UnhedSiaies ' ' EPA-600/R-97-018a
Environmental Protection
Asencv March 1997
&EPA Research and
Development
EVALUATION OF POLLUTION PREVENTION
TECHNIQUES TO REDUCE STYRENE EMISSIONS
FROM OPEN CONTACT MOLDING PROCESSES
Volume I, Final Report
Prepared for
Office of Air Quality Planning and Standards
Prepared by
National Risk Management
Research Laboratory
Research Triangle Park, NC 27711
-------
FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with pro-
tecting the Nation's land, air, and water resources. Under a mandate of national
environmental laws, the Agency strives to formulate and implement actions lead-
ing to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, EPA's research
program is providing data and technical support for solving environmental pro-
blems today and building a science knowledge base necessary to manage our eco-
logical resources wisely, understand how pollutants affect our health, and pre-
vent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for
investigation of technological and management approaches for reducing risks
from threats to human health and the environment. The focus of the Laboratory's
research program is on methods for the prevention and control of pollution to air,
land, water, and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites and groundwater; and prevention and
control of indoor air pollution. The goal of this research effort is to catalyze
development and implementation of innovative, cost-effective environmental
technologies; develop scientific and engineering information needed by EPA to
support regulatory and policy decisions; and provide technical support and infor-
mation transfer to ensure effective implementation of environmental regulations
and strategies.
This publication has been produced as part of the Laboratory's strategic long-
term research plan. It is published and made available by EPA's Office of Re-
search and Development to assist the user community and to link researchers
with their clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
EPA REVIEW NOTICE
This report has been peer and administratively reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22161.
-------
EPA-600/R-97-Ol8a
March 1997
Evaluation of Pollution Prevention
Techniques to Reduce Styrene Emissions
from Open Contact Molding Processes
Volume I, Final Report
By
Emery J. Kong, Mark A. Bahner, Robert S. Wright, and C. Andrew Clayton
Research Triangle Institute
. P.O. Box 12194
Research Triangle Park, NC 27709
EPA Cooperative Agreement CR 818419-03
EPA Project Officer: Geddes Ramsey
Air Pollution Prevention and Control Division
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Prepared for
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
-------
Abstract
Pollution prevention options to reduce styrene emissions, such as new materials and
application equipment, are commercially available to the operators of open molding processes.
However, information is needed about the percent reduction in emissions that is achievable with
these options.
To meet this need, several of these pollution prevention options were examined. Options
examined were operator techniques, air flow velocities in the spraying area, gel coat and resin
formulations, and application equipment. Styrene emission factors calculated from this test
result were compared with the existing AP-42 emission factors for gel coat sprayup and resin
applications.
The study found that using controlled spraying (i.e., reducing overspray), low-styrene and
styrene-suppressed materials, and nonatomizing application equipment can reduce styrene
emissions from 11 to 52 percent. Facilities should investigate the applicability and feasibility of
these pollution prevention options to reduce their styrene emissions. The calculated emission
factors were from 1.6 to 2.5 times the mid-range AP-42 emission factors for the corresponding
gel coat and resin application. These results indicate that facilities using existing AP-42
emission factors to estimate emissions in open molding processes are likely to underestimate
actual emissions.
11
-------
Contents
Section
Page
Abstract ii
Figures v
Tables vi
Acronyms and Abbreviations vii
Conversion Table viii
Acknowledgments ix
Chapter 1 Introduction : 1
1.1 Background 1
1.2 Objectives 2
1.3 Approach 2
1.4 Report Outline 2
References 3
Chapter 2 Experimental Test Design 4
2.1 : Pilot Experiment .4
2.2 Gel Coat Experiment : 5
2.3 Resin Experiment 6
Chapter 3 Facility and Experimental Setup 8
3.1 Total Enclosure System 8
3.2 Emission Sampling Location 8
3.3 Experimental Setup 10
3.3.1 FRPMold 10
3.3.2 Air Flow Baffle 10
3.3.3 Glass Veil and Kraft Paper to Capture Overspray ..: 10
3.3.4 Protective Equipment and Clothing for Operator 12
3.4 Resin Property Testing Laboratory 12
Chapter 4 Materials and Equipment : 13
4.1 Properties of Gel Coat and Resin Materials 13
4.2 Setting of Gel Coat and Resin Application Equipment 15
4.3 Reinforcements 17
Chapter 5 Determination of Emission Quantities 18
5.1 Emission Measurement Method 18
5.1.1 Determination of Styrene Concentrations in Exhausted Air 18
5.1.2 Exhaust Air Flow Rate Measurement 21
5.1.3 Emissions Determined by Emission Measurement Method 22
5.2 Mass Balance Calculation Method Using Gravimetric Measurements 23
111
-------
Contents (continued)
Section Page
Chapter 6 Test Results and Discussion 27
6.1 Distribution of Total Emissions During Application and Curing Stages 32
6.2 Pilot Experiment Results 36
6.3 Gel Coat Experiment Results 39
6.4 Resin Experiment Results 43
6.5 Comparison of Test Results with EPA AP-42 Emission Factors 49
6.6 Comparison of Emission Measurement Method and Mass Balance
Calculation Method 52
6.6.1 Pure Styrene Evaporation Test 52
6.6.2 Comparison of Emissions Measured by Two Test Methods 52
Chapter 7 Data Quality Issues '. 58
7.1 Summation of Project QA Activities 58
7.2 RTI Internal Technical System Audit (TSA) Results 58
7.3 EPA Performance Evaluation ; 59
Chapter 8 Conclusions and Recommendations 60
8.1 Conclusions 60
8.2 Recommendations 61
Appendixes (in Volume 2)
A. A Category HI Quality Assurance Project Plan for the Evaluation of Pollution Prevention
Techniques to Reduce Styrene Emissions from Open Contact Molding Processes
B. Reichhold Standard Test Methods
C. Verification and Intercomparison of Compressed Gas Calibration Standards
D. Summary of Calibration Data, Calibration Error Tests, and Drift Checks
E.. A Summary of Emission Measurements, Gravimetric Measurements, and Calculated
Emission Quantities and Emission Factors for the Test
F. Statistical Analyses of Test Results
G. THC Analyzer Evaluation: Sampling Line Loss and Pressure Effect
H. An RTI Technical System Audit Report
I. EPA Performance Evaluation of Total Hydrocarbon Analyzer
IV
-------
Figures
No. page
3-1 Side view of the Reichhold Chemicals spray booth and the experimental setup in
a permanent total enclosure 9
3-2 Sketch of a male mold 11
5-1 Data recording sheet for mass balance calculation 26
6-1 Distribution of emissions in the pilot and gel coat experiments .... 34
6-2 Distribution of emissions in the resin experiment 35
6-3 Typical emission concentration profiles for normal and controlled gel coat spraying .. 37
6-4 Emission reductions observed in the pilot experiment 38
6-5 Typical emission concentration profiles for regular and low-VOC gel coats 40
6-6 . Typical emission concentration profiles for three types of gel coat spray guns 41
6-7 Emission reductions observed in the gel coat experiment 42
6-8 Emission concentration profiles observed for different resin application techniques
and equipment 45
6-9 Typical emission concentration profiles for various resin formulations 46
6-10 Typical emission concentration profiles for low-profile, neat BPO, and modified
BPO resins 47
6-11 Emission reductions observed in the resin experiment: "controlled spraying"
technique baseline 50
6-12 Emission reductions observed in the resin experiment: "normal spraying"
technique baseline 51
6-13 Comparison of mass balance with EPA Method 25A emissions measurements,
during pure (100 percent) styrene evaporation tests 54
6-14 Comparison of emission measurement methods 56
-------
Tables
No. Page
2-1 Test Runs for Pilot Experiment 5
2-2 Test Runs for Gel Coat Experiment 6
2-3 Test Runs for Resin Experiment .7
4-1 Gel Coat Properties Measured at Reichhold 13
4-2 Resin Properties Measured at Reichhold 14
4-3 Summary of Gel Coat Equipment Setting 15
4-4 Summary of Resin Application Equipment Setting 16
6-1 Summary of Emission Measurements and Calculated Emission Factors 28
6-2 Significance of Variables Affecting Total Styrene Emissions 31
6-3 Distribution of Emissions in Pilot and Gel Coat Experiments 32
6-4 Distribution of Emissions in Resin Experiment 33
6-5 Summary of Emissions for Normal and Controlled Gel Coat Spraying 36
6-6 Summary of Emissions for. Regular and Low-VOC Gel Coats 39
6-7 .Summary of Emissions for Resin Application Equipment 43
6-8 Comparison of Emissions among Various Resin Application Equipment and Normal
Spraying , 44
6-9 Summary of Emissions for Various Resin Formulations 48
6-10 Comparison of EPA AP-42 Emission Factors and Test Results 49
6-11 Comparison of Mass Balance Method and Emission Measurements for Pure
Styrene Evaporation Tests : 53
6-12 Comparison of Mass Balance Method and Emission Measurements 55
6-13 Comparison of Mass Balance and Emission Measurement Test Results 57
VI
-------
Acronyms and Abbreviations
AAA
ANOVA
AP-42
APPCD
AS
BPO
CCP
DCPD
EPA
FID
FRP/C
FS
HVLP
ISO
MACT
MEKP
MMA
MnTAP
NDO
NRMRL
NV
NVS
ORTHO
OSHA
PFA
QA
QAPP
THC
TRI
TSA
VOC
VS
air-assisted airless
analysis of variance
EPA's Compilation of Air Pollutant Emissions Factors
Air Pollution Prevention and Control Division
available styrene
benzoyl peroxide
Cook Composites and Polymers
dicyclopentadiene
Environmental Protection Agency
flame ionization detector
fiberglass-reinforced plastics/composites
full scale
high volume, low pressure
isophthalic acid
maximum achievable control technology
methyl ethyl ketone peroxide
methyl methacrylate
Minnesota Technical Assistance Program
natural draft opening
National Risk Management Research Laboratory
nonvolatile
non-vapor-suppressed
orthophthalic
Occupational Safety and Health Administration
perfluoroalkoxy
quality assurance
Quality Assurance Project Plan
total hydrocarbon
Toxics Release Inventory
technical system audit
volatile organic compound
vapor-suppressed
VII
-------
Conversion Table
The U.S. Environmental Protection Agency policy is to express all measurements in Agency
documents in metric units. In this report, however, to conform to industry convention, English
units are used. Conversion factors from English to metric units are given below.
English Unit
ftVmin
oF
ft
. ft2
gal/min
in. H2O
Ib
psia
ton
Multiply by
0.028314
(°F-32)/1.8
0.304
0.0929
.3.79
1.87
0.454
6.895
0.907
To Obtain
mVmin
°C
m
m2
L/min
. mmHg
kg
kilopascal
Mg
Vlll
-------
Acknowledgments
During the research period, the authors received valuable information from Bob Lacovara
of the Composites Fabricators Association and many representatives in the fiberglass-reinforced
plastics/composites industry. The information helped the authors to design a focused test.
The authors would like to acknowledge the invaluable support and assistance of the
following organizations:
• Reichhold Chemicals for providing a spray booth and all resin materials for the test and
the laboratory support for gel coat and resin properties determination
• Magnum Industries for providing equipment and an experienced operator (Charles Stard)
to conduct the actual gel coat and resin application during the entire test period
• Cook Composites and Polymers for providing all gel coat materials
• PPG Industriesvlnc., for providing all the fiberglass materials for the test.
Without their generous support, the testing described in this report would not have been possible.
IX
-------
Chapter 1
Introduction
1.1 Background
The open contact molding process is one of the most common production processes used
by the fiberglass reinforced plastics/composites (FRP/C) and FRP boat building industry. This
process is used to manufacture boats, bathtubs, shower stalls, truck cabs, body panels for
recreational vehicles and trucks, swimming pools, etc. It is one of the FRP/C processes that
consumes the most polyester resins. It also has the greatest potential of emitting styrene due to
the spraying equipment used and the openness of the process. According to the 1990 Toxics
Release Inventory (TRI) database, 10,600 tons of styrene were emitted from the FRP/C and boat
building industries. More than 50 percent of the total styrene emissions was emitted from the
open molding process.
