-------
REPORT ON
CAN MANUFACTURERS INSTITUTE
CAPTORS EFFICIENCY TEST METHOD STUDY
-------
REPORT ON
CAN MANUFACTURERS INSTITUTE
CAPTURE EFFICIENCY TEST METHOD STUDY
SUBMITTED TO:
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF AIR QUALITY PLANNING AND STANDARDS
RESEARCH TRIANGLE PARK, NC
SUBMITTED BY:
CAN MANUFACTURERS INSTITUTE
1625 MASSACHUSETTS AVENUE N.W.
WASHINGTON, DC
OCTOBER 22,1993
PREPARED BY:
ENVIRONMENTAL RESOURCES MANAGEMENT-NORTH CENTRAL, INC.
1630 HERITAGE LANDING DRIVE, SUITE 100
ST. CHARLES, MISSOURI 63303
PROJECT NO. 92184
-------
EXECUTIVE SUMMARY
FOR CAN MANUFACTURERS INSTITUTE
CAPTURE EFFICIENCY TEST METHOD STUDY REPORT
This report presents data and conclusions from the Can Manufacturers Institute (CMI)
comparative study on capture efficiency test methods. The purpose of this study was to
validate an alternative test method for the measurement of capture efficiency of a
discontinuous metal coating operation (such as the sheet fed metal decorating lines in a
lithography department of a 3-piece can plant). Validation of the proposed test method
is based on a comparison to EPA approved methods that involve the use of a temporary
total enclosure (TTE). This comparison primarily assessed the precision and accuracy of
test results, but also included other evaluation factors such as cost, feasibility, and safety
considerations.
A comprehensive Test Plan was developed for this study with input from EPA's Office
of Air Quality Planning and Standards (OAQPS). The Test Plan included many
enhancements to the liquid/gas mass balance procedure traditionally used to measure
capture efficiency for surface coating lines. For purposes of clarity, this enhanced
liquid/gas mass balance procedure will be referred to as the CMI Mass Balance Test. The
Test Plan was ultimately approved by OAQPS after several rounds of review and
modification.
The cornerstone of this comparative study was the simultaneous measurement of capture
efficiency on a 3-piece sheet fed coating line using three methods: (1) gas/gas TTE test,
(2) liquid/gas TTE test, and (3) CMI mass balance test. This simultaneous testing was
conducted after a TTE was constructed on the coating line. A total of nine 3-hour
consecutive test runs were completed over a five day period while maintaining nearly
identical line operating conditions. This included using a single coating and sheet size
for the entire test duration, as well as very tight control on the film weight application
and other operating parameters. The total VOC input to the line (L), the total gas phase
captured VOC emissions (G), and total gas phase fugitive emissions from the TTE (F)
were measured simultaneously during each test run. These parameters allowed
calculation of capture efficiency in accordance with each of the above listed methods.
Results from the nine consecutive runs were statistically analyzed to assess the precision
of each test method. In addition, the bias (or accuracy) of the proposed CMI mass balance
test was also evaluated using the criteria specified in EPA Method 301. Results from the
statistical analysis of test method precision are summarized as follows:
-------
Statistical Parameter
Range of Capture
Efficiency
Average Capture
Efficiency
Coefficient of Variation
95% Confidence Interval
Error Band*
Results for Each Test Method Evaluated
Gas/Gas TTE
93.6% to 94.8%
94.3%
0.4%
94.0% to 94.5%
0.3%
Liquid/Gas TTE
93.8% to 94.9%
94.2%
0.4%
93.9% to 94.4%
0.3%
CMI Mass Balance
87.6% to 105.1%
95.7%
5.6%
91.6% to 99.9%
4.3%
* Error band calculated as the 95% confidence limit deviation from the mean,
expressed as a percentage of the mean.
The evaluation of bias between the CMI mass balance and TTE test involved a calculation
of a t-statistic in accordance with the procedures outlined in EPA Method 301. The t-
statistic is then compared with the critical value (t = 1.397) for a data set that includes
nine replicate runs to determine if the bias is statistically significant. The t-statistics
calculated for the CMI test data were less than 0.85, indicating no significant bias.
The general conclusions derived from evaluation of the test study data are as follows:
1. The TTE test methods showed very good precision with a coefficient of
variation of approximately 0.4% and a calculated error band of
approximately 0.3%. This extreme precision is partially due to the fact that
the coating line had a high capture efficiency (greater than 94%) which
resulted in relatively low values for the gas phase fugitive emissions (F).
Variability in the TTE test methods is solely due to the variability in the
measurement of F due to the nature of the formulas used to calculate
capture efficiency.
2. The excellent precision achieved by the TTE methods indicates suggest that
the coating line operation was very steady and no significant variation in the
results was caused by process conditions. Therefore, the study provided an
excellent assessment of the test method variability without an interfering
contribution from process variability.
-------
3. The CMI mass balance test method results also demonstrated good
precision. The coefficient of variation and error band calculated for the nine
replicate test runs were less man values generated by a laboratory study of
mass balance closures conducted by the Radian Corporation and reported
in the February 1985 EPA Report No. EPA/600/2-85/011. The CMI test
method precision is also better than the probable error calculated for a
standard VOC emission test (EPA Method 3GB) which has a precision
statement of +/- 7.4%.
4. The error band calculated for the CMI mass balance test meets the +/- 5%
goals set by OAQPS staff. The error band calculated for all nine test runs
was +/- 4.3%. This assessment is consistent with the Method 301
procedures that specify the use of all comparative test study data to validate
an alternative method.
5. The error band calculated for various data sets of multiple replicate run
means were also well within the +/- 5% criteria. Six combinations of
replicate run meas ranging from two 2-hour to four 3-hour runs generated
error bands that ranged from 2.2% to 4.3%. CMI believes that analysis of
the variability of replicate run means is the best indication of the test
method precision, as the means are the data that would be reported for a
compliance test.
6. Accuracy of the CMI test method, based on comparison with the TTE test
methods, was very good. The average capture efficiency for the CMI data
set was within 2% of the average for the TTE data sets. In addition, the
Method 301 analysis also demonstrated that the CMI test did not introduce
any statistically significant bias.
Economic, safety, and feasibility considerations were also evaluated as part of the CMI
test study. The costs expended during the comparative study and cost quotes from a
source testing contractor provided a reliable means of estimating costs for three run, 3-
hour gas/gas TTE and CMI mass balance tests. The cost estimates were based on the
assumption that a facility would be conducting an overall control efficiency test where
capture efficiency would be added to a standard destruction efficiency test. For this case,
the gas/gas TTE test would cost a facility an additional $83,500 to $91,500, depending on
whether cooling zone monitoring as an additional fugitive emission point was necessary.
The corresponding incremental cost of conducting the CMI mass balance test was
estimated to be approximately $10,200. Based on these estimates, a TTE test could cost
up to nine times more than the CMI mass balance test.
-------
Several safety issues were considered in the evaluation of the TTE and CMI mass balance
tests. The CMI study included measurement of breathing zone VOC concentrations before
and after installation of the enclosure. On average, the breathing zone concentrations for
the TTE runs were approximately twice as high as the baseline run that was conducted
without an enclosure. In addition to the increased chemical exposure, the TTE test limits
mobility around the sheet fed coating line which could increase the chance of a physical
injury to the operator. Also, the enclosure will likely increase the probability of a coater
fire by confining a flammable vapor source. Even though the TTE would be ventilated,
dead zones within the enclosure are inevitable. The TTE could intensify the hazard of a
flash fire by limiting emergency egress and concentrating the fire at the vapor source.
Feasibility issues are primarily related to the available space to construct an enclosure in
a plant where clustered coating lines are installed. Many plants have coating lines that
are clustered side-by-side with very little open area between the ancillary equipment
around the belt conveyors, roll coater, and wicket drive. The lack of available space could
prevent construction of a functional TTE around this type of line, or at a minimum,
require that the adjacent lines be shut down during the construction and testing.
When all factors are considered, CMI believes that its proposed mass balance test can
measure coating line capture efficiency with sufficient precision and accuracy, without
imposing extremely high costs or causing undesirable safety conditions. A standard
protocol for the proposed test method is included as Appendix A of this report. CMI is
requesting approval of this protocol as an alternative test to the TTE methods that are
now required.
-------
TABLE OF CONTENTS
List of Figures
List of Tables
1.0 INTRODUCTION 1-1
2.0 BACKGROUND 2-1
2.1 Can Manufacturers Institute 2-1
2.2 Can Coating Process 2-1
2.3 Capture Efficiency Test Methods 2-3
2.4 CMI Position on Capture Efficiency Testing 2-3
3.0 CMI CAPTURE EFFICIENCY TEST METHOD STUDY 3-1
3.1 Purpose of Study 3-1
3.2 Test Procedures 3-2
4.0 TEST METHOD STUDY RESULTS 4-1
4.1 Baseline Testing 4-1
4.2 Transition to TTE Test 4-3
4.2.1 TTE Verification Test 4-3
4.2.2 Balancing Test 4-3
4.3 Capture Efficiency by TTE Methods (Runs 1 through 9) 4-4
4.4 Capture Efficiency by CMI Mass Balance Method
(Runs 1 through 9) 4-5
4.5 Comparison of TTE and CMI Mass Balance Test Results 4-8
4.6 Analysis of Short Duration Test Results 4-10
4.7 Error Band Analysis for Multiple Run Configurations 4-12
4.8 Parametric Study (Runs 10 through 12) 4-15
4.9 TTE Impact on Coating Line Capture 4-16
Environmental Resources Management - North Central, Inc.
-------
TABLE OF CONTENTS
5.0 ECONOMIC, SAFETY, AND FEASIBILITY CONSIDERATIONS 5-1
5.1 Economic Considerations 5-1
5.1.1 Cost of TTE Design and Installation 5-1
5.1.2 Cost of Standard Gas/Gas TTE Test 5-2
5.1.3 Production Losses During a TTE Test 5-3
5.1.4 Cost of CMI Mass Balance Capture Efficiency Test 5-4
5.1.5 Capture Efficiency Cost Comparison 5-5
5.2 Safety Considerations 5-6
5.3 Feasibility Considerations 5-7
6.0 CMI PROPOSAL FOR CAPTURE EFFICIENCY TEST METHOD 6-1
Appendices
A CMI Test Protocol
B Test Run Summary Sheets
C Air Velocity Data Sheets
D Extraction of Short Duration Test Results
E Mostardi-Platt Associates' Quotes on CE/DE Testing
Environmental Resources Management - North Central, Inc.
-------
LIST OF TABLES
Number Description
1 Summary of Capture Efficiency Results
2 TTE Average Face Velocity
3 Comparison of Test Method Results
4 Analysis of Response Factor Variance
5 Summary of Capture Efficiency Results
2-Hour Test Periods Extracted from 3-Hour Runs
6 Comparison of Test Method Results
2-Hour Test Periods Extracted from 3-Hour Runs
7 Summary of Capture Efficiency Results
1-Hour Test Periods Extracted from 3-Hour Runs
8 Comparison of Test Method Results
1-Hour Test Periods Extracted from 3-Hour Runs'
9 Error Band Analysis - Two 3-Hour Runs
10 Error Band Analysis - Three 3-Hour Runs
11 Error Band Analysis - Four 3-Hour Runs
12 Error Band Analysis - Two 2-Hour Runs
13 Error Band Analysis - Three 2-Hour Runs
14 Error Band Analysis - Four 2-Hour Runs
15 Summary of Capture Efficiency Results
Parametric Study
16 Comparison of Capture Hood Performance
17 Comparison of Capture Hood Performance - Normalized Data
18 Distribution of VOC Emissions Captured by Hood System
Environmental Resources Management - North Central, Inc.
-------
LIST OF FIGURES
Number Description
1 Multiple Views of Typical Sheet Fed Coating Line
2 3D Schematic of Typical Sheet Fed Coating Line
3 Multiple Views of TTE on Sheet Fed Coating Line
4 Major Components of TTE on Sheet Fed Coating Line
5 Overview of CE Test Study Measurement Locations
6 Measurement of Liquid VOC Input
7 Measurement of Captured VOC Emissions
8 Cooling Zone Test
9 Method 30 TTE Compliance Summary - Runs 1 through 9
10 Method 30 TTE Compliance Summary - Parametric Study
Runs 10 through 12
11 Measurement Locations for Assessment of Capture Hoods - Run 13
12 Summary of Results Air Velocity Survey
Roll Coater Upper Hood - Front Face
13 Summary of Results Air Velocity Survey
Roll Coater Upper Hood - Rear Face
14 Summary of Results Air Velocity Survey
Roll Coater Upper Hood - Bottom Slot
15 Summary of Results Air Velocity Survey
Roll Coater- Exit Hood
16 Summary of Results Air Velocity Survey
Oven Entrance - Face Opening
17 TTE Workplace Monitoring Locations
18 Typical Ancillary Equipment - Sheet Fed Coating Line
19 Functional TTE Installed on Clustered Sheet Fed Coating Lines
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 1-1
1.0 INTRODUCTION
The U.S. Environmental Protection Agency (USEPA) final guidance on the measurements
of volatile organic compound (VOC) capture efficiency for existing sources requires the
use of temporary total enclosures (TTEs). As opposed to traditional test methods, the
TIE testing requirements will be significantly more costly and may cause unnecessary
worker safety issues for minimal improvements in measurement precision. Moreover,
due to the unique production line layout and operator interaction in a can coating
process, it may not be feasible for many can manufacturers to construct functional TTEs
on their existing coating lines. This situation raises two threshold policy issues:
1. whether mandating the use of a TTE test method as the sole technique to
measure capture efficiency is appropriate for can makers and metal
decorators, and
2. whether an alternative test method can be approved which serves USEPA's
goal of precise capture efficiency measurements without the
disproportionate cost and technical feasibility concerns with the TIE test
method.
This report presents the results of an extensive test study undertaken by the Can
Manufacturers Institute (CMI) to compare the precision of three test methods for
measuring the VOC capture efficiency of a surface coating line. This report also puts
forth CMI's proposed mass balance capture efficiency test protocol. This test protocol
was developed by extracting the pertinent sections from the Test Plan prepared for this
study and then adding several improvements developed during the testing. A copy of
the complete protocol is attached as Appendix A. Discussion of the improvements
developed during the study are presented in Section 6.0.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 2-1
2.0 BACKGROUND
2.1 Can Manufacturers Institute
CMI is the national not-for-profit trade association of metal and composite can producers
and their suppliers. It represents over 45 companies with over 250 can plants located
throughout the continental United States. These companies produce approximately 90
percent of the 125 billion metal food and beverage cans shipped annually. Its members
range from coating suppliers such as Glidden, Grace, and PPG to some of the largest can
manufacturing companies in the world, such as American National Can Company, Ball
Corporation, Crown Cork & Seal Company, Metal Container Corporation, and Reynolds
Metals. CMI has a long history of constructive partnerships with USEPA ranging from
the cooperative development of the can coaters "bubble" policy which authorized plant-
wide daily RACT compliance without SIP revisions, (see 45 Federal Register 80824) to
the development of beverage can effluent limitation guidelines (see 48 Federal Register
52380). CMI remains committed to cost effective and timely compliance with Federal
and State regulations, and looks forward to continuing its positive relationship with
governmental agencies and the general public.
2.2 Can Coating Process
There are essentially two types of metal cans: two-piece (typical 12-oz. beverage can)
and three-piece (normally a food packaging can which varies in size and shape). Both
types of can are normally coated on the inside with a solvent or water-based FDA-
approved coating in order to protect the can contents from interacting with the steel or
aluminum. The coatings used in the three-piece food cans are normally solvent-based
and contain volatile organic compounds (VOCs) which assist in the delivery of the actual
coating material. Over the past ten years, significant strides in the reduction of solvent
material in can and coil coatings have been made. For example, the introduction of
water-based beverage can inside spray coatings reduce the amount of solvent usage by
nearly 70 percent.
Because some can coatings (primarily those used in the three-piece process) still contain
VOCs above the appropriate RACT standards, can manufacturers are required to capture
and remove or oxidize their VOC emissions. In most cases, facilities use a thermal
oxidizer to safely and efficiently burn VOC emissions. Each year, can and coil coating
companies are required by USEPA to determine the amount of VOC being captured and
subsequently destroyed. This determination is expressed as a percentage and is
commonly referred to as "overall control efficiency". Overall control efficiency is the
product of capture efficiency times destruction efficiency. The issue at hand is related
to the measurement of capture efficiency.
Environmental Resources Management-North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 2-2
The three-piece can manufacturing process begins with the decorating of metal sheet
stock which is then formed into can bodies or can ends in subsequent operations. This
sheet fed coating process is the subject of CMI's capture efficiency test method
measurement study. Most of the other can manufacturing coating operations (such as
two-piece inside spray, two-piece overcoat varnish, end sealing compound application,
etc.) comply with the VOC RACT rules through the use of high solids or water-based
compliant coatings.
The typical sheet fed coating line consists of the following operations which are
interconnected by a conveyor system: sheet feeder/depalletizer, roller coating applicator,
wicket conveyor oven, oven cooling section, and sheet stacker. Three-dimensional
schematic drawings of a typical three-piece coating line that identifies key equipment are
included as Figures 1 and 2. These drawings were derived from the coating line actually
tested during the CMI study.
Multiple layers of coatings are applied to each side of the sheet with one layer deposited
during each production run. The coating process consists of the following sequence of
steps:
• palletized sheets are delivered to the sheet feeder by forklift;
• sheets are depalletized and transferred to the belt conveyor;
• coating is applied by a roller coating head;
• coated sheets are conveyed to the oven entry and transferred to a wicket
sheet conveyor;
• coated sheets are cured in a direct-fired wicket oven;
• cured sheets are passed through a cooling zone; and
• finished sheets are palletized by a sheet stacker.
Typically, the curing oven exhaust is ducted to a thermal oxidizer which controls VOC
emissions driven off by the curing process. Many coating lines are also equipped with
capture hoods over or around the roller applicator to assist in the collection of fugitive
losses outside of the curing oven. These capture hood ducts are typically routed to the
thermal oxidizer or back into the curing oven as make-up air. The wicket curing oven
is also operated at a slight negative pressure and is typically designed to draw make-up
air through its entrance at an approximate face velocity of 150 feet per minute (fpm).
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 2-3
2.3 Capture Efficiency Test Methods
There are two general approaches that can be used to measure capture efficiency. These
are commonly referred to as the liquid/gas and gas/gas measurement techniques, and
both are based on the principal that the total amount of VOC delivered to the process
(L) is equal to the amount captured by the control device (G) plus the amount of fugitive
VOC loss (F). Capture efficiency is then calculated as a percentage by dividing the
captured VOCs (defined as G or as (L - F)) by the total VOC input (defined as L or as
(G + F)). All of the above mentioned parameters (L, G, and F) must be expressed in the
same measurement unite. Typically, the total VOC input (L) is converted to a propane
equivalent to match the measurement units from the gaseous phase emission testing (G
and F) which are determined by a flame ionization analyzer (FIA) calibrated against a
propane standard.
The most commonly used capture efficiency test method is known as the liquid/gas
mass balance method. It is conducted by measuring the gas-phase VOC emissions
captured by the control device (G) and the total VOC input to the system (L). This
method does not require the construction and operation of an enclosure around the
coating line, as the gas phase fugitive loss is not measured.
The USEPA mandated methods for measuring capture efficiency require the construction
of a temporary total enclosure (TTE) around the coater area so that fugitive VOC losses
are always measured directly. There are two TTE test options: (1) gas/gas TTE method
where the gas phase captured emissions are measured and capture efficiency is
calculated as G/(G + F), and (2) liquid/gas TTE method where the total VOC input is
measured and capture efficiency is calculated as (L - F)/L.
The CMI test study provided simultaneous measurements of L, G, and F to enable
calculation of capture efficiency by each of the three above listed methods. This
included measurement of capture efficiency for nine 3-hour test runs and an additional
three one-hour test runs for a total of twelve capture efficiency tests. Three-dimensional
schematic drawings of the subject coating line with the TTE installed are included as
Figures 3 and 4. Figures 5 through 7 show the test stations that were used to
simultaneously measure L, G, and F.
2.4 CMI Position on Capture Efficiency Testing
CMI suggests to the USEPA that carefully designed improvements to the traditional
liquid/gas mass balance method will generate capture efficiency results that are
"reasonably consistent" with those measured by the TTE methods. CMI requests that
USEPA reanalyze the exclusive use of TTE capture efficiency test methods in light of the
following issues:
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 2-4
• the cost and complexity of a TTE test is excessive when it is properly
designed using expert consulting services, whereas the CMI mass balance
test is much less of an economic burden (almost a ten-fold difference);
• the TTE limits the access to and egress from the coater area causing a
safety concern in light of the potential for a flash fire;
• enclosing the coater area, even with substantial ventilation, was shown to
increase the workers' exposure to solvent vapors;
• it may not be feasible to construct a functional TTE on many existing lines,
given the close spacing between lines and the need for a significant
clearance within the TTE to allow a worker to operate the line properly;
• the construction/dismantling of the TTE and the actual testing will disrupt
normal production operations.
Given that the use of CMI's improved mass balance test protocol can produce results
that are consistent and compare reasonably well to TTE results, CMI believes that it is
an acceptable alternative to the TTE test methods.
Environmental Resources Management- North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 3-1
3.0 CMI CAPTURE EFFICIENCY TEST METHOD STUDY
3.1 Purpose of Study
The primary goal of the CMI study was to evaluate the repeatability of the following
capture efficiency test methods:
• improved liquid/gas mass balance (CMI mass balance);
• gas/gas using a TTE; and
• liquid/gas using a TTE.
The evaluation of test method repeatability was made to determine if the CMI mass
balance method is a reliable alternative to the TTE methods. The percent deviation
between the liquid/gas TTE method and gas/gas TTE method was also made to
determine if VOC oxidation in the wicket curing oven introduces a significant bias. The
liquid/gas TTE method is the only test which is not effected by this condition.
Other measurements and qualitative tests were made to determine the Tit's impact on
the coating line's VOC capture characteristics. This includes an evaluation of the
percentage of total captured emissions removed by the capture hood system for the
baseline run versus the TTE runs. Also, air velocity surveys and smoke tests were
conducted at and around the oven entrance and capture hood intakes to determine the
TTE's effect on air currents around these capture zones.
In addition to the capture efficiency measurements, the study also addressed the
economic impact of conducting a test using a TTE. Appropriate costs were extracted
from the study to determine the costs for each of the following elements of the standard
gas/gas TTE test:
• engineering design, fabrication, and installation of the TTE;
• production losses related to installation/dismantling of the TUB;
• cost of conducting three 3-hour gas/gas TTE test runs; and
• production losses related to conducting the TTE test.
The cost of running the CMI mass balance test was also estimated by this study. The
cost estimates for both methods were compared to assess the economic impact of
imposing TTE tests on the can manufacturing industry.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 3-2
3.2 Test Procedures
The CMI capture efficiency test method study was conducted in accordance with the
Test Plan submitted to the USEPA Emissions Measurement Branch (EMB) on February
17, 1993 and modifications submitted in the March 15, 1993 correspondence to EMB
responding to their review comments. Generally, the TTE test methods were performed
in accordance with the following test methods listed in USEPA's Method 30 series for
capture efficiency testing (previously designated as the Method 204 series):
• Method 30 - TTE Verification Tests;
• Method SOB - VOC Compound Emissions in Captured Stream (also used
for fugitive stream as all fugitive emissions were directed into a common
TTE exhaust duct; and
• Method 30F - VOC Content in Liquid Input Stream (distillation method).
The test method study included a sequence of tests conducted prior to and after
installation of the TTE on the coating line. Initially, a baseline test involving a single 3-
hour run using the CMI mass balance test method was conducted to define the captured
emissions prior to installation of the TTE. Data from this test run was also used to
determine if the cooling zone needed to be monitored as a fugitive emission point and
to define workplace air quality prior to installation of the TTE. Next, a TTE verification
test was conducted immediately after installation was complete. This test verified that
the TTE met each of the five criteria listed in Method 30. Once the enclosure was
certified as a TTE, a balancing test was conducted to determine if the ventilation was set
so that OSHA PELs were met and that the average VOC concentration measured during
the balancing test did not deviate more than 10% of that measured during the baseline
test. This concluded the initial set of tests needed to ensure that the TTE was
functioning in accordance with Method 30 and other EPA guidance.
The cornerstone of the test method study was the simultaneous measurement of 1) total
liquid VOC input (L); 2) captured VOC emissions (G); and 3) fugitive emissions (F) for
nine 3-hour runs. This allowed calculation of the coating lines capture efficiency by
three different methods and generated a sufficient number of data points to allow a valid
statistical analysis. These test runs also included smoke tests and velocity readings
around the capture hoods and oven entrance to determine the impact of the Tils on air
flow within these capture zones. The workplace VOC concentration within the enclosure
was also measured for comparison with that measured during the baseline test. Lastly,
three 1-hour test runs were conducted at the end of the study with several parameters
changed to determine if there was any significant impact on the capture efficiency. This
parametric study included the installation of a baffle within the enclosure, increasing the
ventilation rate, and the use of a higher VOC coating.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-1
4.0 TEST METHOD STUDY RESULTS
The capture efficiency test results for each of the three methods covered by this study
are summarized on Table 1. This table also presents the weight of VOC (as a propane
equivalent) for the liquid-phase total VOC input into the system (L), the gas-phase
captured VOC (G), and the gas-phase fugitive VOC (F). Summary sheets for each test
run which provide the supporting data used to generate the results for L, G, and F are
included in Appendix B. The air velocity survey data for each run is included as
Appendix C. All of the field and laboratory data used to derive the results presented
on these summary sheets are included in the attached Mostardi-Platt Associates' Data
Summary Report.
4.1 Baseline Testing
A baseline test which included the CMI mass balance procedure as well as monitoring
of the cooling zone discharge was conducted prior to installation of the TTE. The
capture efficiency measured by the CMI mass balance for this baseline run was 90.7%,
with captured emissions at 120.12 Ibs/propane and liquid VOC consumption at 132.4
Ibs/propane. Data from the baseline run was also tabulated on a 1-hour time interval
to allow comparison with data collected during the balancing test on shorter time
intervals. One hour averages for VOC concentrations in the captured emission streams
and the coating consumption are tabulated below. A flow weighted average
concentration for the entire captured gas stream is also calculated. The time interval
from 60 to 120 minutes had a coating consumption rate which most closely matched the
rate measured during the balancing test, and therefore, its flow weighted average VOC
concentration of 796 ppm was used for the comparison.
Time
Interval
(min)
0-60
60 - 120
120 - 180
Average VOC Concentration
(Corrected for Background VOC)
Oven Exhaust
(ppm)
763
854
813
Capture Hoods
(ppm)
570
623
653
Flow Weighted
Average (ppm)
714
796
773
Coating
Consumption
(Ibs/hr)
87.3
96.6
96.1
The baseline testing also included monitoring of the cooling zone exhaust to determine
if it was pulling any VOC from the curing oven and needed to be included as a fugitive
emission point. A schematic diagram of the cooling zone, which depicts air flow
through the system, is shown on Figure 8. The cooling zone intake dampers were
configured so that air was drawn from inside the plant and exhausted to the outside.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-2
Therefore, if no additional VOC was entrained in the cooling zone exhaust, the
background concentration measured inside the plant should have compared reasonably
well with the cooling zone exhaust concentration. A background concentration
measuring point near the cooling zone intake was monitored during the test to make this
comparison. The average VOC concentrations for the 3-hour test period were as follows:
• Cooling Zone Exhaust (TS-4) = 26.7 ppm propane
• Background Near Cooling Zone Intake = 24.6 ppm propane
Since the background concentration was well within the 20% tolerance specified in the
Test Plan, monitoring of the cooling zone exhaust was not required. Additional readings
were also taken during test Runs 8 through 12 to quantify all sources of oven make-up
air. All of the rnake-up air intakes for each oven zone, as well as the combustion air
intake, were located on top of the curing oven. Face velocity readings were taken at
each of these intakes and tike oven entrance to quantify all make-up air except for that
pulled through the oven discharge via the cooling zone. This type of curing oven is
designed to maintain continuous movement of make-up air from the oven discharge
towards the Zone 1 exhaust to prevent a build-up of explosive vapor concentrations.
The oven discharge make-up air flow was estimated by subtracting the sum of each of
the measured make-up air points from the total oven exhaust. A summary of this
analysis is as follows:
Run
8
9
10
11
12
Oven Make-Up Air Flow (scfm)
Entrance
Opening
249
365
251
340
241
Combustion
Ait Intake
3,550
3,260
2,960
3,290
2,750
Zones 1-4
Intakes
1,115
929
980
920
863
Oven
Exhaust
Flow
(scfm)
5,784
5,639
6,192
5,968
6,159
Make-Up Air
Pulled from
Cooling Zone
(scfm)
870
1,085
2,001
1,418
2,305
The make-up air pulled from the cooling zone ranged from 870 scfm to 2,305 scfm. This
condition clearly demonstrates that air flow in the curing oven is continually moved
from the back of the oven to the Zone 1 exhaust, and therefore, there should be no
leakage of VOC from Zone 4 into the cooling zone exhaust.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-3
4.2 Transition to TTE Test
4.2.1 TTE Verification Test
Immediately after construction of the TTE, numerous measurements were made to verify
that the minimum spacing between NDOs and VOC emission sources was achieved. It
was determined that the wicket return channel below the curing oven did not act as
another draft opening, therefore, several additional NDOs were placed along the back
wall of the TTE to provide a sufficient amount of open area. Figure 9 presents the
locations of each NDO, its equivalent diameter, and the measurement to the nearest
VOC emitting point. The separation for all NDOs was greater than 4 times the
equivalent diameter with the closest being NDO N-9, which was 4.27 equivalent
diameters from the nearest VOC emitting point.
