United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA - 454/R-99-026A
August 1999
Air
&EPA
Volume I of II
Asphalt Roofing Industry
Manual Emission Testing
Modified Bitumen
U.S. Intec
Port Arthur, Texas
-------
Asphalt Roofing Industry
Final Report
Contract No. 68-D7-0001
Work Assignment 2-002
U. S. Intec Modified Bitumen Facility
Port Arthur. Texas
Prepared for:
Michael L. Toney
Emission Measurement Center
Emission, Monitoring, and Analysis Division
Office of Air Quality Planning and Standards
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
September 1999
-------
Table of Contents
Page
1.0 INTRODUCTION 1-1
1.1 Objective '. 1-1
1.2 Brief Site Description 1-2
1.3 Emissions Measurement Program 1-3
1.3.1 Test Matrix 1-3
1.3.2 Test Schedule 1-5
1.3.3 Sampling Locations 1-5
1.3.4 Sampling and Analytical Methods 1-6
1.4 Quality Assurance/Quality Control (QA/QC) 1-6
1.5 Test Report 1-7
2.0 SUMMARY OF RESULTS 2-1
2.1 Emissions Test Log 2-1
2.2 PM Emission Results 2-1
2.3 D/F Emission Results 2-6
2.3.1 Overview 2-6
2.3.2 D/F Emission Results 2-6
3.0 PROCESS DESCRIPTION AND PROCESS DATA 3-1
4.0 SAMPLING LOCATIONS 4-1
5.0 SAMPLING AND ANALYTICAL PROCEDURES BY ANALYTE 5-1
5.1 Particulate Matter (PM) Emissions Testing Using EPA Method 5A 5-1
5.1.1 Method 5A Sampling Equipment 5-1
5.1.2 Method 5A Sampling Equipment Preparation 5-3
5.1.2.1 Glassware Preparation 5-3
5.1.2.2 Reagent Preparation 5-4
5.1.2.3 Equipment Preparation 5-4
5.1.3 Method 5A Sampling Operations 5-6
5.1.3.1 Preliminary Measurements 5-6
5.1.3.2 Assembling the Train 5-6
5.1.3.3 Sampling Procedures 5-7
5.1.4 Method 5A Sample Recovery 5-11
5.1.5 Quality Control for Gravimetric Procedures 5-13
5.2 Chlorinated Dibenzo-/?-Dioxin and Chlorinated Dibenzofuran Emissions
Testing Using EPA Method 23 5-14
5.2.1 Method 23 Sampling Equipment 5-14
in
-------
Table of Contents (Continued)
Page
5.2.2 Method 23 Equipment Preparation 5-14
5.2.2.1 Glassware Preparation 5-16
5.2.2.2 XAD-2® Sorbent Resin and Filter Preparation 5-16
5.2.2.3 Method 23 Sampling Train Preparation 5-18
5.2.3 Method 23 Sampling Operations 5-18
5.2.3.1 Preliminary Measurements 5-18
5.2.3.2 Assembling the EPA Method 23 Sampling Train 5-19
5.2.3.3 Sampling Procedures 5-20
5.2.4 PCDD/PCDF Sample Recovery 5-24
5.2.5 PCDD/PCDF Analytical Procedures 5-27
5.2.5.1 Preparation of Samples for Extraction 5-30
5.2.5.2 Calibration of the GC/MS System 5-30
5.2.6 PCDD/PCDF Analytical Quality Control 5-31
5.2.6.1 PCDD/PCDF Quality Control Blanks 5-31
5.2.6.2 Quality Control Standards and Duplicates 5-32
5.2.7 Analytes and Detection Limits for Method 23 5-33
5.3 EPA Methods 1-4 5-35
5.3.1 Traverse Point Location by EPA Method 1 5-35
5.3.2 Volumetric Flow Rate Determination by EPA Method 2 . 5-35
5.3.2.1 Sampling Equipment Preparation 5-35
5.3.2.2 Sampling Operations 5-35
5.3.3 O2 and CO2 Concentrations by EPA Method 3 5-36
5.3.4 Average Moisture Determination by EPA Method 4 5-36
6.0 QUALITY ASSURANCE/QUALITY CONTROL 6-1
6.1 Sampling QC Results ... • 6-1
6.2 D/F Sampling QC 6-1
6.2.1 PM Sampling QC 6-5
6.3 Analytical QC Results 6-7
6.3.1 D/F Analytical Quality Control 6-7
6.3.2 PM Quality Control 6-8
APPENDICES
A Field Data Sheets
B Paniculate Matter Raw Data
C Dioxin/Furan Raw Data
D Summary of Process Data
IV
-------
List of Figures
Page
4-1 Thermal Oxidizer Inlet Sampling Location 4-2
4-2 Thermal Oxidizer Outlet Sampling Location 4-3
4-3 APP and SBS Outlet Sampling Locations 4-4
4-4 Thermal Oxidizer Inlet Traverse Point Layout 4-5
4-5 Thermal Oxidizer Outlet Traverse Point Layout 4-6
4-6 APP and SBS Outlet Traverse Point Layout 4-7
5-1 Schematic of Method 26A Sampling Train 5-2
5-2 Method 26A Sample Recovery Scheme 5-12
5-3 • Method 23 Sampling Train Configuration 5-15
5-4 Method 23 Field Recovery Scheme 5-25
5-5 Extraction and Analysis Schematic for Method 23 Samples 5-29
-------
List of Tables
Page
1-1 V. S. Intec Facility (Port Arthur, Texas) Sampling and Analytical Matrix 1-4
1-2 U. S. Intec Facility (Port Arthur, Texas) Test Schedule 1-5
2-1 Emissions Log, U. S. Intec 2-2
2-2 Sample Volume Collected, dscm 2-3
2-3 Flue Gas Volumetric Flow Rates, dscmm 2-3
2-4 Paniculate Matter Concentration, EPA Method 5A 2-4
2-5 Paniculate Matter Emission Rate, Ib/hr, and Thermal Oxidizer Removal
Efficiency, % 2-5
2-6 Paniculate Matter Emission Rate, Ibs/Ton of Product 2-5
2-7 Thermal Oxidizer Dioxin/Furan Stack Gas Concentrations 2-7
2-8 Dioxin/Furan Stack Emission Rate, ug/hr, Thermal Oxidizer Inlet and Outlet 2-9
2-9 Dioxin/Furan Stack Emission Rate, Thermal Oxidizer Outlet, ug/Mg Product 2-11
2-10 Dioxin/Furan 2,3,7,8-TCDD Toxicity Equivalent Stack Gas Concentrations,
Thermal Oxidizer Inlet and Outlet 2-12
5-1 Glassware Cleaning Procedure for Method 5A Sampling Train Components 5-3
5-2 Sampling Checklist, EPA Method 5A 5-9
5-3 EPA Method 23 Glassware Cleaning Procedure for Train Components, Sample
Containers, and Laboratory Glassware 5-17
5-4 EPA Method 23 Sampling Checklist 5-21
5-5 Method 23 Sample Fractions Shipped to the Analytical Laboratory 5-27
5-6 PCDD/PCDF Congeners To Be Analyzed 5-28
5-7 Method 23 Blanks Collected in the Field 5-31
vi
-------
List of Tables (Continued)
Page
5-8 Analytical Detection Limits (HRGC/HRMS) for Dioxins/Furans 5-34
5-9 PCDD/PCDF Method Detection Limits HRGC/HRMS 5-34
6-1 Summary of Leak Checks Performed, Per Port, Method 23 Testing, Thermal Oxidizer,
Inlet and Outlet 6-2
6-2 Summary of Isokinetic Percentages 6-3
6-3 Dry Gas Meter Post-Test Calibration Results 6-4
6-4 Dioxin/Furan Field Blank Analysis Results 6-5
6-5 Summary of Leak Checks Performed, Per Port, Method 5 A 6-6
VII
-------
1.0 INTRODUCTION
The asphalt roofing industry is among the categories of major sources for which national
emission standards for hazardous air pollutants (NESHAPS) are to be issued by November 2000
pursuant to Section 112 of the Clean Air Act. The asphalt roofing industry category includes
facilities that have the following specific processes: asphalt blowing stills; modified bitumen
production; a saturator/coater process; and fiberglass mat production.
Source tests are required to quantify and characterize the particulate matter (PM) and
hazardous air pollutant (HAP) emissions, and the performance of a thermal oxidizer used to
control emissions associated with a modified bitumen facility.
1.1 Objective
The objective of the testing at the U. S. Intec modified bitumen facility in Port Arthur,
Texas, was to perform all activities necessary to characterize the thermal oxidizer for the
following emission components:
• Determination of particulate matter (PM) using EPA Method 5A; and
• Determination of dioxins and furans (D/F) using EPA Method 23.
The roof top vents for the atactic polypropylene (APP) coater and the -styrene butadiene
styrene (SBS) coater were tested for the following component:
• Determination of particulate matter (PM) using EPA Method 5 A.
In addition, the determination of total hydrocarbons using Method 25A and preliminary
screenings for organic HAPs using a Fourier Transform Infrared (FTIR) monitoring instrument
were conducted by Midwest Research Institute (MRI) under a separate work assignment. Testing
l-l
-------
by ERG and MRI occurred simultaneously. The FTIR element is not included within this final
report.
Testing was performed at the inlet and outlet simultaneously at the thermal oxidizer and
at the outlet only of the two roof top vents (APP and SBS). ERG coordinated all field test
activities with MRI personnel.
1.2 Brief Site Description
The asphalt roofing industry produces roofing and siding materials for weather-proofing
various structures. The process involves the saturation of a substrate with modified bitumen
(asphalt) which is then coated with materials such as sand before being cut to size and packaged.
Typical emission controls include a thermal oxidizer.
The U. S. Intec plant, located in Port Arthur, Texas, produces rolled roofing products by
saturating a polyester substrate or a fiberglass substrate with bitumen that has been modified with
either APP or SBS. The entire manufacturing process is housed under a single roof with the
production beginning on each Monday morning, operating 24 hours per day and ending
production on each Friday night, leaving Saturday and Sunday for maintenance. Production is
greatest during the spring and summer months'to accommodate increases in construction and
decreases in the fall and winter months. A thermal oxidizer is used to control emissions from
various plant processes, such as the holding, mixing, and storage tanks.
The facility produces rolled roofing products by saturating a polyester substrate or a
fiberglass substrate with bitumen (modified with either APP or SBS) on two separate production
lines. Both substrates enter their respective production lines through a web unwind stand and
then go through a dry hooper. Asphalt is loaded from tanker trucks into two 100 ton, 400°F
mixing tanks for the polyester substrate and two 10.5 ton, 390°F mixing tanks for the fiberglass
substrate line. Tanker trucks also unload polymer liquid into two steam jacketed storage tanks.
1-2
-------
Modified bitumen is produced by combining the asphalt with polypropylene and fillers in
the six mixing tanks for the polyester substrate line and the SBS in the first mixing tank and
fillers in the second mixing tank for the fiberglass substrate line. The modified bitumen from the
mixing tanks is transferred to two holding tanks for the polypropylene substrate line and one
holding tank for the fiberglass substrate line before going to the respective coaters. Once the
saturated polyester substrate leaves the vat, it is coated with granules and talc. The saturated
fiberglass substrate is coated with granules and sand. After both products are cooled, the
substrates go through separate finish loopers and roll winders.
Emissions from the mixing tanks and holding tanks for both production lines and the
blended polymer operation go to the thermal oxidizer. The thermal oxidizer has an operating
temperature of 1440°F and a residence time of 0.5 seconds. The emissions from the coaters are
uncontrolled. Emissions from the coater areas are collected by overhead hoods and are each
vented through the roof with separate vents. The flow rate at the outlet of the thermal oxidizer is
approximately 12,000 actual cubic feet per minute (acfm) at a temperature of 479 °F.
1.3 Emissions Measurement Program
This section provides an overview of the emissions measurement program conducted at
the U. S. Intec facility in Port Arthur, Texas. Included in this section are summaries of the test
matrix, sampling locations, sampling methods, and laboratory analysis. Additional detail on
these topics is provided in the sections that follow.
1.3.1 Test Matrix
The sampling and analytical matrix is presented in Table 1-1. Manual emissions tests
were employed; detailed descriptions of these sampling and analytical procedures are provided in
Section 5.0.
1-3
-------
Table 1-1. U. S. Intec Facility (Port Arthur, Texas) Sampling and Analytical Matrix
Sampling
Location
Thermal
Oxidizer
Inlet/Outlel
Thermal
Oxidizer
Inlet/Outlet
Thermal
Oxidizer
Inlet/Outlet
APP and SBS
Roof Vent
Outlets
APP and SBS
Roof Vent
Outlets
Number of
Runs
3
3
3
3
3
Sample Type
Gas Velocity
Volume
Moisture
Total PM
Dioxins/
Furans
Gas Velocity
Volume
Moisture
Total PM
EPA Reference
Method
I -4
5A
23
I -4
5A
Sampling
Period
(Duration)
4 hrs .