Styrene is emitted during the application stage when a catalyzed gel coat or resin is
applied to the surface of an open mold. Styrene continues to be emitted from wet gel coat or
resin during gelation and curing. The open contact molding process usually is conducted in a
facility with ample ventilation to maintain the ambient styrene concentrations under the current
Occupational Safety and Health Administration (OSHA) standard of 100 ppm. Therefore,
styrene emissions from the open contact molding process are difficult to capture and control.
The maximum achievable control technology (MACT) standards for the reinforced
plastics/composites source category and boat building source category are scheduled to be
promulgated by November 15, 1997, and November 15,2000, respectively. For some open
contact molding processes, pollution prevention techniques could be used to reduce styrene
emissions. These pollution prevention techniques include changing application equipment and
environment and using different gel coat or resin formulations. Existing information indicates
that using nonspraying equipment or low-emitting/high-transfer efficiency spray guns, such as
air-assisted airless (AAA) or high-volume, low-pressure (HVLP) spray guns, can reduce
emissions from the application stage. Gel coat and resin manufacturers also have developed
different gel coat and resin formulations to reduce emissions. The effects of these pollution
prevention techniques have not been compared systematically.
Limited studies provide some indications that low-styrene resins can reduce emissions
when compared to regular general-purpose resin. A demonstration project entitled Reducing
Styrene Emissions in Fiber Reinforced Plastics Operations' was conducted by the Minnesota
Technical Assistance Program in the early 1990s. The study found that styrene emissions from
low-styrene resins were reduced by 25 to 45 percent compared to a conventional orthophthalic
(ORTHO)-based general-purpose resin. However, the emissions measured from the simulated
production trials were not a typical open molding process, and the emissions quantified may not
directly apply to actual operations to estimate styrene emissions. A Finland research group2
reports that low-styrene resin reduced total styrene evaporation by 30 to 60 percent compared to
1
-------
standard resin. However, their test was conducted for hand layup operation under a laboratory
hood, and the results cannot be applied to sprayup operations.
1.2 Objectives
This research project has three objectives. The first objective is to quantify and validate
the effectiveness of several pollution prevention techniques, specifically operator techniques, air
flow velocity in the spraying zone, different gel coat and resin formulations, and application
equipment, on styrene emissions from the open contact molding process. The second objective is
to compare a mass balance calculation method with an emission measurement method to quantify
emissions. The third objective is to compare emission factors calculated from this test with the
emission factors for gel coat sprayup and resin applications reported in the U.S. Environmental
Protection Agency (EPA) AP-42 document3 to determine the accuracy of the AP-42 emission
factors. The results of this study are to be presented to the FRP/C and boat building industries so
that individual facilities can identify the most effective and practical pollution prevention
techniques to reduce their styrene. emissions.
1.3 Approach
This test determined the styrene emission reduction from baseline conditions for several
pollution prevention techniques on open contact molding processes. The baseline emissions .
were determined for a typical gel coat and a general purpose resin using a AAA spray gun under
typical environmental and operating conditions. Pollution prevention techniques were evaluated
for gel coat and resin applications under the same environmental conditions. The effectiveness
of these pollution prevention techniques is determined by comparing total styrene emissions (in
grams) and styrene emission factors, expressed as the weight percent of available styrene (% AS)
and as mass per unit mold surface area (g/m2). The former unit (%AS) is the unit used in EPA' s
AP-42 emission factors. The EPA Air Pollution Prevention and Control Division (APPCD)
Category HI quality assurance.(QA) procedures were followed to ensure that the data quality is
sufficient to evaluate the effectiveness of these materials and equipment. The QA project plan
(QAPP) for this testing is included in Appendix A (Volume n).
1.4 Report Outline
This report in divided into two volumes. Volume I documents the planning, execution,
and findings of the pollution prevention technique evaluation test. Chapter 2 presents the
experimental design. Chapter 3 describes the facility and the setup for the testing. Chapter 4
describes the testing procedures used to quantify emissions from the pperation. Chapter 5
presents the materials and equipment used in the testing. Chapters 6 and 7 present the results of
the testing and the associated data quality issues, respectively. Chapter 8 summarizes the
conclusions from the research and presents the recommendations to the industry. Volume n
contains the appendixes to this report detailed supporting documents that are related to data
quality and emission measurement issues.
-------
References
1. Minnesota Office of Waste Management. Reducing Styrene Emissions in Fiber
Reinforced Plastics Operations. Minnesota Technical Assistance Program (MnTAP),
Minneapolis, Minnesota, 1993.
2. Saamamen, A.J., R.I. Miemela, T.K. Blomqvist, and E.M. Nikander. Emission of
Styrene During the Hand Lay-up Molding of Reinforced Polyester. Applied Occupation
Environmental Hygiene, 6(9): 790-793, September 1991.
3. U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards.
Compilation of Air Pollutant Emission Factors (AP-42). Research Triangle Park, NC.
p. 4.12-1, September 1988.
-------
Chapter 2
Experimental Test Design
This testing included a pilot experiment, a gel coat experiment, and a resin experiment.
From the pilot experiment, the linear air flow velocity in the spraying zone and the spraying
technique were evaluated and a set of conditions was selected for the subsequent gel coat and
resin experiments. The gel coat experiment examined two gel coat formulations with three
pieces of gel coating equipment (i.e., spray guns). The resin experiment examined five resin
formulations and three pieces of resin application equipment. Except for a styrene-suppressed
resin with additional wax, the rest of the gel coat and resin formulations and application
equipment selected for the testing are commercially available to the FRP/C and boat building
industries. Each of the experiments is described in the following subsections.
2.1 Pilot Experiment
Before these formulations and equipment were examined, the effects on styrene emission
of the air flow velocity in the spray zone and the spraying technique of the operator were
evaluated in the pilot experiment. The pilot experiment was conducted by spraying a regular
isophthalic acid (ISO)-based gel coat using a AAA spray gun. The gel coat was catalyzed with
methyl ethyl ketone peroxide (MEKP). A low (40 to 50 ft/min) and a high (100 ft/min) air flow
velocity in the spray zone were examined. Air velocities were measured by a hot-wire
anemometer at several locations across the spraying zone. This range represents the low and
high ends of air flow velocity found in an open molding area or in a spray booth. The spraying
technique of the operator was evaluated by asking the operator to spray normally (without.
consciously controlling the spray fan beyond the mold surface and flange) and in a controlled
pattern. Controlled spraying was done by consciously minimizing overspray beyond the flange
of the mold. The effects of spraying techniques were quantified by transfer efficiency, which is
the percentage of gel coat material deposited on the mold right after application. Spraying
techniques, shape of mold (male or female), and size of mold all affect the transfer efficiency of
the material applied.
The number of test runs for air flow velocity and spraying method are summarized as
follows and presented in Table 2-1.
A. Air flow velocity (as measured by a hot-wire anemometer across the spraying zone)
Al. Low air flow velocity (30 to 50 ft/min)
A2. High air flow velocity (90 to 120 ft/min)
B. Spraying method
Ml. Normal technique without conscious control of overspray from flanges
M2. Controlled spraying technique with more conscious control to reduce overspray
-------
Table 2-1. Test Runs for Pilot Experiment
Ml -Normal technique
M2-Controlled technique
Al-Low air flow velocity
3
3
A2-High air flow velocity
3
3
The results were analyzed to determine whether there are any differences in styrene
emissions resulting from different air flow velocities and spraying techniques. Following the
pilot experiment, the low air flow velocity and controlled spraying technique were selected for
the subsequent gel coat and resin experiments.
2.2 Gel Coat Experiment
The gel coat formulations selected were one regular ISO-based gel coat and a low-volatile
organic chemical (VOC), isophthalic acid/neopentyl glycol (ISO/NPG®)-based gel coat. Cook
Composites and Polymers (CCP) provided these two gel coats. For the purpose of this testing,
both gel coats contained straight styrene without any methyl methacrylate (MMA). Typical gel
coats contain only a low percentage of MMA while the styrene content may range from 35 to 50
percent. This minor modification allowed the assumption that total emissions quantified were
styrene emissions. A Reichhold Superox® 46709 MEKP catalyst was used and the catalyst ratios
followed those suggested by CCP.
The gel coat spraying equipment selected included: one AAA spray gun with external
catalyst mixing, one HVLP spray gun with internal catalyst mixing, and one HVLP spray gun
with external catalyst mixing. According to the Composites Fabricators Association's industry
survey, the AAA external mixing spray gun is the major spray gun used by-the industry; :
therefore, it is treated as the baseline condition. The AAA spray gun was compared with the
HVLP spray gun. The effects of internal and external catalyst mixing were evaluated for the
HVLP spray guns. Magnum provided all three spray guns. A pump ratio of 20:1 was selected
for the gel coat pump systems. The spray guns were compared at similar gel coat thicknesses
(about 18 to 24 mil) sprayed on an FRP mold.
The gel coat formulations and application equipment are denoted as follows:
A'. Formulations
GF1. Regular ISO-based gel coat (baseline condition)
GF2. Low-VOC, ISO/NPG®-based gel coat
B. Equipment
GE1. AAA spray gun with external catalyst mixing (baseline condition)
GE2. HVLP spray gun with internal catalyst mixing
GE3. HVLP spray gun with external catalyst mixing.
Table 2-2 shows the number of test runs for each of the gel coat formulation and
equipment combinations in the gel coat experiment.
-------
Table 2-2. Test Runs for Gel Coat Experiment
Formulation
GF1 Regular gel coat
GF2 Low-VOC gel coat
Equipment type
GEl-AAA(ext)
3
3
GE2-HVLP(int)
3
3
GE3-HVLP(ext)
3
3
ext=External catalyst mixing.
int=Internal catalyst mixing.
2.3 Resin Experiment
The resin experiment examined five resin formulations with a AAA spray gun and three
pieces of application equipment with a regular low-profile resin.
The resin formulations selected were one dicyclopentadiene (DCPD)-based low-profile
resin catalyzed with MEKP, one DCPD-based low-styrene resin, one ORTHO-based styrene-
suppressed resin, one DCPD-based resin catalyzed with benzoyl peroxide (BPO), and the same
ORTHO-based styrene-suppressed resin with an additional wax content. All the resin
formulations were sprayed by a AAA spray gun. Reichhold Chemicals, Inc., provided all resin
formulations and catalysts. Reichhold's Superox® 46709 MEKP solution has 9 percent active
oxygen. Reichhold's Superox® 46744 BPO catalyst is a 40 percent BPO dispersion in
nonvolatile plasticizer that has 2.6 percent active oxygen. The catalyst ratios for each of the
resins followed those suggested by Reichhold.
The pieces of resin application equipment selected were one AAA spray gun with .
external catalyst mixing, a flow coater with internal catalyst mixing, and a pressure-fed roller
with internal catalyst mixing. The AAA spray gun has a valve that allows the operator to use
either MEKP or BPO catalyst solution. The AAA external mixing spray gun is considered the
baseline condition of the industry. The AAA spray gun was compared with other nonspraying
equipment (i.e., the flow coater and the pressure-fed roller). Magnum provided all the equipment
for evaluation. A pump ratio of 11:1 was selected for the resin pump systems. The equipment
was compared at similar resin laminate thicknesses (about 70 to 100 mil). Fiberglass roving was
used for the AAA spray gun and 1.5-oz/ft2 chopped strand mat was used for the flow coater and
•pressure-fed roller. Two layers of the chopped strand mat were used for nonspraying lamination;
multiple passes of sprayup were used to give similar laminate thicknesses.
