg
29110166
Test
1
Fuel
Composite
2


Total
?CB
<1 mg/kg
29110167
Test
1
Fuel
Composite
3


Total
PC3
<1 mg/kg
29110183
• Test
2
Fuel
Composite
1


Total
PCB
330,000
mg/kg
29110184
Test
2
Fuel
Composite
2


Total
PCB
380,000
mg/kg
29110185
Test
2
Fuel
Composite
3


Total
PCB
380,000
mg/kg
29110229
Test
3
Fuel
Composite
1


Total
PCB
360,000
mg/kg
29110230
Test
3
Fuel
Composite
2


Total
PC3
380,000
mg/kg
29110231
Test
3
Fuel
Composite
3


Total
PCB
360,000
mg/kg
29110306
Test
4
Fuel
Composite
18:00
:o
19:45
Totai
PCB
325,000
mg/kg
29110305
Test
4
Fuel
Composite
20:00
to
12:45
Total
PCB
340,000
mg/kg
29110307
Test
4
Fuel
Composite
22:00
to
23:45
Total
PCB
310,000
mg/kg
29110308
Test
4
Fuel
Composite
24:00
to
01:08
Total
PCB
310,000
mg/kg
Spacial [natrucuons
Tad raiulli reportin mg/tilar unlau slharwita noiad
120

\
-------

CCPP3CRAT1CN
Rollins Environmental Services, Inc.
Industrial Waste Disposal
P.O. 3ox 609
Deer Park, Tx 77536
Attn: Doug Cooper
ANALYTICAL SERVICES laboratory
SOUTH CENTRAL OPERATIONS
900 GEMINI AV5NU6 • -iOUSTSN. "SXAS 770S8
713-188-5 310
Client No. .
Data Samplad IT /1 "> / 7 c
Dala Racatvad 11/ 13 /79
Oaia Raporttod 11/20/79
N US
Sample
No.
29110171
29110172
29110173
29110174
29110175
29110176
29110162
29110163
29110164
29110168
29110169
29110170
Test No. 1
Well Sample Mo.	1	(3T)
Well Sample Mo.	2	(ET)
Well Sample No.	3	(BT)
Lime Slurry Composite If 1
Lime Slurry Composite It2
Lime Slurry Composite >'!2
Scrub water at TUT //I
Scrub water at TUT 2
Scrub water at TUT #3
F-4 Recirculated Composite //I
F-4 Recirculated Composite f/2
F-4 Recirculated Composite #3
Total
PCS
<0.001
mg/1
Total
PCS
<0.001
mg/1
Total
PC3
<0.001
tng/1
Total
PC3
<0.001
mg/1
Total
PCS
<0.001
mg/1
Total
PC3
<0.001
ag/1
Total
PC3
<0.001
ag/1
Total
PC3
<0.001
mg/1
Total
PC3
<0.001
ag/1
Total
PCS
<0.001
mg/1
Total
PCB
<0.001
mg/1
Total
PCB
<0.001
mg/1
Spacial Inatrucuons
Tatt raiulti raponad m mg/liiar unlait aiharwiia notad	_
ln	•
-------
IMUS
CORPCRATICN
2W
ANALYTICAL SERVICES LABORATORY
SOUTH CENTRAL OPERATIONS
300 GEMINI AVSNUE • -HOUSTON. TEXAS T7C53
713-488-1310
Rollins Environmental Services,
Industrial Waste Disposal
P.O. Box 609
Deer Park, Tx 77536
Attn: Doug Cooper
inc.
C'.iant No. .
Oat* S«mpi«d .
11/13/79
Dmt« R*c*iv*d 11/14/79
Oat* R*port*d 11/20/79
NUS
Sampl*
No.
29110186
29110187
29110188
29110180
29110181
29110182
29110189
29110190
29110191
29110192
29110193
29110194
Test No. 2
Well No. 1
Total
?C3
<0.001 mg/1
Well No. 2
Total
PC3
<0.001 mg/1
Well No. 3
Total
PCS
<0.001^g/l
LS-305 Sample f/1-4
Total
PCB
0.001 mg/1
LS-305 Sample *5-8
Total
PCB
<0.001 mg/1
LS-305 Sample *9-12
Total
PCB
0.0Q4 mg/1
TUT No. 1
Total
PC3
0.001 mg/1
TUT No. 2
Total
PCB
<0.001 mg/1
TUT No. 3
Total
PC3
0.C01 mg/1
F-4 No. 1
Total
PCB
0.002 mg/1
F-4 No. 2
Total
PCB
<0.001 mg/1
F-4 No. 3
Total
PCB
<0.001 ag/1
Special Instructions
T«« r*iuil> r*oon*d in mg/lit*f uni*t» olhwwiM noi*d
122

-------
140
-v-y^CP^CN
ANALYTICAL SERVICES LA BOA A "OR
SCUTH CSNTSAL OPERATIONS
300 GEMINI AVENUE • iQOSTCN. TgXAS 770S3
;i3-»aa-i8io
Rollins Environmental Services, Inc.
Industrial Waste Disposal
P.O. 3ox 609
Deer ?3rk, Tx 77536
Attn: Doug Cooper
Climi No	
D«t» S»«nP1,d -
D«t« R«c»ivtd.
Sat* Repor\«d .
0
1 i /i 1/79
i'/l^
juiri r^.6/7^
IJ2U22.
NUS
Sampi*
Mo.
29110221
29110222
29110223
I 29U0291
29110227
29110228
29110224
I 29110225
I 29110226
29110232
29110233
29110234
Test 3
Well Sample No.	1
Well Sample No.	2
Well Sample No.	3
TUT Sample 1
TUT Sample 2
TUT Sample 3
r4 Simple 1
F4 Sample 2
F4 Sample 3
305 L.S. Composite 1-4
305 L.S. Composite 5-8
305 L.S. Composite 9-14
Total PCB <0.001 ag/1
Total ?C3 0.001 ag/1
Total PCB <0.001 ag/1
Total PCB <0.001 ag/1
Total PCB 0.002 ag/1
Total PCB 0.001 ag/1
Total PCS <0.001 ag/1
Total PCB 0.001 ag/1
Total ?C3 <0.001 ag/1
Total PCB 0.001 ag/1
Total PCB 0.001 ag/1
(Sample accidentia destroyed)
Special Instruction*

-------
2MUS
CORPCRATTCfsj-
ANALYTICAL SERVICES LA80RAT0RV
SOUTH CENTRAL OPERATIONS
300 GEMINI AVENUE • HOUSTON. TEX-AS 77053
71 3-488-1810
Rollins Environmental Services, Inc.
Industrial waste Disposal
P.O. Box 609
Deer Park., Tx 77536
Attn: Doug Cooper
Client No. .
Data Sampled .

i 1
mm
Oat* Received.
~at* Reported 11/20/79
NUS






Samp la






No.
Test 4





29110292
Well Sample 1

Total
PC3
<0.001
mg/1
29110293
Well Sample 2

Total
PC3
0.002
mg/1
29110294
Well Sample 3

Total
PC3
0.001
mg/1
29110295
Well Sample 4

Total
PCB
0.001
ng/1
29110296
F4 Sample 1

Total
PCB
<0.001
mg/1
29110297
F4 Sample 2

Total
PCB
0.001
ng/1
29110298
F4 Sample 3

Total
PCB
0.001
mg/1
29110299
F4 Sample 4

Total
PCB
<0.001
mg/1
29110300
TUT 1

Total
PCB
0.001
rag/1
29110301
' TUT 2

Total
PCB
<0.001
mg/1
29110302
TUT 3

Total
PCB
0.001
ng/1
29110303
TUT 4

Total
PCB
0.002
Bg/1
29110304
TUT 5 (duplicate)

Total
PCB
0.002
mg/1
29110309
Lime (Composite of
1-4)
Total
PCB
<0.001
rag/1
29110310
Lime (Composite of
5-8)
Total
PC3
0.002
mg/1
29110311
Lime (Composite of
9-12)
Total
PCB
0.002
mg/1
29110312
Lime (Composite of
13-15)
Total
PCB
0.001
rag/1
Special Inatructione
Te»t rendu reported m mg/lilar unleu otherwise noted	Cj
124
-------
SMLJS
COPPCPATCN
2
-------
3*5
Division 3
-------
2 ^
EMISSIONS TESTING DURING INCINERATION OF PCBs
AT ROLLINS ENVIRONMENTAL SERVICES, INC., DEER PARK, TEXAS
SEPTEMBER 1981
by
T. L. Sarro, D. R. Moore, and D. G. Ackerman
TRW Inc., Environmental Engineering Division
One Space Park
Redondo Beach, Ca. 90278
EPA Contract No. 68-02-3174, Work Assignment 56
EPA Program Element Mo. C1YL15
Program Manager: Alice C. Gagnon
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Task Officer: Philip C. Schwindt
U.S. Environmental Protection Agency
Region 6 Office
Dallas, TX. 75270
Prepared for:
U. S. Environmental Protection Agency
Office of Pesticides and Toxic Substances and
Region 6 Office
Washington, D. C. 20460
-------
2-45
DISCLAIMER
This report has been reviewed by the Region 6 Office, Office of Toxic
Substances, and the Industrial Environmental Research Laboratory, U. S.
Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and
policies of the U. S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendation
for use.
ii
-------
2%
ABSTRACT
This report describes the emissions testing performed and results
obtained from Rollins Environmental Services, Inc., Deer Park, Texas, while
polycnlorinated biphenyls (PC3s) were being incinerated.
Samples were taken of all incineration system influent and effluent
streams, including stack gas, PCBs fed, auxiliary fuel and waste, scrubber
input, scrubber output, and treated plant effluent. Stack gas samples were
taken with a modified Method 5 Train incorporating a porous polymer adsor-
bent and operating isokinetically. Samples of other streams were taken
manually at specified intervals. Incineration system operating data also
were acquired.
This report was submitted in fulfillment of Contract No. 68-02-3174,
Mork Assignment 56, by TRW Inc., Environmental Engineering Division,
under sponsorship of the U. S. Environmental Protection Agency. This report
covers the period July 15, 1980 to September 15, 1980 (when the testing was
performed) and January 26, 1981 to February 26, 1981, and work was completed
as of February 26, 1981.
111
-------
2 V
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the contributions of Mr. R. W. Korner,
Mr. M. Drehsen, and Mrs. B. L. Riley to this report.
The authors also gratefully acknowledge the contribution of the members
of the field sampling team to this project. Team members were:
Mr. D. R. Moore, Team Leader, Mr. T. L. Sarro, Field Engineer, Mr. W. F. Wright,
Mr. R. W. Korner, Mr. C. W. Stackhouse, Mr. E. P. Banks, Mr. J. 0. Berger, Mr.
J. McReynolds, and Mr. M. Apoian.
1 v
-------

TABLE OF CONTENTS
Page
ABSTRACT	i i i
1.	INTRODUCTION 		1
2.	SUMMARY		3
2.1	Sampling		3
2.2	Process Monitoring 		9
3.	FACILITY DESCRIPTION 		13
3.1	Incinerator 		13
3.2	Scrubber and Discharge 			15
4.	SAMPLING AND MONITORING		17
4.1	Stack Gas Sampling 			17
4.2	Location of Sampling Points 		21
4.3	Process Monitoring 		22
4.4	Quality Assurance 		32
5.	TECHNICAL PROBLEMS AND RECOMMENDATIONS 		40
REFERENCES		41
APPENDIX A - GASEOUS MONITORING 		42
APPENDIX B - FIELD DATA SHEETS			46
APPENDIX C - PROCESS DATA				92
APPENDIX D - CALIBRATION DATA		109
APPENDIX E - CHAIN OF CUSTODY		121
v
-------
2c14
LIST OF TABLES
Table	Page
1	DAILY TEST SUMMARY AT ROLLINS ENVIRONMENTAL SERVICES 		4
2	DAILY DATA SUMMARY 		6
3	GASEOUS SAMPLING SCHEME 		7
4	LIQUID SAMPLING SCHEME 		8
5	DAILY AVERAGES AND STANDARD DEVIATIONS OF COMBUSTION PARAMETERS.	10
6	DAILY MINIMUM AND MAXIMUM VALUES OF COMBUSTION PARAMETERS ...	11
7	SUMMARY OF TEMPERATURE DATA		24
8	SUMMARY OF COMBUSTION GAS CONCENTRATIONS AND COMBUSTION
EFFICIENCIES 		27
9	FUEL CHARACTERIZATION DATA (BASED ON ROLLINS LABORATORY DATA). .	30
10 SUMMARY OF SCRUBBER AND DISCHARGE DATA		33
A1 GAS MONITORING DATA		44
vi
-------
3#o
LIST OF FIGURES
Figure	Pag
1	Schematic of Rollins' incineration system		14
2	Schematic of stack sampling locations		18
3	Schematic of PCS sampling train 		19
4	Schematic of adsorbent traps 		19
5	Schematic of train cleaning operations 		34
6	Calibration equipment set-up procedures 		36
vii
-------
10/
1. INTRODUCTION
TRW Inc., Environmental Engineering Division, under contract to the
U.S. Environmental Protection Agency (Contract No. 68-02-3174, Work
Assignment No. 31), performed emissions testing during incineration of
polychlorbiphenyls (PCBs) at Rollins Environmental Services, Inc., Deer
Park, Texas. The purpose of the testing was to determine whether or not
polychlorinated dibenzo-p-dioxins (PCDDs) and/or polychlorinated dibenzofurans
were emitted from the incineration system while PCBs were being incinerated.
Technical direction of the sampling effort was provided by the U.S.
Environmental Protection Agency's (EPA) Region 6 Office in Dallas, Texas.
Mr. Philip C. Schwindt of Region 6 was the Technical Officer for this work
assignment. Technical content of this effort was coordinated with the
following organizations:
•	EPA's Office of Pesticides and Toxic Substances, Washington, D.C.,
Mr. Bill Gunter, Mr. Glenn Kuntz, and Dr. Don Barnes.
•	EPA's Special Pesticide Review Division, Arlington, VA, Mr. Mike
Dellarco.
•	EPA's Health Effects Research Laboratory, Research Triangle Park,
NC, Dr. Bob Harless, Dr. Bob Lewis, and Mr. Merrill Jackson
•	Wright State University's Brehm Laboratory, Dayton, OH, the
analysis contractor, Dr. Tom Tiernan.
•	Rollins Environmental Services, Deer Park, TX, Messrs. Dave
Stang, Don Matter, Tracy Austin, and Jerry Neel.
At TRW, this effort was managed by Dr. Don Ackerman. Mr. Oave Moore
was the field team leader, and Mr. Tom Sarro was the field engineer.
1
-------
lei-
SamDies were taken of all incineration system influent and effluent
streams, including stack gas, ?C3 feeds (liquids only), auxiliary fuel and
waste feeds, and scrubber inlet and outlet solutions. Testing was performed
during the period 3-23 August 1580.
The following chapters and appendices of this report describe the test-
ing performed and results obtained: Chapter 2, Summary; Chapter 3, Facility
Descriptions; Chapter 4, Sampling and Monitoring, Chapter 5, Technical Prob-
lems and Recommendations; Aopendix A, Gas Monitoring; Appendix 5, Field Data;
Appendix C, Process Data; Appendix D, Calibration Data; and Aopendix E, Chain
of Custody Forms.
2
-------
I**
2. SUMMARY
Z.l SAMPLING
The field test activity took place at Rollins Environmental Services dur-
ing the time frame of August 4, 1980 through August 22, 1980. The third test
was postponed in August because of Hurricane Allen, and this third test was
conducted from August 20 through August 22, 1980. All required tests were
completed, and all recovered samples were sent to Wright State University
for analysis.
Duplicate stack gas samples were taken during incineration of three
different fuel types. The test conditions were as follows:
Test 1 - Incineration of liquid waste at normal incineration
burn fuel rates and temperatures.
Test 2 - Incineration of PC3/liquid waste blend at normal fuel
feed rates but at elevated temperature of 1200i 100°C,
O2 level greater than 32, and a CO content of less than
80 ppm.
Test 3 - Incineration of PCB blended with fuel oil and the
incinerator operating as in test number 2.
A daily field test summary is presented in Table 1. A summary of the
reduced data on a daily basis as calculated from field data sheets is pre-
sented in Table 2. Data listed in Table 2 have been corrected to standard
conditions of 20°C and a barometric pressure of 760 mm Hg (1 atmosphere).
Tables 3 and 4 are a summation of the gaseous and liquid samples collected
and their respective sampling frequencies.
Appendix A presents details of combustion gas monitoring. Field data
sheets are given in Appendix B. Appendix C contains all process data, while
Appendix D is the calibration data. Appendix E contains the Chain of Custody
reports.
3
-------
Date
(1980)
Test
No.
TABLE 1. DAILY TEST SUMMARY AT ROLLINS ENVIRONMENTAL SERVICES
Sampling Locations
Test Comments
8/4
8/5
0
1
Outlet
Outlet - Train 1
8/6
8/7
Outlet - Train 2
Inlet - CO2 Sampling
Outlet - Gas Sampling
Outlet - Train 1
Outlet - Train 2
Inlet - CO2 Sampling
Outlet - Gas Sampling
Outlet - Train 1
Preliminary measurements
Test started at 1320 hours. Resin trap plugged and
impinger filled with condensate, train stopped at 1405
hours. Changed nozzle - restarted with new train at
1650 hours - Resin plugged again. At 2110 hours train
stopped. Test scrubbed.
Test started at 1315 hours, stopped at 1407 to change
impinger, restarted.at 1640 hours, and stopped 2104
hours due to continual resin trap plugging. Test
scrubbed.
4 samples taken. Results are good.
1 composite sample. Results are good.
Test started at	1505 hours. Removed cooling water from
Florisil traps. Made resin trap changes due to plugging
Test quality is	good.
Test started at	1503 hours and ended at 2255 hours.
Test quality is	good.
8 samples taken. Test quality is good.
1st sample had leaks in sample line. Discarded result
from calculations, 2 additional integrated samples taken
Results are good.
Test started at 1420 hours. At 1800 hours train plugged
at resin trap. Replaced XAD-2 trap and continued test.
At 2107 hours, removed Florisil trap and replaced with
empty one. Still plugging occurred. Punched holes in
frits to assist flow. No help, frits were reddish
brown, resin appeared clean. Test stopped at 202
minutes due to plugging. Test quality good.
(continued)
-------
TABLE
Date Test
(1980) No.	Sampling Locations
Outlet - Train l
Inlet - CO^ Sampling
Outlet - Gas Sampling
8/22 4	Outlet - Train 1
Outlet - Train 2
Inlet - CO^ Sampling
Outlet - Gas Sampling
JL	(Continued)
Test Comments
Test started at 1420 hours. Train plugged at 1432
hours. Replaced both traps. At 1736 hours removed
Florisil trap. Replaced with empty trap, punched holes
in frits. No help. Test stopped at 2248 hours due to
plugging problem. Test quality is good.
13 samples taken. Test quality is good.
3 integrated samples taken. Test quality is good.
Test started at 0750 hours - XAD-2 resin plugged at
1100 hours, replaced trap and finished test.
Test quality is good.
Test started at 0745 hours. No apparent resin plugging
problem. Test quality is good.
Sampling probe burned off near port opening - possible
sampling of air. Used control room CO^ readings.
Test quality poor.
2 integrated samples taken. Test quality is good.
-------
TABLE 2. DAILY DATA SUMMARY*
DATE
Test Number
8-6-60 1 Train 1
Train 2
DSCF
99.472
96.919
Nro
2.8
2.7
14.5
14.5
co2 CO
t ppm
N
4.5
4.4
Sti<* Gas Flow
2 Temp Molecular Moisture Velocity
%	 _°F	 Height _ 	1	 ft/sec ACFH	 DSCFM
B1.0
ai.o
162.1 25.99
162.6 25.89
29.30
30.21
35.58 60353.096 36242.759
16.53 62052.662 36753.550
Isokinetic
R*te,E
103.8
98.9
B-7-80
2 Train 1
Train 2
77.232 2.2 13.9 4.3
78.251 2.2 13.9 4.3
81.b 165.2 25.41
fll.8 165.7 25.48
34.12 36.71 62279.183 34672.041 99.3
33.54 35.98 61038.315 34253.612 98.0
8-22-80 3 Train 1
Train 2
92.882 2.6 11.6 7.3
94.713 2.7 11.6 7.3
81.2 158.1 26.21
81.2 156.8 26.23
29.37 36.75 62342.982 36679.587 99.9
29.24 35.61 60406.649 35680.625 108.4
*
Where English units are used, they were provided by ENSCO.
*0ry standard cubic feed sampled, 20°C, 760 sin Hg.
'Norma1 cubic Deters sampled. 20°C, 760 nn Hg.
-------
TABLE 3. GASEOUS SAMPLING SCHEME
Frequency
4 hr
Continuous
Continuous
120 minutes
30 minutes
30 minutes
30 minute
30 minute
30 minute
30 minute
Sample Description
Sawple for
St)|i||jled by Ail-llyzed by
Stack gas
Stack gas
Stack gas
Stack Gas
Hot Duct
Fuel
Scrubber In
Scrubber Out
F-4 Outlet
Well Water
PCBs
PCDDs*
PCDFs**
°2
CO
C02
°2
C09
°2
PCBs
PCDDS, PCDFS
PCBs
PCDDs, PCDFs
PCBs
PCDDs, PCDFs
PCBs
PCDDs, PCDFs
PCB's
PCDDs, PCDFs
TRW
Rol1 ins
Rol1 ins
TRW
TRW
TRW
TRW
TRW
TRW
Wright State
Univers ity
Rol1 ins
Roll ins
TRU (GC)
TRW (GC)
Rollins/TRW EPA Houston
Wright State
University
Wright State
University
Wright State
University
Wright State
Uni versity
PCDDs = polychlorodibenzo-p-dioxins.
PCDFs = polychlorodibenzofurans.
-------
TABLE 4. LIQUID SAMPLING SCHEME
Total Samples
Sampling 4-hour	Sample	Composite Storage
Stream	Frequency Test Period)	Quantity*	Size/Type+
•
Liquid PCB Feed
30
minutes
8
250
ml
per sample
Half-gallon amber glass
•
Well Water Feed
30
minutes
8
250
ml
per sample
Half-gallon amber glass
•
Lime Slurry Feed
30
minutes
8
250 ml
per sample
Half-gallon amber glass
•
Touch Up Tank
30
minutes
8
250
ml
per sample
Half-gallon amber glass
•
Effleunt Discharge
30
minutes
8
250
ml
per sample
Half-gallon amber glass
CO
o
Duplicate samples to be collected from all streams.
+500 ml aliquot of composited PCB feed and touch-up tank to be removed from
half-gallon composites and provided to R.E.S.
-------
lcci
2.2 PROCESS MONITORING
Table 5 provides a summary of critical operating parameters of the
incinerator for each of the three test days including means and standard
deviations whenever appropriate. These parameters include hot duct tem-
peratures, residence times, combustion efficiencies, waste feed rates, and
concentrations of CO^, CO, and 0^ in the flue gas. All of these are based
on 15-minute observations except the waste feed rate which is based on volume
measurements of waste at the beginning and end of each test. All CO2 values
were determined by gas chromatographic (GC) analysis of samples taken at the
scrubber inlet. All CO values were read from a continuous monitor, The O2
values were taken from a continuous monitor for Test 3, and by GC analysis
for test two and three due to technical problems with the 0^ analyzer.
The CO values are all known to be too high, because the on-line CO
monitor could be seen to drift upward during every test. Frequent recali-
bration kept the instrument within tolerable limits, but the means are still
artificially high. This is important, because the combustion efficiencies in
Table 5 were calculated using these values. The average CO level for the
third test day was 6.1 ppm. Since this value is known to contain some
instrument drift, and since every recalibration of the CO monitor revealed a
CO concentration below 4 ppm, an average CO concentration below 5 ppm for any
given test day seems reasonable. This corresponds to a combustion efficiency
of 99.9929%, which is well above the 99.9% requirement set by EPA.
The waste feed rate given in Table 5 represents the volume flow of PCB
waste. The PCB content of this flow is unknown pending laboratory analysis.
These flows were calculated based on volume changes at the beginning and end
of each test, and hence no standard deviations were calculated.
Table 6 presents the minimum and maximum values of selected parameters
for each test. Also, the minima and maxima of these parameters during PCB
mode operation are given. It is worth noting that while the incinerator was
9
-------
TABLE 5. DAILY AVERAGES AND STANDARD DEVIATIONS OF COMBUSTION PARAMETERS
Test 1 (8-6-80)	Test 2 (8-7-80)	Test 3 (8-22-80)
Liquid PCB &	Liquid PCB &
Background Fuel	Background Fuel	Fuel Oil
~* + +
Parameter	x s	x s
a*
CM
©
8.4
0.16
8.2 0.41
7.0 +
CO (ppm)
9.3#
2.50
20.1# 11.95
6.1# 4.93
o2 (%)
ll.O(GC) 0.32
10.5(GC) 0.43
9.6(con- 2.53
tinuous)
Hot duct temperature
2028°F
12.17
2145°F 29.40
2152°F 26.70

1109°C
-
1174°C
1178°C
Residence time (seconds)
2.41
0.0384
2.28 0.1018
2.27 0.0443
Waste feed rate
0
-
7.48GPM -
9.51GPM -

0
-
0.47
1i ters/sfec
0.60
1iters/sec
Combustion efficiency
99.9883
0.0023
99.9753 0.0137
99.9893 0.0071
x and s are, respectively, mean and standard deviation.
+ no standard deviation exists because there was only one observation of CO^ for Test 3
§ these means are too high due to drift of the CO monitor.
-------
TABLE 6. DAILY MINIMUM AND MAXIMUM VALUES OF COMBUSTION PARAMETERS
Test 1 (8-6-80)
Background Fuel
Parameter	High Low
Test 2 (8-7-80)	Test 3 (8-22-80)	PCB Mode
Liquid PCB &	Liquid PCB	Operation
Background Fuel	Fuel Oil	Test 1 & Test 2
High Low	High Low	High Low
C02 (%)
8.7
8.1
9.2
7.6
7.0
7.0
9.2
7.0
CO (ppni)
11.5
4.0
38.0
2.0
16.0
1.0
38.0
1 .0
o2 (%)
11.5
10.5
10.9
9.8
12.1
7.9
12.1
7.9
Hot duct temperature (°F)
2050
2000
2264
2101
2203
2106
2264
2101
Residence time (seconds)
2.47
2.32
2.42
2.10
2.34
2.18
2.42
2.10
Combustion Efficiency
99.9954 99
.9863
99.99 74
99.9529
99.9986 99
.9771
99.9986 99
.9529

-------

in PC3 mode, the hot duct temperature never -"ell below 21013F, rhe .maximum
CO concentration (as seen by the CO monitor) was 38 oDm, and the snortest
residence time was 2.10 seconds. The lowest instantaneous combustion effi-
ciency was calculated to be 99.9529 percent and occurred on the second test
day at the same time the CO peaked at 38 ppm.
12
-------
VI
3. FACILITY DESCRIPTION
The facility described in this section is a hazardous waste incinerator
operated by Rollins Environmental Services in Deer Park, Texas. This
system represents the state of the art for thermal destruction of hazardous
wastes. High destruction efficiency and wet scrubbing of the flue gas
minimize environmental impacts from plant operation. This description is
based on on-site observations, numerous conversations with Rollins
personnel, and a previously prepared test report (1). Figure 1 shows a
schematic of the incineration process.
3.1 INCINERATOR
The incinerator unit at Rollins Environmental Services consists of a
rotary kiln for destroying solids, a Loddby burner for firing liquid or
gaseous waste, and an afterburner which provides a long residence time for
the combustion gases. Although no solids were burned during these tests,
the kiln was always fired with background fuel in order to maintain proper
temperature distribution throughout the incinerator. No hazardous wastes
were fired into the kiln during these tests.
The wastes to be destroyed are fired directly into the Loddby liquid
burner via a waste feed line. The waste fuel is held in a mixing tank, where
several wastes may be blended in order to achieve good thermal properties for
incineration.
Combustion gases from both the Loddby and the kiln enter the afterburner
where a long residence time insures maximum destruction of the waste. The
afterburner has a rather unusual geometry. Rather than having a simple
rectangular shape, it is bent at what is roughly a 45° angle. This provides
some turbulence which promotes more thorough mixing of the combustion gases.
Similarly, an irregularly shaped hot duct carries gases out of the after-
burner. A temperature sensor is located here, as well as probes for
13
-------
ON I.TNE
CAS
MONITORS
SOLI It WASTE
FEEl) CHUTE.
(NOT USED)/
ORSAT
SAMPLING
LOCAT1ON
KOTAKY
K I I.N
VENT
K I LN^-KX I T
SvW CT
nor
DUCT
•	mis r
ABSOKl'T 1 ON El. I Ml tl ATOli
TRAYSV	/
EX 1 T. CAS
1'RESII
WATER
FEEl)
I' U R T S
LODDBY
(WASTE I.KJUID BURN EH )
/	I-1 Oil 10
BACKGROUND BURNER
FUEL BURNERS
VENTUK1
INDUCED
DRAFT FANS
HYDRAT ED I.1ME
SLURRY FEED /
i OIJCII-ll|J
"i ANK U)
() V E R K I. O W
(TO TUCKER liAY(lll)
Figure 1. Schematic of Rollins' incineration system.
-------

monitoring CO and 0?. These are the measurements used at Rollins to
satisfy EPA monitoring requirements for incineration of PC3s. The tempera-
ture at this point must never fall below 1100°C (2012°F) while PCBs are being
ourned or an automatic shutoff system will stop the PC3 feed. Similarly, the
CO level here must never exceed 80 ppm.
Temperatures in the hot duct are typically maintained above 2100°F,
oxygen levels at roughly 10", and CO levels less than lOppm while the
incinerator is in PCS mode. Temperatures in the kiln and afterburner are
usually above 1600°F and 2400°F, respectively. Internal volumes of the
thermal destruction units are as follows:
Loddby volume	=	491 ft3	=	13.3 m3
Kiln	=	1385 ft3	=	38.8 m3 (does not include kiln
Afterburner	=	6285 ft3	=	176 m3	exlt duCt^
Hot duct	=	1700 ft3	=	47.6 m3
The volumes of the Loddby, afterburner, and hot duct combined provide a total
retention time of roughly 2.5 seconds.
3.2 SCRUBBER AND DISCHARGE
A wet scrubber system is employed to clean the combustion gases coming
out of the hot duct. This system consists of a Venturi scrubber in series
with a spray tower scrubber. Hot combustion gas enters the converging sec-
tion of the Venturi and is impacted by an atomized flow of scrubber liquor
just above the throat. This flow varies between 450-500 gallons per minute.
Most of the scrubbing action occurs in the throat section. The liquid
particles have already begun to coalesce as the gas stream passes through
the diverging section of the Venturi. Particulate is removed by impaction
with the liquid,and acidic components (particularly HC1) are removed by
reaction with the alkaline content of the scrubber liquor.
The combustion gases then pass up through the scrubber tower. Although
some scrubbing action does occur here, most of the removal efficiency of the
system is due to the Venturi. The purpose of the tower is to separate the
liquid phase introduced in the Venturi from the flue gases. A spray of fresh
water is introduced at the top of the tower to promote agglomeration of
15
-------
Vb
droplets, and gravitational forces cause the liquid to collect on the
bottom. In addition, a demister pad is mounted at the too of the tower to
prevent the escape of whatever fine droplets may be remaining. Scrubber
liquor collecting on the bottom of the tower is divided into two flows,
one being drained off for discharge and the other being recycled to the
Venturi. Additional hydrated lime slurry is added to the recycle flow in
order to maintain the alkalinity of the scrubber liquor, and fresh make-up
water is also added to this stream to compensate for the volume discharged
and the evaporative losses in the scrubber system.
The combustion gases are cooled substantially by the scrubbing process.
The temperature at the scrubber outlet stays very close to 145®F { 630C),
which is a drop of roughly 2000°F from typical hot duct temperatures. After
scrubbing, the flue gases pass through a pair of induced draft fans in series
and finally escape through a 30 meter stack. Together, the induced draft
fans provide a maximum of 900 Hp of mechanical work on the flue gas. Their
*5	3
rated capacity is 81,000 ft /min (38.1 m /sec) of gas at 170°F and a density
of 0.0472 lb/ft (0.7561 kg/m ). This provides a total pressure differential
of 28 inches water (5.23 cm Hg) over the length of the incinerator system.
Discharged scrubber liquor flows into an open reservoir called the
touch-up tank. This is simply an open tank with an agitator to which chemi-
cals can be added in case additional chemical treatment of the waste flow is
required. The liquid flows from this tank into a series of four settling
lagoons, separated by channels and weirs. The total retention time of this
system is from 2 to 4 days, depending on flow conditions. Final discharge is
to a local body of water known as Tuckers Bayou.
16
-------
Si/
4. SAMPLING AND MONITORING
4.1 STACK GAS SAMPLING
4.1.1	Samplinq Locations
All sampling locations are identified in Figure 1. Figure 2 is a
schematic of the stack including a cross sectional view depicting the
traverse point locations.
' The inorganic gas samples for molecular weight determination were
taken at the stack sampling locations. In addition, an inlet "hot zone"
COj sample was drawn to determine combustion efficiency. This sample was
collected prior to the gas entering the scrubber and undergoing any
compositional change.
4.1.2	Sampling Train Description
. The PCB sampling train, shown schematically in Figure 3, was utilized
to collect the organic constituents of the flue gas. The PCB train is a
modification of a standard EPA Method 5 train. A solid sorbent trapping
system was incorporated downstream of the probe and between the third and
fourth impingers. By placing the sorbent traps in this position, the gas
stream was relatively dry, as moisture was condensed in the first three im-
pingers. Granular XAD-2 resin was used for the primary sorbent material.
As a backup to XAD-2, Florisil was used to adsorb any residual organics or
organochlorine compcrunds that may break through the XAD-2 trap. The two
sorbent traps were water-jacketed, and an exiting gas temperature of 16°C
(60°F) was maintained by circulating the impinger ice bath water through
the jackets (Figure 4). Because of the sensitivity of organics to ultra-
violet light, the sorbent traps were wrapped in aluminum foil, and the
impingers and interconnecting glassware were also shielded from light.
17
-------
It*
TRAVERSE POINT LOCATION
INCHES FROM
OUTSIDE OF NIPPLE
15 .58
POINT
NO
32 .15
FAN
Figure 2. Schematic of stack sampling locations.
18
-------

STACK
W AL
FL0RISI1
a AD-2 TRAP,
TRAPn
PROB
REVERSE-TYPE^,
PTTOT TUBE*
f increased --„c
fitttngsv
KERMOMETER
CHECK
v AL VE
PITOT
MANOMETER
I	IMPINGERS j
THERMOMETER
1
LINE
VACUUM
GAUGE
BY-P-ASS
VALVE
MAIN
VALVE
DRY
TEST
METER
TIGHT
PUMP
ORIFICE
MANOMETER
Figure 3 . Schematic of PCB sampling train.
FLOW DIRECTION
GLASS
WATER, JACKET
ADSORBENT
THERMOCOUPLE
2 8/12'S OCKET
GLASS 1 FRITTED	I
DISK GLASS WOOL
28/12 'SOCKET
PLDG
Figure 4 . Schematic of adsorbent traps.
19
-------
520
4.1.3	Operating Procedures
Sampling for organics was performed at the stack outlet location uti-
lizing two PCS trains. Both trains were operated simultaneously at a rate
approximating the flue gas velocity (isokinetic sampling in accordance with
EPA Method 5, 40 CFR 60). The gas stream was passed through the impinger
solutions (water) to condense any moisture present in the stack gas and to
collect particulates then through the sorbent system to adsorb the organic
constituents. Parameters such as temperatures, pressures, gas volumes were
monitored throughout the sampling period.
Prior to test initiation, the sampling trains were leak checked at 15
inches of mercury. If a leak rate greater than 0.02 cfm was observed, the
sampling train was systematically checked to find the leak. Mo train was
operated with an unacceptable leak rate. No grease was used on any of the
glass fittings. Post-test leak checks were performed at the highest vacuum
encountered during the sampling time. Leak checks were also performed prior
to each port change or any disassembly of the train.
The sampling time was of 240 minutes duration and at a sampling rate of
less than 0.75 cfm.
4.1.4	Sample Recovery
Upon completion of testing, each train was removed to a clean area
(mobile laboratory) for recovery. Each train was recovered separately to
prevent sample mix-up and cross examination. Each train component was re-
covered per the following:
•	Probe and nozzle were wiped to remove residual particulate.
•	After recovering dry particulate from the nozzle and probe, these
parts were rinsed with acetone and hexane. Rinsate was retained
in an amber glass bottle, labeled and sealed.
•	Florisil and XAD-2 resin traps were removed from the train, sealed
with glass caps, and then wrapped for shipment to the analysis
laboratory.
•	Impingers - each of the impingers were weighed, weights were
recorded, and contents were transferred to amber glass bottles.
The impingers were then rinsed with small amounts of acetone and
20
-------
Ml
hexane. Rinses were combined with the impinger catch, and
rinse volumes were recorded.
•	Silica gel impinger was weighed, recorded, and regenerated
for subsequent use.
•	After recovery, each sample was labeled (traps) or sealed
(bottles) with EPA Chain of Custody tags.
4.2 LIQUID/SLURRY SAMPLING LOCATIONS, EQUIPMENT, AND PROCEDURES
This section summarizes the locations in the process where liquid
samples were collected and the equipment and procedures used to collect
the samples at Rollins Environmental Services, Deer Park, Texas.
4.2.1 Location of Sampling Points
Figure 1 illustrates the location of the sampling points in the
incineration system. A description of each sampling location is presented
below.
•	Lime Slurry Feed: Samples of the scrubber lime slurry feed
were collected from a valve in the pipeline leading to the
scrubber from slurry mixing tank No. 305.
t Well Water: Make-up water for the scrubber and lime slurry
originates from three on-site water wells. During the period
of testing, August 5-22, 1980, waters from wells No. 1 and No. 2
were being used in a ratio of 2:1, respectively. Both wells were
sampled during the testing periods. Samples were collected from
a valve in the well water pipelines leading to the scrubber.
•	Touch-up-Tank: The touch-up-tank (TUT) is an open fiberglass
tank, fitted with a mechanical agitator and located adjacent to
the base of the stack. Samples were removed directly from the
tank.
•	F-4 Discharge: Samples of the F-4 discharge were collected from
the flow through the fiberglass weir separating the No. 4
settling lagoon and the receiving stream. This was located
-100 yards south of the stack and averaged 7 inches of discharge
water depth flowing across the weir during each test period.
•	Liquid PCSs - Chlorinated Organic Waste - Fuel Oil: Fuel oil and
liquid wastes or PCBs were premixed prior to the tests. Samples
were collected from a valve in the pipeline just prior to intro-
duction to the Loddby burner.
21
-------
121
4.2.2 Sampling Equipment and Procedures
This subsection describes the equipment and procedures used to collect
the various types of samples during the test periods.
•	Process Stream Sampling Equipment: All liquid process
stream (lime slurry feed, well water, touch-up-tank, F-4
discharge, liquid waste feeds) samples were collected in
duplicate clean, amber glass jars (250 ml) with Teflon
seals in the caps. The duplicate individual 250 ml samples
were transferred to two clean amber glass bottles (2000 ml)
to obtain duplicate composite samples for each stream. Each
sample container and compositing jug was labeled prior to
sample collection.
•	Process Stream Sampling Procedures: All liquid process
streams (liquid waste feeds, well water, lime slurry,
touch-up-tank, F-4 discharge) were sampled in the same manner.
Duplicate 250 ml samples of each liquid stream were collected
and imnediately transferred to 2000 ml compositing jugs. Each
liquid stream had its own sampling jar and compositing jugs
which were sealed with Teflon lined caps when not in use. Each
liquid stream was sampled every 30 minutes during each test. At
the conclusion of a test, the compositing jugs were sealed,
tagged, and readied for shipment to Wright State University.
At the request of R.E.S. and the EPA, sample collection jars,
valves, and lines were not flushed with liquid so as to avoid
generating large volumes of waste which would have required
extra handling as hazardous wastes. In addition, for well water
collection, the sample collection jar was marked at the proper
liquid volume levels, 166.7 ml and 83.3 ml, to provide propor-
tional sample collection volumes of the two well water sources,
well number 1 and well number 2, respectively.
4.3 PROCESS MONITORING
While the three tests were being performed, records were kept of operat-
ing parameters within the Incinerator plant. These included temperatures,
combustion gas concentrations, waste feed rates, and scrubber operations as
described in subsequent subsections. Calculations for residence time and
combustion efficiency are also presented.
The information presented here has been reduced from the raw data sheets
included in Appendix C. Means and standard deviations have been calculated
for each test day whenever possible. In most cases the data were recorded at
15 minute intervals. Any exceptions have been noted.
22
-------
323
4.3.1	Temperature
Three temperatures characterizing the incineration process were measured.
These include the hot duct, afterburner, and kiln temperatures, all of which
were indicated in degrees Fahrenheit by digital readouts in the plant's con-
trol room. In addition, paper copies were made by strip chart recorders.
There was no facility for measuring the Loddby temperature. All furnace tem-
peratures were measured by thermocouples in contact with the gas stream. No
temperature instrumentation problems were encountered during any of the tests.
Table 7 provides a summary of all relevant temperature data, including
mean, standard deviation, minimum value, and maximum value for each of the
three test days. Values are expressed in both degrees Fahrenheit and degrees
Celsius, although the actual calculations were carried out in degrees Fahren-
heit (i.e., only the standard deviation corresponding to °F is given). The
overall mean temperatures (all three test days) for the kiln, hot duct, and
afterburner were 1440°F (782°C), 2108°F (1153°C), and 2403°F (1317°C), respec-
tively. The lowest temperatures were found in the kiln, which is acceptable
since no PCBs were being fired there. The highest temperatures are found in
the combustion chamber, as expected. The maximum combustion chamber tempera-
ture ranged between 2000-2050°F [1093-1121°C) during the background test and
2101-2264°F (1149-1240°C) for the PCB tests. It is required by EPA £40 CFR
761.40) that the minimum afterburner temperature be 2012°F (.1100°C). The
hot duct temperature is considered representative of this value.
4.3.2	Residence Time
Residence times were calculated by dividing the stated volume of the com-
bustion region by the volumetric flow rate of the combustion gases corrected
for temperature (2). Correction for pressure was not made because the pres-
sure in the system was negligibly different from atmospheric. Mathematically,
Residence time = RT = V = V
5" m/o
Where V » volume of the combustion zone (ft ), Q * volumetric flow rate
3	3
(ft /sec), m = mass flow rate (lb/sec), ando » density (lb/ft ) of the com-
bustion gas, respectively.
23
-------
TABLE 7. SUMMARY OF TEMPERATURE DATA





Temperature (°F or °C, as
Indicated)






Background Fuel

PCB
& Background Fuel

PCB & Fuel Oil

temperature
Sensor
(Test 11,
8-6-80)

(Test #2, 8-7-80)

(TestJ3,
8-22-8°)

Location

*
X
s
MAX
MIN
X
s MAX
MIN
X
s
MAX
MIN
Kiln
1542°F
20.99
1577°F
1468°F
1176°F
267.54 1668°F
622 °F
1602°F
59.48
1736°F
1504


839 °C
-
858°C
798°C
636°C
909°C
32 8° C
872°C
-
947°C
818
Hot Duct
2028°F
12.17
2050°F
2000°F
2145°F
29.40 2264°F
2101°F
2152°F
26.70
2203°F
2106