The separation between the ventilation intakes and NDOs was also measured to verify
that the minimum separation was achieved. This analysis is also presented on Figure
9 and demonstrates that all of the ventilation intakes are separated from NDOs by more
than 4 equivalent diameters. Ventilation intakes V-6 (the coater exit hood) was the
closest to this requirement being separated from an NDO by 6.53 equivalent diameters.
The cumulative open area from all NDOs was also compared to the total surface area
of the TTE to determine if it met the 5% requirement specified in Method 30. The
cumulative open area of NDOs was 34.18 square feet and the total surface area of the
enclosure was 1,576 square feet, which yielded an NDO to total area ratio of 2.17%,
which meets the Method 30 requirement.
The last of the TTE verification tests was the assessment of average NDO face velocity.
The total ventilation rate from the TTE was calculated by summing the temporary
exhaust flow rate, capture hood exhaust flow rate, and flow into the oven entrance. This
total ventilation rate (expressed as standard cubic feet per minute) was divided by the
cumulative NDO open area to calculate the average NDO face velocity. This analysis
is summarized on Table 2 and clearly demonstrates conformance with the minimum
average face velocity of 200 fpm specified in Method 30. Plastic streamers were also
placed within each NDO opening to verify that the flow direction was always into the
TTE. Visual observations made during the testing confirmed this condition. Based on
the foregoing analyses, the TTE conformed to each of the criteria listed in Method 30.
4.2.2 Balancing Test
A balancing test was conducted after the coating line began operation with the TTE. The
purpose of this test was to determine if the TTE ventilation rate was set at a level that
maintained a safe workplace environment and an average captured VOC concentration
within 10% of that measured during the baseline run. Initially, the TTE exhaust rate was
set at the design point of 6,900 scfm. Workplace VOC concentrations were measured at
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-4
six points within the enclosure to assess compliance with OSHA PELs. These workplace
concentrations ranged from 35 to 175 ppm, and were below the target concentration of
254 ppm. The target concentration represents the maximum EIA response that complies
with OSHA PELs.
The capture emission streams (the oven exhaust and capture hood ducts) were
monitored for VOC concentrations for approximately 3 hours. Coating consumption
during this balancing run was considerably higher than that measured during the
baseline test. Therefore, a one-hour time period (11:20 to 12:20) which most closely
matched coating consumption during the baseline run was selected for the comparative
analysis. Data from this time period was as follows:
• Coating Feed - 104.1 Ibs/hr
• Oven Exhaust VOC (corrected for background) - 990 ppm
• Oven Exhaust Flow Rate - 5,916 scfm
• Capture Hood VOC (corrected for background) - 508 ppm
• Capture Hood Flow Rate - 2,387 scfm
Since the total captured emissions were comprised of measurements made in two
separate ducts, a flow weighted average was calculated to represent the average VOC
concentration of the entire captured emission stream. The flow weighted average VOC
concentration for the above-listed time period was calculated at 852 ppm. This
concentration was compared to the second hour of operation during the baseline test
which had a flow weighted average VOC concentration of 796 ppm and a coating
consumption rate of 96.6 Ibs/hr. This average VOC concentration was 7.0% lower than
the selected one-hour time period in the balancing run, and was well within the +/-10%
criteria. The concentrations would have likely been much closer if the coating
consumption rates were equal. The coating consumption rate for the selected one-hour
time period from the balancing test was 7.8% higher than that measured in the selected
one-hour time period for the baseline run.
4.3 Capture Efficiency by TTE Methods (Runs 1 through 9)
The TTE test methods produced consistent results with little variability. The following
is a summary of the capture efficiency results for the nine 3-hour test runs determined
by the TTE methods.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-5
Parameter
Range of Capture Efficiency
Average Capture Efficiency
Coefficient of Variation
95% Confidence Interval
Gas/Gas TTE Test
93.6% to 94.8%
94.3%
0.4%
94.0% to 94.5%
Liquid/Gas TTE Test
93.8% to 94.9%
94.2%
0.4%
93.9% to 94.4%
The results obtained by both TIE test methods are very similar, with the observed
deviation within the measurement accuracy for the test procedures. The liquid/gas TTE
test does not include measurement of the curing oven exhaust and is not affected by
VOC oxidation in the direct-fired oven/ whereas the gas/gas TTE test is influenced by
this condition. Since the results between these methods are so similar, it is concluded
that VOC oxidation in the curing oven is an insignificant bias when measuring capture
efficiency for this type of coating operation. Therefore, this potential bias is also an
insignificant issue when measuring capture efficiency with the CMI mass balance test.
4.4 Capture Efficiency by CMI Mass Balance Method (Runs 1 through 9)
Capture efficiencies measured by the CMI mass balance test method were reasonably
consistent with good repeatability. Results for the nine 3-hour test runs are summarized
below:
Parameter
Range of Capture Efficiency
Average Capture Efficiency
Coefficient of Variation
95% Confidence Interval
CMI Mass Balance Test
87.6% to 105.1%
95.7%
5.6%
91.6% to 99.9%
The coefficient of variation calculated for this data set is 5.6%, which indicates good
repeatability. In fact, the USEPA'S Report No. EPA/600/2-85/011, Measurement of
Volatile Organic Compound Efficiency. February 1985, Page 50, states "Precision or
variability of estimates for the two groups are excellent since the coefficients of variation are 5.9%
and 8.9%." This quote was referring to the statistical evaluation of data on laboratory
study results for mass balance closures. The confidence interval limits deviate from the
mean by +/- 4.1%, which represents 4.3% of the 95.7% average capture efficiency. This
analysis indicates precision that is better than a standard VOC emission test (Method
SOB) which has a reported probable error of +/- 7.4%.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-6
The supporting data used to calculate the liquid-phase total VOC input was evaluated
to determine which measurement parameters caused the greatest deviation in the overall
results. The test run that most closely matched the TTE method results (Test Run 7) was
selected for this error margin analysis. Based on review of the liquid-phase summary
sheet for this test run, it is very clear that the field measurements and laboratory analysis
of the coating material are the predominant factors in determining the overall liquid-
phase VOC input. For this test run, the coating material accounted for 97% of the total
VOC input. Potential error in determining the VOC (as a propane equivalent)
contributed by the coating material could be introduced through inaccuracies in one of
the following measurements:
• Measurement of the instantaneous weight of the coating feed bucket. The
coating drum static weight measurement was not considered as it would
only be inaccurate if the scales were not properly calibrated or zeroed.
• Measurement of the VOC content of the coating material
• Measurement of the response factor for the coating material.
The largest amount of error would be introduced if the initial coating measurements
were biased one direction and the final coating measurements biased the other direction.
For purposes of this error analysis, a 5% deviation was introduced for each of the above-
listed measurements with the deviation made in the opposite direction for the initial and
final coating measurements. The results of this error analysis are summarized as
follows:
Error Margin Introduced
into Coating Measurement
Initial Feed Bucket Wt. +5%
Final Feed Bucket Wt. -5%
Initial VOC Content +5%
Final VOC Content -5%
Initial Response Factor -5%
Final Response Factor +5%
Actual
Total
Liquid
VOC Input
(Ibs C3H8)
130.01
130.01
130.01
Adjusted
Total
VOC
Input
(Ibs C3H8)
132.41
140.88
141.23
Percent
Change
to Total
VOC
Input
+1.8%
+8.4%
+8.6%
Percent
Change to
Capture
Efficiency
-1.8%
-7.8%
-8.0%
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-7
This analysis demonstrates that the most critical factors in determining L are the Method
30F measurements of percent VOC and response factor for the coating samples. The
supporting data for test runs where the CMI mass balance deviated the furthest from the
TTE method were evaluated to determine if either of these parameters contributed to the
deviation.
The CMI mass balance capture efficiency for Test Run 2 was 105.1%, which was
approximately 12% higher than the TTE results. Possible reasons for the deviation are
the response factor of the initial and final coating samples. The initial coating response
factor (1.500) is higher than the average for Runs 1 through 9 (1.449) and the final
coating response factor (1.406) is lower than the Runs 1 through 9 average (1.513). If the
average response factors are used in the Run 2 calculations, the resulting capture
efficiency would be 98.9%, which reduces the deviation to approximately 40% of its
original value.
Test Run 6 was also evaluated, as the CMI mass balance capture efficiency of 87.6% was
approximately 7% lower than the corresponding TTE test method results. In this case,
the initial coating response factor of 1.317 was significantly lower than the average initial
coating response factor (1.449) measured for the entire data set. If the average initial
coating response factor is used in the Run 6 calculations, the capture efficiency would
increase from 87.6% to 98.6%. This value is within 5% of the TTE test method results
for Run 6 and much closer to the average capture efficiency measured by the CMI mass
balance testing.
If the adjusted capture efficiency for Runs 2 and 6 (calculated above) are substituted into
the CMI mass balance data set, the coefficient of variation would decrease from 5.7% to
3.5%, and the 95% confidence interval would range from 93.7% to 98.8%. The overall
conclusions derived from the CMI mass balance test results are:
• the repeatability of the results was very good with a coefficient of variation
of less than 6% and the error band calculated from the 95% confidence
interval approximately +/- 4%;
• the VOC content and response factors for the coating samples represent the
most critical measurements in obtaining an accurate measurement of the
total liquid phase VOC input (L); and
• additional quality assurance steps for the measurement of response factors
should result in even better repeatability and accuracy of the test method.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-8
4.5 Comparison of TTE and CMI Mass Balance Test Results
A comparison of test results obtained by the TIE methods and the CMI mass balance
method is presented on Table 3. The precision of each test method was assessed by
calculating the coefficient of variation and 95% confidence interval. Both of the TTE
methods showed excellent precision (or repeatability) with coefficients of variation at
0.4% and 95% confidence limits deviating by less than +/- 0.3% from the mean. The
CMI mass balance results were not as precise as the TTE test methods, however, still
showed good repeatability with a coefficient of variation of 5.6% and the 95% confidence
limits deviating by +/- 4.1% from the mean. CMI believes that the repeatability of this
improved liquid/gas mass balance test is sufficient to justify its use as an alternative to
the TTE methods. We do not dispute the fact that the TTE methods are more precise,
but believe that the extreme level of precision provided by those methods is not
necessary. This conclusion is based on the fact that most facilities are employing thermal
oxidation with destruction efficiencies in excess of 95%, where a capture efficiency of less
than 85% would be required to comply with the RACT emissions limitations. Since
typical capture efficiencies are 90% or higher, there is a reasonable margin for variability
in its measurement. This is a salient issue when considering the costs and potential
technical feasibility issues of the TTE test method. Additional discussion on these factors
is provided in Section 5.0. CMI also believes that the additional quality assurance
measures included in its proposed test protocol for the measurement of the liquid-phase
response factor will further improve test repeatability.
Although not a primary goal of this study, CMI also evaluated the accuracy between the
TTE methods and the CMI mass balance method. The average capture efficiency for
Test Runs 1 through 9 measured by each method are also presented on Table 3. In
addition, percent deviations between the CMI mass balance and both TTE methods for
each run are calculated on Table 2. This assessment indicates that the CMI mass balance
accuracy is close to the TTE methods. On average, the percent deviations of the CMI
mass balance from the gas/gas TTE and liquid/gas TTE test results were +1.6% and
+1.7%, respectively. In addition, individual percent deviations for each test run were all
below 12%, and 6 of the 9 runs were below 5%. Again, CMI believes that this level of
accuracy is more than adequate to justify the use of this test method as an alternative
to the TTE test methods.
An additional assessment of the accuracy of CMI's proposed test method was completed
using procedures specified in EPA Method 301, Field Validation of Pollutant
Measurement Methods from Various Waste Media. Method 301 includes procedures to
calculate a bias value when the field validation procedure is based on the comparison
of a proposed test method to a validated test method. Although EPA's TTE capture
efficiency methods have not been officially validated, the CMI study generates all of the
information needed to complete the bias analysis. This analysis involves the calculation
of a t-statistic that is based on the mean and standard deviation of the difference
between paired sample runs measured during the comparative study. The calculated t-
Enviioiunental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-9
statistic is compared to a critical value to determine if the bias is statistically significant.
For nine data points, the critical value for the t-statistic is 1.397. The t-statistics
calculated when comparing CMI's method to the gas/gas and liquid/gas TTE methods
were 0.84 and 0.83, respectively. Since these calculated t-statistics are less than the
critical value, the bias introduced by the CMI mass balance test is not considered to be
significant. Method 301 also specifies procedures to calculate a correction factor used
to adjust for statistically significant bias. The calculated correction factors for the CMI
test are 0.984 and 0.983 when comparing to the gas/gas and liquid/gas TTE test,
respectively. Although this correction factor calculation is not required, it further
demonstrates the acceptable accuracy of the test method. Method 301 specifies that the
test method is unacceptable if the correction factor is outside the range of 0.9 to 1.10.
Correction factors calculated for the CMI method are well within this required range.
The final element of the comparative analysis was an evaluation of the variation in the
response factor measured in tike liquid-phase analysis. The results from this comparative
analysis are presented on Table 4. The liquid-phase total VOC was calculated excluding
the response factor which yielded total pounds of VOC instead of total VOC as a
propane equivalent. The liquid phase total VOC, excluding the response factor, was
then divided by the liquid phase VOC calculated as a propane equivalent to determine
the overall response factor for the CMI mass balance test method. This overall response
factor represents a weighted average of the six response factors measured for each test
run which were used in the calculation of the liquid phase total VOC as a propane
equivalent. A similar analysis was completed to calculate the overall response factor
that would be needed to match the capture efficiency numbers measured by the liquid/
gas TTE test. In this case, the gas phase captured VOC was divided by the liquid/gas
TTE capture efficiency to generate a liquid phase total VOC (in pounds of propane) that
would generate the same capture efficiency if applied in the mass balance equation.
Next, the liquid phase total VOC (as pounds VOC) was divided by this calculated
propane equivalent to determine the overall response factor needed to precisely match
the TTE capture efficiency.
The results of this analysis demonstrate the importance of the response factors to the
precision and accuracy achieved by the CMI mass balance test. The overall response
factor had to vary from 1.29 to 1.47 in order to exactly match results obtained by the TTE
method. This variation of required response factors suggests that the use of a single
theoretical response factor calculated from the coating and thinner specifications would
increase variation in the capture efficiency measurement results. The comparison of the
actual and required overall response factors shows good agreement. With the exception
of Test Runs 2,3, and 6, the actual overall response factors are within 5% of the required
overall response factors. The greatest deviation occurred in Test Run 2 where the actual
overall response factor deviated from the required overall response factor by +12.2%,
which results in a CMI mass balance capture efficiency that was 11.9% higher than the
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-10
liquid/gas TTE capture efficiency. This underscores the need for additional quality
assurance checks in the measurement of the response factor, as it has the single most
significant effect on the capture efficiency.
4.6 Analysis of Short Duration Test Results
Liquid VOC weight measurements were intentionally taken on 1-hour time intervals to
allow an estimate of the capture efficiency within 1-hour and 2-hour segments that
comprised the 3-hour test run. Since each of the test runs required a change of the roll
wash solvent, this event was selected as the breakpoint to divide the 3-hour test run into
shorter segments. This provided VOC content and response factors for the roll wash
solvent at the start and end of each time interval. The roll wash solvent change typically
occurred approximately 2 hours into the test run which allowed development of capture
efficiencies for the first 2 hours and for the last 1 hour of the 3-hour run. A brief
description of the methodology used to extract the short duration capture efficiencies
from the 3-hour test runs is as follows:
1. Coating consumption versus time was plotted and the net weight of
coating in the system at the breakpoint (roll wash solvent change over) was
calculated by linear interpolation. Graphs presenting this analysis are
included as Section 1 of Appendix D.
2. The response factor and VOC content of the coating material at the
breakpoint were also determined by a linear interpolation of the results
measured for the start and end of the 3-hour test period. Graphs
presenting this analysis are included as Section 2 of Appendix D.
3. The weights, response factors, and VOC contents for roll wash solvent
samples at the breakpoint were measured directly during the test program
and no linear interpolation of data was necessary.
4. The liquid VOC consumption (L) for the time period preceding (2-hour
run) and following (1-hour run) the breakpoint, was calculated using the
original test data and the interpolated breakpoint values for coating
consumption, coating VOC content, and coating response factor. These
liquid VOC calculations for the short duration test runs are included as
Section 3 of Appendix D. In each case, the sum of VOC consumption
calculated for both short duration time intervals equalled the total VOC
consumption measured for the 3-hour test run.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-11
5. The average VOC concentration in the oven exhaust duct, capture hood
duct, fugitive exhaust duct, and background air pulled through the TIE
was determined for the 2-hour time period preceding the breakpoint and
the 1-hour time period following the breakpoint. This data was extracted
from the database of 2-minute averages recorded by the data integration
system on the FIA sample train. This summary of short duration test run
VOC concentrations is included as Section 4 of Appendix D.
6. Values were calculated for the captured emissions (G) and fugitive
emissions (F) for each of the short duration time intervals using the above
listed data. The calculation summary sheets for this analysis are included
as Section 5 of Appendix D. The sum of VOC emissions calculated for the
two short duration time intervals is very close to the total VOC emissions
measured for the 3-hour test run. Minor deviations were likely due to
rounding errors and very minor inaccuracies in defining the average gas
phase VOC concentration for the shorter time periods.
7. The short duration time interval values for L, G, and F were input into the
same calculation matrix used to determine the 3-hour capture efficiencies
by each of the three CE test.methods.
A summary of the capture efficiency results obtained for the short duration test runs for
each of the three CE test methods along with a comparison of the results is presented
on Tables 5 through 8. The TTE test methods still produced consistent results with little
variability over the shorter duration time intervals. Precision in the 2-hour test periods
for both TTE methods was virtually identical to the 3-hour test results. Variability in the
1-hour test periods increased slightly for the TTE test methods with the coefficient of
variation increasing to 0.6% and 0.8% for the gas/gas and liquid/gas TTE methods,
respectively. The 95% confidence interval also expanded to approximately +/- 0.5% as
opposed to +/- 0.3% for the 3-hour test runs. The average capture efficiencies measured
by the 2-hour and 1-hour time intervals were virtually identical to those measured for
the 3-hour test run. This analysis indicates that the sheet-fed coating process emission
rate variability is essentially the same over 2-hour or 3-hour periods, and only slightly
higher for a 1-hour period.
The CMI mass balance results showed an increase in variability with a decrease in the
test run duration. A summary of the statistical analysis on results from the 3-hour, 2-
hour, and 1-hour time intervals is summarized below:
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-12
Parameter
Range of Capture Efficiency
Average Capture Efficiency
Coefficient of Variation
95% Confidence Interval
Results from CMI Mass Balance
3-Hour Runs
87.% to 105.1%
95.7%
5.6%
91.6% to 99.9%
2-Hour Runs
83.1% to 111.4%
96.1%
8.0%
90.2% to 102.1%
1-Hour Runs
80.5% to 114.4%
94.6%
10.3%
86.4% to 102.8%
The 2-hour test run data still showed reasonable repeatability with a coefficient of
variation of 8.0% and the 95% confidence interval limits deviating by +/- 6.0% from the
mean. The deviation of the confidence interval limit around the mean is still less than
the EPA reported probable error for a standard VOC emissions test (Method SOB). It
should also be noted that decreasing the test run duration from 3 hours to 2 hours will
likely eliminate the need to change roll wash solvent during the test and may actually
eliminate one potential source of error.
The 1-hour test run data summary showed significant degradation in repeatability with
the coefficient of variation increasing to 10.3% and the 95% confidence interval limits
deviating by +/- 8.2% from the mean. This represents almost double the variability
measured by the 3-hour test runs. Based on these findings, no further evaluation of 1-
hour test runs was conducted as part of this study.
4.7 Error Band Analysis for Multiple Run Configurations
CMI met with representatives of the Office of Air Quality Planning and Standards
(OAQPS) on July 20, 1993 to discuss the preliminary report submitted on this project.
As a result of these discussions, CMI was asked to present a proposal for the appropriate
number of test runs and run durations to ensure that the test method would generate
repeatable and reliable capture efficiencies. OAQPS defined the acceptable error band
as less than 5% of the capture efficiency mean. The error band is calculated as the
deviation of the 95% confidence limits from the mean, expressed as a percentage of the
mean. OAQPS stated that this 5% error band was necessary to ensure that the reported
capture efficiency was sufficiently reliable to comply with the can coating RACT
standards.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-13
CMI completed the error band analysis using the following methodology:
1. A database of possible multi-run capture efficiency test results was
compiled by extracting the means of various combinations of multiple
consecutive runs. Nine sets of any multiple consecutive run test can be
extracted from the CE test study data. In accordance with Method 301,
these means are the data that a facility would submit to certify that it is
achieving an appropriate capture efficiency to meet its overall control
efficiency requirements, and are based on the average of multiple replicate
analyses.
2. The following test configurations were extracted from the data:
Two 3-hour runs
Three 3-hour runs
Four 3-hour runs
Two 2-hour runs
Three 2-hour runs
Four 2-hour runs
A sufficient number of data points to ensure a valid statistical analysis
were generated by extracting the nine possible combinations of consecutive
multiple run means. In addition, use of the entire data set to validate the
proposed test is consistent with the procedures specified in EPA Method
301.
3. A 95% confidence interval was calculated for each of the test configurations
listed in Item 2. Because each of these confidence intervals were calculated
from the same number of data points, a meaningful comparison between
the results can be made (the student's t value is the same for each data set
and the standard error is calculated using the same denominator in each
case).
4. The deviation between the confidence limit and the mean (the student's t
value times the standard error) was then expressed as a percentage of the
mean to calculate the error band.
Summary tables that identify the block of data from which each multiple run test mean
was extracted and that provide the results of the error band calculations are included
as Tables 9 through 14. For all cases, the error band was less man the EPA criteria of
5% of the mean. The error band ranged from 2.2% (four 3-hour runs) to 4.3% (two 2-
hour runs). A summary of the results is presented below:
Environmental Resources Management-North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-14
Test Configuration
No. of Runs
2
3
4
2
3
4
Run Duration
2-hour
2-hour
2-hour
3-hour
3-hour
3-hour
Error Band
+/- 4.3%
+/- 3.1%
+/- 2.7%
+/- 3.3%
+/- 2.6%
+/- 2.2%
These error bands represent the range around the measured mean for the various
multiple-replicate test results in which the true mean occurs within a 95% certainty. All
of the above-listed test configurations meet the EPA criteria for repeatability. Given that
most facilities should have at least a 5% margin between the actual and required capture
efficiency for their coating lines, these error bands demonstrate adequate precision for
CMI's proposed test method.
CMI is proposing that three 2-hour runs be adopted for its protocol. This test
configuration provides adequate precision (error band = +/- 3.1%) and is consistent with
the follow-up testing requirements specified in EPA Method 301. Method 301 requires
at least nine test runs when using a paired sampling system to validate a proposed
method by comparison to an EPA validated method (Method 301-Section 5.2.1). The
actual validation of the test method is based on statistics generated by the difference
between the paired sample set and it uses all nine data points. Once a method is
validated by this approach, the follow-up compliance testing by a source requires three
test runs (Method 301-Section 11.4). More than three test runs are only required if a
proposed method is validated by the analyte spiking procedures and the relative
standard deviation exceeds 15%. The relative standard deviation is analogous to the
calculation of a coefficient of variation. The statistical analysis of the CMI mass balance
results showed coefficients of variation well below 15% for both the 2-hour and 3-hour
data sets. This comparison also suggests that a three run test series is adequate for the
follow-up compliance testing and should be adapted in the standard protocol for the
CMI method.
CMI also believes that a 2-hour run duration may minimize the probable error margin
for this test procedure. This conclusion is based on the fact that a 2-hour run will reduce
the probability of having to change the roll wash solvent during the test period and
eliminate potential error introduced through additional field measurement and sample
analyses required for a roll wash solvent change.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-15
4.8 Parametric Study (Runs 10 through 12)
After completion of the formal capture efficiency study (nine 3-hour test runs), three
additional test runs were completed to determine the effect of changes in the operation
of the TTE and coating line. This parametric study included the following modifications
and test conditions:
Run 10 - A V-shaped baffle was suspended from the ceiling of the TTE to deflect
flow from NDOs 5 and 10, which appeared to be impinging on the capture hood
above the roll coater. Smoke tests verified a turbulent swirling pattern generated
by the air currents emanating from these NDOs. The baffle was installed to
deflect this flow outward towards the sides of the enclosure to lessen the impact
on normal flow patterns around the roll coater. A three-dimensional drawing of
the TTE showing the approximate location of this baffle is included as Figure 10.
Run 11 - The TTE ventilation was increased to the exhaust fan's maximum rate
(approximately 8,200 cfm) to determine the effect of increased air flow within the
enclosure. The baffle installed for Run 10 was left in place. Additional draft
openings (NDO's 15 through 18) were cut into the TTE to maintain a face velocity
between 200 and 250 fpm. The location and separation to VOC emitting points
for these new NDOs are also shown on Figure 10.
Run12 - A higher VOC coating was run to determine the effect on capture
efficiency when the liquid phase VOC input was significantly increased.
Results from the parametric study are summarized on Table 15 and the data summary
sheets are included in Appendix B. Capture efficiencies measured by the TTE methods
were very consistent. The gas/gas TTE results ranged from 95.3% to 96.6% and the
liquid/gas TTE results ranged from 94.6% to 95.9%. These results are slightly higher
than the average for the nine 3-hour test runs which suggests that the baffle installed
within the TTE improved the capture characteristics of the line. However, the average
increase in capture efficiency was fairly low (approximately 1.5%) and may have been
attributable to increased variability associated with a shorter test run duration.
No significant impact was observed after increasing ventilation through the TTE or when
using a higher VOC coating (Runs 11 and 12, respectively). The capture efficiencies for
both TTE methods remained fairly steady through Runs 10,11, and 12 suggesting that
the magnitude of changes made to these parameters did not cause enough disruption
to air flow within the capture zones or normal evaporation rates at the VOC emitting
points to significantly change the line's capture efficiency.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-16
The CMI mass balance results for the parametric study ranged from 94.1% to 119.2%.
The results for Run 10 correlated very well with the TTE capture efficiencies, however,
the mass balance results for Runs 11 and 12 were skewed high. Unusual results from
the analysis of coating response factors and VOC contents appear to have contributed
to the upward bias in Runs 11 and 12. Coating VOC contents for Run 11 were more
than 1.5% lower than Run 10, where the same coating material was used. Also, the
response factors measured for Run 11 were not consistent with the trend for the
remaining data, where the final coating sample typically had a higher response factor
than the initial sample. If the Run 10 coating VOC contents and the average response
factors for coating samples obtained for Runs 1 through 9 are substituted into the Run
11 calculations, the mass balance capture efficiency is reduced to 98.4%. A similar
comparison cannot be made for Run 12 as a different coating was used. However, the
normal trend of increasing response factors between the initial and final coating was not
observed for this test run. This inconsistency likely contributed to the low liquid VOC
consumption and resulting high capture efficiency.
These results again stress the importance of accurate measurements of coating response
factors and VOC contents. The amount of variability introduced by errors in these
measurements is increased with a shorter test run. This condition was considered in
CMI's decision to eliminate 1-hour test runs from the assessment of appropriate run
duration and number of tests to generate reliable capture efficiency results.
4.9 TTE Impact on Coating Line Capture
The impact of the TTE on the coating line's VOC capture characteristics was evaluated
by comparing capture hood performance during the baseline run to that achieved after
installation of the TTE, and through velocity/smoke tests around the capture hoods and
oven entrance. A comparison of the capture hood performance is presented on Table
16. Performance was evaluated by calculating the percent of total captured VOC
emissions removed by the hood system during each test run. This was possible as VOC
emissions from the oven exhaust duct and capture hood exhaust duct were measured
separately. With the exception of Run 1, the percent of the total captured emissions
removed by the hood system after the TTE was installed (11.2% to 20.6%) was lower
than the percent removed during the baseline tests (21.1%). The average percent
removal by the capture hood system for Test Runs 2 through 12 was 17.1%, which
represents a 19% decrease over the baseline run. This analysis represents the most direct
way to evaluate the TTE's impact on the amount of VOC captured by the hood system.
Since the total VOC delivered to the coating line varied from 38.8 to 63.2 Ibs/hr, an
analysis of the run-by-run capture by the oven or hood system separately is not valid
unless the data is normalized. Table 17 presents an analysis where all runs have been
normalized so that the total VOC captured is equivalent to the baseline run. After this
normalization is completed, the VOC removed by the capture hoods is less than the
baseline test for Runs 2 through 12.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-17
Increased turbulence was observed during the smoke tests in the zones adjacent to the
capture hoods located above and around the roll coater. Prior to installation of the TTE,
the smoke trails were observed to flow in a relatively straight path with a slight vector
towards the left side of the coating line. This flow vector was caused by a slight cross
draft through the plant. After installation of the TTE, the smoke trails indicated a
swirling pattern and dissipated much more rapidly. These observations suggest that
VOC in the vicinity of the capture hoods is more rapidly dispersed in the air flow
moving through the TTE, which would cause a less concentrated stream to be captured
by the hood system. This conclusion is supported by analysis of the concentration
within the capture hood gas stream. All runs conducted after the TTE was installed
showed a lower VOC concentration in the captured gas stream than the baseline run.