4 hrs
4 hrs
I hr
I In-
Analytical
Method
Volumetric/
Gravimetric
Gravimetric
HRGC/HRMS1
Volumetric/
Gravimetric
Gravimetric
Laboratory
Performing
Analysis
ERG
ERG
Triangle
Laboratories
ERG
ERG
'HRGC/HRMS = High Resolution Gas Chromatography/High Resolution Mass Spectrometry.
-------
1.3.2 Test Schedule
The daily test schedule is presented in Table 1-2. The test required two days of set-up,
four test days, and one tear-down day. Each test day was approximately 12-16 hours in length
with a typical working period between 6:00 a.m. and 8:00 p.m. A typical set-up, tear-down day
was 8-10 hours in length, with working hours between 7:00 a.m. and 7:00 p.m. The test schedule
was based on the test duration shown in Table 1-1.
Table 1-2. U. S. Intec Facility (Port Arthur, Texas) Test Schedule
Test Day
1
2
3
4
5
6
7
8
9
Date
9/18/97
9/19/97
9/20/97
9/21/97
9/22/97
9/23/97
9/24/97
9/25/97
9/26/97
Activity
Travel
Travel
Setup
Setup
Test Day #1: Thermal Oxidizer
Test Day #2: Thermal Oxidizer
Test Day #3: Coaler Vents
Test Day #4: Coater Vents
Travel
1.3.3 Sampling Locations
The sampling locations used during the emissions testing program at the U. S. Intec
facility are described below. Flue gas samples were collected at the inlet and outlet of the
thermal oxidizer using two ports at each location and at the outlet of the APP coater vent and the
SBS coater vent.
1-5
-------
The test ports and their locations met the requirements of EPA Method 1. The thermal
oxidizer inlet location is a circular duct with an inside diameter (i.d.) of 36". At the time of the
pre-test survey there were no ports in place at the inlet location, but two 4" i.d. ports had been
installed at 90 degrees to each other, one horizontally and one vertically on top of the duct, by the
time of the test. The thermal oxidizer outlet location is a vertical circular stack with an i.d. of
36". Two new 4" i.d. ports 90 degrees to each other were also installed at this location. The two
roof top vents were similar in size and shape and had an inside diameter of 24". Two 4" i.d.
sampling points, 90 degrees to each other were installed on each vent approximately two feet
above the roof line. A new sampling port for FTIR sampling was installed at the thermal
oxidizer inlet and outlet as well as both roof top vents.
1.3.4 Sampling and Analytical Methods
Total particulate matter emissions were determined using EPA Method 5A, with
paniculate mass emissions collected on a glass fiber filter together with any material that
condensed at or above the filter temperature of 42°C. Particulate concentrations are based on the
weight gain of the filter and any condensible PM recovered from the 1,1,1-trichloroethane (TCE)
rinses of the probe, nozzle, and front half of the glass filter holder.
Flue gas samples for D/F were collected using EPA Method 23. Flue gas was extracted
isokinetieally and any D/F present was collected on the filter, the XAD-2® resin trap, and in the
probe and condenser. The analysis was performed using HRGC coupled with HRMS.
1.4 Quality Assurance/Quality Control (QA/QC)
All flue gas testing procedures followed comprehensive QA/QC procedures as outlined in
the Site Specific Test Plan (SSTP) and the Quality Assurance Project Plan (QAPP). A full
description of the resulting QA parameters is presented in Section 6.
1-6
-------
All post-test and port change leak checks met the criteria prescribed in the manual
methods procedure. The allowable isokinetic QC range of ±10% was met for all PM and D/F
sampling runs. All post-test dry gas meter calibration checks were within 5% of the full
calibration factor. Field blanks (FB) for the D/F and PM tests showed virtually no
contamination.
All analyses were completed under a strict QA/QC regimen. For the D/F results, percent
recoveries of all isotopically-labeled compounds were within the lower and upper limits of
recovery as specified in the method.
The manual flue gas flow rates test data reflected very little variation over the test runs.
The percent relative standard deviation (% RSD) observed during each set of tests runs ranged
from 1 to 6. These values indicate that the process was very stable during the test period.
1.5 Test Report
This final report, presenting all data collected and the results of the analyses, has been
\
prepared in six sections, as described below:
Section 1 provides an introduction to the testing effort and includes a brief
description of the test site, an overview of the emissions measurement program.
and a brief overview of the QC results;
Section 2 gives a summary of the test results for the PM and D/F tests;
Section 3 provides a description of the process and plant operation during the field
test. These data are to be supplied by EPA and are not included in this report;
Section 4 gives a discussion of the sampling locations;
Section 5 presents detailed descriptions of the sampling and analysis procedures;
and
Section 6 provides details of the quality assurance/quality control procedures used
on this program and the QC results.
1-7
-------
The appendices containing copies of the actual field data sheets and the results of the
laboratory analyses are also contained as part of this report.
1-8
-------
2.0 SUMMARY OF RESULTS
This section provides the results of the emissions test program conducted at the U. S.
Intec asphalt roofing facility (Port Arthur, Texas) from September 22 to September 25, 1997.
Included in this section are results of manual tests conducted for PM and D/F.
2.1 Emissions Test Log
Eighteen tests were conducted over a four-day period (6 D/F and 12 PM). Table 2-1
presents the emissions test log which shows the test date, location, run number, test type, run
times and port change times for each test method.
Table 2-2 shows the volume of stack gas sampled for each run in dry standard cubic
meters (dscm) and Table 2-3 shows the stack gas volumetric flow rate during each run in dry
standard cubic meters per minute (dscmm). The percent relative standard deviation (%RSD)
calculated for the three runs for each test method (shown in Table 2-3) was less than 6%,
indicating that the process flow was relatively constant over the four test days. All related field
sheets are given in Appendix A.
2.2 PM Emission Results
Particulate matter emissions were determined from EPA Method 5A sampling trains,
with testing performed at the inlet and outlet of the thermal oxidizer, and at the roof vent outlets
of the APP and SBS production lines. PM collected on the train filter and in the 1,1,1-
trichloroethane probe rinse of the Method 5A sampling train was analyzed gravimetrically. PM
stack concentrations, in grams per dry standard cubic meter (g/dscm), the average and %RSD for
the three test runs at the inlet and outlet of the thermal oxidizer and for the three test runs at each
of the roof top vents are shown in Table 2-4. The % RSD for both the inlet and outlet of the
thermal oxidizer was fess than 32, showing good reproducibility for the sampling and analysis
method as well as relatively constant process conditions over the four-day test period. The
2-1
-------
Table 2-1. Emissions Log, U. S. Intec
Date
9/22/97
9/23/97
9/24/97
9/25/97
Location
Thermal Oxidizer Inlet, Port B
Thermal Oxidizer Inlet, Port A
Thermal Oxidizer Inlet, Port A
Thermal Oxidizer Inlet, Port B
Thermal Oxidizer Outlet, Port B
Thermal Oxidizer Outlet, Port A
Thermal Oxidizer Outlet, Port A
Thermal Oxidizer Outlet, Port B
Thermal Oxidizer Inlet, Port B
Thermal Oxidizer Inlet, Port A
Thermal Oxidizer Inlet, Port B
Thermal Oxidizer Inlet, Port A
Thermal Oxidizer Inlet, Port A
Thermal Oxidizer Inlet, Port B
Thermal Oxidizer Inlet, Port A
Thermal Oxidizer Inlet, Port B
Thermal Oxidizer Outlet, Port B
Thermal Oxidizer Outlet, Port A
Thermal Oxidizer Outlet, Port B
Thermal Oxidizer Outlet, Port A
Thermal Oxidizer Outlet, Port A
Thermal Oxidizer Outlet, Port B
Thermal Oxidizer Outlet, Port A
Thermal Oxidizer Outlet, Port B'
APP Stack, Port Y
APP Stack, Port X
APP Stack, Port X
APP Stack, Port Y
APP Stack, Port Y
APP Stack, Port X
SBS Stack, Port X
SBS Stack, Port Y
SBS Stack, Port Y
SBS Stack, Port X
SBS Stack, Port X
SBS Stack, Port Y
Run
Number
1
1
1
1
1
1
1
1
2
2
3
3
2
2
3
3
2
2
3
3
2
2
3
3
1
1
2
2
3
3
1
1
2
2
3
3
Test Type
PM
PM
D/F
D/F
PM
PM
D/F
D/F
PM
PM
PM
PM
D/F
D/F
D/F
D/F
PM
PM
PM
PM
D/F
D/F
D/F
D/F
PM
PM
PM
PM
PM
PM
PM
PM
PM
PM
PM
. PM
Run Time
1450-1650
1725-1925
1450-1650
1725-1925
1450-1650
1725-1925
1450-1650
1725-1925
1015-1240
1245-1440
1710-1910
1940-2140
1015-1215
1240-1440
1710-1910
1940-2140
1015-1215
1240-1440
1710-1910
1940-2140
1015-1215
1240-1440
1710-1910
1940-2140
1325-1401
1406-1431
1510-1545
1550-1615
1632-1703
1705-1735
1435-1505
1513-1538
0922-0952
0956-1026
1040-1110
1114-1144
2-2
-------
Table 2-2. Sample Volume Collected, dscm1
Location
Thermal Oxidizer Inlet
Thermal Oxidizer Inlet
Thermal Oxidizer Outlet
Thermal Oxidizer Outlet
APP Stack, Outlet
SBS Stack, Outlet
Parameter
PM
D/F
PM
D/F
PM
PM
Runl
4.34
4.45
4.18
3.90
1.24
1.26
Run 2
4.06
4.21
4.00
3.96
1.26
1.41
Run 3
4.04
4.31
3.93
3.87
1.27
1.33
Average
4.15
4.32
4.04
3.91
1.26
1.33
%RSD
4.04
2.79
3.19
1.17
1.21
5.64
'Standard conditions are defined as 1 atm and 68 °F.
Table 2-3. Flue Gas Volumetric Flow Rates, dscmm1
Location
Thermal Oxidizer Inlet
Thermal Oxidizer Inlet
Thermal Oxidizer Outlet
Thermal Oxidizer Outlet
APP Stack, Outlet
SBS Stack, Outlet
Parameter
PM
D/F
PM
D/F
PM
PM
Run 1
227
218
215
212
308
306
Run 2
209
212
207
209
314
329
Run 3
206
213
206
210
317
323
Average
214
214
209
210
313
319
%RSD
5.36
1.72
2.32
0.91
1.45
3.67
'Standard conditions are defined as 1 atm and 68°F.
2-3
-------
Table 2-4. Paniculate Matter Concentration, EPA Method 5A
Location
Thermal Oxidizer, Inlet
Thermal Oxidizer, Outlet
APP Stack, Outlet
SBS Stack, Outlet
gr/dscm
Runl
0.23
0.0093
0.13
0.13
Run 2
0.15
0.0077
0.14
0.056
Run 3
0.12
0.011
0.099
0.068
Average
0.17
0.0093
0.12
0.085
%RSD
31.3
16.7
17.3
46.7
%RSD for the three runs at SBS vent was 46.7. The variability in these results is most likely due
to gravimetric measurements made on very low amounts of PM. Increased variability would not
be unusual in this case. Since no measurable PM was detected in the blanks, the blank correction
was not necessary.
Table 2-5 shows the average PM emission rate in pounds per hour (Ib/hr), for each of the
three outlet test locations. These values were calculated from the average outlet concentration
from Table 2-4 and the average stack flow rate from Table 2-3. Using this value in conjunction
with the equivalent value for the inlet (see Table 2-5), a PM removal efficiency for the thermal
oxidizer was calculated to be 94.6%. Table 2-6 shows the average PM emissions, in pounds per
ton of product, for each of the three test location outlets. These values were calculated from the
results presented in Table 2-5 and the summary of the process data presented in Appendix D,
Tables D-2 through D-4. The average total tons produced during the test period were divided by
the length of the test run. This value was divided into the emission rate from Table 2-5. The
data presented in Appendix D are a summary of the process information extracted from the
information supplied by the facility. Although much of the process data was present, several
clarifications and supplemental data were requested from and supplied by the facility after
several iterations. The PM analytical raw data are given in Appendix B.
2-4
-------
Table 2-5. Participate Matter Emission Rate, Ib/hr, and Thermal Oxidizer Removal
Efficiency, %
Location
Thermal Oxidizer
APP Vent
SBS Vent
Ib/hr
Average Inlet Rate1
0.308
NT3
NT
Average Outlet Rate1
0.0166
0.331
0.228
Removal Efficiency2
%
94.6
NA4
NA
'Average of three test runs.