1109°C
-
1121°C
1090°C
1174°C
1240°C
1199°C
1178°C
-
1206°C
1152
After-
2289°F
44.04
2500°F
2228°F
2455°F
130.79 2696°F
2253°F
2464°F
58.14
2640°F
2370
burner
1254°C

1371°C
1200°C
1346°C
1480°C
1234°C
1351°C
1
i

1449°C
1299
Cleans were calculated using values of °F:
Hence, standard deviations are not given for means of °C
-------
2>2jr
Gas density as a function of temperature can be calculated from the fol-
lowing relationship (2):
p ® Tref	
(13.1) Tcomb
Where p = density in lb/ft3, Tref = 60°F (520°R), the constant 13.1 is the
3
volume (ft ) occupied by 1 lb of air at 60°F and 1 atmosphere pressure, and
Tcomb is the combustion zone temperature.
The above two expressions can be combined into the following relationship:
RT = (V.ft3)(540, °R)
(13.1, ft3/lb)(m, lb/sec)(Tcomb, °R)
During these tests, PCB wastes were fired through the Loddby burner.
Thus, the total combustion volume is the sum of the volumes of the Loddby
3
burner (49/ft ), the afterburner (6285 ft ), and the hot duct up to the point
3	3
of temperature measurement (1700 ft ). The total volume is 8476 ft or
237.4 m3.
The mass flow rate of combustion gas can be determined from the induced
draft fan curves and a knowledge of how much water is added to this flow by
the scrubber. The fans are capable of drawing 81,000 ft/min at a gas den-
sity of 0.0472 lb/ft , which corresponds to a mass flow rate through the fans
of 63.72 lbs/sec. However, this flow contains water evaporated from the
scrubber which was not contained in the original combustion gas flow. The
scrubber approximates an adiabatic saturation process. Therefore, by know-
ing that the scrubber outlet temperature is close to 150°F (65.6°C) and
assuming atmospheric pressure through the fans (a conservative assumption],
the water content of the gas can be found by using thermodynamic principles
(3). The flow of dry. combustion gas under these conditions is 52.71 lbs/sec
(23.91 kg/sec).
At Rollins, the temperature in the hot duct was maintained above the
regulatory minimum of 1200°C + 100°C (at 3» excess oxygen). Thus, the hot
duct temperature is taken as Tcomb. Hot duct temperatures are given in
Tables 5 and 6.
25
-------
-m
Tables 5 and 6 summarize residence times which ranged from 2.10 seconds
to 2.42 seconds and which were above the regulatory minimum of 2 seconds (at
1200°C + 100°C and 3% excess oxygen).
4.3.3 Combustion Gases and Combustion Efficiency
Three parameters are considered important in order to characterize the
combustion gas composition. These are the measurement of concentrations of
carbon monoxide (CO), carbon dioxide (CO^), and oxygen (0^). Combustion
efficiency is related to levels of CO and C02 in the gas stream leaving the
afterburner. The presence of significant levels of 0^ in the gas stream
insures that enough oxygen was present in the combustion chamber for complete
oxidation of the waste. A tabulation of combustion gas data for the three
test days is given in Tafile 8.
Carbon monoxide was measured with a nondispersive, infrared type (EPA
Reference Method 10) continuous monitor (Beckman Model 865). A sample line
provided gas from the breach of the stack to the instrument located in the
instrument room. The CO monitor also provided a signal for a strip chart
recorder, which was located in the control room.
Considerable difficulty was experienced with the CO monitor during these
tests. Upward drift was very noticeable during the first and second test
days; and, although improved, was still present on the third day. In fact,
nearly an entire day of CO data was lost during the first test while repairs
were attempted. The monitor was recalibrated twice during each PCB burn, but
this was not enough to avoid artifically high daily mean CO concentrations.
These means were 20.1 ppm and 6.1 ppm for tests two and three, respectively.
The daily mean for the first test day has been calculated to be 9.3 ppm, but
this is for only 9 observations, as opposed to 34 observations for test two
and 32 observations for test three. Every time the monitor was recalibrated
its output was below 3 ppm, suggesting that even the 6.1 ppm value is too
high.
Regardless of the CO monitor drift problem, the CO monitor never recorded
a concentration higher than 38 ppm; the maximum allowed by EPA for this incin-
erator was 80 ppm. This concentration was calculated from the combustion
efficiency (99.9%) and C02 concentration (8%). Plant operation was permissible
-------
TABLE 8. SUMMARY OF COMBUSTION GAS CONCENTRATIONS AND COMBUSTION EFFICIENCIES
Paraaeter
Test One (8-6-60)
Background Fuel
s* n* Max
Mln
Test Two (8-7-80)
PCB t Background Fuel
x	S	n Max
Hin
9.8
7.0
CO (ppm) Con- 9.3 2.SO 9 11.5 4.0 20.1 11.95 34 38.0 2.0 6.1
tlnuous monitor)
CO. (X) (ORSAT 8.4 0.16 8 8.7 8.1 8.2 0.41 12 9.2 7.6
analysis)
0- (X) (ORSAT 11.0 0.32 8 11.5 10.5 10.5 0.43 11 10.9
analysis)
0, (X) (Con- 13.0 1-89 28 11.2 9.4 -	-	- -
tlnuous Monitor)
(Test Three (8-22-00)
PCB & Fuel Oil
5	n Max Mtn
4.93 31 16.0 1.0
1 7.0 7.0
9.6
Combustion
Efficiency (X)
2.53 32 12.1 7.9
99.9883 0.0023 9 99.9919 99.9862 99.9753 0.0137 34 99.9975 99.9529 99.989 3 0.0071 32 99.9957 99.9801
x, s, and n are, respectively, mean, standard deviation, and nunber of data points.
-------
32 V
while the CO monitor was down since no PCBs were burned on the first test day.
All carbon dioxide measurements were determined by gas chromatographic
(GC) analysis. The procedures involved and a tabulation of results are given
in Appendix A. Because of the time involved with collecting a sample for
analysis, 15 minute sampling intervals were not possible. The C02 data in
Table 3 were all taken from the scrubber inlet location, i.e., the sampling
was performed in the hot duct. This is the appropriate location for combus-
tion gas evaluation. It can be seen in Table 8 that the C02 values ranged
between 7.0 and 8.4%. The first two test days gave fairly similar results,
and the low standard deviations shown that a high level of replication (i.e.,
stable combustion) was achieved. The C02 value for the third test day is
somewhat lower, possibly because fuel oil was being fired instead of chlori-
nated hydrocarbons. It must be pointed out that only one good sample was
analyzed on this day, because leakage through an open port in the hot duct
caused subsequent values to be much lower. However, this is still the lowest
C02 value throughout the three tests. The next lowest was 7.6%, encountered
during test two. The maximum value, 9.18%, was also encountered during the
second test day.
Combustion efficiency, as a function of C02 and CO concentrations, can be
calculated as follows (2):
% co2
Combustion efficiency = CE = 100 X % C02 + % CO
where the levels of C02 and CO are both given as percent concentrations. The
results of this calculation for.the three test days are given in Table 8.
Since low CO levels result in high combustion efficiencies, it is not surpri-
sing to see the highest combustion efficiency occurring on the day on which
drift of the CO monitor was least evident. This was the third test day, and
a mean combustion efficiency of 99.9893% was calculated. The day on which CO
monitor drift was most evident gave the lowest mean combustion efficiency,
i.e., 99.9753% on the second test day. However, even the lowest instantaneous
combustion efficiency of 99.9529% (encountered on the second test day) is
still above the minimum acceptable value, set at 99.9% by the EPA. This value
in fact, corresponds to a CO concentration of 80 ppm, which is the upper limit
28
-------

for CO (assuming an 3% concentration of COg). Thus, the incinerator would
have been well within operating guidelines even if these had been true read-
ings for CO.
Oxygen was measured using a Thermox WDG continuous monitor. A sample
line provided gas from the hot duct, and signals were recorded on a strip
chart as for CO monitoring. Some difficulty was experienced with this equip-
ment also, because of plugging of the O2 probe with particulate matter.
Thorough cleaning of the probe provided good continuous monitoring data for
the third test day, but the GC analyses were considered more accurate for the
first two days. Daily means and standard deviations of oxygen data are given
in Table 8. Low standard deviations for the GC analyses again show a high
level of replication and, hence, stable combustion. The lowest instantaneous
value of Oj recorded was 7.9%. This is well above the 3% minimum required
by EPA regulations (40 CFR 761.40). In general, O2 levels remained fairly
constant throughout the tests. The daily means were 11.0% (GC), 10.5% (GC),
and 9.6% (continuous monitor) for the first, second, and third test days,
respectively.
4.3.4 Waste Feed Data
The only waste feed monitored was the flow of PCB waste through the
Loddby liquid burner. On the first test day, no PCBs were fired, and no feed
rate data were collected because there was no flow through the Loddby. Back-
ground fuel was being fired into the combustion chamber on this day, and a
characterization of this fuel was obtained from the chemistry lab at Rollins.
These data appear in Table 9 with characterizations of the PCB mixtures burned
on the two subsequent test days. The PCB content of these mixtures is unknown
pending an outside laboratory analysis.
The waste flow monitored during the last two test days was the only source
of PCBs into the incinerator. There were a number of ways to measure this
flow rate, including an instantaneous readout in the plant's control room, a
readout of waste volume in the storage tank, and a flowmeter attached to the
feed pipe itself. However, the instantaneous flow readout was not properly
calibrated on the second test day (the first PCB test), and data from the
mechanical flowmeter were not taken until the third test day. Therefore,
29
-------
TABLE 9. FUEL CHARACTERIZATION OATA (BASED ON ROLLINS LABORATORY DATA)
Fuel Property
Specific Gravity
AH, (BTU/lb)
Background Fuel
(Fired Into Combustion
Chamber During
Test One & Test Two)
1.10
9,800
PCB Mixture #1
(Fired Through Loddby
During Test Nunber Two)
1.20
14,400
PCB Mixture #2
(Fired Through Loddby
During Test Number Three)
1.10
12,070
o
CI (Weight *)
Ash (Weight 2)
Scrubbing Efficiency
(lbs lime/lbs fuel)
35%
1.71*
19.3%
0.57 %
0.090
28.2%
0.65%
0.185
if}
o
-------

the volume of waste burned was chosen as the most consistent (and probably
the most accurate) basis for calculating average waste flow rate.
Waste volume level in a given storage tank (T-27 in this case) was
measured by a sensor at the bottom of the tank. This sensor measured the
weight of the column of liquid resting on top of it. A digital readout in
the control room indicated this value but was calibrated to indicate the
total volume of waste in gallons, assuming a specific gravity of one (i.e.,
the readout indicates what the volume of the liquid would be if the tank were
full of water). Therefore, this readout must be divided by the specific
gravity of the waste to find the true volume of liquid in gallons. A digital
readout was also provided to measure density in the tank, but the laboratory
values (Table 9) would be more accurate. The average PCB mixture flow rate
can then be calculated by taking the difference of tank volumes at the begin-
ning and end of each test and dividing by the test duration time. The follow-
ing results were obtained, where Q is the waste feed rate:
^test two, volume method 7,48 9a^/min
^test three, volume method = gal/min
The other two methods were considered fairly accurate for test three and are
included here for comparative purposes. Note that the instantaneous flow
readout was calibrated to read twice the actual flow and that this was the
way the raw data were recorded. The mechanical flow meter data were recorded
directly in terms of actual flow.
^test three, instantaneous flowmeter ~ 9,2 9a^min
(n = 32, standard deviation s 0.04)
^test three, mechanical flowmeter * 9al/m"'n
(n = 26, standard deviation = 0.56)
31
-------
Vi2
4.3.5 Scrubber Operation and Liquid Discharge
Hourly readings of most important scrubber paramters were taken by
Rollins personnel. Means and standard deviations for these over the test
periods are given in Table 10, including scrubber temperature, pressure drop
across the Venturi, Venturi liquid flowrate, touch-up tank pH, and final dis-
charge pH. Parameters of interest that were unavailable include total fresh
water feed to the scrubber system, scrubber recycle flow, lime input, and
discharge flow through the touch-up tank.
The scrubber temperatures can be seen to be around 150°F (65.6°C). The
exiting flue gas is probably a little cooler than this. The pressure drop
across the Venturi throat relates directly to the scrubber efficiency achieved,
larger pressure drops will give higher efficiencies,, and all of the means were
close to 40 inches H20 (7.5 cm Hg) which is the normal operating range for
this scrubber. The Venturi liquid flowrate, which includes recycle and fresh-
water flows into the Venturi throats remained close to 500 GPM (0.032 m^/sec).
The flow through the touch-up tank was, on the whole, slightly acidic.
Large standard deviations indicate that large pH swings occurred, and these
oscillations could in fact be seen on a strip chart recorder. Apparently,
lime is added 'to the scrubber at intervals, rather than on a continuous basis.
The long retention time of the four ponds dampens out these variations, how-
ever, and the final discharge is always slightly alkaline.
4.4 QUALITY ASSURANCE
4.4.1 Test Preparation
All steps needed to clean the train, as well as all glass sample containers
are listed in Figure 5. All impingers used for organic collection were fired
to 450°F, nitric acid washed, distilled water rinsed, acetone rinsed, and hex-
ane rinsed. Prior to initial assembly of the train in the field, all parts
were rinsed with hexane. This solvent was discarded. All subsequent field
cleanings were with glass-distilled water, acetone, and hexane. Only Burdick
and Jackson "Distilled-in-Glass," or equivalent, organic solvents were used.
Visual inspection for contamination was also performed.
32
-------
TABLE 10. SUMMARY OF SCRUBBER AND DISCHARGE DATA 4
Test One (8-6-80)
Background Fuel
Parameter	x +	s +
Scrubber temperature
Venturi liquid flowrate
Venturi Ap
OJ
u
Touch-up tank pH
Outfall pH
* Rollins Environmental Services, Deer Park, Texas, considers the data in Table 10
to be confidential business information as provided for by the Toxic Substances
Control Act (TSCA). A copy of these data is on file with the Document Control
Officer at the U.S. Environmental Protection Agency's Region 6 Office. To obtain
these data, application must be made under the appropriate provisions of TSCA.
Test Two (8-7-80)	Test Three (8-22-80)
PCB & Background Fuel	PCB & Fuel Oil
x	s	x	s
* x and s are, respectively, the mean and standard deviation.
-------
V>H
DISTILLED IN
SLASS
ACETONE RINSE
CAP OR SEAL
0PEMN8&
YES
3&J, OR
EQUIVALENT
ACETONE RINSE
SOAP AND WATEP
CLEANING
SOAK IN 15'i HNO3
MINIMUM OF
3 HOURS
DISTILLED WATER
RINSE
AIR ORY
CLEANING
PROCEDURES
USED "ARTS
TRAIN ONLY
NEW PARTS, TRAIN
PARTS AND ALL
SAMPLE CONTAINERS
DISTILLED IN
SLASS
S&J. OR
EQUIVALENT
HEXANE RINSE
SOAK IN 155 HNO
MINIMUM OF
TAP WATER RINSE
BUROICKi JACKSON
OR EQUIVALENT,
HEXANE RINSE
REASSEttLE TRAIN
distilled water
RINSE
VISUAL INSPECTION
FOR
CONTAMINATION
Figure 5. Schematic of train cleaning operations.
34
-------

To avoid contaminating the trains as they were assembled or disassembled,
as much work as possible was done inside the designated lab area in the mobile
van. PC3 samples were never brought into the van.
SMOKING was prohibited in all but designated areas. Prohibited areas in-
cluded the stack and other sampling locations. Exposure of resin traps was
kept to a minimum. If any equipment was contaminated, it was completely re-
cleaned. All organic solvents were stored in amber glass containers. No
grease of any type was used on the train including the impingers. If a seal
could not be attained, that part was replaced.
The adsorbents used through the PCB test sequences were cleaned by the
following methodology.
•	10 gallons of distilled water was rinsed through the XAD-2
resin to remove residual salts
•	A 24 hour continuously reflux of "Distilled-In-Glass" methanol
•	For XAD-2, a 24 hour reflux with "Distilled-In-Glass" methylene
chloride
•	For Florisil, a 24 hour reflux of a blend of 5% diethyl ether
and 95% hexane to remove organics
•	Both adsorbents were then dried under constant flowing nitrogen
until no residual organic odor was detected
•	After drying, Florisil was heated for 2 hours at 650°C and
allowed to cool
t XAD-2 and Florisil traps were packed and sealed in the
laboratory
4.4.2 Method Five Calibration Data
4.4.2.1 Orifice Meter Calibration—
The orifice meter calibration is performed using a pum'p and metering
system as illustrated in Figure 6 (a). The dry gas meter with attached cri-
tical orifice is run at various orifice flows for a known time. After each
run the volume of the dry gas meter, meter inlet/outlet temperatures, time,
and orifice setting is recorded. The orifice meter calibration factor is
derived by solving the equation:
r (Tw + 460) 9-,2
L vw J
35
-------

	manometer
:rific
MANOMETER
SIR "NLi
HESMOME ER
LEVEL
JATER N
HERMOMETER
(J r RATE METER
POINTER
WATER
EVEL
GAUGE
1EEDLE VALVE
jRY
S METER
JATER OUT
LEVEL AOJUS
AIR OUTLET J
SURGE ANK
¦1RIFTCE
r :mpinger or
I} SATURATOR
Figure 6a. Orifice and dry gas meter calibration.
ELECTS IC MOTCR
STAMWRD PHOT TUBE
ADJUSTABLE DAMPER
VALVE
TEST HOLE
OAKPER
AlRj
AIR FLOW
STEEL TUBE
POINT WHERE TIP 0F PITOT
TUBE WOULD BE WHEN
TAKING A READING-	'
PITOT TUBE BEING CALIBRATED
TOP VIEW
Figure 6b. Pi tot tube calibration.
36
-------
where
iH = Average pressure drop across the orifice meter,
inches H?0
Pb = Barometric pressure, inches Mercury
= Temperature of the dry gas meter, °F
Tw = Temperature of the wet test meter, °F
9 = Time, minutes
Vw = Volume of wet test meter, cubic feet
The aH@ yielded is utilized to adjust the sampling train flow rate by regula-
ting the orifice flow.
4.4.2.2 Dry Gas Meter Calibration—
Meter box calibration consists of checking the dry gas meter for accuracy.
The dry gas meter with attached critical orifice is connected to a wet test
meter, Figure 6 (a), and run at various orifice flows for a known time. Af-
ter each run wet and dry gas meter volumes, temperatures, time, and orifice
readings are recorded. Utilizing the equation:
V ¦ Vw Pb (Td + 460)
Vd (Pb * fjl-y) (T„ ~ 460)
where
V 3 Volume correction factor (.also known as "Y" in EPA Method 5)
Vw = Volume of wet test meter, cubic feet
Pb a Barometric pressure, inches mercury
Td = Temperature of dry gas meter, °F
Vd = Volume of dry gas meter, cubic feet
dH = Average pressure drop across the orifice meter,
inches H^O
T » Temperature of wet test meter, °F
w
A volume factor which compares the dry gas meter with the wet test meter is
obtained.
37
-------
\HQ
5. TECHNICAL PROBLEMS AND RECOMMENDATIONS
The major problem that occurred during the tests at Rollins was plugging
of the sorbent traps. This problem caused delays and loss of one test day.
A brown to red-brown material was distributed throughout the train but
appeared to be concentrated particularly in the glass frit of the XAD-2
sorbent trap. This material could be either organic or inorganic in nature.
It should be noted, however, that the material passed through two water filled,
ice bath cooled impingers before entering the XAD-2 trap.
It was recommended that the material be analyzed, so that this problem
might be avoided in future test projects. These analyses are underway at
Wright State University.
There were minor problems: wrong length probes, occasional leaks in
the trains, and some samples broken in shipment. These problems would have
been lessened had the test preparation period been longer than two weeks and
had a wider selection of shipping companies been available. Also solid ship-
ping containers such as metal ice chests should be used instead of cardboard
shipping cartons.
There were minor facility problems, especially with the CO monitor as
noted in Chapter 4. Once again, a longer lead time for test preparation
might have reduced these problems.
40
-------
•;hi
REFERENCES
1.	Rollins Environmental Services. The PCB Incineration Test Made by
Rollins Environmental Services (TX) at Deer Park, Texas, November
12-16, 1979. Report prepared for U. S. EPA, Region VI.
2.	Beard, J. H, and J. Schaum. Sampling Methods & Analytical Procedures
Manual for PCB Disposal: Interim Report. Report prepared by U. S. EPA,
Office of Solid Waste, February 10, 1978.
3.	Van Wylen, G. J., and R. E. Sonntag. Fundamentals of Classical
Thermodynamics. John Wiley & Sons, New York, 1976.
41
-------
IHX
APPENDIX A
GASEOUS MONITORING
Monitoring of oxygen and carbon dioxide for stack gas molecular weight
determination was performed at the Rollins facility. The sampling positions
were located at ports midway on the stack exit after the incineration and
scrubber units. However, at the Rollins facility, the CO2 concentration
undergoes changes as the gas passes through the caustic scrubber system.
For this reason and to determine combustion efficiency, samples of stack
gas were extracted at the "hot" duct leading from the burner unit to the
scrubber.
The gas sample was drawn through a stainless steel probe, polypropylene
tubing, and an ice bath condenser by means of a small diaphragm pump.
A flow rate meter was used to measure sample flow to a Tedlar bag. The
collected gas sample was then analyzed for carbon dioxide, oxygen, and nitro-
gen with a thermal conductivity detector gas chromatograph. The collection
device consisted of the following components.
•	Probe - stainless steel with a glass wool plug to act as
a filter
•	Condenser - air cooled or ice bath to eliminate excess
moisture
•	Valve - needle valve to regulate sample gas flow rate
•	Pump - leak-free, diaphragm type, or equivalent to transport
sample gas to the Tedlar bag
•	Flow rate meter - meter capable of measuring a flow range
from 0 to 1.0 liter per minute
•	Gas sample bag - Tedlar sample bag or equivalent, with a
capacity of approximately 0.5 - 1.0 cubic foot
•	Manometer - water filled U-tube or dry gas meter to be
used for the Tedlar bag leak check
•	Vacuum gauge - gauge or meter to be used for the sample train leak check
42
-------
w>
Prior to field use, all Tedlar gas sample bags were leak checked. The
bags were leak checked by inflating them to a pressure of 5 to 10 cm H^O
(2-4 in Hg) as determined by an in-line manometer or equivalent. Any dis-
placement in the manometer after a 10-mi'nute time interval was taken as
indicative of a leak. In the field prior to the sampling operation, the
sampling train was also leak checked. The leak check was done by placing
a vacuum gauge at the probe inlet, pulling a vacuum of at least 250 urn Hg
(10 in. Hg), and then turning off the pump. The vacuum should remain stable
for at least one minute.
A sample was taken as follows:
•	Place the probe in the stack at the sampling ooint and
purge the sample line up to the bag.
•	Connect the bag and make certain that all connections are
tight.
•	Fill and evacuate bag twice before taking analysis sample-
•	Sample at a rate that will fill the bag in 10 minutes.
•	Remove full bag, cap opening, transport to lab, and analyze
as soon as possible.
In the lab area, the Tedlar bag was analyzed on the Shimadzu gas
chromatograph, model GC-3BT. The resultant peak heights were compared
to a known standard gas purchased from Scott Environmental Technology, Inc.
Table A-l presents the gas analysis data as collected from both the
"hot zone" and outlet sampling points at the Rollins Facility.
43
-------
TABLE Al. GAS MONITORING DATA
Results
Date	Time Sampled Sampling	CO2	02
(1980)	(hrs)	Location	%	%	Comments
8/5
1338-1358
1703-1720
1915-1935
2028-2048
Hot zone
7.33
8.35
8.80
8.57
10.45
9.49
9.10
9.49
1915-2130
Outlet
4.29
12.58
8/6
1505-1522
1608-1624
1656-1712
1816-1836
1848-1900
1902-1921
2124-2139
2228-2248
Hot Zone
II
8.68
8.35
8.35
8.12
8.35
8.46
8.35
8.35
10.48
10.59
11.03
11.03
10.92
11.47
11.03
1541-1735
1805-2040
2118-2303
Outlet
0.79
4.74
4.29
16.55
14.34
14.56
Leaked in 1ine
8/7
1330
1425
1455
1520
1555
1630
1703
1755
1830-
1920-
llot zone
-1440
1510
-1540
-1610
1645
1715
1820
1645
1935
11
•I
it
11
11
II
M
M
M
8.40
8.28
8.28
7.61
8.71
8.28
8.28
8.06
8.28
8.06
10.69
10.15
10.15
10.79
9.83
10.58
10.15
10.69
10.69
10.90
Preliminary stop time missed
(Continued)
-------
TABLE Al. (Continued)
Date
(1980)
Time Sampled
(hrs)
Sampling
Location
CO2
%
Results
O2
%
Comments
8/7
(cont'd)
2015-2030
2100-2115
2130-2145
Hot zone
7.72
9.18
8.28
11.12
9.94
10.69
1430-1650
1755-1820
1823-1824
2053-2115
Outlet
It
II
II
2.46
1.12
6.04
5.71
15.71
16.78
12.40
12.18
Leak in pump
Leak in sample line
New sample
8/22
tn
0947-1003
1156-1217
1252-1303
0835-1000
1205-1215
Hot zone
II
Outlet
II
7.03
3.05
2.81
7.50
7.03
12.16
14.81
15.23
11.53
11.63
Poor results - when test was
over, found that sample probe
burned off and sampled air
leakage around port
v/-»
-------
bli.
APPENDIX B
FIELD DATA SHEETS
46
-------
7^7
The material on pages 47-92 is not suitable
for reproduction.
These pages can be viewed at the Region 6
Office of the US Environmental Protection
Agency
1201 Elm Street
Dallas, Texas 75270
-------

APPENDIX C
PROCESS DATA
Appendix C contains all of the raw process data sheets for the three
test days at Rollins (8-6-30, 3-7-80, and 8-22-80). The sheets are
organized by test day. There are three different types of data sheets:
The first type contains temperature, combustion gas, and fan data. The
second type contajns PCB feed rate data. The third type 'contains hour-by-
hour readings of plant operations routinely taken by plant personnel.
Note that no PCBs were fed during the-first test day; and, hence, there are
no corresponding data sheets for this day.
The first kind of data sheet contains military time, hot duct tempera-
ture, afterburner temperature, kiln temperature, ambient temperature, CO
concentration, C02 concentration, 02 concentration, forced air fan opening,
and induced draft fan opening. All of these were taken every fifteen
minutes, in the units given on the sheet. The induced draft fans were always
100% open, and the forced draft fan opening did not affect total mass flow
through the incinerator. These, along with ambient temperature, were not
considered important enough to discuss in the report.
The second kind of data sheet contains flow rate data for the PCB waste.
Times are given as before. Column 1 gives the volume of waste in the holding
tank expressed as an equivalent volume of water. This number divided by the
density of the waste gives the true volume of the waste in the tank. Column
2 indicates the density of the waste, in the tank. Column 3 is a direct
reading of instantaneous flowrate; however, it is actually twice the real
flow. It must be pointed out that the numbers in columns 2 and 3 are not
accurate for test two (8-7-80) because neither of these instruments were
properly calibrated at the time. They can be considered accurate for the
third test day (8-22-80). An additional flow measurement was made from a
93
-------

mechanical flow meter on this day. These numbers, read directly in terms of
flow, are given in Column 4.
The third data sheet was filled out hourly by Rollins personnel. Para-
meters of interest are the Venturi flow rate (flow to flexitray, in GPM), the
pressure drop across the Venturi (Venturi AP, in inches HO), the scrubber
temperature (Venturi crossover temperature, in °F), the pH of the touch-up
tank, and the final discharge pH (pH F-4). Other parameters may also be of
interest to some readers.
94
-------

The material on pages 95-109 is not suitable
for reproduction.
These pages can be viewed at the Region 6
Office of the US Environmental Protection
Agency
1201 Elm Street
Dallas, Texas 75270
-------
bs*
APPENDIX D
CALIBRATION DATA
no
-------
wx
The material on pages 111-121 is not suitable
for reproduction.
These pages can be viewed at the Region 6
Office of the US Environmental Protection
Agency
1201 Elm Street
Dallas, Texas 75270
-------
^3
APPENDIX E
CHAIN OF CUSTODY
122
-------

The material on pages 123-12S is not suitable
for reproduction.
These pages can be viewed at the Region 6
Office of the US Environmental Protection
Agency
1201 Elm Street
Dallas, Texas 75270
-------

Division 4
-------
isb
EMISSIONS TESTING DURING INCINERATION OF PCBs
AT ENERGY SYSTEMS COMPANY, EL DORADO, ARKANSAS
SEPTEMBER 1981
by
T. L. Sarro, D. R. Moore, and D. G. Ackerman
TRW Inc., Environmental Engineering Division
Cne Space Park
Redondo Beach, Ca. 90278
EPA Contract No. 68-02-3174, Work Assignment 56
EPA Program Element No. C1YL1B
Program Manager: Alice C. Gagnon
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Task Officer: Philip C. Schwindt
U.S. Environmental Protection Agency
Region 6 Office
Dallas, TX. 75270
Prepared for:
U. S. Environmental Protection Agency
Office of Pesticides and Toxic Substances and
Region 6 Office
Washington, D. C. 20460
-------
>51
DISCLAIMER
This report has been reviewed by the Region 6 Office, Office of Toxic
Substances, and the Industrial Environmental Research Laboratory, U. S.
Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and
policies of the U. S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or reconmendation
for use.
11
-------
S5-*
ABSTRACT
This report describes the emissions testing performed at and results
obtained from Energy Systems Company, El Dorado, Arkansas, while poly-
chlorinated biphenyls (PCBs) were being incinerated.
Samples were taken of all incineration system influent and effluent
streams, including stack gas, PCBs fed, auxiliary fuel and waste, scrubber
input, scrubber output, and treated plant effluent. Stack gas samples were
taken with a Modified Method 5 Train incorporating a porous polymer adsor-
bent and operating isokinetically. Samples of other streams were taken
manually at specified intervals. Incineration systems operating data also
were acquired.
This report was submitted in fulfillment of Contract No. 68-02-3174,
Work Assignment 56, by TRW Inc., Environmental Engineering Division, under
sponsorship of the U. S. Environmental Protection Agency. This report
covers the period July 15, 1980 to September 15, 1980 (when the testing
was performed) and January 26, 1981 to February 26, 1981, and work was
completed as of February 26, 1981.
iii
-------
3^
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the contributions of Mr. M. Drehson
and Mrs. B.L. Riley to this report.
The authors also gratefully acknowledge the contributions of the members
of the fiejd team to this project. Team members were: Mr. D.R. Moore, Team
Leader, Mr. T.L. Sarro, Field Engineer, Mr. W.F. Wright, Mr. M. Drehson, Mr.
J. Mc Reynolds, Mr. M. Apoian, Mr. M. Griffin, and Mr. W. Murphy.
iv
-------
iyo
TABLE OF CONTENTS
Page
ABSTRACT		i ii
1.	INTRODUCTION	•		1
2.	SUMMARY		3
2.1	Sampling		3
2.2	Process Monitoring 		3
3.	FACILITY DESCRIPTION 		12
3.1	Waste Feed Facilities		12
3.2	Incinerator		14
3.3	Scrubber and Discharge System 		15
4.	SAMPLING AND MONITORING		18
4.1	Stack Gas Sampling		18
4.2	Liquid Stream Sampling 		22
4.3	Sampling Locations 		23
4.4	Process Monitoring 		24
4.5	.Quality Assurance 		32
5.	TECHNICAL PROBLEMS AND RECOMMENDATIONS 		40
6.	REFERENCES		41
APPENDIX A - GASEOUS MONITORING 		42
APPENDIX B - FIELD DATA SHEETS		46
APPENDIX C - PROCESS DATA		93
APPENDIX D - CALIBRATION DATA		147
APPENDIX E - CHAIN OF CUSTODY RECORDS 		I59
v
-------
>6/
LIST OF TABLES
Table	Page
1	DAILY FIELD TEST SUMMARY 		4
2	DAILY DATA SUMMARY 		5
3	GASEOUS SAMPLING SCHEME 		6
4	SOLID/LIQUID SAMPLING SCHEME (ENSCO) 		7
5	COMBUSTION PARAMETERS FOR THE THREE TEST DAYS	9
6	MAXIMUM AND MINIMUM VALUES OF COMBUSTION
PROCESS PARAMETERS 		10
7	SUMMARY OF TEMPERATURE DATA	25
8	DAILY AVERAGES OF MASS FLOW RATES THROUGH
THE INCINERATOR 	28
9	DAILY SUMMARY OF INCINERATOR DRAFTS AND PRESSURES . 			31
10 CAUSTIC SOLUTION INPUT DATA 		33
A-1 ENSCO GAS MONITORING DATA 	44
v1
-------
Ibl-
LIST OF FIGURES
Figure	Paae
1	Schematic of incineration system at ENSCO 		13
2	Schematic of ENSCO's stack and sampling points 		19
3	Schematic of PCB sampling train 		20
4	Schematic of adsorbent traps 		20
5	Cleaning procedures 		34
6a Orifice and dry gas meter calibration	36
6b Pi tot tube calibration	36
vt1
-------

1. INTRODUCTION
TRW Inc., Environmental Engineering Division, under contract to the U.S.
Environmental Protection Agency (Contract No. 68-02-3174, Work Assignment
No. 31), performed emissions testing during incineration of polychlorinated
biphenyls (PCBs) at Energy Systems Company (ENSCO), El Dorado, Arkansas. The
purpose of the testing was to determine whether or not polychlorinated dibenzo-
p-dioxins (PCDDs) and dibenzofurans (PCDFs) were emitted from the incineration
system while PCBs were being burned.
Technical direction of the sampling effort was provided by the U.S. En-
vironmental Protection Agency's (EPA) Region 6 Office in Dallas, Texas. Mr.
Philip C. Schwindt of Region 6 was the Technical Officer for the Work Assign-
ment. Technical details were coordinated with the following organizations:
•	EPA's Office of Pesticides and Toxic Substances, Washington, DC,
Mr. Sill Gunter, Mr. Glenn Kuntz, and Dr. Don Barnes
•	EPA's Special Pesticide Review Division, Arlington, VA, Mr. Mike
Dellarco
t EPA's Health Effects Research Laboratory, Research Triangle Park,
NC, Dr. Bob Harless, Dr. Bob Lewis, and Mr. Merrill Jackson
•	Wright State University's Brehm Laboratory, Dayton, OH, the analy-
sis contractor to EPA, Dr. Tom Tiernan
•	ENSCO, El Dorado, AK, Dr. George Combs
At TRW, this effort was managed by Dr. Don Ackerman. Mr. Dave Moore was
the field team leader, and Mr. Tom Sarro was the field engineer.
Samples were taken of all incineration system influent and effluent
streams, including stack gas, PC8 feeds (liquid and shredded capacitors),
auxiliary feed and waste feeds, scrubber inlet solutions, scrubber outlet
solutions, and solid residues from the incinerator. Testing was completed
during the period 25-29 August 1980.
1
-------

The following chapters and appendices of this report describe the
testing performed and results obtained: Chapter 2, Summary, Chapter 3,
Facility Description; Chapter 4, Sampling and Monitoring; Chapter 5,
Technical Problems and Recomnendations: Appendix A, Gaseous Monitoring;
Appendix B, Field Data; Appendix C, Process Data; Appendix D, Calibration
Data; and Appendix E, Chain of Custody Forms.
2
-------
1)4; £
2. SUMMARY
2.1	SAMPLING
The field test effort at Energy System Company (ENSCO) took place during
the period 25-29 August 1980. All required tests were completed and all re-
covered samples were sent to Wright State University for analysis.
Duplicate stack tests were taken during its incineration of the following
fuel types and incinerator conditions:
•	Test 1 - Blended liquid organic wastes, sludges, (ink and paint)
and non-PCB containing capacitors with the incinerator at normal
burn rate and operational temperatures.
•	Test 2 - Blended liquid organic wastes, PCB capacitors and liquid
in place of the sludge. Incinerator operation was at 1200+ 100°C,
Og > 3% and CO content of < 50 ppm.
•	Test 3 - PCB capacitors, liquid PCBs, and fuel oil in place of or-
ganic wastes. Incinerator operating as in Test 2.
Duplicate sets of composite samples were taken from each liquid, solid or
slurry stream.
Gaseous (COg, Og) monitoring was performed at the stack only. ENSCO per-
sonnel monitored CO and O2 at the "hot" or inlet zone.
A daily field test summary is presented in Table 1. A summary of the re-
duced data on a daily basis, as calculated from the field data sheets, is pre-
sented in Table 2. The data listed in Table 2 have been corrected to standard
conditions of 20°C and a barometric pressure of 1 atmosphere (760 mm Hg) as
necessary. Tables 3 and 4, respectively, present descriptions of the gaseous
and solid/liquid samples collected and their respective sampling frequencies.
2.2	PROCESS MONITORING
All parameters critical to the safe operation of the incinerator were moni-
tored once every 15 minutes during the three tests. Among these parameters
are those for which requirements have been established by EPA. These include
3
-------
TABLE 1. DAILY FIELD TEST SUMMARY
Date	Test
(1980)	No.	Sampling Locations
8/26	1	Outlet-Train 1
Outlet-Train 2
Gaseous Monitoring
8/27	2	Outlet-Train 1
-p.
Outlet-Train 2
Gaseous Monitoring
8/28	3	Outlet-Train 1
Outlet-Train 2
Gaseous Monitoring
Test Couments
Test started at 1235 hrs. and ran for 240 minutes.
Numerous leak problems, broken glassware and plugged
XAD-2 traps - all corrected and test quality is good.
Test started at 1240 hours and ran for 240 minutes.
Numerous leak problems, broken glassware and lugged
XAD-2 trap - All problems corrected and list quality
is good.
No problems. Test quality is good.
Test started at 1115 hours, last meter box pump at
1120 hours. Restarted at 1155 hours and ran for total
of 240 minutes. Test quality is good.	V/o
O-
Test started 1200 hours and ran for 240 minutes. Test
quality is good.
No problems. Test quality is good.
Test started at 0915 - FJorisil trap plugged - It was
replaced and test ran for total of 240 minutes. Test
quality is good.
Test started at 1010 hours and ran for 240 minutes.
Test quality is good.
No problems - Test quality is good.
-------
TABLE 2. DAILY DATA SUMMARY
BATE
TEST
NO.
SAMPLING
LOCATION
SAMPLE VOLUME
DSCF
GAS COMPOSITION
CO,
CO
_C£S_
STACK
TEMP
"f
MOLECULAR
UEIGHT
MOI STUNE
VLLOCITY
ft/set ALFM
GAS FlOU
USCIM
IbOKIM III
M It
8-26-80 1 Train 1	128.264 3.6 12.2 5.7
Train 2	105.481 3.0 12.2 5.7
164.3
163.3
25.41	35.00	10./j 45268.4 lb i?4ai /. yyi IU9.il
25.57*	33.53	9.86 41940.631 23462.344 lit.2
in
8-27-80 2
Train I
Train 2
99.545
94.782
2.8
2.7
13.3
13.3
6.2
6.2
165.5
156.4
25.46	35.24	9.23 39271.273 21319.614 99.6
25.50	34.90	9.15 3U931.4UG 215U5.364 95.9
<3-
8-28-80 3
Train 1
Train 2
92.629
91.646
2.6
2.6
13.2
13.2
5.5
5.5
164.6
164.1
25.41	35.05	9.00 38265.27/ 20U92.293 94.b
25.40	35.09	tt.bl 36IU4.475 19/56.40b 101. J
* Dry standard cubic feet of flue gas sampled (20°C, 1 atnosphere).
t Nunul cubic Meters of flue gas sampled (20°C, 1 atmosphere).
-------
TABLE 3. GASEOUS SAMPLING SCHEME
Frequency
4 hours
Continuous
Continuous
120 minutes
Sample
Description
Sample
For
Sampled
By
Stack gases
Flue gas
Flue gas
Flue gas
PCB
dibenzodioxin
dibenzofurans
°2
CO
C0„
TRW
ENSCO
ENSCO
TRW
Analyzed
By
Wright State University
ENSCO
ENSCO
TRW

-------
TABLE 4. SOLID/LIQUID SAMPLING SCHEME (ENSCO)
Stream
Sampling
Frequency
Total Samples
4-hour
Test Period)
Sample
Quanti ty*
LIQUIDS




• Liquid PCB Feed
15
minutes
16
250 ml per sample
• Liquid Waste Feed
30
minutes
8
250 ml per sample
• Well Water
30
minutes
8
250 ml per sample
• Scrubber Liquor
30
minutes
8
250 ml per sample
• Recycled Scrubber
Liquor
30
minutes
8
250 ml per sample
• Caustic Solution
30
minutes
8
250 ml per sample
• Lime Slurry
30
minutes
8
250 ml per sample
SOLIDS