On average, the capture hood VOC concentrations measured during the TTE runs were
7% lower than that measured during the baseline run.
The increased turbulence within the capture zones was also documented by the air
velocity surveys made during each of the test runs. A three-dimensional schematic
diagram showing each of the capture hoods around the roll coater is presented as Figure
11. The most significant turbulence was observed at the front face (the surface facing
the sheet feeder) of the roll coater upper hood. A plot of the average velocity 3 and 6
inches away from the slots on this hood is included as Figure 12. During the baseline
run, the average velocities were 71 fpm and 36 fpm for readings taken 3 inches and 6
inches away from the hood slots, respectively. Average velocities taken after the TTE
was installed and prior to installation of the baffle (Runs 1, 8, and 9) show the average
velocity 6 inches from the slots greater than that measured at 3 inches away. Installation
of the baffle (Runs 10,11, and 12) improved the flow pattern at this hood, however, the
readings 6-inches from the slots were still disproportionally high suggesting increased
dispersion of VOC in the TTE.
Similar plots of me 3-inch and 6-inch velocity readings for other capture hoods are
included as Figures 13 through 15. These figures show similar disruptions in the flow
pattern, although not as pronounced as those measured on the front face of the roll
coater upper hood. The hood that was least affected by the turbulence was the exit
hood located just downstream of the roll coater. The importance of this hood is that it
would tend to catch a significant portion of the VOC vapors that are swept away from
the roll coater upper hood towards the TTE ventilation intakes located at the oven
entrance side of the enclosure. This condition was documented in an additional test
where the VOC concentration and velocity in each leg of the capture hood system was
measured to determine the percent of the total VOC load removed by each individual
hood. The results from this analysis are summarized on Table 18 and indicate that the
roll coater exit hood is capturing nearly 70% of the VOC emissions removed by the hood
system. Therefore, if a coating line was only equipped with an upper hood (as is the
case with many facilities), the TTE's impact on capture hood performance may have
been more pronounced.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 4-18
The other critical area evaluated was the draft into the oven entrance. A velocity grid
of 8 readings across the face of the oven entrance were taken in order to calculate the
air flow pulled into the front end of the oven. The baseline test showed a flow of
approximately 920 scfm, whereas the remaining tests conducted after installation of the
TTE showed oven entrance flows ranging from approximately 240 scfm to 370 scfm. A
plot of the oven entrance flow for the same group of test runs which had full velocity
measurements is included as Figure 16.
The smoke tests conducted at the oven entrance after installation of the TTE also
confirmed this condition. The smoke trace was visibly deflected downward and
outward from the oven entrance as the rube was moved away from the face opening.
Again, this condition could have a more substantial effect on a coating line's overall
capture efficiency if it did not have a multiple-hood capture system (or any capture hood
system) and more heavily relied on VOC capture by the oven entrance negative draft.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 5-1
5.0 ECONOMIC, SAFETY, AND FEASIBILITY CONSIDERATIONS
5.1 Economic Considerations
5.1.1 Cost of TTE Design and Installation
The typical cost to design and install a TTE were derived from the consultants invoices
submitted to CMI for the capture efficiency test method study. ERM-North Central
prepared the engineering design of the TTE and Smith Engineering fabricated and
installed the TTE. A portion of ERM-North Central's invoices and all of Smith
Engineering's invoices are considered appropriate costs for a typical TTE design and
installation.
The TTE design required completion of the following tasks:
1. A facility inspection and meeting to obtain detailed measurements on the
coating line layout, interview line operators, and gather information on the
coating and solvents to be run during the test periods;
2. Calculation of the required ventilation rate to comply with OSHA
permissible exposure limits (PELs) and the equivalent FIA response to
monitor for compliance with OSHA PELs;
3. Preparation of a preliminary layout of the TTE and ventilation system for
facility approval;
5. Preparation of the final TTE and ventilation system design incorporating
facility comments; and
6. Revisions to TTE design to account for USEPA comments.
The following costs were extracted from ERM-North Central's invoices to cover these
tasks:
• Labor for site visit (12 hrs) = $ 1,320
• Expenses for site visit = $ 150
• Engineering labor for ventilation calculations (16 hrs) = $ 1,390
• Engrg. labor for TTE/ventilation system design (40 hrs) = $ 3,210
• Computer drafting labor for TTE design (72 hrs) = $ 2,880
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 5-2
• Computer expenses for TIE CAD design = $ 510
• Labor to redesign TTE based on USEPA comments (6 hrs) = $ 940
SUBTOTAL =$10,400
The costs for fabrication and installation of the TTE roughly broke down into the
following elements:
• Procure ventilation fan (10 hrs) = $ 850
• Rental of ventilation fan = $ 825
• Procure construction materials/
coordinate with contractors (60 hrs) = $ 5,100
• Materials to construct TTE and sheet metal ductwork = $ 5,165
• Work out logistics prior to test (44 hrs) = $ 3,740
• Crown's maintenance department labor for installation
of TTE (288 hrs) =$ 4,410
• Install and dismantle TTE and ductwork
Union Labor (80 hrs S.T. and 25 hrs O.T.) = $ 3,835
• Supervise TTE construction (80 hrs S.T. and 39 hrs O.T.) = $ 11,700
• Modifications to TTE/assistance during testing (38 hrs) = $ 3,230
• Dismantle TTE/demobilize from site (22 hrs) = $ 1.870
SUBTOTAL =$40,725
TOTAL TTE COST =$51,125
5.1.2 Cost of Standard Gas/Gas TTE Test
The cost estimate for conducting the standard gas/gas TTE test method was obtained
by requesting a cost proposal from Mostardi-Platt Associates (MPA), the source testing
contractor used on this CMI project. MPA was asked to quote a control efficiency test
(destruction plus capture efficiency) on a sheet fed coating line that included the
following elements:
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 5-3
• two test stations at the inlet to the thermal oxidizer and one fugitive
emissions test station from the TTE;
• internal VOC monitoring of the TTE;
• background VOC monitoring of the TTE natural draft openings;
• monitoring of the cooling zone intake and discharge stack;
• Method SOB VOC emissions testing at each test station;
• one baseline run at the thermal oxidizer inlet test stations;
• balancing test after construction of the TTE; and
• three 3-hour TTE test runs.
MPA's quote included in Appendix E consisted of the following lump sum prices:
• control efficiency test (including cooling zone monitoring) - $38,875
• control efficiency test (excluding cooling zone monitoring) - $30,850
« destruction efficiency test only - $5,900
In addition to the source testing contractor's costs, the facility would have to provide a
test coordinator at a cost of approximately $30 per hour. The TTE installation and
testing is assumed to require five 12-hour days which results in a facility cost of $1,800.
Therefore, the total incremental cost to add a TTE capture efficiency test to a standard
destruction efficiency test would range from $26,750 to $34,775, depending on whether
the cooling zone had to be monitored as a fugitive emission point source.
5.1.3 Production Losses During a TTE Test
Production losses that were incurred by Crown during the CMI study were .compiled
to calculate unit cost factors for determining the loss that would typically occur during
a standard three run, 3-hour TTE test series. These costs are categorized into two areas:
Manufacturing Volume Loss - The economic impact caused by a reduction in
available hours of coater production time. The unit cost factor derived from the
CMI study was $45 per hour of downtime. This loss was directly attributed to
the TTE test because the line production had to be stopped between test runs
(running with the TIE required workplace air monitoring) and to allow for
construction and dismantling of the enclosure.
Environmental Resources Management- North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 5-4
Increased Spoilage Rate - The cost that results from an increase rate of off-
specification sheets coated by the line. The range of spoilage (percentage of off-
specification sheets) for Coating Line No. 9 when running the same product and
coating used during the test is 0.5% to 1.0%, with an average rate of 0.7%. The
spoilage during the CMI study amounted to 8,800 sheets, over a total production
of approximately 218,000 sheets. However, over half of the spoilage occurred
during the first two runs (approximately 4,500 sheets) and was traced to off-
specification plate received from the mill. Therefore, this spoilage was subtracted
from the total and the spoilage rate during the testing was calculated as [(8,800 -
4,500) / 218,000], or 2.0%. Because the operator cannot respond to a problem on
the coating line as quickly when operating within an enclosure (plus accessibility
to inspect the coating process is also limited by the enclosure), the increase in
average spoilage rate was judged to be directly attributable to the TTE test. On
average, the increased spoilage rate caused by the Tit testing was estimated to
be approximately 1.3%. The unit cost rate for off-specification product derived
from the CMI study was $1.55 per sheet.
The following is a summary of the estimated production losses that would occur on a
typical sheet fed coating line over a standard three run, 3-hour TTE test series:
Manufacturing Volume Loss
TTE construction time = 48 hours
Downtime during TTE testing - 2 days x 12 hours/day = 24 hours
TTE dismantling/line cleaning = 4 hours
Total downtime = 76 hours
Production loss - 76 hours x $45/hour = $3,420
Increased Spoilage Rate
Sheets run with TTE - 24 hrs x 60 min/hr x 75 shts/min = 108,000 sheets
Spoilage caused by TTE - 108,000 sheets x 1.3% = 1,400 sheets
Production loss - 4,320 sheets x $1.55 per sheet = $2,170
The total production loss estimated for a standard TTE test is $5,590.
5.1.4 Cost of CMI Mass Balance Capture Efficiency Test
The cost estimate for conducting CMI's improved liquid/gas mass balance test method
was also obtained by requesting a cost proposal from MPA. MPA was asked to quote
a control efficiency test (destruction plus capture efficiency) on a sheet fed coating line
that included the following elements:
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 5-5
• two test stations at the inlet to the thermal oxidizer;
• Method SOB VOC emissions testing at both test stations;
• Method 30F analysis of 12 liquid samples;
• portable scale weight measurements;
• a 30-minute baseline run before each test run at the thermal oxidizer inlet
test stations to measure the FIA response caused by background VOC and
unburned natural gas; and
• three 3-hour liquid/gas mass balance test runs.
MPA's quote is included in Appendix E and consisted of the following lump sum prices:
• control efficiency test - $15,350
• destruction efficiency test only - $5,900
The CMI mass balance testing is assumed to require two 12-hour days which result in
a facility cost of $720. Therefore, the total incremental cost to add the CMI mass balance
capture efficiency test to a standard destruction efficiency test would be $10,170.
5.1.5 Capture Efficiency Cost Comparison
The total incremental cost of adding a TTE capture efficiency test to a standard
destruction efficiency test is estimated to range from approximately $83,500 to $91,500
based on the summation of the following cost elements:
• TTE design - $10,400
• TTE installation - $40,725
• TTE testing - $26,750 to $34,775 (depending on cooling zone monitoring)
• Production losses - $5,590
The total incremental cost of conducting the CMI mass balance test is estimated to be
approximately $10,200. Based on these estimates, the TIE method could cost up to nine
times more than the CMI mass balance test. Imposing the TTE test method on the can
manufacturing industry could cost a facility an additional $73,000 to $81,000 per line
tested, depending on whether cooling zone monitoring is necessary.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 5-6
5.2 Safety Considerations
Operating a coating line during a TTE test with the coater area inside a relatively small
enclosure will create undesirable safety conditions. The limited clearance between the
enclosure walls and the equipment around the coating line will restrict the operator's
movement and ability to remove or replace a sheet from the belt conveyor system. The
coating line conveyor is typically running at a fairly rapid rate (70 to 90 sheets per min)
which equates to a line speed of over 200 ft/min. There are several mechanical drives
and moving parts associated with the roll coater, belt conveyor system, and wicket
conveyor. Therefore, the reduced mobility and restricted space inside of the enclosure
will increase the chance of a physical injury to the operator.
Chemical exposure to the operator inside of the enclosure will also be increased to due
confining VOC emission points. VOC concentrations within the breathing zone were
monitored as part of the CMI study. Four probes were located at positions that
simulated a time weighted average exposure for the operator. Each of these monitoring
points was located approximately 5' 9" above floor level to simulate the breathing zone
for the operator. Locations of these breaming zone monitoring points are shown on
Figure 17. A summary of the average breathing zone VOC concentration (measured as
a propane equivalent) is presented below:
Breathing
Zone VOC
(ppm QHg)
Percentage of
Baseline Test
Baseline
Test
27.2
100%
TTE Run Number
1
61.6
226%
2
32.8
121%
3
402
148%
4
57.9
213%
5
63.0
232%
6
75.0
276%
7
88.6
326%
8
73.4
270%
9
74.8
275%
As noted on the above-listed summary, the baseline breathing zone VOC concentration
(measured prior to installation of the TTE) was lower than the concentrations measured
during the full 3-hour TTE runs. On average, the breathing zone concentration for Runs
1 through 9 (the full TTE runs) was 63 ppm which is over twice as high as the baseline
breathing zone concentration. Also, if the curing oven balance is upset to the point
where the negative draft at the oven entrance is reversed, the TTE could draw solvent
vapors from inside the oven into the enclosure work space. This could further increase
the VOC concentration within the TTE.
In addition to chemical exposure, there is always a possibility of a flash fire when
working with flammable materials. Based on inquiry with long-term Crown Cork & Seal
lithography personnel, a typical lithograph operation at a can plant will experience a
coater fire on a frequency of approximately once per ten years. Enclosing the coater
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 5-7
(even with adequate ventilation) will increase the chance of a flash fire due to the
potential presence of dead zones within the enclosure. VOC measurements were taken
during the balancing test to try to identify the range of concentrations that may occur
in some of the potential dead zones within the TIE. These readings ranged from 35
ppm up to 180 ppm, confirming the fact that uniform ventilation through the entire
work space was not achieved. The TTE would also intensify the hazard of a flash fire
in two possible ways:
1. It would limit the expansion of the flash fire, with the confinement of
organic vapors, and tend to concentrate and intensify the fire at the vapor
source; and
2. It would limit the emergency egress for coating line operators to the
specific doors or other openings designed into the enclosure.
The primary danger to the operator is the limited emergency egress from the TTE.
Under normal operating conditions, the operator would have unlimited avenues of
escape and could take a direct path away from a flash fire. With a TTE/ the available
escape options are limited to those pathways to the doors or other openings within the
enclosure walls. If a fire would occur between the operator and an available exit point,
there would be no avenue for escape. The potential for an explosion or fire hazard is
also increased by confining a VOC emitting area that has electric wiring, motors, and
controls that may not be explosion proof systems.
5.3 Feasibility Considerations
Construction of a functional TUB on a sheet fed coating process must allow the operator
to work inside the enclosure to perform many of the tasks required to run the line. The
TTE constructed for the CMI study was approximately 25 feet long and ranged in width
from 14 to 15.5 feet. This TTE was constructed on a coating line that had very favorable
spacing as it was not located within a cluster of coating lines and the center to center
spacing between it and the adjacent was over 14 feet. Most facilities have coating lines
clustered with much tighter center line spacing. As an example/ the same Crown Cork
& Seal facility has four coating lines clustered with center line spacing of less than 12
feet.
Construction of a functional TTE on a coating line clustered between others may not be
feasible, or at a minimum, would require the adjacent lines to be shut down to conduct
the test. An example of this condition is shown on Figures 18 and 19. These drawings
were adapted from a FIP Revision submitted by White Cap, Inc., a sheet fed metal
decorating plant located in Chicago, Illinois. Figure 18 shows the typical ancillary
equipment located around a sheet fed coating line. Figure 19 shows a functional TTE
superimposed on the layout of three clustered coating lines. The TTE is 14 feet wide
and the center line spacing between the coating lines is 11.5 feet. The walls of the TTE
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 5-8
actually intersect some of the ancillary equipment located on the adjacent lines. At a
minimum, both of these adjacent lines would have to be shut down to allow the TTE
testing of the line between them. Based on a manufacturing volume loss of $45/hour
and a typical schedule of 5 working days to complete a TTE test (construction, testing,
and dismantling time), the economic impact to the facility would be an additional
$10,800. It should also be noted that this drawing is a simplification of the actual
conditions within a coating operation. Materials may be staged between lines and other
obstacles such as support columns, electrical conduits, and miscellaneous objects may
not have been included. In many cases, it may not be feasible to construct an adequately
sized TTE within a clustered coating line setting.
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 6-1
6.0 CMI PROPOSAL FOR CAPTURE EFFICIENCY TEST METHOD
CMI believes that this capture efficiency test method study has demonstrated that its
improved liquid/gas mass balance method is an acceptable alternative to USEPA's TTE
methods. The CMI mass balance method generated repeatable results (coefficient of
variation = 5.6%) and good accuracy (average capture efficiency within 2% of TTE
methods). The error band analysis also indicates that the reported mean capture
efficiency from the test method will not deviate from the true mean by more than
approximately 3% when three 2-hour tests are conducted. Therefore, CMI is proposing
that USEPA re-evaluate the capture efficiency test requirements set forth in the Chicago
FIP and other SIP revisions. Specifically, CMI is proposing that its mass balance test
protocol (attached as Appendix A) be approved as another acceptable test method for
measurement of sheet fed coating line capture efficiency. The protocol follows portions
of the February 17, 1993 Capture Efficiency Study Test Plan that pertain to the
liquid/gas mass balance procedures and includes the following enhancements:
• Background VOC and uncombusted natural gas contributions to the
captured VOC emission rate will be determined by measuring the FIA
response in the oven exhaust and capture hood exhaust ducts. This
response will be measured over a 30-minute period with an empty oven
and empty coating reservoirs prior to each test run. This procedure will
eliminate the need for converting methane/ethane concentrations to the
FIA propane equivalent.
• Method 30F will include two enhancements to make laboratory
measurements more closely simulate field conditions. These modifications
will include measuring percent volatiles in a convection oven that is set at
the same temperature the wicket oven was run at during the test. Also,
the temperature of the rotary evaporator oil bath will be increased to
ensure that all coating VOC fully evaporates. The optimum temperature
will be selected based on trial runs with the coating material that include
weighing tine residue remaining in the evaporation flask to calculate a
percent volatile that can be compared to the convection oven
measurements.
• Additional quality assurance steps will be added to the Method 30F
response factor measurements. Duplicate bag samples will be prepared at
three different concentrations by varying the amount of solvent extract
injected into the zero-air gas stream. If these duplicates do not agree
within +/- 5%, the analysis will be repeated. The data will be plotted to
generate a curve showing VOC concentration (mg/m3 of VOC) versus FIA
response (mg/m3 of propane). The linearity of the response curve will be
assessed by a linear regression and the analysis repeated until a correlation
coefficient of 0.90 or higher is achieved. The response factor will be
Environmental Resources Management - North Central, Inc.
-------
Can Manufacturers Institute
Capture Efficiency Report
October 22,1993
Page 6-2
defined as the slope of the regression line. In addition, a hexane standard
will be run through the system prior to and after analysis of the samples.
This standard should generate a response factor of 1.0 and is an excellent
quality assurance check on the system.
At the request of OAQPS, CMI also evaluated several test configurations (the number
and duration of runs that comprise a full capture efficiency test) to determine which
combination would produce acceptable repeatability. This analysis is summarized in
Section 4.7 which includes the calculation of error bands for a matrix of six different test
configurations. Based on this analysis and other contributing factors discussed in this
report, it is proposed that the CMI mass balance test method include three 2-hour test
runs. This test configuration has been incorporated into the protocol included as
Appendix A.
Environmental Resources Management - North Central, Inc.
-------
TABLES
-------
TABLE 1
SUMMARY OF CAPTURE EFFICIENCY RESULTS
RUN
NUMBER
B.L.
1
2
3
4
5
6
7
8
9
IU
LIQUID PHASE
TOTAL VOC
Ubs C3H8)
132.43
123.50
116.36
134.53
145.30
136.00
149.83
130.01
146.82
143.53
IG]
GAS PHASE
CAPTURED VOC
(Ibs C3H8)
120.12
118.10
122.28
136.95
135.42
130.63
131.32
120.73
144.28
130.57
IF}
GAS PHASE
FUGITIVE VOC
(Ibs C3H8)
7.43
7.05
8.10
7.41
7.65
8.02
7.79
9.00
8.87
AVERAGE CAPTURE EFFICIENCY (RUNS 1 - 9)
COEFFICIENT OF VARIATION (RUNS 1 - 9)
95% CONFIDENCE INTERVAL (RUNS 1 - 9)
CAPTURE EFFICIENY TEST RESULTS
TTE TEST METHODS
GAS/GAS
I G/(G+F) ]
94.1%
94.5%
94.4%
94.8%
94.5%
94.2%
93.9%
94.1%
93.6%
94.3%
0.4%
94.0%
to
94.5%
LIQUID/GAS
t (L-F)/L) ]
94.0%
93.9%
94.0%
94.9%
94.4%
94.6%
94.0%
93.9%
93.8%
94.2%
0.4%
93.9%
to
94.5%
LIQUID/GAS
MASS BALANCE
[G/L]
90.7%
95.6%
105.1%
101.8%
93.2%
96.1%
87.6%
92.9%
98.3%
91.0%
95.7%
5.6%
91.6%
to
99.9%
-------
TABLE 2
TTE AVERAGE FACE VELOCITY
RUN
NUMBER
1
2
3
4
5
6
7
8
9
10
11
12
TTE
EXHAUST
(scfm)
6830
6859
6855
6754
6717
6811
6717
6787
6724
6867
8282
8077
CAPTURE
HOOD
EXHAUST
(scfm)
2179
2083
2063
2038
1936
2047
2067
2065
1993
2134
2134
2140
FLOW INTO
OVEN
ENTRANCE
(scfm)
327
256
318
363
278
252
262
249
365
251
340
241
TOTAL TTE
VENTILATION
RATE
(scfm)
9336
9198
9236
9155
8931
9110
9046
9101
9082
9252
10756
10458
TOTAL
NDO AREA
(sq. ft)
34.18
34.18
34.18
34.18
34.18
34.18
34.18
34.18
34.18
34.18
43.99
43.99
AVERAGE
NDO FACE
VELOCITY
(fpm)
273
269
270
268
261
267
265
266
266
271
245
238
-------
TABLE 3
COMPARISON OF TEST METHOD RESULTS
RUN
NUMBER
1
2
3
4
5
6
7
8
9
AVERAGE
MAXIMUM
MINIMUM
COEFF OF VARIATION
CAPTURE EFFICIENY TEST RESULTS
GAS/GAS
TTE
94.1%
94.5%
94.4%
94.8%
94.5%
94.2%
93.9%
94.1%
93.6%
94.3%
94.8%
93.6%
0.4%
LIQUID/GAS
TTE
94.0%
93.9%
94.0%
94.9%
94.4%
94.6%
94.0%
93.9%
93.8%
94.2%
94.9%
93.8%
0.4%
LIQUID/GAS
MASS BALANCE
95.6%
105.1%
101.8%
93.2%
96.1%
87.6%
92.9%
98.3%
91.0%
95.7%
105.1%
87.6%
5.6%
LIQUID/GAS MASS BALANCE
PERCENT DEVIATION FROM
GAS/GAS
TTE
1.6%
11.1%
7.8%
-1.7%
1.7%
-7.0%
-1.1%
4.4%
-2.8%
1.6%
11.1%
-7.0%
LIQUID/GAS
TTE
1.7%
11.9%
8.3%
-1.8%
1.8%
-7.4%
-1.2%
4.7%
-3.0%
1.7%
11.9%
-7.4%
-------
TABLE 4
ANALYSIS OF RESPONSE FACTOR VARIANCE
RUN
NUMBER
1
2
3
4
5
6
7
8
9
LIQUID PHASE
TOTAL VOC
(Ibs VOC)
162.46
171.14
191.59
208.69
191.84
187.48
179.03
215.44
205.02
LIQUID PHASE
TOTAL VOC
(Ibs C3HS)
123.50
116.36
134.53
145.30
136.00
149.83
130.01
146.82
143.53
GAS PHASE
CAPTURED VOC
(Ibs C3H8)
118.1
122.28
136.95
135.42
130.63
131.32
120.73
144.28
130.57
LIQUID PHASE
VOC TO MATCH
TTE RESULTS
(Ibs C3H8)
125.66
130.17
145.72
142.70
138.42
138.75
128.43
153.70
139.17
CAPTURE EFFICI
LIQUID/GAS
MASS BALANCE
95.6%
105.1%
101.8%
93.2%
96.1%
87.6%
92.9%
98.3%
91.0%
ENCY RESULTS
LIQUID/GAS
TTE
94.0%
93.9%
94.0%
94.9%
94.4%
94.6%
94.0%
93.9%
93.8%
AVERAGE
MAXIMUM
MINIMUM
ACTUAL
OVERALL R. F.
TO MATCH C.E. BY
L/G MASS BALANCE
1.32
1.47
1.42
1.44
1.41
1.25
1.38
1.47
1.43
1.26
1.47
1.25
REQUIRED
OVERALL R. F,
TO MATCH C.E. BY
L/G MASS TTE
1.29
1.31
1.31
1.46
1.39
1.35
1.39
1.40
1.47
1.24
1.47
1.29
-------
TABLE 5
SUMMARY OF CAPTURE EFFICIENCY RESULTS
2-HR TEST PERIODS EXTRACTED FROM 3-HR RUNS
RUN
NUMBER
B.L.
1
2
3
4
5
6
7
8
9
[L]
LIQUID PHASE
TOTAL VOC
(Ibs C3H8)
N.A.
90.15
92.36
106.66
78.83
93.42
108.98
114.76
108.25
88.79
[G]
GAS PHASE
CAPTURED VOC
(Ibs C3H8)
N.A.
83.65
102.89
105.12
74.91
91.03
90.54
106.93
109.76
82.02
[F]
GAS PHASE
FUGITIVE VOC
(Ibs C3H8)
5.26
5.53
6.32
4.21
5.49
6.31
6.96
6.54
5.54
AVERAGE CAPTURE EFFICIENCY (RUNS 1 - 9)
COEFFICIENT OF VARIATION (RUNS 1 - 9)
95% CONFIDENCE INTERVAL (RUNS 1 - 9)
CAPTURE EFFICIENY TEST RESULTS
TTE TEST METHODS
GAS/GAS
[ GAG+F) ]
94.1%
94.9%
94.3%
94.7%
94.3%
93.5%
93.9%
94.4%
93.7%
94.2%
0.5%
93.8%
to
94.5%
LIQUID/GAS
[ (L-F)/L) ]
94.2%
94.0%
94.1%
94.7%
94.1%
•94.2%
93.9%
94.0%
93.8%
94.1%
0.3%
93.9%
to
94.3%
LIQUID/GAS
MASS BALANCE
[GfL]
N.A.
92.8%
111.4%
98.6%
95.0%
97.4%
83.1%
93.2%
101.4%
92.4%
96.1%
8.0%
90.2%
to
102.1%
-------
TABLE 6
COMPARISON OF TEST METHOD RESULTS
2-HR TEST PERIODS EXTRACTED FROM 3-HR RUNS
RUN
NUMBER
1
2
3
4
5
6
7
8
9
AVERAGE
MAXIMUM
MINIMUM
COEFF OF VARIATION
CAPTURE EFFICIENY TEST RESULTS
GAS/GAS
TTE
94.1%
94.9%
94.3%
94.7%
94.3%
93.5%
93.9%
94.4%
93.7%
94.2%
94.9%
93.5%
0.5%
LIQUID/GAS
TTE
94.2%
94.0%
94.1%
94.7%
94.1%
94.2%
93.9%
94.0%
93.8%
94.1%
94.7%
93.8%
0.3%
LIQUID/GAS
MASS BALANCE
92.8%
111.4%
,_ 98.6%
95.0%
97.4%
83.1%
93.2%
101.4%
92.4%
96.1%
111.4%
83.1%
6.9%
LIQUID/GAS MASS BALANCE
PERCENT DEVIATION FROM
GAS/GAS
TTE
-1.4%
17.4%
4.5%
0.4%
3.3%
-11.1%
-0.8%
7.4%
-1.4%
2.0%
17.4%
-11.1%
LIQUID/GAS
TTE
-1.5%
18.5%
4.8%
0.4%
3.5%
-11.8%
-0.8%
7.9%
-1.5%
2.2%
18.5%
-11.8%
-------
TABLE 7
SUMMARY OF CAPTURE EFFICIENCY RESULTS
1-HR TEST PERIODS EXTRACTED FROM 3-HR RUNS
RUN
NUMBER
B.L.
1
2
3
4
5
6
7
8
9
[L]
LIQUID PHASE
TOTAL VOC
(Ibs C3H8)
N.A.
N.A.
24.00
27.87
66.48
42.25
40.85
15.26
38.57
54.75
[G]
GAS PHASE
CAPTURED VOC
(Ibs C3H8)
N.A.
N.A.
19.32
31.89
61.31
39.40
41.30
13.93
35.34
50.56
[F]
GAS PHASE
FUGITIVE VOC
(Ibs C3H8)
N.A.