2Percent Removal Efficiencey = (Inlet - Outlet) / Inlet x 100
3NT = Not Tested
4NA = Not Applicable
Table 2-6. Participate Matter Emission Rate, Ibs/Ton of Product
Location
Thermal Oxidizer Outlet
APP Vent Outlet
SBS Vent Outlet
Ibs/ton of Product
1.9x 10'3
4.5 x 10'2
2.9 x 10'1
2-5
-------
2.3 D/F Emission Results
2.3.1 Overview
Six 4-hour D/F emission test runs were completed at U.S. Intec during the week of
September 21, 1997. Three test runs were completed at the inlet and at the outlet of the thermal
oxidizer which is associated with various plant processes, such as holding, mixing, and storage
tanks. The sample collection protocol followed EPA Method 23, while the analysis protocol
followed EPA Method 23/8290. The sample preparation and analytical procedure is discussed in
Section 5 of this report.
2.3.2 D/F Emission Results
Table 2-7 presents the concentration, in nanograms per dry standard cubic meter
(ng/dscm) for the selected D/F congeners by run number, the average concentration over the
three runs and the %RSD. All results were determined by HRGC/HRMS using a DB-5 capillary
gas chromatographic column.
Any compound mat was not detected is reported as a "less than" value with this value
being the instrumental detection limit. A "less" than" value rather than "0" is used in all
appropriate calculations. The % RSD values reported in Table 2-7 for each test run by
compound range widely, but are not indicative of poor sampling/analytical reproducibility
because almost all values are reported as "less than" values. Increased variability is not unusual
in these cases.
Table 2-8 shows the D/F stack emission rates, in ug/hr, from the thermal oxidizer inlet
and outlet. These values were calculated from the average concentrations from Table 2-7 and the
average stack flow rate from Table 2-3.
2-6
-------
Table 2-7. Thermal Oxidizer Dioxin/Furan Stack Gas Concentrations
Congener
ng/dscm
Runl
Run 2
Run 3
Average
1
% RSD 1
Inlet [
2,3,7,8-TCDD
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,6,7,8-HpCDD
OCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDF
<0.001
<0.002
<0.002
<0.002
<0.002
<0.002
0.005 '
<0.001
<0.002
<0.002
<0.002
<0.001
<0.002
<0.002
<0.002
<0.002
<0.004
<0.001
<0.002
<0.002
<0.002
<0.002
<0.002
<0.005
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.002
<0.002
<0.002
<0.001
<0.002
<0.002
<0.002
<0.002
<0.002
<0.005
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.002
<0.002
<0.002
<0.002
<0.001
<0.002
<0.002
<0.002
<0.002
<0.002
<0.005
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.002
<0.002
<0.002
<0.003
°'°
o.o 1
0.0
0.0
0.0
o.o 1
0.0 |
0.0
63.2
63.2
63.2
63.2
66.7
0.0
0.0
Outlet
2,3,7,8-TCDD
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,6,7,8-HpCDD
OCDD
0.002'
<0.002
<0.002
<0.002
<0.002
<0.002
<0.02
<0.002
<0.003
<0.003
<0.003
<0.003
0.0051
<0.005
<0.003
<0.005
<0.005
<0.005
<0.005
<0.005
<0.008
<0.002
<0.003
<0.003
<0.003
<0.003
<0.004
<0.011
28.9
50.9
50.9
50.9 I
50.9 1
37.6 |
72.1 j
2-7
-------
Table 2-7. Continued
Congener
2,3,7,8-TCDF
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDF
ng/dscm
Runl
<0.001
<0.001
<0.001
<0.001
<0.001
0.0021
<0.001
<0.002
<0.002
<0.002
Run 2
<0.002
<0.002
<0.002
<0.002
<0.002
<0.002
<0.003
<0.003
<0.005
<0.005
Run 3
<0.002
<0.002
<0.002
<0.002
<0.002
<0.003
<0.003
<0.005
<0.005
<0.005
Average
<0.002
<0.002
<0.002
<0.002
<0.002
<0.002
<0.002
<0.003
<0.003
<0.003
% RSD
83.3
50.0
50.0
50.0
50.0
37.6
57.7
50.9
96.2
96.2
'Amount detected is less than 5 times the detection limit and should be considered only an
estimate.
2-8
-------
Table 2-8. Dioxin/Furan Stack Emission Rate,
Outlet
, Thermal Oxidizer Inlet and
Congener
2,3,7,8-TCDD
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,6,7,8-HpCDD
OCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDF
Average
Concentration
ng/dscm
Average
Emission
Rate
jig/hr
Inlet
<0.001
<0.002
<0.002
<0.002
<0.002
<0.002
<0.005
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.002
<0.002
<0.002
<0.003
<0.013
<0.026
<0.026
<0.026
<0.026
<0.026
<0.064
<0.0i3
<0.013
<0.013
<0.013
<0.013
<0.013
<0.026
<0.026
<0.026
<0.039
Average
Concentration
ng/dscm
Average
Emission
Rate
Mg/hr 1
Outlet [
<0.002
<0.003
<0.003
<0.003
<0.003
<0.004
<0.011
<0.002
<0.002
<0.002
<0.002
<0.002
<0.002
<0.002
<0.003
<0.003
<0.003
<0.025 1
<0.038
<0.038
<0.038
<0.038 J
<0.050 j
<0.139 |
<0.025 I
<0.025 1
<0.025
<0.025
<0.025
<0.025
<0.025
<0.038
<0.038
<0.038
2-9
-------
Table 2-9 shows the average D/F emissions, in micrograms per megagram (ug/Mg) of
product, for the thermal oxidizer cutlet. These values were calculated from the results presented
in Table 2-8 and the summary of the process data presented in Appendix D, Tables D-land D-2.
The average total tons produced during the test period were divided by the length of the test run.
This value was divided into the emission rate for each congener from Table 2-8. The data
presented in Appendix D are a summary of the process information extracted from the
information supplied by the facility. Although much of the process data was present, several
clarifications and supplemental data were requested from and supplied by the facility after
several iterations.
Table 2-10 shows the congener concentrations in ng/dscm converted to 2,3,7,8-
tetrachlorodibenzo-p-dioxin toxicity equivalents as well as a summation of the values presented
as total chlorinated dioxins and total chlorinated furans. All D/F analytical raw data can be found
in Appendix C.
2-10
-------
Table 2-9. Dioxin/Furan Stack Emission Rate, Thermal Oxidizer Outlet,
ug/Mg Product
Congener
2,3,7,8-TCDD
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1, 2,3,4,6,7, 8-HpCDD
OCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDF
Average
Emission
Rate
Mg/Mg Product
<0.0032
<0.0048
<0.0048
<0.0048
<0.0048
<0.0064
<0.018
<0.0032
<0.0032
<0.0032
<0.0032
<0.0032
<0.0032
<0.0032
<0.0048
<0.0048
<0.0048
2-11
-------
Table 2-10. Dioxin/Furan 2,3,7,8-TCDD Toxicity Equivalent Stack Gas
Concentrations, Thermal Oxidizer Inlet and Outlet
Congener
2,3,7,8-
TCDD
TEF1
ng/dscm
Runl
Run 2
Run 3
Average
Inlet
2,3,7,8-TCDD
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,6,7,8-HpCDD
OCDD
1.0
0.5
0.1
0.1
0.1
0.01
0.001
<0.001
<0.001
<0.0002
<0.0002
<0.0002
<0.00002
0.0000052
<0.001
<0.001
<0.0002
<0.0002
<0.0002
<0.00002
<0.000005
<0.001
<0.001
<0.0002
<0.0002
<0.0002
<0.00002
<0.000005
Total PCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDF
0.1
0.05
0.5
0.1
0.1
0.1 •
0.1
0.01
0.01
0.001
<0.0001
<0.0001
<0.001
<0.0002
<0.0001
<0.0002
<0.0002
<0.00002
<0.00002
<0.000004
<0.0001
<0.00005
<0.0005
<0.0001
<0.0001
<0.0001
<0.0002
<0.00002
<0.00002
<0.000002
<0.0001
<0.00005
<0.0005
<0.0001
<0.0001
<0.0001
<0.0002
<0.00002
<0.00002
<0.000002
Total PCDF
<0.001
<0.001
<0.0002
<0.0002
<0.0002
<0.00002
<0.000005
<0.0026
<0.0001
<0.00005
<0.0007
<0.0001
<0.0001
<0.0001
<0.0002
<0.00002
<0.00002
<0.000003
<0.0014
Outlet
2,3,7,8-TCDD
1,2,3,7,8-PeCDD
1.0
0.5
0.0022
<0.001
<0.002
<0.0015
<0.003
<0.0025
<0.002
<0.0015
2-12
-------
Table 2-10. Continued
Congener
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,6,7,8-HpCDD
OCDD
2,3,7,8-
TCDD
TEF1
0.1
0.1
0.1
0.01
0.001
ng/dscm
Runl
<0.0002
<0.0002
<0.0002
<0.00002
<0.000002
Run 2
<0.0003
<0.0003
<0.0003
0.000052
<0.000005
Run 3
<0.0005
<0.0005
<0.0005
<0.00005
<0.000008
Total PCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1, 2,3,6,7, 8-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDF
0.1
0.05
0.5
0.1
0.1
0.1
0.1
0.01
0.01
0.001
<0.0001
<0.00005
<0.0005
<0.0001
<0.0001
0.00022
<0.0001
<0.00002
<0.00002
<0.000002
<0.0002
<0.0001
<0.001
<0.0002
<0.0002
<0.0002
<0.0003
<0.00003
<0.00005
<0.000005
<0.0002
<0.00015
<0.0005
<0.0003
<0.0003
<0.0003
<0.0003
<0.00005
<0.00005
<0.000005
Total PCDF
Average
<0.0003
<0.0003
<0.0003
<0.00004
<0.000005
<0.0044
<0.0002
<0.0001
<0.001
<0.0002
<0.0002
<0.0002
<0.0002
<0.00003
<0.00004
<0.000003
<0.0022
'TEF, Toxicity Equivalence Factor
2The amount detected is less than 5 times the detection limit and should be considered an
estimated value.
2-13
-------
3.0 PROCESS DESCRIPTION AND PROCESS DATA
The facility produces rolled roofing products by saturating a polyester substrate and a
fiberglass substrate with modified bitumen on two separate production lines. Both substrates
enter their respective production lines through a web unwind stand and then go through a dry
looper. Asphalt is loaded from tanker trucks into two 100-ton, 350°F asphalt storage tanks.
Asphalt from the storage tanks is distributed to six 10.5-ton, 400°F mixing tanks for the
polyester substrate line and two 10.5-ton, 390°F mixing tanks for the fiberglass substrate line.
Tanker trucks also load polymer liquid into two steam jacket storage tanks.
Modified bitumen is produced by combining the asphalt with polypropylene and fillers in
the six mixing tanks for the polyester substrate line and SBS in the first mixing tank and fillers in
the second mixing tank for the fiberglass substrate line. The modified bitumen from the mixing
tanks is transferred to two holding tanks for the polyester substrate line and one holding tank for
the fiberglass substrate line before going to their respective coaters. The coaters are
impregnation vats which saturate both substrates with the modified bitumen. Once the-saturated
polyester substrate leaves the vat it is coated with granules and talc. The saturated fiberglass
substrate is coated with granules and sand. After both products are cooled, the substrates go
through separate finish loopers and roll winders.
Emissions from the mixing tanks and holding tanks for both production lines and the
blended polymer operation go to the thermal oxidizer. The thermal oxidizer has an operating
temperature of 1400°F and a residence time of 0.5 seconds. The talc applicator for the polyester
substrate production line is controlled by a baghouse. The baghouse (fabric filter) has an air-to-
cloth ratio of 5:1 actual cubic feet per minute (acfm), an inlet volumetric flow rate of 5000 dry
standard cubic feet per minute (dscfm), and an inlet gas temperature of 100°F. Emissions from
the coaters are uncontrolled and vent to the atmosphere through roof top vents.
3-1
-------
4.0 SAMPLING LOCATIONS
The sampling locations used during the emission testing program at the U. S. Intec, Port
Arthur. Texas, plant are described in this section. Flue gas samples were collected at the inlet
and outlet of the thermal oxidizer scrubber using two ports at each location and at the outlet of
the APP coater vent and the SBS coater vent. The configurations of the sampling locations are
shown in Figures 4-1,4-2, and 4-3. The configurations for the two roof top vents are similar.