Resin formulations and application equipment are denoted as follows:
A. Formulations
RF1. DCPD-based low-profile resin catalyzed with MEKP (baseline condition)
RF2. DCPD-based low-styrene resin catalyzed with MEKP
.RF3. ORTHO-based styrene-suppressed resin catalyzed with MEKP
-------
RF4. DCPD-based resin catalyzed with BPO
RF5. Water-emulsified resin (included in the test plan but the manufacturer
withdrew from the test)
RF6. Same ORTHO-based styrene-suppressed resin (RF3) with an additional
0.1 percent wax
B. Equipment
RE1. AAA spray gun with external catalyst mixing (baseline condition)
RE2. Flow coater with internal catalyst mixing
RE3. Pressure-fed roller with internal catalyst mixing
RE4. Same AAA with external BPO catalyst mixing for the resin catalyzed with
BPO
Table 2-3 shows the number of test runs for the resin formulation and equipment
examined in the resin experiment.
Table 2-3. Test Runs for Resin Experiment
Formulation
RF1. DCPD-based low-profile
resin with MEKP catalyst
RF2. DCPD-based low-styrene
resin with MEKP catalyst
RF3. ORTHO-based styrene-
suppressed resin with MEKP
catalyst
RF4. DCPD-based low-profile
resin with BPO catalyst
RF6. ORTHO-based styrene-
suppressed resin + 0.1 % of wax
with MEKP catalyst
Equipment type
REl-AAA(ext)
6
3
3
NA
3
RE2-flow
coater(int)
3
NA •
NA
NA
NA
RE3-pressure-
fed roller(int)
3
NA
NA
NA
NA
RE4-A. A A (ext)
for BPO resin
NA
NA
' NA
3
NA
ext=External catalyst mixing.
int=Internal catalyst mixing.
NA = Not included in the experiment.
Note: RFS is a water-emulsified resin that was not tested because the manufacturer withdrew from the test.
-------
Chapters
Facility and Experimental Setup
The evaluation test was conducted in an isolated spray booth in Reichhold Chemicals'
physical testing laboratory, located in Research Triangle Park, North Carolina. This type of spray
booth is commonly used in an FRP/C facility. Reichhold Chemicals' physical testing laboratory
is used to perform testing for their resin users. It is not a production facility; therefore, the
background VOC concentration can be minimized.
3.1 Total Enclosure System
The spray booth is situated in an enclosed room with a double door leading to the
physical testing laboratory. The laboratory is air-conditioned; therefore, the room temperature
and humidity were very stable during the entire period of testing. The stable conditions reduced
the variability of temperature effect. Most facilities are not air-conditioned in the summer,
however, they do have winter heating to maintain product quality.
Figure 3-1 shows the side view of the spray booth. The room is 12 feet wide, 19 feet
high, and 15 feet deep, which can be considered a permanent total enclosure. The double door
measures 6 feet wide by 7 feet high, which can be considered the natural draft opening (NDO) to
the enclosure. Inward linear air flow velocity at the door (i.e., NDO) during the testing was
always above 200 ft/min. The spray booth and the enclosed room meet the criteria for a total
enclosure as prescribed in EPA Method 204-Criteria for and Verification of a Permanent or
Temporary Total Enclosure. Therefore, the emissions from the operations in the spray booth can
be assumed 100 percent captured.
The spray booth is 7 feet high, 11.5 feet wide, and 7.5 feet deep from the front edge to the
filter bank. The filter bank is 6 feet high by 11 feet wide. The distance between the front; edge of
the spray booth to the double door is 4 feet 10 inches. The air-conditioned makeup air flows
through the double door. The exhaust air flows through the filter bank at the end of the spray
booth and is exhausted upward by a duct 34 inches in diameter. The exhaust flow rate from the
spray booth averaged 8,670 ftVmin during the testing.
3.2 Emission Sampling Location
Emission measurements and exhaust air flow rate were monitored from the exhaust duct.
The sampling location is 6 diameters downstream of the last bend as shown in Figure 3-1. EPA
Methods 1 and 2 were used to determine the exhaust gas velocity and volumetric flow rate. EPA
Method 25A was used to determine total gaseous organic emissions. The emission sampling
procedures are outlined in Section 5.1.
-------
_Emission and air flow rate
monitoring location
Baffle
7'x6' (W)
Double door
Floor scale
Application equipment
(placed outside the enclosure
on a floor scale)
Figure 3-1. Side view of the Reichhold Chemicals spray booth and the experimental
setup in a permanent total enclosure (19'H x 12'W x 15'L).
-------
3.3 Experimental Setup
33.1 FRPMold
Three identical FRP molds were used for this evaluation test. Figure 3-2 shows a sketch
of the male FRP mold. The male mold has five exposed smooth surfaces similar to a rectangular
box. The mold measures 2 feet high, 2.5 feet long, and 2 feet wide. A 2-inch wide flange
surrounds the bottom of the mold for ease of part removal. The total surface area, including
flange, equals 24.5 ft2 (2.28 m2). The mold is constructed of traditional reinforced plastics
material to represent actual tooling material used by the industry. These empty molds weighed
about 34 kg. The mold was placed on a turntable mounted on a cart with casters. The turntable
allows the operator to spray on all mold surfaces by turning the mold and without moving his
position to the downwind location. The cart allows easy transfer of the mold from a preparation
area to the spray booth.
3.3.2 Air Flow Baffle
The exhaust flow rate from the spray booth could not be adjusted because the spray booth
had a constant speed exhaust fan. Therefore, a baffle was used between the double door and the
spraying zone to divert the air flow to the sides of the spray booth so that the air flow velocity in
the spraying zone could be reduced. The 6.5-foot by 4-foot baffle was constructed from lattice
board on a frame built from 2-inch by 4-inch studs. It has two additional pieces measuring 6.5
feet by 2 feet on either side of the baffle. These two side pieces can swing open like a screen.
Two layers of 15-mil-thick glass veil were attached to the centerpiece of the baffle to reduce the
air flow velocity through the baffle. In the pilot experiment, the baffle was used to maintain the
low air flow velocity and was removed for the high air flow velocity. Using this baffle, the linear
air flow velocity in the spraying zone can be reduced from more than 100 ft/min to 40 ft/min.
The baffle was used throughout the gel coat and resin experiments to maintain a low air velocity
in the spraying zone.
33.3 Glass Veil and Kraft Paper to Capture Overspray
Gel coat and resin sprayup generate overspray. To account for the materials not adhered
to the mold, glass veil was used on the filter bank and kraft paper was used on the ground surface
and side walls to capture overspray. The veil is 15 mils thick of A-type glass with non-styrene-
soluble binder. Two layers of the veil were used on the filter bank. The test results showed that
almost all airborne droplets were trapped on the first layer of veil. The kraft paper used for
ground cover was 50# weight. These veil and kraft papers were replaced every test run so that
overspray for each test run could be accounted for accurately.
10
-------
2" flange surrounding
the bottom of the mold
to ease part removal
Figure 3-2. Sketch of a male mold.
11
-------
3.3.4 Protective Equipment and Clothing for Operator
The operator wore a respirator with activated carbon cartridges to protect him from
exposure to styrene during the application stage of a test run. The operator also wore safety
glasses at all times. In order to account for any materials that might come into contact with the
operator, clean disposable gloves, coveralls, and shoecovers were used in each test run. This
protective gear was weighed before and after the test run to determine the amount of materials on
them.
3.4 Resin Property Testing Laboratory
Reichhold Chemicals has a resin property testing laboratory located in the same building
as the spray booth. The laboratory has all the instrument and equipment necessary to determine
the styrene contents and curing characteristics for the gel coat and resin formulations. Reichhold
personnel followed their standard procedures to measure properties for every gel coat and resin
formulation examined in the test. These Reichhold Standard Test Procedures are No. 18-001,
Determination of Non-Volatile Content of Polyester Resins; No. 18-021, Determination of
Brookfield Viscosity & Thixotropic Index of Polyester Resins; No. 18-501, Determination of
Room Temperature Gel, Time to Peak, and Peak Exotherm Characteristics of Polymer Resin;
and No. 18-152, Determination of Static Styrene Emissions for Compliance with SCAQMD Rule
1162. Copies of these Reichhold standard test procedures are provided in Appendix B
(Volume H).
12
-------
Chapter 4
Materials and Equipment
4.1 Properties of Gel Coat and Resin Materials
The properties of the gel coats and resins were analyzed by Reichhold personnel in their
laboratory using Reichhold standard test methods. These properties are shown in Table 4-1 for
gel coats and Table 4-2 for resins. A large sample (about 200 g) was collected for each material
when it was first used and a smaller sample (about 20 g) was collected when the material was last
used. At the end of the testing, the large samples were analyzed for the listed properties and the
small samples for percent nonvolatile (%NV). The final %NV measurement was to verify
whether the material lost styrene over the test period or not. The results indicated that no
noticeable styrene was lost from the container because proper procedures were used to minimize
styrene evaporation loss.
Generally, the measured properties were in agreement with the properties listed in the
manufacturers' data sheets. However, major differences were found for the low-VOC gel coat,
the low-styrene resin, and the BPO resin. At the same catalyst ratio, the measured cup gel time
for the low-VOC gel coat (27 min) is longer than CCP's listed gel time (14-17 min). Measured
gel time for low-styrene resin (30 min) is also longer than a typical gel time (15 min) listed in the
Reichhold data sheet: The longer gel time might have an effect on total emissions because the
wet surface had a longer time to emit styrene.
Table 4-1. Gel Coat Properties Measured at Reichhold
CCP product code (color)
Density, Ib/gal
% NV, average (range)
% Styrene (by difference)
Viscosity, cps (LVF #4 @ 60 rpm)
Thix index
Catalyst Superox* No. and type
Catalyst ratio, weight %
Cup gel time, min
Total time to peak, min
Peak exothenn, "F
Rule 1 162 static emissions, g/m2
GF1 Regular gel coat
944-W-005SP (base white)
10.6-10.9
61.3(61.1:61.4)
38.7 '
3,040
6.1
46709 MEKP
1.8
17
35
353
133
GF2 Low-VOC gel coat
962-WA-196SP (pink)
11.3-11.6
74.6 (74.6-74.7)
25.4
2,970
3.9
46709 MEKP
1.8
27
51
251
83
13
-------
Table 4-2. Resin Properties Measured at Reichhold
Reichhold's Polylite® No.
Density, Ib/gal
% NV, average
(range)
% Styrene (by difference)
Viscosity, cps (Brookfield LVF
#3 @ 6/60 rpm)
Thix index
Catalyst Superox® No. and type
Catalyst ratio, weight %
Cup gel time, min
Total time to peak, min
Peak exotherm, °F
Rule 1 162 static emissions, g/in2
RF1
Low-
profile
33233-05
9.0
61.7
(60.9-62.2)
38.3
585
3.5
46709
MEKP
1.5
20
31
333
69.0
RF2
Low-
styrene
33234-17
9.2
64.7
(63.8-65.5)
353
421
2.3
46709
MEKP
1.4
30
40
316
73.6
RF3
Styrene-
suppressed.
33099-08
9.1
56.5
(56.5-56.6)
43.5
534
2.6
46709
MEKP
1.5
17
32
309
54.0
RF6
Styrene-
suppressed+
0.1% wax
33099-08
9.1
56.7
(56.7-56.7)
43.3 -
394
2.5
46709
MEKP
1.5
17
33
318
49.4
RF4
Neat BPO
33146-17
9.1
57.4
(57.3-57.4)
42.6
163
2.2
46744
BPO
2.1
(slow gel)
30
39
231
160.9
RF4 (modified)
BPO+0.5% fume
silica as thickening
agent
33146-17
9.1
57.4
(57.3-57.4)
42.6
386
3.8
46744
BPO
3.1
(fast gel)
17
26
277
96.6
-------
When the neat (unfilled) BPO resin was first used, it had a much longer gel time (30 min)
than the listed gel time (12 min). Its viscosity was also low, so the resin did not stay on the mold
until it cured. After two test runs on the neat BPO resin, a 0.5 weight percent of ftime silica was
added to thicken the resin and the BPO catalyst ratio was increased from 2.1 to 3.1 percent to
shorten the gel time. Two additional test runs were conducted for the modified BPO resin. The
test results for neat BPO resin and modified BPO resin were analyzed separately.