1.29
1.84
3.24
2.08
2.65
0.82
2.51
3.33
AVERAGE CAPTURE EFFICIENCY (RUNS 2 - 9)
COEFFICIENT OF VARIATION (RUNS 2 - 9)
95% CONFIDENCE INTERVAL (RUNS 2 - 9)
CAPTURE EFFICIENY TEST RESULTS
TTE TEST METHODS
GAS/GAS
[ G/(G+F) ]
N.A.
93.7%
94.5%
95.0%
95.0%
94.0%
94.4%
93.4%
93.8%
94.2%
0.6%
93.7%
to
94.7%
LIQUID/GAS
[ (L-F)/L) ]
N.A.
94.6%
93.4%
95.1%
95.1%
93.5%
94.6%
93.5%
93.9%
94.2%
0.8%
93.6%
to
94.8%
LIQUID/GAS
MASS BALANCE
IG/L]
N.A.
N.A.
80.5%
114.4%
92.2%
93.3%
101.1%
91.3%
91.6%
92.3%
94.6%
10.3%
86.4%
to
102.8%
-------
TABLE 8
COMPARISON OF TEST METHOD RESULTS
1-HR TEST PERIODS EXTRACTED FROM 3-HR RUNS
RUN
NUMBER
1
2
3
4
5
6
7
8
9
AVERAGE
MAXIMUM
MINIMUM
COEFF OF VARIATION
CAPTURE EFFICIENY TEST RESULTS
GAS/GAS
TTE
N.A.
93.7%
94.5%
95.0%
95.0%
94.0%
94.4%
93.4%
93.8%
94.2%
95.0%
93.4%
0.6%
LIQUID/GAS
TTE
N.A.
94.6%
93.4%
95.1%
95.1%
93.5%
94.6%
93.5%
93.9%
94.2%
95.1%
93.4%
0.8%
LIQUID/GAS
MASS BALANCE
N.A.
80.5%
114.4%
92.2%
93.3%
101.1%
91.3%
91.6%
92.3%
94.6%
114.4%
80.5%
10.3%
LIQUID/GAS MASS BALANCE
PERCENT DEVIATION FROM
GAS/GAS
TTE
N.A.
-14.1%
21.0%
-2.9%
-1.8%
7.6%
-3.3%
-1.9%
-1.6%
0.4%
21.0%
-14.1%
LIQUID/GAS
TTE
N.A.
-14.9%
22.5%
-3.1%
-1.9%
8.1%
-3.5%
-2.0%
-1.7%
0.4%
22.5%
-14.9%
-------
TABLE 9
ERROR BAND ANALYSIS
TWO 3-HOUR RUNS
RUN
NUMBER
CMIMASS
BALANCE
GAS/GAS
TTE
TWO RUN TEST SERIES MEANS
RUNS 1-2
RUNS 3-4
RUNS 5-6
RUNS 7-8
RUNS 2-3
RUNS 4-6
RUNS 6-7
RUNS 8-9
RUNS 9 & 1
95.6%
105.1%
101.8%
93.2%
96.1%
87.6%
92.9%
98.3%
91.0%
94.1%
94.5%
100.4%
94.3%
94.4%
94.8%
97.5%
94.6%
103.5%
94.5%
94.5%
94.2%
91.9%
94.3%
94.7%
94.6%
93.9%
94.1%
95.6%
94.0%
90.3%
94.1%
93.6%
94.6%
93.9%
93.3%
93.9%
ERROR BAND FOR TWO-RUN MEANS
SUCESSIVE
TWO-RUN
SERIES
1THRU2
3THRU4
5THRU6
7THRU8
2THRU3
4THRU5
6THRU7
8THRU9
9&1
AVERAGE
U.L.95%CI.
LL95%C.L
ERROR BAND
MEANCE.
CMIMASS
BALANCE
100.4%
97.5%
91.9%
95.6%
103.5%
94.7%
90.3%
94.6%
93.3%
95.7%
98.9%
92.5%
3.33%
MEAN C.E.
GAS/GAS
TTE
94.3%
94.6%
94.3%
94.0%
94.5%
94.6%
94.1%
93.9%
93.9%
94.3%
94.5%
94.0%
0.24%
BLOCK DATA LEGEND
100.4%
94.3%
2-RUN MEAN FOR CMI TEST METHOD
2-RUN MEAN FOR TTE TEST METHOD
-------
TABLE 10
ERROR BAND ANALYSIS
THREE 3-HOUR RUNS
RUN
NUMBER
CMI MASS
BALANCE
GAS/GAS
TTE
THREE RUN TEST SERIES MEANS
RUNS 1-3
RUNS 4-6
RUNS 6-9
RUNS 2-4
RUNS 5-7
RUNS 3-5
RUNS 6-8
RUNS 8,9,1
RUNS 94,2
95.6%
105.1%
101.8%
93.2%
96.1%
87.6%
92.9%
98.3%
91.0%
94.1%
94.5%
100.8%
94.3%
94.4%
100.0%
94.6%
94.8%
94.5%
92.3%
94.5%
:>,'2g**i'f,<
vfefefrj^,,,.^ _?•,„.<
97.0%
94.6%
97.2%
94.1%
94.2%
92.2%
94.2%
92.9%
93.9%
94.1%
94.0%
93.9%
94.1%
93.6%
94.9%
94.0%
ERROR BAND FOR THREE-RUN MEANS
SUCESSIVE
THREE-RUN
SERIES
1THRU3
4THRU6
6THRU9
2THRU4
5THRU7
3THRU5
6THRU8
8,9,1
9,1,2
AVERAGE
U.L.95%C.I.
L.L. 95% CI.
ERROR BAND
MEAN CE.
CMI MASS
BALANCE
100.8%
92.3%
94.0%
100.0%
92.2%
97.0%
92.9%
94.9%
97.2%
95.7%
98.2%
93.2%
2.60%
MEAN CE.
GAS/GAS
TTE
94.3%
94.5%
93.9%
94.6%
94.2%
94.6%
94.1%
94.0%
94.1%
94.3%
94.4%
94.1%
0.21%
BLOCK DATA LEGEND
100.8%
94.3%
3-RUN MEAN FOR CMI TEST METHOD
3-RUN MEAN FOR TTE TEST METHOD
-------
TABLE 11
ERROR BAND ANALYSIS
FOUR 3-HOUR RUNS
RUN
NUMBER
CMIMASS
BALANCE
GAS/GAS
TTE
FOUR RUN TEST SERIES MEANS
RUNS 1-4
RUNS 5-8
RUNS 2-5
RUNS 6-9
RUNS 3-6
RUNS 4-7
RUNS 7,8,9,
RUNS 8,9,1,2
RUNS 9,1,2,3
95.6%
105.1%
101.8%
93.2%
96.1%
87.6%
92.9%
98.3%
91.0%
94.1%
94.5%
98.9%
94.5%
94.4%
99.1%
94.6%
98.4%
94.2%
94.8%
94.7%
94.6%
94.5%
94.2%
93.7%
94.2%
92.5%
94.5%
93.9%
92.4%
94.1%
2L
94.1%
\ %#*&"># *£'
"^"oXnf-
94.4%
94.0%
93.6%
'/.
97.5%
94.1%
ERROR BAND FOR FOUR-RUN MEANS
SUCESSIVE
FOUR-RUN
SERIES
1THRU4
5THRU8
2THRU5
6THRU9
3THRU6
4THRU7
7A9,1
8,9,1,2
9,1,2,3
AVERAGE
U.L.95%C.I.
L.L.95%CI.
ERROR BAND
MEANCE.
CMIMASS
BALANCE
98.9%
93.7%
99.1%
92.4%
94.7%
92.5%
94.4%
97.5%
98.4%
95.7%
97.8%
93.6%
2.19%
MEANCE.
GAS/GAS
TTE
94.5%
94.2%
94.6%
94.1%
94.6%
94.5%
94.0%
94.1%
94.2%
94.3%
94.5%
94.1%
0.19%
BLOCK DATA LEGEND
98.9%
94.5%
4-RUN MEAN FOR CMI TEST METHOD
4-RUN MEAN FOR TTE TEST METHOD
-------
TABLE 12
ERROR BAND ANALYSIS
TWO 2-HOUR RUNS
RUN
NUMBER
CMI MASS
BALANCE
GAS/GAS
TTE
TWO RUN TEST SERIES MEANS
RUNS 1-2
RUNS 3-4
RUNS 5-6
RUNS 7-8
RUNS 2-3
RUNS 4-6
RUNS 6-7
RUNS 8-9
RUNS 9 & 1
92.8%
111.4%
98.6%
95.0%
97.4%
83.1%
93.2%
101.4%
92.4%
94.1%
94.9%
102.1%
94.5%
92.6%
94.3%
94.7%
96.8%
94.5%
105.0%
94.6%
94.3%
93.5%
90.3%
93.9%
96.2%
94.5%
93.9%
,'", '//•„?
-------
TABLE 13
ERROR BAND ANALYSIS
THREE 2-HOUR RUNS
RUN
NUMBER
CMIMASS
BALANCE
GAS/GAS
TTE
THREE RUN TEST SERIES MEANS
RUNS 1-3
RUNS 4-6
RUNS 6-9
RUNS 2-4
RUNS 5-7
RUNS 3-5
RUNS 6-8
RUNS 8,9,1
RUNS 9,1,2
92.8%
111.4%
98.6%
95.0%
97.4%
83.1%
93.2%
101.4%
92.4%
94.1%
94.9%
100.9%
94.4%
94.3%
101.7%
94.6%
98.9%
94.2%
94.7%
94.3%
91.8%
94.2%
97.0%
94.4%
93.5%
91.2%
93.9%
92.6%
93.9%
94.4%
95.7%
94.0%
93.9%
93.7%
95.5%
94.1%
ERROR BAND FOR THREE-RUN MEANS
SUCESSrVE
THREE-RUN
SERIES
1THRU3
4THRU6
6THRU9
2THRU4
5THRU7
3THRU5
6THRU8
8,9,1
9,1,2
AVERAGE
U.L. 95% CI.
L.L.95%CL
ERROR BAND
MEANCE.
CMIMASS
BALANCE
100.9%
91.8%
95.7%
101.7%
91.2%
97.0%
92.6%
95.5%
98.9%
96.1%
99.1%
93.2%
3.07%
MEAN C.E.
GAS/GAS
TTE
94.4%
94.2%
94.0%
94.6%
93.9%
94.4%
93.9%
94.1%
94.2%
94.2%
94.4%
94.0%
0.21%
BLOCK DATA LEGEND
100.9%
94.4%
3-RUN MEAN FOR CMI TEST METHOD
3-RUN MEAN FOR TTE TEST METHOD
-------
TABLE 14
ERROR BAND ANALYSIS
FOUR 2-HOUR RUNS
RUN
NUMBER
1
2
3
4
5
6
7
8
9
CMI MASS
BALANCE
92.8%
111.4%
98.6%
95.0%
97.4%
83.1%
93.2%
101.4%
92.4%
GAS/GAS
TTE
94.1%
94.9%
94.3%
94.7%
94.3%
93.5%
93.9%
94.4%
93.7%
FOUR RUN TEST SERIES MEANS
RUNS 1-4
99.5%
94.5%
-"tp^i;':<
&r,Vc <^;V
/•&"• " * ?/''?~:
<&''*-' *'<%?•
"$'•£»•' ' ,'/, \
RUNS 5-8
,w ''**$$&
''$&4$&$
-/, '-^/:'"t
93.8%
93.9%
s*- '*-„''&'.
RUNS 2-5
^f^l^l.
100.6%
94.6%
n',j$%,, >-f.
$&%(&%'%
:,- *ff^'*
*< 'xjpr1^
RUNS 6-9
^,«^^
££&.% 38?
4^1?f/i;
^f'-'x?^'
^^..^. . <;..'.. .-^^>> .
92.5%
93.9%
RUNS 3-6
&i&»*<"J%'
*&:$&%&*.
93.5%
94.4%
%&£$&>&"
&'/"$%$%>
;••-' '\>& .
RUNS 4-7
?*y#$M$'*%&'i'' ''.
RUNS 7,8,9,
J%,';'C.A^
y?.*9 ;<&*;"
y\ v ;?,
< -^'/^ -'(^5
/-^ ;- /;;, -/
: ""^"^^r^.
i", ^^'#,-
',/.* xH?4"''
99.5%
94.3%
RUNS 9,1,2^
98.8%
94.3%
K*ijE;/ ;-^v^;
'.^r^*r^7
"%',^^V~/i-
? ^' ^s >^- ^
/nS?*i«.^ 'y^V* **
^:^.':!.^1.,'
ERROR BAND FOR FOUR-RUN MEANS
SUCESSIVE
FOUR-RUN
SERIES
1THRU4
5THRU8
2THRU5
6THRU9
3THRU6
4THRU7
7,8,9,1
8,9,1,2
9,1^
AVERAGE
U.L.95%CI.
L.L. 95% CI.
ERROR BAND
MEANCE
CMI MASS
BALANCE
99.5%
93.8%
100.6%
92.5%
93.5%
92.2%
95.0%
99.5%
98.8%
96.1%
98.7%
93.5%
2.71%
MEANCE.
GAS/GAS
TTE
94.5%
93.9%
94.6%
93.9%
94.4%
94.2%
94.0%
94.3%
94.3%
94.2%
94.4%
94.0%
0.20%
BLOCK DATA LEGEND
99.5%
94.5%
4-RUN MEAN FOR CMI TEST METHOD
4-RUN MEAN FOR TTE TEST METHOD
-------
TABLE 15
SUMMARY OF CAPTURE EFFICIENCY RESULTS
PARAMETRIC STUDY
RUN
NUMBER
10
11
12
[L]
LIQUID PHASE
TOTAL VOC
(Ibs C3H8)
57.80
45.15
105.31
[G]
GAS PHASE
CAPTURED VOC
(Ibs C3H8)
54.39
49.25
125.51
[F]
GAS PHASE
FUGITIVE VOC
(Ibs C3H8)
2.39
2.42
4.37
AVERAGE CAPTURE EFFICIENCY (RUNS 10 - 12)
CAPTURE EFFICIENY TEST RESULTS
TTE TEST METHODS
GAS/GAS
[ G/CG+F) 1
95.8%
95.3%
96.6%
95.9%
LIQUID/GAS
f (L-F)/L) ]
95.9%
94.6%
95.9%
95.5%
LIQUID/GAS
MASS BALANCE
fG/L]
94.1%
109.1%
119.2%
107.5%
-------
TABLE 16
COMPARISON OF CAPTURE HOOD PERFORMANCE
TEST RUN
NUMBER
BASELINE
1
2
3
4
5
6
7
8
9
10
11
12
CAPTURED EMISSIONS (Ibs C3H8)
OVEN EXHAUST
94.77
91.79
100.20
111.76
113.80
106.33
106.72
95.89
119.47
105.97
46.47
41.08
111.5
CAPTURE HOODS
25.36
26.31
22.08
25.19
21.62
23.96
24.59
24.84
24.81
24.61
7.92
8.17
14.01
TOTAL
120.13
118.10
122.28
136.95
135.42
130.29
131.31
120.73
144.28
130.58
54.39
49.25
125.51
CAPTURE HOODS
PERCENT OF TOTAL
21.1%
22.3%
18.1%
18.4%
16.0%
18.4%
18.7%
20.6%
17.2%
18.8%
14.6%
16.6%
11.2%
-------
TABLE 17
COMPARISON OF CAPTURE HOOD PERFORMANCE
(NORMALIZED DATA)
TEST RUN
NUMBER
BASELINE
1
2
3
4
5
6
7
8
9
10
11
12
CAPTURE HOOD EMISSIONS (Ibs C3H8)
OVEN EXHAUST
ACTUAL
94.77
91.79
100.20
111.76
113.80
106.33
106.72
95.89
119.47
105.97
46.47
41.08
111.50
NORMALIZED
94.77
93.37
98.44
98.03
100.95
98.04
97.63
95.41
99.47
97.49
102.64
100.20
106.72
CAPTURE HOODS
ACTUAL
25.36
26.31
22.08
25.19
21.62
23.96
24.59
24.84
24.81
24.61
7.92
8.17
14.01
NORMALIZED
25.36
26.76
21.69
22.10
19.18
22.09
22.50
24.72
20.66
22.64
17.49
19.93
13.41
TOTAL
ACTUAL
120.13
118.10
122.28
136.95
135.42
130.29
131.31
120.73
144.28
130.58
54.39
49.25
125.51
NORMALIZED
120.13
120.13
120.13
120.13
120.13
120.13
120.13
120.13
120.13
120.13
120.13
120.13
120.13
NORMALIZED CAPTURE
HOOD VOC AS PERCENT
OF BASELINE RUN
100.0%
105.5%
85.5%
87.1%
75.6%
87.1%
88.7%
97.5%
81.5%
89.3%
69.0%
78.6%
52.9%
-------
TABLE 18
DISTRIBUTION OF VOC EMISSIONS
CAPTURED BY HOOD SYSTEM
TEST STATION
DUCTDIA.(in)
VELOCITY (fpm)
AVG. VELOCITY (fpm)
FLOW (acfm)
FLOWCscfm)
VOC CONC (ppm C3H8)
VOC EMISSIONS
(lbsC3H8/hr)
% of TOTAL HOOD
CAPTURED VOC EMISSIONS
RWS
TROUGH
HOOD
1
6
3290
2904
2860
3520
3144
617
607
95
0.40
3.9%
ROLL
COATER
UPPER
HOOD
2
6
4150
4920
3670
5370
4528
889
874
337
2.02
19.9%
COATER
EXIT
HOOD
3
6
4200
4150
4350
3940
4160
817
803
1264
6.96
68.4%
SUM OF
BUCKET
FLOOR
SWEEPS
4
6
330
255
465
890
485
95
94
1245
0.80
7.8%
TOTAL HOOD CAPTURED VOC EMISSIONS: 10.17
RWS
BUCKET
FLOOR
SWEEP
5
3.25
"79
77
1684
0.89
7.7%
COATING
BUCKET
FLOOR
SWEEP
6
3.25
390
275
250
240
289
17
16
167
0.02
0.1%
* Recalculated from difference in individual flow
measurement(Sum of Floor Sweeps - Floor Sweep Coating
Bucket)
-------
FIGURES
-------
CURING
OVEN
I I 1 I I
I I I I
1 ' ' .I..1 ' '
COATING Of—1
ROLL " '
APPLICATOR
IIUUI1~
SHEET
FEEDER
i!
I*
i!
FIGURE 1
MULTIPLE VIEWS OF
TYPICAL SHEET-FED COATING LINE
fl
-------
i!
EXHAUST TO
THERMAL OXIOIZER
ON ROOF
OPENING
CAPTURE
HOOD
COATING ROLL
APPLICATOR
SYSTEM
ACCESS DOOR TO
WICKET DRIVE
OATINO DRUM
DATING FEED PIPE
aOOR SWEEP
DATING FEED BUCKET
CHILL ROLLER WASH
SOLVENT BUCKET
FIGURE 2
3D SCHEMATIC OF TYPICAL
SHEET-FED COATtNQ LINE
-------
PTURE HOOO
EXHAUST OUCT
11
NO SCALE
NO SCALE
FIGURE 3
MULTIPLE VIEWS OF TTE ON
SHEET-FED COATINQ LINE
-------
18" X 2V NATURAL
DRAFT OPENINGS (TTP)
TT£ SIOCS CONSTRUCTED
Of CAS080A«0
PANELING
•y x ?• ACCESS
OOOR (TYP)
i!
FIGURE 4
MAJOR COMPONENTS OF TTE
ON SHEET-FED COATINQ LINE
-------
l!
UOUIO-PHASE VQC INPUT [L]
MEASURED (SEE FIGURE 6)
_ S-PHASE CAPTURE
EMISSION [G]
MEASURED (SEE FIGURE 7}
CAS-PKASE
FUGITtvt EMISSIONS
IF] MEASURED
ROW & VOC
CONCENTRATION
d
FIGURE 6
OVERVIEW OF C.E.
TEST STUDY
MEASUREMENT LOCATIONS
-------
i!
li!
6
SAMPLES OBTAINED FROM
ROLL WASH SOLVEOT
FEED BUCKET
PORTABLE SCA1E BELOW
UOUIO VOC RESERVOIRS
FIGURE 6
MEASUREMENT OF
LIQUID VOC INPUT
-------
vl on I uc* I erf I
RECUPRATIVE
THERMAL
OXIDI2ER
I 1
FLOW, VOC
CONCENTRATION *
METHANE/ETHANE
MEASURED
FLOW & VOC-
CONCENTRATION
MEASURED
OVEN
EXHAUST
DUCT
VLX
CAPTURE
HOOD
EXHAUST
DUCT
tfffxa.
o
FIGURE 7
MEASUREMENT OF
CAPTURED VOC
-------
MCVJ L JK* I
INTAKE •
DUCT
-EXHAUST STACK
BACKGROUND —
VOC MEASUREMENT
POINT
k\ INTAKE /
\FAN/ I
x^ X.X
INTAKE FAN
PLATFORM
PLAN VIEW
ELEVATION
APPRO*.
0
FIGURE 8
TEST
STATION
TS-4
STACK
DIAMETER
36-
EQUIVELANT OIA. TO
DISTURBANCE
UPSTREAM
1.0
DOWNSTREAM
4.7
COOLING ZONE TEST
-------
NOO
NUMBER
.N-1
N-2
N-3
N-4
N-5
N-6
N-7
N-8
N-9
N-10
N-11
N-12
N-13
N-14
DIMENSION
(INCHES)
18 X 24
24 X 18
18 X 24
18 X 24
24 X 18
22 X 20
9 X 46
18 X 24
54 X 9
24 X 18
18 X 24
48 X 1
36 X 1.5
48 X 0.5
EQUIVELANT
NUMBER
(FEET)
1.95
1.95
1.95
1.95
.95
.97
.91
.95
2.07
.95
.95
0.65
0.69
0.46
DISTANCE TO VOM SOURCE
FEET
8.58
9.17
9.25
9.33
9.58
8.67
8.25
8.92
8.83
9.17
9.25
3.50
11.92
8.83
NUMBER OF EQUIVALENT
DIAMETERS
4.40
4.70
4.74
4.78
4.91
4.40
4.32
4.57
4.27
4.70
4.74
• 5.38
17.28
19.20
VENTILATION
INTAKE
NUMBER
V-1
V-2
V-3
V-4
V-5
V-6
V-7
DIMENSION
(INCHES)
36 X 36
36 X 36
48 X 1
48 X 1
6 OIA
24 X 48
48 X 1
EQUIVELANT
NUMBER
(FEET)
3.38
3.38
0.65
0.65
0.5
3.19
0.65
DISTANCE TO NOO
FEET
24.00
23.73
8.61
4.25
4.35
21.20
7.45
NUMBER OF EQUIVALENT
DIAMETERS
7.10
7.02
13.25
6.53
8.70
6.64
11.46
FIGURE 9
METHOD 30 TTE COMPLIANCE SUMMARY
RUNS 1-9
-------
NOO
NUMBER
N-1
N-2
N-3
N-4
N-5
N-6
N-7
N-8
N-9
N-10
N-11
N-12
N-13
N-14
N-15
N-16
N-17
N-18
DIMENSION
(INCHES)
18 X 24
18 X 24
18 X 24
18 X 24
18 X 24
20 X 22
9 X 46
18 X 24
54 X 9
18 X 24
18 X 24
48 X 1
36 X 1.5
48 X 0.5
18 X 24
12 X 36
15 X 18
I5.5 X 18
EOUIVELANT
NUMBER
(FEET)
.95
.95
.95
.95
.95
.97
1.91
1.95
2.07
1.95
1.95
0.65
0.69
0.46
1.95
1.95
1.55
1.57
DISTANCE TO VOM SOURCE
FEET
8.58
9.17
9.25
9.33
9.58
8.67
8.25
8.92
8.83
9.17
9.25
3.50
11.92
8.83
8.50
8.00
9.83
9.75
NUMBER OF EQUIVALENT
DIAMETERS
4,40
4.70
4.74
4.78
4.91
4.40
4.32
4.57
4.27
4.70
4.74
.5,38
17.28
19.20
4.36
4.10
6.34
6.21
VENTILATION
INTAKE
NUMBER
V-1
V-2
V-3
V-4
V-5
V-6
V-7
DIMENSION
(INCHES)
36 X 36
36 X 36
48 X 1
48 X 1
6 OIA
24 X 48
48 X 1
EQUIVELANT
NUMBER
(FEET)
3.38
3.38
0.65
0.65
0.5
3.19
0.65
DISTANCE TO NDO
FEET
24.00
23.73
8.61
4.25
4.35
21.20
7.45
NUMBER OF EQUIVALENT
DIAMETERS
7.10
7.02
13.25
6.53
8.70
r 6.64
11.46
FIGURE 10
METHOD 30 TTE COMPLIANCE SUMMARY
PARAMETRIC STUDY
RUNS 10-12
-------
i!
L
I
ll
a!
L
li
FIGURE 11
MEASUREMENT LOCATIONS
FOR ASSESSMENT OF
CAPTURE HOODS (RUN 13)
-------
MRjjtrr.
93184GV
1 wcnr 1 GP
1MTD I^U
6/14/93 I
HC9
IOttXT MHD
Clfl
140
9 10
RUN NUMBER
11
12
3-IN. FROM SLOT 6-IN. FROM SLOT
FIGURE 12
SUMMARY OF RESULTS
AIR VELOCITY SURVEY
ROLL COATER UPPER HOOD
FRONT-FACE
-------
.184C\
F^r i
I =~., I -CTT-T- I
100
9 10
RUN NUMBER
11
12
3-IN. FROM SLOT
6-IN. FROM SLOT
FIGURE 13
SUMMARY OF RESULTS
AIR VELOCITY SURVEY
ROLL COATER UPPER HOOD
REAR-FACE
-------
02184GV
IMWH Icxnxa IMTO ln_c HMD lama M
MOW I GFV I e/u/gg I noio I cm
lOOn
BL
9 10
RUN NUMBER
11
3-IN. FROM SLOT I
6-IN. FROM SLOT
FIGURE 14
SUMMARY OF RESULTS
AIR VELOCITY SURVEY
ROLL COATER UPPER HOOD
BOTTOM SLOT
-------
p1
wj I
6/14/Wj I KIO10 I CMJ
160-f
140
9 10
RUN NUMBER
11
3-IN. FROM SLOT
6-IN. FROM SLOT
12
FIGURE 15
SUMMARY OF RESULTS
AIR VELOCITY SURVEY
ROLL COATER EXIT HOOD
-------
92ia4GVl I MC1T I QPV I 6/14/93 I FIOIO I CM
1000
900"
800-
700-
600-
I
2 3 4 5 6 7 8 9 10 11 12
500-
400-
300-
200-1 r
BL 1
FIGURE 16
SUMMARY OF RESULTS
AIR VELOCITY SURVEY
OVEN ENTRANCE
FACE-OPENING
-------
BHI I mnr I IB Itg/rgyreI rmn ifln
APPROX. ELEVATION
OF AIR SAMPLES
SYMBOL LEGEND
APPROX. LOCATION
OF AIR SAMPLE
FIGURE 17
TIE WORK-PLACE MONITORING LOCATIONS
CROWN, CORK AND SEAL
PHILADELPHIA, PENNSYLVANIA
-------
i!
L
COATING
UN£
i!
L
FIGURE 18
TYPICAL ANCILLARY EQUIPMENT
SHEET-FED COATING LINE
-------
NO SCALE
FIGURE 19
FUNCTIONAL TTE INSTALLED ON CLUSTERED
SHEET-FED COATING LINES
COATING LINE
-------
APPENDIX A
CMI TEST PROTOCOL
-------
CAN MANUFACTURERS INSTITUTE'S
CAPTURE EFFICIENCY TEST PROTOCOL
1.0 INTRODUCTION
1.1 Applicability
This procedure is applicable to sheet fed metal coating operations such as those
conducted in a lithography department of a can manufacturing facility. It was
developed through an extensive comparative study of a liquid/gas mass balance and
total temporary enclosure test methods. Most of the test protocol is derived from the
original Test Plan written for this comparative study conducted by the Can
Manufacturers Institute. The procedure incorporates a liquid VOC input measurement
that relies on instantaneous weights of coating and solvent feed reservoirs and captured
VOC emission measurements at the inlet to the coating line's VOC control device. Most
of the procedures are patterned after the EPA Method 30 test series for surface coating
capture efficiency.
1.2 Principal
The capture efficiency measured by this protocol is calculated by dividing the total
captured VOC emissions as a propane equivalent (G) by the total liquid VOC input
which is converted to a propane equivalent (L). The amount of VOC captured is
calculated as the sum of the products of the VOC concentration (C), the gas flow rate
(Q), and the sampling time (t) from the exhaust duct(s) that feed the VOC control device.
Other procedures are also included to quantify the amount of background VOC and
unburned natural gas from the curing oven which comprise part of the FIA response in
the captured emission stream. These contributions are subtracted from C to quantify the
total captured emissions (G).