The test ports and their locations met the requirements of EPA Method 1. The thermal
oxidizer inlet is a circular duct with an i.d. of 36". At the time of the pre-test survey there were
no ports in place at the inlet location, but ERG requested that two 4" i.d. ports be installed at
;90 degrees to each other, one horizontally and one vertically on top of the duct; these ports were
in place for the field testing effort. The thermal oxidizer outlet is a vertical circular stack with an
i.d. of 36" with two 4" ports installed at 90 degrees to each other. The two roof top vents have an
inside diameter of 24" and each has two 3" outside diameter (o.d.) ports installed at 90 degrees
to each other. The 3" ports were enlarged to 4" prior to the testing effort in order to
I
Accommodate the sampling probes. The position and number of traverse points for the outlet and
inlet locations are shown in Figures 4-4,4-5, and 4-6, respectively.
4-1
-------
FTIR
v_
n
Flow
0
15'Above Grade
" I.D.
36" I.D,
T.O
-------
FTIR
4" I.D.
U>
Thermal Oxidizer
36" I.D.
30'
Figure 4-2. Thermal Oxidizer Outlet Sampling Location
-------
FTIR 4" I.D.
24" I.D,
4" I.D.
Roof Top
50' Above Grade
-------
36" I.D.
L/l
Figure 4-4. Thermal Oxidizer Inlet Traverse Point Layout
-------
36" I.D.
-------
24" I.D.
Figure 4-6. APR and SBS Outlet Traverse Point Layout
-------
5.0 SAMPLING AND ANALYTICAL PROCEDURES BY ANALYTE
The sampling and analytical procedures used for the asphalt roofing test program are
the most recent revisions of the published EPA methods. In this section, descriptions of each
sampling and analytical method by analyte are provided.
5.1 Particulate Matter (PM) Emissions Testing Using EPA Method 5A
Sampling for Paniculate Matter (PM) was performed according to the EPA Method 5A
protocol. This method is applicable to the determination of particulate mass emissions collected
on a glass fiber filter and any material that condenses at or above the filter temperature of 42 °C
from various types of process controls and combustion sources associated with the asphalt
industry.
Particulate concentrations are based on the weight gain of the filter and any condensible
PM recovered from the 1,1,1 -trichloroethane (TCE) rinses of the probe, nozzle, and front half of
the glass filter holder.
5.1.1 Method 5A Sampling Equipment
The Method 5A methodology uses the sampling train shown in Figure 5-1.
The 4-impinger train consists of a borosilicate glass nozzle/probe liner followed by a
heated filter assembly with a Teflon® filter support, a series of impingers and a meter box and
vacuum pump as specified in EPA Method 5. The sample is not exposed to any metal surfaces in
this train.
5-1
-------
to
o CLj
TEMPERATURE SENSOR
PROBE
/ \
PITOT TUBE
TEMPERATURE
SENSOR
\
REVERSE TYPE
PITOT TUBE
IMPINGER TRAIN OPTIONAL, MAY BE REPLACED
BY AN EQUIVALENT CONDENSER
STACK WALL
THERMOMETER
HEATED AREA
PROBE
6-7,
THERMOMETER
T CHECK VALVE
FILTER
HOLDER
PITOT
MANOMETER
ORIFICE
VACUUM
LINE
THERMOMETERS
T ~
i
DRY GAS
METER
IMPINGERS
BYPASS
VALVE
MAIN VACUUM
VALVE Q GAUGE
ORIFICE
MANOMETER
AIR TIGHT
PUMP
-------
The design and contents of the sequential impingers are:
• Impinger 1 (modified Greenburg-Smith) and Impinger 2 (Greenburg-Smith),
known volumes of deionized water (nominal 100 mL);
• Impinger 3 (modified Greenburg-Smith), empty; and
• Impinger 4 (modified Greenburg-Smith), silica gel.
5.1.2 Method 5A Sampling Equipment Preparation
5.1.2.1 Glassware Preparation
Glassware is washed in soapy water, rinsed with hot tap water, rinsed with Type n
water, rinsed with acetone to remove the water, rinsed three times with TCE and then air-dried.
This procedure is used on all the glass components of the sampling train including the glass
nozzles plus any sample bottles, Erlenmeyer flasks, petn dishes and graduated cylinders. Non-
glass components (such as the Teflon®-coated filter screens and seals, tweezers, Teflon® squeeze
bottles, Teflon® probe brushes) are cleaned following the same procedure. The cleaning
procedure is summarized in Table 5-1.
Table 5-1. Glassware Cleaning Procedure for Method 5A Sampling Train
Components
NOTE: USE DISPOSABLE GLOVES AND ADEQUATE VENTILATION
1. Soak all glassware in hot soapy water (laboratory detergent).
2. Rinse with tap water to remove soap.
3. Rinse three times with HPLC-grade water.
4. Rinse three times with acetone.
5. Rinse three times with TCE.
6. Air dry. •
7. Cap glassware with clean glass plugs or Parafilm®.
8. Mark cleaned glassware with color-coded identification sticker.
5-3
-------
5.1.2.2 Reagent Preparation
The filters in the sampling train were Whatman 94 AH glass fiber filters, without
organic binders, that met the criteria specified in EPA Method 5 Section 3.1.1. These filters were
used as received. The TCE was purchased as HPLC-grade to ensure a low residue after
evaporation.
The reagent water is HPLC-grade or equivalent. The lot number, manufacturer and
grade of each reagent that was used was recorded.
The analyst wears both safety glasses (with side shields) and protective gloves when the
reagents are mixed or handled. Each reagent had its own designated transfer Teflon® squeeze
bottle and was marked for identification and used only for the reagent for which it was
designated.
5.1.2.3 Equipment Preparation
The remaining equipment preparation included calibration and leak checking of all of
the train equipment, including meter boxes, thermocouples, nozzles, pitot tubes, and umbilicals.
Referenced calibration procedures were followed when available, and the results were properly
documented and retained. A discussion of the techniques used to calibrate the sampling
equipment is presented below.
Type S Pitot Tube Calibration. The EPA has specified guidelines concerning the
construction and geometry of an acceptable Type S pitot tube. If the specified design and
construction guidelines are met, a pitot tube coefficient of 0.84 is used. Information pertaining to
the design and construction of the Type S pitot tube may be found in detail in Section 3.1.1 of
EPA Document 600/4-77-027b. Only Type S pitot tubes meeting the required EPA
specifications were used. Pitot tubes were inspected and documented as meeting EPA
specifications prior to field sampling.
5-4
-------
Sampling Nozzle Calibration. Glass nozzles are used for isokinetic sampling.
Calculation of the isokinetic sampling rate requires that the cross-sectional area of the sampling
nozzle be accurately known. All nozzles were thoroughly cleaned, visually inspected and
calibrated according to the procedure outlined in Section 3.4.2 of EPA Document 600/4-77-
027b.
Temperature Measuring Device calibration. Accurate temperature measurements
are required during stationary source sampling. Bimetallic stem thermometers and thermocouple
temperature sensors are calibrated using the procedure described in Section 3.4.2 of EPA
Document 600/4-77-027b. Each temperature sensor was calibrated at a minimum of two points
over the anticipated range of use against an NBS-traceable mercury-in-glass thermometer. All
sensors were calibrated prior to field sampling.
Drv Gas Meter Device Calibration. Dry gas meters (DGMs) are used in the
Method 5A sampling trains to monitor the sampling rate and to measure the sample volume. All
DGMs were calibrated to document the volume correction factor just prior to shipping of the
equipment to the field. Post-test calibration checks are performed as soon as possible after the
equipment has been returned to the ERG laboratory. Pre- and post-test calibration should agree
to within 5%.
Prior to calibration, a positive pressure leak check of the system was performed using
the procedure outlined in Section 3.3.2 of EPA Document 600/4-77-023b. The system was
placed under approximately 10 inches of water pressure and a gauge oil manometer was used to
determine if a pressure increase could be detected over a one-minute period. If leaks are detected
at this point, they are repaired before actual calibration are performed and a successful leak check
is obtained for each system.
After the sampling console was assembled and leak checked, the pump ran for 15
minutes to allow the pump and DGM to warm up. The valve was then adjusted to obtain the
desired flow rate. For the pre-test calibration, data were collected at orifice manometer settings
5-5
-------
(AH) of 0.5, 1.0, 1.5, 2.0, 3.0 and 4.0 in H,O. Gas volumes of 5 ft' were used for the two lower
orifice settings, and volumes of 10 ft3 were used for the higher settings. The individual gas meter
correction factors (Yj) were calculated for each orifice setting and averaged. The method
requires that each of the individual correction factors fall within ±2% of the average correction
factor or the meter must be cleaned, adjusted, and recalibrated. In addition, ERG requires that
the average correction factor be within 1.00±1%. For the post-test calibration, the meter was
calibrated three times at the average orifice setting and vacuum which were used during the
actual field test.
5.1.3 Method 5A Sampling Operations
5.1.3.1 Preliminary Measurements
Prior to sampling, preliminary measurements are required to ensure isokinetic sampling.
These preliminary measurements include determining the traverse point locations, performing a
preliminary velocity traverse and a cyclonic flow check. These measurements are used to
calculate a "K factor." The K factor is used to determine an isokinetic sampling rate from stack
gas flow readings taken during sampling. Preliminary measurement datasheets from the field test
are shown in Appendix A.
Measurements are then made of the duct inside diameter, port nozzle length, and the
distances to the nearest upstream and downstream flow disturbances. These measurements are
then used to determine sampling point locations by following EPA Reference Method 1
guidelines. The distances are then marked on the sampling probe using an indelible marker.
5.1.3.2 Assembling the Train
Assembly of the EPA Method 5A sampling train components was begun in the recovery
trailer and final train assembly was performed at the specific stack testing location. First, the
empty, clean impingers were assembled and laid out in the proper order in the recovery trailer.
5-6
-------
Each ground glass joint was carefully inspected for hairline cracks. After the impingers were
loaded, each impinger was weighed, and the initial weight and contents of each impinger were
recorded on a recovery data sheet (see Appendix A).
The impingers were connected together using clean glass U-tube connectors and
arranged in the impinger bucket. The height of all the impingers is approximately the same to aid
in obtaining a leak-free seal. The open ends of the train were sealed with Parafilm® or clean
ground glass caps.
The pre-weighed filter was then loaded into the filter holder in the recovery trailer. The
filter holder was then capped off and placed into the impinger bucket. To avoid contamination of
the sample, sealing greases were not used. The train components were then transferred to the
sampling location and assembled as shown in Figure 5-1.
5.1.3.3 Sampling Procedures
After the train is assembled, the heaters for the probe liner and filter box were turned
on. When the system reached the appropriate temperatures, the sampling train was ready for pre-
test leak checking. The gas stream exiting the heated filter was maintained at a temperature of
42±10°C (108±18°F). To monitor this temperature during sampling, a thermocouple was
inserted directly into the gas stream at the end of the heated filter. The filter temperature was
initially set at 42±10°C (108±18°F) and the probe temperature at 100°C (212°F). The
temperature of these two heated zones was regulated as necessary to maintain the proper
temperature of the gas exiting the filter. Due to the physical constraints of vertical sampling at
the thermal oxidizer inlet, a flexible Teflon® line was used to connect the exit of the filter to the
first impinger. The heated probe and filter box were a single unit mounted on a vertical
monorail. The chilled impinger box was placed on the floor of the scaffolding.
The sampling trains were leak checked at the start and finish of sampling. EPA Method
5 protocol requires post-test leak checks and recommends pre-test leak checks. ERG protocol
5-7
-------
also incorporates leak checks before and after every port change. An acceptable pre-test leak rate
is less than 0.02 acfm (ftYmin) at approximately 15 inches of mercury (in. Hg). If, during
testing, a piece of glassware needed to be emptied or replaced, a leak check was performed
before the glassware piece was removed and after the train was re-assembled.
To perform a leak check of an assembled train, the nozzle end was capped off and a
vacuum of 15 in. Hg was pulled in the system. When the system was evacuated, the volume of
gas flowing through the system was timed for 60 seconds. After the leak rate was determined.
the cap was slowly removed from the nozzle end until the vacuum dropped off, and then the
pump was turned off. If the leak rate requirement was not met, the train was systematical!}
checked by first capping the train at the filter, at the first impinger. etc., until the leak was located
and corrected.
After a successful pre-test leak check had been conducted, all train components were at
their specified temperatures and initial data were recorded (DGM reading), the test was initiated.
Sampling train data were recorded every five minutes on standard data forms.
A checklist for sampling is included in Table 5-2.
The leak rates and sampling start and stop times are recorded in the sampling task log.
Any other events that occur during sampling are also recorded on the sampling task log, such as
pitot cleaning, thermocouple malfunctions, heater malfunctions, or any other unusual
occurrences.