4.2 Setting of Gel Coat and Resin Application Equipment
The application equipment was prepared for each test run in a separate spray booth. An
experienced operator from Magnum Industries operated the application equipment for the entire
5-week period. He adjusted the setting on the equipment in the preparation area until a good
spray pattern was acquired. Then the equipment was disconnected from the central compressed-
air line, moved to the spray booth where the test was conducted, and reattached to the central
compressed-air line. The setting on the equipment was recorded after the application was
completed. Tables 4-3 and 4-4 show the range of the setting for the gel coat and resin application
equipment, respectively. Table 4-3 shows that the low-VOC gel coat required more air pressure
and larger spray tips to achieve a spray pattern similar to the regular gel coat.
Table 4-3. Summary of Gel Coat Equipment Setting
Equipment type
Magnum model No.
Pump ratio
Type of gel coat
Air supply pressure, psi
Catalyst atomizing
pressure, psi
Spray tip No.
Catalyst ratio setting,
volume %
Deliver rate, g/min
Deliver rate, gal/min
GE1-AAA (external
catalyst mixing)
ATG-3500
20:1
GF1
Regular
42-44
26
418
2.1
784-794
0.16
GF2
Low-VOC
52-60
26
518/718
2.1-2.4
746-839
0.14-0.16
GE2-HVLP (internal
catalyst mixing) '•.-.?•
HVLPF-5500
20:1
GF1
Regular
38-45
NA
418 .
2.1
670-760
0.14-0.16
GF2
Low-VOC
60-68
NA
518/718
2.1-2.4
779-933
0.15-0.18
GE3-HVLP (external
catalyst mixing) :
HVLPF-5500
20:1
GF1
Regular
42-44
26
418
2.1
780-809
0.16-0.17
GF2
Low-VOC
54-64
26
518/718
2.1-2.4
774-927
0.15-0.18
15
-------
Table 4-4. Summary of Resin Application Equipment Setting
Equipment type
Magnum model No.
Pump ratio
Type of resin
Air supply pressure,
psi
Catalyst atomizing
pressure, psi
Spray tip No.
Catalyst ratio setting,
volume %
Deliver rate, g/min
Deliver rate, gal/min
RE1-AAA (external MEKP catalyst
mixing)
ATC-4000
11:1
RF1
low-
profile
37-39
18-19
443
1.5-1.6
2332-
2559
0.57-0.63
RF2
low-
styrene
37-38
18
443
1.6
2316-
2335
0.56-0.56
RF3
styrene-
suppress.
37-41
17-18
443
1.5-1.6
2254-
2409
0.55-0.58
RF6
styrene-
suppress.
+ wax
36-38
18-26
443
1.6
2286-
2569
0.55-0.62
RE4-AAA (external
BPO catalyst mixing)
ATC-4000
11:1
RF4
neat BPO
26
43-45
443
2.0-3.0
2065-
2376
0.50-0.58
RF4
BPO with
thickener
24-25
41-43
443
3.5
2205-2269
0.53-0.55
RE2-Flow
coater
(internal
catalyst
mixing)
Flo-6000
11:1
RF1
low-profile
40-42
NA
#2
1.5-1.6
3388-3605
0.83-0.88
RE3-Pressure-
fed roller
(internal
catalyst
mixing)
MRD resin
roller
11:1
RF1
low-profile
40-42
NA
NA
1.6
Not measured
Not measured
-------
4.3 Reinforcements
PPG Industries, Inc., provided all the reinforcements for the testing. The HYBON® 700
HTX roving material was used for resin sprayup and a GPM chopped strand mat (1.5 oz/ft2) was
used for nonspraying resin lamination. The chopped strand mat was cut into proper sizes before
the test so that the operator could apply the mat, piece by piece, on the mold surface.
17
-------
Chapters
Determination of Emission Quantities
Two separate test methods were used to quantify emissions from the test. The first test
method was an emission measurement method that uses EPA Method 25A to determine a real-
time, continuous emission concentration in the stack and EPA Methods 1 through 4 to determine
the exhaust flow rate. Because the enclosed room with the spray booth met the criteria for a total
enclosure, the emissions within the enclosed room could be assumed 100 percent captured. The
emission quantities were calculated from the product of an average emission concentration and
an average exhaust flow rate during a test run.
The second test method was a mass balance calculation method using gravimetric
measurements. This method measures the weights of all materials, overspray, mold, and part at
the beginning and end of each test run. The difference between the total initial weight and the
total final weight is considered the weight loss due to emissions.
Before and during the test campaign, pure styrene evaporation tests were used to
compare the emission quantities determined by these two test methods. This involved measuring
weight loss due to evaporation from terry cloth towels soaked in pure styrene using a high-
precision scale. The measured weight loss was compared with emissions determined by the
emission measurement method. If the results are close, it implies that the emission measurement
method was accurate.
Both test methods determine total emissions. Because these gel coat and resin materials
contained only styrene monomer, the total emissions measured could be considered styrene
emissions. Other VOC emissions were excluded from the surrounding environment. The
background (i.e., baseline) VOC concentration in the laboratory was measured before each test
run and subtracted from the average emission concentration so the net increase in concentration
could be attributed to the test. A test run began when gel coat or resin application started and
ended when curing was complete and the monitored concentration returned to the baseline
concentration. .Most of the test runs lasted from 60 to 90 minutes, depending on the time
required for complete curing. Two test runs for the neat BPO resin were longer than 100 minutes
because of unusually long curing time.
5.1 Emission Measurement Method
The emission measurement method determines styrehe concentrations in the exhaust air
and exhaust air flow rate, then uses these results to calculate total emissions during the test run.
5.1.1 Determination of Styrene Concentrations in Exhausted Air
Styrene emissions in the exhaust stack of the total enclosure were measured according to
EPA Method 25A as given in Code of Federal Regulations, Chapter 40, Part 60. The
18
-------
measurements were made using a Ratfisch Instruments model RS55CA total hydrocarbon (THC)
analyzer. This analyzer has a flame ionization detector (FID) that responds to hydrocarbons
approximately in proportion to the number of carbon atoms entering the detector. The total
hydrocarbon measurements can be attributed to the styrene emissions because the only
hydrocarbon emission source was the gel coat or resin application and styrene is the only
monomer used in the formulations. Measurements were made immediately before gel coat or
resin application started to determine the levels of background hydrocarbons (e.g., natural
methane and other trace VOCs) in ambient air. These background levels were subtracted from
the levels measured during the test run.
The THC analyzer was connected to the exhaust stack by a sampling line fabricated from
a 12-foot length of 1/4-inch ID perfluoroalkoxy (PFA) Teflon tubing. The tubing was capped
inside the exhaust stack. Eight holes were drilled in the tubing at various points across the 34-
inch diameter of the exhaust stack to obtain a representative sample of its contents. Sample was
drawn through the sampling line into the analyzer at a flow rate of 7 Um\i\. Most of the sample
was vented to the atmosphere, but a small portion of the sample entered the FID through a
capillary. A backpressure regulator maintained a constant sample pressure, which maintained a
constant sample flow rate in the capillary. The sample was oxidized by a hydrogen/air flame and
the ionized carbon atoms produced in the flame were detected by an electrometer.
The output signal (0 to 10 V dc full scale [FS]) from the analyzer was recorded by an
Omega Engineering model OM-170 microprocessor-based portable data logger and a Hewlett- «•
Packard model 7132A strip chart recorder. Both instruments were operated on their 0- to 1Q-V
FS ranges. The data logger recorded the voltages at 2-second intervals throughout each test run.
At the end of each day, the voltage measurements were transferred to a laptop computer
containing a spreadsheet program. Hie strip chart recorder provided a visual indication of the
styrene emissions measurements during each test ran and provided physical documentation for
each test ran. .
The THC analyzer was operated on Range 2 (0 to 200 ppm styrene) for most of the test
runs, but was operated on Range 1 (0 to 20 ppm styrene) for three resin experiment runs in which
low styrene concentrations were expected. These ranges are also equivalent to 0 to 53 ppm
propane and 0 to 533 ppm propane. The THC analyzer was calibrated prior to each test run
using compressed gas calibration standards. A calibration drift check was done at the end of
each test run. Styrene calibration standards could not be used directly for routine calibrations
during test runs because of cylinder pressure limitations associated with styrene's dewpoint.
Instead, propane in air calibration standards without such pressure limitations were used for the
routine calibrations. The calibration gases were 16,27,45, 160,267, and 453 ppm propane.
These calibration gases corresponded to 30 percent, 50 percent, and 85 percent of the two full-
scale ranges, as called for in EPA Method 25A.
Calibration data obtained from measurements of propane calibration standards could be
used for the styrene emissions determination because RTI developed a correction factor for
19
-------
converting propane concentrations into the equivalent styrene concentrations prior to the testing.
A styrene molecule has eight carbon atoms and a propane molecule has three carbon atoms. As a
first approximation, one would expect the propane-to-styrene correction factor to equal the ratio
of the carbon atoms (i.e., 8/3 = 2.667). The measured propane-to-styrene correction factor was
2.686 for the 0- to 200-ppm styrene range on the THC analyzer. This measured correction factor
was used in emission calculations.
The propane calibration standards were verified by RTI using propane analytical
reference standards. They were intercompared with styrene calibration standards to obtain the
propane-to-styrene correction factor. The styrene calibration standards were verified by RTI
using styrene analytical reference standards. The details of these measurements are presented in
Appendix C (Volume H).
In general, six propane calibration standards, including the high-level, mid-level, and
low-level calibration standards for the selected analyzer range, and zero air were measured
during the calibration for the first test run of each day. The three propane calibration standards
for the selected analyzer range and zero air were measured during the calibrations for subsequent
test runs. Calibration data for one test run were used for the preceding test run's drift check,
except for the last test run of the day when a separate drift check was conducted. The analyzer's
zero and span pots were not adjusted during the entire 5-week testing period. The details of these
calibrations are presented in Appendix D (Volume n).
The propane calibration standards were connected to the analyzer's calibration port via an
8-foot length of 1/8-inch ID PFA Teflon tubing. An in-line pressure regulator set to 5 psig and a
needle valve maintained a constant flow rate in the tubing. Quick-connect fittings were used to
switch from one calibration standard to the next. The analyzer required 15 minutes or more to
yield a stable analyzer response for the first standard to be analyzed during a calibration.
However, the stabilization period for subsequent calibration standards was only a few minutes.
The cause for this long initial stabilization period was never determined, but it represented only a
minor impediment to the calibrations.
After the voltage readings from the styrene emissions measurements were transferred to
the computer-based spreadsheet, they were converted into an average voltage for the test run.
The voltage associated with the background air measurement from the start of the test run was
subtracted to yield a net average voltage. An average styrene concentration for the test run was
obtained by multiplying the net average voltage by a styrene calibration factor. This calibration
factor was obtained by dividing the equivalent styrene concentration for the high-level calibration
standard by the difference between the voltages from the measurements of the high-level
calibration standard and zero air.
RTI checked for concentration stratification inside the exhaust stack during a pure styrene
evaporation test by sampling the exhaust stream across two perpendicular traverses. These
measurements were made at 10 points on each traverse at distances corresponding to equal
20
-------
subdivisions of the exhaust stack's area. The sampling line was modified so that sample was
collected from a single point in the exhaust stack. The styrene concentration was not stable
during this pure styrene evaporation test, and the analyzer response declined from 2.17 to 1.32 V.
Least squares regression was used to fit these data to an exponential decay curve with a
coefficient of determination (i.e., r-squared) equal to 0.98. Individual measurements deviated
from the regression curve by an average of 1.6 percent of the predicted value. The maximum
deviation was 5.2 percent. This statistical analysis suggests that there is little concentration
stratification in the exhaust stack.