The total amount of liquid VOC introduced to the process is calculated as the difference
between the total VOC (as a propane equivalent) at the start and finish of the defined
test period. These liquid VOC quantities are calculated as the sum of the products of
the instantaneous weight (W) of each VOC containing liquid (the coating and roll wash
solvent) and its corresponding VOC content (V). This product is then divided by the
response factor (RF) measured via EPA Method 30F (with several enhancements). The
Method 30F testing involves distilling a sample of the VOC liquid to separate the solvent
fraction. A precise weight of the distillate is vaporized and used to prepare known gas
concentration samples, which are measured with the same flame ionization analyzer
(FIA) used in the field to measure captured VOC emissions.
-------
1.3 Testing Requirements
The capture efficiency test will consist of three replicate sampling runs. Each sampling
run will be conducted over a minimum production period of 2-hours. This test
configuration was developed by a statistical analysis of confidence intervals during the
CMI test method study. A three run, 2-hour test method configuration was shown to
have an error band of +/- 3.1%.
The testing should be conducted over a time period in which the operation is as uniform
as possible so that the test runs are truly replicate analyses and are not reflecting
variability caused by changes in operating parameters. The same product and coating
material should be used for each test run. Other quality assurance checks (such as film
weight thickness) on the production control should be carried out in accordance with
normal operating practices to maintain steady VOC consumption.
-------
2.0 SUMMARY OF CAPTURE EFFICIENCY METHOD
2.1 Pretest Procedures
2.1.1 Oven Balance
The curing oven will be balanced prior to performing compliance testing on the
production line. A balanced oven is one that has a negative pressure differential, higher
pressure outside the oven than inside, and thus generating net air flow into the oven.
The pre-test oven balance will be performed one to four weeks in advance of
performance testing. Once set, major changes in the system will not occur.
2.12 Coating Line Inspection
The coating line equipment, especially the coater support equipment, will be inspected.
The pumps, piping, hoses and all accessories "will be checked for leaks, and any
problems corrected. The knife blade scraper on the chill roller will be removed,
inspected, and replaced, if necessary. Duct work in the vicinity of all test port locations
will be inspected and, if necessary, cleaned so that duct shape and diameter are known
and consistent throughout a sufficient distance. At a minimum, ductwork will be
inspected, and if necessary, cleaned to distances of 3 diameters upstream and 1 diameter
downstream of each test station.
2.2 Baseline Testing
Accurate measurement of the VOC concentration in the captured emission stream is
essential in determining capture efficiency by this method. During the normal operation
of a coating line, background VOC from the plant atmosphere and unburned natural gas
will contribute to the overall FLA response measured at the control device inlet. These
contributions would tend to over estimate the amount of VOC captured by the line and
must be accounted for. The CMI protocol includes a baseline run that will be conducted
at the beginning of each day of testing. During this baseline run, the wicket curing oven
will be operating with the gas burners and exhaust fan running at normal capacity,
however, the liquid VOC reservoirs will be dry and no coated plate will be traveling
through the oven. Therefore, the only organic vapors in the captured emission stream
will be the background VOC pulled in from the plant atmosphere and the unburned
natural gas emanating from the oven. The FLA response measured during this baseline
run will be as a propane equivalent and can be directly subtracted from the average,
calibration-corrected, VOC concentration measured during the 2-hour run.
The baseline testing will consist of a 30-minute run. Testing will be conducted in
accordance with EPA Method SOB: Volatile Organic Compound Emissions in Captured
Stream. Method SOB includes the measurement of VOC concentrations using a flame
ionization analyzer (FIA). Method SOB also includes measurement of the volumetric
flow rate in accordance with EPA Methods 2 through 4. Since the baseline FLA response
will only be used to correct the average VOC concentration measured during the normal
test runs, this flow measurement is not necessary and will not be included as part of the
baseline run.
-------
2.3 Measurement of Liquid VOC Consumption
Liquid VOC input during the 2-hour test runs will be quantified by measuring the
weight of coating and roll wash solvent that are consumed during each test run. The
coating consumption is anticipated to represent more than 95% of the total VOC
consumed by the coating line. Coating is manually dispensed from a 55-gallon drum
into a feed bucket. A pump on the coating line withdraws coating from the bucket and
distributes it to the roller applicator. Coating is evenly distributed across a series of
rollers which apply it to the sheet at a uniform film weight. The application roller is
equipped with a scraper mechanism that removes excess coating into a catch pan which
then drains back into the feed bucket. After the line has been running for a short period
of time, the volume of coating within the feed system stabilizes and little change is
expected for the duration of the test. Therefore, the coating consumption will be
measured by instantaneously weighing the 55-gallon drum and feed bucket at the
beginning and end of the test run. These instantaneous weights, along with the roll
wash solvent feed bucket, will be measured with portable scales positioned at the roll
coater.
2.3.1 Emissions Testing Start-up
The test will begin upon the agreement of the test equipment operators and plant test
assistant. The test start time will be recorded by all parties. For correlation, the time of
significant events (such as pallet changes, sheet feed jams, and film weight checks)
during the tests will also be written directly on the FIA chart paper. Operating data will
be recorded on a periodic basis, including oven bake temperature and time, sheet speed
(feed rate), sheet size and sheet count. A downtime record will be kept which records
any reason for a line stoppage, the line's stop and restart times, and the corresponding
sheet count. During the test, production will be at a normal rate, and every effort will
be made to reduce downtime to an absolute minimum. A test run will be voided if the
production efficiency is less than a level that is 10% below the coating line's normal
efficiency when it is applying the same coating used for the CE test.
2.3~2 Measurement of VOC Consumption
The quantity of coating consumed will be determined by instantaneous weight
measurements at the beginning and end of each run. Portable scales will be placed
under the coating drum and feed bucket to obtain these weights. Both scales will be
certified prior to the test The coating line will be operated for at least 15 minutes prior
to starting a test run. This will allow the volume of coating within the delivery system
and the VOC concentration in the oven exhaust to stabilize. A sample of the "as-
applied" coating will be collected from the feed system just before the test begins.
Tare weights of the 55-gallon drum and feed bucket will be required to determine the
net weight of liquid VOC in the reservoirs. Coating will be transferred into a clean, pre-
weighted drum prior to the test. Coating feed bucket tare weights will also be measured
prior to the testing. The VOC content of the initial coating will be measured by EPA
Method 30F to enable calculation of the starting weight of VOC in the coating. A final
-------
weight reading for the coating drum and feed bucket will be made at the end of the test
run. A sample of the "as-applied" coating will be collected from the feed system
immediately after flie ending weight measurements are made. These measurements will
allow calculation of the ending weight of VOC in the coating. The same procedure will
be used to determine the starting and ending weight of VOC in the roll wash solvent.
The total VOC consumption will be computed as the difference between the sum of the
starting combined weight of VOC in the coating and roE wash solvent and the sum of
the ending combined weight of VOC in the coating and spent roll wash solvent.
23.3 Measurement of VOC Content
The VOC content of the coating and roll wash solvent samples will be measured in
accordance with EPA Method 30F. An independent analytical laboratory will perform
this analysis, which includes distillation of the coating solvent constituents, evaporation
of a precise quantity of the solvent blend into a known volume of zero air, and
measurement of the gas VOC content using a flame ionization analyzer. This data will
enable calculation of the VOC content as a propane equivalent and direct comparison
with the gaseous emissions measured by Method SOB. Details on the Method 30F
analysis are presented in Section 3.1.5 of this protocol.
2.4 Measurement of Gaseous Phase VOC Emissions
2.4.1 Captured VOC Emissions
Captured VOC emissions during the full test runs will be measured in accordance with
EPA Method 30B: Volatile Organic Compound Emissions in Captured Stream. The
following measurements will be made at each appropriate test station in the control
device inlet ductwork: (1) VOC concentration, (2) average velocity pressure, (3) average
gas temperature, (4) O2/CO2 concentrations, and (5) gas moisture content. Schematic
diagrams of the sampling trains and test equipment used for the gaseous phase testing
are presented on Figures 1 through 4. A list of the equipment and detailed procedures
for this testing is presented in Section 3.2 and 3.3 of this protocol.
2.4.2 Calculation of Capture Efficiency
The measurements taken during each 2-hour test run will aEow calculation of a separate
capture efficiency for each of the three replicate analyses:
CE = (G / L) x 100%
Where: CE = Coating Line Capture Efficiency (%)
G = Gaseous captured VOC emissions (Ibs/propane)
L = Total liquid VOC consumption (Ibs/propane)
Detailed procedures for calculating all test parameters are presented in Section 4.0 of this
protocol.
-------
3.0 MEASUREMENT PROCEDURES
3.1 Liquid VOC Consumption
3.1.1 Test Preparation
Production staff and testing personnel will complete the following steps prior to
beginning the test
• Tare weights will be measured for dean 55-gallon drums, the coating feed
bucket, and roll wash solvent bucket.
• Transfer several 55-gallon drums (at least three) of coating into pre-
weighed, clean 55-gallon drums.
• Portable scales will be placed adjacent to the coater for measuring
instantaneous weights of the coating drum, coating feed bucket, and roll
wash solvent bucket. Each of the scales will be calibrated and certified by
the equipment supplier. The coating drum scale should be capable of
weighing up to 1,000 pounds to an accuracy of +/- 0.2 pounds. The two
feed bucket scales should be capable of weighing up to 200 pounds to an
accuracy of +\- 0.04 pounds.
3.1.2 Coating Line and Test Start-up
The line operator and testing crew will complete the following steps to prepare the
coating line for testing:
1. A new bucket of roll wash solvent will be delivered to the line.
2. A film weight sample sheet will be fed through the coating line. The
weight of coating solids deposited per unit area will be measured, and the
coating feed adjusted until this film weight is within the acceptable
tolerance range.
3. Coating will be fed into the coating reservoirs and the line will be run for
at least 15 minutes prior to the start of the test run.
4. A representative sample of the coating and roll wash solvent will be
collected from the coating drum and both feed buckets just before the test
begins in accordance with procedures listed in Section 3.1.4 of this
protocol.
5. The line sheet counters will be set to zero at the exact time the test run
begins.
-------
6. The weight of the coating drum, coating feed bucket, and roll wash solvent
feed bucket will be measured with portable scales at the exact time that the
two-hour test run begins.
7. The following operating data will be collected for each the test period:
sheet count, sheet size, oven bake temperature, oven bake time, and a log
of all line down time.
3.1.3 Post Test Measurements
The following measurements and samples will be taken at the end of each two-hour test
run:
1. The weight of the coating drum, coating feed bucket, and roll wash solvent
feed bucket will be measured at the exact time that the two-hour test run
ends.
2. The sheet count will be recorded at the exact time that the test run ends.
3. A representative sample of the coating and spent roll wash solvent will be
collected from the coating drum and both feed buckets just after the weight
measurements.
3.1.4 Liquid Feed Sampling
Representative samples will be collected from the coating drum, coating feed bucket, and
roll wash solvent feed bucket at the beginning and end of each test run. A new drum
of coating will be used for each 2-hour test run. Prior to collecting the initial sample,
the coating drum and feed bucket will be weighed to determine the percentage of each
sample that needs to be combined into a sample representative of the coating in the
reservoir system at the start of the test. For example, if the drum net weight is ten times
higher than the feed bucket net weight, the representative sample would consist of 90%
volume from the drum and 10% volume from the feed bucket.
The coating and roll wash solvent samples will be collected using a glass drum pipette.
This sampler will extract a continuous column of liquid from the top to the bottom of
the drum or feed bucket, therefore, accounting for any stratification or non-uniform
dispersion of coating solids. The roll wash solvent bucket will be mixed prior to
sampling to uniformly suspend coating solids that have become entrained in the wash
solvent. Two, 100 ml glass jars with teflon lined lids will be filled with each coating
sample. One jar will be the primary sample to be analyzed by the laboratory and the
second jar will be a split sample that is retained for possible duplicate analysis. The jars
will be completely filled with no headspace to eliminate any volatilization of the carrier
solvent prior to analysis.
-------
3.1.5 Analysis for VOC Content
The VOC content of each of the primary coating and roll wash solvent samples collected
during this program will be analyzed in accordance with EPA Method 30F-Volatile
Organic Carbon Content in Liquid Input Stream "Distillation Approach". This method
determines the VOC content of a liquid material in accordance with EPA Method 24,
then measures a response factor to convert this VOC concentration to a propane
equivalent. A sample of the coating is distilled to separate the VOC fraction. The
distillate is used to prepare a known gas standard for analysis with the same flame
ionization analyzer (FIA) used in the field to measure the captured emissions VOC
concentration. The FIA is calibrated against propane to determine the response factor.
This response factor is used to convert the weight of VOC to an equivalent weight of
propane. The procedures to be used are as follows:
1. Distillation of VOC - A vacuum rotary evaporator equipped with a folded
inner coil, a vertical style condenser, rotary speed control, and teflon sweep
gas delivery tube with a valved inlet as specified in the method is to be
assembled. A schematic diagram of the distillation apparatus is included
as Figure 5. Prior to running samples each day, the rotary evaporation
system should be leak checked by aspirating a vacuum of approximately
20 mm Hg from absolute. The system should be closed and monitored for
one minute. A vacuum fall of greater than 125 mm Hg in one minute will
indicate a failed leak check. Leaks should be repaired and the leak check
repeated until successful. The oil bath temperature will be verified at
room temperature before beginning a distillation.
Solids within the liquid will be uniformly dispersed by placing the jar in
a mechanical shaker to agitate the sample without releasing any VOC.
Approximately 20 ml of sample will be transferred using a wide bore
graduated pipette into the rotary evaporator distillation vessel.
Throughout the entire distillation, the evaporation flask will be rotated at
a constant rate and the condenser temperature held at -10° C. A vacuum
will be gradually applied to the evaporation unit to within 20 mm Hg of
absolute. While maintaining this vacuum, the distillation vessel should be
heated at the rate of 2 to 3° C per minute until the optimum temperature
is achieved. The optimum oil bath temperature will be determined prior
to sample analysis (see modification #2). Once the sample reaches dryness,
it should be held constant at this vacuum and temperature for an
additional ten minutes. At this point, a nitrogen sweep gas will be slowly
introduced through the purge tube into the distillation flask while carefully
preserving the 125 mm Hg vacuum. This sweeping should be continued
for ten minutes to remove remaining solvent from the distillation flask and
condenser assembly. After sweeping is completed, the apparatus will be
disassembled, and the distillate transferred to a labeled sample vial with
a valved cap. All samples will then be placed in storage at 4° C pending
preparation of the VOC standard bag sample.
8
-------
2. Preparation and Analysis of Bag Samples - The distilled solvent sample to
be analyzed will be removed from the storage refrigerator and the solvent
temperature allowed to equilibrate. A 30 ul (nominal) sample of the
solvent will then be taken up in a 100 ul glass syringe. After weighing, the
solvent is to be injected into the volatilization vessel and the syringe
quickly reweighed. Additional analyses will be performed with 60 ul and
90 ul of distillate (see modification #9).
A schematic diagram of the VOC bag sample apparatus is included as
Figure 6. Zero air will be passed through the system into a Tedlar sample
bag, at a rate of 2 L/minute (nominal) for ten minutes. The flow of air
will be begun before solvent injection and pre- and post-test meter
readings taken for each run.
Immediately after preparation, the bag sample will be analyzed with the
same FLA used in the field. Steady readings will be obtained for at least
30 seconds and recorded. The FIA will be calibrated at least once every
four hours using zero air and high, mid, and low level span gases. Certain
samples will be prepared and analyzed in duplicate as a quality control
measure.
Based on experienced gained during the CMI comparative study of CE test methods, the
following modifications and quality assurance checks are incorporated into Method 30F
for this protocol:
1. The oven bake temperature used in determining percent VOCs should be
set at the same temperature that the curing oven was run at during the test
procedure.
2. The oil bath temperature for the rotary evaporator should be set at a level
high enough to ensure all coating VOC fully evaporates. The optimum
temperature should be selected based on trial runs with the coating
material that include weighing the residual remaining in the evaporation
flask. This will enable calculation of a percent volatile that can be
compared to the Method 24 results.
3. A leak check of the VOC bag sample generation systems should be
conducted at least daily or each time the apparatus is disassembled,
cleaned, or modified. The delivery end of the apparatus should be sealed
and a moderate positive pressure (approximately 10 inches w.c.) put on the
system. All joints, connections, and valves should be checked with soap
solution for leaks.
4. A zero check should be performed on the VOC bag sample generation
apparatus prior to beginning the analysis and after each ten samples run
through the apparatus. This will prevent sample bias and cross-
contamination.
-------
5. A hexane standard should be run through the VOC bag sample generation
apparatus at the same frequency as the zero checks. The hexane standard
should generate a response factor of exactly 1.0 and is an excellent quality
assurance check on the performance of the system.
6. The gas meter used in the VOC bag sample apparatus should be calibrated
prior to the test runs following procedures specified in USEPA Method 5,
Section 5.3. The meter correction factor, Y, should be used in equation
30F-1 when calculating the bag sample volume.
7. The vial used to store the distilled solvent should be a glass, septum
capped vial and should be completely filled without headspace. If bag
samples are not immediately prepared, the vials should be refrigerated at
4° C. If a distillate sample is refrigerated, it should be allowed to
equilibrate to room temperature before a syringe is prepared and weighed.
8. An injection volume larger than 10 ul will likely be desirable for most
applications, to more closely approximate the captured VOC concentration
and to maximize the mass weighed (and, therefore, reduce the effects of
weighing errors). Accordingly, a 100 ul syringe will be used in lieu of the
10 ul syringe specified in Section 2.2.7 of Method 30F.
9. The response factor will not be based on the result of the preparation of a
single bag sample. Duplicate bag samples will be prepared at three
different concentrations by varying the amount of solvent extract injected
into the zero air gas stream. If these duplicate analyses do not agree
within +/- 5%, the measurement will be repeated. The individual runs
will be plotted to generate a calibration curve of VOC concentration
(mg/m3 of VOC) versus FIA response (mg/m3 of propane). The linearity
of tine calibration curve will be assessed by a regression analysis and the
test will be repeated if the correlation coefficient is 0.90 or less. The
response factor will then be defined as the slope of the linear regression
line.
Coating and roll wash solvent instantaneous weights will be converted to propane
equivalents by multiplying the net weight (W) by the ratio of VOC content (V) divided
by response factor (RF). Example calculations for computing the total liquid VOC
consumption are presented in Section 4.1.
3.2 Measurement of VOC Gas Concentrations
33.1 Test Method Overview
Captured VOC emissions will be measured in accordance with EPA Method SOB -
Volatile Organic Compound Emissions in Captured Stream. This test method will be
used for quantifying the average VOC concentration and volumetric flow rate of the
entire volume of gas directed into the VOC control device. The average VOC
10
-------
concentration will be adjusted for background VOC and unburned natural gas
contributions measured during the baseline run. The amount of VOC captured (G) is
calculated as the sum of the products of the VOC content (Q;), the flow rate (Qc), and
the sample time (t) from each captured emissions measurement point.
3.2.2 Apparatus and Reagents
A schematic of the measurement system is shown in Figure 1. The main components
are listed below:
• Sample Probe - Stainless steel probe that is heated to prevent VOC
condensation.
• Sample Line - Teflon tubing transports the sample gas to the analyzer. The
sample line is heated to prevent condensation.
• Sample Pump - A leak-free pump inside the FIA pulls the sample gas
through the system at a flow rate sufficient to minimize the response time
of the measurement system. The components of the pump that contact the
gas stream are constructed of stainless steel and the pump is heated to
prevent VOC condensation.
• Sample Flow Rate Control - A sample flow rate control valve and back
pressure gauge are installed after the sample pump to maintain a constant
sampling rate within 10 percent. The flow rate control valve and rotameter
are heated to prevent condensation.
• Organic Concentration Analyzer - An FIA with a span value of 1.5 times
the expected concentration as propane must be used. The analyzer must
be capable of meeting or exceeding the following specifications:
Zero Drift - Less than ± 3.0 percent of the span value.
Calibration Drift - Less than ± 3.0 percent of the span value.
Calibration Error - Less than + 5.0 percent of the calibration gas
value.
Response Time - Less than 30 seconds.
• Integrator/Data Acquisition System - An analog strip chart recorder
and/or computerized data acquisition system will be used to integrate the
FIA response, compute the average response, and record measurement
data. The minimum data sampling frequency for computing average or
integrated values of measurement value every 5 seconds will be met or
exceeded.
11
-------
• Calibration and Other Gases - Gases used for calibration, fuel, or
combustion air (if required) are contained in compressed gas cylinders. All
calibration gases will be traceable to National Institute of Standards and
Technology standards and shall be certified by the manufacturer to ± 1%
of the tag value.
• Fuel - A 60% helium, 40% hydrogen gas mixture will be used.
• Combustion Air (if required) - High purity air with less than 1 ppm of
organic material (as propane or carbon equivalent) or less than 0.1 percent
of the span value will be used.
• FIA Linearity Calibration Gases - Low-, mid-, and high-range gas mixture
standards will be used with nominal propane concentrations of 20-30, 45-
55, and 80-90 percent of the span value in air, respectively.
• Particulate Filter - An out-of-stack metal filter inside the FIA will be used
to filter particulates. The out-of-stack filter is heated to prevent any VOC
condensation.
3.2.3 Measurement of Captured Emissions VOC Concentration
The VOC response of the stack gas in each captured emission stream test station(s) will
be measured by the following steps:
1. A sample train will be assembled as shown in Figure 1. The FIA will be
calibrated according to the procedure outlined in Section 3.2.4.
2. A system check will be conducted according to the procedure outlined in
Section 3.2.4.
3. Zero Gas will be injected at the end of the sample probe allowing the
measurement system response to reach zero. The system response time
will be measured as the time required for the system to reach the effluent
concentration after the calibration valve has been returned to the effluent
sampling position.
4. A system check will be conducted before and a system drift check after
each sampling run according to the procedures in Section 3.3.4. If the drift
check following a run indicates unacceptable performance, the run is not
valid and will be repeated. System drift checks may be performed during
the run not to exceed one drift check per hour.
5. The sample lines, filter, and pump temperatures will be heated to prevent
condensation.
12
-------
6. The sample probe will be installed so that the probe is centrally located in
the duct at each test station and will be sealed tightly at the test port
connection.
7. Sampling will begin at the start of the test period and will continue during
the entire run. Starting and ending times will be recorded as well as any
required process information. Measurements will be disregarded at each
sampling location until two times the response time of the measurement
system has elapsed. Sampling will be continuous unless a system drift
check is performed.
EPA Protocol No. 1, Calibration Gases, will be used for the linearity checks with the
following range of concentrations used for the low level, mid level, high level
calibrations:
• Low Level - 25% to 30% span value
• Mid Level - 45% to 55% span value
• High Level - 80% to 90% span value
3.2.4 Calibration and Quality Assurance
FIA Calibration and Linearity Check - The necessary adjustments to the air and fuel
supplies for the FIA will be made and the burner ignited. The FIA will be allowed to
warm up for the period recommended by the manufacturer. A calibration gas will be
injected into the measurement system and the back pressure regulator adjusted to the
value required to achieve the flow rates specified by the manufacturer. The zero- and
high-range calibration gases will be injected and the analyzer calibration adjusted to
provide the proper response. The low- and mid-range gases will then be injected and
the responses of the measurement system recorded. The calibration and linearity of the
system will be acceptable when the responses for all four gases are within 5 percent of
the respective gas values. If the performance of the system is not acceptable, the system
will be repaired or adjusted and the linearity check repeated. A calibration and linearity
check will be conducted after assembling the analysis system and at the beginning of
each test day.
System Drift Checks - The calibration gas that most closely approximates the
concentration of the captured emissions will be selected for the conducting the drift
checks. The zero and calibration gas will be introduced at the calibration valve assembly
and the appropriate gas flow rate and pressure present at the FIA will be verified. The
measurement system responses to the zero and calibration gases will be recorded. The
performance of the system is acceptable when the difference between the drift check
measurement and the value obtained in the FIA calibration and linearity check is less
than 3 percent of the span value. At a mirumum, system drift checks will be conducted
at the end of each run.
13
-------
System Check - The high range calibration gas will be injected at the inlet of the
sampling probe and the response recorded. The performance of the system is acceptable
when the measurement system response is within 5 percent of the value obtained in the
FIA calibration and linearity check for the high range calibration gas. A system check
will be conducted before and after each test run.
Analysis Audit - Immediately before each test, the audit cylinder will be analyzed as
described in the system drift check. The analysis audit must agree with the audit
cylinder concentration within 10 percent.
3.3 Measurement of Gas Volumetric Flow Rates
3.3.1 Test Method Overview
The volumetric flow rates for the captured emission stream test stations(s) will be
measured in accordance with EPA Methods 1 through 4. Test stations will be selected
to conform with the minimum requirements of EPA Method 1 (at least two equivalent
diameters to the nearest upstream flow disturbance and at least 0.5 equivalent diameters
to the nearest downstream flow disturbance). The appropriate number of velocity
traverse points will also be selected in accordance with EPA Method 1.
The average velocity pressure, average temperature, static pressure, moisture content,
and O2/CO2 concentrations will be measured at each test station. This information,
along with the cross-sectional area of each test station, will be used to calculate the
volumetric flow rate expressed as standard cubic feet minute (SCFM).
The initial velocity traverse at each station will also include a check for cyclonic flow.
An angle finder will be mounted to the stauscheibe pitot tube. The pitot will be rotated
in the gas flow until a zero reading is obtained, which indicates that the pitot tube is
aligned perpendicular to the gas flow. This check will be completed at each traverse
point with the angle recorded to determine if a cyclonic flow condition exists at the test
station.
3.3.2 Apparatus and Reagents
EPA Method 2 Apparatus will be used for deterrnining volumetric flow rate. This
includes a stauscheibe pitot tube, inclined manometer, and thermocouple temperature
sensor (see Figure 2).
EPA Method 3 Apparatus and Reagents will be used for determining molecular weight
of the gas stream. An integrated gas sample will be collected using the sampling train
shown on Figure 3. Components include a sample probe with filter, air cooled
condenser, vacuum gauge, leakless sample pump, surge tank, rotameter, and Tedlar gas
sampling bag. Concentrations of oxygen and carbon dioxide within the integrated gas
sample will be analyzed with an Orsat Analyzer, or approved equivalent device.
14
-------
EPA Method 4 Apparatus and Reagents will be used for determining the gas stream
moisture content. The moisture sampling train is shown on Figure 4 and includes a
heated and filtered sampling probe, midget impingers containing deiorazed water or
silica gel, temperature sensors, vacuum gauge, leakless sample pump, surge tank,
rotameter, and dry gas meter.
3.3.3 Measurement of Captured Emissions Volumetric Flow Rate
The following measurements will be made at each of the captured emission stream test
stations:
• Velocity pressure will be measured using a stauscheibe pitot tube and
inclined manometer at each traverse point.
• Temperature will be measured using a thermocouple sensor and pyrometer
at each traverse point.
• Concentrations of oxygen and carbon dioxide in an integrated gas sample
will be measured by an Orsat Analyzer, or approved equivalent device, in
accordance with Method 3.
• Flue gas moisture content will be measured by the Standard Method 4
Sampling Train, which includes water impingers and silica gel.
• Static pressure will be measured with a U-tube manometer.
The above listed equipment and procedures will enable determination of the average
velocity of the stack gas. The average velocity will be multiplied by the corresponding
cross-sectional area of the test station to calculate volumetric flow rate. The flow will
be corrected to standard conditions based on temperature and pressure. Example
calculations for determining the volumetric flow rate of a test station are presented in
Section 4.2 of this protocol.
15
-------
4.0 CALCULATIONS TO DETERMINE CAPTURE EFFICIENCY
The following section presents example calculations for the determination of key
parameters needed to measure capture efficiency.