At the conclusion of the test run, the sample pump (or flow) was turned off. the probe
was removed from the duct, a final DGM reading was taken, and a post-test leak check was
completed. The procedure was identical to the pre-test procedure. However, the vacuum was at
least one inch Hg higher than the highest vacuum attained during sampling. An acceptable leak
5-8
-------
Table 5-2. Sampling Checklist, EPA Method 5A
Prior to Test:
1. Check the set of impingers to ensure that the number of impingers is correct and that the
impingers are in the right order. Verify that the markings on the probe are correct; re-
mark, if necessary.
2. Check that the set of glassware is complete and that all the pieces are correct and in the
proper order and orientation.
3. Check that a sufficient number of the correct datasheets are available, and check
barometric pressure.
4. Check that sampling equipment is ready for Method 3 analysis.
5. Leak check pilot tubes.
6. Examine meter box - level it and confirm that the pump is operational.
7. Assemble train to the filter and leak check at 20 in. Hg. Attach probe to train and
perform final leak check; record leak rate and pressure on sampling log.
8. Check out thermocouples to verify that they are all reading correctly.
9.. Turn on heaters and ensure that their temperatures are increasing.
10. Check that cooling water is on and flowing (if required). Add ice to impinger bucket(s).
11. Check isokinetic K Factor to be sure it has been calculated correctly. Refer to previous
results to confirm assumptions. Two people should perform this calculation
independently to double check the accuracy.
12. Have a spare probe liner, probe sheath, meter box, and filter prepared at the testing
location.
During the Test:
1. Notify Test Crew Leader of any sampling problems ASAP. Fill in sampling log.
2. Perform simultaneous/concurrent testing with other locations (if applicable). Maintain
filter temperature at 42±10°C (108±18°F), and keep temperature as steady as possible.
Maintain impinger temperatures below 68°F. Maintain probe temperature above 212°F
or at the temperature required to maintain the correct filter temperature.
3. Perform leak checks between ports and record on sampling log.
4. Record sampling rate, times, ad location for the fixed gas (CO2, O,) sample (if
applicable).
5. Blow back pitot tubes every 15 minutes.
5-9
-------
Table 5-2. Continued
6. Change filter if pressure drop exceeds 20 in. Hg.
7. Check impinger containing silica gel every 0.5 hour. If indicator changes color, request a
pre-filled impinger from the recovery trailer and replace.
8. Check manometer fluid levels and zero every hour.
After Test is Completed:
1. Record final meter reading.
2. Check data sheet for completeness.
3. Perform final leak check of sampling train at maximum vacuum achieved during test.
4. Leak check each leg of pitot tubes.
5. Disassemble train. Cap sections. Take sections to recovery trailer.
6. Probe/cyclone recovery (use 500 mL bottles)
For TCE rinses (all trains)
Attach flask to end of probe (use separate recovery containers for each
sampling location)
Add 50 mL TCE.
Put a brush down the probe, and brush back and forth
Rinse back and forth in probe with TCE
Empty TCE rinse into sample jar
Repeat brushing and TCE rinse three times, so that the final combined TCE
rinse volume is < 150 mL.
7. Reattach nozzle and cap for next day. Store in dry safe place.
8. Make sure that data sheets are filled out completely and give to test crew leader.
rate is less than 4% of the average sampling rate or 0.02 acfm (whichever is lower). All final
leak rates on site met the acceptance criterion.
5-10
-------
5.1.4 Method 5A Sample Recovery
Recovery procedures were initiated as soon as the probe was removed from the stack
and the post-test leak check was completed.
To facilitate transfer from the sampling location to the recovery trailer, the sampling
train was disassembled into three sections:
• Nozzle/probe liner;
• Filter holder; and
• Impingers in their bucket.
Each of these sections was capped with Teflon® tape before removal to the recovery trailer. All
train components were rinsed and the samples were collected in separate, pre-labeled, pre-
cleaned sample containers to avoid cross-contamination of inlet and outlet samples. Trains of
each type were recovered in separate areas of the recovery trailer to avoid cross-contamination or
mistakes in recovery sequences.
Once in the recovery trailer, the sampling trains were recovered as separate front and
back half fractions. A diagram illustrating front half and back half sampling train recovery
procedures is shown in Figure 5-2.
No equipment with exposed metal surfaces was used in the recovery procedures. The
weight gain in each of the impingers was recorded to determine the moisture content in the flue
gas. Following weighing of the impingers, the front half of the train was recovered, including the
front one half of the filter holder, filter and all sample-exposed surfaces forward of the filter. The
probe liner was rinsed with TCE by tilting and rotating the probe while squirting TCE into its
5-11
-------
t—>
to
Probe Liner &
Nozzle
Front Half of Filter _.„
., . Filter
Housing
Impinger Water Silica Gel
|
|
Rinse with TCE
Brush liner with
Teflon® brush
and rinse with
TCE
Rinse with TCE
F. H.
Container #2
Rinse 3x with TCE
Remove filter from
support with
tweezers and place
in petri dish
Brush loose
particulate onto
filter
Seal petri dish with
Teflon® tape
Container #1
Measure volume in
impingers 1-3
Discard
Measure weight gain
Discard
Figure 5-2. Method 5A Sample Recovery Scheme
-------
upper end so that all inside surfaces were wetted. The TCE was quantitatively collected into the
appropriate bottle. This rinse was followed by two additional brush/rinse sequences, using a
non-metallic brush; the probe was held in an inclined position and TCE was squirted into the
upper end as the brush was pushed through with a twisting action. All of the TCE and particulate
material were caught in the sample container. The brush/rinse procedure was repeated until no
visible particulate matter remained and finished with a final TCE rinse of the probe and brush.
The front half of the glass filter holder was also rinsed with TCE until all visible particulate
matter was removed. These rinses were added to the probe rinses. After all front half TCE
washes were collected, the cap of the container was tightened, the liquid level was marked, and
the bottle was weighed to determine the TCE rinse volume. This sample was labeled as
Container #2. EPA Method 5 A specifies that a nominal volume of 100 mL of TCE must be used
for rinsing these sampling train components. For blank correction purposes, the exact weight or
volume of TCE used was measured. A TCE reagent blank of approximately the same volume as
the TCE rinses was analyzed with the samples.
The filter was carefully removed from the filter support and placed in a clean, well-
marked glass petri dish (labeled Container #1) and sealed with Teflon® tape.
After the volume of liquid in the first three impingers was measured and the weight
gain of the desiccant in the fourth impinger was determined, the contents of the impingers were
discarded. The measured volume of water was used to adjust the moisture content of the
sampled gas stream calculations.
5.1.5 Quality Control for Gravimetric Procedures
All quality control procedures specified in the test method were followed. All field
reagent blanks were processed and analyzed as specified in the test method. Prior to each
gravimetric determination, the balance calibration was verified using a series of certified weights
covering the range of weights encountered for the samples.
5-13
-------
5.2 Chlorinated Dibenzo-p-Dioxin and Chlorinated Dibenzofuran Emissions
Testing Using EPA Method 23
The sampling and analytical method for determining flue gas emissions of
polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans (PCDD/PCDF) is EPA
Method 23.
5.2.1 Method 23 Sampling Equipment
EPA Method 23 requires the sampling train shown in Figure 5-3.
The EPA Method 23 sampling train is similar to a Method 5 sampling train, with the
following exceptions:
All components (glass probe/nozzle, probe liner, all other glassware, filters) are
pre-cleaned using solvent rinses and extraction procedures; and
A condensing coil and XAD-2® resin sorbent module are located between the
filter and the impinger train.
All sampling equipment specifications are detailed in EPA Method 23.
5.2.2 Method 23 Equipment Preparation
In addition to the standard EPA Method 5 requirements, Method 23 includes several
unique preparation steps which ensure that the sampling train components are not contaminated
with organic compounds that may interfere with analysis. The glassware, glass fiber filters, and
sorbent resin were cleaned and the filters and resin were checked for 42 residuals before they
were packed for shipment to the test site.
5-14
-------
i
Gas Flow
i
Ui
stack wall
heated glass liner
A
"S" type
pilot
manometer
gas exit
manometer
temperature
sensor
* condenser
heat
-T
i
i
-d
i
1
i
^
"
filler
holder
recirculation
temperature
/
1
pump ^^^
1
calibrated
empty
orifice
L
t
dry gas
sensor
!
^ XAD-2trap
^
temperature
sensor
T
I
FJ fJ jfJ T-J ifJ
I • : "1
i ice
' i bath i 1
1 | i
1
< W A A !
100ml .
empty siica
HPLC Water y
gel
fine
meter
1
i
j
S
I
t
> ^vacuum
X( } "^^H
^ gauge
J Y
V
coarse
-^^ vacuum pump
/ \
Figure 5-3. Method 23 Sampling Train Configuration
-------
5.2.2.1 Glassware Preparation
Glassware was cleaned as shown in Table 5-3. Glassware was washed in soapy water,
rinsed with distilled water, baked, and then rinsed with acetone followed by methylene chloride.
Clean glassware was allowed to dry under a hood loosely covered with aluminum foil to prevent
laboratory contamination. Once the glassware was dry, the ends exposed to air were sealed with
methylene chloride-rinsed aluminum foil. All the glass components of the sampling train,
including the glass nozzles plus any flasks, petri dishes, graduated cylinders, and pipets that are
used during sampling and recovery were cleaned according to this procedure. Non-glass
components (such as the Teflon®-coated filter screens and seals, tweezers, Teflon® squeeze
bottles, Teflon® probe/nozzle brushes) were cleaned following the same procedure except that no
baking was performed.
5.2.2.2 XAD-2® Sorbent Resin and Filter Preparation
XAD-2® sorbent resin and glass fiber filters were pre-cleaned by separate procedures
according to EPA Method 23. Only pesticide-grade solvents and HPLC-grade water were used
to prepare for D/F sampling, and to recover these sampling trains. The lot number, manufacturer,
and grade of each reagent were recorded in the laboratory notebook.
To prepare the filters, a batch of 50 filters was placed in a Soxhlet extractor pre-cleaned
by extraction with toluene. The Soxhlet extractor was charged with fresh toluene and refluxed
for 16 hours. After the extraction, the toluene was analyzed as described in Sections 5.2 and 5.3
of EPA Method 23 for the presence of polychlorinated dibenzodioxins and dibenzofurans. If
these analytes were found, the filters were re-extracted until the analyte was not detected. The
cleaned filters were then dried under a clean nitrogen stream. Each filter was individually
checked for holes, tears, creases or discoloration and, if any were found, was discarded.
Acceptable filters were stored in a pre-cleaned petri dish, labeled by date of analysis, and sealed
with Teflon® tape.
5-16
-------
Table 5-3. EPA Method 23 Glassware Cleaning Procedure for Train Components,
Sample Containers, and Laboratory Glassware
1. Soak all glassware in hot soapy water (laboratory detergent).
2. Rinse with tap water to remove soap.
3. Rinse three times with distilled/deionized water.
4. Bake at 450°F for 2 hours.3
5. Rinse three times with pesticide-grade acetone.
6. Rinse three times with pesticide-grade methylene chloride.
7. Cap rinsed glassware with clean glass plugs or methylene chloride-rinsed
aluminum foil.
8. Mark cleaned glassware with color-coded identification sticker.
9. Rinse glassware immediately before using with acetone and methylene chloride
(laboratory proof).
aStep (4) has been added to the cleanup procedure to replace the dichromate soak specified in the
reference method. ERG has demonstrates in the past that Step (4) sufficiently removes organic
artifacts. Step (4) is not used for probe liners and non-glass components of the sampling train
that cannot withstand a temperature of 450°F (i.e., Teflon®-coated filter screens and seals,
tweezers, Teflon® squeeze bottles, Teflon® probe/nozzle brushes).
To prepare the sorbent resin, the XAD-2® polymer resin was cleaned in the following
order:
• Rinse with HPLC-grade water; discard water;
• Soak in HPLC-grade water overnight; discard water;
Extract in a Soxhlet extractor with HPLC-grade water for 8 hours; discard water;
Extract in a Soxhlet extractor with methanol for 22 hours; discard methanol;
5-17
-------
Extract in a Sbxhlet extractor with methylene chloride for 22 hours; discard
methylene chloride;
Extract in a Soxhlet extractor with methylene chloride for 22 hours, retain an
aliquot of the solvent for analysis for PCDDs/PCDFs by HRGC/HRMS; and
Dry resin under a clean nitrogen stream.
Once the resin was completely dry, it was checked for the presence of methylene
chloride, PCDDs, and PCDFs as described in Section 3.1.2.3.1 of EPA Method 23. If any
analytes were found, the resin was re-extracted. If methylene chloride was found, the resin was
dried until the excess solvent was removed. The sorbent was used within four weeks of cleaning.