• Shredded Capacitors
14
minutes
16
1 pint per sample
• Kiln Ash
30
minutes
8
1 quart per sample
Composite Storage
Size/Type+
Half-gallon amber glass
Half-gallon amber glass
Half-gallon amber glass
Half-gallon amber glass
Half-gallon amber glass
Half-gallon amber glass
Half-gallon amber glass
1 gallon steel can
250 ml amber glass
** Duplicate samples were collected from streams.
It 500 ml aliquot of composited PCB feed and shredded capacitors were removed from composite
storage containers and provided to ENSCO.
t Samples were taken approximately one-half hour to one and one-half hours after the other
samples to allow for the residual time of the capacitors in the kiln between the time they
entered the kiln to the time they reached the sampling location at the exit from the kiln.
-------
17 0
the combustion chamber (thermal oxidation unit, TOU) temperature, concentra-
tions of 02 and C0' and the PCS fue1 feed rate- C02 was a1'so monitorecl be-
cause it is required to calculate the combustion efficiency, which also is a
required parameter. Residence time, also a required parameter, is strongly
related to the volume of the combustion zone to the temperature within that
region and less strongly to the amount of excess air in the combustion zone.
All of these parameters, and also the kiln temperature, have been summarized
on a daily average basis for the three test days in Table 5. Table 6 presents
the minima and maxima for these same parameters, based on the 15-minute in-
stantaneous values. Raw data have been included in Appendix C.
The average TOU temperature during the three tests was 2239°F (1226°C).
This is well above the minimum value of 2012°F (1100°C) required by EPA (40
CFR 761.40). In fact, this temperature never fell below 2150°F (1177°C) dur-
ing the entire sequence of tests. Similarly, the residence time of the process
never fell below 2 seconds, well above the minimum allowable (40 CFR 761.40).
Residence times on the order of 4 seconds or above were typical, and the low-
est instantaneous value was 3.27 seconds. No problems were encountered with
any of the temperature instrumentation.
Continuous monitors were used to measure oxygen and carbon monoxide levels.
A gas chromatograph was also available to measure carbon dioxide, so that all
of these values were recorded on a 15-minute basis. All of the analyzers per-
formed well, as evidenced by the generally small standard deviations for CO,
COg, and 02 (evidence also for stable combustion). The only exception was a
period of high CO values during the background (first) test. A CO spike oc-
curred (Table 6) in which the highest recorded value was 150 ppm. This ex-
plains the unusually high standard deviation for CO on this day. While PCBs
were being burned, 50 ppm was the highest allowable CO concentration without
requiring shutdown of PCB feed. It can be seen that CO never exceeded 4.5 ppm
during the two PCB tests. The average C02 concentration over all three tests
was 5.9%, the range extending from 3.5 - 10.02. The lowest acceptable 02 value
for this facility was 4.52, and the lowest 02 value measured was 8.0%. Com-
bustion efficiency is a function of CO and C02 concentrations. The lowest In-
stantaneous combustion efficiency was 99.7506%, occurring simultaneously with
the 150 ppm CO spike during the background test. The lowest value encountered
a
-------
TABLE 5. COMBUSTION PARAMETERS FOR THE THREE TEST DAYS
Test One, 8-26-80
Background Fuel
Test Two, 8-27-80
PCB & Background Fuel
Test Three, 8-28-80
PCB & Fuel Oil
Parameter
X+
s+
TOU Temperature
Kiln Temperature
o2 (%)
co2 (%)
CO (ppm)
2234°F
1223°C
1375°F
746°C
11.2
6.0
12.8
Combustion Efficiency (%) 99.9779
Residence Time (Seconds) 2.60
Liquid PCB Feed Rate*
PCB Capacitor Feed*
Rate
0
0
42.84
95.74
1.06
1.54
29.21
0.0489
0.42
2235°F
1224°C
1360°F
738°C
12.7
6.1
1.6
99.9970
2.80
520 lb/hr
236 kg/hr
1213 lb/hr
546 kg/hr
30.96
80.46
0.86
1.22
1.46
0.0046
0.43
2247°F
1231°C
1412°F
767°C
12.6
5.7
0.76
99.9988
2.54
459 lb/hr
208 kg/hr
1128 lb/hr
512 kg/hr
28.89
127.66
1.34
1.56
0.87
0.0052
0.61
* Total Flow. Exact PCB content of these materials will be known pending laboratory analysis.
+ Y and s are, respectively, the average and standard deviation.
-------
TABLE 6. MAXIMUM AND MINIMUM VALUES OF COMBUSTION PROCESS PARAMETERS
Test One, 8-26-80 Test Two, 8-27-80 Test Three, 8-28-80 PCB Mode Operation
Background Fuel PCB & Background Fuel PCB & Fuel Oil (Tests Two & Three)
Parameter	Max. Min. ~ Max.	Min.	Max. MirT. Max. Min.
TOU Temperature
2350°F
1288°C
2150°F
1177°C
2300QF
1260°C
2180°F
1193°C
* 2300°F
1260°C
2200°F
1204°C
2300°F
1260°C
2180°F
1193°C
Kiln Temperature
1500°F
816°C
1300°F
704°C
1500°F
816°C
1150°F
621°C
1725°F
941 °C
1150°F
621 °C
1725°F
941°C
1150°F
621 °C
o2 (%)
13.0
8.0
13.5
10.5
14.7
8.0
14.7
8.0
co2 (%)
10.0
3.5
8.3
4.5
10.0
3.5
10.0
3.5
CO (ppm)
150.0
0
4.5
0
4.0
0
4.5
0
Combustion Efficiency
(%)
99.9999 99.7506
99.9999
99.9900
99.9999
99.9954
99.9997 99.9900
Residence Time
2.49
2.68
2.73
2.86
2.49
2.58
2.86
2.49
vJ
(Seconds)
-------
•> -7 -f
bis
during the PCB incineration tests was 99.99", above the 99.9% required by EPA.
PCS feed rates can only be given in terns of total mass flow of those
wastes containing PCB, since the exact PCB content of these materials is not
yet known. No PC8s were fed during the background test (first test day).
Feed rates for PCB liquid (actually an oil) were 236 kg/hr (520 lb/hr) and
208 kg/hr (451 lb/hr) for the second and third test days. The corresponding
rates of solid PCB capacitors were 550 kg/hr (1213' lb/hr) and 512 kg/hr
(1128 lb/hr).
Other mass flow rates through the incinerator facility are also of in-
terest. The total fuel flow rates were 2244 kg/hr (4948 lbs/hr), 2177 kg/hr
(4800 lbs/hr), and 1650 kg/hr (3637 lbs/hr), respectively, for the three test
days. Amounts of injected water ranged from 1424 to 2150 kg/hr (.3140 to 4740
lbs/hr) and mass flow rates of ambient air were on the order of 39,000 kg/hr
(85,000 lbs/hr). Consumption of caustic solution by the scrubber ranged from
945 kg NaOH/hr (2084 lbsNaOH/hr) for the second test day to 372 kg NaOH/hr
(821 lbs NaOH/hr) for the third test day. This is logical since fuel oil was
used instead of chlorinated organic waste on the third day, drastically re-
ducing the chlorine content of the composite fuel for that day.
11
-------
3 7H
3. FACILITY DESCRIPTION
The facility described in this section is a hazardous waste incinerator
operated by the Energy Systems Company (ENSCO), El Dorado, Arkansas. This
system represents the state of the art for thermal destruction of hazardous
wastes. High destruction efficiency and wet scrubbing of the flue gas mini-
mize environmental impacts from plant operation. This description is based
on on-site observations, numerous conversations with ENSCO personnel, and a
previously prepared test report (1). Figure 1 shows a schematic of the in-
cineration process.
3.1 WASTE FEED FACILITIES
The equipment in use at ENSCO is capable of handling a large variety of
liquid, solid, and semi-solid wastes. A shredder and hammermill are used to
reduce objects as large as utility capacitors to very small pieces which will
burn easily. Liquid wastes can be piped from large tanks or pumped out of
smaller containers, such as 55 gallon drums. The drums can be steam cleaned
or destroyed by the shredder and incinerated.
Shredding prepares solid objects for incineration in a rotary kiln. Two
hammermills are employed in series to produce fragments roughly 0.5 inch square
The primary hammermill, rated at 30 Hp, is mounted directly over the secondary
hammermill, rated at approximately 50 Hp. Solids (such as electronic capa-
citors) are stored in 55 gallon drums. Each drum is weighed and hoisted up
to the level of the primary hammermill, where an operator slowly empties the
barrel into a hopper. Snreddings from the primary hammermill fall through the
secondary hammermill and collect on the bottom of the shredder enclosure In a
form referred to as "fluff". The fluff is deposited into the rotary kiln
through a screw worm auger. Release of PCBs from the shredding operation is
minimized by maintaining the hammermill hoppers at subtamospheric pressure
(0.1-0.2 inches H20 or 0.019-0.037 cm Hg). This subatmospheric pressure is
12
-------
SHALL
ELECTtOMlC
CAPACITORS IM
1S-CAL. OtUHS
LIQUID ORQANIC WASTES
NATURAL CAS
COMBUSTION AIR
bT ACK
TOTALLY ENCLOSED
snbeddebs
STEAM.
1000 CU. FT
PRIMARY
COMBU ST1ON
CHAMBER
NATURAL CAS
TlMI
n«n:KAnjat
¦WELL WATER
LIQUID OBCAMIC WASTES
ciiMi-kt sbi:o
AIR
^"STEAH
ROTARY KILN
—^VENTURI THROAT
SI ACK
ASH
otor
DRAIN
1000 CU. FT
SECONDARY
COMBUSTION Alt
SIEVE TkAYS
AtfCCB
COMBUSTION
CHAMBER
UATtk
CXNAUST AIR
TEHPOBABY ASH
STOBACE ONSITfc
SLUDCE LACOON
TANK
Figure 1. Schematic of incineration system at ENSCO.
-------

maintained by a blower (rated 3000 cfm at 1 psi) which exhausts into the
rotary kiln, providing a source of combustion air for incineration of the
sol ids.
Most liquid wastes can be fired directly into any of several burners
in the incineration system. However, very hazardous wastes such as PCBs
are fed only into the rotary kiln because this affords a somewhat longer
residence time. Such wastes are stored in drums which are opened one at a
time, eliminating the possibility of a widespread spill. Liquid PCB feed
is accomplished by placing a full 55 gallon drum onto a scale. The cap is
removed from the drum and a hand operated gravity fed pump developing
up to 50 psi of suction is used to draw the liquid out. The pump operator
can feed a given rate of liquid PCB by watching the weight change on the
scale and turning the pump on or off-as necessary. Rather than injecting
this liquid directly into the kiln, it is first mixed with the fluff
settling on the bottom of the shredder enclosure. Thick organic liquids
such as PCBs adsorb readily onto the fluff, so that there is no residual
liquid in the fluff mixture that is brought into the kiln by the auger.
Heavy, non-solid materials such as sludges can be fed manually through
a door at the base of the shredder enclosure. These would then be lifted
into the rotary kiln by the auger. The opening is large enough to accom-
modate a 55 gallon drum, so that waste drums can be cleaned by placing them
upside down into the enclosure and applying a high pressure steam jet to the
inside of the drum.
3.2 INCINERATOR
The incinerator is comprised of three combustion units in series.
These are the rotary kiln, the primary combustion chamber, and the secon-
dary combustion chamber. All have refractory linings and maintain tempera-
tures adequate for the destruction of even the most stable compounds. The
greatest volume of liquid wastes are fired into the primary combustion
chamber. The kiln is primarily intended for destruction of solids although
liquids are also fired there.
Fluff from the shredding operation is carried by the auger into the
front of the kiln. This unit is 34 feet long, has a 7 foot inside diameter,
14
-------
377
and rotates once every two minutes. The internal geometry of the kiln is
such that solids will be "turned over" as they traverse its length. Reten-
tion times are approximately 60 minutes for solids and 3.7 seconds for
gases. In addition to solid fluff, natural gas, water, and liquid organic
waste can be fired in the kiln. It is desirable to maintain a temperature
of 1300°F (704°C), and the flow rate of liquid organic waste can be adjusted
accordingly. A control console is located near the front of the kiln, where
an operator can monitor kiln temperature, kiln draft, draft at the top of
the primary hammermill, and pressure drop across the auger. An ash drop at
the opposite end collects the solid residue, and a drag chain lifts it out
into containers for disposal.
Gases leaving the rotary kiln are literally pushed into the primary
combustion chamber by a jet of steam situated so as to impart momentum to the
gas stream (steam ejector). The primary combustion chamber and afterburner
make up a pair of identically shaped furnaces connected by a crossover duct.
In addition to liquid waste, natural gas, water, or aqueous waste may be
fired in the first chamber, where essentially complete combustion takes
place. Ambient air inlets are also provided here to insure adequate excess
oxygen levels.
Any ash content of the fuel collects on the furnace floor in molten
state and can be removed by a slag tap. A thermocouple is mounted in the
crossover duct, and this is taken as a conservative measure of temperature
in the primary combustion chamber. This is the temperature reading used to
satisfy EPA requirements for monitoring of the incineration process.
The secondary chamber functions to increase residence time of the
combustion gases and thus insure maximum destruction efficiency. This
chamber is baffled to promote thorough mixing. A thermocouple mounted at
the outlet of the secondary chamber measures the lowest temperature. This
is roughly 300°F lower than the primary combustion chamber (crossover duct),
Which is usually maintained a few degrees above the EPA recommended value of
2192°F. These temperatures, as well as the crossover duct draft, are moni-
tored in the plant's control room. The combustion chambers have internal
volumes of about 3000 cubic feet (85 m3) each, which provides a total resi-
dence time of approximately 4.5 seconds in both combustion chambers.
15
-------
S7%
3.3 SCRUBBER AND DISCHARGE SYSTEM
Combustion gas leaving the secondary chamber is cooled by a liquid
quench, enters the wet scrubber, and eventually exits through the stack.
The scrubber is a closed-cycle system in which spent scrubber liquor is dis-
charged to a settling lagoon and clarified surface water is recycled to the
scrubber. There is no surface water discharge from the lagoon. A vertical
tower scrubber is used, and although there is a Venturi device followint
the tower, its principal function is not to scrub the flue gas but to pro-
vide draft for the incineration system.
Combustion gases leaving the secondary chamber enter a quench section at
the bottom of the scrubber tower. Quenched gases pass up through the tower,
and recycled scrubber liquor is sprayed down from the top. A series of sieve
trays are coated with the flow of scrubber liquor and provide a large contact
area between the combustion gas and the highly alkaline scrubber liquor.
Thus, acids in the gas dissolve into the liquid and are neutralized (particu-
larly HC1). Particulates are entrained into the liquor phase. Liquid collect-
ing on the bottom of the scrubber is also used as the quench, is ultimately
drained off as spent scrubber liquor, and flows by gravity to the sludge
lagbon.
Solids settle out to the bottom of the sludge lagoon. The relatively
clear surface water is treated with lime and/or caustic solution to restore
proper alkalinity and then pumped back into the scrubber tower.
After leaving the scrubber, gases enter a duct 25 feet long which connects
the scrubber to the base of the stack. This duct contains a high velocity jet
which points directly into the converging section of a Venturi throat. Com-
pressed air, steam, and stack drain liquid are injected into the gas stream.
The gas stream is entrained into this "liquid jet" and acquires some of its
velocity. The Venturi converges very rapidly, then diverges gradually so
that the velocity energy of the gas is transformed into pressure energy which
provides a driving force for release through the stack. A measurement of jet
draft (typically 12 cm Hg or 6 in H20) is taken just before the gas stream
reaches the point of jet release. Gas samples for the CO and 02 monitors
are withdrawn in the duct between the scrubber and compressed air jet.
16
-------
V)
Although the principal purpose of the Venturi is to provide draft, some
scrubbing occurs here. For instance, the steam provides a supersaturated
atmosphere so that water vapor will condense on submicron size particles,
increasing their mass and thus facilitating removal from the gas stream. The
converging section of the Venturi promotes agglomeration of small liquid par-
ticles so that they can be removed by gravitational forces.
Flue gas exits the Venturi duct and rises to escape through the stack.
Large liquid droplets fall quickly to the bottom of the stack and are drawn
off to the sludge lagoon. A demister pad mounted inside of the stack cap-
tures any small liquid particles remaining in the gas stream. A spray of
fresh water is maintained over the top of the demister pad to keep it clean
and to inhibit any upward motion of the water collecting on it. Make-up
water for the entire system is added at the bottom of the stack. The pres-
sure drop across the demister pad is 0.1 inches water. The stack pressure
on top of the demister pad is -0.5 inches water (-0.09 cm Hg).
17
-------

4. SAMPLING AND MONITORING
4.1 STACK GAS SAMPLING
4.1.1	Samplinq Locations
All sampling locations are identified in Figure 1. Figure 2 is a sche-
matic of the stack, including a cross sectional view depicting the traverse
point locations.
The inorganic gas samples for molecular weight determination were taken
at the stack sampling locations. ENSCO monitored the gases for CO and 02 at
a point between the scrubber and the compressed air jet.
4.1.2	Sampling Train Description
The PC8 sampling train, shown schematically in Figure 3, was utilized to
collect the organic constituents of the flue gas. The PC3 train is a modifi-
cation of a standard EPA Method 5 train. A solid sorbent trapping system was
incorporated downstream of the probe and between the third and fourth impin-
ger. By placing the sorbent traps in this position, the gas stream was rela-
tively dry as moisture was condensed in the first two water filled impingers.
Granular XAD-2 resin was used for the primary sorbent material. As a backup
to XAD-2, Florisil was used to adsorb any residual organics or organochlorine
compounds that may break through the XAD-2 trap. The two sorbent traps are
water jacketed, (Figure 4), and an exiting gas temperature of 16°C (60°F)
was maintained by recirculating the impinger ice bath water through the
jackets. Because of the sensitivity of organics to ultraviolet light, the
sorbent traps were wrapped in aluminum foil, while the impingers and inter-
connecting glassware were shielded from light.
4.1.3	Operating Procedures
Sampling for organics was done concurrently at the stack outlet location
utilizing the PCB train. Two identical samplng trains were operated at a rate
18
-------

30
TRAVERSE POINT LOCATION
INCHES FROM
OUTSIDE OF NIPPLE
POINT
NO.
22.64
39. 74
INCINERATOR 8UIL0IN8
Figure 2. Schematic of ENSCO's stack and samDling points.
19
-------
1*2
LOR ISIL
TRAP
ALL
AD-2
RA?'
THERMOMETER
PROB
REVERSE-TYPE
PT.TOT TUBE
PITOT
MANOME
FLOW
VACUUM
LINE
THERMOMETER
3Y-PASS
VALVE
MAIS
VALVE
DRY
TEST
METER
ORIFICES
MANOMETER
Figure 3. Schematic of PCB sampling train;
FLOW DIRECTION
THERMOCOUPLE
28/12 'SOCKET
GLASS
WATER.JACKET
ADSORBENT

GLASS FRITTED
SOCKET
DISK
GLASS WOOL
PLUG
Figure 4. Schematic of adsorbent traps.
20
-------
approximating the flue gas velocity (isokinetically according to EPA Method 5,
40 CFR 60}. The gas stream was passed through the impinger solutions (water)
to condense any moisture present in the stack gas and to remove all particulate
matter and any associated organics. The gas stream then passed through the
sorbent traps to adsorb the remaining organic constituents. Process parameters
such as temperatures, pressures, gas volumes were monitored throughout each
sampling period.
Prior to test initiation, the sampling trains were leaked checked at 15
inches mercury. If a leak rate greater than 0.02 cfm was present, the samp-
ling train was systematically checked to find the leak. No train was opera-
ted with an unacceptable leak rate. No grease was used on any of the glass
fittings. Post-test leak checks were performed at the highest vacuum encoun-
tered during the sampling time. Leak checks were also performed prior to each
port change or any disassembly of the train.
The sampling time was of 240 minutes duration and at a sampling rate of
less than 0.75 cfm.
4.1.4 Sample Recovery
Upon completion of testing, each train was removed to a clean area (mobile
laboratory) for recovery. Each train was recovered separately to prevent sam-
ple mix-up and cross contamination. Each train component was recovered per
the following:
•	Probe and nozzle were wiped to remove residual particulate.
•	After recovering dry particulate from the nozzle and probe,
these parts were rinsed with acetone and hexane. Rinsate
was retained in an amber glass bottle (labeled, then sealed)
•	Florisil and XAD-2 resin traps were removed from the train,
capped, and then wrapped for shipment to the analysis lab-
oratory.
t Impingers - each of the impingers was weighed, weights were
recorded, and contents were transferred to amher glass bot-
tles. The impingers were then rinsed with small amounts of
acetone and hexane. Rinses were combined with the impinger
catch, and rinse volumes were recorded.
•	Silica gel impingers were weighed, recorded, and regenerated
for subsequent use.
21
-------

$ All retained samples were sealed and labeled with an EPA Chain
of Custody tags,
« PCB feed samples were not brought into the van for recovery
in order to prevent contamination of the train samples.
4.2 LIQUID STREAM SAMPLING
This section summarizes the liquid sampling locations, the procedures
used to collect these samples, and the methods by which the samples were re-
covered.
4.2.1	Samp!inq Locations
The location of the sampling points can be seen in Figure 1. A descripti
of each sampling point is presented in the following subsections:
•	Liquid Organic Waste - Sample Stream 1 - Samples of the chlori-
nated organic waste stream were taken from a valve in the pipe-
line prior to introduction into the rotary kiln.
•	Liquid PCBs - Sample Stream 2 - Liquid PCBs were pumped into the
hammermill hopper before the auger system. Samples were taken
from the line running between the liquid PCB containers and the
hammermill.
•	Well Water - Sample Stream 5 - Well water samples were taken from
a valve in the pipeline leading to the scrubber.
•	Scrubber Liquor - Sample Stream 6 - Spent venturi scrubber liquor
samples were collected from an open portion of the pipeline trans-
porting the liquor to the sludge lagoon.
•	Recycled Scrubber Liquor - Sample Stream 7 - Recycled scrubber
liquor samples were taken directly from the pipeline leading to
the scrubber.
•	caustic Solution - Sample Stream 8 - Caustic solution samples were
collected from the pipeline leading from the caustic tank to the
venturi scrubber.
•	Lime Slurry - Sample Stream 9 - Lime slurry samples were collected
from a valve in the pipeline leading from the lime mixing tank to
the scrubber.
4.2.2	Operating Procedures
Table 4 illustrates that sampling frequencies and sample quantities were
the same for all of the liquid streams except the liquid PCB feed. The differ
ence with the liquid PCB stream was that it was sampled twice as often (15-
minute intervals). This was done because there was no premixing of capacitors
22
-------
3>*r
and a wide variation in PCB concentration was expected. Liquid stream samp-
ling frequencies were performed at times concurrent with stack sampling.
Each liquid process stream (liquid PCB feed, liquid waste feed, well
water, recycled scrubber liquor, scrubber liquor, lime slurry and caustic
solution) sample was collected in a hexane rinsed amber glass bottle (.250 ml)
and transferred to a hexane rinsed amber glass bottle (1/2 gal) to obtain a
composite sample for each test. Duplicate sets of composite samples were ob-
tained for each type stream.
4.3 SOLID STREAM SAMPLING
This section summarizes the solid sampling locations, the procedures used
to collect these samples, and the methods by which these samples were recovered.
4.3.1	Sampling Locations
The location of the sampling points is shown in Figure 1. A description
of each site is presented below:
Shredded Electronic Capacitors - Sample Stream 3 - Samples of the
shredded electronic capacitors (fluff) were collected from the
haiTmermill hopper prior to the fluff falling into the auger feed
system.
Rotary Kiln Ash - Sample Stream 4 - The rotary kiln ash samples
were taken directly from the ash hopper.
4.3.2	Operating Procedures
Table 4 illustrates the sampling frequencies and quantities of samples col-
lected. Solid sampling frequencies were performed concurrent with stack samp-
ling.
4.3.3	Sample Recovery
The shredded solid capacitor samples were collected in a new, hexane rin-
sed 1-pint steel can and composited in a new hexane rinsed 1-gallon steel can.
Duplicate sets of composite samples were obtained. The rotary kiln ash was
sampled with a hexane rinsed steel can (1 qt) attached to a clean steel handle.
The individual samples of ash were composited in a hexane rinsed, galvanized
steel can (20 gal). At the end of 6ach sampling test, a composite sample was
mixed by rolling the 20-gallon can. This method of mixing prevented the sample
from becoming contaminated by a stirring rod. Duplicate 250-ml samples of the
23
-------

mixed ash were then transferred to hexane-rinsed amber glass bottles.
Each solid/liquid sample container was labelled prior to sample collection
to avoid mislabeling. Following sample collection, each container was sealed,
tagged, and packaged in accordance with EPA Chain of Custody procedures. The
samples were shipped to Wright State University for analysis via various
freight carriers. Each sample set was sent via a different carrier to prevent
all samples from being lost or damaged in the same shipment.
4.4 PROCESS MONITORING
While the three tests were being performed, records were kept of operating
parameters within the incinerator plant. These included temperatures, combus-
tion gas concentrations, waste feed rates, and scrubber operations as described
in subsequent subsections. Calculations for residence time and combustion ef-
ficiency are also presented.
The information presented here has been reduced from the raw data sheets
included in Appendix C. Means and standard deviations have been calculated
for each test day whenever possible. In most cases the data were recorded at
15 minute intervals. Any exceptions have been noted.
4.4.1 Temperature
Four temperatures were monitored on a 15-minute basis: thermal oxidation
unit (TOU), combustion chamber outlet, kiln, and scrubber. All were measured
with thermocouples, which provided signals to strip chart recorders in the cen-
tral control room. TOU temperatures were measured in the passageway between
the primary and secondary combustion chambers (Figure 1). Outlet temperatures
were measured in the duct leading to the scrubber from the secondary combustion
chamber. Kiln temperatures were measured in the front of the kiln. Scrubber
temperatures were measured just below the sieve trays in the scrubber. No
problems occured with any of the temperature instrumentation throughout the •
three day test period.
Table 7 provides a summary of all temperature data for the three test days
The TOU passageway is the location for determining whether EPA incineration cri-
teria (temperature and residence time) are being met. The minimum acceptable
temperature was 1100°C (2012°F) while PCBs were being burned, without requiring
24
-------
TABLE 7. SUMMARY OF TEMPERATURE DATA
Test One. 8-26-80	Test Two. 8-27-80	Test Three. 8-28-BO
Location of	Background Fuel	PCB I Background Fuel	PCU t Fuel Oil
Teaperature Measurement jj* s* n*	Hin	X	s n Hax Mln	X	s n Max Mm
TOU Passageway
"F
2234
42.84
30
2350
2150
2235
30.96
29
2300
2)80
2247
28.89
29
23(10
22(1(1

"C
1223
-
-
1286
1177
1224
-
-
1260
1193
1231
' -
-
1260
12(14
Coafcustton Charter Outlet
"F
1773
55.78
30
1825
1500
1775
19.59
29
1810
1720
1805
28.8b
29
1900
1720
(Scrubber Inlet)
"C
967
-
-
996
816
968
-
-
988
938
9U5
-
-
1038
938
Kiln
"F
1375
95.74
4
1500
1300
1360
80.46
29
1500
1150
1412
127.66
26
1725
115(1

"C
746
-
-
816
704
738
-
-
816
621
767
-
-
941
621
Scrubber Tower
"F
162
4.31
30
175
158
162
2.45
29
171
160
161
4.23
29
170
151

"C
72
-
-
79
70
72
-
-
77
7l
72
-
-
77
t>6
* _
x, s, and n are. respectively, the Mean, standard deviation, and nuriier of data points.
-------
Vi%
shutdown of PCB feed. It can be seen that this temperature never fell below
1193°C (2180°F) while PCBs were being incinerated. Outlet temperatures were
somewhat lower than TOU temperatures, falling in the range between 816-1033°C
(1 500-1900°F). Kiln temperatures ranged from 621°C-941°C (1150°F to 1725°F).
Temperatures downstream of the quench (scrubber tower) were roughly 66-79°C.
4.4.2 Combustion Gas Concentrations & Combustion Efficiency
Levels of 0^, CO^, and CO in the combustion gas were determined every 15
minutes by reading continuous monitors or by performing simple analyses in the
control room. An electrolytic fuel cell type analyzer (Ecolyzer Model 3103)
was used to measure CO levels. An automatic shutoff device was set to stop
PCB flow to the kiln if this instrument read a CO level higher than 50 ppm.
Oxygen was also measured continuously by a paramagnetic type analyzer (Taylor-
Servomax Model OA-289). Both of these instruments supplied signals to strip
chart recorders. Carbon dioxide was measured using a Fyrite analyzer, a small
instrument which contained a column of liquid selectively absorptive for COg.
Sample gas was drawn from the scrubber tower, so that C02 measurements are all
for "clean" flue gas.
Tables 5 and 6 present all the summarized data for CO, C02, and 02> CO
levels averaged 5.1 ppm over all three tests and 1.2 ppm over the two PCB tests
A large CO peak occurred during the first day, with measured CO levels as high
as 150 ppm. However, CO levels were consistently low during the PCB tests; the
highest recorded value was 4.5 ppm. The automatic shutoff system could allow
CO concentrations up to 50 ppm. Some CO spikes did occur during these tests.
The most notable one resulted from a flame failure in the kiln. However, this
spike, as well as the others, lasted for ortly a very short period of time.
Oxygen and carbon dioxide levels averaged 12.2 percent and 5.9 percent,
respectively, over the three tests. An excess oxygen level of 4.5 percent is
considered adequate for this facility. The higher values of C02 generally in-
dicate better combustion efficiency, which is given by the.formula:
Combustion Efficiency « CE • 100X [CO2]/([C02] + [002]),
Where [C02] and [CO] are the concentrations of carbon dioxide and carbon mon-
oxide, respectively, as percents.
26
-------
3>
-------
TABLE 8. DAILY AVERAGES OF MASS FLOW RATES THROUGH THE INCINERATOR

Test One, 8-26-80
Background Fuel
Test Two, 8-27-80
PCB & Background Fuel
Test Three,
PCB & Fuel
8-28-80
Oil
Flow
lb/hr
kg/hr
lb/hr
kg/hr
lb/hr
kg/hr
Total Fuel
4948
2239
4800*
2172*
3637*
1646*
Inert Fraction
228
103
456
206
360
163
Combustible Fraction
4720
2136
4344
1966
3277
1483
Liquid Organic Waste
3871
1752
3048
1379
0
0
Capacitors
198
90
1213*
549*
1128*
510*
Waste Oil
37
17
520*
235*
459*
208*
Sludge, Ink & Paint
810
367
0
0
0
0
Natural Gas
32
14
19.3
8.7
158
71
Fuel Oil
0
0
0
0
1892
856
Water Injected
3645
1649
3140
1421
4740
2145
Ambient Air
85837+
38840*
79722*
36073+
87754*
39 708+
Combustion Gas
94430+
42729*
87662+
39666+
96131+
43498*
Contains PCBs
+Based on combustion equation calculation
-------
3*/
From Table 3 it can be seen that total fuel feed rates range from roughly
1584 to 2262 kg/hr (3500 to 5000 Ibs/hr). Also, from this table it is obvious
which fuels were burned on which days and their relative amounts. The combus-
tible and inert fractions were calculated from the ash content of the composite
fuel, known from ultimate analysis to be 4.6, 9.5, and 9.9 percent for the
first, second, and third test days, respectively. The total fuel input, added
to the flows of water and ambient air, gives a range of average total combus-
tion gas flow rates from 39,666 kg/hr (87,662 Ib/hr) to 43.498 kg/hr (96,131
lb/hr).
4.4.4 Residence Time
Residence times were calculated by dividing the stated volume of the com-
bustion region by the volumetric flow rate of the combustion gases corrected
for temperature (2). Correction for pressure was not made because pressures
in the system were negligibly different from atmospheric. Mathematically,
Residence time = RT = V _ V
$ ~ m7p
where V = volume of the combustion zone (ft^), Q ¦ volumetric flow rate
•a	3
(ft /sec), m = mass flow rate (lb/sec), and o = density (lb/ft ) of the combus-
tion gases, respectively.
At ENSC0, the T0U temperature (Figure 1) is normally maintained above the
regulatory minimum temperature. Thus, it is the T0U temperature and the volume
up to that point (4624 ft^) that were used in calculating residence times.
Gas density as a function of temperature can be calculated from the follow-
ing relationship (2):
- _ Tref
p " (13.1)Tcomb
where p s density in lb/ft^, Tref = 60°F (520°R), the constant 13.1 is the
3
volume (ft ) occupied by 1 lb of air at 60°F and 1 atmosphere pressure, and
Tcomb is the T0U temperature. These two expressions can be combined to give
the following relationship:
RT a (4624 ft^)(520°R)	
(13.1 ft3/lb)(m lb/sec)(Tcomb °R)
29
-------
m
Temperatures in the TOU are given in Tables 3 and 6. Combustion gas
flow rates are given in Table 8. Calculated residence times are given in
Tables 5 and 6. All residence times are of the order of 2.5 seconds, in ex-
cess of the minimum allowed by the PC8 Regulations.
4.4.5 System Drafts and Pressures
Table 9 presents a summary of most of the important system drafts and
pressures for the three test days. These include the draft due to the forced
air fan in the grinder enclosure, the pressure differential across the screw
worm auger, the kiln draft, the combustion chamber (TOU) draft, the jet draft
(measured just upstream from the jet nozzle, the pressure differential across
the demister pad, and the stack draft (measured just above the demister pad).
Scrubber draft measurements were not taken on a 15 minute basis,.
Since the venturi jet provides most of the motive power for this system,
the highest draft measurements were expected to occur here. Reference to Table
9 shows that they did. Jet drafts of 7.10, 6.62, and 6.42 inches of water
(1.33, 1.24, and 1.20 cm Hg) occurred on the three test days. All of the other
draft measurements were under 0.5 inches of water (0.09 cm Hg). Although scrub-
ber draft measurements were not taken every 15 minutes, the following two ob-
servations from test three show that the scrubber operates at a slightly lower
draft than is present at the jet:
Time	Jet Draft	Scrubber Draft.
The demister &p is usually about 0.1 "HjO (0.02 cm Hg), which is a low pressure
drop for a demister pad.
4.4.6 Scrubber Operation
Except for scrubber temperatures (Table 7), it was not required to monitor
other scrubber operational parameters at 15-minute intervals. Most of the
scrubber parameters were fairly constant, except for the caustic usage rate,
which varied proportionately with the amount of chlorine in the fuel. Scrubber
drafts were somewhat less than the jet draft (the jet draft averaged 6.71 "HjO
or 1.25 cm Hg over the three test days). Scrubber temperature averaged 72°C
9:30
10:15
8.0	"H20 (1.49 cm Hg)
6.1	"H20 (1.14 cm Hg)
7.3 "HgO (1.46 cm Hg)
4.8 "H20 (0.90 cm Ha)
30
-------
TABLE 9. DAILY SUMMARY OF INCINERATOR DRAFTS AND pressure



Test One,
8-26-80


Test Two,
8-27-80


Test Three,
8-26-80

Location


Background Fuel


PCB I Background Fuel


PCS &
Fuel
011

of Measurement

1*
s*
n
Max
Min
X
s
n
Max
Min
I
s
n
Max
Min
Grinder Enclosure
In H20
0.50
0
4
0.50
0.50
0.09
0.0267
29
0.19
0.06
0.14
0.0529
21
0.23
0.06

ca Hg
0.09
-
-
0.09
0.09
0.02
-
-
0.04
0.01
0.03
-
-
0.04
0.01
AP Across Auger
In H20
0.12
0.0245
4
0.15
0.10
0.13
0.0583
29
0.24
0.04
0.17
0.0793
21
0.30
0

ca Hg
0.02
-
-
0.03
0.02
0.02
-
T
0.04
0.01
0.03
-
-
0.006
0
Kiln Draft
In HgO
0.14
0.0455
4
0.20
0.10
0.18
0.0456
29
0.26
0.11
0.23
0.0495
21
0.35
0.15

ca Hg
0.03
-
-
0.04
0.02
0.03
-
-
0.05
0.02
0.04
-
-
0.07
0.03
TOU Draft
In H20
0.25
0.0413
30
0,30
0,18
0.27
0.0824
28
0.38
0.16
0.26
0.0736
27
0.41
0.15

ca Hg
0.05
-
-
1.59
1.10
0.05
-
-
0.07
0.03
1.20
-
-
1.49
0.84
AP Across Demister
in H20
0.09
0.0268
5
0.12
0.06
-
-
0
-
-
0.09
0.0190
25
0.14
0.06

at Hg
0.02
-
-
0,02
0.01





0.02
-
-
0.03
0.01
Stack draft
in H20
0.33
0.0389
22
0.41
0.23
0.35
0.0279
28
0.41
0.31
0.32
0.0212
2
0.33
0.30

ca Hg
0.06
-
-
0.08
0.04
0.07
-
-
0.0B
0.06
0.06
-
-
0.06
0.06
*x, s, and n are, respectively, the Dean, standard deviation, and nunber of data points.
-------

(112°F) over the three test days, covering a range from 66	to 79°C (151®F to
1753F). This represents a drop of 884°C (1622°F) from the	average combustion
chamber exit temperature, recycle flowrate to the scrubber	tower is unknown,
as well as the flowrate of material from the high velocity	jet.
The amount of hydrochloric acid in the flue gas can be determined from the
amount of chlorine in the fuel. For instance, the chlorine content of the com-
posite fuel for test one was 26.6 percent. Since the total fuel flowrate was
4948 Ibs/hr, this corresponds to 1316 Ibs/hr of chlorine or 37.6 lb moles of
chlorine. Similarly, 37.6 lb moles of hydrochloric acid corresponds to 1353
Ibs/hr of hydrochloric acid. Since an equal number of moles of NaOH (caustic)
would be required to neutralize this number of moles of HC1, the theoretical
caustic requirement would be about 15jp0 Ibs/hr of NaOH. This compares to an *
actual caustic consumption of 1789 Ibs/hr (810 kg/hr) for the first test day.
Equivalent information for the three tests is given in Table 10. The actual
caustic usage for the second test was 2084 Ibs/hr (943 kg/hr) and 821 Ibs/hr
(372 kg/hr) for the third test day.
4.5 QUALITY ASSURANCE
4.5.1 Test Preparation
All steps needed to clean the train, as well as all glass sample containers
are listed in Figure 5. All impingers used for organic collection were fired to
450°F, nitric acid washed, distilled water rinsed, acetone rinsed, and hexane
rinsed. Prior to initial assembly of the train in the field, all parts were
rinsed with hexane. This solvent was discarded. All subsequent field clean-
ings were with glass-distilled water, acetone, and hexane. Only Burdick and
Jackson Distilled-in-Glass, or equivalent, organic solvents were used. Visual
inspection for contamination was also performed.
To avoid contaminating the trains as they were assembled or disassembled,
as much work as possible was done inside a designated lab area.
SMOKING was prohibited in all but designated areas. Prohibited areas in-
cluded the stack and other sampling locations. Exposure of resin traps was
kept to a minimum. If any equipment was contaminated, it was completely re-
cleaned. All organic solvents were stored in amber glass containers. No grease
of any type was used on the train including the impingers. If a seal could not
32
-------
TABLE 10. CAUSTIC SOLUTION INPUT DATA
(a>
Ca>
Test One, 8-26-80	Test Two, 8-27-80	Test Three, 8-28-80
Description of Data	Background Fuel	PCB & Background Fuel	PCB & Fuel Oil
Chlorine Content of Fuel
(weight percent)	26.6	29.3	9.2
Hydrochloric Acid Produced
lbs/hr	1353	1446	345
kg/hr	612	654	156
Theoretical Caustic
Requirement
lbs NaOH/hr	1485	1587	379
Kg NaOH/hr	672	718	172
Actual Consumption
lbs NaOH/hr	1789	2084	821
Kg NaOH/hr	810	943	372

-------
y/6
CLEANING
PROCEDURES
I
NEW ?ARTS, TRAIN
PARTS AND ALL
SAMPLE CONTAINERS
J
SOAP ANO WATEP
CLEANING
I
SOAK. IN 15% HMO,
MINIMUM OF J
3 HOURS
I
TAP WATER RINSE
I
DISTILLED 'WATER
RINSE
I
DISTILLED IN
GLASS
^ »™SE
1
OISTILLEO IN
SLASS
ACETONE RINSE
I
SURDICX& JACKSON,
OR EQUIVALENT,
HSXANE RINSE
I
AIR DRY
I
CAP OR 5EAL
OPENINGS
YES
1
USED tWS
TRAIN ONLY
I
SOAK IN 1S=J HNO?
MINIMUM OF
3 HOURS
I
OISTILLEO WATER
RINSE
I
SiJ, OR
EQUIVALENT
ACETONE RINSE
I
8SJ, OR
EQUIVALENT
HEXANE RINSE
I
AIR DRY
I
VISUAL INSPECTION
FOR
CONTAMINATION
NO -fc REASSEMBLE TRAIN
Figure 5. Cleaning procedures.
34
-------
'5*7
be attained, that part was replaced in the sampling train.
The resins used through the PCB test sequences were cleaned by the follow-
ing methodology,
$ 10 gallons of distilled water is rinsed through the resin to re-
move residual salts.
•	A 24 hour continuous reflux of Distil 1ed-In-Glass methanol to
remove water.
•	For XAD-2, a 24 hour reflux with Oistilled-In-Glass methylene
chloride to remove organic solvent.
•	For Florisil, a 24 hour reflux of a blend of 5% diethyl ether
and 95% Hexane to remove organics.
•	Both resins were then dried under constant flowing nitrogen un-
til no organic residual odor is noticed.
Upon completion of cleaning a composite sample of each resin type was submitted
for a quality control check which consisted of extraction, concentration of ex-
tract, and total chromatographable organics (TCO), and gravimetric residues.
If the resin did not meet the 1.5 mg/150 g of resin for TCO and 3.8 mg/150 g of
resin for residue weight, it was rejected and returned for additional clean-
ing.
4.5.2 Method Five Calibration Data
This section describes the calibration procedures used prior to conducting
the field test at ENSCO's incinerator facility. Figure 6 shows the calibration
equipment and how it was set up.
4.5.2.1 Orifice Meter Calibration—
The orifice meter calibration is performed using a pump and metering sys-
tem as illustrated in Figure 6 (a). The dry gas meter with attached critical
orifice is run at various orifice flows for a known time. After each run the
volume of the dry gas meter, meter inlet/outlet temperatures, time, and ori-
fice setting is recorded. The orifice meter calibration factor is derived by
solving the equation.
.na _ 0.317 AH	p(Tw + 460) 9t2
-H® Pb (T. + 466) c'vS	 ]
35
-------
59*
THERMOMETER
LEVEL
.POINTER
WATER
LEVEL
GAUGE
ORY \
IS METER
MET TEST METER
o
HATER "3UT
LEVEL ADJUST
PUMP AIR OUTLET
SURGE TANK
ORIFICE
MANOMETER
Figure 6a. Orifice and dry gas meter calibration.
ELECTRIC MOTOR
STAfOARD PITOT TUBE
TEST HOLE
ADJUSTABLE DAMP®
VALVE
/— OAriPER
AIR FLOW
AIRj
STEEL TUBE
POINT WHERE TIP OF PITOT
TUBE MOULD BE WEN
TAKING A READING	1
PITOT TUBE BEING CALIBRATED
TOP VIEW
Figure 6b. P1tot tube calibration.
36
-------

where
aH =	Average pressure drop across the orifice meter, inches
h2o
Pb =	Barometric pressure, inches mercury
=	Temperature of the dry gas meter, °F
Tw s	Temperature of the wet test meter, °F
9 =	Time, minutes
Vw =	Volume of wet test meter, cubic feet
The aH@ yielded is utilized to adjust the sampling train flow rate by regula-
ting the orifice flow.
4.5.2.2* Dry Gas Meter Calibration-
Meter box calibration consists of checking the dry gas meter for accuracy.
The dry gas meter with attached critical orifice is connected to a wet test me-
ter (see Figure 6 (a)) and run at various orifice flows for a known time. Af-
ter each run wet and dry gas meter volumes, temperatures, time, and orifice
readings are recorded. Utilizing the equation:
,, _ Vw Pb (Td + 460)
V	" Vd (Pb + AH) fr + 460)
T3.6 w
where
V	=
Vw	=
Pb	=
Td	=
Vd	=
AH	=
Tw =
a volume factor
tained.
37
Volume correction factor (also known as "Y" in EPA Method 5)
Volume of wet test meter, cubic feet
Barometric Dressure, inches mercury
Temperature dry gas meter, °F
Volume of dry gas meter, cubic feet
Average pressure drop across the orifice meter,
inches H20
Temperature of wet test meter, °F
which compares the dry gas meter with the wet test meter is ob-
-------
400
4.5.2.3	Pi tot Tube Ca1i bration—
Pi tot tubes are calibrated on a routine basis utilizing two methods.
The type S pitot tube specifications are illustrated and outlined in the
Federal Register, Standards of Performance for New Stationary Sources, [40 CFR
Part 60], Reference Method 2 (refer to Figure 6 (b)). When measurement of
pitot openings and alignment verify proper configuration, a coefficient value
of 0.84 is assigned to the pitot tube.
If the measurements do not meet the requirements as outlined in the Federal
Register, a calibration is then performed by comparing the S type pitot tube
with a standard pitot tube (known coefficient of 1.0). Under identical condi-
tions, values of P, for both S type and standard pitot tube are recorded using
various velocity flows (14 fps to 60 fps). The pitot tube calibration coeffi-
cient is determined utilizing the following equation,
Pitot Tube Calibration = (Standard Pitot Tube X rAP reading of std. pitot ,
Factor (CP)	Coefficient)	aP reading of S type pitot"^
The coefficient assigned to the pitot tube is the average of calculated values
over the various velocity ranges.
4.5.2.4	Nozzle Diameters —
The nozzle diameters were calibrated with the use of a vernier caliper.
If the nozzle showed excessive wear or was considered not fit for use, it was
discarded.
4.5.3 Gas Chromatograph Calibration
Manufacturer's recommended calibration procedures were followed using the
following standard gas mix which had an accuracy of + 2%:
Carbon Dioxide	15%
Oxygen	5.15%
Carbon Monoxide	7.23%
Nitrogen	Balance
(All gases contained in one cylinder)
Scott Environmental Technology Inc., Specialty Gas Division
38
-------