5.1.2 Exhaust Air Flow Rate Measurement
Air velocity in the 34-inch diameter circular exhaust stack of the total enclosure was
measured according to EPA Methods 1 and 2 as given in Code of Federal Regulations, Chapter
40, Part 60. The measurements were made using a Dwyer series 160 stainless steel pitot tube
(standard type) and a Dwyer series 2000 Magnehelic differential pressure gauge. The
Magnehelic gauge was compared to an inclined manometer (a primary standard). Magnehelic
readings were 95 percent (0.95) of manometer readings. This 0.95 correction factor was used in
air flow rate calculations. The velocity was measured at a distance of approximately five to six
stack diameters downstream of two right-angle bends in the exhaust stack.
Velocity across two perpendicular traverses was measured weekly. These measurements
were made at 12 points on each traverse at distances corresponding to equal subdivisions of the
exhaust stack's area. These data were recorded in a data sheet and were transcribed into a
computer-based spreadsheet for data reduction. In general, the weekly velocity measurements
indicated that the exhaust air flow rate remained relatively constant for the entire 5-week testing
period. The average exhaust air flow rate was 8,685 ftVmin for the entire testing period and
individual weekly measurements varied from 8,358 to 9,034 ftVmin. The Reynolds number for
the air flow in the 34-inch diameter exhaust stack was 3.97X105, which places the flow in the
turbulent regime. Therefore, any concentration stratifications were not likely to persist for long
distances inside the exhaust stack. This conclusion is consistent with the negative results of the
concentration stratification measurements.
RTI checked for off-axis flow on one occasion by rotating the pitot tube inside the
exhaust stack and recording the velocity head at various angles. The results of these
measurements indicate that off-axis flow was not a problem in the exhaust stack.
Velocity head (Ap) measurements at the centerline of the exhaust stack were usually
performed at 15-minute intervals during each test run. The Magnehelic differential pressure
gauge was used to obtain these measurements. In general, the centerline Ap remained relatively
constant throughout each test run although there were short-term fluctuations on the Ap. The 15-
minute data were recorded in a data sheet and in a laboratory notebook during each test run and
were transcribed into a computer-based spreadsheet for data reduction. The average centerline
Ap for a test run was used as a scaling factor for calculating an estimated average exhaust air
21
-------
flow rate for that test run. In general, the average centerline Ap remained relatively constant for
the entire testing period. Individual values ranged from 0.109 to 0.133 inch of water with an
overall average value of 0.120 inch of water.
The exhaust flow rate during the test run was calculated according to the following
formula:
Q run = [avg (Ap run)0 V(Ap weekly)0'] x Q weekly (5-1)
where
Q run = exhaust flow rate during a test run (acfm)
Q weekly = exhaust flow rate determined by weekly velocity traverse (acfm)
avg (Ap run)05 = average square root of 15-minute Aps recorded at centerline
during the test run
(Ap weekly)0 5 = square root of Aps recorded during weekly velocity traverse.
The relative humidity and temperature of the air in the spray booth were measured by a
sling psychrometer during each test run. The measurements showed that the ambient air
temperature and relative humidity were very stable at 73±1 °F and 58±2 percent, respectively.
5.1.3 Emissions Determined by Emission Measurement Method
Styrene emission quantity (Em) for each test run was calculated by the following
equation:
Em, Ib = 2.6x 10'9 x Q. x MW x C x T (5-2)
Em,g =1.18xlO-6xQxMWxCxT (5-3)
where
2.6x10"', 1. ISxlO"6 = conversion factors to standard conditions (68°F and 29.92 inches
mercury) in English and metric units, respectively
Q = average exhaust air flow rate (actual cubic feet per minute)
MW = molecular weight of styrene (104)
C = average styrene emission concentration during the entire test run
(ppmv actual)
T = duration of test run (minutes).
Using the emission concentration profile and the exact duration of the application stage
(Ta), an average emission concentration (Ca) could be calculated for the application stage in each
test run. The total emissions during the application stage (Ea) could be calculated from the same
equation:
22
-------
= 2.6xlO-9xQxMWxCaxTa (5-4)
Ea,g =1.18xlO-6xQxMWxCaxTa (5-5)
The difference between total emissions and emissions during the application stage is the
emissions from the wet-out stage (for resin lamination), stagnant (curing) stage, and overspray.
5.2 Mass Balance Calculation Method Using Gravimetric Measurements
The mass balance calculation method involves weighing all materials, overspray, mold,
and part in the beginning and at the end of a test run. The difference between the total initial
weight and the total final weight is the weight loss due to emissions.
Weight losses due to styrene emissions were determined using two floor-type, high-
precision scales (Sartorius Corporation, Model F150S). The scales have a 150,000-g capacity
and 1-g readability. These two scales were calibrated with subsequent additions of standard
weights - 1 g, 5- g, 10 g, 20 g, 50 g, 100 g, 500 g, 1,000 g, and 2,000 g - daily. The calibration
procedures were performed on an empty scale and with a heavy object (i.e., an empty mold) on
the scale. This dual calibration procedure ensured that the scales had the same sensitivity in the
range of weights encountered in the tests. These calibration procedures showed that the scales
precisely indicate the standard weights added. Scale drift was checked periodically by leaving a
1,000-g or 2,000-g standard weight on the scale overnight. The drift check showed that these
two scales were very stable and the overnight drifts were within ±2 g.
The first scale was used to measure the initial and final weights of gel coat or resin
materials, catalyst, fiberglass reinforcement, glass veil, protective clothing, and kraft paper for
ground cover. A second scale was specifically used for the mold and cart. The mold and cart
were left on the second scale for the entire test run.
The gravimetric measurement procedures for the mass balance calculation method are
outlined as follows:
A. Before Application
1. Determine the initial weight of ground cover (e.g., kraft paper) and thin glass veil
(used to capture overspray droplets on filter bank) - W1.
2. Determine the initial weight of tools and other items (e.g., wet-out rollers, gloves,
booties, coverall) that will come into contact with the materials during the
application - W2.
3. Determine the initial weights of materials (i.e., gel coat, resin, catalyst, fiberglass
roving, or chopped strand mat) to be used for part production. Weigh the pump
23
-------
system including gel coat or resin container on the first scale to determine the initial
weight of gel coat or resin material and catalyst - W3.
4. Determine the initial weight of the empty mold with the cart and protective skirt on
the second scale - W4.
B. Preparation for Gel Coat or Resin Application
1. Place ground cover on the floor around the second scale in the spray booth and attach
glass veil on the filter bank to capture droplets from overspray.
2. Place the empty mold and the cart (with protective skirt attached) on the second scale
in the center of the spray booth.
3. Apply gel coat or resin in a manner consistent with a typical operation.
C. After Application
1. Determine the final weights of materials (i.e., gel coat, resin, catalyst, fiberglass
roving, or chopped strand mat) used for application. Take the reading of the whole
pump system from the first scale to determine the final weight of gel coat or resin
material and catalyst - W5.
2. Take a measurement reading of the second scale for the mold with the wet gel coat
or resin after application is completed - W6.
3. For resin lamination, take another measurement reading of the second scale for the
mold with the part, after wet-out rolling is completed - W7.
D. After Complete Curing
1. Determine the final weight of the mold with cured gel coat or resin laminate on the
second scale - W8.
2. Determine the final weight of ground cover and thin glass veil - W9.
3. Determine the final weight of tools and items (e.g., wet-out rollers, gloves, booties,
coverall) that came into contact with the materials during the application. (Wet-out
rollers can be weighed right after they are used so that resin residue can be cleaned
from the rollers) - W10.
24
-------
E. Transfer Efficiency
From the weights of materials used and materials applied on the mold, calculate the
transfer efficiency for each test run. Transfer efficiency = (W6-W4)/(W3-W5).
F. Emissions from Part During Stagnant (Curing) Period
Emissions from the gel coat on the mold during the stagnant (curing) period can be
determined from the weight loss after the spraying and complete curing = W6-W8.
Emissions from the resin laminate during the stagnant (curing) period can be determined
from the weight loss between when the wet-out rolling was completed and curing was
completed. During this period the mold remained on the second scale. Emissions from
the resin laminate only during the stagnant period = W7-W8.
G. Total Emissions
Total emissions (including emissions from the application stage, the wet-out rolling stage,
the stagnant [curing] period, and overspray) = Materials used (W3-W5) - Materials on the
mold (W8-W4) - Materials not on the mold (W9+W10-W1-W2).
A data recording sheet used for the testing is shown in Figure 5-1. This data recording
sheet records more detailed measurements so that glass ratio and catalyst ratio can also be
calculated.
25
-------
Data Recording Sheet for Mass Balance Calculation
Date: Time: Recorded by:
Test run#:
Wet bulb:
Material/container #:
Dry bulb:
Equipment:
RH%:
Application technique: normal/controlled
Air velocity: low/high Baffle: with/without
Initial (g) Final (g) Change (g)
1 Mass of rear veil with a roll of wide masking tape
2. Mass of ground cover/front veil/masking tape/booties/gloves/pants/stool
2a. Mass of ground cover and a roll of narrow masking tape
2b. Mass of front veil with existing masking tape
2c. Mass of booties/gloves/pants/stool
3.
4.
5.
6a.
6b.
7a.
7b.
8a.
8b.
9a.
9b.
Mass of wet-out rollers in container
Mass of fiberglass roving or chopped strand mat
Mass of resin or gel coat material
Volume of catalyst, ml (catalyst type/#:
Mass of catalyst
Sp.gr.
Weight(g) Time
Mass of scale cover + empty mold + skirt + duct tape (fan off)
Mass of scale cover + empty mold + skirt + duct tape (fan on before application)
Mass of scale cover + wet mold + skirt + duct tape (at the end of application)
Mass of scale cover + wet mold + skirt + duct tape (at the end of rolling)
Mass of scale cover + cured mold + skirt + duct tape (fan on)
Mass of scale cover + cured mold + skirt + duct tape (fan off)
10. Mass of material used (M4 + M5 + M6b)
11. Mass of cured material on mold + skirt + scale cover (M9b-M7a)
12. Mass of cured materials deposited elsewhere (M1 + M2 + M3)
13. Total emissions (M10-M11-M12)
14. Curing emissions from mold and skirt (M8a-M9a)
15. Glass ratio (M4/(M4 + M5))
16. Catalyst ratio (M6b/M5)
17. Transfer efficiency l(M8a-M7b)/M10]
Note:
Figure 5-1. Data recording sheet for mass balance calculation (used in the testing at Reichhold).
BLANKRES.XLS
-------
Chapter 6
Test Results and Discussions
A Spreadsheet containing the emission measurements, gravimetric measurements, and
calculated emission quantities and emission factors is provided in Appendix E (Volume II).
Table 6-1 summarizes the emission measurements and calculated emission factors for three
experiments. The test results were subject to analysis of variance (ANOVA) to determine the
significance of variables on styrene emissions. The detailed results of this statistical analysts are
included in Appendix F (Volume IT). Table 6-2 compares variables that affect total emissions as
measured by THC. Information in Table 6-2 is extracted from Table 17 in Appendix F.
The level of significance of the comparison is presented by the number of asterisks --'the
more asterisks, the more confidence that there is a significant difference in the comparison. The
majority of the comparisons are significant at the 0.001 level (or 99.9 percent confidence
interval), noted by ***. This means that if the test is repeated 100 times, 99.9 percent of the test
results will show a difference between the two variables compared. A few variables are
significant at 0.01 or 0.05 levels (99 or 95 percent confidence intervals), noted by ** .and *,
respectively. If the comparison is not statistically significant at the 0.05 level, it is noted by "ns."
Table 6-2 shows that, in the pilot experiment, normal and controlled spraying techniques
caused a difference in total emissions and the difference is significant at the 0.001 level. The
high and low air velocity in the spray booth did not make a difference in total emission and the
difference was not significant at the 0.05 level.
The gel coat experiment shows that the regular and low-VOC gel coat made a difference
in total emissions and the difference is significant at the 0.001 level. The total emissions from
three gel coat spray guns are not significantly different at the 0.05 level.