The cleaned XAD-2® resin was spiked before shipment to the field test site with five
PCDD/PCDF internal standards. Due to the special handling considerations required for the
internal standards, the spiking was performed by Triangle Laboratories. For convenience and to
minimize the potential for contamination, Triangle Laboratories also performed the resin and
filter cleanup procedures and loaded the resin into the glass sampling modules.
5.2.2.3 Method 23 Sampling Train Preparation
The remaining preparation of the Method 23 sampling train included calibration and
leak checking of all sampling train equipment, including meter boxes, thermocouples, nozzles,
pilot tubes, and umbilicals. Referenced calibration procedures were followed when available.
The results were properly documented and retained.
5.2.3 Method 23 Sampling Operations
5.2.3.1 Preliminary Measurements
Prior to sampling, preliminary measurements were taken to ensure isokinetic sampling.
These preliminary measurements included determining the traverse point locations, performing a
5-18
-------
preliminary velocity traverse, cyclonic flow check and moisture determination. These
measurements were used to calculate a K factor. The K factor was used to determine an
isokinetic sampling rate from stack gas flow readings taken during sampling. Measurements
were then made of the duct inside diameter, port nozzle length, and the distances to the nearest
upstream and downstream flow disturbances. These measurements were then used to determine
sampling point locations by following EPA Reference Method 1 guidelines. The distances were
then marked on the sampling probe using an indelible marker.
5.2.3.2 Assembling the EPA Method 23 Sampling Train
The components of the EPA Method 23 sampling train components were gathered in
the recovery trailer and final assembly of the sampling train was performed at the stack location.
First, the empty clean impingers were assembled and laid out in the proper order in the recovery
trailer. Each ground glass joint was carefully inspected for hairline cracks. The first impinger
was a knockout impinger which has a short tip. The purpose of this impinger was to collect
condensate which formed in the coil and XAD-2® resin sampling module. The next two
impingers were modified-tip impingers, which each contained 100 mL of HPLC-grade water.
The fourth impinger was empty, and the fifth impinger contained 200-300 grams of blue
indicating silica gel. After the impingers were loaded, each impinger was weighed, and the
initial weights and contents of each impinger were recorded on a recovery data sheet. The
heights of all the impingers were approximately the same to aid in obtaining a leak-free seal. The
open ends of the train were sealed with methylene chloride-rinsed aluminum foil, or clean ground
glass caps.
The filter was loaded into the filter holder in the recovery trailer. The filter holder was
then capped off and placed with the sorbent resin sampling module and condenser coil (capped)
into the impinger bucket. A supply of pre-cleaned foil and socket joints was also placed in the
bucket inside a clean plastic bag for the convenience of the samplers. Sealing greases were not
used to avoid contamination of the sample. The train components were transferred to the
sampling location and assembled as shown in Figure 5-3.
5-19
-------
5.2.3.3 Sampling Procedures
After the train was assembled, the heaters were turned on for the probe liner and heated
filter box, and the sorbent module/condenser coil recirculating pump was turned on. When the
system reached the appropriate temperatures, the sampling train was ready for a pre-test leak
check. The temperature of the sorbent resin sampling module must not exceed 50°C (120°F) at
any time and during testing the temperature must not exceed 20°C (68°F). The filter temperature
was maintained at 120±14°F (248±25°F). The probe temperature was maintained above 100°C
(212°F).
The sampling trains were leak checked at the start and finish of sampling. (EPA
Method 5/23 protocol requires only post-test leak checks and recommends pre-test leak checks.)
ERG protocol also incorporates leak checks before and after every port change. An acceptable
pre-test leak rate is less than 0.02 acfm (ftVmin) at approximately 15 in. Hg. If, during the
testing, a piece of glassware needed to be replaced or emptied, a leak check was performed
before the glassware piece was removed, and after the train was re-assembled.
To perform a leak check of the assembled sampling train, the nozzle end was capped off
and a vacuum of 15 in. Hg was pulled in the system. When the system was evacuated, the
volume of gas flowing through the system was timed for 60 seconds. After the leak rate was
determined, the cap was slowly removed from the nozzle end until the vacuum dropped off, and
then the pump was turned off. If the leak rate requirement was not met, the sampling train was
systematically checked by first capping the train at the filter, at the first impinger, etc., until the
leak was located and corrected.
After a successful pre-test leak check has been conducted, all train components were at
their specified temperatures, and initial data and DGM readings were recorded, the test was
initiated. Sampling train data were recorded every five minutes on standard data forms. A
checklist for PCDD/PCDF sampling is shown in Table 5-4. In performing PCDD/PCDF
5-20
-------
Table 5-4. EPA Method 23 Sampling Checklist
Prior to Testing:
1. Check impinger set to verify the correct number, order, and orientation of impingers.
Verify probe markings, and re-mark if necessary.
2. Check glassware to ensure that the number of pieces of each type is correct. Have a
spare probe liner, probe sheath, meter box, and filter prepared at the test location.
3. Check that a sufficient number of the correct data sheets are available; check barometric
pressure.
4. Prepare sampling equipment for CO2/O2, unless CEMs are being used for CO,/O2
determination.
5. Examine meter box: level, zero the manometers, and confirm that the pump is
operational.
6. Verify that the filter is loaded correctly and as tightly as possible. Place filter in line with
the train and leak check at 15 in. Hg.
7. Add probe to sampling train.
8. Check thermocouples; make sure that they are reading correctly.
9. Conduct pitot leak check and recheck manometer zero.
10. Perform final leak check: record leak rate and vacuum on sampling log.
11. Turn on heaters and verify that heat is increasing.
12. Check that cooling water is on and flowing. Add ice to impinger bucket(s).
13. Check isokinetic K factor to be sure that it has been calculated correctly. Refer to
previous results to confirm assumptions. Two people should perform this calculation
independently to double check it.
During Test:
1. Notify Test Crew Chief of any sampling problems ASAP. Train operator should fill in
sampling log and document any abnormalities.
2. Perform simultaneous/concurrent testing with other locations (if applicable). Maintain
filter temperature between 248°F±25°F. Keep temperature as steady as possible.
Maintain the sorbent sampling module and impinger temperatures below 68 °F. Maintain
probe temperature above 212°F.
3. Perform leak checks between ports and record on data sheet. Perform a leak check if the
test is stopped to change silica gel, to decant condensate, or to change filters.
5-21
-------
Table 5-4. Continued
4. Record sampling times, rate, and location for the fixed gas bag sampling (CO, CO2, O;),
if applicable.
5. Blow back pilot tubes periodically if moisture entrapment is expected.
6. Change filter if vacuum suddenly increases or exceeds 20 in. Hg.
7. Check impinger solutions every 0.5 hour. If the knockout impinger is approaching full,
stop the test and empty the knockout impinger into a pre-weighed bottle and replace the
empty knockout impinger in the train. Perform a leak check, and continue testing.
8. Check the silica gel impinger every 0.5 hour. If the color of the indicator begins to fade,
request a pre-filled, pre-weighed impinger from the recovery trailer.
9. Check the ice in the impinger bucket frequently. If stack gas temperatures are high or the
moisture content of the emission stream is high, the ice at the bottom of the impinger
bucket will melt rapidly. Add more ice, as needed. Maintain the temperature of the gas
at the condenser coil and silica gel impinger below 20°C (68°F).
After Test is Completed:
1. Record final DGM reading.
2. Perform final leak check of the sampling train at the maximum vacuum achieved during
the test.
3. Perform final pilot leak check, check the functions (level, zero, etc.), and inspect for tip
damage.
4. Check thai the data sheet has been filled oul completely. Verify lhal ihe impinger bucket
identification is recorded on Ihe dala sheels. Nole any abnormal conditions.
5. Disassemble sampling irain, cap sections, and lake each section and all data sheels down
lo Ihe recovery irailer.
6. Use 500 mL bottles lo recover the probe.
a) Bring probes inlo Ihe recovery irailer (or olher enclosed area).
b) Wipe ihe exterior of Ihe probe lo remove any loose material lhal could
coniaminale Ihe sample.
c) Carefully remove ihe nozzle/probe liner and cap il off wilh aluminum foil lhat
has been rinsed with melhylene chloride.
d) For acetone rinses (all trains)
Attach a pre-cleaned cyclone flask lo Ihe probe to calch rinses.
5-22
-------
Table 5-4. Continued
Wet all sides of the probe interior with acetone
While holding the probe in an inclined position, put pre-cleaned probe
brush down into probe and brash it in and out
Rinse the brush, while in the probe, with acetone
Repeat the brush/rinse sequence at least three times until all of the
particulate material has been removed
Recover acetone into a pre-weighed, pre-labeled sample container
e) Repeat the brash/rinse sequence at least three times using methylene chloride.
Recover the methylene chloride into the same acetone recovery bottle.
f) Repeat the brush/rinse sequence at least three times using toluene. Recover the
toluene into a separate pre-weighed pre-labeled-sample container.
7. Cap both ends of the nozzle/probe liner for the next day, and store in a dry safe place.
8. Make sure that data sheets are completely filled out and legible, and give completed data
sheets to the Test Crew Chief.
sampling, the temperature of the gas entering the resin sampling module must be below 20 °C
(68 °F). The gas was cooled by a water jacket condenser through which ice water was circulated.
The leak rates and sampling start and stop times were recorded on the field sampling
data sheets. Also, any other events that occurred during sampling were recorded on the task log
(e.g., sorbent module heat excursions, pitot cleaning, thermocouple malfunctions, heater
malfunctions, or any other unusual occurrences).
At the conclusion of the sampling run, the sample pump (or flow) was turned off, the
probe was removed from the duct, a final DGM reading was taken, and a post-test leak check
was completed. The leak test procedure is identical to the pre-test procedure. However, the
vacuum for the post-test leak check should be at least one inch Hg higher than the highest
vacuum attained during sampling. An acceptable leak rate is less than 4% of the average
sampling rate or 0.02 acfm (whichever is lower).
5-23
-------
5.2.4 PCDD/PCDF Sample Recovery
To facilitate transfer from the sampling location to the recovery trailer, the sampling
train was disassembled into the following sections:
• Probe liner;
• Filter holder;
• Filter to condenser glassware;
• Condenser and sorbent module; and
• Impingers in their bucket.
Each of these sections was capped with ground glass caps or aluminum foil that had
been rinsed with methylene chloride before removal to the recovery trailer. Once in the recovery
trailer, field recovery followed the scheme shown in Figure 5-4.
The samples were recovered and stored in cleaned amber glass bottles to prevent
degradation by light.
The probe and nozzle were first rinsed with approximately 100 mL of acetone and
brushed to remove any paniculate matter. This first rinse was followed with a rinse of methylene
chloride. Both of these rinses were collected in the same bottle. The same two solvents were
used to rinse the cyclone, front/back half of the filter holder, filter support, connecting glassware,
and condenser. These rinses were added to the probe rinse bottle. All of the components listed
above were again rinsed with toluene and the toluene rinses were collected in a separate
container.
The contents of Impingers 1-4 (water) were discarded.
5-24
-------
to
Probe Probe Cyclone Front Half of ' Filler Back Half of Connecting Condenser
N
Rinse w
>zzle L
th Acetone A
ncr
(ach
Filter 1
lousing Sup
port Filter He
•Rising Lii
e
Unlil all Paniculate 250 mL flask Brush and Brush and
is Removed to Ball Joint rinse with rinse with Rinse with Rinse with Rinse with Rinse with
acetone (3x) acetone (3x) acetone (3x) acetone (3x) acetone (3x) acetone (3x)
Rinse wi
Rinse with
Acetone
Empty Flask
into 950 inL
Bottle
Brush Liner and Rinse
with 3 Aliquols of
Acetone
Check Liner to see if
Paniculate is
removed; if not, repeat
h 3 Aliquots Rin
se with R
use with Rins
c with
of Melhylene 3 Aliquols 3 Aliquots 3 Aliquots
Chloride ofMethylene ofMethylene ofMcthylene
Chloride Chloride Chloride
Reco
Rinse with incthylcne Rinse with Rinse with
chloride (3x) (at least methylcnc incthylcne
once let the linsc stand chloride chloride (3x)
5 inin in unit) (3x) (at least once let
the rinse stand
5 inin in unit)
vcr into
preweighed
b
Mile
PR 1
Rinse with Rinse with Rinse with Rinse with Rinse with Toluene (3x) Rinse with Toluene (3x) (at
Toluene (3x)* Toluene (3x) Toluene (3x) Toluene (3x) (at least once let the rinse Toluene (3x) least once let the
stand 5 inin in unit) rinse stand 5 mm
*This fraction should not be combined with the other toluene
fractions
PRT/CRT
in uni
0
Figure 5-4. Method 23 Field Recovery Scheme
-------
Filter
Resin Trap
1st Impinger
(knockout)
2nd Impinger
3rd Impinger
4ih Impinger
5lh Impinger
(silica gel)
Carefully
remove filter
from support
with tweezers
Secure XAD
trap openings
with glass balls
and clamps
Weigh
Impinger
Weigh
Impinger
Weigh
Impinger
Weigh
Impinger
Weigh
Impinger
ON
Brush loose
Paniculate
onto filter
Seal in
pelri dish
Place in cooler
for storage
Record weight
and
calculate gain
Record weight
and
calculate gain
Record weight
and
calculate gain
Record weight
and
calculate gain
Record weight
and
calculate gain
Save for
regeneration
Discard
Note: See Table 5-5 for Sample Fractions Identification
Figure 5-4. Method 23 Field Recovery Schedule (continued)
-------
All solvents used for sampling train recovery were pesticide-grade. The use of the
highest grade of reagents for sampling train recovery was essential to prevent the introduction of
chemical impurities which interfere with the quantitative analytical determinations.