Calibrations were performed at the start of each test day. Replicate in-
jections of the standard gas mixture were made until a comparison of corres-
ponding peak heights were in agreement to within 5%. At this point, the gas
chromatograph was considered calibrated and ready for sample analysis. A
strip chart recorded was used to record both the calibration and test data.
4.5.4	Blanks
For each test conducted at the ENSCO facility, a blank sampling train was
set up. The blank train consisted of the back half of the PCB train, impin-
gers, impinger solutions and resin traps. Prior to the start of a specific
test, the blank train was transported to the stack where it remained, protected
from light, for the duration of the test. Upon completion of the test, the
blank was recovered utilizing the same recovery procedure as the test sample.
In addition to the blank train, a small portion of the solvents, approxi-
mating the amounts used in rinsing the train, was composited during each test
cleanup and submitted as a solvent blank.
4.5.5	Chain of Custody
Once the samples were recovered from the sampling trains, liquid and solid
streams, they were labeled, taped and shipped according to EPA's chain of cus-
tody requirements.
The samples were individually wrapped with bubble-pak or other suitable
packaging material, placed in shipping containers and then given to Federal
Express or United Parcel Service for shipment to Wright State Laboratory. A
chain of custody form accompanied each sample shipment and each sample/shipment
was sealed and labeled with EPA chain of custody tags.
See Appendix E for the daily chain of custody record forms.
39
-------

5. TECHNICAL PROBLEMS AND RECOMMENDATIONS
The major problem that occurred during the tests at ENSCO was plugging
of the sorbent traps. This problem at ENSCO was less serious than that at
Rollins, but plugging did cause delays. A brown to red-brown material was
distributed throughout the train but appeared to be concentrated particularly
in the glass frit of the XAD-2 sorbent trap. The material could be either
organic or inorganic in nature. It should be noted, however, that the
material passed through two water filled and ice bath cooled impingers before
entering the XAD-2 trap. Perhaps it would be better to use a glass wool plug
instead of a glass frit in the sorbent traps in order to avoid plugging prob-
lems.
It was recommended that the material be analyzed, so that this problem
might be avoided in future test projects. These analyses are underway at
Wright State University.
There were also other, minor problems: occasional leaks in the trains
and some samples damaged in transit to Wright State University. These problems
would have been lessened had the test preparation period been longer than two
weeks and had there been a wider choice of shipping companies available. It
is also recommended that all samples be shipped in metal ice chests instead of
cardboard cartons.
There were minor facility problems: plugged line slurry valve, inadequa-
tely shredded capacitors, and jamming of the shredder. These sorts of prob-
lems are routine, and there is little than can be done to prevent them from
occurring.
40
-------
4
-------

A. GASEOUS MONITORING
Monitoring of oxygen and carbon dioxide for stack gas molecular weight
determination was performed at the ENSCO facility. The sampling positions
were located at ports midway on the stack exit after the incineration and
scrubber units.
The gas sample was drawn through a stainless steel probe, polypropylene
tubing, and an ice bath condenser by means of a small diasphragm pump. A
flow rate meter was used to measure sample flow to a Tedlar bag. The collect-
ed gas sample was then analyzed for carbon dioxide, oxygen, and nitrogen with
a thermal conductivity detector gas chromatrograph. The collection device con-
sisted of the following components.
•	Probe - stainless steel with a glass wool plug to act as
a filter
•	Condenser - air cooled or ice bath to eliminate excess
moisture
•	Valve - needle valve to regulate sample gas flow rate
•	Pump - leak-free, diaphragm type, or equivelent to transport
sample gas to the Tedlar bag
•	Flow rate meter - meter capable of measuring a flow range
from 0 to 1.0 1i ter per mi nute
t Gas sample bag - Tedlar sample bag or equivalent, with a
capacity of approximately 0.5 - 1.0 cubic foot
t Manometer - water filled (J-tube or dry gas meter to be
used for the Tedlar bag leak check
•	Vacuum gauge - gauge or meter to be used for the sample train
leak check.
Prior to field use, all Tedlar gas sample bags were leak checked.	The bags
were leak checked by inflating them to a pressure of 5 to 10 cm (2-4 in Hg)
as determined by an in-line manometer or equivalent. Any displacement in the
manometer after a 10-minute time interval was taken as indicative of a	leak, in
the field prior to the sampling operation, the sampling train was also	leak
42
-------
yes
checked. The leak check was done by placing a vacuum gauge at the Drobe inlet,
pulling a vacuum of at least 250 tot Hg (10 in. Hg), and then turning off the
pump. The vacuum should remain stable for at least one minute.
A sample was taken as follows:
•	Place the probe in the stack at the sampling point and
purge the sample line up to the bag
•	Connect the bag and make certain that all connections are
tight
•	Fill and evacuate bag twice before taking analysis sample
t Sample at a rate that will fill the bag in 10 minutes
•	Remove full bag, cap opening, transport to lab, and analyze
as soon as possible.
In the lab area, the Tedlar bag was analyzed on the Shimadzu gas chromato-
graph, model GC-3BT. The resultant peak heights were compared to a known
standard gas purchased from Scott Environmental Technology, Inc.
Table A-l presents the gas analysis data as collected from both the "hot
zone" and outlet sampling points at the ENSCO Facility.
43
-------
TABLE A-l. ENSCO GAS MONITORING DATA
Date
Time
Sampled
Sampling
Location
Results
CO.
0„
Comments
8/26
1320-1327
1356-1408
1435-1449
1615-1626
1642-1657
1720-1730
1750-1805
1830-1840
1905-1915
1938-1949
Outlet
Hot Zone
Outlet
II
la
it
7.41
8.16
6.53
5.90
5.65
6.15
5.40
5.40
6.03
6.41
11.60
10.30
12.40
12.50
12.70
12.30
12.60
13.10
12.60
12.30
Compare with ENSCO 0^ Readings
8/27
1247-1305
1356-1402
1425-1435
1500-1510
1530-1544
1635-1648
1712-1722
1758-1808
Outlet
Hot Zone
Outlet
6.43
5.33
4.97
6.06
7.64
5.58
7.64
5.94
13.09
13.89
14.16
13.30
13.95
13.73
11.59
12.98
Compare with ENSCO 0^ Readings
(Continued)
-------
TABLE A-l. (Continued)
Date
Time
Sampled
Sampling
Location
Results
CO,
0,
Comments
8/28
1132-1147
1226-1238
1320-1330
1412-1427
1730-1740
1755-1805
Outlet
Hot Zone
5.11
4.87
5.23
6.69
4.50
6.57
13.73
13.94
13.78
12.47
13.63
11.83
Compare with ENSCO 0^ Readings
a
cn
-s:
Q
-------
ho*
APPENDIX B
FIELD DATA SHEETS
46
-------
ijO'f
The material on pages 47-92 is not suitable
for reproduction.
These pages can be viewed at the Region 6
Office of the US Environmental Protection
Agency
1201 Elm Street
Dallas, Texas 75270
-------
mo
C. PROCESS DATA
Appendix C presents raw process monitoring data sheets taken by ENSCO per-
sonnel and checked by TRW while the three tests were in progress. The sheets
are organized by days, corresponding to the three test days (8-26-80, 8-27-80,
and 8-28-80). Each day contains at least five kinds of data sheets. At the
end of the appendix is a set of notes from ENSCO's Engineering Manager, listing
some fuel feed data and a set of sample calculations for the third test day.
These calculations are the same as those used for the other two days.
The five data sheets for each day include the control room sheet (TOU op-
erations), a "Kiln Operations" sheet, a "Liquid PCB Operations" sheet, a "Ham-
mermill Operations" sheet, and a "Forklift Operations" sheet. The TOU opera-
tions sheet contains drafts in inches H^O, Og and CO2 concentrations in percent,
CO concentration in ppm, temperatures in °F, gas meter readings in hundreds of
cubic feet, and various storage tank levels in inches. The kiln operations
sheets contains the kiln temperature in "F, drafts and pressures in inches H 0,
and water meter readings in tens of gallons. The Hammermill operations sheet
contains the weights in pounds of barrels of solid capacitors, and the time
frame in which they were fed to the Hammermill. The liquid PCB sheet contains
analogous data for the liquid feed. The forklift operations sheet duplicates
information on the Hammermill operations sheet. In addition to these, a main-
tenance operations sheet is included for the first day, and a record of fuel
oil feed is included for the third day.
93
-------
Hi/
The material on pages 94-146 is not suitable
for reproduction.
These pages can be viewed at the Region 6
Office of the US Environmental Protection
Agency
1201 Elm Street
Dallas, Texas 75270
-------
Hi 2
APPENDIX D
CALIBRATION DATA
147
-------
HI 3
The material on pages 148-138 is not suitable
for reproduction.
These pages can be viewed at the Region 6
Office of the US Environmental Protection
Agency
1201 Elm Street
Dallas, Texas 75270
-------
HH
APPENDIX E
CHAIN OF CUSTODY RECORDS
159
-------

The material on pages 160-171 is not suitable
for reproduction.
These pages can be viewed at the Region 6
Office of the US Environmental Protection
Agency
1201 Elm Street
Dallas, Texas 75270
-------
i y
Division 5
-------
l] »7
"DETERMINATION OF POLYCHLORINATED DIBENZO-p-
DIOXINS (PCwOs), DIBENZCFURANS (PCDFs), AND
3XPHENYLS (PC3s) IN STACK EFFLUENTS AND OTHER
SAMPLES FROM PCS INCINERATION TESTS AT
ROLLINS ENVIRONMENTAL SERVICES, DEER PARK,
TEXAS AND ENERGY SYSTEMS COMPANY (ENSCO), EL
DORADO, ARKANSAS"
Interim Report Under
U.S. EPA Cooperative Agreement No. CR806846
Submitted To
Mr. Michael Dellarco,' U. S. EPA
Washington, D.C. 20460
-------
V 18
DETERMINATION OF PQLYCHLORINATED DIBENZO-P-OIOXINS (PCDDs).
DIBENZOFURANS (PCDFs), AND 31 PHENYLS (PC3s) IN STACK EFFUJENTS AND
OTHER SAMPLES RESULTING FROM ASSESSMENTS OF PCS INCINERATION TESTS AT
ROLLINS ENVIRONMENTAL SERVICES, DEER PARK. TEXAS AND
ENERGY SYSTEMS COMPANY (ENSCO), ELDORADO, ARKANSAS
INTERIM REPORT ON WORK ACCOMPLISHED UNDER U.S. EPA COOPERATIVE
AGREEMENT NO. CR806846
Prepared By
T.O. Tlernan, J.G. Solch, G.F. VanNess, J. Garrett, M. Porter and M.L. Taylor
BREHM LABORATORY
WRIGHT STATE UNIVERSITY
DAYTON, OHIO 45435
Submitted To
Mr. Michael Dellarco
U.S. Environmental Protection Agency
Office of Pesticides and Toxic Substances
Special Pesticide Review Division
Washington, D.C. 20460
February 15, 1981
-------

TABLE OF CONTENTS
Table of Contents
I.	Introduction
II.	Description of Incineration Tests and Sampling
Procedures Utilized By TRW
III.	Analytical Methodology
IV.	Results
V.	Sunraary
Page Number
i
1
2-9
10-27
27-32
32
i
-------
1
I. INTRODUCTION
In July, 1980 the U.S. Environmental Protection Agency completed planning
for a series of sampling and monitoring activities to be conducted at two
commercial incinerators, one operated by Rollins Environmental Service, and
located at Deer Park, Texas, and the other operated by Energy Systems Company
(ENSCO), and located at El Oorado, Arkansas. These source assessments,
sponsored by the EPA's Office of Pesticides and Toxic Substances and managed
by EPA's Region VI office in Dallas, Texas, were aimed primarily at deter-
mining the concentrations and mass emission rates of chlorinated d1benzo-p-
dioxlns (CDDs) and chlorinated dlbenzofurans (CDFs) in the stack effluents
from each incinerator. The sampling and field measurement tasks associated
with these tests were conducted under contract with EPA by TRW, Environmental
Engineering Division, Redondo Beach, California, while all analyses of
collected samples were accomplished by The Brehm Laboratory, Wright State
University, Dayton, Ohio, under EPA Cooperative Agreement No. CR806846. The
present interim report describes a portion of the analytical results obtained
by Wright State in the course of this project.
-------
2
II. DESCRIPTION OF INCINERATION TESTS AND SAMPLING
PROCEDURES UTILIZED SV 1W
A. Incinerator Systems at Rollins Environmental Services and Energy Systems
Company CENSCO).
The incineration facilities at the Rollins and Ensco plants have been
described in some detail in the plan for testing developed &y TRW,1 and will
be only briefly mentioned here, in order to provide a basis for discussion
of the various types of samples which were collected during the monitoring
tests.
The incineration system utilized by Rollins Environmental Services
consists of a rotary kiln and a liquid injection burner, both feeding a
comnon afterburner. Temperatures attainable 1n these three burners are
normally 130Q-1500°C» The overall retention time of the incineration system
is 2-3 seconds. Solid wastes can be fed into the rotary kiln by conveyor,
(.although no solids were burned during the tests described here), while
liquids and sludges are pumped into the kiln or into the liquid injection
burner. Auxiliary fuel, natural gas, is used for initial heat«up and to
provide supplemental heat for the incineration of low heat capacity wastes.
Fuel oil (Number 2) 1s also used to provide heat for burning materials such
as PCBs, and is mixed with the PCBs prior to the waste Being, fed Into the
Incinerator. Gaseous emissions from the combustion of solid and liquid
wastes are controlled by a Venturf scrubber. Lime 1s injected to neutralize
the scrubber water, which 1s then admitted to settling ponds, where the
-------
^13-
3
water is analyzed and subjected to further treatment, if necessary, prior to
discharge into the Houston-ship channel. Exhaust gases are also routed
through absorption traps and a mist eliminator before-entering, the 30-m«t«r
exhaust stack.
The incineration system utilized By Energy Systems Company (.ENSCQ) 1s
also capable of combusting both solid and liquid wastes. This incinerator
consists of a rotary kiln, in which combustion temperatures of 1200-1 50Q°f
are attained, and 1n which the retention time for solid material is approximately
one hour and the gas retention time is approximately 4 seconds. Liquid wastes,
water and natural gas can also Be injected into the front end of the kiln.
Ash is removed from the end of the kiln By a drag chain and placed 1n steel
drums for storage. Gases which are evolved from the rotary kiln pass Into an
afterburner which consists of two separate combustion chambers. Residence
time in the primary chamber is approximately 2.2 seconds and temperatures of
2250-2340®F are typically maintained there. Most of the organic waste materials
are destroyed in this region. Molten slag resulting from the combustion
settles on the chamber floor and 1s removed through a slag trap. The. secondary
combustion chamber is baffled to facilitate gas mixing, and the residence time
1n this chamber is approximately 2 seconds, while the temperature is typically
1400°F. The total residence time for gases in both chambers is about 4.5
seconds. The scrubber at the ENSCO facility 1s also of the Venturl jet type.
Gases evolved from the combustion chamber are quenched with lime and caustic
solution as they enter the bottom of the scrubber. This treatment removes
particulates and neutralizes acidic gases such as HC1. Spent scrubber liquor
-------
4
1s transferred to a holding tank which then overflows to a sludge lagoon.
Spent scrubber solution is available for recycling. After neutralization,
the combustion gases pass through the Venturi section of the scrubber, in
which recycled scrubber water is sprayed to facilitate particulate removal.
Ultimately, sludge and liquids from the lagoon are discarded by deep-well
injection.
B. Incineration Tests Conducted at Rollins and ENSCO.
Sampling and monitoring were accomplished by TRW during three different
modes of operation at each of the Incinerators tested. The materials
combusted and the conditions for each of these tests are as follows:
Test 1. For Test 1, each incinerator was operated under the
conditions at which incineration 1s normally conducted. Temperatures 1n
this case were not necessarily maintained at 1200*0 as required for PCB
Incineration. At the Rollins plant, the materials combusted were liquid
chlorinated hydrocarbon wastes, such as vinyl chloride still bottoms. At
the ENSCO plant, shredded capacitors (.solids, but not containing PCBs) were
Incinerated, along with chlorinated hydrocarbon wastes, such as pesticide
process wastes, as well as paint and ink manufacturing wastes.
2. Test 2. For Test 2, the materials combusted at the Rollins
Incinerator, consisted of the same types of materials as described for Test 1,
along with liquid PCB wastes. Similarly, at ENSCO, the Test 2 incinerator
-------
V*i
feed included the normally processed wastes (as in Test 1) along with both
liquid PCBs and shredded PCB-containing capacitors. Combustion temperatures
for Test 2 were maintained near 1200°C at both plants, as required by EPA
regulations for incineration of PCBs.
Test 3. For Test 3, the materials combusted at Rollins consisted
of a mixture of liquid PCB wastes and clean fuel oil. At the ENSCO plant,
the materials incinerated during Test 3 included both liquid PCB wastes and
PC8-conta1ning capacitors, mixed with diesel fuel. Again, temperatures were
maintained at the levels required for PCS incineration.
C. Sampling Procedures and Types of Samples Collected.
The sampling procedures and apparatus utilized in the monitoring tests
at the Rollins and ENSCO incinerators have been described in detail 1n
documentation prepared by TRW,1 and only those aspects which relate to the
samples ultimately analyzed by Wright State will be briefly summarized here.
1. Stack Effluent Samples. Stack effluents were collected by using a
modified EPA Method 5 sampling train, which consisted of a glass probe
(capable of being heated to 250°F), followed by a series of three glass
1mp1ngers (the first containing 100 ml. of distilled water, the second
containing 200 ml. of distilled water, and the third being empty), a
sorbent trap containing XAO-2 resin, a sorbent trap containing Florlsll, and
a fourth impinger containing silica gel. The train was backed by appropriate
gauges, valves, meters, and a pump, for the purpose of controlling and
-------
a. j'
monitoring flow through the train. No filter was inserted between the probe
and Impingers, so any particulates collected entered the impingers and were
immersed in the liquids. The impingers were cooled by immersing them in an
ice bath for the purpose of condensing moisture from the stack effluent
stream. The sorbent traps were also cooled by circulating chilled water
through the exterior jackets. The sorbent traps incorporated a glass fritted
disc on one end, which retained and supported the sorbent bed, and the sorbent
was held in place by a glass wool plug at the other end of the trap. The
traps, as well as the probe and the impingers, were interconnected using
standard glass ball and socket joints, which were not greased for these tests.
All components of the sample train were rigorously cleaned, using standard
procedures, prior to use.
Upon termination of a stack sampling event, the probe was removed from
the stack and the train was disassembled. The nozzle and glass probe were
rinsed with approximately 100 ml. of acetone and 100 ml. of hexane, and the
interior surfaces were scrubbed with a nylon brush. The rinses and any
materials removed from the probe were transferred to a pre-rinsed amber glass
bottle and the bottle was sealed with a teflon-lined cap. The contents of the
first three impingers were pooled by pouring the liquids into a single amber
glass bottle. The Interior chamber of each Impinger was rinsed successively
with 30 ml. of acetone and 30 ml. of hexane and these rinse solutions were
placed in the same bottle containing the collected impinger liquids. Rinsings
of connector tubing were also placed 1n this container, which again, was
sealed with a teflon-lined I1d. The XAD-2 and Florisil sorbent tubes were
-------
7
removed and sealed by placing mating ground glass caps on the end joints of
each tube. Appropriate solvent and water blank samples were also collected
during the tests and placed in sealed amber glass bottles. Several method
blank samples were also obtained by charging and assembling a complete train,
then dissassembling the train components (unused) and collecting the
appropriate samples.
From the foregoing, it is seen that five types of samples requiring
analyses for PCDDs, PCDFs and PCBs result from the stack sampling. These
types of samples are:
*	Probe Wash
*	Impinger catch and rinses (composite)
*	XAD-2 resin traps
*	Florisil traps
*	Solvent and method blanks
2. Solids and Liquid Samples. Concurrent with the stack sampling,
samples of appropriate solids and liquids relevant to the incineration process
(feed materials, combustion products, emission control equipment samples)
were also collected by stmpite grab, sampling. Such samples were collected at
appropriate intervals during the incineration tests,1 and were placed in
prerinsed amber glass bottles and sealed with teflon-lined caps. These
samples are listed below.
-------
4 3.7
9
a.	Rollins. Only liquid process stream samples were collected at the
Rollins facility. These included:
*	Liquid PCB and waste feed.
*	We11-water feed
*	Lime slurry feed
*	Touch-up tank
*	Effluent discharge
b.	ENSCO. Both solid and liquid process stream samples were collected
at the ENSCO facility. These included:
*	Shredded capacitors
*	K1ln ash
*	Liquid PCB feed
*	Liquid waste feed
*	Well water
*	Scrubber liquor
*	Recycled scrubber liquor
*	Caustic solution
*	Lime slurry
-------
9
D. Shipment of Samples to Wright State.
Sampling activities at the Rollins and ENSCO facilities were initiated
by TRW in early August* 1980, and shipment of samples to Wright State
University for analyses began on August 5, 1980. Wright State continued to
receive shipments through September 17, 1980. A complete listing of all
samples received in the nine separate shipments, including dates of arrival
at Wright State, and information on sample condition, was provided in writing
to Mr. Michael Dellarco, EPA Dioxln Monitoring Program Coordinator, on
September 26, 1980. Photographs of some of the samples and resin traps were
taken by Wright State and these were also provided to Mr. Michael Dellarco
in a communication on Oecember 15, 1980. As discussed in these communications,
while most of the samples shipped to Wright State were received in good
condition, several of the sample bottles containing liquids leaked, several
other sample vessels were broken resulting in loss of the sample, and a few
samples were lost entirely in the course of shipment and did not arrive at
Wright State. In some of the shipments during which breakage occurred during
transit, the shipper had been forced to repackage some of the sample vessels
since the original shipping box was disintegrated or destroyed. In the latter
cases, chain of custody records were lost and sample integrity could not be
ensured. In spite of these proBlems, since duplicate sets of train samples
were collected in all of the tests (dual trains were utilized), an essentially
complete set of samples in reasonably good condition was available from each
test for analysis.
-------
10
III. ANALYTICAL METHODOLOGY
A. Sample Handling and Preparation.
As noted already, some of the liquid samples from the sampling trains
Cprobe rinses and impinger samples) contained two distinct layers, and in one
case suspended solid material was visible. After allowing such samples to
stand for a period sufficient to effect complete separation of the layers,
each layer of these samples was removed to a separate amber glass sample
container fitted with a teflon-lined lid, so that each phase could be analyzed
separately. The quantities of the liquid phases varied considerably from sample
to sample, and 1n a few instances, only one layer was present. It is
not known whether these variations are due entirely to leakage of the sample
containers Cwhich did occur 1n many samples) or to variations in the sampling
train operation and rinsing procedures followed by TRW. In the one sample
where particulate could be seen, the particulate was removed separately by
filtration. In subsequent attempts to remove and and extract this particulate,
the latter was lost owing to the extremely small quantity available and the
small particle size. The separated liquid layers were subsequently processed
as described below and analyzed.
The XAO-2 resin and the Florist! traps were removed from the sampling
trains by TRW, the ends were sealed with glass caps, and the traps were shipped
Intact to the Brehm Laboratory. Prior to analysis of a given trap at the
Brehm Laboratory, the XAD-2 resin or Florisil was removed and transferred to
a clean amber glass vessel fitted with a teflon-lined cap. Aliquots of the
sample were subsequently removed for processing as described below. The
-------
11
total sample was manually mixed prior to taking an aliquot, in an effort to
ensure homogeneity and representative sampling. In all cases, both the XAD-2
resin and Florisil appeared to contain substantial quantities of water (although
the quantity varied from. sampTe- to sample) and the solid sorbents were agglomerated
and difficult to remove from the traps. In fact, some quantities of the sorbents
adhered tenaciously to the inner surfaces of the traps and could not be removed.
The quantities remaining could not be determined and related to the original
weight of resin because the weights of the individual traps prior to packing,
and the weight of the water absorbed by the solid sorbent in each trap could
not be reliably determined. Consequently, there is an uncertainty in the
reported analytical results arising from the fact that these data do not take
into account the residues of the sorbents left in the traps. The quantities
of these residues were generally small however. No attempt was made to dry
the XAD-2 or Florisil samples prior to analysis, since it was thought that
this might also lead to loss of volatile chemical compounds trapped on these
sorbents. The total weight of the sorbent removed from the trap was determined
prior to analysis. Aliquots of the sorbents taken for analysis were also
weighed, of course, and by relating these aliquots to the total quantity of
sorbent removed, the total burden within the trap of the compounds analyzed
could be determined.
Another factor which bears on the accuracy of the reported data for the
XAD-2 resin and Florisil traps should also be mentioned here. It was observed
that the glass frits of the traps received by the Brehm Laboratory had been
perforated, so that they exhibited large non-uniform holes of varying sizes.
Apparently, according to information received from TRW, holes were punched
-------
Hi)
12
during the actual sampling tests because the glass frits plugged and the
desired flow rate through the sampling train could not otherwise have been
achieved. These observations raise questions about the uniformity of exposure
of the solid sorbents to the gas stream during the sampling process. The
colorations observed in various parts of the traps also raise similar
questions. Several of the traps from the tests at the ENSCO plant exhibited
a pronounced red-brown coloration on the frit and on the glass wool plugs
at the other end of the solid sorbent bed, but the sorbent itself appeared
to have retained its characteristic stark white color. This was observed
for both the XAD-2 resin and the Florisil traps. Similarly, many of the
Rollins traps exhibited a bright yellow color on the frits and the glass wool
plugs at the other end of the sorbent bed, but the sorbents themselves appeared
to be stark white with no indication of the yellow color. In the case of traps
originating from Test 3, at the ENSCO site, the frit was covered with an oily
residue Cgray in color) but again the resin appeared to be "clean". Again,
these observations suggest that the sorbents may not have been uniformly
exposed to the sampled stack gas stream. If this is true, the analytical data
obtained for the sorbent traps represent, at best, lower limits in terms of the
concentrations of CDDs, CDFs and PCBs which were present in the stack effluents.
Photographs of the sorbent traps from these tests which document the
observations mentioned above, were provided earlier to EPA by- the Brehm
Laboratory 1n our communication of Dec. 15, 198Q.
As noted above, small quantities of the solid sorbents could not He
removed from the traps. Also, the colored contaminants just mentioned which
appeared on the glass frits of the traps as well as other solid deposits on
trie trap walls renalned in the traps after the Bulk of the sorBents had been
-------
430-
13
removed. Since it is important to know whether or not these trap residues
may include CDDs, CDFs, or PCBs, attempts were made to extract these residues
from one trap by Immersing the entire trap (.after the sorbent had been
removedl in a very large Soxhlet apparatus and employing the extraction procedures
described below. The extract was su&sequently analyzed to determine content
of the compounds of Interest.
Other solid samples (.feed materials, etc.) were analyzed as received.
AHquots. of these were removed from the sample vessels in which the samples
were received, and weighed. Again, attempts were made to ensure that a
reasonably representative sample was selected but with the course solid feed
materials (.for example, shredded capacitors} and the viscous scrubber samples
there is no certainty that this was achieved.
B. Analytical Methods Development and Testing.
At the outset of this program, analytical methods to accomplish
quantitative measurements of the CDDS, CDFs and PCBs in samples of the type
collected in this assessment had not been demonstrated. In fact, a complete
survey of combustion effluents for all classes of CDDs and CDFs had not been
reported by any workers in this country. In previously reported studies,z
Wright State had obtained quantitative data on CDDs, including some Information
on Isomeric composition of the TCDDs, for effluents from a municipal refuse
incinerator. Dow Chemical scientists had also reported the results of
comprehensive studies of CDDs in various combustion products.3.1* The samples
-------
4^3
14
collected in the Roll ins/ENSCO tests however, encompass a much broader array
of sample types than have previously been examined in such surveys. Accordingly,
it was necessary to develop and test the efficacy of analytical procedures
which could be applied for the required analyses.
The approach taken for developing the needed analytical procedures by our
laboratory entailed initial application of methodology similar to that which
we have applied previously for combustion products and other hazardous waste
samples containing CDDs. In order to gauge the efficacy of these procedures
portions of samples were spiked with labelled 37Cllf^2,3,7,8-TCDD, and other
labelled CDDs, and the samples were then analyzed to determine the recovery
of these CDDs. If recoveries were not acceptable (.usually a minimum recovery
of 5G55 of the added labelled Internal standard was used as a criterion of
acceptability) then the extraction and/or cleanup procedures were modified
and another spiked sample aliquot was analyzed using the modified methods.
This procedure was repeated until recovery of the added CDD was acceptable.
It was not possible to accomplish the same type of tests for the CDFs because
no labelled CDF was available or could be obtained for used as an internal
standard In the studies reported here. A limited number of experiments were
conducted with CDFs using the method of standard additions, in which a given
sample was analyzed before and after addition of a known quantity of an
unlabelled CDF standard. The methods ultimately developed and validated for
the determinations reported here are described in the following sections.
The extent to which information on the CDD and CDF isomers present in
the samples could be obtained was limited by the availability of pure standards
-------
V 3V
15
of the Isomers which could be used for comparisons of GC retention times and
mass spectral response. Of the total 75 CDD and 132 CDF isomers which are
possible only 33 CDOs and 10 CDFs are available 1n our laboratory. A listing
of these is given in Tables 1 and 2. In addition, it 1s seen that four other
1sotopically labelled COOs are on hand.
C. Extraction of Samples and Preliminary Separation of CDDs, CDFs and PCBs
From Other Matrix Constituents.
The procedures developed and applied for extraction of the samples and
preliminary separation of the compounds of interest are as follow*.
1» Extraction.
a.	Liquid samples. Measure the total volume of the liquid sample.
Transfer approximately one-half of the total sample (or up to 40 ml)
to a precleaned 125 ml amber glass bottle. Add appropriate quantities
of the internal standards, 37CU-2,3,7,3-TCDD and 37C1i*-l,2,3,4,6,7,8-HpCDD.
Add 40 ml of petroleum ether, seal the vessel tightly with a teflon-
lined cap, and agitate the vessel vigorously on a laboratory shaker
for a period of one hour. Proceed with step 2.
b.	Solid samples. Accurately weigh an aliquot of the XAD-2 or
F1oris11 sorbent (or other solid sample) corresponding to approximately
one-half of the total sample (.typically 5-13 g), place this 1n a
precleaned glass thimble, and add appropriate quantities of the Internal
-------
V3*r
16
standards, 37C11^-2,3,7,8-TCDD and 37C"U-1,2,3,4,6,7,8-HpCDD, directly
to the sample in the thimble. Insert the thimble into a precleaned
Soxhlet apparatus, charge the Soxhlet reservoir with 100 ml of benzene,
and apply heat to the reservoir to extract the sample. Continue
extraction for a period of 16 hours. Remove the extract and
concentrate to a volume of about 5 ml using a Snyder column. Transfer
the concentrate to a precleaned 125 ml glass bottle and add 40 ml
of petroleum ether. Proceed with step 2.
2. Clean-up and liquid chromatographic separation.
a.	Add 50 ml of doubly distilled water to the vessel containing the
sample extract, reseal the vessel and agitate for 10 minutes. Allow
the vessel to stand for a period sufficient for the aqueous and
organic layers to separate completely, and remove and discard the
aqueous layer.
b.	Using the same procedure as applied in 2a., wash the extract
successively with 5Q ml portions of 202 KQH, doubly distilled water,
concentrated HjSOit Cexcept in this case agitate mixture for 15 min.),
and doubly distilled water, in each case discarding the washing
agent. The acid washing procedure is repeated until the acid layer
is visually colorless.
c.	Add 5 g of anhydrous sodium sulfate to the washed extract and
allow to stand in order to remove residual water. Transfer the
extract to a centrifuge tuBe and concentrate to near dryness
-------
^ 3£
17
fay placing the tube in a water bath at 55°C, and passing a gentle
stream of filtered, prepurified Nj over the solution.
d.	Prepare a glass macro-column, 20 ran OD x 230 mm in length,
tapered to 6 mm OD on one end. Pack the column with a plug of
silanized glass wool, followed successively by 1.0 g silica,
2.0 g silica containing 33% (w/wl 1M NaOH, 1.0 g silica, 4.0 g
silica containing 44% (w/w) concentrated H2S0i* and 2.0 g silica.
Quantitatively transfer the concentrated extract from Step c. to
the column and elute with 45 ml hexane. Collect the entire
eluent and concentrate to a volume of 1-2 ml in a centrifuge
tube, as before.
e.	Construct a disposable liquid chromatography mini-column by
cutting off a Pyrex 10 ml disposable pipette at the 2.0 ml mark
and packing the lower portion of the tube with a small plug
of silanized glass wool, followed by one gram of Moelm basic
alumina, which has been previously activated for at least 16 hours
at 600°C 1n a muffle furnace, and cooled in a dessicator for 30
minutes just prior to use. Quantitatively transfer the concentrate
from Step d. onto the liquid chromatography column, rinse the
centrifuge tube consecutively with two 0.3 ml portions of 3%
CHjCla -in-hexane, and also transfer the rinses to the chromatography
column.
-------
18
f. El lite the column with 10 ml of 3% (v/v) CH2C12-1n-hexane and
retain the eluent for PCB analysis.
9- Elute the column with 10 ml of 50% (.v/v) ^Cli-in-hexane
and retain the eluent for analyses for CDDs and CDFs.
h.	Elute the column'with 5 ml CH2C12 and retain the eluent
to check for retention of CDDs and CDFs on the column.
i.	Concentrate each of the retained fractions to a volume of
approximately 1 ml by heating the tubes 1n a water bath while
passing a stream of prepurified N2 over the solutions, as
described above. Quantitatively transfer the concentrated
fractions into separate 2 ml micro-reaction vessels, splitting
eacir fraction so that one-half goes Into each of two sample
vessels. (The contents of one vessel are analyzed for CDFs,
and the contents of the other for CDDs}. Evaporate the
solutions in each of the micro-reaction vessels almost to
dryness, using the procedures just mentioned, rinse the walls of
each vessel down with Q.5 ml CH2C12, and reconcentrate just to
dryness.
j- Approximately 1 hour before gas chromatographic-mass spectrometric
(JSC-MS) analysis, dilute the residue 1n each micro-reaction vessel
with an appropriate quantity of benzene (.depending upon the
anticipated quantities of analytes in each vessel 1 and gently swirl
-------
19
the solvent in the vessel to ensure dissolution of CDDs, CDFs and
PCBs.
k. If preliminary GC-MS screening analysis of the sample indicates
the presence of potential interfering compounds (that is, which
obscure the CDD, CDF or PCS signals) or other sample matrix
constituents which elute from the GC at very long times, then
additional sample clean-up or fractionation is required using high
performance liquid chromatography (HPLC). The HPLC used for this
purpose is described below. The sample is injected into the HPLC
and appropriate fractions are collected, as determined in advance,
by injecting pure COD and CDF standards and measuring the retention
times of these.
D. Analysis of Sample Extracts Using Gas Chromatography-Mass Spectrometry fsc-aro
Two separate gas chromatographic-mass spectrometric methods of analysis
were employed in this program, one utilizing low resolution gas chromatography,
hlgh resolution mass spectrometry CLKGC-HRMSl, and the other utilizing high
resolution gas chromatography-low resolution mass spectrometry (HRGC-LRMS).
For determination of total TCODs in the extracts of the samples LR6C-HRMS
employed. This technique yields essentially unequivocal quantitative data on
the total TCODs present in the analyte, (.assuming that the instrument response
is the same for all TCDD Isomers 1 but does not yield information regarding
the quantities of specific TCDD isomers which are present. For these analysts,
a modified MS-30 mass spectrometer, operated 1n selected-ion monitoring mode,
is used. This technique utilizes a specialized step-scan circuit and
-------
20
associated electronic hardware developed by our laboratory. Both m/z 319.8966
and 321.8936 are monitored as indicators of TCDD during the period when
TCDDs elute from the gas chromatograph and thus, quantitation of the TCODs
detected can be based upon the signal observed at either mass. The
theoretical ratio of nr/z 32Q:m/z 322 resulting from TCDD (based on the known
isotonic abundances of 35C1 and 37C1 and the numbers of CI substituents in
the molecular ion) is 0.27, and the experimentally observed ratio should be
essentially the same as the theoretical ratio. This is another criterion
which the data should satisfy in order to certify with confidence that TCDD
is Indeed detected. Since the ion signal at m/z 327.8846, which arises from
the 37Cllf-2,3,7,8-TCDD Internal standard added to all samples prior to
processing, is also monitored concurrently with the two masses typical of
native TCDDs, and the quantitation of native TCDDs. is actually based upon
the ratios of the signals at m/z 32Q and m/z 322 to that at m/z 328, the TCDD
data obtained are inherently "recovery corrected". This is, of course, one
of the chief reasons for utilizing an internal standard, and results in
Improved accuracy. The percent recovery of the internal standard is
specified 1n the data listings solely for the purpose of Illustrating the
overall efficiency of the analytical procedure.
For determining the total concentrations of each of the various other
classes .(pmTtachlorinated through octachlorinated) of chlorinated dibenzo-p-
dioxlns and dibenzofurans 1n the sample extracts,a sophisticated HRGC-LRMS
technique was employed. A complex computer-controlled selected, ion monitoring
scheme was utilized for this purpose and the compounds 1n each extract were
quantitated during two separate GC-MS runs. The ions monitored during these
-------
21
GC-MS analyses for each group of CDDs and CDFs are listed 1n Table 3, while
the sequence of GC and MS operations are shown in Tables 4 and 5 for the two
runs. These procedures permit the monitoring of 1on masses characteristic
of each class of CDDs and CDFs during the appropriate gas chromatographic
retention time interval. As expected, the monochlorinated CDDs and mono-
chlorinated CDFs have similar retention times. However, use of a 50 M (WCOT)
fused silica gas chromatographic column provides optimum separation and
minimizes overlap of individual compounds. As mentioned earlier, of the 75
possible CDDs and 135 possible CDFs, authentic standards of only a limited
number are available 1n our laboratory for use 1n calibration (.see Tables
1 and 2). Therefore, most of the CDD and CDF peaks observed 1n the analyses
of the sample extracts cannot be assigned to a specific isomer. Thus, 1n
arriving at a quantitative value for a given class of CDDs or CDFs, the areas
of the mass chromatographic peaks appearing at the appropriate retention times,
and having the appropriate mass spectral response, were summed and compared
to the corresponding area observed from injection of a known quantity of a
calibration standard of the same class. In general, a single CDD or CDF
isomer of each chlorinated group (.1.8., monochlorfnated, d1 chlorinated, etc.)
was used 1n the calibration process. However, since all 22 TCDD Isomers are
available in our lab, 1t 1s possible to make a reasonably definitive Identi-
fication of most of the TCDD Isomers present 1n the samples, because the
techniques employed permit complete separation of many of these isomers.
Sample extracts expected to contain PC8s were also subjected to HRGC-LRMS
analysis, using selected 1on monitoring. PC8 standards representative of several
classes of PC8s Gnonchlorinated through decachlor1nated) were used to calibrate
the GC-MS and to determine appropriate GC retention time parameters and sensitivities
-------
22
for each PCB class. The mass spectral 1ons indicative of each PC8 class which
were monitored at the appropriate retention time are shown in Table 6. The same
monitoring sequence was applied to the sample extracts in .order to-estimate the
quantities of total PC8s of each class which were present.
Specific details of the apparatus, operating conditions and experimental
parameters for the GC-MS analyses reported herein are given 1n the following*
1. Parameters For Low Resolution Gas Chromatoqraphic-H1qh Resolution Mass
Spectrometrlc (LRGC-HRMS) Analysis of Sample Extracts.
a.	Instrumentation: Varlan 3740 Gas Chromatograph coupled through
an AE! silicon membrane separator to a modified AEI MS-30 Mass Spectrometer.
Modifications to the MS-30 Include a new ESA power supply and
Incorporation of a custom built, step-scan circuit which is driven by a
Nlcolet 1074 Signal Averaging Computer- Four masses are rapidly scanned
at the retention time of the dloxln or furan of interst.
b.	Conditions for the Gas Chromatograph:
Column:	1.8m x 2mm 1.0. glass column packed with 1.52 0V-101
on Gas Chrom Q 000/120 mesh)
Carrier Gas: Helium at a flow rate of 30 ml/m1n
Temperature: Injector: 2SQ°C
Column: 220°C
Transfer Lines: 285CC
c.	Conditions for the Mass Spectrometer:
Ionizing Voltage:
Static Resolution:
70 eV
1:12 , 500 (10X Valleyl
-------

23
Source Envelope Pressure:	5 x 10"5 torr
Analyzer Pressure:	5 x 10"® torr
Source Temperature:	250°C
Membrane Separator Temperatures: 215°C
Transfer Line Temperature:	270°C
Ions Monitored: m/z 319.8966, 321.8936, 325.8805, 327.8846
2. Parameters Far tffgtr ftesoTatlon: Gas Cftromatoqraphic-low Resolution
MSagSbectrometric (HRGC-LRM3) Analysis of Sample Extracts.
a. Instrumentation: Perkin Elmer Sigma III Gas Chromatograph Coupled
through a custom-fabricated interface Including
a single-stage glass jet separator to a Kratos
MS-25 Mass Spectrometer equipped with a DS-50SM
Data System.
t>. Conditions For the Gas Chromatograph:
Column:	5QM WCOT C0V-101) Silica Capillary
Carrier Gas: Hydrogen, 30 lb head pressure
Column Temp.: Programmed from 190°C to 220°C @
5°C/min, hold at 220°C for 20 min.
Interface Temp: 25Q°C
Injector: Split ratio 50:1
-------
24
c. Conditions for the Mass Spectrometer
Selected Ion Monitoring Mode (for m/z's monitored and details
of instrumental procedures see Tables 1-3)
Ionizing Voltage: 70 eV
Accelerating Voltage: 4 KV
E. High Performance Liquid Chromatography CHPLC).
The apparatus and instrumental parameters applied for HPLC fractionation
of sample extracts are specified below.
1. Parameters For High Performance Liquid Chromatographic (HPLC)
Fractionation Of Sample Extracts.
a.	Instrumentation: Varian Model 5021 Microprocessor Controlled High
Performance Chromatograph equipped with CDS-111L
Data System
b.	Parameters:
Pressure: Minimum: 10 atm
Maximum: 250 atm
Injection Loop: 25 yl
Column: Guard: 37u Vydac SC Reverse Phase
4.0 cm x 0.4 cm 1.0.
Analytical: 2-Dupont Zorbax-ODS
25.0 cm x 0.6 an 1.0.
-------
25
Temperature: Guard Column: Ambient
Analytical Column: 50°C
Detector: Fixed UV: 254 nm, 0.01 A.U.F.S.
Variachrom UV-Vis: TCDD, 235 nm, 0.01 A.U.F.S.
HxCDD, HpCDD, 245 nm, 0.Q1
A.U.F.S.
Program: Time
Code
Value
.a
Of
/o
100 Methanol
.a
Flow
2.5 ml/min
.0
Event
Hold
.1
Event
Inject
2Q.0
Event
Reset
F. Reagents and Chemicals.
The following reagents and chemicals were utilized in the procedures
outlined above. Potassium hydroxide, anhydrous sodium sulfate, and sulfuric
acid were all Reagent Grade and were obtained from J.T. Baker-Chemical Co. or Ftsh«r
Scientific Co., Fairlawn, N.J. Methanol, hexane, methylene chloride and
benzene were "Distilled in Glass" quality obtained from Burdlck and Jackson,
Muskegon, Michigan. Petroleum ether (low boiling) was Omnisolve Quality from
Matheson, Coleman, and Bell, Cincinnati, Ohio. Woelm basic alumina (Activity
Grade I) was obtained from ICN Pharmaceuticals, Cleveland, Ohio. Doubly
distilled water was obtained using the all-glass distillation apparatus in
the Brehm Laboratory. Prepurified nitrogen was obtained from Airco, Inc.
Montvale, New Jersey.
-------