4.1 Liquid VOC Consumption (L)
VOC consumption over a specific test run will be determined by subtracting the ending
weight of VOC (as propane equivalent) in the liquid reservoir system from the starting
weight of VOC (as propane equivalent). The following equations will be used to
calculate each of these parameters:
Beginning Weight of VOC in Coating
LCI = (Wa x Va) / RFCI
Where: LCJ = The beginning weight of VOC in the coating reservoirs
expressed as a propane equivalent (pounds of propane)
WCJ = Net weight of coating in reservoirs at the beginning of the
test calculated as the instantaneous weight measurements of
the coating drum and feed bucket minus the tare weight of
an empty drum and empty feed bucket (pounds of coating)
VQJ = The Method 30F VOC content of the initial coating expressed
as a weight percentage (pounds VOC/pound coating)
RFCI = Method 30 response factor for the initial coating sample
(pounds VOC/pound propane)
Beginning Weight of VOC in Roll Wash Solvent
I* = (Ws,i x VSJ) / RFW
Where: L^ = The beginning weight of VOC in the roll wash solvent
reservoirs expressed as a propane equivalent (pounds of
propane)
WS/I = Net weight of roll wash solvent in reservoirs at the beginning
of the test calculated as the instantaneous weight
measurement of the feed bucket minus the tare weight of an
empty feed bucket (pounds of solvent)
VS4 = The Method 30F VOC content of the roll wash solvent
expressed as a weight percentage (pounds VOC/pound
solvent)
RFS4 = Method 30F response factor for the initial roll wash solvent
sample (pounds VOC/pound propane)
16
-------
Ending Weight of VOC in Coating
= (WC/F x VC(F)
Where: LQF = The ending weight of VOC in the coating reservoirs
expressed as a propane equivalent (pounds of propane)
WCJ: = Net weight of coating in reservoirs at the end of the test
calculated as the instantaneous weight measurements of the
coating drum and feed bucket minus the tare weight of an
empty drum and empty feed bucket (pounds of coating)
Vcf = The Method 30F VOC content of the final coating expressed
as a weight percentage (pounds VOC /pound coating)
RFCjF = Method 30F response factor for the final coating sample
(pounds VOC/pound propane)
Ending Weight of VOC in Spent Roll Wash Solvent
x VS/F) / RFS/F
Where: L^ = The ending weight of VOC in the roll wash solvent
reservoirs expressed as a propane equivalent (pounds of
propane)
WSF = Net weight of roll wash solvent in reservoirs at the end of
the test calculated as the instantaneous weight measurements
of the feed bucket minus the tare weight of an empty feed
bucket (pounds of solvent)
Vs F = The Method 30F VOC content of the spent roll wash solvent
expressed as a weight percentage (pounds VOC/pound
solvent)
RFSJ! = Method 30F response factor for the spent roll wash sample
(pounds VOC/pound propane)
Calculation of VOC Consumption for Test Run
L = (L0 + Ly)
Where: L = Total weight of VOC consumed during the test run measured
as a propane equivalent (pounds of propane)
17
-------
Example calculation of determining the VOC consumption for a hypothetical test run:
Tare Weights
Coating Drum
Coating Feed Bucket
Roll Wash Feed Bucket
Instantaneous Weights at Beginning of Test
Coating Drum
Coating Feed Bucket
RoU Wash Feed Bucket
Instantaneous Weights at End of Test Run
Coating Drum
Coating Feed Bucket
Roll Wash Feed Bucket
Method 30F Data
VOC Content in Initial Coating
Response Factor for Initial Coating
VOC Content of Initial RoU Wash Solvent
Response Factor for Initial Roll Wash Solvent
VOC Content of Final Coating
Response Factor for Final Coating
VOC Content for Spent RoU Wash Solvent
Response Factor for Spent RoU Wash Solvent
= 35.3 pounds
= 10.4 pounds
= 6.3 pounds
457.8 pounds
78.1 pounds
43.5 pounds
256.6 pounds
50.9 pounds
48.6 pounds
= 56.25%
= 1.24
= 99.3%
= 1.22
= 55.80%
= 1.35
= 93.7%
= 1.28
Example Calculations
LCJ = [[(457.8 + 78.1) - (35.3 + 10.4)] x 0.5625] / 1.24 = 222.37 pounds propane
Ly = [(43.5 - 6.3) x 0.993] / 1.22 = 30.28 pounds propane
LCF = [[(256.6 + 50.9) - (35.3 + 10.4)] x 0.558] / 1.35 = 108.21 pounds propane
LSJ = [(48.6 - 6.3) x 0.937] / 1.28 = 30.96 pounds propane
L = (222.37 + 30.28) - (108.21 + 30.96) = 113.48 pounds propane
18
-------
4.2 Response Factor Calculations
The following formulas are used to calculate the response factor generated by Method
30F:
Bag Sample Volume (By)
By = [ Mv x TOT x Y x PM ] / [ TM x PS™ ]
Where: Bv = Bag sample volume (standard liters)
Mv = Dry gas meter volume (liters)
Y = Meter correction factor (dimensionless)
TSTD = Standard temperature (293° K)
TM = Gas meter temperature (° K)
PM = Gas meter pressure (mm Hg absolute)
Psw = Standard pressure (760 mm Hg)
Bag Sample VOC Concentration, as VOC (C,^
CVQC = ML / BV
Where: Cvoc = Bag sample concentration (mg VOC/standard liters)
ML = Weight of VOC liquid injected (mg)
Ba Samle VOC Concentration, as Proane (
Where: CPROP = Bag sample concentration (mg C3H8/standard liter)
RPROP = FIA reading for bag gas sample (ppm C3H8)
K = Conversion Factor [(0.00183 mg C3H8/liter)/(ppm C3H8)]
Resonse Factor (RF)
>ROP
Where: RF = Response factor (mg VOC/mg QHs)
19
-------
Example Calculations
Example data for Method 30F test run:
Mv = 24.95 liters
TM = 297.4 ° K
PM, = 747 mm Hg
ML = 29.4 mg
RPROP = 530 ppm
Y = 1.011
By = [ 24.95 x 293 x 1.011 x 747 ] / [ 297.4 x 760 ] = 24.43 liters
= [ 29.4 / 24.43 ] = 1.204 mg/1
= [ 530 x 0.00183 ] = 0.970 mg/1
RF = [ 1.204 / 0.970 j = 1.241
For the full test method, duplicate analyses of the distillate would be made at three
concentration levels where approximately 30, 60, and 90 ul of VOC would be injected
to form the VOC bag samples. The duplicate response factors at each injection level
would have to agree within 5% to be accepted. Next, all of the data points would be
plotted on a graph of CVoc versus CPROP, and a linear regression analysis performed. If
the linear regression correlation coefficient is greater than 0.90, the overall response
factor will be the slope of the regression line. If the correlation coefficient is less than
0.90, the analysis will be repeated.
4.3 Volumetric Gas Flow Rate
4.3.1 Moisture Content of Gas Stream
The moisture content of the gas stream will be measured in accordance with EPA
Method 4 and calculated using the following equation:
"wo = " WSTD / \ *MSTO "•" * WSTD/
Where: VWSID = volume of water vapor, standard conditions (ft3)
= (0.04707) x VLC
= volume liquid collected (ml)
20
-------
VMSTD = volume gas sampled, standard conditions ((ft3)
V^ = [17.64 x VM x (P, + (dH / 13.6)) x YD] / [460 + TJ
VM = volume of gas sampled, meter conditions (ft3)
PB = barometric pressure (in. Hg)
dH = pressure drop across meter orifice (in. H2O)
YD = gas meter correction factor (dimensionless)
TM = average gas meter temperature (°F)
Example Method 4 Data:
VM = 21.73 ft3
PB = 29.52 in. H2O
VLC = 27 ml
YD = 0.9952
dH = 1.8 in. H2O
TM = 66°F
VWSTD = 0.04707 x (27) = 1.271 ft3
VMSID = [17-64 x (21.73) x (29.52 + (1.8 / 13.6)) x 0.9952] / [460 + 66]
VMSTO = 21.51 dscf
Bwo = (1.271) / (1.271 + 21.51) = 0.0558 = 5.58%
4.3.2 Molecular Weight of Dry Gas Stream
The molecular weight of the gas stream for will be determined in accordance with EPA
Method 3 and calculated by the following formula:
MD = 0.44 ("/oCCy + 0.32 (%O2) + 0.28 (%CO + %N2)
21
-------
Example data from Orsat Analysis at Test Station No. 1:
CO2 = 4.3%
O2 = 17.1%
CO = 0.1%
N2 = 100% - (4.3 + 17.1 + 0.1) = 78.5%
MD = 0.44 (4.3) + 0.32 (17.1) + 0.28 (78.6) = 29.37 Ib/lb mole
Molecular Weight (Ms) of Wet Gas Stream
MS = MD (1 - Bwo) + 18 (Bwo)
MS = [29.37 (1 - 0.0558)] + [18 (0.0558)] = 28.73 Ib/lb mole
4.3.3 Velocity of Gas. Stream
The velocity of the gas stream will be measured in accordance with EPA Method 2 and
will be calculated using the following formula:
Vs = [85.49 CP DP (Ts + 460)m] / [Mg Ps]1/2
Where: Vs = velocity of gas stream (ft/sec)
CP = Pitot tube coefficient (dimensionless)
DP = average square roots of velocity pressure (in. H2O)1/2
Ts = average gas stream temperature (°F)
Ps = absolute gas stream static pressure (in. Hg)
Ps = PB + (SP / 13.6)
SP = gas stream static pressure (in. H2O)
Example Method 2 Data:
CP = 0.84
SP = 3.4 in. H2O
DP = 0.489 (in. H2O)1/2
PB = 29.52 in Hg
22
-------
Ts = 375 °F
Ps = 29.52 + (-3.4 / 13.6) = 29.27 in Hg
Vs = [85.49 x (0.84) x (0.489) x (375 + 460)1/2] / [28.73 x 29.27]172
Vs = 34.99 ft/sec
4.3.4 Volumetric Flow of Gas Stream
The volumetric flow of the gas stream will be calculated based on the actual velocity
times the cross-sectional area and then will be adjusted to standard conditions using the
following formulas:
60As Vs
= [QA Ps (17.64)] / [Ts + 460]
Where: AS = area of stack or duct (ft2)
Example Data Duct Size = 30-inch diameter
AS = 4.91 ft2
QA = 60 (4.91) (34.99) = 10,308 acfm
- [(10,308) (29.27) (17.64)] / [375 + 460]
= 6,374 scfm
4.4 VOC Emission Rates
4.4.1 VOC Concentration at Actual Conditions
The VOC concentration at actual conditions will be measured in accordance with EPA
Method 30B and calculated using the following formula:
C = [(CAVG - C0) x CMA, / [CM - QJ
Where: C = average corrected concentration at actual conditions (ppm propane)
CAVG = average concentration from test station, wet basis (ppm propane)
C0 = average response for zero gas calibration for initial and final
system check (ppm)
23
-------
CM = average response for drift check calibration gas for initial and
final system check (ppm)
CMA = certified concentration of drift check calibration gas from
supplier (ppm)
Example Data:
CAVG = 1/287 ppm
Initial C0 = 0 ppm, Final C0 = 16 ppm, Avg. Q, = 8 ppm
Initial CM = 2,650 ppm, Final CM = 2,678 ppm, Avg. CM = 2,654 ppm
CMA = 2,682 ppm
C = [(1,287 - 8) x (2,682)] / [2,654 - 8]
C = 1,296 ppm propane
4.4.2 Correction for Background VOC and Unburned Natural Gas
Background VOC and unburned natural gas will be measured during the 30-minute
baseline run before each test run. The measured VOC concentration will be adjusted to
correct for background VOC and unburned natural gas using the following formula:
CG = C-CBL
Where: CG = Background corrected VOC concentration (ppm propane)
CBL = Average corrected VOC concentration from 30-minute baseline
run (ppm propane)
Example Data for Background Correction:
Captured VOC C = 1,296 ppm propane
Baseline VOC CBL = 87 ppm propane
CG = 1,296 - 87 = 1,209 ppm propane
4.4.3 Total Captured VOC Emissions
The total captured VOC emissions will be calculated based on the corrected VOC
concentrations (at standard conditions), and the volumetric gas flow rate (at standard
conditions). The following formula will be used to calculate this emission rate:
24
-------
G = (Cc x Qsro x MVVpRQp x t) / 385.3 x 106
Where: G = captured VOC emissions (Ibs propane)
QSTD = volumetric flow rate at standard conditions (scfm)
MWpRQp = molecular weight of propane (44 Ibs/lb-mol)
t = total minutes of sampling
Example Data from Test Run:
GG = 1,209 ppm
QSTD = 6,374 scfm
t = 120 minutes
G = (1,209 x 6,374 x 44 x 120) / 385.3 x 106 = 105.60 Ibs propane
4.5 Capture Efficiency
The measurements taken during each of the replicate test runs will enable calculation
of capture efficiency using the following formula:
CE = G / L
Example data from the test run illustrated in the foregoing example calculations is as
follows:
L = 113.48 Ibs propane
G = 105.60 Ibs propane
CE = 105.60 / 113.48 = 0.931 = 93.1%
25
-------
5.0 QUALITY ASSURANCE/QUALITY CONTROL PROCEDURES
5.1 Quality Control
Quality control procedures for all aspects of field sampling; sample preservation and
holding time; reagent quality; analytical method, analyst training and safety; and
instrument cleaning, calibration and safety will be followed. These procedures are
generally consistent with EPA guidelines documented in "Quality Assurance Handbook
for Air Pollution Measurement Systems;" Volume HI, "Stationary Source Specific
Methods: (EPA-600/4-77-027b).
On those projects where more than 10 personnel are required, a separate quality
assurance officer will be appointed. This officer will report through an independent
chain of command. The test coordinator will have overall authority and responsibility
for quality assurance.
All appropriate field equipment will be calibrated at the testing contractor's laboratory
prior to shipment to the job site. Copies of all calibration certification sheets will be
included in the final test report. Calibration certificates will include as a minimum:
• Unique identification of equipment.
• Calibration procedure used.
• Acceptance criteria (if applicable).
• Person performing calibration.
• Date of calibration.
• Calibration due date (if applicable).
• Standard or natural physical constant used.
The emissions testing will strictly conform to all calibration and quality assurance
procedures outlined in the respective EPA test methods. For example, measurement of
VOC concentrations under EPA Method SOB will include the FIA calibration and
linearity check, systems drift checks, system checks, and analysis audit that are specified
in Section 3.4.4 of this protocol. Other typical quality assurance checks will include
duplicate analysis for Method 24, calibration of Pitot tubes, calibration of resistive
temperature devices, calibration of gas meters, and the Method 30F quality assurance
elements included in Section 3.1.5 of this Protocol. The schedule for calibration and
corrective action is presented in Section 5.3 of this protocol.
26
-------
5.2 Chain of Custody
Documentation of the Chain-of-Custody of samples and data obtained during the test
program is essential for insuring the validity of the test program results. Chain-of-
Custody procedures are followed during sampling, sample and data transport, sample
preparation and analysis, storage of data, as well as with archived samples and reported
results. The testing contractor will follow the protocol listed in SW 846, Section 1.3
during field sampling and in-house laboratory analysis.
5.3 Calibration Frequency Chart
Apparatus
Wet Test Meter
Dry Gas Meter
Thermometers
Probe Heating
System
Barometer
Type S Pilot
Tube
Stack Gas
Temperature
Measurement
System
Acceptance Limits
Capacity >3.4m3/h (120
ftVhr); accuracy within +/-
0.02Y.
Yi = Y +1-2%.
Impinger thermometer +/-
1°C; dry gas meter
thermometer +/-3°C over
range; stack temperature
sensor
+1-1.5% of absolute
temperature.
Capable of maintaining
248°F +/-25°F at a flow of
0.71 ftVmin.
+/-2.5mm (0.1 in.) Hg of
Hg in glass barometer.
All dimension
specifications are met
Capable of measuring
within +1-1.5% of stack
temperature
Frequency and Method of
Measurement
Calibrate initially and then
yearly by liquid
displacement
Calibrate vs. wet test meter
initially and when post-test
check exceeds
±0.05Y.
Calibrate each initially as
separate component against
a Hg in glass thermometer,
then before each field trip
compare each as part of
train with the Hg in glass
thermometer.
Calibrate component
initially by APTD-0576, if
constructed by APTD-0581,
use published calibration
curves.
Calibrate initially vs. Hg in
glass barometer; check
before and after each field
use.
Calibrate initially; visually
inspect before and after
each field test
Calibrate initially and after
each field test
Action if Requirements
are Not Met
Adjust until specifications
are met.
Repair or replace;
recalibrate.
Adjust; determine constant
correction factor or reject.
Repair or replace; reverify
the calibration.
Adjust to agree with
certified barometer.
Repair or replace pilot
tubes that don't meet face
opening specs.
Adjust to agree with Hg
bulb temperature or
calibration curve to correct
the readings.
27
-------
FIGURES
-------
HEATED PROBE
DATA
AQUISITION
SYSTEM
CHART
RECORDER
L_T_J
HEATED
F.I.A
NOTE: PUMP. FILTER.
AND BACK PRESSURE
GAUGE INCLUDED. O
CALIBRATION GAS
LINE ATTATCHED
TO END OF PROBE
FLOW
METER
CALIBRATION GAS
FIGURE 1
FLAME IONIZATION DETECTOR
ANALYZER METHOD SAMPLING SYSTEM
NO SCALE
-------
I
1.90 - 2.54 CM
(.75 - 1.0 IN.)
l!
FLEXIBLE TUBING
6.25 CM
(0.25 IN)
TEMPERATURE SENSOR
LEAK-FREE CONNECTIONS
MANOMETER
NO SCALE
FIGURE 2
TYPE S TUBE-MANOMETER ASSEMBLY
-------
•STACK WALL
• PROBE
(END PACKED
WITH GLASS
WOOL)
VACUUM
GAUGE
_
TT
BY-PASS
VALVE
MAIN
VALVE
RATE
METER
PUMP
SURGE TANK
AIR-COOLED
CONDENSER
TEDLAR GAS - SAMPLING BAG
*
I ca
FIGURE 3
INTEGRATED GAS SAMPLING TRAIN
NO SCALE
5
-------
i!
/
STACK WALL
HEATING ELEMENT
TEMPERATUR
SENSOR
- PROBE"
/
/
>
/
/
tp
MIDGET -^
IMPINGER
/jYpN
\__/
>
P
0
>
>°
(END PACKED
WITH GLASS
WOOL)
D.I. WATER
EMPTY
SILICA-
GEL
TEMPERATURE
SENSOR
\
1
/ DRY \
/ GASS \
nwiwrwriwi'l
BY-PASS
VALVE
RATE
METER
SURGE
TANK
FIGURE 4
MOISTURE SAMPLING TRAIN
NO SCALE
VACUUM
GAUGE
-------
I ^^ro I ^^a
0-3CT HG
MONOMETER
OIL TEMPERATURE
PROBE
ETHYLENE GLYCOL
COOLING/CIRCULATING
BATH
TO ASPIRATOR
DISTILLATE
COLLECTION
FLASK
HOT OIL BATH
STIRRING HOT PLATE
FLASK WITH
TUBULATION
FIGURE 5
VOC DISTILLATION SYSTEM
APPARATUS
NO SCALE
-------
|iS5^^^|p5^^^TiiS^^^^^^I^5^^^^|33!T
MOW I QPV I 6/30/83 I FIQ5 I CM
ACTIVATED CHARCOAL
MOLECULAR SIEVE
METER OUTLET-
TEMPERATURE
PROBE
SYRING
VOLITILIZATION
VESSEL
NEEDLE
VALVE
cc
o
on
Ul
N
J
1
i
-a
U
GLASS W
U-TUBE
MANOMETER
S.S. G
DRY
GAS
METER
WATER BATH
TEMPERATURE
PROBE
HOT BATH WATER
O
STIRRING
HOT PLATE
o
FIGURE 6
TEDLAR GAS BAG GENERATION
SYSTEM APPARATUS
NO SCALE
-------
APPENDIX B
TEST RUN SUMMARY SHEETS
-------
Baseline Run
-------
SUMMARY OF GAS PHASE MEASUREMENTS
BASELINE
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS {Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
845.0
17.0
127.0
4.0
45.0
783.0
5888
180
94.77
CAPTURE
HOODS
(TS-2)
641.0
17.0
624.0
1977
180
25.36
TrE
EXHAUST
(TS-3)
120.12
-------
CALCULATION OF LIQUID VOC CONSUMPTION
BASELINE
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:00:00
2:03:00
3:00:00
TARE WEIGHT
DRUM
(LBS)
531.00
410.50
312.50
213.00
98.40
FEED
BUCKET
(LBs)
26.90
60.10
61.50
64.90
4.80
NET
WEIGHT
(LBs)
454.70
367.40
270.80
174.70
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
41.20
49.10
53.90
44.30
4.20
BUCKET
#2
(LBs)
NET WEIGHT
BUCKET #1
(LBs)
37.00
44.90
49.70
40.10
NET WEIGHT
BUCKET #2
(LBs)
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
RWS BUCKET 2
COATING
RWS BUCKET 1
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
454.70
37.00
174.70
40.10
[V]
VOC
CONTENT
(WT %)
60.70%
84.52%
ER]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.483
1.461
BEGINNING TOTAL:
57.78%
67.16%
1.854
1.305
END TOTAL:
NET LIQUID VOC CONSUMPTION:
IL]
WEIGHT OF
VOC
(LBS C3H8)
186.11
21.40
207.52
54.45
20.64
75.08
132.43
-------
Test Run 1
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN1
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ F 1 TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
796.0
25.7
144.0
3.1
50.1
720.2
6142
181.7
91.79
CAPTURE
HOODS
(TS-2)
607.0
25.7
581.3
2181
181.7
26.31
TTE
EXHAUST
(TS-3)
78.1
25.7
52.4
6830
181.7
7.43
118.10
7.43
-------
CALCULATION OF LIQUID VOC CONSUMPTION
RUN1
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:00:00
2:00:00
2:08:50
2:08:55
2:36:45
3:01:40
FARE WEIGHT
DRUM
(LBs)
458.00
366.00
264.00
FEED
BUCKET
(LBs)
34.54
33.48
43.00
SOLVENT ADDED TO
RWS BUCKET #1
RWS BUCKET CHANGE
191.00
98.40
32.42
4.14
NET
WEIGHT
(LBs)
390.00
296.94
204.46
PRE-ADD (PRE)
POST-ADD (POST)
120.88
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
22.18
26.02
27.48
27.72
35.08
35.52
4.06
BUCKET
#2
(LBs)
29.26
30.68
4.12
NET WEIGHT
BUCKET #1
(LBs)
18.12
21.96
23.42
23.66
31.02
31.46
NET WEIGHT
BUCKET #2
(LBs)
25.14
26.56
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS B-l (PRE)
RWS B-l (POST)
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
390.00
18.12
31.02
25.14
[V]
VOC
CONTENT
(WT %)
58.65%
90.89%
77.67%
94.04%
[R]
RESPONSE
FACTOR
(LBSVOC/LBSC3H8)
1.392
1.358
1.303
1.303
BEGINNING TOTAL:
COATING
RWS B-l (PRE)
RWS B-l (POST)
RWS BUCKET 2
120.88
23.66
31.46
26.56
55.40%
71.96%
77.89%
82.77%
1.514
1.337
1.425
1.426
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
164.32
12.13
18.49
18.14
213.08
44.23
12.73
17.20
15.42
89.58
123.50
-------
Test Run 2
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN 2
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppmCSHS)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (oom)
UNBURNED NATURAL GAS
CONCENTRATION (ppmCSHS)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppmCSHS)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS Obs C3H8)
[ G I TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
913.0
11.2
125.0
2.7
43.5
858.3
5679
180
100.20
CAPTURE
HOODS
(TS-2)
527.0
11.2
515.8
2083
180
22.08
TTE
EXHAUST
(TS-3)
612
112
50.0
6859
180
7.05
122.28
7.05
-------
CALCULATION OF LIQUID VOC CONSUMPTION
RUN 2
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:01:05
2:07:00
2:29:30
3:00:00
TARE WEIGHT
DRUM
(LBS)
405.00
324.50
210.50
FEED
BUCKET
(LBs)
67.14
45.74
61.60
NET
WEIGHT
(LBs)
370.38
268.48
170.34
RWS BUCKET CHANGE
125.50
98.40
50.88
3.36
74.62
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
27.00
33.82
37.56
43.18
4.10
BUCKET
#2
(LBs)
25.98
29.05
4.12
NET WEIGHT
BUCKET #1
(LBS)
22.90
29.72
33.46
39.08
NET WEIGHT
BUCKET #2
(LBs)
21.86
24.93
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
370.38
22.90
21.86
[V]
VOC
CONTENT
(WT %)
60.29%
90.58%
99.83%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.500
1.315
1.278
BEGINNING TOTAL:
COATING
RWS BUCKET 1
RWS BUCKET 2
74.62
39.08
24.93
57.74%
68.28%
84.67%
1.407
1.388
1.361
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
148.87
15.77
17.08
181.72
30.62
19.22
15.51
65.36
116.36
-------
Test Run 3
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUNS
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G 1 TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
986.0
15.2
114.0
2.5
39.7
931.1
5807
181
111.76
CAPTURE
HOODS
(TS-2)
606.0
152
590.8
2063
181
25.19
TIE
EXHAUST
(TS-3)
72.7
152
575
6855
181
8.15
136.95
8.15
-------
CALCULATION OF LIQUID VOC CONSUMPTION
RUNS
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:00:00
2:03:00
2:19:21
3:01:05
TARE WEIGHT
DRUM
(LBS)
459.00
344.50
238.50
FEED
BUCKET
(LBs)
58.74
62.85
56.30
BUCKET CHANGE
176.50
98.40
19.48
4.76
NET
WEIGHT
(LBs)
414.58
304.19
191.64
92.82
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
22.18
28.54
34.00
37.84
4.10
BUCKET
#2
(LBs)
23.50
27.76
4.06
NET WEIGHT
BUCKET #1
(LBs)
18.08
24.44
29.9
33.74
NET WEIGHT
BUCKET #2
(LBs)
19.44
23.7
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
RWS BUCKET 2
IW]
NET
WEIGHT
(LBS)
414.58
18.08
19.44
IV]
VOC
CONTENT
(WT %)
60.50%
99.89%
99.95%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.482
1.349
1.355
BEGINNING TOTAL:
COATING
RWS BUCKET 1
RWS BUCKET 2
92.82
33.74
23.70
57.59%
70.56%
82.12%
1.606
1.541
1.420
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
169.24
13.39
14.34
196.97
33.28
15.45
13.71
62.44
134.53
-------
Test Run 4
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN4
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
1061.0
21.9
118.0
2.5
41.0
998.1
5547
180
113.80
CAPTURE
HOODS
(TS-2)
538.0
21.9
516.1
2038
180
21.62
TTE
EXHAUST
(TS-3)
75.3
21.9
53.4
6754
180
7.41
135.42
7.41
-------
CALCULATION OF LIQUID VOC CONSUMPTION
RUN 4
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:00:50
1:38:00
2:00:00
3:00:00
TARE WEIGHT
DRUM
(LBS)
385.50
287.00
FEED
BUCKET
(LBs)
77.40
61.88
BUCKET CHANGE
185.00
103.50
98.40
57.96
24.60
4.64
NET
WEIGHT
(LBs)
359.86
245.84
139.92
25.06
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
22.80
30.22
32.70
4.10
BUCKET
#2
(LBs)
21.68
23.40
26.70
4,04
NET WEIGHT
BUCKET #1
(LBs)
18.70
26.12
28.60
NET WEIGHT
BUCKET #2
(LBs)
17.64
19.36
22.66
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
359.86
18.70
17.64
[V]
VOC
CONTENT
(WT %)
62.13%
99.96%
99.95%
[RI
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.483
1.268
1.099
BEGINNING TOTAL:
COATING
RWS BUCKET 1
RWS BUCKET 2
25.06
28.60
22.66
59.64%
65.26%
77.70%
1.481
1.409
1.364
END TOTAL:
NET LIQUID VOC CONSUMPTION:
. [LI
WEIGHT OF
VOC
(LBS C3H8)
150.76
14.74
16.04
181.55
10.09
13.25
12.91
36.25
145.30
-------
Test Run 5
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUNS
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfin)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
( F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
1004.0
19.8
126.0
2.8
43.9
940.3
5501
180
106.33
CAPTURE
HOODS
(TS-2)
622.0
19.8
602.2
1963
180
24.30
TTE
EXHAUST
(TS-3)
752
19.8
55.4
6717
180
7.65
130.63
7.65
-------
CALCULATION OF LIQUID VOC CONSUMPTION
RUNS
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:00:00
2:01:00
2:06:45
3:00:00
TARE WEIGHT
DRUM
(LBs)
465.50
370.50
230.50
FEED
BUCKET
(LBs)
29.44
43.30
67.42
RWS BUCKET CHANGE
167.50
98.40
25.60
4.96
NET
WEIGHT
(LBs)
391.58
310.44
194.56
89.74
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
24.52
25.66
26.16
27.90
4.10
BUCKET
#2
(LBs)
30.20
29.94
4.04
NET WEIGHT
BUCKET #1
(LBs)
20.42
21.56
22.06
23.80
NET WEIGHT
BUCKET #2
(LBs)
26.16
25.90
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
391.58
20.42
26.16
[V]
VOC
CONTENT
(WT %)
61.07%
94.92%
99.95%
[Rl
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.470
1.373
1.239
BEGINNING TOTAL:
COATING
RWS BUCKET 1
RWS BUCKET 2
89.74
23.80
25.90
58.61%
75.43%
86.02%
1.550
1.535
1.369
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
162.68
14.12
21.10
197.90
33.93
11.70
16.27
61.90
136.00
-------
Test Run 6
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN 6
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE ( oom)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
917.0
26.5
39.8
1.0
13.9
876.6
5923
180
106.72
CAPTURE
HOODS
(TS-2)
611.0
26.5
584.5
2047
180
24.59
TTE
EXHAUST
(TS-3)
83.8
26.5
57.3
6811
180
8.02
131.32
8.02
-------
CALCULATION OF LIQUID VOC CONSUMPTION
RUN 6
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:00:40
2:00:00
2:00:30
3:00:00
TARE WEIGHT
DRUM
(LBS)
464.50
326.50
227.50
FEED
BUCKET
(LBs)
34.82
62.33
65.86
NET
WEIGHT
(LBs)
395.74
285.25
189.78
RWS BUCKET CHANGE
157.00
98.40
46.54
5.18
99.96
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
19.76
19.78
19.38
19.40
4.10
BUCKET
#2
(LBs)
22.30
23.26
4.04
NET WEIGHT
BUCKET #1
(LBs)
15.66
15.68
15.28
15.3
NET WEIGHT
BUCKET #2
(LBS)
18.26
19.22
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
395.74
15.66
18.26
[V]
VOC
CONTENT
(WT %)
60.86%
96.79%
99.70%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.317
1.248
1.183
BEGINNING TOTAL:
COATING
RWS BUCKET 1
RWS BUCKET 2
99.96
15.30
19.22
57.66%
79.29%
88.23%
1.500
1.354
1.285
END TOTAL-.