The sampling train components recovered in the field are listed in Table 5-5. The
XAD-2® sorbent module was stored in coolers on ice at all times. The samples were delivered to
the analytical laboratory upon return to ERG, accompanied by written information designating
the analyses to be performed.
Table 5-5. Method 23 Sample Fractions Shipped to the Analytical Laboratory
Container
1
2
3
4
Code
F
Pra
PRT
CRT
SM
Fraction
Filter(s)
Acetone and methylene chloride rinses of nozzle/probe, cyclone,
front half/back half filter holder, filter support, connecting
glassware, condenser
Toluene rinse of nozzle/probe, cyclone, front half/back half filter
holder, filter support, connecting line and condenser
XAD-2® sorbent -module
"Rinses include acetone and methylene chloride recovered into the same sample bottle.
5.2.5 PCDD/PCDF Analytical Procedures
The analytical procedure used to obtain analyte concentrations from a single flue gas
sample incorporates HRGC and HRMS, with mass spectrometric resolution from 8000-10000.
The PCDD/PCDF congeners that were designated as target analytes are listed in Table 5-6. The
analyses were performed by Triangle Laboratories, Inc., using EPA Method 23/8290.
.*
The Method 23 samples were prepared and analyzed according to the scheme shown in
Figure 5-5. The XAD-2® and filter (along with the concentrated acetone/methylene chloride
5-27
-------
Table 5-6. PCDD/PCDF Congeners To Be Analyzed
DIOXINS:
2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD)
l,2,3,7,8-pentachlorodibenzo-p-dioxin(l,2,3,7,8-PeCDD)
1,2,3,4,7,8-hexachlorodibenzo-/?-dioxin (1,2,3,4,7,8-HxCDD)
l,2,3,6,7,8-hexachlorodibenzo-p-dioxin(l,2,3,6,7,8-HxCDD)
l,2,3,7,8,9-hexachlorodibenzo-/7-dioxin(l,2,3,7,8,9-HxCDD)
1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin (1,2,3,4,6,7,8-HpCDD)
Total octachlorinated dibenzo-p-dioxin (OCDD)
FURANS:
2,3,7,8-tetrachlorodibenzofuran(2,3,7,8-TCDF)
1,2,3,7,8-pentachlorodibenzofuran (1,2,3,7,8-PeCDF)
2,3,4,7,8-pentachlorodibenzofuran(2,3,4,7,8-PeCDF)
1,2,3,4,7,8-hexachlorodibenzofuran (1,2,3,4,7,8-HxCDF)
1,2,3,6,7,8-hexachlorodibenzofuran (1,2,3,6,7,8-HxCDF)
2,3,4,6,7,8-hexachlorodibenzofuran(2,3,4,6,7,8-HxCDF)
1,2,3,7,8,9-hexachlorodibenzofuran (1,2,3,7,8,9-HxCDF)
1,2,3,4,6,7,8-heptachlorodibenzofuran (1,2,3,4,6,7,8-HpCDF)
1,2,3,4,7,8,9-heptachlorodibenzofuran (1,2,3,4,7,8,9-HpCDF)
Total octachlorinated dibenzofuran (OCDF)
5-28
-------
Toluene
Rinses
Rotovap
to 1 mL
Add to SOX
for Toluene
Extraction
AC/MeCl2
Rinses
XAD-2®
Prespiked with
five
PCDD/PCDF
Standards
KDto
ImL
Add to SOX for Toluene
Extraction
Filter +
XAD-2® Add
to Soxhlet
Spike with
PCDD/PCDF Standards
Soxhlet in Toluene
Split 1:1
50% Toluene Extract to
Dioxins
Do PCDDs/Fs Cleanup
Analyze for PCDDs/Fs Method 8290X
Figure 5-5. Extraction and Analysis Schematic for Method 23 Samples
5-29
-------
rinses) were extracted with toluene, the extract processed through a column cleanup procedure,
concentrated, and analyzed. For the D/F analysis, isotopically-labeled surrogate compounds and
internal standards and surrogates that were used are described in detail in EPA Method 23/8290.
Data from the mass spectrometer were recorded and stored on a computer file as well as
printed on paper. Results such as amount detected, detection limit, retention time, and internal
standard and surrogate standard recoveries were calculated by computer. The chromatograms
were retained by the analytical laboratory with copies included in the analytical report delivered
to ERG.
5.2.5.1 Preparation of Samples for Extraction .
Upon receipt of the sample shipment at Triangle Laboratories, the samples were
checked against the Chain of Custody forms and then assigned an analytical laboratory sample
number. Each sample component was re-weighed to determine if leakage had occurred during
travel. Color, appearance, and other particulars of the samples were noted. Samples were
extracted within 21 days of collection and processed through cleanup procedures before
concentration and analysis.
5.2.5.2 Calibration of the GC/MS System
A five-point calibration curve for the GC/MS was prepared to demonstrate the linearity
of the instrument response over the concentration range of interest. Relative response factors
were calculated for each congener shown in Table 5-6; the response factors from the multipoint
curve were verified on a daily basis using a continuing calibration standard consisting of a mid-
level calibration standard. Instrument performance was acceptable only if the measured response
factors for the labeled and unlabeled compounds and the ion abundance ratios were within the
allowable limits specified in EPA Method 23.
5-30
-------
5.2.6 PCDD/PCDF Analytical Quality Control
All quality control procedures specified in Method 23 were followed. Blanks were used
to determine analytical contamination, calibration standards were used for instrument calibration
and linearity checks, internal standards were used to determine isomer recoveries and adjust
response factors for matrix effects, surrogate standards were used to measure the collection
efficiency of the sampling methodology, and an alternate standard was used as a column
efficiency check.
5.2.6.1 PCDD/PCDF Quality Control Blanks
Four different types of sample blanks were collected for D/F analysis. The types of
blanks that were required are shown in Table 5-7.
Table 5-7. Method 23 Blanks Collected in the Field
Blank
Field Blanks
Glassware
Proof Blank
Method Blank
Reagent Blanks
Collection
One blank collected and analyzed for each sample
location
Each train to be used (2) will be loaded and
quantitatively recovered prior to sampling
At least one for each analytical batch (laboratory
generated)
One 500 mL sample for each reagent and lot
Analysis
Analyze with field
samples
Archive for
potential analysis
Analyze with each
analytical batch of
field samples
Archive for
potential analvsis
Reagent blanks consisting of 1000 mL of each reagent used at the test site were saved
for potential analysis. Each reagent blank was of the same lot as was used during the sampling
program. Each lot number and reagent grade was recorded on the field blank label and in the
laboratory notebook (acetone, methylene chloride, toluene HPLC-grade water, filter, XAD-2®).
5-31
-------
A glassware blank (proof blank) was recovered from each set of sampling train
glassware that was used to collect the samples in the field. The pre-cleaned glassware, which
consists of a probe liner, filter holder, condenser coil, and impinger set, was loaded as if for
sampling and then used to check the effectiveness of the glassware cleaning procedure only if
sample analysis indicated a potential contamination problem.
A field blank was collected from a set of D/F glassware that had been used to collect at
least one sample and had been recovered. The sampling train was re-loaded, leak checked, and
left at a sampling location during a test run. The train was then recovered. The purpose of the
field blank was to measure the level of contamination that occurs from handling, loading,
recovering, and transporting the sampling train. The field blanks were analyzed with the flue gas
samples. If the field blank results are unsatisfactory in terms of contamination, reagent blanks
may be analyzed to determine the specific source of the contamination.
In addition to the three types of blanks that are required for the sampling program, the
analytical laboratory analyzed a method blank with each set of flue gas samples. This method
blank consisted of preparing and analyzing an aliquot of toluene by the exact procedure used for
the field sample analysis. The purpose of the method blank was to verify that there was no
laboratory contamination of the field samples.
5.2.6.2 Quality Control Standards and Duplicates
Recoveries of the internal standards must be between 40 and 130% for the tetra-
through hexachlorinated congeners and in the range of 25 to 130% for the hepta- and
octachlorinated homologs. If these requirements are not met, the data will be acceptable if the
signal-to-noise ratio is greater than or equal to 10. If these requirements are met, the results for
the compounds sampled can be adjusted according to the internal standard recoveries.
5-32
-------
Surrogate standard recoveries must be between 70 and 130%. If the recoveries of all
surrogate standards are less than 70%, the surrogate standard recovery results can be used to
adjust the results for the sampled species.
5.2.7 Analytes and Detection Limits for Method 23
The compounds of interest are the tetra- through octachlorinated dibenzodioxins and
tetra- through octachlorinated dibenzofurans. The detection limit for the individual compounds
is dependent upon the detection limit of the analytical method, the volume of the final extract,
and the total volume of gaseous sample collected in the sampling train. Following the
Method 23 protocol, the fractions to be analyzed from each sampling train include:
• Fraction 1—Filter;
Fraction 2—XAD-2® sorbent module;
• Fraction 3—Acetone and methylene chloride rinses of all train components prior
to sorbent module; and
• Fraction 4—Toluene rinses of all train components prior to the sorbent module.
Following the sample preparation protocol of EPA Method 23, a single combined
sample is prepared for analysis for D/F by HRGC/HRMS. The individual components of this
sample are no longer available for analysis. The final volume of this combined sample was
200 uL; the injection into the HRGC/HRMS was 2 uL. Using an instrument detection limit of
50 pg for tetra-, 250 pg for penta- through hepta-, and 500 pg for octachlorinated compounds, the
total minimum detectable amounts were calculated and are shown in Table 5-8. Using a four-
hour sampling time as selected by the EPA Work Assignment Manager at an assumed sampling
rate of 0.75 cfm, the Method Detection Limits shown in Table 5-9 were achieved. The sampling
flow rate at the outlet location was dictated by the flow rate of the stack gas since isokinetic
sampling was performed. .
5-33
-------
Table 5-8. Analytical Detection Limits (HRGC/HRMS) for Dioxins/Furans
Analyte
Tetra CDDs
Penta CDDs
Hexa CDDs
Hepta CDDs
Octa CDD
Tetra CDFs
Penta CDFs
Hexa CDFs
Hepta CDFs
Octa CDF
Total Detectable Amount, ng
5
25
25
25
50
5
25
25
25
50
NOTE: D/F analysis by High Resolution Mass Spectrometry assumes a 2 //L injection of a
200 yuL sample extract.
Table 5-9. PCDD/PCDF Method Detection Limits HRGC/HRMS
Sampling Time. Hours
Sampling Rate, cfm
Sample Volume, m3
Tetra CDDs
Penta CDDs
Hexa CDDs
Hepta CDDs
Octa CDD
Tetra CDFs
Penta CDFs
Hexa CDFs
Hepta CDFs
Octa CDF
4
0.75
5.1
MDL, ng/m3
0.98
4.9
4.9
4.9
9.8
0.98
4.9
4.9
4.9
9.8
5-34
-------
5.3 EPA Methods 1-4
5.3.1 Traverse Point Location by EPA Method 1
The number and location of sampling traverse points necessary for isokinetic and flow
sampling were dictated by the EPA Method 1 protocol. These parameters were based upon the
duct distance separating the sampling ports from the closest downstream and upstream flow
disturbances. The minimum number of traverse points for a circular duct with an i.d. of 12 feet
is 12.
5.3.2 Volumetric Flow Rate Determination by EPA Method 2
Volumetric flow rate was measured according to EPA Method 2. A Type K
thermocouple was used to measure flue gas temperature and a Type S pitot tube was used to
measure flue gas velocity.
5.3.2.1 Sampling Equipment Preparation
The pitot tubes were calibrated before use following the directions presented in EPA
Method 2. The pitot tubes were leak checked before and after each sampling run.