26
Standards employed in this work were obtained from the sources listed
below.
1-chlorodibenzo-p-dioxin-Analabs,	Inc., North Haven, Conneticut
2-chlorodibenzo-p-d1oxin-Analabs,	Inc., North Haven, Conneticut
2.7-dichTorodibenzo-p-dioxin-Analabs,	Inc., North Haven, Conneticut
2.3-dichlorodibenzo-p-dioxin-Analabs,	Inc., North Haven, Conneticut
1,2,4-trichlorodibenzo-p-dioxin-Anala5s, Inc., North Haven, Conneticut
1,2,3,4-tetrachlorodibenzo-p-diox1n-Analabs, Inc., North Haven,
Conneticut
37C1 (*-2,3,7,8-TCDD-KOR Isotopes, Cambridge, Mass.
2,3,7,8-TCDD and other TCDD isomers-Oow Chemical Co., Midland, Michigan,
and H.R. Buser, Swiss Federal Research Station, Wadenswil, Switzerland.
1,2,3,7,8-PCDD-KOR Isotopes, Cambridge, Mass.
1.2.3.4.7.8-HxCDO	- Or. A. Poland, Univ. Rochester, N.Y.
1.2.4.6.7.9-HxCDD-Or.	A. Poland, Univ. Rochester, N.Y.
1,2,3,4,6,7,8-HpCDD-KOR Isotopes, Cambridge, Mass.
OCDD-Analabs, Inc. North Haven, Conneticut
37C1g-OCDD-KOR Isotopes, Cambridge, Mass.
2.4-dichlorodibenzofuran	- All CDFs obtained from U.S. FDA, Wash., D.C.
2.8-di	ch1orodi benzofuran
1,2,4-trichlorodibenzofuran
1,2,4,8-tetrachlorodibenzofuran
2,3,7,8-tetrach1orodibenzofuran
2,3,6,8-tetrachIorod1benzofuran
1,2,4,7,8-pentachlorodibenzofuran
-------
27
1,2,4,6,7,9-hexachlorodlbenzofuran
1.2,3,4,6,3,9-heptachlorodlbenzofuran
Octachlorodlbenzofuran
6. Quality Assurance.
As Is the case for all analytical programs conducted at our laboratory,
all analytical measurements were accomplished 1n accordance with good
laboratory practice. An extensive quality assurance program 1s established
and followed for all projects such as those described herein. This program
Includes the analyses of solvent and method blanks, analyses of Internally
spiked control samples, determinations to validate the efficacy of the analytical
procedures applied, as already described. Quality Control for the Individual
analyses reported here was also provided on a continuing basis for COO
determinations because of the incorporation of known quantities of labelled
TCOO and HpCOO Internal standards Into each sample prior to analysis.
IV. RESULTS
A. Total TCDOs and TCSFs
Results obtained from the determinations of total TCDOs and TCOFs 1n the
stack effluents (that 1s, the Individual sampling train samples) from the testa
at ENSCO are reported 1n Tables 7 and 8, respectively. Corresponding data for
the tests at Rollins are reported 1n Tafiles 9 and 10. The data for total TCDOs
and TCOFs are reported separately because these results were considered by EPA
-------
<-H7
28
to have the greatest relevance for the estimation of health risks 1n connection
with the Incineration tests, and were the first results obtained by the Breton
Laboratory 1n this study. As mentioned earlier, the total TCDO results were
obtained by LRGC-HRMS, while the total TCOF data were obtained by HRGC-LRMS.
As noted 1n the previous section of this report, the results obtained for
total TCODs do not permit any definitive conclusions with respect to the
composition of TCOO positional isomers. Such information was obtained however,
for both TCDDs and TCDFs from the high resolution (capillary-column) GC-LRMS
measurements. These results are discussed below. The data for total TCDDs
shown 1n Tables 7 and 8 Indicate that these compounds were detected 1n the
samples from Tests 1 and 2, but not In Test 3 samples, at both Rollins and
ENSCO. One possible conclusion which can be drawn from these observations 1s
that the "normal" chlorocarbon wastes Incinerated by these plants are the
source of the TCDDs (since these were burned 1n both Tests 1 and 2), whereas
Incineration of the PCBs (which were the only chlorocarbons burned in Test 3)
does not yield TCDOs. There Is no obvious rationale for the curious pattern
of the TCDD results (.for example, the observation of TCDDs 1n the impinger,
probe wash, and XAD-2 resin samples from one test, whereas only the probe
wash sample or the Impinger sample contained TCDDs 1n other tests). Possibly,
this reflects erratic sampling procedures Cas discussed earlier) or variations
1n the washing of the train components by TRW. Alternatively, temperature
¦fluctuations 1n various sections of the train might account for some of these
observations. Similar peculiarities are apparent 1n the pattern observed for
the total TCOF data (Tables 9 and 101.
Data for both the TCDDs and TCDFs are reported as total quantities present
1n each type of sample analyzed. Since the volumes and weights of the various
-------
29
samples varied considerably, and since these could not be related to original
volumes and weights (these were not provided to us, and 1n any case, leakage
and/or loss of unknown quantities of most samples occurred during shipment)
1t was not possible to calculate concentrations of TCDDs and TCOFs present 1n
the samples from the analytical data determined by our laboratory. Also, the
quantities of TCDOs and TCOFs determined could not be related to the volume
of air sampled, since this information was unknown to us.
B. Isomeric Composition of TCDDs and TCDFs
Data bearing on the isomeric composition of the TCDDs and TCDFs present
1n representative Rollins and ENSCO samples were also obtained 1n the present
study from the HRGC-LRMS measurements. The latter utilize a 50-meter capillary
GC column which Is capable of resolving many of the tetrachlorinated isomers.
For example, Figure T shows selected-1on mass chromatograms obtained for a
mixture of 12 TCDD isomers by monitoring sequentially nominal m/z 257 and 259
Otfte sum of these is displayed); nominal m/z 320 and 322 (sum of these 1s
displayed); and nominal m/z 328 in one GC-MS run. Corresponding displays
obtained from the HRGC-LRMS analyses of TRW sample number 6-0606 (Rollins Test 2,
Train 1 Probe Wash sample) and sample number 6-0560 (ENSCO Test 2, Train 1
Impinger sample) are shown in Figures 2 and 3 (the numbering of peaks
corresponds to that shown 1n Figure 1). By comparing the observed sequence of
peaks 1n these runs with that observed for the standard Isomer mixture (Figure T),
5
and by utilizing related GC data reported by Buser and Rappe it 1s possible to
positively Identify four of the TCDD Isomer peaks detected 1n the analyses of
these two samples. It is known from the results cited above that each of
these four Isomers 1s completely resolved from the other 21 TCDD isomers, and
-------
30
so Identification of these Isomers 1s unequivocal using the experimental conditions
employed here. These Isomers and their relative quantities (percents of the total
TCDDs) are listed 1n Table 11 for the two samples Indicated. Several of the
other isomer peaks shown 1n Figures 2 and 3 also correspond in retention time
to specific Isomers included 1n our 12-isomer mixture, but other data obtained
e
both In our lab and elsewhere indicates that these isomers are not uniquely
resolved from the other TCDD Isomers, and so the identifications of these other
peaks ^indicated. 1iv Table 11) are tentative. Under the presently utilized
experimental conditions, the 2,3,7,8-TCDD Isomer 1s also resolved from all other
TCDD Isomers, and 1t is therefore possible to state that, within the limits
of detection applicable here (50 pptl 2,3,7,8-TCDD is not detected in the
effluent samples analyzed. It seems probable that the isomeric composition of
the TCDDs 1n the other effluent samples would be essentially the same as those
detected for the two samples just discussed.
Less definitive data bearing on the Isomeric composition of the TCDFs could
be obtained because only two TCDF Isomer standards were available for cali-
bration purposes 0.2,4,8-TCDF and 2,3,7,8-TCDF). Thus, while a TCDF peak 1s
observed at the retention time corresponding to that of 2,3,7,8-TCDF 1n the
mass chromatograms of many of the samples, completely unequivocal assignment
of this peak 1s not possible. The quantities of the TCDF component which 1s
"apparently" 2,3,7,8-TCDF are listed In Tables 9 and 10, along with the data
for total TCDFs. Again, 1t must be emphasized that this assignment 1s tentative.
-------
V So
31
C. Total Higher Chlorinated CDDs and CDFs
Using the complex sequence of HRGC-IRMS procedures described 1n an earlier
section, the complete series of higher chlorinated CDDs and CDFs (total penta-
through octachlorinated) were determined 1n representative Rollins and ENSCO
samples. Typical mass chromatograms obtained in the course of these complete
GC-MS scans resulting from two injections of a mixture of such CDD and CDF
standards are shown in Figures 4 through 13 The CDD and CDF Isomers used in
obtaining this calibration data were l-chlorodibenzo-p-d1oxin; 2,7-d1chloro-
d1benzo-p-d1oxin; 1,2,4-trichlorod1benzo-p-diox1n; 2,3,7,8-tetrachlorodl-
benzo-p-d1oxin; 37C1if-2,3,7,8-tetrachlorodibenzo-p-d1ox1n; 1,2,3,4,7,8-hexa-
chlorodibenzo-p-d1oxin; 1,2,3,4,6,7,8-heptachlorodibenzo-p-dloxln; octa-
chlorod1benzo-p-d1ox1n; 2,4-d1chlorod1benzofuran; 1,2,4-trichlorodlbenzofuranj
2,3,7,8-tetrachlorodibenzofuran; 1,2,4,8-tetrachlorodlbenzofuran; 1,2,4,7,8-
pentachlorodibenzofuran; 1,2,4,6,7,9-hexachlorodlbenzofuran; and octachloro-
dlbenzofuran. Corresponding mass chromatograms resulting from the full GC-MS
scan of two Injections of the Rollins Test 2, Train 1 Probe wash sample (TRW
Number 6-0606) for all classes of CDDs and CDFs are shown in Figures 14
through 22. The quantities of CDDs and CDFs 1n the Test 1, Train 1 Probe Mash
sample from Rollins (TRW Number 6-0502) and 1n the Test 2, Train 1 Impinger
sample from ENSCO (TRW Number 6-0560), as determined from similar HRGC-LRMS
data are listed 1n Tables 12 and 13, respectively. It appears that the
quantities of CDFs in these samples generally exceed the quantities of the
corresponding CDDs. The significance of these observations and their bearing
on the combustion mechanisms is not currently apparent, and 1s difficult to
determine 1n the absence of data on other chemical constituents 1n the effluents
and feed materials.
-------
L)S)
32
D. PCBs
In the limited time which was permitted for the analyses of samples
collected 1n this survey, 1t was possible to analyze only a single sample for
content of PCBs. The sample analyzed was a Flor1s1l trap sample from Rollins
Test 2, Train 1 (TRW Number 6-0609), and the results are shown fn Table 14.
The mass chromatograms obtained for a mixture of PC3 standards and for this
sample 1n the course of these analyses are shown 1n Figures 23 through 28.
Clearly, several classes of PCSs were detected 1n the sample (the estimated
quantities are given 1n Table 14). In the absence of Information on the
volumes of stack gas sampled 1n this particular test, we are unable to relate
this data to actual effluent concentrations.
V. SUMMARY
The analytical data obtained by our laboratory which are presented 1n
this report, when related to sampling data provided by TRW, should permit
EPA to make preliminary determinations of the mass emission rates of toxic
effluents (CCDs and CDFs) from the Incineration tests monitored at Rollins
Environmental Services and ENSCO. These data 1n turn should be applicable
for accomplishing health-risk assessments for the population exposed to the
effluents from these plants. Characterization by the Brehra Laboratory of
still other samples resulting from these tests, particularly feed materials
and control equipment samples, 1s still 1n progress, and such data will
hopefully provide much useful Information relevant to the mechanisms of
formation of toxic compounds such as the CDOs and COFs 1n combustion
tnvlronments.
-------
2-
33
REFERENCES
1.	D.R. Moore, R.W. Korner, W.F. Wright and D.G. Ackerman, "Plan for
Emissions Testing at Two Commercial Incinerators: Rollins Environmental
Services, Inc., Deer Park, Texas, and Energy Systems Company, El Dorado,
Arkansas", TRW Environmental Engineering Division, July, 1980, prepared
for Industrial Environmental Research Laboratory, U.S. Environmental
Protection Agency, Research Triangle Park, N.C. 27711, Contract No.
68-02-3174.
2.	T.O. Tlernan, M.L. Taylor, J.G. Solch, G.F. VanNess and J.H. Garrett,
"Determination of Polychlorinated Dibenzodioxins in Incinerator
Effluents", Report on U.S. Environmental Protection Agency Order No.
D2832 NAEX, prepared for U.S. Environmental Protection Agency,
Environmental Sciences Research Laboratory, Research Triangle Park,
N.C. 27711, June 18, 1980.
3.	R.R. Bumb, W.B. Crummett, S.S. Cutie, J.R. Gledhill, R.A. Hummel,
R.O. Kagel, L.L. Lamparski, E.V. Luoma, D.L. Miller, T.J. Nestrick,
L.A. Shadoff, R.H. Stehl, and J.S. Woods, "Trace Chemistries of Fire:
A Source of Chlorinated Dioxins" Science, 210, 385 (.1980).
4.	L.L. Lamparski and T.J. Nestrick, Anal. Chem. 52, 2045 (.1980).
5.	H.R. Buser and C. Rappe, Anal. Chem. 52, 2257 0980).
-------

TABLE 1
LIST OF CHLORODIOXIN ISOMER STANDARDS CURRENTLY AVAILABLE
AT THE BREHM LABORATORY-WRIGHT STATE UNIVERSITY
1-Chlorodibenzo-p-dioxin
2-Chlorodil3en20-p-dioxin
2,7-Dichlorodibenzo-p-dioxin
2,3-Dichlorodibenzo-p-dioxin
1,2,4-Trichlorodibenzo-p-dloxin
Te-trachlorodibenzo-p-dioxins - all 22 isomers
3 7Cltf-2,3,7, a-Tetrachlorodibenzo-p-dioxin
13Ci+ - 2,3,7, a-Tetrachlorodib«nzo-p-dioxin
1^C^-2,3,7,8-Tatrachlorodibanzo-p-dioxin
1,2,3,7,S-Pentachlorodlbenzo-p-dioxin
1,2,4,6,7,9-Haxachlorod±benzo-p-dioxin
1,2,3,4,7,8-H«xachlorodibenzo-p-dioxin
1,2,3,4,6,7-Haxachlorodibenzo-p-dioxin
1,2,3,4,6,7,3-H«ptachlorodibenzo-p-dioxin
37C1^-1,2,3,4,6,7,a-H«ptachlorodibenzo-p-d±oxin
Octachlorodibenzo-p-dloxin
37Clg-Octachlorodibanzo-p-diaxin
-------
TABLE 2
LIST OF CHLORINATED DIBENZOFUSAN STANDARDS CURRENTLY AVAILABLE
AT THE BREHM LABORATORY-WRIGHT STATE UNIVERSITY
2,4-DichXorodibeiizofuran
3,6-QichJLorodibenzofuran
2,8-Dichlorodibenzofuran
1,2,4-Trichlorodibenzo£uran
1,2,4,3-Tetrachlorodibanzo£uran
2,3,7,8-Ta-tracblorodibenzofuran
1,2,4,7, 3-Pentaciilorod±benzafuran
1,2,4,6,7,9-Hexachlorodibenzofuran
1,2,3,4,3,3,9-Heptachlorodibenzo£uran.
Oc-tachlorodibanzofuran
-------
36
tABL2 3
LIST OF ION MASSES MONITORED OSIMG GC-SSLECTSD-ION
MONITCHIN6 MASS SPECTROMETRY FOR SIMULTANEOUS DETERMINATION OF
MONO-, 01-, TRI-t TST3A-, P5NTA-, KE2CA-, HEPTA-, and OCTA-
CHLOSINATSD DIBENZO-o-OIOXINS AND DIBENZOFOBANS
dasa of
dlorlaaead
Oibanzedioxin
ar Olbanzefuran
Number of
Chlorina
Substituants
U)	
Monitorsd m/z for
Diiienzo furaia
Oi2H8 .jgOCly
Monitored at/z fcr
Dlbanzo-p-dioxina
Ci2H9_x02Clx 	
Approximate
Thaorattical Ratio
Sxpactad on 3asia
of Isotocie Ahundancs
Mono-	1	202.019 4	213.013 4	1,00
204.016	220.011	0.35
Di-	2	235.980 4	251.974 *	1.00
237.977	253.972	0.69
Xt±-	3	26S. 941 4	285.940 4	0.99
271.938	287.937	1.00
T*er»-	4	303.902 4	319.397 4	0.74
305.899	321.394	1,00
327,385 6	—
1236.933]	0.21
[258,930]°	0.20
5	337.863 4	353.358 4	0.S7
339.860	355.855	1,00
go*a-	6	373.821	389.316	1.00
375.318	391.813	0,87
H«pca-	7	407.782	423,777	1.00
409.779	425.774	1.00
431.765 3	—
OeS*-	8	441.743	457.738	0.36
443.740	459.735	1,00
*" Mol«nilax Ion peak.
a" 3^Cl<»-l*J5«ll«d apCBD standard oaaJt.
a'ton» which can ba aonitorad in TCDD analysaa for confirmation purpoaaa.
-------
TAB LB 4
SEQUENCE OF OPERATIONS IN GC-MS (MS-25) ANALYSES OF
CIILORODXBEHZODIOXXNS AND CHLOROP1BENZOFURANS
IN FIRST INJECTION OF SAMPLE EXTRACT
Elapsed
Tine
(Min)
0.00
1.50
2.00
4.50
5.00
Event
Injection, splitless
Turn on split valve
Begin Temp. program to 220*C
0[>en column flow to Mass Spec.
Start PROGRAM |1
Sweep = 100 ppm
Tine on each mass = 0.15 sec
GC
Column
Temperature
190*C
190*C
190*C
202*C
205°C
Temperature
Program
Rate
( *C/MIN)
5*C/mln
5*C/mln
Ions
Monitored By
Mass Spectrometer
(w/z)	
5*C/min
202,019
204.016
218.013
220.011
Compounds
Monitored
Mono-Cl furans
Mono—CI furans
Mono-Cl dioxins
Mono-Cl dioxins
8.00	Column reaches Isothermal hold
8.75	Stop PROGRAM II
9.50	Start PROGRAM 13
Sweep = 100 ppm
Time on each mass 0.15 sec
16.00	Stop PROGRAM 13
16.75	Start PROGRAM 15
Sweep = 125 ppm
Time on penta mass » 0.2 sec
Time on^Cl-TCDD = 0.07 sec
25.00	Begin temp program to 235*C
28.00	Stop PROGRAM 15
220*C
220*C
220°C
220*C
220'C
235*C
S'C/eec
269.941
271.938
285.940
289,937
327,885
337,863
339.860
353.858
355,855
Clj furans
Clj furans
CI3 dioxins
CI3 dioxins
37C1 labelled TCDD
CI5 furans
CI5 furans
CI5 dioxins
CI5 dioxins
ST"
-------
TABLE 4 (cont)
Elapsed
Tine
(Mln)
45.00
60.00
95.00
Event
Start PROGRAM #7
Sweep " 175 ppsi
Tine on each mass ¦> 0.35 sec
GC
Coluam
Temperature
Temperature
Program
Rate
(A'C/Hln)
235*C
Stop PROGRAM 17
Return GC to initial temp
Ions
Monitored By
Mass Spectrometer
(w/g)	
407.782
409.779
423.777
425.774
431.765
Compounds
Monitored
CI7 furans
CI7 furans
CI7 dloxlns
CI7 dioxins
37C1 labelled HpCDD
vl\
-------
TABt.fi 5
Ok
«
SEQUENCE Of OPERATIONS IM GC-HS (MS-25) ANALYSES Of
ClUjOROfl IB EHZOO IPX I MS AMD CHIOROOrBEHZOFORAHS
IN SECOND INJECTION OF SAMPLE EXTRACT
Temperature
Ions
Sweep * 100 !>(>¦>
Tlae on each nasa ¦ 0.15 sec
237.977
251.974
Elapsed

GC
Program
Monitored By

Tine

Column
Rate
Mass Spectrometer
Compuuds
(Nln)
Bvent
Temperature
( *C/Mln)
(«/z)
Monitored
o.oo
Injection, spiltless
190*C

	
—
1.50
Turn on split valve
190*C
	
—
—
2.00
Begin teap program to 220*C
190*C
	
—
—
4.50
0|*en flow to Mass Spec.
202*C
5'C/wln


6.00
Start PROGRAM 12
210*C
5#C/nln
235.980
CIj furans
CI j furans
Cl2 dloxlns




253.972
Cl2
dloxlns
8.00
Column reaches 220*C
220*C
—
—



liold isothermal





12.00
Stop PROGRAM 12
220*C
—
—

—
12.75
Start PROGRAM 14
220*C
—
3O3.902
CI,,
furans

Sweep - 100 ppm


305.899
CI,,
furans

Time on CI Mass » 0.15 sec


319.897
CI,,
dloxlns

Tine .on C1-TCD0 *=0.05 sea


321.894
CI,.
dloxlns




327.885
®'C1 labelled
24.OO
Stop PROGRAM 14
220*C
--
—

	
25.00
Begin temp program to 235*C
220*C
5*C/mln
—

—
26.00
Start PROGRAM 16
225*C
5*C/i»ln
373.821
C16
furans

Sweep ¦ 150 ppm


375.818
C16
furans

Time on each mass =0.25 sec


389.816
C16
dloxlns




391.813
C16
dloxlns
45.OO
Stop PROGRAM 16
235*C
—
—

— —
46.00
Start PROGRAM 17
235*C
—
407.702
Cl7
furans

Sweep ¦ 175 ppm


409.779
Cl7
furans

Tlae on each Haas -0.035 sec


423.777
ci7
dloxlns




425.774
Cl7
dloxlns




431,765
37C1 labelled
60.00
Stop PROGRAM 17
235*C
—
441.743
ciB
furans
70.00
Start PROGRAM |8
235*C
—
443,740
ci8
furans

Sweep » 225 ppm


457,738
CI j
dloxlns

Tlae on each masf - 0.55 see


459.735
CI j dloxlns
90.00
Stop PROGRAM 18
235'C




95.00
Return to initial teqp




-t
«A
-------
4*?
TABLE 6
LIST or ICN MASSES MGNITCRED JSEtG GC-SELECTED-ICN
MCNITCRI^IG MRSS SPECTROMETR? ?OR STMOX.TBHEOPS DETERMINATION
OF MONO- THR0PG3 DECA- CHLORINATED P0LYC3L0RXHRTE0 BXgHSMYLS (PC3a)
Class of	Number of
Chlorinatad	Chlorine	Ions Monitored (m/z)
PC3	Suba-cltaents	For PC3CCi2Hk),vC1*]
Mono*	I	188.039
190.036
Di-	2	222.000
223.997
Tri-	3	255.961
257.958
Tatra-	4	289.922
291.919
?«nta-	5	325.880
327.877
Hwca-	6	359.341
361.839
Htpta-	7	393.302
396.800
Octa—	3	427.763
429.761
Hon*-	9	463.722
465.719
Dsca-	10	497.683
499.680
-------
TABLE 7
WRIGHT STATE UNIVERSITY-BREI1M LABORATORY
DAYTONt OHIO 45435
RESULTS OF ANALYSES OF STACK EFFLUENT SAMPLES FROM
INCINERATOR LOCATED AT EMSCO. EL DORADO, ARKANSAS.
FOR TETRACHLORODIBENZO-p-DIOXmS (TCDDs)
(PCB DESTRUCTION TESTS)
TEST AND
SAMPLE TRAIN
NUMBERS
TEST 1, TRAIN 2
TEST 1, TRAIN 2
TEST 1, TRAIN 2
TEST 1, TRAIN 2
TEST 2, TRAIN 1
TEST 2, TRAIN 1
TEST 2, TRAIN 1
TEST 2, TRAIN 1
TEST 3, TRAIN 1
TEST 3, TRAIN 1
TEST 3, TRAIN 1
TEST 3, TRAIN 1
TEST 3, TRAIN 1
b.
TRW
SAMPLE
NUMBER	SAMPLE TYPE
6-0528	Probe wash
6-0529	XAD-2 resin
6-0530	XAD-2 resin
6-0531	Florisil
6-0560	Iwpinger °*
6-0561	Probe wash
6-0562	XAD-2 resin
6-0563	Florisil
6-0574A	Iapinger °*
6-0574B	Probe wash
6-0574C	XAD-2 resin
6-0574D	Florisil (1 of 2)
6-0574B	Florisil (2 of 2)
d.
NATIVE TCDDs DETECTED
IN ENTIRE SAMPLE
w/e 320 n/z 322 Average
o.ltl
O.BH ng 0.149 ng 0.158 ng
0.088 ng 0.122 ng 0.105 ng
0	0	0
0	0	0
0.400 ng 0.551 ng 0.476 ng
0
0
0
0
0
0
0
0
0
0
0
0
o
0
0
0
0
0
0
0
o
0
0
0
MINIMUM
DETECTABLE
QUANTITY
>5 ?
0.062 ng.
0.062 ng
0.073 ng
0.062 ng
0.22 ng
0.25 ng
0.38 ng
0.66 ng
0.35 ng
0.21 ng
0.20 ng
0.20 ng
0.45 ng
RECOVERY
146%
104%
85%
120*
45%
99%
134%
57%
71%
98%
94%
122%
46%
-fc
-------
TABLE 7 (cont)
'Based on J7C1i,-2,3,7,B-TCDD added as internal standard.
'Xnpinger sample from this series was broken when received by Wright State.
'Upper layer (apparently hexane fraction) only.
'The other XAD-2 resin trap was not received by Wright State.
-------
TABLE 8
n
WRIGHT STATE UNIVBRSITY-BREHM LABORATORY
DAYTOH, OHIO 4543S
RESULTS OF ANALYSES OF STACK EFFIAJEHT SAMPLES FROM
INCINERATOR LOCATED AT ROLLINS ENVIRONMENTAL SERVICES, INC., DEER PARK.
TEXAS, FOB TETRACIILORODIBENZO-p-DIOXINS (TCDDa)
(PCB DESTRUCTION TESTS)
TEST AND
SAMPLE TRAIN
NUMBER
TRH
SAMPLE
NUMBER
SAMPLE TYPE
NATIVE TCDDa DETECTED
IN ENTIRE SAMPLE
njz 320 b/e 322 Average
MINIMUM
DETECTABLE
QUANTITY
«
RECOVERY
TEST 1, TRAIN 1
6-0581
Inpinger
1.61 ng
1.97 ng
1.79 ng
0.43 ng
100%
TEST 1, TRAIN 1
6-0582
Probe wash
3.16 ng
4.51 ng
3.84 ng
0.43 ng
100*
TEST 1, TRAIN 1
6-0583
XAD-2 resin
0.53 ng
2.09 ng°
"1.31 ng
0.43 ng
100%
TEST 1, TRAIN 1
6-0584/
6-0585
(cowposite)
Florisil
0
0
0
0.70 ng
44%
TEST 2, TRAIN 1
6-0605
Inpinger
0
0
0
0.27 ng
75%
TEST 2, TRAIN 1
6-0606
Probe wash
1.32 ng
1.52 ng
1.42 ng
0.27 ng
129%
TEST 2, TRAIN 1
6-0607/
6-0608
(cowposite)
XAD-2 resin
0
0
0
0.38 ng
107%
TEST 2, TRAIN 1
6-0609
Florisil
0
0
0
0.48 ng
79%
TEST 3, TRAIN 2
6-0643
Inpinger
0
0
0
0.90 ng
100%
TEST 3, TRAIN 2
6-0644
Probe wash
—
~
—
—
—
TEST J, TRAIN 2
6-0645
MAD-2 reain
0
0
0
0.30 ng
eot
TEST 3, TRAIN 2
6-0646
Florisil
0
0
0
0.35 ng
58%
-------
NT
"'Based on 37Cl(,-2#3,7,6-TCDO added as internal standard.
k*Upper layer (apparently liexane fraction) only.
c*Incompletely resolved Interference at m/z 322 results In elevated signal.
^"Sasple contained only one layer which appeared to be the acetone-water layeri upper liexane layer observed In
other sanplea was completely absent.

-------
m
«*
TEST AND
SAMPLE TRAIN
NUMBERS
TEST 1, TRAIN 26
TEST 1, TRAIN 2
TEST 1, TRAIN 2
TEST 1,	TRAIN 2
TEST 2, TRAIN 1
TEST 2,	TRAIN 1
TEST 2,	TRAIN 1
TEST 2, TRAIN 1
TEST 3,	TRAIN 1
TEST 3,	TRAIN 1
TEST 3,	TRAIN 1
TEST 3,	TRAIN 1
TEST 3,	TRAIN 1
TABU! 9
WRIGHT STATE UNIVERSITY-BREHM LABORATORY
DAYTON. OHIO 45435
RESULTS OF ANALYSES OF STACK EFFLUENT SAMPLES FROM
INCINERATOR LOCATED AT EMS CO. EL DORADO, ARKANSAS,
FOR TETRACHLOROOIBENZOFURANS (TCDFa)
(PCB INCINERATION TESTS)
TRW
SAMPIE
NUMBER
6-0528
6-0529
6-0530
6-0531
6-0560
6-0561
6-0562
6-0563
6-0574A
6-0574B
6-0574C
6-0574D
6-0574E
SAMPLE
TYPE
QUANTITY OF
TOTAL TCDFs
IN ENTIRE
SAHPIJE
a.
Probe vash
XAD-2 resin
XAD-2 resin
Florisil
lmpinger ^'
Probe Nash
XAD-2 reBln
Florisil
Inpinger
Probe wash
XAD-2 resin
Florisil (1 of 2) 0
Florisil (2 of 2) 0
0
0
0
2.0 ng.
4.0 ng.
0.5 ng.
1.5 ng.
0
0
0
0
APPARENT
NUMBER
OF TCDF
ISOMERS *»•
0
O
0
4
3
2
5
0
0
0
0
0
0
QUANTITY OF
APPARENT
2,3,7,8-TCDF
0
0
0
0.5 ng.
1.0 ng.
0.2 ng.
0.3 ng.
0
0
0
0
0
0
MINIMUM
DETECTABLE
QUANTITY OF
TCDF 	
1	ng.
2	ng.
0.1 ng.
0.4 ng.
0.5 ng.
0.09 ng.
0.05 ng.
0.08 ng.
0.5 ng.
0.2 ng.
O.OB ng.
0.4 ng.
0i09 ng.
.£
c-
RECOVERY
50%
h.
70%
100%
45%
75%
100+%
100%
100+%
100+%
100%
51%
85%
-------
TKBLB 9 (cont)

'Based on Buawation of the areas of all UB8 chromatographic peaks observed In a selected TCDF retention tine
window, while Monitoring ¦/z 304 and 306| It Is assumed that the instrtanent response determined by calibrating
with an authentic 2,3,7,8-TCDF standard Is the sane for all TCDF Isomers.
k'The number of discrete mass chromatographic peaks observed In the TCDF window. Since all TCDF Isomers are not
available for determination of GC retention times, It Is not known whether or not each peak represents more than
one Isomer. The number olted Is therefore the minimum number of TCDF Isomers which can be present In the sample.
£•
'Determined on the basis of the mass chromatographic peak which has a retention time corresponding to that of the
2,3,7,0-TCDF Isomer. Other TCDF Isomers may also be Included In this peak.
d'Based on '7C1i,-2, 3,7,8-TCDO added as an Internal standard.
e"Impinger sample from this series was broken when received by Wright State.
''Upper layer (apparently hexane fraction) only.
'"The other XAD-2 resin trap was not received by Mright State,
^'Recovery could not be determined on basis of low resolution MS analysis because of unresolved interference at m/r 326.
-c.
u\
-------
1
TABLE 10
WRIGHT STATE UNIVERSITY-BREHM LABORATORY
DAYTON, OHIO 45435
RESULTS OF ANALYSES OF STACK EFFLUENT SAMPLES FROM
INCINERATOR LOCATED AT ROLLINS ENVIRONMENTAL SERVICES, INC., PEER PARK,
TEXAS, FOR TETRACHLORIDIBENZOFURANS (TCDFb)
(PCB DESTRUCTION TESTS)
TEST AND
SAMPLE TRAIN
NUMBERS
TEST 1, TRAIN 1
TEST 1, TRAIN i
TEST 1, TRAIN 1
TEST 1, TRAIN 1
TEST 2, TRAIN I
TEST 2, TRAIN 1
TEST 2, TRAIN 1
TEST 2, TRAIN 1
TEST 3, TRAIN 2
TEST 3, TRAIN 2
TEST 3, TRAIN 2
TEST 3, TRAIN 2
TRW
SAMPLE
NUMBER
6-0581
6-0582
6-0583
60584/
6-0585
(coapoaite)
6-0605
6-0606
6-0607/
6-0608
(composite)
6-0609
6-0643
6-0644
6-0645
6-0646
SAMPLE TYPE
Iapinger e"
Probe wash
XAD-2 resin
Florisil
Inplnger e"
Probe wash
XAD-2 resin
Florisil
Inpinger
Probe wash '
XAD-2 resin
Florisil
QUANTITY OF
TOTAL TCDFb
IN ENTIRE
SAMPLB '
3.5 ng.
5.0 ng.
5.0 ng.
0
6.0 ng.
14 ng.
2.0 ng.
0
0
0
2.0 ng.
0
APPARENT
NUMBER
OF TCDF
ISOMERS
6
7
8
0
9
8
9
0
0
0
8
0
QUANTITY OF
APPARENT
2,3,7,8-TCDF C
0.7 ng.
0.5 ng.
0.3 ng.
0
0.8 ng.
1.4 ng.
0.25 ng.
0
0
0
0.2 ng.
0
MINIMUM
DETECTABt£
QUANTITY OF
TCDF 	
0.08 ng.
0.15 ng.
0.15 ng.
0.08 ng.
0.1 ng.
1.0 ng.
0.07 ng.
j
0.09 ng.
0.2 ng.
0.5 ng.
0.1 ng.
0.4 ng.
d.
RECOVERY
100+%
100%
100%
1% \Q-
72%
122%
100%
85%
100+%
60%
72%
50% l r
-------
TABLE 10 (cont)
a"Based on summation of the areas of all mass chromatographic peaks observed in a selected TCDF retention time
window, while monitoring m/z 304 and 306. It is assumed that the instrument response determined by
calibrating with an authentic 2,3,7,8-TCDF standard is the same for all TCDF isomers.
b*Tlte number of discrete mass chromatographic peaks observed in the TCDF window. Since all TCDF isomers are
not available for determination of GC retention times, it is not known whether or not each peak represents
more than one isomer. The number cited Is therefore the minimum number of TCDF isomers which can be present
in the sample.
Q
'Determined on the basis of the mass chromatographic peak which has a retention time corresponding to that of
the 2,3,7,8-TCDF isomer. Other TCDF isomers may also be included in this peak.
^"Based on "cl|,-2, 3,7,8-TCDD added as an internal standard.
0 ^
'Upper layer (apparently hexane fraction) only.
Sample contained only one layer which appeared to be the acetone-water layeri upper hexane layer observed in
other samples was completely absent.
c*
-------

TABLE 11
WRIGHT STATS ONXTSSSIT?, 3KEHM LABORATORY, DAYTON. OHIO 45435
TCDD ISOMERS PfcTmMIUKD d STAC3C slgVUXENT SAMPLES FROM
DicsreaAncN tests at ensco, el dorado, askansas, amp
ROLXJKS ENVTROKMENTftl. SERVICES, DEER PARK. TEXAS
TEST AMD SAMPIfi
TRAIN ilCMBERS
TSW
SAMPLE
OTMBER
ORIGIN
TCDD ISOMERS
DHTKLTill
* OF TOT3U,
TCSDs
Test 2, Train 1
6—0606
Rolling
Environmental
Sarvicaa
1,3,6,3-
1,3,7,9
1,3,6,9
1,2,7,9
Others
a.
25
22
9
9
35
Test 2, Train 1
6-0560
ENSCO
1,3,6,3
1,3,7,9
1,3,6,9
1,2,7,9
Others
b.
20
18
3
5
54
a* TCDEa are also observed at retention times corresponding to- tha 1,3,7,8-,
1,2,6,3-, and 1,2,6,9- isomers, but thai a isomers ara apparently not completely
rssolved from several othar TCSD iaomera which could also account for these
^'TCDOa ara also observed at retention tiaas corresponding to tha 1,3,7,8-, 1,2,6,3—
and 1,2,3,4- isomers, but thasa pea*a may alao ba accounted for by aavaral other '
isomers.
-------
V 69
so
TABLE 12
WRIGHT STATS UNIVERSITY, BKEHM LABORATORY, DAYTON, OHIO 45435
HIGH RESOLOTION GAS CHROMATOGRAPHIC-LCW 3ESOLUTION MASS SPECTROMETRY
DATA ON TOTAL CDDa AND CDFs (TST5ACHL0RINATSD THROUGH OCTACHLORINATSD)
m T5ST 1, TRAIN 1 PROBE WASH SAMPLE (TRW NUMBER 6-0582)
FROM INCINERATION TSSTS AT ROLLINS ENVIRONMENTAL SERVICES, DESB o&Bg. TFvag
QUANTITY OF	MINIMUM
NUMBER OF	TOTAL CDD/CDF	DETECTABLE	.
CDD/CDF APPARENT ISOMERS DKTaL'l'^ IN TOTAL SAMPLE (no) C QUANTITY (no) % RECOVERY
TCDD	a	— —	100
PCDD	4	4	0.3
HxCDD	2	1	0.5	— -
HpCDD	1	2	2	100
OCDO	1	5	3
TCSF	7	5	0.2
PCDF	6	30	0.3
HxCDF	4	25	0.5
BpCDF	2	12	2
OCDF	1	10	3
** Prefix daaignationa axa : T ¦ tetra-; P ¦ penta-j Hx * haxa-> Hp * hepta; 0 ¦ octa-
b"Sinca only a limited number of CDS and CDF isomers ara available (see Tables 1 and 2),
it ia not possible to demonstrate for moat: of the elaaaas of CODa and CDFs that all
isomers of a given chlorinated group (for example* the pentachlorinatad isomers; there
are 14 PCDDa) are completely resolved by using the GC conditions employed here. Therefore,
it is not certain that each peaJc of the family observed for a given chlorinated group
corresponds to a single isomer. Hence, tha terminology "apparent" is used which
designates the number of GC peaks observed for a given chlorinated group of CDOs or CDFs.
c* Results corrected for recovery on the basis of recovery of added internal standards.
Recoveries for the CDDa were estimated on the basis of recovery of two internal standards
added to tha samples prior to processing. These standards ara 37Clu-2,3,7,3-TCDO, the
recovery of which was assumed to be indicative of the recoveries of native PCDDs and lower
chlorinated CDDs (that is, mono- through penta- CCDs), and 37Clu-l,2,3,4,5,7,3-HpCDO,
the recovery of which was assumed to be indicative of the recoveries of native HxCDDs
and higher chlorinated CDOs (that is, hexa- through acta- CDOs). No isotopically-
labellad CDFs were available, and so the recoveries of CDFs could not be directly
assessed. However, recoveries of CDFs are probably on tha same order as the corresponding
CSOs.
-------
L)10
TABL2 12 Ceoat)
51
e*Mar dataxininad in ttha HRGC-LSMS analyses; thasa data vara obtained by L2GC-HRMS
aire reported in Tabla 7.
""Wafers to recovery of *7C1-labelled standards nentioned in dots d.
-------
52
TABLE 13
WRIGHT STATS UNIVERSITY, BRHHM LABORATORY, DAYTON, OHIO 45435
HIGH RESOLUTION GAS CHROMATOGRAPHIC-LOW RESOLUTION MASS SPECTROMETRY
DATA ON TOTAL CDDs AND CDF3 (TBTRACHLORINATED THROUGH 0CTAC3L0RINATSD)
IN TSST 2, TRAIN 1 IMPINGER SAMPLE (TRW NUMBER 6-0S60) ?ROM INCINERATION
TESTS AT 3NSC0, EL DORADO, ARKANSAS
CDP/CDF
TCSD
PCDD
axcDO
HpCDD
OCDD
NUMBER OF
APPARENT
ISOMERS
QUANTITY OP
TOTAL CED/CDF 0ET5C7SD
IN TOTAL SAMPLE (ng) C
MINIMUM
DETECTABLE
QUANTITY (nq)
% RECOVERY
100 f
iee f
TCBF	3 4	O.S
PCDF	6 20	1
fflecnp	~ —	2
BfcCSF	—	6
OCSF	—	8
*" Prefix designations arc T ¦ tatra-; P » penta-; Hx - haxa-; Hp - hepta-; 0 - octa-
b"Since only a limited number of CSO and CDF isomers are available (sm Tables 1 and 2),
it is not possible to demonstrate for oiost of the classes of CSDs and CSFs that all
isomers of a given chlorinated group (for example, the pentaehlorinated isomers; there
are 14 PCSDs) are completely resolved by using the GC conditions employed here. Therefore,
it is not certain that each peak of the family observed for a given chlorinated group
corresponds to a single isomer. Hence, the terminology "apparent" is used which
designates the number of GC peaks observed for a 71van chlorinated group of CSDs or CSFs.
c*Results corrected for recovery on the basis of recovery of added internal standards.
^"Recoveries for the CDDs were estimated on the basis of recovery of two internal standards
added to the samples prior to processing. These standards are 37Cl*-2,3,7,a-TCDD, the
recovery of which was assumed to be indicative of the recoveries of native PCSDs and lower
chlorinated CSDs (that is, aono- through penta- CSDs), and 37Cl^-l,2,3,4,o,7,3-apCDD,
the recovery of which was assumed to be indicative of the recoveries of native SxCSDs
and higher chlorinated CDDs (that is, hexa- through octa- CSDs). No isotopically-
labelled CSFs ware available, and so the recoveries of CSFs could not be directly
assessed. However, recoveries of CSFs are probably on the same order as the corresponding
CSDs.
-------
Lj-J Z-
TABLS 13 (cent)
S3
®*Sot detexained in the HBGC-CBMS analyses? these data vera obtained by LSGC-SBMS and
are reported in Table 3.
Zm£*za to recovery of 37C1-labelled standards mentioned in Mota d.
-------
4 73
54
TABLE 14
WRIGHT STATS UNIVERSITY, BREHM LABORATORY, DAYTON, OHIO 45435
HIGH RESOLUTION GAS CHROMATOGRAPHIC-LOW RESOLUTION MASS SPECTROMETRY
DATA ON ?C3s IN TEST 2. TRAIN 1 ?LORISIL SAMPLE (TRW NUMBER 6-0609)
TB.GU INCINERATION TESTS A* aoT.T.TMq ENVIRONMENTAL SERVICES, DEER PARK, TEXAS
CLASS OP CHLORINATED QUANTITY CF	MINIMUM
PCS NUMBER OP APPARENT PC3S DETECTED	DETECTABLE
(NO. OF CHLORINE SUBSTITT7SNTS) ISOMERS a	 IN TOTAL SAMPLE(nc) QUANTITY tnc)
Monoehlorinatad (1)
—
—
0.1
Dichlorinated (2)
3
10
0.3
Trichlorinatad (3)
a
is
0.3
Tetrachlorinated (4)
4
3
0.5
Pentachlorinatad-
—
c

Decachlorinatad (5-10)
a*Only a limit ad number of PCS is ours are available for use in astablimbing GC retention
times and sensitivitias for PCBs, and so it could not be demonstrated tiiat all isomers
of a given chlorinated claas ara resolved using the GC conditions employed here. Therefore,
it is not cartain that each peak of the family observed for a given chlorinated group
corresponds to a single Isomer. Hence, the terminology "apparent" is used, which,
designates the number of GC peaks observed for a given chlorinated group of ?C3s.
b"Assumes that the sensitivities of all isomers of a given chlorinated class are the same.
These data are not corrected for recovery, since no isotopically-labelled PCBs were
available to establish recoveries for the method used.
c*Some CI5- Clio PCBs are probably present, but there were too many interferring MS peaks
to reliably determine these.
-------
BS--30 CROSS SCAN REPORT, RUM, VElU000a3 i/7^1
* 236-233 * 313-322 0 327-323
55
31 33
3t 13
5, 33
I9i 33
iaa
9. 13
30J(*12S9
130X»3«4I3
I9d%»2692
U
iaa
TCDO
rSQMEKS
1.
1.3,6.3
2.
1.3,7,9
3.
1.4.6,9
4.
1,3,7,8
5.
1,2,6,8
6.
1,2,7,9
7.
1,2,6,3
3.
1,2,3,4
9.
2,3,7,8
10.
1,2,7,3
U.
1,2,6,7
12.
1,2,8,9
239
Scan. Munte«r_.
Figura 1. WSU-Braha laooratory, Dayton, Ohio	___
Salactad-Ioa Haaa drcaatooraaa Obtaiaad 3y SHGC-mMS
T__ ii-xsomcr TCDO standard Hixtura. Maaaaa Moaitorad Ara ladicatad
It ®» of «aca Along with Carraapondiw? Symbol* ?or 2aeh Tsraea Shown.
Sunbaring of ?aak* Corraaponda To taonar Idantificationa Shown Right.
499
-------
« 256-233 * 313-322 0 327-323
V7 
-------
~ 313-322 * 327-323
V7 L
57
6. 03
1 005s»3«93;
Iu0";»lfl741
A

*Vu
l ' ^
109
l '
130
290
230
Sean Xuaber
390
'till T"
>30
490
?igur« 3. Salactad Ion Monitoring »ui Ciromatocraaia Obtainad Froa HRGC-LSMS
Analysis of 2NSC0 Teat 2, train 1 tmpingar Sampla (TSW 5-0560) Showing
TCDD Isomar PaaJcs. the Susdsaring tJsad to Idantify taomar ?sajca
Corraaponds To That: Indicatad la ?iguza L. 37C1u-2,3,7,a-TCaD was Addad
To Sampla As Internal standard.
-------
9a
99
79
69
59
49
39
29
18
a
.58
>30 1.13 2,33 3. S3 3.13 S, 33 7.33 9,13 19.33

sa
400
130
139
330
290
Sean Number
Figure 4. 3BGC-MS Hui Chramatograna Basalts Obtained For Calibration Standard
Containing Tatrachlorodibanzoforans (2,3,7,8-TCOF, 1,2,4,3-TCST,
2,3,6,3-TCDP). Zona Monitored And Corresponding Identification
Symbol For Traces Za Shown At Top of Page.
3-304 » 303-306
^77
-------
~319-329 # 321-322 0 327-329	^7%
*^9
9,00 1.13 2.33 3. 59 3.13 S. 33 7,53 9.13 13.38
90
33
S3
29
13
138
2S8
330
1
Scan Mtrab«r
?igura 5. HSGC-I3MS Mass Chromatograaa Obtained ?or Calibration standard
Containing 37Cl(f-2,3,7,a-tCDO.
-------
* 337-338 # 333-340	^7 
-------
• 327-328 ~ 333—354 0 333-336 y 7q
3 > 98
S< 82
9: 84
12: 36
108JJ»S?
180%«43

i!jfl|?!n f" rf [VI ,

UOTmm

18
a
iAmU
188
288	388
Sean Numhcr
1 ¦ '
488
i" 1 1 1 1-p
399
Tiquxa Jl. asOC-tSMS Mass Chranatogrims Obtained ?or Calibration Standard
Containing 37CIu-2#3,7,a-TC3D with InsttroMat *lao Configured For
Detection of ?«ntac£ilororfihonzo-p-Qloxiaa.
-------
• 373-374 » 373-37S
^ ? J
62
a< 90
2.37	3.36
a. 33
11.34	14,33
17;
iaa \\\t			
t I t t i t
taegaiea
1002*212


ft
scan NUDcar
$80
Figura 3. HBGC-I2tt!S Mass Chroma toyiams Obtained For Calibration Standard
Containing Haxaciilorodiianzofurana (l,2,4,S,7,9-a*CDF).
-------
OS-30 CROSS SCAN REPORT* RUN, ACGS3Q093 'J 7 3-
63
« 339-393 • 391-392
0« ®®	2i 37	3»36	9: 35	lli34	14,33	17,32
99
88
70
29
10
400
I
<09
?iguz« 3. aaSC-LUMS Mass Chroaatogrsma Obrainad ?or Calibration Standard
Containing- Saxachlorodiiaaao-p-Oioxias Including (L,2,3,4,7,a-&c3D)
-------
• 407-493 » 403-413
0,30	4,2?	3.37	13.27	17.37	22,
198
99
90
78
w
sa
20

10
•dUi A>»ii
t
Scan Naabar
Figura 10. aaGC-UOfS Mui Chrcoacograms Qbeaiaad ?or Calibration standard
Comuin1n
-------
* 423-424 # 425-426 0 431-432	H¥V
65
9, 98
4, 27
9. 37
13. 27
17, 37
22.
130
99
39
70
50
>. 3Q
0
c
V
u
s
H 49
41
>
9
3 39
x
29
19
1092*iS€3

1003*321 1

I 002*394
L
U^Vv^J^ui,
VvJu^_ .