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
182.88
12.15
15.39
210.41
38.42
8.96
13.20
60.58
149.83
-------
Test Run 7
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN 7
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (pom C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
860.0
31.4
38.0
1.0
13.3
815.3
5722
180
95.89
CAPTURE
HOODS
(TS-2)
616.0
31.4
584.6
2067
180
24.84
TTE
EXHAUST
(TS-3)
87.8
31.4
56.4
6717
180
7.79
120.73
7.79
-------
CALCULATION OF LIQUID VOC CONSUMPTION
RUN 7
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:00:00
2:00:00
2:36:45
3:00:00
TARE WEIGHT
DRUM
(LBs)
451.50
355.50
258.00
FEED
BUCKET
(LBs)
60.24
66.22
69.71
RWS BUCKET CHANGE
165.00
98.40
60.96
5.02
NET
WEIGHT
(LBs)
408.32
318.30
224.29
122.54
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
23.12
23.42
23.88
24.45
4.12
BUCKET
#2
(LBs)
24.24
24.30
4.06
NET WEIGHT
BUCKET #1
(LBs)
19
19.3
19.76
20.33
NET WEIGHT
BUCKET #2
(LBs)
20.18
20.24
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
408.32
19.00
20.18
[V]
VOC
CONTENT
(WT %)
60.97%
99.86%
99.93%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.448
1.484
1.609
BEGINNING TOTAL:
COATING
RWS BUCKET 1
RWS BUCKET 2
122.54
20.33
20.24
61.09%
76.72%
91.92%
1.644
1.762
1.448
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
171.93
12.79
12.53
197.25
45.54
8.85
12.85
67.24
130.01
-------
Test Run 8
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUNS
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (pom C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfml
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAFTURED VOC
EMISSIONS (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
1041.0
22.5
39.0
1.0
13.7
1004.8
5784
180
119.47
CAPTURE
HOODS
(TS-2)
607.0
225
584.5
2065
180
24.81
TTE
EXHAUST
(TS-3)
87.0
22.5
64.5
6787
180
9.00
144.28
9.00
-------
CALCULATION OF LIQUID VOC CONSUMPTION
RUNS
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:00:00
2:00:00
2:10:30
3:00:00
TARE WEIGHT
DRUM
(LBs)
458.50
316.00
219.00
FEED
BUCKET
(LBs)
43.70
67.76
48.75
NET
WEIGHT
(LBs)
398.88
280.44
164.43
RWS BUCKET CHANGE
140.00
98.40
25.38
4.92
62.06
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
21.52
20.82
19.66
19.64
4.10
BUCKET
#2
(LBs)
24.18
20.98
4.04
NET WEIGHT
BUCKET #1
(LBs)
17.42
16.72
15.56
15.54
NET WEIGHT
BUCKET #2
(LBs)
20.14
16.94
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
398.88
17.42
20.14
[V]
VOC
CONTENT
(WT %)
60.91%
99.80%
99.89%
[R]
RESPONSE
FACTOR
(LBSVOC/LBSC3H8)
1.511
1.242
1.230
BEGINNING TOTAL:
COATING
RWS BUCKET 1
RWS BUCKET 2
62.06
15.54
16.94
60.39%
78.57%
90.52%
1.512
1.548
1.316
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
160.79
14.00
16.36
191.15
24.79
7.89
11.65
44.33
146.82
-------
Test Run 9
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN 9
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscftn)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
969.0
29.6
42.8
1.0
14.9
924.5
5639
178
105.97
CAPTURE
HOODS
(TS-2)
637.0
29.6
607.4
1993
178
24.61
TTE
EXHAUST
(TS-3)
94.5
29.6
64.9
6724
178
8.87
130.57
8.87
-------
CALCULATION OF LIQUID VOC CONSUMPTION
RUN 9
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:07:00
1:52:00
2:02:15
2:58:00
TARE WEIGHT
DRUM
(LBS)
458.00
296.50
FEED
BUCKET
(LBs)
44.02
87.62
BUCKET CHANGE
216.00
160.00
98.40
76.66
29.34
4.94
NET
WEIGHT
(LBs)
398.68
280.78
189.32
86.00
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
22.18
21.40
22.57
4.12
BUCKET
#2
(LBs)
30.31
29.35
28.44
4.04
NET WEIGHT
BUCKET #1
(LBs)
18.06
17.28
18.45
NET WEIGHT
BUCKET #2
(LBs)
26.27
25.31
24.40
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
RWS BUCKET 2
COATING
RWS BUCKET 1
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
398.68
18.06
26.27
86.00
18.45
24.40
[V]
VOC
CONTENT
(WT%)
62.16%
99.72%
99.99%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.437
1.329
1.316
BEGINNING TOTAL:
58.64%
82.87%
87.52%
1.422
1.440
1.306
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
172.46
13.55
19.96
205.97
35.46
10.62
16.35
62.43
143.53
-------
Test Run 10
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN 10
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (pom)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
1147.0
37.6
14.0
1095.4
6192
60
46.47
CAPTURE
HOODS
(TS-2)
579.0
37.6
541.4
2134
60
7.92
TTE
EXHAUST
(TS-3)
88-3
37.6
50.7
6867
60
2.39
5439
2.39
-------
CALCULATION OF LIQUID VOC CONSUMPTION
RUN 10
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:00:00
TARE WEIGHT
DRUM
(LBs)
452.50
303.50
98.40
FEED
BUCKET
(LBs)
48.32
76.34
4.92
NET
WEIGHT
(LBs)
397.50
276.52
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
21.58
21.80
4.12
BUCKET
#2
(LBs)
4.04
NET WEIGHT
BUCKET #1
(LBs)
17.46
17.68
NET WEIGHT
BUCKET #2
(LBs)
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
[W]
NET
WEIGHT
(LBS)
397.50
17.46
[V]
VOC
CONTENT
(WT %)
59.96%
99.56%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.409
1.312
BEGINNING TOTAL:
COATING
RWS BUCKET 1
276.52
17.68
58.57%
87.12%
1.420
1.460
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
169.16
13.25
182.41
114.05
10.55
124.60
57.80
-------
Test Run 11
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN 11
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G 1 TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
1045.0
26.4
14.0
1004.6
5968
60
41.08
CAPTURE
HOODS
(TS-2)
585.0
26.4
558.6
2134
60
8.17
TTE
EXHAUST
(TS-3)
69.1
26.4
42.7
8282
60
2.42
49.25
2.42
-------
CALCULATION OF LIQUID VOC CONSUMPTION
RUN 11
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:00:00
TARE WEIGHT
DRUM
(LBs)
208.50
145.50
98.40
FEED
BUCKET
(LBs)
72.24
22.60
4.92
NET
WEIGHT
(LBs)
177.42
64.78
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
24.32
24.62
4.04
BUCKET
#2
(LBs)
4.12
NET WEIGHT
BUCKET #1
(LBs)
20.20
20.50
NET WEIGHT
BUCKETS
(LBs)
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
[W]
NET
WEIGHT
(LBS)
177.42
20.20
[V]
VOC
CONTENT
(WT %)
59.07%
99.99%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.512
1.559
BEGINNING TOTAL:
COATING
RWS BUCKET 1
64.78
20.50
57.77%
88.23%
1.446
1.609
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
69.31
12.96
82.27
25.88
11.24
37.12
45.15
-------
Test Run 12
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN 12
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (DDHV)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscftn)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
MEASUREMENT LOCATIONS
OVEN
EXHAUST
(TS-1)
1633.0
33.7
14.0
1585.3
6159
100
111.50
CAPTURE
HOODS
(TS-2)
607.0
33.7
573.3
2140
100
14.01
TTE
EXHAUST
(TS-3)
81.1
33.7
47.4
8077
100
4.37
125.51
4.37
-------
CALCULATION OF LIQUID VOC CONSUMPTION
RUN 12
COATING WEIGHT
MEASUREMENTS
ELAPSED
TIME
START
1:09:20
1:40:00
TARE WEIGHT
DRUM
(LBs)
431.00
296.00
287.5
98.40
FEED
BUCKET
(LBs)
76.72
80.12
16.1
5.04
NET
WEIGHT
(LBs)
404.40
272.80
200.28
ROLL WASH SOLVENT
WEIGHT MEASUREMENTS
BUCKET
#1
(LBs)
22.46
21.88
21.28
4.08
BUCKET
#2
(LBs)
4.12
NET WEIGHT
BUCKET #1
(LBs)
18.34
17.76
17.16
NET WEIGHT
BUCKETS
(LBs)
BEGINNING
OF
TEST
END
OF
TEST
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
[W]
NET
WEIGHT
(LBS)
404.40
18.34
[V]
VOC
CONTENT
(WT %)
65.84%
99.00%
[RJ
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.197
1.596
BEGINNING TOTAL:
COATING
RWS BUCKET 1
200.28
17.16
63.14%
86.55%
1.080
1.301
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
222.44
11.38
233.81
117.09
11.42
128.51
105.31
-------
APPENDIX C
AIR VELOCITY DATA SHEETS
-------
SUMMARY OF VELOCITY READINGS
BASELINE
ROLL
COATER
CAPTURE HOOD
UPPERHOOD
FRONTPAGE
UPPERHOOD
REAR FACE
UPPERHOOD
BOTTOM
EXIT HOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
0
0
3
6
0
0
3
6
0
3
6
VELOCITY READINGS ffom)
LEFT
900
65
40
1000
1800
95
30
1165
850
100
55
|_ 2050
140
35
CENTER
700
92
28
1165
1300
55
20
950
1200
75
45
2150
150
30
RIGHT
850
55
40
1100
1100
80
25
850
1150
60
25
1900
145
35
AVG.
VELOCITY
(fpm)
817
71
36
1088
1400
77
25
988
1067
78
42
2033
145
33
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VELOCITY READINGS (fom)
LEFT
95
130
LEFT
CENTER
110
105
RIGHT
CENTER
110
115
FLOW INTO OVEN
RIGHT
135
125
(ACFM)
(SCFM)
AVG.
VELOCITY
(fpm)
118
115
933
923
-------
SUMMARY OF VELOCITY READINGS
RUN1
ROLL
COATER
CAPTURE HOOD
UPPER HOOD
FRONT FACE
UPPER HOOD
REAR FACE
UPPER HOOD
BOTTOM
EXTTHOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
0
0
3
6
0
0
3
6
0
3
6
VELOCITY READINGS (fpm)
LEFT
1350
45
90
1300
70
35
680
30
25
2100
55
30
CENTER
1150
140
120
1300
1150
70
35
1350
1050
30
25
1950
130
40
RIGHT
1250
35
45
1250
65
35
1200
60
20
1850
125
85
AVG.
VELOCITY
(fpm)
1250
73
85
1300
1233
68
35
1350
977
40
23
1967
103
52
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VELOCITY READINGS (torn)
LEFT
15
55
LEFT
CENTER
40
25
RIGHT
CENTER
60
45
FLOW INTO OVEN
RIGHT
30
55
(ACFM)
(SCFM)
AVG.
VELOCITY
(fpm)
36
45
325
327
-------
SUMMARY OF VELOCITY READINGS
RUN 2
ROLL
COATER
CAPTURE HOOD
UPPER HOOD
FRONT FACE
UPPER HOOD
REAR FACE
UPPER HOOD
BOTTOM
EXIT HOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
6
0
3
6
6
0
3
6
0
3
6
VELOCITY READINGS (fpm)
LEFT
81
132
32
30
25
8
CENTER
85
99
23
35
28
70
RIGHT
53
38
25
88
AVG.
VELOCITY
(fpm)
83
116
36
34
26
55
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VEt
LEFT
27
46
OCTTY READINGS (torn)
LEFT
CFJNTER
33
42
RIGHT
CENTER
25
25
FLOW INTO OVEN
RIGHT
22
36
(ACFM)
(SCFM)
AVG.
VELOCITY
(fpm)
27
37
256
256
-------
SUMMARY OF VELOCITY READINGS
RUNS
ROLL
COATER
CAPTURE HOOD
UPPERHOOD
FRONT FACE
UPPERHOOD
REAR FACE
UPPERHOOD
BOTTOM
EXIT HOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
6
0
3
6
6
0
3
6
0
3
6
VELOCITY READINGS (fpm)
LEFT
120
105
32
47
18
110
CENTER
130
110
40
40
30
90
RIGHT
150
112
35
50
25
93
AVG.
VELOCITY
(fpm)
133
109
36
46
24
98
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VELOCITY READINGS ffom)
LEFT
30
63
LEFT
CENTER
45
54
RIGHT
CENTER
25
48
FLOW INTO OVEN
RIGHT
20
33
(ACFM)
(SCFM)
AVG.
VELOCITY
(fpm)
30
50
318
318
-------
SUMMARY OF VELOCITY READINGS
RUN 4
ROLL
COATER
CAPTURE HOOD
UPPER HOOD
FRONT FACE
UPPER HOOD
REAR FACE
UPPER HOOD
BOTTOM
EXIT HOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
6
0
3
6
6
0
3
6
0
3
6
VELOCITY READINGS (fan)
LEFT
101
122
38
28
39
87
CENTER
123
47
26
54
66
118
RIGHT
87
39
28
37
20
112
AVG.
VELOCITY
(fpm)
104
69
31
40
42
106
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VELOCITY REAI
LEFT
34
89
LEFT
CENTER
44
36
DINGS (fpm)
RIGHT
CENTER
53
47
FLOW INTO OVEN
RIGHT
21
42
(ACFM)
(SCFM)
AVG.
VELOCITY
(fpm)
38
54
366
363
-------
SUMMARY OF VELOCITY READINGS
RUNS
ROLL
COATER
CAPTURE HOOD
UPPER HOOD
FRONT FACE
UPPER HOOD
REAR FACE
UPPER HOOD
BOTTOM
EXIT HOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
6
0
3
6
6
0
3
6
0
3
6
VELOCITY READINGS (fpm)
LEFT
124
120
55
27
19
52
CENTER
70
124
77
123
41
86
RIGHT
112
124
32
48
36
100
AVG.
vELoary
(fpm)
102
123
55
66
32
79
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VELOCITY READINGS (fom)
LEFT
33
72
LEFT
CENTER
18
40
RIGHT
CENTER
23
49
FLOW INTO OVEN
RIGHT
19
27
(ACFM)
(SCFM)
AVG.
VELOCITY
(fpm)
23
47
281
278
-------
SUMMARY OF VELOCITY READINGS
RUN 6
ROLL
COATER
CAPTURE HOOD
UPPER HOOD
FRONTPAGE
UPPER HOOD
REAR FACE
UPPER HOOD
BOTTOM
EXIT HOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
6
0
3
6
6
0
3
6
0
3
6
VELOCITY READINGS (fpm)
LEFJ.
100
84
.
36
38
21
115
CENTER
160
160
45
42
55
97
RIGHT
95
145
52
56
57
75
AVG.
VELOCITY
(fpm)
118
130
44
45
44
96
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VELOCITY REAI
LEFT
38
42
LEFT
CENTER
33
43
DINGS ffpm)
RIGHT
CENTER
25
35
FLOW INTO OVEN
RIGHT
16
23
(ACFM)
(SCFM)
AVG.
VELOCITY
(fpm)
28
36
255
252
-------
SUMMARY OF VELOCITY READINGS
RUN 7
ROLL
COATER
CAPTURE HOOD
UPPER HOOD
FRONTPAGE
UPPER HOOD
REAR FACE
UPPER HOOD
BOTTOM
EXIT HOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
6
0
3
6
6
0
3
6
0
3
6
VELOCITY READINGS (fpm)
LEFT
139
60
43
38
25
88
CENTER
118
73
43
45
75
62
RIGHT
160
90
35
52
50
95
AVG.
VELOCITY
(fpm)
139
74
40
45
50
82
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VELOCITY READINGS Ifom)
LEFT
27
62
LEFT
CENTER
18
32
RIGHT
CENTER
28
45
FLOW INTO OVEN
RIGHT
18
35
(ACFM)
(SCFM)
AVG.
VELOCITY
(fpm)
23
44
265
262
-------
SUMMARY OF VELOCITY READINGS
RUNS
ROLL
COATER
CAPTURE HOOD
UPPER HOOD
FRONTPAGE
UPPER HOOD
REAR FACE
UPPER HOOD
BOTTOM
EXIT HOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
0
0
3
6
0
0
3
6
0
3
6
VELOCITY READINGS ffpm)
LEFT
1050
110
140
1150
1250
80
45
1180
1100
65
35
2400
120
105
CENTER
1350
130
125
1300
1480
120
55
1320
920
70
55
2650
145
75
RIGHT
1100
95
110
1250
1000
75
55
1240
850
50
20
2500
115
70
AVG.
VELOCTIY
(fern)
1167
112
125
1233
1243
92
52
1247
957
62
37
2517
127
83
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VELOCITY REAI
LEFT
35
55
LEFT
CENTER
20
40
DINGS ffpm)
RIGHT
CENTER
25
25
FLOW INTO OVEN
RIGHT
30
20
(ACFM)
(SCFM)^
AVG.
VELOCITY
(fpm)
28
35
250
249
-------
SUMMARY OF VELOCITY READINGS
RUN 9
ROLL
COATER
CAPTURE HOOD
UPPER HOOD
FRONTPAGE
UPPER HOOD
REAR FACE
UPPER HOOD
BOTTOM
EXIT HOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
0
0
3
6
0
0
3
6
0
3
6
VELOCITY READINGS (fpm)
LEFT
1100
45
65
1250
1200
50
45
1150
1220
20
25
2500
125
75
CENTER
1450
135
115
1100
1200
65
60
1350
1500
25
25
2550
135
90
RIGHT
1250
90
90
1150
1000
75
40
850
1200
70
35
2600
125
65
AVG.
VELOCITY
(fpm)
1267
90
90
1167
1133
63
48
1117
1307
38
28
2550
128
77
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VELOCITY READINGS f fonrt
LEFT
65
80
LEFT
CENTER
40
35
RIGHT
CENTER
50
35
FLOW INTO OVEN
RIGHT
30
35
(ACFM)
(SCFM)
AVG.
VELOCITY
(fpm)
46
46
370
365
-------
SUMMARY OF VELOCITY READINGS
RUN 10
ROLL
COATER
CAPTURE HOOD
UPPER HOOD
FRONTPAGE
UPPER HOOD
REAR FACE
UPPER HOOD
BOTTOM
EXIT HOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
0
0
3
6
0
0
3
6
0
3
6
VELOCITY READINGS ffpm)
LEFT
1050
55
35
1100
1200
35
30
1050
1300
95
30
2300
75
20
CENTER
1250
70
45
1300
1350
25
35
1250
1450
25
30
2450
95
45
RIGHT
1150
90
55
1200
1350
35
25
850
1250
100
20
2550
115
25
AVG.
VELOCITY
(fpm)
1150
72
45
1200
1300
32
30
1050
1333
73
27
2433
95
30
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VELOCITY READINGS (torn)
LEFT
20
45
LEFT
CENTER
35
25
RIGHT
CENTER
30
35
FLOW INTO OVEN
RIGHT
25
40
(ACFM)
(SCFM)
AVG.
VELOCITY
(ftsn)
28
36
255
251
-------
SUMMARY OF VELOCITY READINGS
RUN 11
ROLL
COATER
CAPTURE HOOD
UPPER HOOD
FRONT FACE
UPPER HOOD
REAR FACE
UPPER HOOD
BOTTOM
EXIT HOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
6
0
3
6
6
0
3
6
0
3
6
VELOCITY READINGS (fom)
LEFT
1120
110
65
1150
1200
65
30
1200
1300
115
45
2250
110
30
CENTER
1300
100
80
1300
1350
75
25
1250
1250
55
25
2700
125
70
RIGHT
1250
95
105
1250
950
35
20
980
1300
115
30
2550
125
50
AVG.
VELOCITY
(fpm)
1223
102
83
1233
1167
58
25
1143
1283
95
33
2500
120
50
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VELOCITY REAI
LEFT
40
65
LEFT
CENTER
55
35
DINGS (fpm\
RIGHT
CENTER
45
50
FLOW INTO OVEN
RIGHT
30
25
(ACFM)
(SCFM)
AVG.
VELOCITY
(fpm)
43
44
345
340
-------
SUMMARY OF VELOCITY READINGS
RUN 12
ROLL
COATER
CAPTURE HOOD
UPPER HOOD
FRONT FACE
UPPER HOOD
REAR FACE
UPPER HOOD
BOTTOM
EXIT HOOD
SLOT
LOCATION
UPPER
SLOTS
LOWER SLOTS
UPPER
SLOTS
LOWER SLOTS
BOTTOM
SLOTS
UPPER
SLOTS
DISTANCE
FROM SLOT
(inches)
0
3
6
6
0
3
6
6
0
3
6
0
3
6
VELOCITY READINGS (fom)
LEFT
1100
70
65
1150
1100
45
35
1150
1320
100
40
2450
120
60
CENTER
1400
110
30
1250
1320
55
40
1200
1350
65
35
2600
120
55
RIGHT
1300
65
95
1250
1050
30
35
1050
1250
45
25
2750
120
55
AVG.
VELOCITY
(fpm)
1267
82
63
1217
1157
43
37
1133
1307
70
33
2600
120
57
OVEN ENTRANCE
AT FACE
OPENING
UPPER ROW
LOWER ROW
VELOCITY REA
LEFT
25
35
LEFT
CENTER
35
40
DINGS (fpm)
RIGHT
CENTER
25
40
FLOW INTO OVEN
RIGHT
20
25
(ACFM)
(SCFM)
AVG.
VELOCITY
(fpm)
26
35
245
241
-------
APPENDIX D
EXTRACTION OF SHORT DURATION TEST RESULTS
-------
APPENDIX D
SECTION 1
COATING CONSUMPTION AT BREAKPOINT
-------
450
RUN1
COATING CONSUMPTION VS. TIME
TIME (MIN.)
t = 128.7 min.
-------
450
RUN2
COATING CONSUMPTION VS. TIME
20
40
60
80 100
TIME(MIN.)
120
140 160 180
t = 149.5 rain
-------
450
20
RUN3
COATING CONSUMPTION VS. TIME
80 100 120
TIME(MIN.)
t = 139.3 min
-------
RUN4
COATING CONSUMPTION VS. TIME
20
40
60
80
TIME (ME"-
.00
•)
120
140
160
180
t = 98.0 min
-------
450
0
20
RUNS
COATING CONSUMPTION VS. TIME
40
60
80 100
TIME (MIN.)
120 140
t = 126.8 nun
160
180
-------
450
20
RUN 6
COATING CONSUMPTION VS. TIME
40
80 100
TIME (MIN.)
t = 120.5 min
-------
450
20
RUN7
COAUNG CONSUMPTION VS. TIME
40
80 100
TIME (MEN.)
t = 156.8 min
-------
20
RUNS
COATING CONSUMPTION VS. TIME
40
60
80 100
TIME (MIN.)
120 140 160
t = 130.5 rain
180
-------
20
RUN9
COATING CONSUMPTION VS. TIME
40
60
80 100
TIME(MIN.)
120
t = 112.0 min
160
180
-------
APPENDIX D
SECTION 2
COATING RESPONSE FACTOR AND VOC CONTENT AT BREAKPOINT
-------
60
RUN1
VOC CONTENT & RESPONSE FACTOR VS. TIME
55-
20
40
60
80 100
TIME (MM.)
120
140
160
t = 128.7 min
180
1.38
PERCENT VOC
-B-- RESPONSE FACTOR
-------
RUN 2
VOC CONTENT & RESPONSE FACTOR VS. TIME
1.55
60.5-
U
g
w
g
w
PH
59.5-'
58.5- •
57.5-
20
40
60
80 100
TIME (MIN.)
t = 149.5 min
PERCENT VOC
-o- RESPONSE FACTOR
-------
RUNS
VOC CONTENT &: RESPONSE FACTOR VS. TIME
1.65
57-
20
40
60
80 100
TIME (MIN.)
180
1.4
t = 139.3 min
PERCENT VOC - & - RESPONSE FACTOR
-------
RUN 4
VOC CONTENT & RESPONSE FACTOR VS. TIME
1.487
CO
*?y
20
180
1.48
t = 98.0 min
PERCENT VOC
-a- RESPONSE FACTOR
-------
RUNS
VOC CONTENT & RESPONSE FACTOR VS. TIME
1.58
58-
20
40
80 100
TIME (NUN.)
160
180
1.44
t = 126.8 mi
PERCENT VOC
-a- RESPONSE FACTOR
-------
RUN 6
VOC CONTENT & RESPONSE FACTOR VS. TIME
1.55
57
20
40
80 100
TIME (MDSI.)
160
180
1.3
t = 120.5 min
PERCENT VOC
-o~ RESPONSE FACTOR
-------
RUN7
VOC CONTENT & RESPONSE FACTOR VS. TIME
61
1.65
oo
60.94
1.4
20
40
60
80 100
TIME (MIR)
180
t = 156.8 min
PERCENT VOC
-a~ RESPONSE FACTOR
-------
RUNS
VOC CONTENT & RESPONSE FACTOR VS. TIME
1.513
y*^
0
60.2'
20
40
60
80 100
TIME (MIN.)