In the resin experiment, the difference in total emissions from low-profile (general
purpose) resin and low-styrene resin is significant at the 0.01 level. The differences between
low-profile resin and styrene-suppressed resin or styrene-suppressed resin with additional wax
are significant at the 0.001 level. This implies that emission reductions achieved by using low-
styrene resin, styrene-suppressed resin, and styrene-suppressed resin with additional wax are
statistically significant. However, total emissions between styrene-suppressed resin and styrene-
suppressed resin with additional wax are not significantly different. This implies that the
emission reduction achieved by adding wax to the styrene-suppressed resin is not statistically
significant at the 0.05 level.
Similarly, emissions generated from the flow coater and pressure-fed roller are
significantly different from those from the AAA spray gun at the 0.001 level. However, the
emissions generated from the flow coater and the pressure-fed roller are not significantly
different.
27
-------
Table 6-1. Summary of Emission Measurements and Calculated Emission Factors
1
Test run #lMat6 Equip.
| Icontainer #
Application
Method
EXPERIMENTAL RUNS
i , '
RF1-EXPJRF1IRE1
GFI-EXPIGF1IGE1
= ;
: :
PILOT EXPERIMENT
;
PI iGFUGEI
P2 iGFIIGEI
P3 IGFIIGEI
P4 IGFllGE!
P5 IGFIIGEI
P6 .GFIIGEI
P7 iGFIIGEI
P8 IGFIIGEI
P9 |GF1|GE1
PIO ' iGFIIGEI
Pll iGFIIGEI
P12 JGF1
!
GEI
Average (12 runs)
i
Normal
Normal
Controlled
Normal
Controlled
Normal
Controlled
Normal
Controlled
Normal
Controlled
Controlled
Normal
Normal
Ml -Normal spraying (6 runs)
Mi-Controlled spraying (6 runs)
1
1
A2-High air velocity (6 runs)
A 1 -Low air velocity (6 runs)
1
Ml/A 1(3 runs)
M2/A1 (3 runs)
Ml/A2(3runs)l
M2/A2 (3 runs)
j -
Air
Vel.
High
High
Low
High
High
High
Low
Low
High
Low
High
Low
High
Low
Total I Emissions | Total
emissions | during
byTHCg
631
456
322
536
382
526
397
506
application, g
(byTHQ
32S
228
134
310
138
253
134
219
. 42?| 157
5541 271
422
396
479
476
452
513
391
462
442
512
372
514
410
176
145
214
• 204
196
245
147
208
184
231
138
259
157
emissions
tyMB.g
886
478
380
607
442
415
412
506
466
499
389
457
535
442
463
. 501
424
476
449
482
416
519
432
Material
used. g
9287
1564
1564
2261
1695
2067
1964
2204
1900
2056
2064
2021
2077
2049
1994
2119
1868
2011
1976
2103
1850
2135
1886
• Transfer
eff.%
90.1
75.5
79.8
66.1
80.0
74.0
85.1
75.7
81.2
73.8
85.0
82.6
74.5
77.7
78.0
73.6
82.3
76.8
79.1
75.7
82.5
71.5
82.1
Siyrene
content. %
38.33
38.75
38.75
38.75
-38.75
38.75
38.75
38.75
38.75
38.75
38.75
38.75
38.75
38.75
38.75
38.75
38.75
38.75
38.75
38.75
38.75
38.75
38.75
Emission Factors
%AS
byTHC
17.7
75.2
53.1
61.2
58.2
65.7
52.2
59.2
58.0
69.6
52.7
50.6
59.5
60.0
58.3
-62.5
54.1
59.2
57.4
62.9
52.0
62.1
56.3
Note: THC=THC emission measurement. MB=mass balance measurement. EF=emission factor, %AS=% available sryrene.
g/g=g of emission per g of material used.
«/«n2
tft
by THCl by THC
277
200
0.068
0.291
1
• 141
236
168
231
175
222
0.206
•0.237
0.225
0.254
0.202
0.230
1881 0.225
244
-185
174
210
209
199
225
172
203
194
225
163
226
0.270
0.204
0.196
0.231
0.232
0.226
0.242
0.210
0.229
0.223
0.244
0.201
0.241
I80| 0.218
(con.)
28
SUMSUM.XLS
-------
Table 6-1. Summary of Emission Measurements and Calculated Emission Factors
i
1 I
Test run # |Matd Equip.
I 'container #
Application
Method
GELCOAT EXPERIMENT
let
!G2
IGJ
'G4
[OS
GFIIGE3
GF1IGE2
GFIIGE1
GF2IGEI
GF1IGE1
!G6 iGFIlGEl
107
GF2IGE3
G8 .GF11GE3
G9
CIO
Gil
GF1IGE2
GF1
GE3
GF2)GE2
;G12 'GF2IGE2
iGI3
G14
G15
G16
G!7
GI8
I
GF2IGE2
GF2IGEI
GFIIGI2
GF21GEI
GF2JG13
GF21GE3
Avenge (18 runs)
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
GF1 .-Regular gel coat (9 runs)
GF2-Low VOC gel coat (9 runs)
j
GE1-AAA ext mix gun (6 runs)
GE2-HVLP int mix gun (6 runs)
GE3-H VLP ext mix gun (6 runs)
!
GFI/GE1 (3 runs)
GFI/GE2 (3 runs) ~l
GFI/GE3 <3 runs)
GF2/GE1 (3 runs)
GF2/GE2 (3 runs)
GF2/GE3 (3 runs)
Air
Vel.
Total i Emissions
emissions i during
byTHC. g] application, g
! (byTHC)
Low
Low
Low
Low!
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
3951 154
3951 164
4091 158
2731 91
4031 154
387| 142
2711 80
3821 143
386J 155
384 i 134
2941 94
2491 94
291! 112
2671 92
3391 162
2741 99
2981 100
282 93
I
3321 123
, , , |
3871 152
278 95
336 123
326) 130
3351 1 17
4001 151
3731 160
3871 143
2721 94
2781 100
2841 91
Total
emissions
byMB.g
Material
used, g
Transfer
Styrene
eff. %| conteni. «
3481 1723
406
404
227
355
410
273
472
398
360
327
346
374'
332
387
302
310
286
351
393
309
338
373
342
390
397
39*
287
349
290
1808
1756
1929
1765
1817
1876
1891
1787
1723
1940
1940
2245
1933
1776
2128
2275
1963
1904
1783
2025
1888
.1916
1909
1779
1790
1779
84.6
81.4
79.0
88.8
84.3
82.9
87.3
80.0
82.4
84.7
83.3
81.91
Emission Factors i
%AS
byTHC
i
38.75
38.75
38.75
59.2
56.4
60.2
25.351 S5.8
38.75| 58.9
38.75
25.35
38.75
38.75
38.75
25.35
25.35
81.11 25.35
. 81.9
82.0
87.4
85.7
85.4
83.6
82.4
84.8
84.1
. 82.0
84.6
82.1
81.9
54.9
56.9
52.1
55.7
57.6
59.8
50.6
51.2
25.351 54.6
38.75
25.35
25.35
25.35
32.05
38.75
25.35
32.05
32.05
• 32.05
38.75
38.75
49.2
50.9
51.7
56.6
55.1
56.0
54.2
55.9
! 53.8
55.7
58.0
53,»
83. 1 1 38.751 56.3
19971 86.0
2042T 82.1
2038| 86.1
1
25.35| 53.7
25.35
25.35
53.9
55.1
Note: THC=THC emission measurement. MB=mass balance measurement. EF=cmission factor, %AS=% available styrene.
g/g=g of emission per g of material used.
i
g/m2|. .g/g
byTHC
174
174
180
120
byTHC
0.229
0.218
0.233
0.141
I77| 0.228
170
119
168
169
169
129
109
128
117
149
120
131
124
146
170
122
• 147
143
147
176
164
170
119
122
124
0.213
0.144
0.202
0.216
0.223
0.152
0.128
0.130
0.138
0.191
0.129
0.131
0.144
0.177
0.217
0.137
0.180
0.172
0.179
0.225
0.208
0,218
0.136
0.137
0.140
(con.)
29
SUMSUM.XLS
-------
Table 6-1. Summary of Emission Measurements and Calculated Emission Factors
(con.)
1
Test run #|Mauj Equip.
;
j container #
RESIN EXPERIMENT
Rl
R2
R3
R4
1R5
!R6
R7
R8
R9
RIO
Rll
R12
R13
R14
LR15
R16
R17
R18
R19
RF6IRE1
RF1|RE2
RFHRE3
RFI IRE3
IRFliREl
RF3RE1
RFI REI
IRF2 REI
RF4|RE4
RFllREI
RF6|RE1
RF6JRE1
RF2REI
RF2REI
RFI REI
RFI RE3
RF4RE4
RFI RE2
RF3REI
R20 IRF3 REI
R21
RF1JRE1
JR22 IRF1 REI
R23
R24
R25
RF4RE4
RF1IRE2
RF4IRE4
i
Application
Method
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Normal
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
ControRed
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Controlled
Average (25 runs, w "Normal" run RIO)
Air
Vel.
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Average (24 runs, w/o "Normal" run RIO)
! 1.
RE1/RF1-AAA ext mix gun (6 runs, w/ RIO)
RE1/RF1-AAA ext mix gun (5 runs, w/o RIO)
RE2-Flow coater (3 runs) |
RE3-Pressure-fed roller (3 runs)
Total
emissions
by THC. g
316
305
286
279
440
296
404
389
743
636
267
217
. 403
394
453
332
.762
293
291
272
468
459
524
322
500
402
392
477
445
306
299
RE4-AAA ext mix gun for BPO system (see RF4)
RF2-Low styrene Resin (3 runs)
RF3-Styrene suppressed resin (3 runs)
RF4-BPO-Catalyzed Resin (4 runs)
RF4-BPO-Catalyzed Resin (2 runs, slow gel)
RF4-BPO-Catalyzed Resin (2 runs, fast gel)
RF5-Water emulsified resin (Not tested))
RF6-Styrene suppressed resin plus wax (3 runs)
i
395
286
632
752
512
NA
266
Emissions
during
application, g
(by THC)
196
169
136
142
212
193
207
195
266
310
153
97
192
191
228
22-1
356
126
166
180
216
207
192
134
193
195
190
230
214
'' 143
166
192
180
252
311
193
NA
149
Total
emissions
byMB.g
284
335
267
280
434
302
424
361
694
607
247
203
374
366
442
336
732
265
262
273
428
431
488
289
488
384
375
461
432
296
294
367
279
601
713
488
NA
245
Material
used, g
7712
5445
4919
5041
5978
5663
6160
5979
6116
6133
5152
4872
6566
6870
7000
5328
7467
5371
6689
6423
6958
7256
5606
6040
5698
6098
6096
6581
6670
5619
5096
6472
6258
6222
6792
5652
NA
5912
Transfer
eff. %
92.7
96.2
97.2
97.3
90.8
91.7
91.0
91.4
88.7
80.5
89.9
92.7
91.4
93.2
92.6
96.7
79.5
97.7
92.9
92.9
92.2
93.5
91.7
97.9
91.9
92.2
92.7
90.1
92.0
97.3
97.1
92.0
92.5
88.0
84.1
91.8
NA
91.8
Styrene
content. %
43.29
38.33
38.33
38.33
38.33
43.45
38.33
35.34
42.61
38.33
43.29
43.29
35.34
35.34
38.33
38.33
42.61
38.33
43.45
43.45
38.33
38.33
42.61
38.33
42.61
39.87
39.93
38.33
38.33
38.33
38.33
35.34
43.45
42.61
42.61
42.61
NA
43.29
Emission Factors
%AS
by THC
9.5
14.6
15.2
14.5
19.2
12.0
17.1
18.4
28.5
27.1
12.0
10.3
17.4
16.2
16.9
16.3
23.9
14.2
10.0
9.8
17.5
46.5
21.9
13.9
20.6
16.5
16.1
19.1
17.5
14.2
15.3
17.3
10.6
23.7
26.2
21.3
NA
10.6
Note: THO=THC emission measurement. MB=mass balance measurement. EF=emission factor. %AS=% available styrene.
g/g=g of emission per g of material used.
g/m2
by THC
138
134
125
123
193
130
' 177
170
• 326
279
117
95
177
173
199
146
334
129
127
119
205
201
230
141
219
175
172
209
195
134
131
173
126
277
330
225
NA
117
g/g
by THC
0.041
0.056
0.058
0.055
0.074
0.052
0.066
0.065
0.122
0.104
0.052
0.044
0.061
0.057
0.065
0.062
0.102
0.055
0.043
0.042
0.067
0.063
0.094
0.053
0.088
0.065
0.064
0.073
0.067
0.055
0.059
0.061
0.046
0.101
0.112
0.091
NA
0.046
30
SUMSUM.XLS
-------
Table 6-2. Significance of Variables Affecting Total Styrene Emissions
Comparison
Significance
Pilot experiment
Normal vs. Controlled spraying
High vs. Low air velocity
***
ns
Gel coat experiment
Regular vs. Low-VOC
AAA (ext) vs. HVLP (int)
AAA (ext) vs. HVLP (ext)
HVLP (int) vs. HVLP (ext)
***
ns
ns
ns
Resin experiment
Low-profile vs. Low-styrene
Low-profile vs. Styrene-suppressed
Low-profile vs. Styrene-suppressed+wax '
Low-styrene vs. Styrene-suppressed
Low-styrene vs. Styrene-suppressed+wax
Styrene-suppressed vs. Styrene-suppressed+wax
AAA vs. Flow coater
AAA vs. Pressure-fed roller
Flow coater vs. Pressure-fed roller
Regular vs. BPO-catalyzed (fast gel)
Regular vs. BPO-catalyzed (slow gel)
BPO-catalyzed slow gel vs. fast gel
*
***
***
***
***
ns
***
***
ns
**
***
***
ns = Not statistically significant at O.OS level.