299
399
Sean KUmbar
499
S9«
rigura U. aHGC-LUMS Mass Cbronatograas Obtained For Calibration Standard
Containing amp-tzzchlorvdlZuma-p-Oiaxln* (1,2,3,4,6,7,3-a&CDn
37Cl,»-l,2,3,4,6,7,3-BpCDO).	****
-------
3S—38 CROSS SCAN REPORT. RUN, ACC388883 ^ Pgfi
* 44 i—442 • 443—444
8,99 2,2$ 4, 3« 7.23 9, 33 12.24 14.34 17:23 19
<50
13
238
138
288
1
Seu. MuBh*ir
Figura 12. HBSC-UIMS Hin Chromatograma Obtained For Calibration Standard
Containing Octacfalorodibanzofuran.
-------
•437-433 # 439-453
67
9. 90 2- 26 4,36 7.25 9:33 12.24 14,34 17:23 13, 33
"3 4J
130
30
230
1
Figure 13. HHGC-L8MS Mass Chroma tocxajn* Obtained ?or Calibration standard
Containing Octachlorodiiaiuo-o- Dioxin.
-------
DS-53 CROSS SCAN HSPOHT< RUN. RES3FC0003
Ji 7	««
* 303-336
0
2, 38
3. S3
3; 13
S, 33
7. 33
3 ¦ 13
1005{»32733
70
£50
10

0
1
9
230
350
Tigtiza 14. HSCC-L3M3 Hua Chromaeeqna Obeaiaod Za Analraia of. TCSFa In n«inw«
Kit 2, trmia 1 Tzefaa Huh Saapla (2381 Slater 6-0606).
-------
3S-S5> CROSS SCHH REPORT, RUN. RE3BF30002
>i?2
• 337-340
0:00	1:09	21 17	3,27	4, 36	3.46
1005S"3249
u 3i
25
30
1
?tgnr« IS. BXSC-UMS Him Ctrenweegrm Obcmiaad Ca Aulysi* Of PCCFs la
Cue 2, Sfcmia 1 ?rai» Kaah 3anpl« CTSW Hmhmr <-0<06).
-------
DS-53 CROSS SCAN REPORT* RUN. RESBFB0302
* 333-335
70
9, 9<3
1 ® ^ r i i > I I If
1 : 33
' ' 1 ' ' 1 1 ' 1
17
3: 27
4: 35
3 i 45
¦ ¦ 1 1 ' 1 1 1 ' ' 1 ' 1 1 ' 1 1 ' ¦ ' 1 1 1 ¦1 ' » ' ' ' ' 1 ' 1 ' ' 1 ' 1
t00y.»«733
9t.
ac
figar* IS. HSGC-&8M3 M&aa Oreoatogzia Cbtaiaad In Anal^si* of PCSDs la Sailisa
?ue 2, Train I frobm Wmaii Ssmplm (TRW Mmbar 6-0606},
-------
D3-55 UKOSS SCAN REPORT* SUNi RE3DEF0001	__
Hlo
m 373-37S
0:00	2sl7	4j3fi	5:33	3:14	ix,
100*««S01I<9
?(.
S«
U
1
?igur« 17. sSGe~E3Ml Maaa anuteqru Obeaiaed la Analysis of axCSTs ta
Tut 2/ teain 1 Prote Hub Swpla (SSW Suobor 6-0606).
-------
D3-S3 CROSS SCAN REPORT. RUN. RES2EF9031
» 33^-392

72
9: 90
Z> 17
4 : 3S
6, 33
3: 14.
11 . 33
35

-*
I
s
5 50
4€
1 1 11 '
1005!»10933
199
299	399
Sean suaber
499
399
rigors 18. aXSC-£3M8 Hui Crcomatograa Obtained Za Analyaia of HaCBPa Xa Sellina
Taafc 2, train 1 Praia* ftaab Sanpla CTSW Misobar 6-4606).
-------
DS-33 CROSS SCAN REPORT. RUK, RESBFP0002
73
« 407-4L0
0:00 2 s 08 4,02 6:03 3=07 10.10 12s 12 14,13 IS, 17
100 I		 ¦ ¦ I ¦ ¦ ¦ ' , ¦ ' ' ' I ' ' i ¦ ' 1 ' ' i ' ' ' ' ' i i ' ' ' ' ' ' ' i I ' i ' I I ' ' ' ' ' ' ' ¦» ' I t ' ' ' I ' ' I ' I I ' ' I i ' I i t I
90
30
5* 79
s
M
e
M
« €3
1 00J{»737S0
1
50
40

1.09
130 200	250
Seas MMBhwr
300 330	499
in i |MI,11 nji am Obcadaed la Aaaly«i* ot SpCSfs In aolllnt
	MKK-L2MS Hui Csseaaxoqm eowa*,	»r-
Ttgur»	^ ^ ; prob4	Sasqpl* (flHI »»to«r 6-C«0«).
-------
OS-33 CROSS SCAN REPORT, RUN. RE3DF-Q002
i 13	74
* 423-425 » 431-432
5: 93
3: a?
12: 12 14; 13
IS.
199J«»3S1320
90
100J«»294«2
39
39
40
0
400
339
1
39
Soma Sgffihr
Figur« 20. SBSC-LSMS Mass Otreoato^xaa Obtained ta Analysis of 3pC2Ds la Rollins
Ts*C 2, Train 1 Proba Wash Sampla (T3W Muabar 6-0686).
-------
03-55 C 3 0 S '•> 'iCHU REPORT- RUN-. ft E3 OF G'3 y ki2
* 441-444
4ii
75
0:39	2:15 4:36	S-.33	9,13	1 1:34 13,34
100
» * i » » > i » i » »
100%*42321
i


Si
90
m
e
«
m
K
4 a
20
t*
33	199
139	209	230
Scas tUsfait
389
?igor« 21. 8RSC-IJWS Mu« Ojroaatogram Obtained la Analysis Of ocsr la ¦«¦)¦»<	
Tmwt 2, Traia 1 Probm tfub S*apl« (taw Missb«r 6-C606).
-------
0 3 ~ 33 CIV 0 3 3 3 C«il RETORT/ R U t < t RE3S7"S3i3'ji
# 437-453
* ^ c&
76
9:30	2:15 4,35	5=33	3=13	11:34 13:34
13Q.,,.,,,.,i..,.,...,i,..,,.,,,ri.
1303*9041
i l I i t > > 1
3«
2S
li
y
i . ii , i ¦ i i | i i m i i i i i i i i i i i : i i i | i i i i i i i i i i i i i i . ii i i | i i n ii i i i | . i m i i i ¦ i'
30	100
139	200
Scan ttaariMr
230	300
Fiqura 22. axtGC-LSMS Mass Omaategraa Obtained Za Analyaij of OCSO la aoJLlins
Smc 2, train 1 ?rota vtaab Sanpl« (TSH tfuabar 6-0606).
-------
5-33 CROSS SCAN REPORT* RUN, PC3130091
133-192 * 222-223 0 233-233 & 239-292
9:99	2i 39	5:21	9:94	19,46
130^-27159593
90
1005j»37S29a0
100J:»49S9472
19055*87321 S3
SO
33
49
30
30
1
Sean Humbar
,3- s«l«eesd Ion Mass Ojroaatogrms Obtained ?rem HHCC-L2MS Analysis Of
Figurs 23".	Qi.(	„d Tstwchloriaatad PCS Standards. ions
Mnitorsd and Corresponding Symbols Ossd To Iab«l Trae«s Ars Listsd At
Top. 3« Tabls 5 ?or Listing of ion husm Corrsspending To Sach ?C3
Class.
-------
33 CROSS SCAN REPORT. RUN, PCS373902
497
233-233 t 233-232 0 323-323 4 333-362 * 333-336
78
0, 00	1:31	2,06	3,1 1	4,16	3,21	6,26
9 l ¦ i r . ; . . t I i l ¦ t l . ¦ i . I i i i i f l t i i t l ¦ ¦ ¦ i i ¦ ' ¦ I i i ¦ t l t ¦ ¦ ¦ I i ¦ > i , ' ¦ . i ' ' i ¦ ¦ > ¦ ¦
1002*1139304
1092*1383039
1002*2302734
1002*300132
1002*3318!
j ii i
60
Scan number
Figur« 24. Salactad-Ion Mass Chromatogxams Obtained From 3BCC-LSHS Analysis Cf
a Mixwira of Tri-, Tatra-, Panta-, Haxa-, and Haptachlorinaead PCS
Standards.
-------
03-33 CROSS SCAM REPORT, RUN, PC3S100003 ^^?
339-352 # 393-39S 0 427-439 4 453-46S * 497-500
<*, 00 1,01 2,0s 3:11 4, IS 3,21 " l" 143 tS®
Scan ttuabcr
- i j t__ u..a rsirtsnattaeraaui Qbfe*in«d Proa KBGC-tilWS Aa*iy*ia OH
Standards.
-------
OS-!
• 2:
l
100
90
30
70
€0
se
40
30
20
10
0
ao
S CROSS SCAN REPORT * RUN. RES34P0002 l}G) Cj
2-223 » 224-223
S: 59
«. 2b
3 t 34
1305S-33310
i005«»5?2aa
T-f-T-T-r-f-r^—r
I 1 1 1 11 1 I 11 I 1 I 1 . I 1 I I 1 I 1 I I II I I I I I 1 I 1 I II I I I I 1 I I
1 13
120
130	140
Scan Numbar
130
isa
.gura 26. Salacrad Ion Maaa Chranatograaa Obealnad ?rca HSGC-LSMS Analysis of
SHSCO Tast 2, T3rain 1 ?lorisil Sampla (TSW 5-0609) Tor Diehlorinatad
PC3s.
-------
BS-55 CROiS 3ChH REPORT/ RUN: R£53y«.
S'o o
253-23S * 237-253
3.94 $> 35
3:41
13. 46
11.51
31
12-. 36
1 ggx»291 S3
1 303S»29574
T-l
139 1 
-------
* 539-230 # 231-292
fo I
S3
19.46	11,31	12>56
183 ¦ ¦ ¦ i ¦ i i ¦ i t * ¦ ¦ i ' * ¦ * ¦ * 1 ' ' ' '
3?
6?
2 33
0
JJ
c
41
a
x
30
14, 91	13, 96
' ' 1 1 : 1 1 1 1 1 1 ' 1 ! 1 1 1 1
16.11
130:'.-7S24
100*»S044
\Av
11 i
229
j r i r i i1 i i i r j . . t1 i i i
Scan Number
T—1 • 1 1 1 ' • I
233	300
"igure 28, Selected Ion Maaj Chronaeograas Obtained ?rom HSGC-LSf"
Analysis of EHSCO Teat 2, Train 1 Florisil Sample
(TSW 6-0609) For Teerachloriaatad ?C3s.
-------

Division 6
-------
Soi
Air Quality Modeling Analysis of
Dioxin and Furan Emissions from Rollins
Prepared by:
Joe Winkler
Air Programs Branch
EPA Reqion 6
January 1981
-------

Statement of Assignment
The task entails the evaluation of the ambient impact of dioxin and furan
emission from Rollins. The following requirements were specified for
the concentration estimates:
1.	One year of National Weather Service meteorological data was used
in the analysis.
2.	Two concentration estimates were requested:
a.	Maximum annual concentration
b.	Maximum annual residential concentration
3.	Model input assumptions
a.	Dioxins and	furans behave as nonreactive gases and have infinite
half-lifes
b.	The average	stack gas velocity (10.97 m/sec) and stack gas
temperature	(345.0 K), based on tests for Oay 1, Day 2, and Day
3 were used	as representative of annual average values
c.	The emission rates Identified in Table 1 were used as
representative of annual average values.
Selection of Modeling Technique
A dispersion model 1s a commonly used technique for relating pollutant
emissions to ambient air quality. It 1s a mathematical description of
pollutant transport, dispersion and transformation processes that occur 1n
the atmosphere.
The current EPA guidance on selection and use of air quality models states
that the gausslan plume model is state-of-the-art and recommended for
pollutants that are nonreactive. The gausslan plume equation is a solution
to the simplified conservation of mass equation.
The Rollins plant has one stack which is located on a site for which
urban diffusion modeling parameters and urban mixing height assumptions
were determined to be appropriate. This decision was based on using
the land use procedure recommended in the proposed revisions to EPA's
"Guideline on Air Quality Models" (Reference 2) and aerial photographs
of the area around the stack.
Annual average concentrations are desired. The use of RAM (Gaussian -
Plume Multiple-Source A1r Quality Algorithm) 1n the long-term single
point source mode is recommended in an urban area where there are no
significant meteorological or terrain complexities. RAM 1s used to
calculate concentrations for an entire year. The model output shows
the annual arithmetic mean concentration at each of the receptor
points of a radial grid and identifies the maximum annual concentration.
-------
£04
Consistent with the most recent EPA modeling guidance, the new version
of RAM (12-30) was employed. Adequate verification of the new version
was performed in the later part of 1980 by EPA's Office of Research
and Development.
The basic time increment for the model is one hour. One-hour concen-
trations are computed for each of the hours in a given year. The
one-hour values are averaged to obtain concentrations for longer averaging
times.
Successful application of RAM is dependent on recognition of limitations
Imposed by mathematical assumptions associated with the model, the
structure of the computer program, and the availability and form of
meteorological data.
The assumptions Incorporated into RAM can be summarized as follows:
1.	The pollutant exhibits the dispersion behavior of a non-reactive
gas.
2.	Gravitational effects and chemical transformation of the pollutant
are not considered.
3.	Transformations of a pollutant resulting in Its loss throughout
the entire depth of each plume is accomplished by an exponential
decrease with travel time from the source. The Input parameter
1s the length of time expected for loss of SOS (half-Hfe) of
the emitted pollutant. Although this view of chemical or
physical depletion processes 1s overly simplistic, 1t may be
useful under certain circumstances.
4.	Complete reflection of the plume takes place at the earth's
surface, and the plume 1s completely reflected at the mixing
height.
5.	No consideration 1s given to the cumulative effects from consecutive
1-hour periods, and no allowance 1s made for changes in the wind
with time to affect the path taken by the plume downwind.
6.	No Initial concentration of pollutant.
7.	Continuous uniform emission rate.
8.	Homogeneous horizontal wind field.
9.	No directional wind shear 1n the vertical.
10.	Dispersion coefficients are based on aerometrlc measurements
taken 1n open, level to gently rolling terrain.
11.	The pollutant within the plume takes on a gaussian distribution
in both the horizontal crosswlnd and vertical directions.
Dilution dominates diffusion in the downwind direction.
-------
$"06
13.	No consideration of aerodynamic effects resulting from the
interaction of the wind with the physical structures nearby the
plant.
14.	The input wind direction includes a random variation (-4 to *5°)
superimposed on the value recorded by the National Weather Service
(NWS).
15.	The wind speeds and directions should be hourly averages which
are representative of the region being modeled. (In reality, NWS
hourly observations are not hourly averages but are one minute
averages at the time of observation, usually about 5 minutes
before the hour).
16.	A single mixing height and a single stability class for each hour
are assumed representative of the area.
Please refer to the RAM Users' Manual (Reference 1) for a more complete
and thorough treatment of RAM assumptions and limitations.
RAM presents a specific set of input data requirements, and yields output
data consisting of pollutant concentrations for a specific averaging
time and receptor locations. The input data requirements can be broadly
classified as.meteorological factors, source factors and site factors.
As specified in the "Statement of Assignment," only one- year of meteorolo-
gical data (1964) was used in the RAM Modeling. In view of the priority
placed on the modeling effort, the decision was made to use one year of
Houston (Hobby) surface data and one year of Lake Charles, Louisiana upper
air data (for mixing heights) as being the most representative data
readily available. Meteorological inputs to the model consisted of the
following observations from each day of 1964:
1.	hourly wind speed
2.	hourly wind direction sector (tens of degrees)
3.	hourly ambient air temperature
4.	hourly total opaque cloud cover
5.	twice daily mixing heights
6.	hourly cloud celling height
The source data required for calculation of an annual average concentration
consists of an annual average emission rate, annual average stack gas
exit velocity and annual average stack gas temperature. (See Attachment 1.)
In addition, the following emission source parameters were input into RAM:
1.	Plant elevation and locations
2.	Stack exit diameter
3.	Physical stack height
A description of the site and Its surroundings 1s provided in Attachment 2.
RAM has available a radial receptor grid. This grid provides for receptors
at five different distances on thirty-six (10° increment) radlals or a
-------
$0 1
total of 180 receptors for each model run. Based on model test runs,
a dense receptor grid near the absolute maxima and a line of arid points
along the southern boundary of the residential area to the north and north
northwest were employed.
III. Model Application
The latest EPA modeling guidance recommends that the following options
and parameters be applied to the RAM Modeling:
1.	The urban option was employed.
2.	Terrain in the area modeled was assumed to be flat.
3.	Infinite half-life was assumed.
4.	Stack tip downwash was not used.
5.	Final plume rise was used rather than transitional plume rise.
6.	Buoyancy Induced dispersion was not used. (As stated in
Reference 3, buoyancy Induced dispersion should only be used
with complex terrain models).
7.	The anemometer height was assumed to be at 10 meters.
8.	The following urban wind profile law exponents were used: For
stability A: 0.15, B: 0.15, C: 0.20, D: 0.25, E: 0.40,
F: 0.60. (See page 13 of Reference 1.)
9.	1964 Houston (Hobby) surface data was used.
10.	1964 Lake Charles upper air data was used.
11.	Receptors were placed along each of 36 radlals originating
at the stack location at the Intersections of the following
ring distances (km): 0.4, 0.5, 0.6, 0.7, 0.3, 0.9, 1.0, 1.2,
3.2, and 4.5. In order to locate the absolute maximum, a dense
receptor network was placed close to the point source around
the 340° radial.
The following emission source parameters were used in this RAM modeling:
1.	Stack base elevation: 0 feet.
2.	Stack location (long/Tat): 95° 05' 45729° 43' 45" (Texas
Control Board Permit)
3.	Annual average stack gas exit velocity: 10.97 m/sec.
4.	Annual average stack gas temperature: 345.0 K.
-------
$0%
5. Physical Stack Height: 30.48 rn.
5. Stack Exit Diameter: 1.33 m.
7. Annual Average Emission Rates: (See Table 1.)
IV. Consideration of Background
The contributions due to background were not considered in the air quality
analysis.
V. Presentation of Results
The results are presented in Table 1. When evaluating these results it
is important to consider the assumptions incorporated into RAM which
were discussed above. In particular, the dispersion coefficients are
valid only for downwind distances greater than 100 meters from the stack.
Beyond a few kilometers downwind, the estimates are based on limited data
and so may be less accurate. It should be remembered in this application
of RAM that the accuracy of the results is dependent upon not only the
model, but on the quality of the input information used in the application
(including both emission information and meteorological data).
VI. Acknowledgements
The author wishes to recognize the major contributions of F. Hall
and J. Yarbrough to this study.	F, Hall's preliminary work and
prudent advice were invaluable.	Special recognition is due to
J. Yarbrough who did the bulk of	the computer modeling. The author
also wishes to thank L. Stevens,	J. Sales and P. Schwindt for their
contributions.
VII. References
1.	User's Guide for RAM, November 1978, (EPA-600/8-78-016A).
2.	Proposed Revisions to "Guideline on Air Quality Models,"
October 1980.
3.	DRAFT Regional Workshops on Air Quality Modeling: A Summary Report,
July 1980.
-------
Test Day	Annual Average Emission Rates
TCDD (lxlO-9 grains/second)
1.	Adding in detection limits for
ND samples
Day I	43.74
Day 2	17.83
Day 3	9.64
2.	Without detection limits
Day 1	43.74
Day 2	10.34
Day 3	9.64
TCDF (1x10 grams/second)
1.	Adding in detection limits for
ND samples
Day 1	90.13
Day 2	166.67
Day 3	19.28
2.	Without detection limits
Day 1	85.08
Day 2	160.13
Day 3	12.44
~These are annual arithmetic averages
Table I
Absolute Maximum Annual 3
Concentrations (grams/meter )*/
Distance and Direction from Stack
Maximum Annual Concentrations
(grams/meter )* in the Jtesidentia
Area/Distance and Direction from
the Stack
4.68xl0~}1/0.24 km N,	0.8 km W
1.91x10,2/0.24 km N,	0.8 km W
1.03x10/0.24 km N,	0.8 km W
4.68xl0~|1/0.24 km N,	0.8 km W
1.11x10,2/0.24 km N, 0.8 km W
1.03x10/0.24 km N,	0.8 km W
2.19x10"^/4.3 km N, 1.2 km W
8.92x10 !?/4.3 kiu N, 1.2 km U
4.82x10 /4.3 km N, 1.2 km U
2.19xl0~Jlj/4.3 km N, 1.2 km W
5.17xl0~|?/4.3 km N, 1.2 km W
4.82x10 /4.3 km N, 1.2 km U
9.64x10"jl/0.24 km N, 0.8 km W
1.78xl0"J:/0.24 km N,	0.8 km W
2.06x10/0.24 km H,	0.8 km W
9.10xl0~}l/0.24 km H, 0.8 km W
1.71x10"J?/0.24 km N, 0.8 km W
1.33x10/0.24 km N, 0.8 km W
4.51xl0~jc/4.3 km N,	1.2	km U
8.33xl0,c/4.3 km N,	1.2	km W
9.64x10/4.3 km N,	1.2	km M
4.25xl0~j|/4.3 km N.	1.2	km W
8.01xl0"J;/4.3 km N,	1.2	km W
6.22x10 /4.3 km N,	1.2	km U
-------
£'o
Attachment 1
Sampling and Analysis for PCBs, Dioxins, and Furans
In order to determine the emission rates of the various air pollutants,
EPA required the facilities to conduct stack emission tests. Tests were
conducted by an independent testing firm using methods and procedures
prescribed by EPA.
During all emission tests, EPA representatives were on-site observing the
tests and recordinq information concerning the operation of the incinerator
and the air pollution control equipment.
Sampling for PC8s, dioxins and furans was performed in the stack utilizing
a modified EPA Method 5 sampling train. The sampling train incorporated
two (2) resin cartridges to adsorb the organic vapors. The resin cartridges
contained XAD-2 resin and florisil resin.
The resin cartridges were located downstream from the probe and between the
third and fourth impingers. The sampling train was designed to collect
both particulates and organic compounds in the stack gas. Both resins were
used so that the second resin would absorb any residual organic compounds
which may have passed through the first resin. For dioxin and furans, the
cartridges which house the resins were water jacketed. Cold water from the
ice bath surrounding the impingers was pumped into the jackets which
maintained the resins at a temperature of less than 70°F. Because of the
sensitivity of organic compounds to ultraviolet lighti the resin cartridges,
impingers, and interconnecting glassware were protected from the sunlight by
aluminum foil and opaque plastic.
The stack gases were sampled isokinetlcally for a period of four hours.
At the completion of each sampling period, the stack samples were transported
to the laboratory for analysis. The stack samples were extracted with an
appropriate solvent and analyzed using capillary column gas chromatography and
high resolution mass spectrometry. For dioxins and furans, and added quality
control technique was used. The stack samples were spiked with a labeled
dioxin and furan which allowed the analyst to determine the extraction
efficiency of each sample.
-------
sn
Attachment 2
Rollins at Deer Park, Texas
Surroundinqs
The Rollins Environmental Services facility is located on the south side
of the Houston Ship Channel, about 4,000 feet to the southwest of the
shoreline. It is 1n the middle of one of the most densely industrialized
areas in the nation. From the Ship Channel's Turning Basin at the edge of
downtown Houston, to the Bayport Bridge, the ship channel 1s lined on
both sides with industrial complexes. Because of the complete lack of
zoning regulations 1n the area, these industrial districts are closely
flanked by residential neighborhoods and small incorporated townships.
Within the area of concern for emissions from Rollins there are two
townships: Deer Park and Deepwater. These are located to the south and
southwest of the Rollins plant site, with the most populous portions of
both towns lying within three miles of the site.
The ecological character of the region is that of a coastal short-grass
prairie. There are no known characteristics of the flora or fauna which
would make them especially susceptible to or immune from contamination by
dloxins or furans.
Distribution of Population
The Agency estimates that up to 25,000 people could experience exposure
to some level of emissions from the incinerator. Within a radius of one-
mile the only exposed population 1s a group of employees of the nearby
Industries. These would be subject to an eight-hour exposure. Within
a two-mile radius is the most densely populated center of Deer Park. A
three-mile radius takes in about 1/3 of the most densely populated section
of the town of Deepwater.
-------

Division 7
-------
JTJ3
Air Quality Modeling Analysis of
Dioxin and Furan Emissions from ENSCO
Prepared by:
Joe Winkler
Air Programs Branch
EPA Region 6
January 1981
-------
511
ENSCO
I.	Statement of Assignment
The task entails the evaluation of the ambient impact of dioxin and furan
emissions from ENSCO. The following requirements were specified for
the concentration estimates:
1.	One year of National Weather Service meteorological data was used
in the analysis.
2.	Two concentration estimates were requested:
a.	Maximum annual concentration.
b.	Maximum annual residential concentration.
3.	Model input assumptions:
a.	D1ox1ns and furans behave as nonreactive gases and have infinite
half-lifes.
b.	The average stack gas velocity {2.86 m/sec) and stack gas
temperature (346.6 K), based on tests for Day 1, Day 2, and Day
3, were used as representative of annual average values.
c.	The emission rates Identified in Table 1 were used as
representative of annual average values.
II.	Selection of Modeling Technique
A dispersion model 1s a commonly used technique for relating pollutant
emissions to ambient air quality. It 1s a mathematical description of
pollutant transport, dispersion and transformation processes that occur in
the atmosphere.
The current EPA guidance on selection and use of air quality models states
that the gausslan plume model 1s state-of-the-art and recommended for
pollutants that are nonreactive. The gausslan plume equation is a solution
to the simplified conservation of mass equation.
The ENSCO plant has one stack which is located on a site for which
rural diffusion parameters and rural mixing height assumptions were deter-
mined to be appropriate. This decision was based on using the land use
procedure recommended 1n the proposed revisions to EPA1s "Guideline on Air
Quality Models" (Reference 2) and aerial photographs of the area around the
stack.
Annual average concentrations are desired. The Single Source (CRSTER)
Model is recommended for point source applications 1n rural areas where
there is one stack and there are no significant meteorological or terrain
-------

complexities. The CRSTER model calculates concentrations for an entire year
and prints out the annual arithmetic mean concentration at each of the
receptor points of a radial grid and lists the maximum annual concentration.
The CRSTER model is based on a modified form of the gaussian plume equation
which uses empirical dispersion coefficients and includes adjustments for
plume rise, limited mixing height and elevated terrain. The basic time
increment for the model is one hour. One hour concentrations are computed
for each of the hours in a given year. The 1-hour values are averaged to
obtain concentrations for longer averaging times.
Successful application of the CRSTER model is dependent on recognition of
limitations imposed by mathematical assumptions associated with the model,
the structure of the computer program, and the availability and form
of source data and meteorological data.
The assumptions Incorporated into the CRSTER model can be summarized as
fo11ows:
1.	The pollutant exhibits the dispersion behavior of a non-reactive gas.
2.	Gravitational effects and chemical transformation of the pollutant
are not considered.
3.	None of the pollutant is removed from the plume (by, e.g., rainout,
washout and dry deposition) as the plume advects and diffuses
downwind.
4.	Complete reflection of the plume takes place at the earth's
surface, and the plume 1s completely reflected at the mixing
height.
5.	No consideration 1s given to the cumulative effects from consecutive
1-hour periods, and no allowance is made for changes in the wind
with time to affect the path taken by the plume downwind.
6.	No Initial concentration of pollutant.
7.	Continuous uniform emission rate.
8.	Homogeneous horizontal wind field.
9.	No directional wind shear 1n the vertical.
10.	Dispersion coefficients are based on aerometrlc measurements
taken 1n open level to gently rolling terrain.
11.	The pollutant within the plume takes on a gaussian distribution
1n both the horizontal crosswlnd and vertical directions.
Dilution dominates diffusion In the downwind direction.
12. Constant eddy d1ffus1vit1es.
-------

13.	No consideration of aerodynamic effects resulting from the
interaction of the wind with the physical structures nearby the
pi ant.
14.	The input wind direction includes a random variation (-4 to +5°)
superimposed on the value recorded by the National Weather Service
(NWS).
15.	The wind speeds and directions should be hourly averages which
are representative of the region being modeled. (In reality, NWS
hourly observations are not hourly averages but are one minute
averages at the time of observation, usually about 5 minutes
before the hour).
16.	A single mixing height and a single stability class for each hour
are assumed representative of the area.
Please refer to the CRSTER Users' Manual (Reference 1) for a more complete
and thorough treatment of the CRSTER model assumptions and limitations.
The CRSTER model presents a specific set of input data requirements, and
yields output data consisting of pollutant concentrations for a specific
averaging time and receptor location. The input data requirements can be
broadly classified as meteorological factors, source factors and site
factors.
As specified in the "Statement of Assignment," only one year of meteorolo-
gical data (1964) was used in the CRSTER Modeling. Shreveport, Louisiana
was selected as the most representative location for which one year of
meteorological data was immediately available. Meteorological inputs to
the model consisted of the following observations from each day of 1964:
1.	hourly wind speed
2.	hourly wind direction sector (tens of degrees)
3.	hourly ambient air temperature
4.	hourly total opaque cloud cover
5.	twice daily mixing heights
6.	hourly cloud ceiling height
The source data required for calculation of an annual average concen-
tration consists of an annual average emission rate, annual average
stack gas exit velocity and annual average stack gas temperature. (See
Attachment 1). In addition the following emission source parameters were
input into the CRSTER model:
1.	Plant elevation and location
2.	Stack exit diameter
3.	Physical stack height
A description of the site and surroundings is provided 1n Attachment 2.
The effect of terrain was taken Into account. The CRSTER model Includes
a terrain adjustment procedure to simulate the effect of elevated terrain.
-------
sn
For a recaDtor location above the base elevation of the stack, but below
the effective plume height of the stack, the plume centerline is lowered.
In other words, the terrain adjustment procedure decreases the effective
plume height by an amount equal to the difference in elevation between
the plant site and the specific receptor site. The model then uses the
adjusted plume height in estimating concentrations at the receptor. Also
the model considers receptors below base elevation of the stack to be at
base elevation.
CRSTER has available a radial receptor grid. This grid provides for
receptors at five different distances on thirty-six (10° increment)
radials or a total of 180 receptors for each model run. Attachment 3
provides an explanation on how the terrain heights for the receptors
were determined.
. Model Application
The latest EPA modeling guidance recommends that the techniques used
1n the CRSTER model, specifically those included in the model's
default options, should generally be used for rural applications. The
following options and parameters were employed in the CRSTER modeling
for ENSCO's air quality impact analysis:
1.	Final plume rise was used rather than transitional plume rise.
2.	The anemometer height was assumed to be at 7 meters.
3.	Stack tip downwash was not accounted for.
4.	The default set of wind profile law exponents were used.
5.	The emissions were assumed to originate from a hot, bouyant
stack plume.
6.	The rural option was selected.
7.	1964 surface Shreveport data was used.
8.	1964 upper air Shreveport data was used.
9.	Terrain receptors were used. Receptors were placed along 36
radlals originating at the stack location at the intersections
of the following ring distances (km), 0.4, 0.5, 0.55, 0.6, 0.65,
0.7, 0.8, 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, and 1.7.
The following emission source parameters were used 1n this CRSTER
model1ng:
1. Stack base elevation: 244 feet
2. Stack location: Longitude 92°37'48"
Latitude 33a12'17"
-------
51%
3.	Annual stack gas exit velocity: 2.86 m/sec
4.	Annual average stack gas temperature: 346.6K
5.	Physical Stack Height: 60.96 m
S. Stack Exit Diameter: 2.83 m
7.	Annual Average Emission Rates: (See Table 1)
IV. Consideration of Background
The contributions due to background were not considered in the air quality
analysis.
V. Presentation of Results
The results are presented in Table I. When evaluating these results it 1s
Important to consider the assumptions incorporated into the CRSTER model
which were discussed above. In particular, the dispersion coefficients
are valid only for downwind distances greater than 100 meters from the
stack. Beyond a few kilometers downwind, the estimates are based on
limited data and so may be less accurate.
As reported 1n Reference 3, validation studies of the CRSTER model
were performed for five power plants. Until further studies become
available, it may be concluded from these validation studies for
power plants that the CRSTER model is generally accurate withTn a
factor of two. It should be remembered in this application of
CRSTER that the accuracy of the results is dependent upon not only
the model, but on the quality of the input information used in the
application (including both emission information and meteorological
data).
VI.	Acknowledgements
The author wishes to recognize the major contributions of F. Hall
and J. Yarbrough to this study. F. Hall's preliminary work and
prudent advice were Invaluable. Special recognition 1s due to
0. Yarbrough who did the bulk of the computer modeling. The author
1s grateful to D. Speak who provided the terrain elevations. The
author also wishes to thank L. Stevens* J. Sales, and P. Schwindt
for their contributions.
VII.	References
1. Users Manual for Single-Source (CRSTER) Model, July 1977 (EPA-
4-50/2-77-013)
-------

2.	Proposed Revisions to "Guidelines on Air Quality Models," October 1980
3.	Applications of the Single Source (CRSTER) Model to Power Plants:
A Summary, J. A. Tikvart and C. E. Mears, Proceedings of the
Conference on Environmental Modeling and Simulation, July 1976
4.	Validation of a Single Source Dispersion Model, Russell F. Lee and
Michael T. Mills, Proceedings of the Sixth International Technical
Meeting on Air Pollution Modeling and its Application MATQ/CCMS,
September 1975
-------
Table I
Test Day	Annual Average Emission Rates
TCDD (1*10"^ grains/second)
1. Adding in detection limits
NO samples
Absolute Maximum Annual ~
Concentrations (grains/meter )*/
Distance and Direction from Stack
Maximum Annual Concentrations
(grams/meter )* in the Residential
Area/Distance and Direction from
the Stack
for
Day 1
1.46
2.09x10:5/1.2
8.95xlO"{J/1.2
7.52x10/1.2
km
at
310 ***
3l0o
Same
as
absolute
maximum
**
Day 2
6.26
km
at
Same
as
absolute
maximum
A A
Day 3
5.26
km
at
310°
Same
as
absolute
maximum
A A

2. Without detection limits
1.36x10" J*?/1.2
2.42x10"J?/1.2
7.52x10/1.2


310°
310°
310°





Day 1
0.95
km
at
Same
as
absolute
maximum
A A
Day 2
1.69
km
at
Same
as
absolute
maximum
A A
Day 3
5.26
TCDF (1*10 grain/second)
km
at
Same
as
absolute
maximum
A A











1. Adding in detection limits for










NO samples
2.67xl0~|5/1.2
3.44x10 ,c/l.2
1.49x10/1.2


310°
310o





Day 1
18.67
km
at
Same
as
absolute
maximum
A*
Day 2
24.07
km
at
Same
as
absolute
maximum
A A
Day 3
10.44
km
at
310°
Same
as
absolute
maximum
A A

2. Without detection limits
1.05x10"}®/1.2
3.04x10"J1/1.2
1.49x10/1.2


310°
310°
310°





Day 1
7.32
km
at
Same
as
absolute
maximum
A*
Day 2
21.24
km
at
Same
as
absolute
maximum
A*
Day 3
10.44
km
at
Same
as
absolute
maximum
A A
"These are annual arithmetic averages.
"The absolute maximum is located within a residential neighborhood. Therefore, the two maximum concentrations will be at
one location.
***Where 360° is due north of the stack.
-------
£1)
Attachment 1
Sampling and Analysis for PC3s, Dioxins, and Furans
In order to determine the emission rates of the various air pollutants,
EPA required the facilities to conduct stack emission tests. Tests were
conducted by an independent testing firm using methods and procedures
prescribed by EPA.
During all emission tests, EPA representatives were on-site observing the
tests and recording information concerning the operation of the incinerator
and the air pollution control equipment.
Sampling for PCBs, dioxins and furans was performed in the stack utilizing
a modified EPA Method 5 sampling train. The sampling train incorporated
two (2) resin cartridges to adsorb the organic vapors. The resin cartridges
contained XAD-2 resin and florlsil resin.
The resin cartridges were located downstream from the probe and between the
third and fourth impingers. The sampling train was designed to collect
both particulates and organic compounds 1n the stack gas. Both resins were
used so that the second resin would absorb any residual organic compounds
which may have passed through the first resin. For dloxin and furans, the
cartridges which house the resins were water jacketed. Cold water from the
1ce bath surrounding the Impingers was pumped into the jackets which
maintained the resins at a temperature of less than 70°F. Because of the
sensitivity of organic compounds to ultraviolet light, the resin cartridges,
impingers, and Interconnecting glassware were protected from the sunlight by
aluminum foil and opaque plastic.
The stack gases were sampled 1sok1netically for a period of four hours.
At the completion of each sampling period, the stack samples were transported
to the laboratory for analysis. The stack samples were extracted with an
appropriate solvent and analyzed using capillary column gas chromatography and
high resolution mass spectrometry. For dioxins and furans, and added quality
control technique was used. The stack samples were spiked with a labeled
dloxin and furan which allowed the analyst to determine the extraction
efficiency of each sample.
-------