120
140
160
t = 130.5 min
180
1.51
PERCENT VOC
-Q-- RESPONSE FACTOR
-------
RUN9
VOC CONTENT & RESPONSE FACTOR VS. TIME
1.445
58'
20
40
60
80 100
TIME (MIN.)
t =
120 140 160
112.0 rain
180
1.42
PERCENT VOC
-S-- RESPONSE FACTOR
-------
APPENDIX D
SECTION 3
CALCULATION OF SHORT DURATION LIQUID VOC CONSUMPTION
-------
RUN 1 - START TO FIRST RWS ADD
(2 HOUR RUN)
ELAPSED
TIME
(min)
0
128
LIQUID FEED
RESERVOIR
^COATING
RWS BUCKET 1
[W]
NET
WEIGHT
(LBS)
390.00
18.12
m
VOC
CONTENT
(WT%)
58.65%
90.89%
[R]
RESPONSE
FACTOR
(LBSVOOLBSC3H8]
1.392
1358
BEGINNING TOTAL-
COATING
RWS BUCKET 1
192.95
23.66
56.35%
71.96%
1.478
1337
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
164.32
12.13
176.45
73.56
12.73
8630
90.15
RUN 2 - START TO RWS BUCKET CHANGE
(21/2 HOUR RUN)
ELAPSED
TIME
(min)
0
149
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
[W]
NET
WEIGHT
(LBS)
370.38
22.90
[V]
VOC
CONTENT
(WT%)
60.29%
90.58%
[R]
RESPONSE
FACTOR
(LBSVOC/LBSC3H8)
1.500
1315
BEGINNING TOTAL:
COATING
RWS BUCKET 1
129.70
39.08
58.17%
68.28%
1.422
1.388
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
148.87
15.77
164.64
53.06
19.22
72.28
92.36
RUN 3 - START TO RWS BUCKET CHANGE
(21/3 HOUR RUN)
ELAPSED
TIME
(min)
0
139
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
[W]
NET
WEIGHT
(LBS)
414.58
18.08
[V]
VOC
CONTENT
(WT%)
60.50%
99.89%
[R]
RESPONSE
FACTOR
(LBSVOOLBSC3H8)
1.482
1.349
BEGINNING TOTAL:
COATING
RWS BUCKET 1
163.82
33.74
58.26%
70.56%
1.577
1.541
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBSC3H8)
169.24
13.39
182.63
60.52
15.45
75.97
106.66
-------
RUN 4 - START TO RWS BUCKET CHANGE
(12/3 HOUR RUN)
ELAPSED
TIME
(min)
0
98
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
[W]
NET
WEIGHT
CLBS)
359.86
18.70
COATING
RWS BUCKET 1
179.08
28.60
[V]
VOC
CONTENT
(WT%)
62.13%
99.96%
[R]
RESPONSE
FACTOR
(LBS VOOLBS C3H8)
1.483
1.268
BEGINNING TOTAL-
60.77%
65.26%
1.482
1.409
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
150.76
14.74
165.50
73.43
13.25
86.68
78.83
RUN 5 - START TO RWS BUCKET CHANGE
(2 HOUR RUN)
ELAPSED
TIME
(min)
0
127
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
[W]
NET
WEIGHT
(LBS)
391.58
20.42
[V]
VOC
CONTENT
(WT%)
61.07%
94.92%
[R)
RESPONSE
FACTOR
(LBS VOOLBS C3H8)
1.470
1.373
BEGINNING TOTAL:
COATING
RWS BUCKET 1
184.34
23.80
59.34%
75.43%
1.526
1.535
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
162.68
14.12
176.80
71.68
11.70
83.38
93.42
RUN 6 - START TO RWS BUCKET CHANGE
(2 HOUR RUN)
ELAPSED
TIME
(min)
0
120
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
[W]
NET
WEIGHT
(LBS)
395.74
15.66
[V]
VOC
CONTENT
(WT%)
60.86%
96.79%
[R]
RESPONSE
FACTOR
(LBS VOOLBS C3H8)
1.317
1.248
BEGINNING TOTAL-
COATING
RWS BUCKET 1
189-03
1530
58.72%
7929%
1.440
1.354
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
182.88
12.15
195.02
77.08
8.96
86.04
108.98
-------
RUN 7 - START TO RWS BUCKET CHANGE
(2 2/3 HOUR RUN)
ELAPSED
TIME
(min)
0
157
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
[W]
NET
WEIGHT
(LBS)
40832
19.00
COATING
RWS BUCKET 1
161.97
20.33
[V]
VOC
CONTENT
(WT%)
60.97%
99.86%
IR]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.448
1.484
BEGINNING TOTAL:
61.08%
76.72%
1.619
1.762
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
171.93
12.79
184.71
61.11
8.85
69.96
114.76
RUN 8 - START TO RWS BUCKET CHANGE
(21/4 HOUR RUN)
ELAPSED
TIME
(min)
0
131
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
[W]
NET
WEIGHT
(LBS)
398.88
17.42
[V]
VOC
CONTENT
(WT%)
60.91%
99.80%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.511
1.242
BEGINNING TOTAL:
COATING
RWS BUCKET 1
146.52
1554
60.53%
78.57%
1.512
1.548
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
160.79
14.00
174.79
58.66
7.89
66.54
108.25
RUN 9 - START TO RWS BUCKET CHANGE
(2 HOUR RUN)
ELAPSED
TIME
(min)
0
112
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 1
COATING
RWS BUCKET 1
[W]
NET
WEIGHT
(LBS)
398.68
18.06
20629
18.45
m
VOC
CONTENT
(WT%)
62.16%
99.72%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.437
1329
BEGINNING TOTAL:
59.95%
8Z87%
1.428
1.440
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
172.46
13.55
186.01
86.60
10.62
9722
88.79
-------
RUN 1 REMAINING DATA NOT INCLUDED
DUE TO SOLVENT ADDTIONS TO RWS
RUN 2 - RWS BUCKET CHANGE TO END
(1/2 HOUR RUN)
149
180
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 2
[WJ
NET
WEIGHT
(LBS)
129.70
21.86
[V]
VOC
CONTENT
(WT%)
58.17%
99.83%
IR]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.422
1.278
BEGINNING TOTAL:
COATING
RWS BUCKET 2
74.62
24.93
57.74%
84.67%
1.407
1.361
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
53.06
17.08
70.13
30.62
15.51
46.13
24.00
RUN 3 - RWS BUCKET CHANGE TO END
(2/3 HOUR RUN)
139
181
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
163.82
19.44
[V]
VOC
CONTENT
(WT%)
58.26%
99.95%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.577
1.355
BEGINNING TOTAL:
COATING
RWS BUCKET 2
92.82
23.70
57.59%
82.12%
1.606
1.420
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L3
WEIGHT OF
VOC
(LBS C3H8)
60.52
14.34
74.86
33.28
13.71
46.99
27.87
-------
RUN 4 - RWS BUCKET CHANGE TO END
(11/3 HOUR RUN)
98
180
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 2
COATING
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
179.08
17.64
[V]
VOC
CONTENT
(WT%)
60.77%
99.95%
[R]
RESPONSE
FACTOR
(LBSVOOLBSC3H8)
1.482
1.099
BEGINNING TOTAL:
25.06
22.66
59.64%
77.70%
1.481
1.364
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
73.43
16.04
89.48
10.09
12.91
23.00
66.4S
RUN 5 - RWS BUCKET CHANGE TO END
(1 HOUR RUN)
127
180
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
18434
26.16
[V]
VOC
CONTENT
(WT%)
5934%
99.95%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.526
1.239
BEGINNING TOTAL:
COATING
RWS BUCKET 2
89.74
25.90
58.61%
86.02%
1.535
1.369
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
71.68
21.10
92.79
34.26
16.27
50.54
42J25
RUN 6 - RWS BUCKET CHANGE TO END
(1 HOUR RUN)
120
180
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 2
COATING
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
189.03
18.26
99.96
19.22
W
VOC
CONTENT
(WT%)
58.72%
99.70%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.440
1.183
BEGINNING TOTAL:
57.66%
8823%
1.500
1.285
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
77.08
1539
92.47
38.42
mo
51.62
40.85
-------
RUN 7 - RWS BUCKET CHANGE TO END
(1 HOUR RUN)
157
180
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 2
COATING
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
161.97
20.18
12254
20.24
[V]
VOC
CONTENT
(WT%)
61.08%
99.93%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.619
1.609
BEGINNING TOTAL:
61.09%
91.92%
1.644
1.448
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
61.11
1253
73.64
45.54
12.85
5838
15.26
RUN 8 - RWS BUCKET CHANGE TO END
(3/4 HOUR RUN)
131
180
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
14652
20.14
[V]
VOC
CONTENT
(WT%)
6053%
99.89%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1512
1.230
BEGINNING TOTAL:
COATING
RWS BUCKET 2
62.06
16.94
6039%
9052%
1.512
1.316
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
58.66
16.36
75.01
24.79
11.65
36.44
38.57
RUN 9 - RWS BUCKET CHANGE TO END
(1 HOUR RUN)
112
178
LIQUID FEED
RESERVOIR
COATING
RWS BUCKET 2
[W]
NET
WEIGHT
(LBS)
20629
2627
m
VOC
CONTENT
(WT%)
59.95%
99.99%
[R]
RESPONSE
FACTOR
(LBS VOC/LBS C3H8)
1.428
1316
BEGINNING TOTAL:
COATING
RWS BUCKET 2
86.00
24.40
58.64%
8752%
1.422
1.306
END TOTAL:
NET LIQUID VOC CONSUMPTION:
[L]
WEIGHT OF
VOC
(LBS C3H8)
86.60
19.96
106.56
35.46
1635
51.82
54.75
-------
APPENDIX D
SECTION 4
SHORT DURATION VOC GAS CONCENTRATIONS
-------
JUL-09-1993 15:56
FROM
TO 13149282050
P. 032/885
Baseline
03/24/93
0-60 min
785
879
840
03/27/93
0-128 nrin
796
03/2g/93
0-149 mm
930
149-180 imn
828
03/28/93
0-139 min
984
139-181 min
993
03/29/93
O-98 min
1084
98-180 sain
1047
03/29/93
O-127 nriB
996
127-18O nrin
1020
03/30/93
O-120 mill
949
120-180 nrin
864
O3/3O/93
O-156 nnn
875
156-180 mirt
766
03/31/93
0-130 min
1098
130-180 mio
915
03/31/93
0-1 12 milt
971
112- 178 rain
1013
-------
JIL.-09-1993 15 = 56
FROM
TO 13149282050
P.083/005
An-HxvKtvM"^ ^•^v-rx:*.:-*>v-:<*«'- '•><»VOK>K-><:>I<->;•:• <-•.
^^^..^^^^^fcsgSSx^C.SSj-'^^J.^c::::^;,
Baseline
03/24/93
0-60 nrin
592
60-120 min
648
120-180 nrin
680
03/27/93
0-128 inin
607
03/28/93
0-149 min
517
149-1 80 min
575
03/28/93
0-139 min
603
139-181 min
620
03/29/93
0-98 min
535
98-180 mm
548
03/29/93
0-127 min
613
127-180 nrin
644
03/30/93
0-120 min
609
120-180 mm
625
03^0/93
0-156 min
622
156-18O mm
578
03/31/93
0-13Omin
613
130-180 min
592
03/31/93
0-112 min
628
112-178 nan
652
-------
JUL-09-1993 15=57
FROM
TO 13149282050
P.004/005
Baseline
03/24/93
0-60 min
N/A
6O-120 mm
N/A
N/A
03/27/93
0-128 mia
78.1
03/28^93
0-149 min
58.1
149-180 min
67.8
03/28/93
0-139 nun
73.2
139-181 min
71.4
03/29/93
0-98 min
79.7
98-180 min
71.0
03/29/93
O-127 rain
77.1
127-18O min
70.8
03/30/93
0-120 min
93.1
12O-J8Omin
86.8
03/30/93
0-156 nrin
89.3
156-180 min
77.4
03/31/93
O-13Q min
86.7
130-180 min
8S.8
03/31/93
0-112 mm
94.5
112-178 min
94.5
-------
JUL-09-1993 15--57
FROM
TO 13149282050
"'
Baseline
03/24/93
22.5
6O-J2Omifl
24.9
120-180 mm
27.1
03/27/93
O-128 min
25.7
03/28/93
0-149 nrin
10.9
149-180 min
13.6
03/28/93
0-139 min
15.2
139-181 mm
15.1
03/29/93
0-98 mm
24.0
98-180 min
19.7
03/29/93
0-127 iron
20.7
127-180 min
19.8
03/30/93
0-120 mto
25.8
12O-J8Omin
29.6
03/30/93
0-156 HBD
31.4
156-1 80 mia
31.3
03/31/93
O-13Omiii
22-0
23.5
03/31/93
0-112 mill
30.1
112-178 mia
28.7
TDTflL P.005
-------
APPENDIX D
SECTION 5
CALCULATION OF SHORT DURATION VOC GASEOUS EMISSIONS
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN1
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppmC3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppmCSHS)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppmC3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ L 1 TOTAL LIQUID VOC
CONSUMPTION (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
LIQUID/GAS MASS BALANCE
CAPTURE EFFICIENCY
FIRST SEGMENT OF TEST RUN
START TO RWS BUCKET CHANGE
(2Hrs-9Min)
OVEN
EXHAUST
(TS-1)
796.0
25.7
144.0
3.1
50.1
720.2
6142
128.7
65.02
CAPTURE
HOODS
(TS-2)
607.0
25.7
581.3
2181
128.7
18.63
FUGITIVE
DUCT
(TS-3)
78.1
25.7
52.4
6830
128.7
5.26
83.65
90.15
5.26
92.8%
SECOND SEGMENT OF TEST RUN
RWS BUCKET CHANGE TO END
OVEN
EXHAUST
(TS-1)
CAPTURE
HOODS
(TS-2)
FUGITIVE
DUCT
(TS-3)
NOT REPORTED
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN 2
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppmC3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G 1 TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ L ] TOTAL LIQUID VOC
CONSUMPTION (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
LIQUID/GAS MASS BALANCE
CAPTURE EFFICIENCY
FIRST SEGMENT OF TEST RUN
START TO RWS BUCKET CHANGE
(2Hre-30Min)
OVEN
EXHAUST
(TS-1)
930.0
10.9
125.0
2.7
43.5
875.6
5679
149.5
84.90
CAPTURE
HOODS
(TS-2)
517.0
10.9
506.1
2083
149.5
18.00
FUGITIVE
DUCT
(TS-3)
58.1
10.9
47.2
6859
149.5
5.53
102.89
92.36
5.53
111.4%
SECOND SEGMENT OF TEST RUN
RWS BUCKET CHANGE TO END
(30 Min)
OVEN
EXHAUST
(TS-1)
828.0
13.6
43.5
L 770.9
5679
30.5
15.25
CAPTURE
HOODS
(TS-2)
575.0
13.6
561.4
2083
30.5
4.07
FUGITIVE
DUCT
(TS-3)
67.8
13.6
54.2
6859
30.5
1.29
19.32
24.00
1.29
80.5%
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUNS
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ L ] TOTAL LIQUID VOC
CONSUMPTION (Ibs C3H8)
1 F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
LIQUID/GAS MASS BALANCE
CAPTURE EFFICIENCY
FIRST SEGMENT OF TEST RUN
START TO RWS BUCKET CHANGE
(2 Hrs - 19 Min)
OVEN
EXHAUST
(TS-1)
984.0
15.2
114.0
2.5
39.7
929.1
5807
139.3
85.83
CAPTURE
HOODS
(TS-2)
603.0
15.2
587.8
2063
139.3
19.29
FUGITIVE
DUCT
(TS-3)
73.2
15.2
58.0
6855
139.3
6.32
105.12
106.66
6.32
98.6%
SECOND SEGMENT OF TEST RUN
RWS BUCKET CHANGE TO END
(41 Min)
OVEN
EXHAUST
(TS-1)
993.0
15.1
39.7
938.2
5807
41.7
25.94
CAPTURE
HOODS
(TS-2)
620.0
15.1
604.9
2063
41.7
5.94
FUGITIVE
DUCT
(TS-3)
71.4
15.1
56.3
6855
41.7
1.84
31.89
27.87
1.84
114.4%
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN 4
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ L ] TOTAL LIQUID VOC
CONSUMPTION (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
LIQUID/GAS MASS BALANCE
CAPTURE EFFICIENCY
FIRST SEGMENT OF TEST RUN
START TO RWS BUCKET CHANGE
(lHr-38Min)
OVEN
EXHAUST
(TS-1)
1084.0
24.0
118.0
2.5
41.0
1019.0
5547
98
63.26
CAPTURE
HOODS
(TS-2)
535.0
24.0
511.0
2038
98
11.65
FUGITIVE
DUCT
(TS-3)
79.7
24.0
55.7
6754
98
4.21
74.91
78.83
4.21
95.0%
SECOND SEGMENT OF TEST RUN
RWS BUCKET CHANGE TO END
(lHr-22Min)
OVEN
EXHAUST
(TS-1)
1047.0
19.7
41.0
986.3
5547
82
51.23
CAPTURE
HOODS
(TS-2)
548.0
19.7
528.3
2038
82
10.08
FUGITIVE
DUCT
(TS-3)
71.0
19.7
51.3
6754
82
3.24
61.31
66.48
3.24
92.2%
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUNS
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppmC3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ L ] TOTAL LIQUID VOC
CONSUMPTION (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
LIQUID/GAS MASS BALANCE
CAPTURE EFFICIENCY
FIRST SEGMENT OF TEST RUN
START TO RWS BUCKET CHANGE
(2H»-7Min)
OVEN
EXHAUST
(TS-1)
996.0
20.7
126.0
2.8
43.9
931.4
5501
126.8
74.19
CAPTURE
HOODS
(TS-2)
613.0
20.7
592.3
1963
126.8
16.84
FUGITIVE
DUCT
(TS-3)
77.1
20.7
56.4
6717
126.8
5.49
91.03
93.42
5.49
97.4%
SECOND SEGMENT OF TEST RUN
RWS BUCKET CHANGE TO END
(53 Min)
OVEN
EXHAUST
(TS-1)
1020.0
19.8
43.9
956.3
5501
53.2
31.96
CAPTURE
HOODS
(TS-2)
644.0
19.8
624.2
1963
53.2
7.44
FUGITIVE
DUCT
(TS-3)
70.8
19.8
51.0
6717
53.2
2.08
39.40
42.25
2.08
93.3%
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN 6
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppmCSHS)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
I G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ L J TOTAL LIQUID VOC
CONSUMPTION (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
LIQUID/GAS MASS BALANCE
CAPTURE EFFICIENCY
FIRST SEGMENT OF TEST RUN
START TO RWS BUCKET CHANGE
(2Hrs-lMin)
OVEN
EXHAUST
(TS-1)
949.0
25.8
39.8
1.0
13.9
909.3
5923
120.5
74.11
CAPTURE
HOODS
(TS-2)
609.0
25.8
583.2
2047
120.5
16.43
FUGITIVE
DUCT
(TS-3)
93.1
25.8
67.3
6811
120.5
6.31
90.54
108.98
6.31
83.1%
SECOND SEGMENT OF TEST RUN
RWS BUCKET CHANGE TO END
(59 Min)
OVEN
EXHAUST
(TS-1)
864.0
29.6
13.9
820.5
5923
59.5
33.02
CAPTURE
HOODS
(TS-2)
625.0
29.6
595.4
2047
59.5
8.28
FUGITIVE
DUCT
(TS-3)
86.8
29.6
57.2
6811
59.5
2.65
41.30
40.85
2.65
101.1%
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN 7
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppmC3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ L ] TOTAL LIQUID VOC
CONSUMPTION (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
LIQUID/GAS MASS BALANCE
CAPTURE EFFICIENCY
FIRST SEGMENT OF TEST RUN
START TO RWS BUCKET CHANGE
(2Hrs-37Min)
OVEN
EXHAUST
(TS-1)
875.0
31.4
38.0
1.0
13.3
830.3
5722
156.8
85.07
CAPTURE
HOODS
(TS-2)
622.0
31.4
590.6
2067
156.8
21.86
FUGITIVE
DUCT
(TS-3)
89.3
31.4
57.9
6717
156.8
6.96
106.93
114.76
6.96
93.2%
SECOND SEGMENT OF TEST RUN
RWS BUCKET CHANGE TO END
(23 Min)
OVEN
EXHAUST
(TS-1)
766.0
31.3
13.3
721.4
5722
23.2
10.94
CAPTURE
HOODS
(TS-2)
578.0
31.3
546.7
2067
23.2
2.99
FUGITIVE
DUCT
(TS-3)
77.4
31.3
46.1
6717
23.2
0.82
13.93
15.26
0.82
91.3%
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUNS
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppm C3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppm C3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppmC3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
t L ] TOTAL LIQUID VOC
CONSUMPTION (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
LIQUID/GAS MASS BALANCE
CAPTURE EFFICIENCY
FIRST SEGMENT OF TEST RUN
START TO RWS BUCKET CHANGE
(2 Hrs - 11 Min)
OVEN
EXHAUST
(TS-1)
1098.0
22.0
39.0
1.0
13.7
1062.3
5784
130.5
91.57
CAPTURE
HOODS
(TS-2)
613,0
22.0
591.0
2065
130.5
18.19
FUGITIVE
DUCT
(TS-3)
86.7
22.0
64.7
6787
130.5
6.54
109.76
108.25
6.54
101.4%
SECOND SEGMENT OF TEST RUN
RWS BUCKET CHANGE TO END
(49 Min)
OVEN
EXHAUST
(TS-1)
915.0
23.5
13.7
877.8
5784
49.5
28.70
CAPTURE
HOODS
(TS-2)
592.0
23.5
568.5
2065
49.5
6.64
FUGITIVE
DUCT
(TS-3)
88.8
23.5
65.3
6787
49.5
2.51
35.34
38.57
2.51
91.6%
-------
SUMMARY OF GAS PHASE MEASUREMENTS
RUN 9
MEASUREMENT
PARAMETER
AVERAGE CORRECTED VOC
CONCENTRATION (ppmC3H8)
AVERAGE BACKGROUND VOC
CONCENTRATION (ppmC3H8)
METHANE (ppm)
ETHANE (ppm)
UNBURNED NATURAL GAS
CONCENTRATION (ppm C3H8)
ADJUSTED AVERAGE VOC
CONCENTRATION (ppm C3H8)
VOLUMETRIC GAS FLOW
RATE (wscfm)
SAMPLING TIME (min)
VOC EMISSIONS (Ibs C3H8)
[ G ] TOTAL CAPTURED VOC
EMISSIONS (Ibs C3H8)
[ L 1 TOTAL LIQUID VOC
CONSUMPTION (Ibs C3H8)
[ F ] TOTAL FUGITIVE VOC
EMISSIONS (Ibs C3H8)
LIQUID/GAS MASS BALANCE
CAPTURE EFFICIENCY
FIRST SEGMENT OF TEST RUN
START TO RWS BUCKET CHANGE
(lHr-52Min)
OVEN
EXHAUST
(TS-1)
971.0
30.1
42.8
1.0
14.9
926.0
5639
112
66.78
CAPTURE
HOODS
(TS-2)
628.0
30.1
597.9
1993
112
15.24
FUGITIVE
DUCT
(TS-3)
94.5
30.1
64.4
6724
112
5.54
82.02
88.79
5.54
92.4%
SECOND SEGMENT OF TEST RUN
RWS BUCKET CHANGE TO END
(lHr-8Min)
OVEN
EXHAUST
(TS-1)
1013.0
28.7
14.9
969.4
5639
66
41.20
CAPTURE
HOODS
(TS-2)
652.0
28.7
623.3
1993
66
9.36
FUGITIVE
DUCT
(TS-3)
94.5
28.7
65.8
6724
66
3.33
50.56
54.75
3.33
92.4%
-------
APPENDIX E
MOSTARDI-PLATT ASSOCIATES' QUOTES ON CE/DE TESTING
-------
MOSTARDI-PLATT ASSOCIATES, INC.
Environmental Consultants
RECEIVED
June 14, 1993
sff I 6 1993
Proposal No. 007219 ERM-NORTH CENTRAL
ST. CHARLES, MISSOURI
Can Manufacturers Institute
1625 Massachusetts Avenue, N.W.
Washington, D. C. 20036
Attention: Mr. Matt Middaugh
Gentlemen:
VOC Capture and Destruction Efficiency Testing
In response to your request for professional testing services to be performed at a three piece
can plant in the Chicagoland area, MOSTARDI-PLATT ASSOCIATES, INC. (MPA) is
pleased to submit the following proposal no. 007219.
Proposal Outline
I.
A. Capture Efficiency Tests Gas/Gas Method using a Total Temporary Enclosure
(TIE).
1. Baseline: One one-hour baseline test will be performed at the oven exhaust,
coating hood exhaust, cooling zone, and breathing zone near the coating area
in accordance with EPA Methods 1, 2, (3 & 4 at the oven only) and 25A.
2. Balancing: Following the enclosure fabrication , approximately one hour of
monitoring will be performed to verify that the VOC concentration in the
enclosure is below the specified acceptable level and to determine if the
afterburner inlet VOC concentration is within 10% of the baseline
measurements.
3. Capture Efficiency Tests: Three, three-hour capture efficiency test runs will
be performed at the oven exhaust duct, coating hood exhaust duct, fugitive
exhaust duct, six natural draft openings, four breathing zone ambient points
near the coating area, and possibly the cooling zone exhaust stack, in
945 OaWawn Avenue • Elmhurst, IL 60(26-1012 • (70S] 993-9000 « Facsimile: (708J 993-9017
-------
Can Manufacturers Institute
June 14, 1993
PN 007219
Page 2
accordance with EPA Methods 1, 2 (3 & 4 at the oven only) and 25A,
40CFR, Part 60, Appendix A as currently revised.
a. One integrated tedlar bag sample will be taken during each test run at
the oven exhaust and analyzed for methane and ethane concentrations
by (AC/FD).
B. Capture Efficiency Tests - Liquid to gas phase material balance without TTE.
1. One hour of baseline background VOC and methane measurements will be
performed at the oven exhaust duct in accordance with EPA Methods 1-4
and 25A,, 40CFR, Part 60, Appendix A.
a. Two integrated tedlar bag 'samples will be taken during the baseline
test period and analyzed for methane and ethane concentrations by
AC/FD.
2. Three three-hour VOC test runs will be performed simultaneously at the
oven exhaust and capture hood exhaust ducts in accordance with proposed
EPA Method 30B, 40CFR, Part 60, Appendix A and EPA Methods 1-4, 40
CFR, Part 60, Appendix A.
3. Two coating and two solvent samples (one of each before and after each
run) will be taken and analyzed for VOC content in twelve samples total in
accordance with proposed EPA Method 30F, 40 CFR, Part 60, Appendix
A.
C. Destruction Efficiency Tests
1. Three one-hour VOC test runs will be performed simultaneously at the oven
exhaust, capture hood exhaust and afterburner stack in accordance with EPA
Methods 1-4 and 25A, 40 CFR, Part 60, Appendix A.
D. Per diem and travel expenses are included in the pricing.
E. Equipment calibration and rental are included in the pricing.
-------
Can Manufacturers Institute
June 14, 1993
PN 007219
Page 3
D. Quality control procedures used for this project can be found in but are not limited
to the following:
1. Item LA. above.
2. Q. A. Handbook, EPA 600/9-76-005, Vol. I.
3. Q. A. Handbook, EPA 600/4-77-027b, Vol. m.
F. 20 copies of the final report will be forwarded.
G. Pricing
1. Lump Sum Gas/Gas TTE Capture Testing per scope and
schedule $28,400.00
Additional if cooling zone requires monitoring $8,025.00
Additional for Destruction Efficiency Testing during program $2,450.00
Additional for test coordinator to monitor process line
during program $2,502.00
2. Lump sum Liquid to Gas phase capture testing per scope
and schedule $11,335.00
Additional for Destruction Efficiency Testing during program $2,450.00
Additional for testing coordinator to monitor process line and take samples
during program $1,562.00
3. Lump sum Destruction Efficiency Testing per scope and
schedule $5,900.00
. Test Program Schedule
The prices quoted above are based on the following schedule:
A. Gas/Gas TTE Capture Tests
-------
Can Manufacturers Institute
June 14, 1993
PN 007219
Page 4
Day
1
2
3
4
5
6
N < ' Task "
Set-up equipment and run baseline tests and
cooling zone evaluation
Set-up enclosure by others
Set-up enclosure by others
Set-up equipment and conduct balancing test
Run first test
Run second and third test (and Destruction
Test if required)
On Site Hours
7
~
~
8
8
10
B. Liquid/Gas without the capture test
Day
1
2
Task
Set-up equipment and perform background
measurements and first test run
Run second and third test (and Destruction
Test if required)
On Site Hours
8
10
C. Destruction Test Only
Day
1
Task
Set-up equipment and run all tests
On Site Hours
8
(All test days are considered consecutive 8 hour weekdays generally scheduled between the
hours of 8:00 A.M. and 4:30 P.M. unless otherwise stated.)
-------
Can Manufacturers Institute
June 14, 1993
PN 007219
PageS
, Project Delays
Project delays caused by circumstances not within the control of MPA and project extensions
resulting from client requests are chargeable according to the enclosed price list.
A change order for monies chargeable due to delays, extensions or changes will be required
prior to the submittal of the final report.
IV.
Gas/Gas TTE
1 Project Manager
1 Test Supervisor
1 Field Chemist
1 Test Engineer
1 Test Technician
Liquid/Gas without TTE
1 Test Supervisor
1 Field Chemist
1 Test Engineer
1 Test Technician
Destruction
1 Test Supervisor
3 Test Engineers
V. Indemnification
To the fullest extent not prohibited by law, CMI, its member companies and employees shall
indemnify and hold harmless Mostardi-Platt Associates, Inc., (MPA) its directors, officers, agents
and employees from and against any and all claims, damages, losses and expenses (including but
not limited to attorney's fees) arising by reason of any act or failure to act, negligent or otherwise,
of contractor, of any subcontractor (meaning anyone, including but not limited to consultants
having a contract with contractor or a subcontractor for part of the services), of anyone directly
or indirectly employed by contractor or by any subcontractor, or of anyone for whose acts the
contractor or its subcontractor may be liable, in connection with providing these services, including
the testing and operation of the temporary total enclosure for capture efficiency testing.
VI. Client Assistance
MPA must be supplied with the following items in order to complete this test program:
-------
Can Manufacturers Institute
June 14, 1993
PN 007219
Page 6
Note: It is the responsibility of Can Manufacturing, Inc. to convey these responsibilities
to the plant at least one week prior to the scheduled test week.
1) A process operation description and flow diagram and layout drawings of process and
test location must be provided if a test protocol is required.
2) Safe access to test positions.
3) Multiple circuits of electrical power 1 lOv and 30 amp service at the test locations (180
amp total).
4) Two inch test ports cleaned and loose prior to arrival of test crew.
5) Any scaffolding required to perform the tests.
6) Sufficient lighting at the test site.
7) Safety belts, if required.
8) Elevators safety checked and certified in good operation.
9) Hoist equipment, if required.
10) Plant or pollution control equipment operating data if required for report.
11) Fuel samples, if required. (Analyses cost not included unless listed)
12) Washroom facilities for use by members of the test crew.
13) Laboratory facilities (if required) including a small workspace for processing samples.
14) A shelter at the test location if weather conditions warrant.
15) Plant assistance in hoisting equipment to and from test site.
16) Communication between the test location and the control room.
17) Continuous coating line operation for each test run each day.
18) Pre and post test coating and solvent samples and exact usage during each test run for
capture efficiency determinations.
Note: It is the responsibility of Can Manufacturing, Inc. to notify all appropriate regulatory
agencies of the scheduled test date(s). MPA can notify the required agencies if
requested in writing.
-------
MOSTARDI-PLATT ASSOCIATES, INC.
Environmental Consultants
Can Manufacturers Institute
June 14, 1993
PN 007219
Page 7
All invoices are billed net Payments not received within thirty (30) days from the date billed are
subject to a late payment charge of 1.5% per month until payment is received.
MOSTARDI-PLATT ASSOCIATES, INC. (MPA) has enjoyed being of service to Can
Manufacturers Institute in the past and looks forward to working with you on this project.
If you have any questions regarding this proposal, or require additional information, please contact
David A. Ozawa at (708) 860-5900.
Respectfully submitted,
MOSTARDI-PLATT ASSOCIATES, INC.
Thomas D. Mostardi, PE
Associate
THIS PROPOSAL NO. 007219 IS GOOD IF ACCEPTED WITHIN 30 DAYS AND FOR WORK
PERFORMED WITHIN 30 DAYS OF ACCEPTANCE.
FURTHERMORE, THIS PROPOSAL IS FOR THE ONE TIME COMPLETION OF WORK AS
SCHEDULED AND FTXED QUOTED PRICES LISTED DO NOT APPLY TO ADDITIONAL
TESTS OR RETESTS UNLESS OTHERWISE STATED HEREIN.
DAO/blt
------- |