* = Statistically significant at the 0.05 level.
** = Statistically significant at the 0.01 level.
*** = Statistically significant at the 0.001 level.
The emissions from the BPO-catalyzed resin (either neat resin with slow gel or modified
resin with fast gel) are different from the regular resin under controlled spraying. The emissions
from the neat BPO-catalyzed resin are also significantly different from those from the modified
BPO-catalyzed resin.
The following results and discussions are based on total emissions quantified by the
emission measurement method. The percent reduction is based on the averages of total
emissions for the number of test runs conducted under the same conditions. The same
31
-------
percentage of reduction is also achieved when the emissions are expressed in grams/square meter
because the same mold surface area is used in the denominator to calculate the emission factors.
6.1 Distribution of Total Emissions During Application and Curing Stages
Tables 6-3 and 6-4 show the distribution of emissions in gel coat and resin experiments.
The emission quantities are the averages of the number of test runs conducted for that condition.
Total emissions and emissions during the application stage were determined by emission
measurement method using the procedures outlined in Section 5.1.3. The difference between
total emissions and emissions during the application stage is postapplication emissions. The
postapplication emissions included emissions from the wet-out rolling (for resin lamination),
stagnant (curing) period, and the curing of overspray.
Table 6-3. Distribution of Emissions in Pilot and Gel Coat Experiments
Gel coat application condition
Pilot experiment (gel coat spraying)
Normal spraying/high air velocity (3 runs)
Controlled spraying/high air velocity (3 runs)
Normal spraying/low air velocity (3 runs)
Controlled spraying/low air velocity (3 runs)
Gel coat experiment (controlled spraying)
Regular gel coat/AAA-extemal mix (3 runs)
Regular gel coat/HVLP-internal mix (3 runs)
Regular gel coat/HVLP-external mix (3 runs)
Low-VOC gel coat/AAA-external mix (3 runs)
Low-VOC gel coat/HVLP-internal mix (3 runs)
Low-VOC gel coat/HVLP-external mix (3 runs)
Range
Average
Total
emissions
g
514
410
512
372
400
373
387
272
278
284
Emissions
during
application
g
259
157
231
138
151
160
143
94
100
91
% total
50
38
45
37
38
43
37
35
36
32
32-50
39
Post-
application
emissions
g
255
253
281
234
249
213
244
178
178
193
% total
50
62
55
63
62
57
63
65
64
68
50-68
61
32
-------
Table 6-3 shows that 32 to 50 percent (average 39 percent) of total emissions was
emitted during the gel coat spraying stage and the remainder was emitted during the
postapplication (curing) stage. Figure 6-1 shows the amount of styrene emitted for each of the
gel coat application conditions. It is apparent that controlled spraying emitted less styrene than
normal spraying, and the low-VOC gel coat emitted less styrene than the regular gel coat. The
pilot experiment also showed that low and high linear air velocities in the spray booth (between
40 and 100 ft/min) did not have an effect on emissions. Figure 6-1 also shows that there is no
significant difference in total emissions for three different spray guns (i.e., AAA spray gun with
external catalyst mixing and HVLP spray gun with internal and external catalyst mixing). More
detailed discussion is presented in Section 6.3.
Table 6-4 shows that 38 to 63 percent (average 50 percent) of total emissions was
emitted during the resin application stage and the remainder was emitted during the
postapplication stage. The postapplication stage included wet-out rolling and stagnant (curing)
periods. Figure 6-2 shows the styrene emissions quantified for each of the resin application
conditions. It is apparent that the flow coater and the pressure-fed roller resulted in less
emissions than normal or controlled resin sprayup. Low-styrene and styrene-suppressed resins
also emitted less styrene than the low-profile resin. More detailed discussions for the resin
experiment are presented in Section 6.4.
Table 6-4. Distribution of Emissions in Resin Experiment
Resin application condition
AAA-normal spraying (1 run)
AAA-controlled spraying (5 runs)
Flow coater (3 runs)
Pressure-fed roller (3 runs)
Low-styrene resin (3 tuns)
Styrene-suppressed resin (3 runs)
Styrene-suppressed resin + wax (3 runs)
Neat BPO resin - slow gel (2 runs)
BPO resin + thickener - fast gel (2 runs)
Range
Average *
Total
emissions
. g
636
445
306
299
395
286
266
752
512
Emissions during
application
• g
310
214
143
166
192
180
149
311
193
% total
49
48
47
56
49
63
56
41
38
38-63
50
Postapplication
emissions
g.
326
231
163
133
203
106
117
441
319
% total
51
52
53
44
51
37
44
59
62
37-62
50
33
-------
Styrene emissions, g
-.
O
§ O
O>
q
«'
r*
O1
c
Si
O
O
•*
O
w*
10
o"
"5.
5"
ii»
w
a
(Q
2.
o
o
o>
(0
X
•a
ID
(D
I
Normal/high velocity
Controlled/high
velocity
Normal/low velocity
Controlled/low
velocity
CD
2.
o
o
0)
TJ
•g.
I
5'
Q.
PM^
o"
3
Regular/AAA-ext •
RegularWVlP-Int
Regular/HVLP-ext
Low-VOC/AAA-ext
Low-VOC/HVLP-int
Low-VOC/HVLP-ext
-------
Styrene emissions, g
-» M
o o
o o
o
o
O> -J
§ §
Tl
(5*
at
o
O
55*
O
3
O
•*•
§
w
w
o'
to
(D
W
5'
(D
X
•o
-------
6.2 Pilot Experiment Results
Typical emission concentration profiles recorded by the THC analyzer for various test
runs are presented and their resultant emission quantities are used in the following discussions.
A comparison of typical emission concentration profiles provides a clear picture of what
happened when different techniques, equipment, and materials were used. A test run began when
spraying or nonspraying application started and ended when the curing was completed. The end
of the zigzag-like concentration profile indicates the end of the application stage. The average
concentration and the duration of application are used to calculate emissions during the
application stage. As soon as the application is completed, the emission concentration gradually
returns to the baseline concentration during the wet-out rolling and curing stages.
Figure 6-3 shows the typical emission concentration profiles for normal and controlled
spraying test runs for the regular gel coat. Average styrene concentration during the application
stage of the normal spraying test run P6 (59 ppm) is higher than that of the controlled spraying
test run P10 (41 ppm).
Using the total emissions data presented in Table 6-3 and normal spraying/high air
velocity as the baseline condition, Figure 6-4 shows the emission reduction observed in the pilot
experiment. Controlled spraying reduced emissions by 27 percent and 20 percent at low and high
air velocity conditions, respectively.
Table 6-5 shows the summary of emissions for the pilot experiment. Transfer efficiency
increased and gel coat usage decreased when spraying technique improved. Total emissions and
emission factors also reduced when spraying technique changed from normal to controlled
spraying. However, the effects of air velocity under each spraying technique are not significant.
Table 6-5. Summary of Emissions for Normal and Controlled Gel Coat Spraying
Spraying
technique/air
velocity
Normal/High
(3 runs)
Controlled/
High (3 runs)
Normal/Low
(3 runs)
Controlled/
Low (3 runs)
Transfer
efficiency
71.5
82.1
75.7
82.5
Materials used
g
2,135
1,886
2,103
1,850
Reduc.
(%)
—
12
-
12
Total emissions
g
514
410
512
372
Reduc.
(%)
—
20
—
27
Emission factor
%AS
62.1
56.3
62.9
52.0
Reduc.
(%)
—
9
—
17
.Emission factor .
g/g
0.241
0.218
0.244
0.201
Reduc.
(%)
-
10
-
18
Note: Material usage and emission quantities are the averages of the number of test runs for that condition.
%AS = percent of available styrene in gel coat.
g/g = gram of styrene emitted per gram of gel coat material used.
36
-------
a
a
§
a
u
o
U
90
80
70
60
50
40
30
20
10
0
90
80
70
60
50
40
30
20
10
0
Normal spraying of gel coat (run P6) '
•a
a
_
1
o
U
Controlled spraying of gelcoat (run P10)
10
20 30 40
Elapsed time, minutes
50
60
Figure 6-3. Typical emission concentration profiles for normal and controlled gel coat
spraying.
37
-------
30 T
OJ
o
0
c
o
2
•a
a>
25
20
OJ
00
o
I
•o
£
.2
M
J2
til
10
Normal
spraying/Low
air. velocity
Controlled
spraying/Low
air velocity
Normal
spraying/High
air velocity
(baseline)
Controlled
spraying/High
air velocity
Condition
Figure 6-4. Emission reductions observed in the pilot experiment
(based on total emissions, or emission factors in grams per square meter).
-------
6.3 Gel Coat Experiment Results
Figure 6-5 shows the typical emission concentration profiles for a regular gel coat and a
low-VOC gel coat test run. Average styrene concentration during the application stage of the
regular gel coat was 45 ppm, which is higher than the 26 ppm for the low-VOC gel coat. The
emission reduction from the low-VOC gel coat is evident.
Figure 6-6 shows the emission concentration profiles of the AAA spray gun with external
catalyst mixing, the HVLP spray gun with internal catalyst mixing, and the HVLP spray gun with
external catalyst mixing. These concentration profiles are similar in magnitude and the total
emissions are not significantly different among these three spray guns.
Using the total emissions data presented in Table 6-3 and the regular gel coat/AAA spray
gun controlled spraying as the baseline condition, Figure 6-7 shows the emission reductions
observed in the gel coat experiment. The low-VOC gel coat reduced total emissions by 28 to 32
percent when compared to the regular gel coat. The changes in emission among these three
spray guns are not significant.
Because the effect of spray guns on total emissions was not significant, nine regular and
nine low-VOC gel coat test runs were combined to compare the effects of different gel coat
formulations. Total emissions and emission factors were calculated from the averages of nine
test runs in each gel coat formulation and presented in Table 6-6. Table 6-6 shows that the low-
VOC gel coat reduced total emissions by 28 percent when compared to the regular gel coat under
controlled spraying conditions. When the emission factors are presented as percent available
styrene (%AS), there is a little or insignificant reduction of 3 percent. This is because the styrene
contents of the gel coats canceled out the effects of emission reduction in the emission factor
calculation. When the emission factors are presented as g of styrene emitted per g of .gel coat
. applied, the low-VOC gel coat achieved 37 percent reduction. SCAQMD Rule 1162 static
emission measurements for these two gel coats indicate that there is a 37 percent reduction.