5.3.2.2 Sampling Operations
The parameters that were measured include the pressure drop across the pitot tubes,
stack temperature, stack static and ambient pressure. These parameters were measured at each
traverse point, as applicable. A computer program was used to calculate the average velocity
during the sampling period.
5-35
-------
5.3.3 O2 and CO2 Concentrations by EPA Method 3
The O2 and CO2 concentrations were determined by Fyrite following EPA Method 3.
Flue gas was extracted from the duct for analysis. The Method 3 analysis for O2 and CO2 were
performed approximately every 30 minutes as a grab sample at the outlet and at the inlet.
5.3.4 Average Moisture Determination by EPA Method 4
The average flue gas moisture content was determined according to EPA Method 4.
Before sampling, the initial weight of the impingers was recorded. When sampling was
completed, the final weights of the impingers were recorded, and the weight gain was calculated.
The weight gain and the volume of gas sampled were used to calculate the average moisture
content (%) of the flue gas. The calculations were performed by computer. Method 4 was
incorporated in the technique used for the Method 5A manual sampling method that was used
during the test.
5-36
-------
6.0 QUALITY ASSURANCE/QUALITY CONTROL
Specific Quality Assurance/Quality Control (QA/QC) procedures were strictly followed
during this test program to ensure the production of useful and valid data throughout the course
of the project. A detailed presentation of QC procedures for all sampling and analysis activities
can be found in the Site Specific Test Plan and Quality Assurance Project Plan for this program.
This section reports the results of all QC analyses so that the degree of data quality can be
ascertained.
In summary, a high degree of data quality was maintained throughout the project. All
sampling train leak checks met the QC criteria as specified in the methods. Isokinetic sampling
rates were maintained within 10% of 100% isokinetic for all test runs. Good spike recoveries
and close agreement between duplicate analyses were shown for the sample analyses.
6.1 Sampling QC Results
The following sections discuss the QC results of the specific sampling methods employed
during this program.
6.2 D/F Sampling QC
Table 6-1 lists the pre- and post-test and port change leak check results. The acceptance
criteria are that all post-test leak checks must be less than 0.02 cfm or 4% of the average
sampling rate, whichever is less. All D/F leak checks met this criterion.
Table 6-2 presents the isokinetic sampling rates for the D/F sampling runs. The
acceptance criterion is that the average sampling rate must be within 10% of 100% isokinetic.
All D/F sampling runs met this criterion.
6-1
-------
Table 6-1. Summary of Leak Checks Performed, Per Port, Method 23 Testing,
Thermal Oxidizer. Inlet and Outlet
Date
Run #/Port
Initial Leak
Check
Leak Check
Final Leak
Check
Inlet
9/22/97
9/23/97
I /A
l/B
2/A
2/B
3/A
3/B
0.004® 15"Hg
O.Ol @ 10" Hg
0.008 @ 15"Hg
Passed1'
Passed11
Passed*
Passed11
Passed11
Passed3
0.011 @ 12" Hg
Passed"
0.002® 15"Hg
Outlet
9/22/97
9/23/97
I /A
l/B
2/A
2/B
3/A
3/B
0.012 @ 15" Hg
0.017 @ 10" Hg
0.014® 10" Hg
0.007 @ 9" Hg
0.011 @5"Hg
0.007 @ 10" Hg
0.008 @ 10" Hg
0.009 @ 9" Hg
0.009 @ 9" Hg
0.011 @ 5"Hg
0.008 @ 10" Hg
0.009 @ 9" Hg
aActual value of leak check not recorded on field data sheet.
6-2
-------
Table 6-2. Summary of Isokinetic Percentages
Date
Run#
Percent Isokinetic
Method 23, Thermal Oxidizer Inlet
9/22/97
9/23/97
9/23/97
1
2
3
105
102
108
Method 23, Thermal Oxidizer Outlet
9/22/97
9/23/97
9/23/97
1
2
3
98
101
98
Method 5A, Thermal Oxidizer Inlet
9/22/97
9/23/97
9/23/97
1
2
3
98
103
104
Method 5A, Thermal Oxidizer Outlet
9/22/97
9/23/97
9/23/97
1
2.
3
104
103
102
Method 5A, APP Stack
9/24/97
9/24/97
9/24/97
1
2
3
100
99
99
Method 5A, SBS Stack
9/24/97
9/25/97
9/25/97
1
2
3
102
106
102
6-3
-------
All dry gas meters are fully calibrated every six months against an EPA-approved
intermediate standard. The full calibration factor is used to correct the actual metered sample
volume to the true sample volume. To verify the full calibration, a post-test calibration is
performed. The full and post-test calibration coefficients must be within 5% to meet ERG's
internal QA/QC acceptance criterion. As shown in Table 6-3, the meter boxes used for the D/F
and PM testing met this criterion.
Table 6-3. Dry Gas Meter Post-Test Calibration Results
Sampling Train
D/F, TO Inlet
D/F, TO Outlet
PM, TO Inlet
PM, TO Outlet
PM, APP Stack
PM, SBS Stack
Meter Box
Number
A-39
A-38
A-40
A-37
A-40
A-39
Full
Calibration
Factor
0.996
0.984
0.984
0.999
0.984
0.996
Post-Test
Calibration
Factor
0.974
0.981
0.980
0.988
0.980
0.974
Post Test
Deviation3
%
-2.2
-0.30
-0.41
-1.1
-0.41
-2.2
Tost-Test Deviation (%) = (Post Test Factor - Full Factor) / Full Factor x 100.
Field blanks are collected to verify the absence of any sample contamination. A D/F train
was assembled as if for sampling, leak checked at the sampling location, left at the sampling
location for the duration of a test run, and then recovered. Table 6-4 presents the analytical
results for the field blank and reagent blank as well as the laboratory method blank. No analytes
were observed above method detection limits in any of the blank samples.
6-4
-------
Table 6-4. Dioxin/Furan Field Blank Analysis Results
Congener
2,3,7,8-TCDD
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,6,7,8-HpCDD
OCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDF
Reagent
Blank
ng Detected
<0.006
<0.009
<0.009
<0.007
<0.008
<0.01
<0.02
<0.004
<0.006
<0.006
<0.006
<0.004
<0.005
<0.006
<0.008
<0.01
<0.01
Field Blank 1
ng Detected
<0.009
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.006
<0.009
<0.009
0.008
<0.006
<0.008
<0.008
<0.01
<0.01
<0.01
Field Blank 2
ng Detected
<0.005
<0.007
<0.009
<0.007
<0.007
<0.01
<0.01
<0.003
<0.005
<0.005
<0.005
<0.004
<0.005
<0.006
<0.007
<0009
<0.01
Cleanup
Blank
ng Detected
<0.002
<0.003
<0.004
<0.003
<0.003
<0.004
<0.007
<0.001
<0.002
<0.002
<0.002
<0.002
<0.006
<0.003
<0.003
<0004
<0.005
6.2.1 PM Sampling QC
Table 6-5 lists the pre- and post-test and port change leak check results for the thermal
oxidizer inlet and outlet, the APP Stack, and the SBS Stack. The acceptance criterion of less
than 0.02 cfm or 4% of the sampling rate, whichever is less, was met by all of the sampling
trains.
6-5
-------
Table 6-5. Summary of Leak Checks Performed, Per Port, Method 5A
Date
Run #/Port
Initial Leak
Check
Leak Check
Final Leak Check
Thermal Oxidizer Inlet
9/22/97
9/23/97
1/B
I/A
2/B
2/A
3/B
3/A
0.016 @ 15" Hg
0.01 @ 10" Hg
0.001 @ 10" Hg
Passed"
Passed"
Passed11
Passed1'
Passed"
Passed"
Passed'1
Passed"
0.002 @ 10" Hg
Thermal Oxidizer Outlet
9/22/97
9/23/97
1/B
I/A
2/B
2/A
3/B
3/A
0.007 @ 15" Hg
0.01 0@ 10" Hg
0.006 @ 9" Hg
0.010 @ 5" Hg
0.020 @ 5" Hg
0.010 @ 10" Hg
0.006 @ 10" Hg
0.007 @ 7" Hg
0.007 @ 7" Hg
0.020 @ 5" Hg
0.006 @ 10" Hg
0.007 @ 7" Hg
APP Stack
9/24/97
1/Y
1/X
2/X
2/Y
3/Y
3/X
0.008 @ 15"Hg
0.012® 15" Hg
0.011@15"Hg •
Passed"
Passed"
Passed"
0.014 @ 15" Hg
Passed"
Passed"
0.005 @ 15" Hg
0.014® 15" Hg
0.005 @ 15" Hg
SBS Stack
9/24/97
9/25/97
1/X
1/Y
2/Y
2/X
3/X
3/Y
0.010 @ 15" Hg
0.006 @ 15" Hg
0.008 @ 15"Hg
Passed"
Passed"
Passed"
Passed"
Passed"
Passed"
0.016 @ 15" Hg
0.006 @ 15" Hg
0.007 @ !5"Hg
"Numerical value not recorded.
6-6
-------
Table 6-2 presents the isokinetic sampling rates for the PM sampling runs. The sampling
rate acceptance criterion of being within 10% of 100% isokinetic was met for all sampling runs
at the thermal oxidizer inlet and outlet, at the APP Stack, and at the SBS Stack.
As shown in Table 6-3, the calibration coefficients of the meter boxes used for the PM
testing were within 5% of their full calibration coefficient, thus meeting the acceptance criterion.
6.3 Analytical QC Results
The following section reports QC parameters for the D/F and PM analytical results.
6.3.1 D/F Analytical Quality Control
One sample was generated for D/F analyses for each flue gas sample collected; the
sample was subjected to a full screen analysis conducted using a DB-5® capillary gas
chromatographic column which allows the separation of each class of chlorinated dioxins and
furans (i.e., tetra-. penta-, etc.) and fully resolves 2,3,7,8-TCDD from the other TCDD isomers.
Since no D/F were observed at a level five times the method detection limit, confirmation
analysis to resolve the 2,3,7,8-TCDF from the other TCDF isomers (performed, if required, on a
DB-225to capillary gas chromatographic column) was not required.
A component of the D/F QC program consists of the addition of isotopically labeled
standards to each sample during various stages of analysis to determine recovery efficiencies and
to aid in the quantitation of sampled D/F species. Four different types of standards are added:
Surrogate standards are usually spiked onto the XAD-2® sorbent prior to shipment
to the test site. Recovery of these compounds allows for the evaluation of overall
sample collection efficiency and analytical matrix effects;
• Internal standards are spiked after sampling but prior to extraction.
Alternate standards are also spiked after sampling but prior to extraction.
6-7
-------
Recovery percentages of internal standards are used in quantifying the D/F
sampled from the stack gas. Recovery of alternate standards is used for
determination of extraction/fractionation efficiencies.
Recovery standards are added after fractionation, just prior to analysis by
HRGC/HRMS.
The recovery of each of the spiked isotopically-labeled compounds was within the
acceptance criteria of Method 23 (see Appendix C for analytical data).
6.3.2 PM Quality Control
All filters and TCE probe rinse residues were weighed to a constant weight following the
procedures given in EPA Method 5A. The TCE probe rinse residues were blank-corrected using
a known volume of TCE reagent. The five-place analytical balance calibration was verified prior
to use by weighing a series of Class S weights which covered the range of weights encountered
with the samples.
6-8
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1. REPORT NO.
EPA- 454/R-99-026JA
3. RECIPIENTS ACCESSION NO.
4. TITLE AND SUBTITLE
Asphalt Roofing Emissions Teat Report Modified Bitumen Process
Final Report
U.SJntec, Port Arthur Te
5. REPORT DATE
August 1999
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
EMAD
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-D-7-0001
12. SPONSORING AGENCY NAME AND ADDRESS
Director
Office of Air Quality Planning and Standards
Office of Air and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT The asphalt roofing industry is among the categories of major sources for which national
emission standards for hazardous air pollutants (NESHAPS) are to be issued by November 2000
pursuant to Section 112 of the Clean Air Act. The asphalt roofing industry category includes
facilities that have the following specific processes: asphalt blowing stills: modified bitumen
production; a saturator/couter process: and fiberglass mat production.
Source tests are required to quantify and characterize the paniculate matter (PM) and
hazardous air pollutant (HAP) emissions, and the performance of a thermal oxidizer used to
control emissions associated with a modified bitumen facility.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COS ATI Field/Grow
Hazardous Air Pollutants (NESHAP)
Method 5A, Method 23, PM
16. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (Report)
Unclassified
21. NO. OF PAGES
88
20. SECURITY CLASS (Page)
Unclassified
22. PRICE
------- |