Attachment 2
ENSCO at El Dorado, Arkansas
Surroundings
El Dorado, Arkansas is known as the capital of the petroleum and chemical
industry in the State. This is a relative term, since the bulk of Arkansas
is strictly rural in nature. Ninety percent of the area around El Dorado
is forest.
The area presents a topography of gently rolling hills rising to an elevation
of about 260 feet above sea level, with the intervening valleys at about
180 to 200 feet elevation. The ENSCO facility itself is located to the
east of El Dorado in a small industrial park. There are several small
communities, including churches and other public facilities, adjacent to
the plant boundaries.
The surrounding forest 1s a mixture of deciduous and coniferous trees with
typical upland animal species. There is a small area (3% of the total)
of wetlands to the south. While the effects of chlorinated dibenzodioxins
and furans on such environmental systems have not been studies in detail,
the persistence and potential for bloaccumulation of these materials have
been documented.
Distribution of Population
The population of El Oorado 1s about 25,000. Another 24,000 people live
within the maximum limits to which the air dispersion modeling of ENSCO's
emissions was applied.
El Dorado has the fourth largest per capita income in Arkansas, with
most of this coming from manufacturing, wholesale and retail sales, and
services.
-------
b"Z3
Attachment 3
MAPPING GRID SUMMARY
FOR THE ENSCO INCINERATOR
AIR MODEL DEVELOPMENT
BY
DEBORAH K. SPEAK
U.S. ENVIRONMENTAL PROTECTION AGENCY
REGIONAL ANALYTIC CENTER
DECEMBER 1980
-------

MAPPING ACCURACY
Maps used for plotting the 36 point grid for the application of the air model
were accessed from the Department of Interior, United States Geological Survey
(USGS). The USGS is in the process of processing and editing 7.5 minute
quadrangles for the £1 Dorado area which was utilized in this mapping exercise.
This map was then enlarged further for establishing the grid used in the model.
The National Standards for the horizontal and vertical accuracy of topographies
maps were adopted in 1941. The standards for horizontal accuracy require that
no more than 10 percent of the well-defined map points tested shall he. more
than.1/50 inch (0.5 mm) out of correct position at publication scales or
1:20,000 or smaller. This tolerance corresponds to 40 feet on the ground for
the 1:24,000 scale map that was enlarged by a factor of 2. The standards for
vertical accuracy requires that no more than 10 percent of the elevations of
test points interpolated from contours be in error more than half the contour
interval.
The air model uses elevations from the surrounding £1 Dorado area, tg predict
concentrations expected from the incinerator stack emissions, The elections
are taken at 504 points. Elevation points are the intersections between
circles drawn from around the stack and 36 lines at 10 degree Intervals drawn
from the stack.
Distances drawn from the stack location were 0,4, 0,5, 0,55, 0,6, 0,65, 0.7, 0.8,
1.1, 1.2, 1.3, 1.4, 1.5, 1.6, and 1.7 kilometers, Corresponding eleyatigns
depicting the intersection of the 36 lines at 10 degree interyals with the
concentric circles representing the above distances are attached.
The elevation of the stack base is at 244' MSL, The stack location was
pin-pointed by using an aerial photograph to interpret the stack location on
the topographic map.
Elevations below the stack's base elevation were plotted at the 244' MSL
elevation. Thus, all 244' elevations are at stack ground eleyation or below.
MAPPING METHODS
MAP SCALE*
Series
Scale
One Inch Represents
0,61 km
0,305 km
Draft of 7.5-minute
Enlarged Scale
1:24,000
* Map was enlarged to twice this scale.
-------
km
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
DISTANCES FROM STACK
0.4 0,3 ].1 1.2 1.3 1.4 1.5 1.5 1.7
244
250
244!
244
244
244
244
244
250
244
244
244
244
244
244
244
255
255
244
244
244
244
255
260
250
250
244
244
244
244
244
265
250
255
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
250
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
2*4
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
214
244
244
244
244
244
244
244
244
2*4
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
2*4
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
244
•244
244
244
244
244
244
244
250
250
244
244
244
250
244
244
244
250
250
244
244
244
250
270
265
244
270
255
255
260
265
260
244
244
244
265
244
244
244
250
244
260
244
244
255
255
244
244
244
244
244
244
244
244
255
250
244
244
244
244
244
244
244
250
244
244
244
244
2S0
260
244
244
270
255
244
244
244
244
244
244
265
270
255
244
244
250
260
265
244
244
270
265
250
244
250
260
260
244
244
270
265
260
265
270
270
270
244
24.4
285
290
280
280
290
290
270
244
244
270
275
290
285
270
275
270
244
244
270
265
260
270
265
250
260
244
260
255
244
244
244
250
260
275
244
260
244
244
260
265
265
260
250
244
270
244
244
244
244
244
255
244
2
2
4
1
3
5
6
3
2
2
3
2
6
6
7
7
7
7
5
5
5
0 10 14 9 9	9 13 14 13
Elevations Higher
Than Stack Elevation
At Ground Level
-------
DISTANCES FROM
km
.5
.55
.6
.65
.7
1
244
244
244
250
250
2
244
244
244
244
244
3
244
244
244
244
244
4
244
244
244
244
244
5
244
244
244
244
244
6
244
244
244
244
244
7
244
244
244
244
244
8
244
244
244
244
244
9
244
244
244
244
244
10
244
244
244
244
244
n
244
244
244
244
244
12
244
244
244
244
244
13
244
244
244
244
244
14
244
244
244
244
244
15
244
244
244
244
244
16
244
244
244
244
244
17
244
244
244
244
244
18
244
244
244
244
244
19
244
244
244
244
244
20
244
244
244
244
244
21
244
244
250
255
260
22
244
250
260
265
270
23
250
250
244
250
260
24
244
244
244
244
250
25
244
244
244
244
244
26
244
244
244
244
244
27
244
244
244
244
244
28
244
244
250
250
250
29
244
244
250
250
250
30
244
244
250
250
244
31
244
244
244
244
244
32
244
244
244
244
244
33
244
244
244
244
244
34
244
250
250
250
244
35
255
265
270
270
270
36
250
260
265
265
265
-------
$11
Division 8
-------
52S
Considerations of Risks and
Benefits/Alternatives Concerning PCB
Incineration in Region 6
-------

NOTE: Final analytical data (Appendix S) contains
some small differences from the data used in this
analysis. The effect is to reduce the anticipated
risks in most cases, but in no instance is the
change substantial.
-------

I. Risk Sisman-cs
A. Introduction
Central to the operation of TSCA is the concept of
unreasonable risk to human health or the environment. Figure I
illustrates the process by which this determination is made.
First, the inherent toxicity hazards posed by a chemical or
process are identified. This information is generated by
laboratory experiments, field studies, epidemiological
investigations, etc.
Second, data are gathered on the exposures which are likely
to result. This includes information on the populations which
might be exposed and the frequency, duration, and the intensity
with which such exposures are likely to occur. These data are
observed by direct monitoring and/or mathematical modeling of the
local situation.
Third, the risk is determined as a result of a consideration
of toxicity and exposure. If either factor is zero, the risk is
zero. In the case of cancer toxicity, the Agency has adopted a
method of analysis which permits a quantitative estimate of the
associated risk. In the case of other toxicities, the Agency
presents a qualitative risk assessment.
-------
^31
Fourth, the benefits associated with the use of a chemical
are usually considered. In this instance, however, Congress has
mandated that any benefits of PC3s are outweighed by the risk
they present and that the PCBs will be removed and go through
some sort of a disposal process. The benefits referred to here
are the benefits (or reduction of risks and/or costs) to be
gained from the incineration of PC3s, compared to alternative
nodes of handling the materials, including storage. Such an
analysis should include such factors as economics and timeliness.
Finally, the judgment of unreasonable risk is made by
weighing the risk and the benefits involved. Since a)only some
elements of the risk analysis are quantitative in nature and b)
the benefits are sometimes measured in non-commensurable terms
(e.g./ dollar costs for disposal vs. predicted number of lives
saved), a quantitative statement of unreasonable risk using the
same units of measurement is impossible. From an examination of
the risk/benefit analysis, however, the decision-maker can render
a judgment concerning the "unreasonableness" of any risk.
A. Toxicity Hazard
1. Structure-Activity Relationships between TCDD and TCDF
Since the late 1960s scientific and public interest has
grown as information has become available about the extreme
toxicity of 2,3,7,3 tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD)
(McConnell, 1978; Lucier, 1973).
-------
£ 3>"
This concern has resulted from such events as:
a.	the suspension of the use of Agent Orange herbicide
(contaiminated with 2,3,7,8-TCDD) in Viet Nam;
b.	the cancellation of the use of 2,4,5-T (contaminated
with 2,3,7,3-TCDD) in Sweden;
c.	industrial accidents of the type that occurred in
Seveso, Italy in 1976 which resulted in human expsoure to
2,3,7,8-TCDD?
d.	the suspension of certain uses of 2,4,5-T
(contaminated with 2,3,7,8-TCDD) in the U.S.
While considerable attention has been directed at the
toxicity of 2,3,7,3- and other isomeric forms of TCDDs, only a
limited amount of information has been gathered on the analogous
TCDF compounds; e.g., 2,3,7,3-TCDF.
The data which have been generated on the furan derivative,
however, suggest comparable and parallel toxicities between
related isomeric forms of TCDF and TCDD. This similarity in
toxic properties should be expected in light of the similarity in
Structure
2, 3/7,8-TCDD
2/3/7,8-TCDF
-------
SS3
chemical structures involved. In fact, the work of Poland puts
forth a model of toxicity of certain halogenatad ring systems
vfaich links potency to geometric dimensions of the molecule ana
position of halogen substitution on the rings (Poland,
197371976;1979). This model predicts that the toxic behavior of
2,3,7,8-TCDD and 2,3,7,8-TCDF will be closely related.
Moore cites the simliarity in the types of toxic effects
produced by 2,3,7,3-TCDD and 2,3,7,8-TCDF (Moore, et al, 1979).
At the same time, however, there is a quantitative difference in
their acute toxicities with the furan derivative being less
potent than the comparable dioxin. (See Table 2-1) The table
also illustrates the marked, similar, species variation
associated with the toxicity of these compounds.
Likewise, animal species are also less sensitive to the
chronic effects of 2,3,7,8-TCDF than they are to those of
2,3, 7,8-TCDD. For example, the disease pattern produced by the
two substance is the sane in mice, but the dose of the TCDF
required to produce certain organ weight effects is approximately
30 times greater for the furan than for the dioxin. In addition,
the ability of TCDF to induce hepatic aryl hydrocarbon
hydroxylase (AHH), a biochemical response which has been
associated with the mechanism of toxicity (Poland, 1979), is
about 100-fold lower than that of TCDD (6) (Bradlaw, et al,
1979).
-------
fit
The cause for this reduced susceptibility to TCDF compared
to TCDD is being investigated. Recent speculation suggests that
the variance is related to differences in rcetabolisn, with the
TCDF being enzyraatically attacked and cleared from the body more
rapidly than TCDD.
As noted above, the quantity of data which has been
generated on the furans is less than that on the dioxins. This
is particularly true in the case of cancer. While the evidence
cited here might suggest that TCDF would be a weaker carcinogen
than TCDD, the upper limit of risk (ULOR) assessment developed i«
this paper will utilize TCDD data as a surrogate for missing TCDF
information. The Agency believes that this is a conservative
assumption (possibility erring on the side of public safety),
given the extreme toxicity of 2,3,7,8-TCDD and the reduced
potency of 2,3,7,8-TCDF in the effects which have been studied t&.
date. The toxicity of both classes of substances will be
reviewed below.
2. Relative Toxicity of Isomers
Most of the toxicity hazard information available on TCDDs
is centered on the 2,3,7,8-isomer. The data that do exist on ttm
relative toxicities of the other isomeric forms (Reggiani, 198 0)
suggest that the 2,3,7,8.-isomer is the most toxic and that the
major toxic manifestations appear when at least three of the
positions—
-------
Sir
2,3,7,and 3—are substituted with chlorine (Poland, 1979).
Consequently, in order to more accurately assess the toxicity
associated with TCDD (or TCDF), it is desirable to ascertain, the
distribution of isomeric forms in the sample.
In recent years synthetic and analytical chemists have
succeeded in synthesizing and characterizing all 22 of the
possible TCDD isomers (Nestrick, 1979;1980). The use of these
isomers as standards has made possible the analysis of
environmental samples for the specific amounts of 2,3,7,8-TCDD
and other isomers present.
Unfortunately, the same situation does not extend to the
TCDFs. In this case there are a total of 35 isomers possible.
EPA analysts have fewer than 5 of these standards. Fortunately,
one of the isomers in hand is the 2,3/7,3-TCDF isomer, which is
thought, by analogy to the TCDD isomers, to be the most toxic.
The availability of this standard allows the analyst to determine
if a particular peak in his chroraatogram is possibly the 2,3,7,3-
TCDF isomer. Such a peak, however/ could result from a subset of
the TCDF isomers which have chromatographic properties identical
to the 2,3,7,3-isomer. The number and identity of isomers in
this subset cannot be determined unless all 35 standards are
available. In this paper the Agency will make the conservative
assumption that all TCDFs eluting from the chromatograph with the
2,3,7,8-TCDF isomer are exclusively the 2,3,7,3-isomer.
-------
SSL
3. Toxicity Hazard
As noted in Table 2-1, the acute LD50s of the TCDD and TCXJlT
are similar, with the furan being 3onewhat leas potent that t&s
dioxin. In each case the syndrome associated with death is a
generalized wasting away. Attempts to identify the precise
biochemical mechanises) involved have been unsuccessful (Neal,
1979). As noted earlier, there is considerable variation in th%
responses of different animal species to sinilar doses. Othmx
toxic effects of TCDD and TCDF observed in non-humans are found
in Table 2-2.
Dermal adminstration of these materials at doses as low a*
10-20 ug can result in a persistent acne-like condition, call«<|
chloracne, in both rabbits and humans, which, while not life
threatening, can be disfiguring. Chloracne in hunans has been
considered to be unequivocal evidence of exposure to certain
halogenated hydrocarbons, including TCDD and TCDF.
The involution observed in the thymus of aninals strong
suggests that the TCDD may have an effect on the organism's
ability to defend itself through an immunological response.
There is clear evidence of damage to the liver, reminiscent olf
the action of other chlorinated hydrocarbons..
-------
S*7
The reproductive effects produced in rats (reduced numbers
and weights of offspring) are of particular concern in light of
the low doses at which effects are observed (0.001 ug/kg-day).
In fact, the Agency has taken the position that no "no observed
effect level" (NOEL) has been demonstrated for reproductive-
effects.
Mutagenicity studies have resulted in a range of responses
in the hands of different investigators. A carefully controlled
study from the National Institutes of Environnental Health
Sciences (NIEHS) is about to be released.
Recent studies by the National Cancer Institute have
demonstrated the carcinogenic activity of TCDD in both rats and
mice at low doses (0.001 ug/kg—day) (National Cancer Institute,
1980). Similar data obtained by researchers at Dow Chemical Co.
(Kociba, 1978) have formed the basis of the risk assessment
developed by the Agency's Carcinogen Assessment Group(CAG). The
Dow data have been used in a multi-stage linear extrapolation
model which relates any exposure level with a predicted excess
risk of cancer {Carcinogen Assessment Group, 1980a).
Table 2-3 is a summary of the toxic effects which have been
attributed to TCDD—related exposures in humans.
-------

rhe Agency's position on the toxicity of 2,3,7,3 -tchd is
mora fully developed in the position document on 2,4,5-T (I?A,
1978) and in the 2,4,5-T cancellation hearings now in progress.
It Exposure Assessment
This ULOR/benefit assessment addresses two particular sites
in Region o. While the toxic emissions of interest are the sane
in both cases, the elements of exposure differ. This section
discusses the unique features of the two locations, the
populations at risk, and the emissions.
A. Populations at Risk
1. Rollins at Deer Park, Texas
a. Surroundings
The Rollins Environmental Services facility is located on
the south side of the Houston Ship Channel, about 4,000 feet to
the southwest of the shoreline. It is in the middle of one of
the most densely industrialized areas in the nation. Prom the
Ship Channel's Turning Basin at the edge of downtown Houston, to
the Bayport Bridge, the ship channel is lined on both sides with
industrial complexes. Because of the complete lack of zoning
regulations in the area, these industrial districts are closely
flanked by residential neighborhoods and small incorporated
townships. Within the area of concern for emissions from Rollins
-------
S 3?
there are two townships: Deer Park and Deepwater. These are
located to the south and southwest of the Rollins plant site,
with the most populous portions of both towns lying within three
miles of the site.
The ecological character of the region is that of a coastal
short-grass prairie. There are no known characteristics of the
flora or fauna which would make them especially susceptible to or
immune from contamination by dioxins or furans.
The land use figures shown in Table 3-1 are based on a water
quality planning survey which did not include any of the nearby
townships. The estimates of the total area, including the towns,
are given in parentheses.
b. Distribution of Population
The Agency estimates that up to 25,000 people could
experience exposure to some level of emissions from the
incinerator. Within a radius of one mile the only exposed
population is a group of employees of the nearby industries.
These would be subject to an eight-hour exposure. Within a two-
mile radius is the most densely populated center of Deer Park. A
three mile radius takes in about 1/3 of the most densely
populated section of the town of Deepwater.
2. ENSCO at El Dorado, Arkansas
a. Surroundings
-------
J* *i °
21 Dorado, Arkansas is known as the capital of the petroleum
and chemical industry in the State. This is a relative term,
since the bulk of Arkansas is strictly rural in natura. Table 3-
2 shows that 90% o£ the area around El Dorado is forest.
The area presents a topography of gently rolling hills
rising to an elevation of about 260 feet above sea level, with
the intervening valleys at about 13 0 to 200 feet elevation. The
ENSCO facility itself is located to the east of 31 Dorado in a
small industrial park. There are several small communities,
including churches and other public facilties, adjacent to the
plant boundaries.
The surrounding forest is a mixture of deciduous and
coniferous trees with typical upland animal species. Th«re is a
small area (3% of the total) of wetlands to the south. While the
effects of chlorinated dibenzodioxins and furans on such
environmental systems have not been studies in detail, the
persistence and potential for bioaccumulation of these materials
have been documented.
b. Distribution of Population
The population of El Dorado is about 25,000. Another 24,000
people live within the maximum limits to which the air dispersion
modeling of ENSCO's emissions will be applied.
-------
El Dorado has the fourth highest per capita income in
Arkansas, with most of this coming from manufacturing, wholesale
and retail sales, and services.
5. Exposure
1. Intensity
a. Sampling
In the fall of 198 0 EPA personnel at Headquarters and in
Region 5 devised a stack sampling scheme to assess the presence,
if any, and extent of emissions of tetrachlordibenzo-p-dioxins
(TCDDs) and tetrachlorodibenzofurans(TCDFs) from the Rollins and
SNSCO incinerators. The report prepared by the contractor for
the project contains the details of this activity (TRW, 198 0b)-
At each site a series of three test burns was conducted.
Test 1—Regular toxic wastes with which PCBs would be
burned.
Test 2—Regular toxic wastes .burned along with PC3s.
Test 3—PC3s burned with special supplementary fuel.
The intent of the study was to identify and quantify any
contribution of PCBs to TCDD and/or TCDP production.
-------
JW*-
After the sampling, the probe was washed off and the
resulting liquid analyzed. The impinger catch consisted of cvo
phases: an organic phase and an aqueous phase, each of which was
analyzed. The resins were extracted with hot organic solvent to
give an extract which was also analyzed.
b. Analysis
All of the samples were sent to the Wright State
University laboratory of Dr. Thomas Tiernan in Dayton, Ohio, for
analysis by high resolution gas chromatography, coupled to a high
resolution mass spectrometer (HR GC/MS). The samples were first
analyzed for total TCDD and then specifically for 2,3 ,7,3 -TCDD.
Next, the samples were analyzed for total TCDP on a low
resolution mass spectrometer. Finally, quantitation was made of
2,3,7,8-TCDF.
Table 3-3 summarizes the findings from each test at each
site. Table 3-4 converts these data into the implied mass
emission rates (Region 6, 1980a).
c. Modeling
Humans, of course, are not exposed to incineration
products as they make their way up the stack. Human exposure
occurs when the stack emissions are dispersed through their
encounter with the air and make their way down to the levels
which constitute the breathing zones of people.
-------

Sines it is this breathing zone that is of primary
importance in assessing risk to human health, it would seen that
sampling the ambient environment would provide a more direct
measure than would stack sampling. As a practial matter,
however, even the parts per trillion (ppt) level of sensitivity
achieved by HR GC/MS is not sufficient to detect TCDD and TCDF at
the levels of concentration anticipated to be present in the
breathing zone, without sampling massive volumes of air, a
procedure which itself is open to question for these chemicals.
(Lewis, 198 0).
Consequently, computer modeling has been used to predict the
dispersion of the components found in the stack sampling, with
the assumption that the nature and relative composition of the
components remain unchanged.
Region 6 has conducted the computer modeling for each of the
two sites (Region 6, 198 Ob). Table 3-5 summarizes these results
by citing the predicted maximum annual average concentrations of
TCDD and TCDF a) anywhere within the area and b) in the residential
communities within the area. (In the case of ENSCO these points
coinefde.) Note that these maximum concentrations exit for only a
small portion of the area surrounding the stack. For example, only
people living directly downwind of the prevailing winds will
experience this maximum concentration. In addition, people living
closer to the stack or farther from it will be subjected to lower
concentrations than those living at a distance where the plume
first comes to ground.
-------
fit
This paper makes the assumption that all people (upwind,
downwind, near the stack, far from the stack, etc.) are subjected
to the sane maximum average concentration. While this assunption
is obviously extreme, it does contribute to making this an upper
limit of risk estimation.
2. Frequency and Duration
It is unclear at this point just how frequently ?C3s
will be burned at Rollins and/or ENSCO. As related in Section V
below (Benefits/Alternatives), however, there is only a finite
amount of PCBs 3)currently awaiting dispoal or b)scheduled for
disposal during the next twenty years when the electrical
equipment in which they are contained are removed from service.
It has been estimated (Gunter, 1981) that if Rollins devoted
itself for 20 hours/day to burning liquid PCBs, the entire
inventory of the nation's supply of these materials would be
removed in 35 years. Since the feed rate at ENSCO is less, the
time to incinerate the nation's inventory would be longer by a
factor of 10. Although such a senario is highly unlikely, this
paper will make the worst case assumption that the exposed
population will be exposed for their full lifetimes. Such an
assumption again errs on the side of public safety.
-------

IV. Risk Assessment
A.	Assumptions
As the risk assessment is developed, it is important to keep
in mind some of the underlying, as well as explicit, assumptions
which are being made. These are summarized in Table 4-1.
B.	Qualitative risk assessment
Table 4-2 summarizes
a.	the toxic effects observed in animals treated with
2,3,7,3-TCDD
b.	the lowest dose tested at which ah effect is observed
c.	the dose expected to be received by a 60 kg female
breathing the predicted maximum average annual
concentration of TCDD and TCDF emitted by the ENSCO
f ac il i ty.
d.	the same value at Rollins
e.	the ratio of the animal dose to the ENSCO dose
f.	the ratio of the animal dose to the Rollins dose
These data suggest that a large margin of safety exists
between the animal doses and the maximum conceivable doses
resulting from PC3 incineration at ENSCO and Rollins.
C.	QuantitativeRisk Assessment for Cancer
-------
Recently the Agency's Carcinogen Assessment Group (CAG) has
prepared a quantitative risk assessment for 2,3,7,8 TCDD for use
in the on-going 2,4,5-T cancellation hearing. In addition, CAG
has used some of the same data to prepare a 2,3,7,8-TCDD unit risk
assessment for EPA's Office of Air Quality Planning and Standards
(OAQPS). This latter document estimates that the risk factor
associated with, breathing air containing lxlO"9 g (1 nanogram) of
2,3,7,8-TCDD per cubic meter (1 ng/m^) is .091. This means that
if 1000 people breathe air containing this concentration of TCDD
through a 70-year lifetime, 91 cases of cancer are predicted to
occur in excess of what otherwise would have occured.
Applying this unit risk factor to the predicted
concentrations at 2NSCO and Rollins, Table 4-3 displays the
resulting predicted cancer risks associated with the maximum
annual concentration of TCDD and TCDF during the incineration of
PCSs (Test 2). At ENSCO this risk is between .4 in a million (4
in 10,000,000 chances) and .1 in a million (1 in 10,000,000). At
Rollins the risk of an additional cancer is between 13 in a
million (130 in 10,000,000) and .1 in a million (1 in
10,000,000), depending on whether one is concerned with the
maximum risk in the entire area or just within the residential
communities.
-------
fil
These are the risks which would arise at the point of
aaxinun concentration. People living in other locations would be
subjected to lesser risks
Assuming the worst case, Table 4-4 shows the risks for the
populations involved.
-------
fl?
V. Benefits/Alternatives
As rioted above, by legislating the disposal of ?C3s from the
environment. Congress effectively decided that there were no
benefits associated with these materials that would justify the
risks they posed. Consequently, the only point at issue here is
whether the risks associated with incineration are justified in
terns of alternative forms of disposal. The alternatives to be
considered are:
a.	landfiiling
b.	storage
c.	chemical destruction
As background it is important to keep in mind that the
nation has an inventory of 750,000,000 lbs of PC3s, virtually all
of which is in "totally enclosed" forms (capcitors and
transformers) which will come out of service over the next 20
years. There are 532,000,000 pounds of PC3s which will require
special disposal. 153,000,000 pounds are in capacitors of types
that will not require special disposal. 300 of the 58 2 million
pounds of pure PCBs are in transformers and comprise about 25% of
the total weight of the liquid material from such transformers
-------
S1
-------

3, Storage
The strategy hera is the sane as that for land filling:
reduce the risk by reducing the exposure. A similar list of
risks, costs, and limitations exist for this option.
a.	Risks: The storage of PC3s as outlined in the rule is
intended to be preliminary to disposal; not a substitute for
disposal. With approximately 2 million gallons of ?C3 liquids
presently in storage, and more PC3s being removed from service
everyday, the potential for a spill during handling or transport
associated with storage continues to increase.
In the absence of some permanent disposal method, persons
storing PCBs could limit this expense in two ways: 1)illegal
disposal or 2)legal action against SPA for having issued an
unworkable regulation and asking a court to weaken the existing
disposal regulation to allow such alternatives as the landfilling
of liquids with high PC3 concentrations.
b.	Expense: The estimated costs are $17,00Q/utility for 6
months of storage for transformers and $3 4,000/utility for S
months of storage for.capacitors (Gunter, 19R1).
c.	Limited capcaity; All current capacity has been
utilized* New storage facilities are having to be built.
C. Chemical destruction methods
At least two chemical destruction methods have recently been
developed which hold considerable promise. Advantages assocated
with one or both of these systems include:
-------
SS'i
!• Reduction of risk by removal of hazard. 3y transfornir.n
'•h.e toxic PCBs into innocuous materials, the risk is absolutely
Jnd permanently eliminated.
2. Promise of portability: Portable rigs can be constructed
to go to the PC3s, rather than visa versa. This removes the risk
associated with the transport' of PCBs.
These advantages are accompanied by certain limitations.
1.	Current technology is designed to only handle dilute
(less than 50 0 ppm) solutions of ?C3s The problem of
concentrated PCB systems would remain.
2.	The methods cannot handle the "solid problem"; i.e.,
destruction of the PCBs in capacitors.
3.	Limited accessibility: No group has come forward to
develop the Goodyear process. Sunohio is developing mobile units
which will service PCBs in transformers, but they will be limited
in their ability to visit all sites.
4.	Capacity: The Sunohio unit can handle 1000 gal/day.
Therefore, while chemical destruction methods can be
anticipated to destroy a certain fraction and type of PCBs in the
coming years, the technology does not have the capability nor the
capacity to deal, by itself, with the problem currently
confronting the country.
-------
S-i'X'
Refarencas
Bradlaw, J.A. and J.L. Casterline,Jr., "Introduction
of Enzyme Activity in Cell Culture: A Rapid Screen for
Detection of Planar Polychlorinated Organic Compounds,
J. Assoc. Off. Anal. Chem., _62.* (1979) 904.
Carcinogen Assessment Group, SPA, "Risk Assessments on
2,4,5-T, Silvex and TCDD", 198 0a
Carcinogen Assessment Group, SPA, "Unit Risk Assessment
on TCDD in Air", 193 0b
Courtney, K. and J.A. Moore, "Teratology Studies with
2,4,5-T and TCDD," Tox. and Appl. Pharm, _20 (1971) 396-400
Dellarco, M., Director of the Dioxin Monitoring Program
(DMP), memorandum to Dr. Donald G. Barnes, Co-Chairrian
of the Chlorinated Dioxins Work Group, 1981.
EPA# Position Document on 2,4,5-T, Federal Register, _43_
(1978) 17116-17157
Gupta, B. N. et al, "Patholgoci Effects of 2,3,7,8-TCDD
in Laboratory Animals," Environmental Health Prespectives,
Experimental Vol 5, (1973) 125.
-------
JY3
Gunter, W., PC3 Team Leader at SPA Headquarter3, rr.smcrandun
to Dr. Donald G. Barnes, Co-Chairman of the Chlorinated
Dixoins Work Group, 1981
Xociba, R.J. et al, "Results of a Two-Year Chronic Toxicity
and Oncogenicity Study of 2,3,7,8-TCDD in Rats," Tox. and
Appl. Pharm., A6_ (1978 ) 279-303.
Lewis, R., Branch Chief, SPA Health Effects Research Lab
Research Triangle Park, NC, personal commmunication to
Dr. Donald G. 3arnes, Science Advisor to AA/OPTS at SPA
Headquarters, 198 0.
Lucier, G. W., Ed., Envrionmental Health Perspectives,
Experimental Issue No. 5, DHEW Publication No. (NIH) 74-213,
Sept. , 1973.
Luster, M.I. "Effects of 2,3,7,3-TCDF on the Immune System
in Guinea Pigs," Drug and Chem Tox (1979) 49-6 0.
McConnell, E.E. et al, "The Compartive Toxicity of
Chlorinated Dibenzo-p-dioxins," Tox. and Appl. Pharm.,
44 (1978) 33S-356.
-------
sr-i
Moors, J.A. et al, "Comparative Toxicity cf Three
Haloaenatad Di'cenzofurans in Guinea Pigs, Mice and
Rhesus Monkeys," Annals NY Acad Sci, 320 (1979) 151-153.
Murray, F.J., et al, "Three Generation Reproduction Study of
Rats Ingesting 2,3 ,7 ,3 -TCDD, " Tox. and Appl. Pharn. ^50_
(1979) 241-252.
National Cancer Institute, "3ioassay of 2,3,7,8-TCDD for
Possible Cacinogenicity (Gavage Study)," US Dept of HHS
Publication No. (NIH) 80-1755, (1980).
Neal, Robert A., et al, "Studies of the Mechanism
of Toxicity of 2,3,7,8-TCDD," Annals NY Acd Sci, 320
(1979) 204-213.
Nestrick, T.J. et al, "Synthesis and Identification of the
22 TCDD Isomers by EPLC and GC," Anal Chem J31 (1979)
2273-2281.
Nestrick, T. J. and L. L. Lamparski, "Identification of
TCDD Isomers at 1 Nanogram Level by Photolytic
Degradation and Pattern Recongition Techniques,"
Anal Chem 52 (198 0) 13 65-18 74.
-------

Poland, A. and S. Glover, "Studies on the Mechanism of-
Toxicity of the Chlorinated Di'censo-p-dioxins,"
Environmental Health Perspectives, Experimental
Issue No. 5, DHEW Publication No. 74-213, (1973)
245-251.
Poland, A. and E. Glover, "Stereospecific, High Affinity
Finding of 2,3,7,3-TCDD by Hepatic Cytosol," J Biol Chem
25 (1976) 4936-4945.
Poland, A. et al, "Studies on the Mechanisn of Action of
the Chlorinated Dibenzo-p-dio:cins and Related Compounds,"
Annal3 NY Acad Sci 320 (1979) 214-230.
Reggiani, G. "Toxicology of TCDD and Related Compounds,"
Delivered at the Workshop in the Impact of Chlorinated
Dioxins and Related Compounds on the Environment, Rome,
Italy, Oct 22-24, 1980.
Region 6, "Calculation of Mass Emission Rates at ENSCO
and Rollins," 1981a
Region 6, "Air Dispersion Modeling for ENSCO (El Dorado,
Arkansas) and Rollins (Deer Park, Texas), 1981b.
Smith, P.A. et al, "Teratogenicity of 2,3,7,8-TCDD in
CP1 Mice," Tox. and Appl. Pharn. 33 (197 6) 517-527.
-------
-------
LD (rag/kg
60
Table 1-1 (from Rgggiani,
bodyweight) of various PCBs I
, 1980)
(Ardors)
Species (route)
1221
1232
1242
1248
1254
1260
1262
Mouse (oral)




2*000


Rat (oral)
4*000
4*500
8*700
11*000

10'000
11*300
Adult rat (oral)




4*000-10'000
4*000-10'000

Weaning rat (oral)




1'300
1 '300

Adult rat (i.v.)




350


30 day old rat (oral)




1' 300-1'400


120 day old rat (oral)




2'000-2'500


Rabbit (skin)
501 solution MID
2'000-3'200
1'300-2'000
800-1'300
800-1'300

1 '300-2'000
I' 300-3*200
-------
sfi
Tacla 1-2. Sinrary of Chronic Effects cf FCEs
(Ada?tad fraa National research Council Report en ?C2s)
Health Concerns
Oncogenicity
Mitagenicity
Teratogenicity
Other ^Effects in
Maonals
Environnental Concerns
Ecotoocicity
High chlorinated compounds produced
tgnors in rats and mica, relationship
with EC3 not, always clear
Pines test - 1221, 4 chlorcbiphenyl
significantly mutagenic
Daninant lethal nutations - negative
results
Chranoscnal abnormalities - negative
results.
Effects seen in avian species,
50-200 ug/g
Misuse - sane liver .change with exposure
to high chlorine containing products,
300-500 ug/g
Sat - seme liver changes, niniroal
reproductive effects, 100-500 ug/g
Monkey - Yusho symptoms, altered
reproduction cycles, hyperplastic
gastritis and ulceration, *2.5-5 ug/g
Chicken - seme morphologic deforaity,
reprcduction decline, subcutaneous
edgna, 20-50 ug/g
Dogs - reduced growth, scree liver
changes, 100 ug
Mink - dose response relationship in
growth and reproduction, 10 ug/g"
at levels of 2-5 ug/1
Threshold effects in eg? hatchiiility
of aquatic vertebrates and insertearatss
at levels of 2-5 ug/1
-------
sn
Ecotsxicity ( cont'd.)
Ag^tic sncryo to:cicity evident at 50 ug/1
Pelicsn - scse hepatocellular charges,
100 ug
Wildfcwl - sane reproduction changes,
varies with species, 50-200 ug/g
Additional Trouble-
sane properties
Persistant
Bicacsurmlatss
-------
SCO
TABLE 2-1 Comparative Acuta Oral Toxicity of 2,3,7,3-TCDD
and 2,3,7,S-TCDF
(Moors, 1979 as quoted in Reggiani, 19 80)
Species
Guinea pig
Monkey
Mouse
LD50 TCDD (ug/kg)
0.5~1.5
30-70
150-300
LD50J3
LD50 TCDF (ug/kg.) Approx. ld:>0 j 1
3-10
1000
6000
4
20
25
-------
re i
TA3LZ 2-2 Other Toxic	Effects of 2,3,7,3-TCDD and 2,3,7,3-?CDF
Observed in Non-Humans
Effect	TCDD	TCDF
Chloracne	(Reggiar.i, 1930)	(Reggiani, 1980)
Thymus atrophy	(Gupta, 1973)	(luster, 1979)
Liver effects	(Neal, 1979)	(Luster, 197&)
Reproductive effects	(Murray, 197 8)	——	
Weight loss	(Neal, 1979)	(Moore, 1979)
Mutagenicity				
Cancer	(Kociba, 1978)		
Teratogenic effects	(Courfoey, 1971)	——.—-
-------

TA3L2 2-3 Toxi
-------
TABLE 3-1 Land Use Patterns Near Deer Park, Texas
Land Use
Total
Residential and Built-Up
Conmercial and Service
Industrial
Open (Parks, etc.)
Agricultural
Vacant
Water
Acres
4491 (30,000)
0 (5,000)
I
10 (500)
2658 (15,000)
122 (300)
0
1402 (8,000)
299 (1,200)
% of Total
17
2
50
1
9
27
4
(l)froia Houston-Galveston Area Council 208 planning documents
-------
TABLE 3-2 Land Use Patterns for
El
Dorado,
Arkansas*
Land Use
Acres
% of Total
Total
2,165,374

Urban! and Built-Up
89,963
2.1
Residential
14,896

Agricultural
232,119
6.6
Forests
1,759,129
90
Hater Bodies
13,096

Hetlands
71,166

(1) Prom 208 planning
document

-------
TABLE 3-3A Rollins Analytical Results in Nanograms
(detection limits in parentheses)
Sample Mo. Sample
6-05-81
6-05-82
g=8§=8ii
6-05-85
(1) Composited
Impinger
Probe wash
XAD-2
Florisil
6-06-05 Impinger
fK8^Awash
6-06-081 XAP-2B
6-06-09 Florisil
(1) Composited
6-06-43
6-06-44
6-06—45
6-06-46
Impinger
Probe wash
XAD-2
Florisil
Total
TCPP
1.79(.4)
3.84( .4)
1, 31(.4)
N P( )
6.941-
Test 1
2,3,7,8
TCDD
Test 2
NP(.27)
1.42(.27)
ND(.38)
NP(.48)
1,42t (1.03)
Test 3
ND( .9)
ND( )
NP( .3)
NP(.3)
NPt (1.55)
Train 1
Percent
Recovery
100
100
100
20
Train 1
75
129
107
79
Train
100
80
58
Total
Percent
TCDF
TCPF Recovery
3.5( .8)
.7
lOOt
5(.15)
.5
100
5(.15)
.3
100
ND( .8)
NP
100
13.5H.fi)
1.5t (-8)

6.0(.1)
.8
72
14(1)
1.4
122
2(.07)
.25
100
NP(.9)
NP(.9)
85
22H.9)
2. 41- (.9)

NP(.2)
NP
1004
NP(.5)
NP
60
2(.l)
.2
72
NP(.4)
NO
15
24- (1.1) .2+ (1.1)
-------
TABLE 3-3B EHSCQ Analytical Results in Nanograms
(detection limits in parentheses)



Test 1
Train 1



Sample

Total
2,3,7,8
Percent
Total 2
,3,7,8
Percent
Sample
TCDD
TCDD
Recovery
TCDF
TCDF Recovery
6-05-27
Impinqer
(Sample
expended
in method development)

6-05-28
Probe wash
.16(.06)

146
ND( 1)
ND
50
6-05-29
XAD-2A
.10(.06)

104
ND( 2)
ND
15
6-05-30
XAD-2B
ND(.07)

85
ND( . 1)
ND
70
6-05-31
Florisil
ND(.07)

128
2( .2)
.5
30

26f(.14)


2f (3.1)
. 51- ( 3.1)




Test 2
Train 1



6-05-60
Impinger
• 40 ( )

41
4( .5)
1
100
6-05-61
Probe wash
ND(,25)

97
.5(.09)
.2
75
6-05-62
XAD-2
ND(.38)

134
1.5(.5)
.3
lOOt
6-05-63
Plorisil
ND(.66)

51
ND(.8)
ND
100

.48+ (1.29)


r,l ( .fl)
1.5h ( .8 )




Test 3
Train 1



6-05-74A
Impinger
ND(.35)

71
ND( .5)
ND
lOOt
6-05-74B
Probe wash
ND(.21)

98
ND(.5)
ND
loot
6-05-74C
XAD-2
ND(.2)

94
ND( .8 )
ND
100
6-05-740
Florisil-A
ND(.2)

122
ND( . 4)
ND
15
6-05-74E
Florisil-B
ND(.45)

46
ND(1.9)
ND
88


(1.41)


(2.8)
(2.8)

-------
TABLE 3-3C Summary of Analytical Results at ENSCO and Rollins
in Nanograms
(Totals including detection limit values for NDs in ())
Total 2,3,7.8 Total	2,3,7,8	Total	Total TCDD +
Site	TCDD 	TCnP TCDF	TCDF	TCDD » TCDF	2,3,7,8-TCPF
ENSCO
Test 1	.261- C.4IH-)	2(5.1)	.5(3.6)	2.39(5.5)	.76h
Test 2	.48(1.77)	6(6.8)	1.5(2.3)	6.48(8.6)	2.0
Test 3	ND(1.41)	NO(2.8)	ND(2.8)	NO(4.2)	HD
Rollins
Test 1	6.94(6.941-) 13.5(14.3)	1.5(2.3)	20.4(21.2*)	8.4
Test 2	1.42(2.45)	22(22.9)	2.45(3.4)	23.4(25.3)	3.87
Test 3	N0(1.55)	2(3.1)	.2(1.3)	2(4.7)	.2
-------
Site
ENS CO
Test 1
Test 2
Test 3
Rollins
Test 1
Test 2
Test 3
TABLE 3-4 Mass Emission Rates (in nq/sec) of TCDO and TCDP
Pound at Rollins and ENSCO
(Results including detection limit values for NDs in ())
Total
TCDO
Total
TCDP
Possible
2378 TCDF
Total TCDD
Plus TCDP
Total
possible
TCDD plus
2370 TCDF1
.95* (1.46* )
1.69(6.26)
0(5.26)
7.321- (18 .7+ )
21.24(24.1)
0(10.44)
1.8 3* (6.6* )
5.3(8.1)
0(10.44)
8 .27* (28 .16* ) 2.78*
22.93 (30.4) 6.99
0(15.70)	0
43.74(43.74*) 85.05(90.3)	9.45(14.5) 128.79(134)	53.19
10.34(17.8) 160.13(166.7) 17.8(24.8) 170.47(184) 28.14
0(9.6)	12.44(19.3)	1.2(8.1)	12.44(29.2)	1.2
(1) Sum of columns 2 and 3
-------
TABLE 3-5A Predicted Annual Average Concentration in (10~6ng/m )
of TCDD and TCDP in the Breathing Zone
at Point of Highest Exposure in the Dispersion Area
(Results including detection limit values for NDs in ())
Site
ENSCO
Test 1
Test 2
Test 3
Rollins
Test 1
Test 2
Tes t 3
Total
TCDD
(.209)
.242( .895)
0(.752)
46.8(46.8)
11.1(10.1)
0(10)
Total
TCDP
(2.67)
3.04(3.44)
0(1.49)
91(96.4)
172(178)
13(21)
Possible
Possible
2378 TCDP
.758(1.16)
10(15)
19(27)
1(9)
Total TCDD
plus TCDP
(2.87)
3.28(4.35)
(2.24)
138 (143)
183(197)
12(31)
Total
Possible
TCDD plus
2378 TCDF
2.9
1.4
61.8
30
1
B. Predicted Annual Average Concentration (in 10 ® ng/m^) in the breathing
sone in residential area
ENSCO — Sane as above
Rollins
Test 1
Test 2
Test 3
2.18 ( 2.18+ )
.52( .89)
0(.48)
4.24(4.50)
7.98(8.31)
.62(.96)
.47(.72)
.89(1.24)
.06((.40)
6.42(6.68)
8.50(9.17)
.62(1.45)
2.65
1.41
.06
-------
Table 4-1 Assmuptions
1.	Identity in the carcinogenicity of 2,3,7,fl~TCDD and
2,3,7,8 r-TCDP
2.	All TCDDa and TCDFs have the same carcinogenicity potential
as 2,,3,7,8-TCDF
3.	All enissions from the stack are respirable and are retained
in the body
4.	All the TCDD and TCDF that is respired is retained and is
biologically available to the organism even though some
significant fraction is tenaciously bound in or on the
emission .particles
5.	The composition of the emission products found in the stack
is identical to the composition (but not the concentration)
found at the ground level
6.	The combustion of the PCBs would continue for 24 hours a day
Cor a 70 year 1i fet iroe
7.	The air dispersion modeling is representative of what happens
at the two sites
R. The entire population is subjected to the maximum average
annual concentration found at the point of highest
concentration.
9.	The multi-stage, one-hit, linear extrapolation method of
cancer risk assessment is valid.
10.	Laboratory animals are faithful surrogates for humans
11.	The other assumptions for cancer risk assessment as found in
FH 44 131, July 6, 1979 are valid
12.	The assumptions made by the Cancer Assessment Group in
determinimg unit risk assessments for the Air Programs
are valid (Cancer Assessment Group, 1980b).
-------
Table 4-2 Qualitative Risk Estimate
Effect
Thymus atrophy
Reproductive
Teratogenic!ty
Lowest
Lowest 237B
TCDD Dose
(ng/kg/day)
* 3
4
Maximum
TCDD + TCDP
BN3CO Dose1
jng/kg/day)
1.45xl0~6
1.45xl0~6
»~6
1.45x10'
(1)	(Table 3-5A value of 4.35xl0~6
(2)	(Table 3-5A value of 197x10
(3)	Vos, 1973
(4)	Murray, 1978
(5)	Smith, 1976
n
ng/m
TCDD + TCDP
Rollins Dose Column 2 value Column 2 value
(ng/kg/day) Column 3 value Column 3 value
66x10"®
»-6
66x10'
66x10
-6
1,000,000
200,000
200,000
g/m3) x (20 m3/day) x (1/60 kg)
/m ) x (20 m /day) x (1/60 kg)
30,000
5,000
5,000
-------
TABLE 4-3 Range of Risks
Expressed in. chances out of a million
A. At point of Maximum concentration in the dispersion model area
Maximum possible values of Analytical values of
Site	total TCDP ami TCPF1 TCDP plus possible 2,3,7,8-TCPF2
ENS CO
Vest 1	.3	.03
Test 2	.4	.1
Test 3	.1	0
Rollins	^
Test 1	13	6
Teat 2	18	3
Test 3	3	.1
B. At point of maximum conentration in residential area
ENS CO — Same' as above
Rollins
.2
.1
.005
Test 1	.6
Test 2	.8
Test 3	.1
(1)	Based on values of column 5 in parentheses in Table 3-5
(2)	Based on values of column 6 in Table 3-5
-------
TABLE 5-1 Amount of Materials with PCBs >500 ppin
(Gunter, 1981)
Type
Transformer liquids
Capacitors
Ready for	Retiring within
Immediate disposal 20 years
20,000,000 lbs 280,000,000 lbs
10,000,000 lbs 440,000,000 lbs
-------
Control
Option
Unreasonable
"\Rlsk?
Benefits
Pig I Unreasonable risk determination
-------

EXTERNAL REVIEW COPY
JULY 31, 133C-
United Stdtas
Environmental Promotion
Agency
Off leg of Research ar.q Development
PR08LEM ORIENTED REPORT
METHOD FOR DETERMINING THE UNIT RISK
ESTIMATE FOR AIR POLLUTANTS
Offlca of Air Quality Planning tnd Standard*
Off1ca qf Air, Noisa and Radiation
U. S« Environmental Protection Agtncy
Carclncgtn Aas«ssm«nt Group
Offlca of Haalth and Invironmmtal Assassnent
Off 1c# of Rtsaarch and Oevel oprrant
U. S. Environmental Protection Agwcy
Prtpartd for:
Praparad by:
-------
S1(o
THE CARCINOGEN ASSESSMENT GROUP'S
METHOD FOR OETER.MIMIMS THE UNIT RISK ESTIMATE
PARTICIPATES MEMBERS
Elizabeth L. Anderson, Ph.O.
Larry Anderson, Ph.O.
Oolph Am-fcar, 8.A.
Steven 3ayard, Ph.O.
David L. Sayllss, M.S.
Chao W. Chen, Ph.O.
John R. Fowle III, Ph.O.
Bernard HabeTman, O.Y.M., M.S.
Charalfngayya Hirwwth, Ph.O.
Chang S. Lao, Ph.O.
Robert McSaughy, Ph.D.
Jeffrey Rosenblatt, 8.S.
Ohara V. Singh, O.Y.M., Ph.O.
Todd W. Tftorslund, Sc.O.
FOR AIR POLLUTANTS
fi!**oy £. Aioer:,
Chai man
July 31, 1980
-------
S 77
METMOO FOR DETERMINING THE UNIT RISK ESTIMATE
FOR AIR POLLUTANTS
INTxCOUCTIQN
The unit risk astimata far an air pollutant is defined as the lifetime
cancer risk occurring in a hypothetical population in which all individuals are
exposed continuously from birth throughout their lifetimes to a concentration of
1 ug/m^ of the agent in the air which they breathe. This calculation is done
to estimate in quantitative tarms the effectiveness (potency) of the agent as a
carcinogen. Unit risk estimates are used for two purposes: 1) to compare the
carcinogenic potency of several agents with each other, and 2) to give a crude
indication of the population risk which might be associated with air exposure to
these agents, if the actual exposures art known. These two uses have different
limitations.
In order to use these estimates intelligently, the nature of tht source data
used and the assumptions necessary ts derive- tht estimates must be clearly
understood. This appendix discusses the general approach and tht assumptions
common to most unit risk estimates. Tht cr«d*nce ont can ascribe to each risk
estimate depends heavily on tht quailof tht studies on which tht estimate fs
based and tht relevance of these studies to the evaluation of air exposure in
humans.
PROCEDURES FOR DETERMINATION OF UMtT RISK
Th* data used for the quantitative estimate is one or both of two types: 1)
lifetime animal studies, and 2) human studies where excess cancer Hsk has been
assodated wi th exposure to .tht agent. In animal studies 11 1 j assumed, unless
evidence exists to the contrary, that if a carcinogenic response occurs it the
dose levels used in the study, then responses will at so .occur at *U Uwerioses
Repro
-------
S7?
with an incidence detamined by the axtrjpola- on modal discussed below.
A. Choice of Model
There is no really solid scientific basis for any »iathasiatica] extrapo!atio"
modal which relates carcinogen exposure to cancar risks at the extremely low
concentrations that must be dealt with fn evaluating environmental hazards* ?°r
practical reasons such low levels of rfsk cannot be measured directly either by
animal experiments or by epidemiologic studies. We must, therefore, depend on
our current understanding of the mechanisms of carcinogens for guidance as to
which risk model to use. At the present-time the dominant view of the
carcinogenic process involves the concept that most agents which cause cancer
also cause irreversible damage to DMA. This position is reflected by the fact
that a very large proportion of agents which cause cancer are also mutagenic-
There is reason to expect the quanta! type of biological response that is
characteristic of mutagenesis 1s associated with a linear non-threshold
dose-response relationship. Indeed, there is substantial evidence from
mutagenesis studies with both 1on1rfng radiation and a wide variety of chemical
that this type of dose-response model is the appropriate one to use. This is
particularly true at thef 1ower end of the dose-response curve; at higher dos*s»
there can be an upward curvature probably reflecting the effects of mul tl stag*
processes on the mutagenic response. The linear non-threshold dosa-response
relationship is also consistent with the relatively few epidemiological stud4«*
of cancer responses to specific agents that contain enough information to
the evaluation possible (e.g., radiation induced leukemia, breast and thyroid
cancer, skin cancer induced by arsenic in drinking water, liver cancer indue#*
by-a frataxin in the diet). There is also sotte evidence from animal
e#e|Hments that is consistent with the linear non-threshold aodel (e.g., 1 iv#r
1
-------

tumors induced in falca by 2-acatyl aminofluo rsne in the Urge scale. E33oi study
at tiie National Canter for Toxicologies! Research and the initiatton stage of
the two-stage cardnogenesi 5 mode' in rat liver and mouse skin).
Secause it has the best, albeit limited, scientific basis of any of the
current mathematical extrapolation models, the linear non-threshold model has
been adopted as the primary basis for risk extrapolation to low levels of the
dose-response relationship. The risic estimates made with this model should be
regarded as conser/ative, representing the most plausible upper limit for the
risk, i.e., the true risk is not likely to be higher than the estimate, but it
could be smaller.
The mathematical formulation chosen to describe the linear, non-threshold
dose-response relationship at low doses is the improved multistage model
developed by Or. X. Crump. This model employs enough arbitrary constants to be
able to fft almost any mcnotonically increasing dose-response data and it
incorporates a procedure for estimating the largest possible linear slope (in'
the 955 confidence limit sense) at low extrapolated doses that is consistent
with the data at all dose levels of the experiment.
3. Description of the Extrapolation Model
Let P(d) represent the lifetime risk (probability) of-cancer at dose d. The
multistage model has the form

P(d) • I - exp C-(<50 + * «2<*2 * ... +
where
*1 £ Q, f • 0, 1. 2, ..., k
-------
S2o
Equivalsntly,
A(d) • I - exp C-Uid + qz^z * ... * qjtd^j]
where
A( d) * P (d) • ? (o),
i - HIO)
is the extra risk over background rata at dose d.
The point estimata of the coefficients qf, i * 0, l, 2, ..., k and
consequently the extra risk function A(d) at any given dose d is calculated by
maximizing the likelihood function of the data.
The point estimate and the 952 upper confidence limit of the extra risk A(d)
are calculated by using the computer program GLOBAL 79 developed by Cramp and
Watson (1979). The calculation proceeds as follows: Let La be the maximum
value of the log-liklihood function. The 9$ upper confidence limit for the
. extra ri sk A{d) has the form
Ay(d) ¦ .1 - exp C-{
-------

quantities qz> ^3» • ••. 5k an maximum likelihood estimates of the
other coefficients given aj, equal ta q^*. This approach of computing the
upper confidence limit for the extra risk A(d) is an improvement on the Crump at
al. (1977) nodal. At low dosas, the exponent g(d} * q^d * q2a^ * ...
¦^kd* is dominated by the linear tars qx*d and hence
Ay{d) * 1 - exp(- qi*d) « qj*d
Therefore, the upper confidence limit for the extra risk A(d) at low dosas is
always linear. This is conceptually consistent with the linear non-threshold
cancapt discussed earlier. The slope qj* is a measure of the potancy of the
chemical 1n inducing cancer at low doses.
In fitting the dose-response model, the number of tarns in the polynomial
g{d), is chosen equal to (h-1) where h is the number of dose groups in the
experiment including the' control. group.
Whenever the multistage model does not fit the data sufficiently well, data
at the highest dose is deleted and the model is refitted ta the rest of the
data. This is continued until an acceptable fit ta the data is obtained. To
determine whether or not a fit is acceptable, the chl-square statistic
h
X2 * . (X< - IhM*
* Vt tl -1 #,)'
1»1
is calculated where is the number of animals In the 1^ dose group, Xf
is the number of animals in the 1th dose group with a tumor response, ?\ Is
the probability of a response in the 1th dose group estimated by fitting the
multistage model to the data, a nd h is the number of remaining groups, the fit
is determined ta be unacceptable whenever X2 is larger-than the cumulative 99%
-------
S'Zl-
point of the chi-square 
-------

Ai, , ..., ^ i s defi ned as
(A-t x % x„,.. x
3.	If two or rcar® significant tumor sites are observed in the same study and
if the data are available to us, the number of animals with at least one of the
specific tumor sites under consideration is used as incidence data in the model
4.	Following the suggestion of Mantel and Schneiderman (1977) we assume
that mg/surface area/day is an equivalent dose between species. Since to a
close approximation the surface area is proportional to the 2/3 rds. power of the
weight as would be the case for a perfect sphere, the exposure in rag/day per 2/3
rds power of the weight is also considered to be an equivalent exposure. In an
animal experiment this equivalent dose is computed in the following manner.
Let
* duration of experiment
le * duration of exposure
m » average dose per day in mg during administration of the agent
(I.e., during 18)
W * average weight of the experimental animal
7
-------
Then, the lifetime average exposure is
d
Often exposures are not given in units of mg/day and it becomes necessary *3
convert the given exposures into mg/day. For example in most feeding studies
exposure is in terns of ppm in the diet. In this case the exposure in rag/day
m » ppm x F x r
where ppm is parts par arm ion in the diet of the carcinogenic agent and F is
the weight of the food consumed per day in kgms and r is the absorption
fraction. In the absencs of any data to the contrary r is assumed to be equal
to one. For a uniform diet the weight of the. food consumed is proportional to
the calories required which in turn is proportional to the surface area or
Z/Znis power of the weight, so that
m orppm x Vf2/3 x r or
¦ a PP»
rtp7T"
As a result, ppm in the diet is often assumed to be an equivalent exposure
between species. However, we feel that this is not justified since the
caloriesAg of food is very different in the diet of man compared to laboratory
animals primarily due to moisture content differences. Instead we use an
empirically derived food factor f * F/W which is the fraction of a species feo#
a
-------
S'9S*
we/ght that is consumed per day as food. We use the rates given be w.
Species W	f
Man	70	0.023
Rat	0.35	0.05
Mice	0.03	0.13
Thus, when the exposure 1s given as a certain dietary concentration fn pom the
exposure in mg/Vf2/3 1s
« . otxF . |»Pm x fx* ,	f wl/3
r xW3	W*'4
When exposure 1s given 1n terms of mg/kg/day * m/Yfr * s the conversion is
simpl y
rtf
arr
When exposure is via inhalation, the calculation of dose can be considered
for two cases where I) the carcinogenic agent 1s etthar a completely water
soluble gas or an aerosal and 1s absorbed proportionally to the amount of air
breathed in, and 2} where the carcinogen 1s a poorly wetar soluble gas which
reaches an equilibrium between the air breathed and the body comparttfcnts.
After equilibrium 1s reached, the rate of absorption of these agents is expected
to be proportional to the metabolic rate, which in tum is proportional to the
rate of oxygen consumption, which in turn is a function of surface area.
S
-------
J-/6
Case I
Agents that are in tM form of parti cul ate'inattar or virtually csmpietaly
absorbed gasas such as SO? can reasonably be expected to be absorbed
proportional to the breathing rate. In this case the exposure in fig/day siay ttai
expressed as
m * I x x r
where
I • inhalation rata per day in
v * mg/m3 of the agent in air
r * the absorption, fraction
The inhalation rates, I, for various species can be calculated from the
observations (FASE3 1974) that 25 gm :tries breathe 34.5 liters/day and 113 gn»
rats breathe 105 liters/day. For mice and rats of other weights, W (in
kilograms), the surface area proportionality can be used to find breathing
in m^/day as follows:
For nice, 1 « 0.0345 (W/0.025)2/3 in3/day
For rats, I « Q-105 (W/Q.113)2/3 m3/day
For hunans, the values of 20 m^/dayf is adoptad as a standard breathing rata
(ICS? 1377).
The equivalent exposure 1n iag/w2/3 for thes* agents can be derived frea
the air intake data in way analogous to the food intake data.
•From "Recommendation of the International Commission on Radiological
Protection", page 3, the average breathing rate is 107 car per 3 hour work
day and 2 x 10' caH in 24 hours.
10
-------
jrn
The empirical factor? for the air intake pe-r kg per day, i * I/W based upon
trie previous stated relationships are tabulated below
Soecies	V	i » t/U
"*7Ta n	ITS iJT2T"
Rat	0.35 Q.o4
Mice	0.03 1.3
Therefore, for particulates or completely absorbed gases, the equivalent
exposure in *g/w2/3 U
^ ' yZTT" -jr * ^ -
In the absence of experimental information or a sound theoretical argument to
the contrary, the fraction absorbed, r, is assumed to be the same for all
species.
Case Z
The dose in mg/day of partially soluble vapors is proportional to the
consumption which in tarn is proportional to W^/3 and al so -proportl onal to the
solubility of the gas in body fluids, which can be expressed as an absorption
coefficiennt r for the gas- Therefore, expressing the 0^ consumption as 0^
* '
-------
fgs
concentration in pptn or ug/m^ in experimental animals is equivalent to the-
same ccncantratian in humans. This is supported by the observation that the
minimum alveolar concentration that is necessary to produce a given "stage" of
anesthesia is similar in-man and animals (Oripps, et al. 1975). When the
animals were exposed via the oral route and human exposure is* via inhalation or
vice-versa, the assumption is made, unless there is pharmacofcenetlc evidence tfl
the contrary, that absorption is equal by either exposure route.
5. If the duration of experiment (Ls) is less than the natural lifespan of
the test animal (I), the slope qj* or more generally the exponent g(d) is
increased by multiplying a factor (l/l€}3. v<« assume that if the average
dose, 0, is continued, the age specific rate or cancer will continue to increase
as a constant function of the background rata. The age specific rates for
humans increases at least by the 2nd power of the age and often by a
considerably higher power as demonstrated by Qoll (1971). Thus, we would exp*1-
the cumulative tumor rats to increase by at least the 3rd power of age. Us1«9
this fact we assume that the slope qj* or more generally the exponent g(d),
would also increase by at least the 3rd power of age. As a result, if the sloPe
qi* Co** gU)I fs calculated at age La, we would expect that if the
experiment had been continued for the full lifespan, L, at the given average
exposure, the slope q^* Cor g(d)] would have been increased by at least
(l/U)3.
This adjusttnent is conceptually consistent to the proportional hazard model
proposed by Cox (1972) and the time-to-tumor model considered by Crump et al.
(1979) where the probability of cancer at age t and dose d 1s given by
P(d,t) * 1 - expC-f(t) x 3(d)]
•*»
-------
rsg
This rcors refined approach would be used in the calculations of unit risk wh n
the data are available.
CALCULATION CF OF THE UNIT RISK
The risk associate with d mg/kg2/3/day as noted previously is
Au(d) » l-ex?C-(qi* d + % d2'+ ... +• q^d^)]
A "unit risk" in units X is simply the risk corresponding to an exposure of X *
I. To estimate this value we simply find the number of mg/kg^/Vday
correspondlng to one unit of X and substitute this value into the above
relationship. Thus, for example if X 1s 1n units- of ug/tn3 in the air we have
that for
case (I) d ¦ 0.23 x 70^ x ^q-3 , x i0-3
and for case (2) d * 1 when ug/ra^ 1s unit used to canputa parameters in
animal experiment.
If exposures ar« given in terms of ppm in air we may simply use the fact
that
1 ppm * 1.2 x molecular weight (gas) mg/m^
moiecuiar weignt lair)
Mote, an equivalent method of calculating unit risk would be to use mg/kg for
the animal exposures and then increase the jtft polynomial coefficient by an
amount
(Wh/W,)j/3 J • 1,2,...,k
and use ag/kg equivalents for the unit risk values.
-------
Slo
ESTIMATION GF UNIT RISK 3AScD ON HUMAN QATA
If human apidenrialacy studies and sufficiently valid exposure information
are available for the compound, they are always used 111 S 0*t!3 way. If they show
a carcinogenic effect, the data are analyzed ta give an estimate of the linear*
dependence of cancer rates on lifetime average dose, which is equivalent to the
factor &h< If they show no carcinogenic effect when positive animal evidence
is available, then it is assumed that a risk does exist but it is smaller than
could have been observed in the epidemiology study, and an upper limit of the
cancer incidence is calculated assuming hypothetical ly that the true incidence
is just b«low the level of detection in the cohort studied, which is determine^
largely by the cohort size. Whenever possible, human data ar« used in
preference ta animal bioassay data.
In human studies, the response is measured in terms of the relative risic
the exposed cohort of individuals compared ta the control group. In the
analysis of this data it is assumed that the excess risk, or relative risk minus
one,	- I, is proportional to the lifetime average exposure, X^, and
that it is the same for all ages. It follows that the lifetime risk in the
general population exposed to a lifetime average concentration Xg, P(X2h is
equal ta CR(X^) - UX^/X^ multiplied by the lifetime risk'at that site in
the general population. The unit risk estimate 1s the value of P when X2 is ^
ug/m^. Except for an unusually well documented human study, the confidence
limit for the excess risk P is not calculated, due to the difficulty of
accounting far the uncertainty inherited in the data {exposure and cancer
response).
Li
-------
ft) i
JNTSKPREWTCN CF UNIT RISC
Tfie unit risk estimate is a rough indication of the relative patency of a
given agent compares with other carcinogens. The comparative potency of
different agents is more nil aisle when the comparison is based on studies in the
same test species, strain, and sex and by the same route of exposure, preferably
by inhalation.
For several reasons the unit risk estimate is only an approximate indication
of the absolute risk in populations exposed to fcncwn air concentrations. First,
there are important species differences in uptake, metabolism, and organ
distribution of carcinogens, as well as species differences 1n target site
susceptibility, immunological responses, hormone function, dietary factors, and
disease. Secondly, the concept of equivalent doses for humans compared to
animals on a mg/surfaca area basis is virtually without experimental
verification regarding carcinogenic response. Finally, human populations an
variable with respect ta genetic constitution and diet, living environment,
activity patterns, and other cultural factors.
15
-------

REFSRENCES
Albert, R.E.at aT. 1977.. Rationale'developed by the Environmentar Protection
Agency for the assessment of carcinogenic risks. J. Mat]. Cancer Inst.
53:1537-1541.
Cordle, r., ?. Corneliussen, C. Jellinek, 3. Hack!ay, R. Lahman, J.
McLaughlin, R. Rhoden, and R. Shapiro. 1373. Human exposure to
polychlorinated biphenyls and polyorominated biphenyls. environ. Health
Perspect. 24:157-172.
Cox, C.R. 1972. Regression model and life tables. J. Roy. Stat. Soc. 3
34:187-220.
Crump, K.S., .4.A. Guess, and l.L. Deal. 1277. Confidence intervals and test of
hypotheses concerning dose-respansa relations inferred from animal
carcinogenicity data. Biometrics 33:437-451.
Crump, K.S. 1979. Qose-response problems in carcinogenisis. 3iometrics
35:157-167.
Crump, X.S., W.W. Watson. 1979. GLOBAL 79. A fortran program to extrapolate
dichotcmous animal carcinogenicity data to low dose. Natl. Inst, of
Environ. Health Science. Contract Mo. l-SS-2123.
Crump, X.S. 1980. An improved procedure for low-dose carcinogenic risk
assessment from animal data. J. of Environ. Pathology and Toxicology (in
preparation).
Doll, R. 1971., Weibull distribution of cancer. Implications for models of
carcinogenesis, J; Roy.. Statistical Soc. A 13:133-156.
Qrfpps, Robert 0., J.E. Eckenhoff, and L.D. Yandara. 1977. Introduction to
anesthesia, the principles of safe practice. 5th Ed. VI.3. Saunders Company,
Phil. Pa. pp. 121-123.
FASE3. 1974. Biological Data Books, 2nd ed. Vol. III. Edited by Philip L.
Altaian and Gorothy. S. Olttaen. Federation of American* Societies for
Experimental Biology. Bethesda, MO. Library of Congress Mo. 72-37738.
Guess, H., et al. 1977. Uncertainty estimates for low dose rata extrapolation*
of animal carcinogenicity data; Cancer Res. 37:3475-3483.
Interagency Regulatory Liaison Group. 1979. Scientific bases for identifying
potential carcinogens and estimating their risks. Feb. 6, 1979.
International Commission on Radiological Protection. 1977. Recommendation of
the International Commission on Radiological Protection, Pub. Mo. 26,
adopted Jan. 17, 1977. Pergammon Press, Oxford, England.
Mantel, M., and M.A. Schneideman. 1975. Estimating "Safe Levels, A Hazardous
Undertaking. Cancer Res. 35:1379-1336.
1 e
-------

HClm~ '-375.^Guidelines !»q- carcinogen bicassay in small rodents. National
uancer institute Carcinogenesis. jechnical Report Series, no.. L "sorjary
U.S. "A. 1975. Interim procedures and impact assessments of susce-"^
carcinogens, -sderal Register '4-1:2liQ2, May 25, 1375.	~
17
-------

Division 9
-------

f	- UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
\	/	REGION VI
1201 BUM STREET
DALLAS, TEXAS 7S270
Mr. William B. Philipbar, Jr., President
Rollins Environmental Services, Inc.
One Rollins Plaza
Wilmington, Delaware 19899
Dear Mr. Philipbar:
Based on the results of test burns conducted atr'your facility in October
1979 and August 1980, I hereby grant approval to you to incinerate PCBs
at the Rollins Environmental Services facility at Oeer Park Texas
under the authority of Section 6 of the Toxic Substances Control Act
The approval is subject to the conditions specified in the attachment to
this letter.
Violation of any condition inciuaea as part of the approval will subject
Rollins to enforcement action under the appropriate statute and/or
termination of the approval. Furthermore, this approval may be with-
drawn or further conditions may be added anytime I feel that the oper-
ation of the incinerator presents an Unreasonable risk of injury to
health or the environment from PCB incineration. This approval shall
be effective on March 4, 1981.
If you have any questions, please contact me.
Sincerely,
Adlene Harrison
Regional Administrator
cc: Governor Bill Clements
State of Texas
Mr. Bill Stuart, Director
Texas Air Control Board
Mr. Harvey Davis, Director
Texas Department of Water Resources
-------

Condition for Incineration of Liquid PCBs
at Rollins Environmental Services
of Deer Park, Texas
At all times during PCB incineration the incinerator shall meet
the requirements specified in 40 CFR 761.40(a).
The PCB feed rate shall not exceed an average of 2439 lbs/hr
calculated over the actual period of incineration of the PCB
material.
The CO- level shall be measured every 24 hours during PCB
incineration.
The flow of PCBs to the incinerator shall stop automatically
under any of the following conditions:
a.	the temperature drops below USQ^C at the exit end
of the combustion zone,
b.	the CO level shall not exceed 10 times the CO-level,
and in any event shall not exceed 100 ppm (calibrate
monitors every 24 hours during PCB incineration by zero
and span gas), or
c.	the 0. drops below 35 (monitors shall be calibrated
every 24 hours by zero and span gas).
The water scrubber shall at least achieve 99% HC1 removal efficiency,
and shall not emit greater than 40 lbs/hr HC1. Chorine input shall
be determined by the appropriate ASTM method. Stack gases shall
be sampled using impingers containing NaOH, and analyzed for
chloride using the Argentimetric method as specified in Standard
Methods for Water and Waste Water, current edition.
Total particulate emission shall not exceed State emission limit
for particulate emissions during PCB incineration.
All records and data shall	be maintained in accordance with
40 CFR 761.45(b), (c), and	(f).
All PCB storage facilities	shall comply with 40 CFR 761.42.
These requirements include	implementation of an SPCC Plan for
PCB storage and feed tanks	as described under 40 CFR 761.42(7)(ii).
Any PCB container used for PCB transport, storage, or disposal
shall not be used for any other purpose unless decontamination of
the container complies with 40 CFR 761.43. All decontamination
wash fluids shall be incinerated*
All PCB articles, equipment, and containers shall be properly
marked according to 40 CFR 761.44.
-------
vT3"7
11.	Rollins shall comply with NPDES requirements for PCS discharges.
12.	Rollins shall comply with all State and local permits for the
incineration of wastes during PCB incineration.
13.	Rollins personnel safety requirements and procedures for PCB
handling, storage, transport, and disposal shall comply with
OSHA requirements.
14.	Any violation or noncompliance with the conditions referenced
herein may result in suspension or revocation of the final
approval for incineration of PCBs.
15.	Rollins must comply with 40 CFR Part 265 and Part 122 of the
May 19, 1980, Hazardous Waste Regulations ("Interim Status
Standards" and "RCRA Hazardous Waste Permit Regulations").
16.	EPA may require annual testing or monitoring of the facility for
PCBs, HC1, particulates and up to 3 organics identified by EPA.
Written reports discussing the results of the testing or monitoring
shall be submitted to EPA within 180 days. Any modification to the
approved facility which may result in increased or changes 1n types
of emissions may require additional testing or monitoring.
17.	Rollins must maintain a negative draft sufficient to prevent
fugitive emissions from the kiln or after burners.
18.	Should an Air Stagnation Advisory be Issued which includes Harris
County, Texas the facility shall upon notification by EPA or
the State cease incineration of PCBs until an "all clear" is issued
in accordance with the Texas SIP.
19.	Rollins will through contractural agreement, hire a private engin-
eering company to conduct regular monitoring of the Rollins fa-
cility. The selection of the private engineering company must
be approved by EPA, Region 6. The contract scope of work shall at
least include the scope of work prepared by EPA, Region 6 1n order
to ensure that all monitoring as a condition of the approval 1s
included. The private enginering company will monitor the Rollins
facility and will report the monitoring results to EPA. The contract
scope of work and reporting will be 1n accordance with the attached
Scope of Work for Monitoring. This condition must be satisfied
prior to incineration of PC8s under this approval.
20.	The conditions of this authorization are severable, and if any
provision of this authorization, or any applications of any pro-
vision, Is held invalid, the remainder of this authorization shall
not be affected thereby.
-------

Scope of Work for Monitoring the.Rollins Environmental Services Facility
Deer Park, Texas
I. Background
The purpose of this scope of work is to ensure that the incineration
of PCBs is continuously accomplished consistent with the conditions
set forth in the letter of approval dated	 from EPA,
Region 6 regarding the incineration of PCBs at Rollins Environmental
Services in Deer Park, Texas. Rollins will contract with and com-
pensate a registered professional engineer who is mutually accept-
able to both Rollins and EPA. The contractor will monitor Rollin's
operations with regard to the conditions specified in the above
referenced letter. The monitoring will be conducted by the con-
tractor at random intervals, on an unannounced basis. Rollins will
provide the contractor immediate access at any time to its entire
facility. The monitoring under the proposed contract will be
consistent with the principles set forth in EPA's Toxic Substance
Control Act (TSCA) PCB Monitoring and Disposal Regulation Interim
Inspection Guidance regarding TSCA inspections for the incineration
of PCBs. This contractual arrangement 1s not and should not be
construed as a substitute for periodic TSCA inspections conducted
by EPA personnel or inspections conducted by EPA's authorized
representatives.
Prior to awarding the contract, Rollins has agreed to provide EPA
an opportunity to review and comment on the terms to be Included
therein and, if necessary, EPA may require Rollins to modify the terms
of the contract to ensure consistency with the January	, 1981,
letter of approval mentioned above.
II. Specific Requirements of Monitoring
The monitoring shall Include at a minimum:
1.	A general monitoring of the Incinerator Including the
PCB shredder area, rotary kiln, the condition and main-
tenance of the main combustion chamber area, scrubber and
stack (including plume opacity readings). The contractor
will report on the overall operating condition of the
incinerator (eg., Instrument conditions, scrubber performance,
stack opacity, evidence of PCB spillage, negative draft,
maintained to prevent fugitive emissions).
2.	The contractor shall examine and record the feed rate
records for the past 48 hours of PCB Incineration and
determine the actual feed rate of PCBs fed to the Incin-
erator. This Item will include copies (If any) of
PCB analytical results retained by Rollins of all PCB
materials fed to the Incinerator during the previous
48 hours.
-------

3.	The contractor shall examine the record and CCL level in
the combustion gases upstream of the scrubber, the PCB
cutoff level set for the CO, 0^, and temperature instru-
ments, and shall observe that the cutoff mechanisms are
operable. Also, the contractor shall check the continuous
0- and CO monitors with zero and span gas to ensure proper
calibration of the monitors. The contractor shall also
check recorders for CO, Og, and temperature.
4.	The contractor shall determine if Rollins ceased burning
PCBs upon an issuance of an Air Stagnation Advisory by
NOAA for the surrounding area since the last monitoring
activity.
5.	Upon completion of the inspection, the contractor will
call the Project Officer (Region 6 Office) as soon as
possible to give a verbal report on the monitoring.
6.	The contractor shall monitor the incinerator 5 times
per month at variable intervals-and at variable times.
During each month the monitoring visits will take place
at least one time during each shift.
7.	A full written report shall be prepared every 90 days,
and submitted to the Region 6 Office within 30 days of
the expiration of each 90 day reporting period. The
reporting format shall follow the format as shown In
Attachment 1.
Qualifications of the Contractor
The contractor shall be a Registered Professional Engineer, know-
ledgable of combustion theory and Incineration. The contractor
shall also be certified 1n evaluating plume opacity 1n accordance
with EPA method 9 (40 CFR Part 60).
-------
6 flo
ATTACHMENT 1
Rollins Environmental Services Reporting Format
I.	Statement of Date and Time of monitoring and person conducting
monitoring.
II.	General summary of observations during each monitoring date high-
lighting compliance and/or non-coinpliance to the approval conditions.
III.	Detailed summary of comparison of the operating conditions and
limitations specified in the approval conditions. This summary
shall include appropriate copies of records and strip chart data
from CO, O2 and temperature instruments. *¦
IV.	Summary of observations of areas where PCBs or PCB items were
transported, handled, processed, or decontaminated for evidence
of PCB contamination. The contractor will record facility pro-
cedures that assure incineration of all PCBs. This section shall
include observations of the disposition of used PCB containers,
articles, and equipment.
V.	Summary of incinerator records and combustion products monitoring
records. Record whether CL, CO, C0-» and temperature were monitored
at all times during the PCB incineration period. This section shall
Include what PCB cutoff levels were in use for CO during the
previous 48 hours of PCB Incineration, and a summary of whether
PCB cutoff mechanisms were operable and 1f CO and 02 continuous
monitors were calibrated according to requirements. Obtain feed
rate data and determine if PCB feed rate during previous 48 hours
of PCB incineration from each monitoring date was within the
required limit.
VI.	A summary of observation of the water scrubber, scrubber lagoon,
and stack plume opacity. State the condition of the scrubber
lagoon's freeboard, and the operability of pumps and compressors
during the monitoring.
-------
(o o I
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION VI
1201 ELM STREET
DALLAS, TEXAS 7SZ7Q
Mr. Melvyn Bell, President
Energy Systems Company
P. 0. Box 1975
El Dorado, Arkansas 71730
Dear Mr. Bell:
Based on the results of test burns conducted at your facility 1n
October 1979 and August 1980, I hereby grant approval to you to
incinerate PC8s at the ENSCO facility at El Dorado, Arkansas, under
the authority of Section 6 of the Toxic Substances Control Act.
The approval is subject to the conditions specified in the enclosure
to this letter.
Violation of any condition Included as part of the approval will subject
ENSCO to enforcement action under the appropriate statute and/or
termination of the approval. Furthermore, this approval may be with-
drawn or further conditions nay be added at any time I feel that the
operation of the incinerator presents an unreasonable risk of Injury
to health or the environment from PCB Incineration. This approval
shall be effective on March 12, 1981.
If you have any questions, please contact me.
Sincerely,
/b/ Adlene Harriaoa
Adlene Harrison
Regional Administrator
Enclosure
cc: Governor Frank D. White
State of Arkansas
Jarrell E. Southall, Director
Arkansas Department of Pollution Control
and Ecology
-------
Condition for Incineration of Liquid and Solid PCBs
at Energy Systems Company (ENSCO)
El Dorado, Arkansas
At all times during PCB incineration, the incinerator shall meet
the requirements specified in 40 CFR 761.40(a).
The PC8 feed rate shall not exceed an average of 289 kg/hr
calculated over the actual period of incineration of the PCB
material.
The CO2 level shall be measured in the "hot duct" every 24 hours
during PCB incineration.
The flow of PCBs to the incinerator shall stop automatically
under any of the following conditions:
a.	the temperature drops below 1150°C at the exit end
of the combustion zone,
b.	the CO level shall not exceed 10 times the CO2 level, and
in any event shall not exceed 100 ppm (calibrate monitors
every 3 hrs. during PCB incineration by zero and span
gas), or
c.	the Q« drops below 4% (calibrate monitor every 24 hrs.
with zero and span gas).
The water scrubber shall at least achieve 99% HC1 removal efficiency,
and shall not emit greater that 37 lbs/hr HC1. Chlorine Input will
be determined by the appropriate ASTM method. Stack gases shall be
sampled using impinger containing NaOH, and analyzed for chloride
using the Argentimetric method as specified in Standard Methods for
Examination of Water and Waste Water, current edition.
Total particulate emissions shall not exceed State emission limit
for particulate emissions.
All records and data shall be maintained in accordance with 40 CFR
761.45(b), (c), and (f).
All PCB stor&ge facilities shall comply with 40 CFR 761.42. These
requirements include implementation of an SPCC Plan for PCB storage
and feed tanks as described under 40 CFR 761.42(7)(ii).
Any PCB container used for PCB transport, storage, or disposal shall
not be used for any other purpose unless decontamination of the
container complies with 40 CFR 761.43. All decontamination wash
fluids shall be Incinerated.
All PCB articles, equipment, and containers shall be properly
marked according to 40 CFR 761.44.
-------
(o 0 3
11.	ENSCO shall comply with NPDES requirements for PCB discharges.
12.	ENSCO shall comply with all State and local permits for the
incineration of wastes during PCB incineration.
13.	ENSCO personnel safety requirements and procedures for PCB
handling, storage, transport, and disposal shall comply with
OSHA requirements.
14.	Any violation or noncompliance with the conditions referenced
herein may result in suspension or revocation of the final approval
for incineration of PCBs.
15.	ENSCO must comply with 40 CFR Part 265 and Part 122 of the
May 19, 1980, Hazardous Waste Regulations ("Interim Status
Standards" and "RCRA Hazardous Waste Permit Regulations") and
any subsequent amendments.
16.	EPA may require annual testing or monitoring of the facility for
PCBs, HC1, particulates and up to 3 organics identified by EPA.
Written reports discussing the results of the testing or monitoring
shall be submitted to EPA within 180 days. Any modification to the
approved facility which may result in increased or changes in types
of emissions may require additional testing or monitoring.
17.	ENSCO must maintain a negative draft sufficient to prevent
fugitive emissions from the kiln or after burners.
18.	Kiln ash containing less than 50 ppm PCBs may be placed in a
permitted landfill; ash containing PCB concentrations greater than
50 ppm will require the ash to be recycled through the kiln, or
disposed in an approved PCB landfill.
19.	Should an Air Stagnation Advisory be issued which includes Union
County, Arkansas, the facility shall upon notification by EPA or
the State cease incineration of PCBs until an "all clear" is
issued in accordance with the Arkansas SIP.
20.	ENSCO will through contractual agreement, hire a private engineering
company to conduct regular monitoring of the ENSCO facility. The
selection of the private engineering company must be approved by
EPA, Region 6. The contract scope of work will be prepared by EPA,
Region 6. The private engineering company will monitor the ENSCO
facility and will report the monitoring results to EPA. The
contract scope of work and reporting will be in accordance with the
attached Scope of Work for Monitoring . This condition must be
satisfied prior to incineration of PCBs under this approval.
21.	The conditions of this authorization are severable, and 1f any
provision of this authorization, or any applications of any pro-
vision, Is held invalid, the remainder of this .authorization shall
not be affected thereby.
22.	Conditions 6 and 20 must be satisfied prior to incineration of
PCBs under this approval.
-------
t>"i
Scope of Work for Monitoring the ENSCO Facility
El Dorado, Arkansas
Background
The purpose of this scope of work is to ensure that the incineration
of PCBs is continuously accomplished consistent with the conditions
set forth in the letter of approval dated January 26, 1981, from
EPA, Region 6 regarding the incineration of PCBs at Energy System
Company (ENSCO) in El Dorado, Arkansas. ENSCO will contract with
and compensate a registered professional engineer who is mutually
acceptable to both ENSCO and EPA. The contractor will monitor
ENSCO's operations with regard to the conditions specified in the
above referenced letter. The monitoring will be conducted by the
ccntractor at random intervals, on an unannounced basis. ENSCO will
provide the contractor immediate access at any time to its entire
facility. The monitoring under the proposed contract will be
consistent with the principles set forth in EPA's Toxic Substance
Control Act (TSCA) PCB Monitoring and Disposal Regulation Interim
Inspection Guidance regarding TSCA inspections for the incineration
of PCBs. This contractual arrangement is not and should not be
construed as a substitute for periodic TSCA inspections conducted
by EPA personnel or inspections conducted by EPA's authorized
representatives.
Prior to awarding the contract, ENSCO has agreed to provide EPA
an opportunity to review and comment on the terms to be included
therein and, if necessary, EPA may require ENSCO to modify the terms
of the contract to ensure consistency with the January 26, 1981,
letter of approval mentioned above.
Specific Requirements of Monitoring
The monitoring shall Include at a minimum:
1.	A general monitoring of the incinerator including the
PCB shredder area, rotary kiln, the condition and main-
tenance of the main combustion chamber area, scrubber and
stack (including plume opacity readings). The contractor
will report on the overall operating condition of the
incinerator (eg., instrument conditions, scrubber performance,
stack opacity, evidence of PCB spillage, negative draft
maintained to prevent fugitive emissions, disposition of
kiln ash).
2.	The contractor shall examine and record the feed rate
records for the past 48 hours of PCB Incineration and
determine the actual feed rate of PCBs fed to the incin-
erator. This item will Include copies (if any) of
PCB analytical results retained by ENSCO of all PCB
materials fed to the incinerator during the previous
48 hours.
-------
6 or
3.	The contractor shall examine the record and CO- level in
the combustion gases upstream of the scrubber, the PCB
cutoff level set for the CO, 0-, and temperature instru-
ments, and shall observe that the cutoff mechanisms are
operable. Also, the contractor shall check the continuous
0, and CO monitors with zero and span gas to ensure proper
calibration of the monitors. The contractor shall also
check recorders for CO, 0g» and temperature.
4.	The contractor shall determine if ENSCO ceased burning
PCBs upon an issuance of an Air Stagnation Advisory by
NOAA for the surrounding area since the last monitoring
activity.
5.	Upon completion of the inspection, the contractor will
call the Project Officer (Region 6 Office) as soon as
possible to give a verbal report on the monitoring.
6.	The contractor shall monitor the incinerator 5 times
per month at variable intervals and at variable times.
During each month the monitoring visits will take place
at least one time during each shift (the operation of
the incinerator entails 3 shifts per day).
7.	A full written report shall be prepared every 90 days,
and submitted to the Region 6 Office within 30 days of
the expiration of each 90 day reporting period. The
reporting format shall follow the format as shown in
Attachment 1.
Qualifications of the Contractor
The contractor shall be a Registered Professional Engineer, know-
ledgeable of combustion theory and Incineration. The contractor
shall also be certified in evaluating plume opacity in accordance
with EPA method 9 (40 CFR Part 60).
-------
(o o (o
ATTACHMENT 1
ENSCO Reporting Format
I.	Statement of Date and Time of monitoring and person conducting
monitoring.
II.	General summary of observations during each monitoring date high-
lighting compliance and/or non-compliance to the approval conditions.
III.	Detailed summary of comparison of the operating conditions and
limitations specified in the approval conditions. This summary
shall include appropriate copies of records and strip chart data
from CO, Og and temperature instruments.
IV.	Summary of observations of areas where PCBs or PCB items were
transported, handled, processed, or decontaminated for evidence
of PCB contamination. The contractor will record facility pro-
cedures that assure incineration of all PCBs. This section shall
include observations of the disposition of used PCB containers,
articles, and equipment.
V.	Summary of incinerator records and combustion products monitoring
records. Record whether 0,, CO, CO-, and temperature were monitored
at all times during the PCB incineration period. This section shall
include what PCB cutoff levels were in use for CO during the
previous 48 hours of PCB incineration, and a summary of whether
PCB cutoff mechanisms were operable and if CO and O2 continuous
monitors were calibrated according to requirements. Obtain feed
rate data and determine if PCB feed rate during previous 48 hours
of PCB Incineration from each monitoring date was within the
required limit.
VI.	A summary of observation of the water scrubber, scrubber lagoon,
and stack plume opacity. State the condition of the scrubber
lagoon's freeboard, and the operability of pumps and compressors
during the monitoring.
-------