Table 6-6. Summary of Emissions for Regular and Low-VOC Gel Coats
Type of gel
coat
Regular
(9 runs)
Low-VOC
(9 runs)
Materials
used
g
1,783
2,025
Total
emissions
g
387
278
Reduc.
(%)
—
28
Emission factor
%AS
56.0
54.2
Reduc.
(%)
—
3
Emission factor
g/g
0.217
0.137
Reduc.
(%)
—
37
Rule 1162 static
emissions
g/m2
133
83
Reduc.
(%)
—
37
Note: Material usage and emission quantities are the averages of the number of test runs for that material.
%AS = percent of available styrene in gel coat.
g/g = gram of styrene emitted per gram of gel coat material used.
39
-------
o
1
«*
o
o
o
o
i
*
o
•55
o
O
90
80
70
60
50
40
30
20
10
0
90
80
70
60
50
40
30
20
10
0
Controlled spraying of regular gel coat (run G6)
Controlled spraying,of low-VOC gelcoat (run G16)
10
20 30 40
Elapsed time, minutes
50
60
Figure 6-5. Typical emission concentration profiles for regular and low-VOC gel
coats.
40
-------
a
a
*
o
!
80
70
i
AAA, external-mix spray gun (run Q5) i
HVLP, external-mix spray gun {run G8J
a
a,
%
o
s
O
u
a
a
i
u
o
u
HVLP, internal-mix spray gun {run G9)
Elapsed time, minutes
Figure 6-6. Typical emission concentration profiles for three types of gel coat
spray guns.
41
-------
Regular
gel
coat/AAA
ext. mix
Regular
gel coat/
HVLP int.
mix gun
Regular
get coat/
HVLP ext.
mix gun
Low VOC
gel
coat/AAA
ext. mix
Low VOC
gel
coat/AAA
ext. mix
Low VOC
gel
coat/AAA
ext. mix
Condition
Figure 6-7. Emission reductions observed in the gel coat experiment
(based on total emissions, or emission factors in grams per square meter).
-------
6.4 Resin Experiment Results
Emission quantities for different resin application equipment are summarized and
presented in Table 6-7. Nonspraying equipment (i.e., flow coaler and pressure-fed roller)
reduced total emissions by 31 to 33 percent when compared to controlled resin sprayup. As
shown in Table 6-1, the statistical analysis indicates that total emissions between the flow coaler
and pressure-fed roller are not statistically significant at the 95 percent confidence level.
Table 6-7. Summary of Emissions for Resin Application Eq
Type of
equipment
AAA spray gun
(controlled
spraying, 5 runs)
Row coaler (3
runs)
Pressure-fed
roller (3 runs)
Materials
used
g
6,670
5,619
5,096
Total emissions
g
445
306
299
Reduc. (%)
BL
31
33
Emission factor
%AS
17.5
14.2
15.3
Reduc. (%)
BL
19
13
uipment
Emission factor
g/g
0.067
0.055
0.059
Reduc. (%)
BL
18
12
Note: Material usage and emission quantities .are the averages of the number of test runs for that equipment.
BL = Baseline condition for emission reduction calculation.
%AS = Percent of available styrene in resin.
g/g = Gram of styrene emitted per gram of resin material used.
The test plan originally called for six runs of controlled spraying. Accidentally, one of
the six test runs was conducted in normal spraying. Table 6-8 compares the emission quantities
for three types of application equipment to the normal spraying test run. Controlled resin . •
sprayup reduced total emissions by 30 percent when compared to normal (uncontrolled) resin
sprayup. Flow coater and pressure-fed roller achieved 52 to 53 percent total emission reduction
when compared to normal resin sprayup.
Figure 6-8 shows the typical emission concentration profiles for various resin application
techniques and equipment. Average styrene concentration during the application stage of the
normal spraying test run RIO was 88 ppm, which is higher than the 60 ppm of the controlled •.:
spraying test run R15. The flow coater and the pressure-fed roller took longer to complete the
lamination but the magnitude of concentration profiles (less than 10 ppm) is much lower than
that of resin sprayup.
43
-------
Table 6-8. Comparison of Emissions among Various Resin Application Equipment and
Normal Spraying
Type of
equipment
AAA spray gun
(normal spraying,
1 run)
AAA spray gun
(controlled
spraying, 5 runs)
Flow coaler (3
runs)
Pressure-fed roller
(3 runs)
Transfer
efficiency
%
80.5
92.0
97.3
97.1
Materials
used
g
6,133
6,670
5,619
5,096
Total emissions
g
634
445
306
299
Reduc.
(%)
BL
30
52
53
Emission factor
%AS
27.1
17.5
14.2
15.3
Reduc.
(%)
BL
35
48
44
Emission factor
g/g
0.104
0.067 '
0.055
0.059
Reduc.
(%)
BL
36
47
43
Note: Material usage and emission quantities are the averages of the number of test runs for that equipment
BL = Baseline condition for emission reduction calculation.
%AS = Percent of available styrene in resin.
g/g = Gram of styrene emitted per gram of resin material used.
Figure 6-9 shows the typical emission concentration profiles for various resin formulation
test runs. The magnitude of the emission profiles decreases as the styrene content in the resin
decreases or the styrene-suppressant content increases. Figure 6-10 compares emission profiles
of the low-profile resin, neat BPO-catalyzed resin, and modified BPO-catalyzed resin. The neat
and modified BPO resins had higher and longer concentration profiles than the low-profile resin
because of higher styrene content or longer gel time. The BPO-catalyzed resin was formulated
for filled application, but it was used in this testing without any filler. Fume silica was added to
the BPO resin to keep the resin material on the mold until it was cured.
Emission quantities for different resin formulations are summarized and presented in
Table 6V9. Low-styrene resin reduced total emissions by 11 percent when compared to controlled
resin sprayup. Styrene-suppressed resin with or without additional wax reduced total emissions
from controlled resin sprayup by 36 to 40 percent. Emission reductions are even higher when the
comparison is based on the normal spraying test run. Statistical analysis indicates that total
emissions from styrene-suppressed resin with and without additional wax are not significantly
different at the 95 percent confidence interval. The neat and modified BPO-catalyzed resins
44
-------
0.
a.
*
0
i
E
O
u
a
a
*
1
S
a
u
o
u
i
*
5
+«
i
I
IUV T
190 -
mo -i.
on _
fiO -
Af\
20 -
I- AAA sorav aun normal soravina techniaue frun R10) I
~ Application ends
^
'
VH^^____
0 . 10 20 30' 40 50 60
160
140 --
: AAA *5r*rsiv niin pnntrnllf*H Qnrauinn t^^hnintio (nin F71*%\ '
4 on i x Application ends
100 '' ^
80 i]
60 -r
40 •
20 - -
iilK
1
fr
I1 , . .
• WS*-*^Y**«^-»*-__ .
0 10 ..20 30 40 50 60
Application ends
on • ^ — !
jrA*u*^-***^*^^^ - F'°W C03ter
-------
160
140
\ i\\ i
'Si
iii
i -
i Low profile resin, normal spraying technique (run RIO) —
IIIIT. ~ • • — — •
• <
V
Low-profile resin, controlled spraying technique (run R15) L
80
Low-styrene resin (run R14)
s
i ill
Styrene-suppressed resin (run R20)
j—
K HJW
/F*'''
I i •
\- . . •
Styrene-suppressed resin plus wax (run R1) j
20 30
Elapsed time, minutes
40
50
60
Figure 6-9. Typical emission concentration profiles for various resin formulations
(applied by AAA spray gun).
46
-------
E
a
a
*
I
g
J
I
s
I
o
I
§
I
160
140
120
100
80
60
40
20
0
160
140
120
100
80
60
40
20
0
160
140
120
100
80
60
40
20
0
Low profile resin, controlled spraying technique (run R15)
,
~ - --~ ~ ~ MAat-RPQ fAcin ^Inw OA! tirvtA /run R171
y i
BPO resin with thickener; fast gel time (run R25)
10
20 30 40
Elapsed time, minutes
50
60
Figure 6-10. Typical emission concentration profiles for low-profile, neat-BPO,
and modified BPO resins (applied by AAA spray gun).
47
-------
Table 6-9. Summary of Emissions for Various Resin Formulations
Type of resin (applied
by controlled
spraying, except as
noted) •
Low-profile (normal
spraying, 1 run)
Low-profile
(5 runs)
Low-styrene (3 runs)
Styrene-suppressed (3
runs)
Styrene-suppressed
+wax (3 runs)
Neat BPO resin (2 runs)
BPO resin + thickener
(2 runs)
Materials
used
g
7,710
6,670
6,472
6,258
5,912
6,792
5,652
Total emissions
g
634
445
395
286
266
752
512
Reduction
(%)
BL
30
38
55
58
-19
19
—
BL
11
36
40
-69
-15
Emission factor
%AS
27.1
17.5
17.3
10.6
10.6
26.2
21.3
Reduction
(%)
BL
35
36
61
61
3
21
—
BL
1
39
39
-50
.-22
Emission factor
g/g
0.104
0.067
0.061
0.046
0.046
0.112
0.091
Reduction
(%)
BL
36
41
56
56
-8
13
~
BL
9
31
31
-67
-36
Rule 1162 static
emission
g/m2
69
69
74
54
49
161
97
Reduction
(%)
BL
BL
-7
22
29
-133
-41
Note: Material usage and emission quantities are the averages of the number of test runs for that equipment.
BL = Baseline condition for emission reduction calculation.
%AS = Percent of available styrene in resin.
g/g = Gram of styrene emitted per gram of resin material used.
-------
emitted more styrene than the low-profile resin. These results contradict the original assumption
that resin catalyzed by BPO might reduce styrene emissions. Possible explanation for this
situation is that the BPO resin was developed for filler application, not for neat resin sprayup and
the gel time for BPO resin was longer than that of the low-profile resin.
Figure 6-11 shows the overall emission reductions or increases observed in the resin
experiment. The reductions or increases are calculated based on the total emission quantities
presented in Tables 6-6,6-7, and 6-8 for each of the conditions. Using low-profile resin and
controlled spraying as the baseline, emission reductions or increases for various equipment and
materials are presented. Except for the BPO-catalyzed resin, all other resin formulations resulted
in emission reduction when compared to the low-profile resin.
Figure 6-12 shows the emission reductions or increases when using low-profile resin and
normal spraying as the baseline. Higher emission reductions were achieved by flow coater,
pressure-fed roller, low-styrene resin, and styrene-suppressed resin. If the normal spraying
technique represents actual practices in the industry, Figure 6-12 shows the potential reduction
that could be achieved by changing to flow-coater and pressure-fed roller and other low-styrene
or styrene-suppressed resins.
6.5 Comparison of Test Results with EPA AP-42 Emission Factors
Emission factors derived from the test results are compared to relevant EPA AP-42
emission factors in Table 6-10. These emissions factors are 1.6 to 2.S times the respective :
midpoints of AP-42 emission factors. The implication of this finding is that current EPA AP-42
emission factors for gel coat and resin sprayup and hand layup operations may underrepresent
actual emissions for these operations.
Tabie 6-10. Comparison of EPA AP-42 Emission Factors and Test Results (in % AS)
Type of material and
operation
Gel coat sprayup (NVS)
Resin sprayup (NVS)
Resin sprayup (VS)
Resin hand layup (NVS)
AP-42 emission
factor range
26-35
9-13
3-9
5-10
AP-42 EF
midpoint
30.5
11
6
7.5
Emission factors from test
results
62.5 (normal spraying)
56 (controlled spraying)
54.2 (low-VOC gel coat,
controlled spraying)
27.1 (normal spraying)
17.5 (controlled spraying)
10.6 (styrene-suppressed
resin, controlled spraying)
15.3 (pressure-fed roller)
Ratio
2.0
1.8
1.8
2.5
1.6
1.8
2.0
49
-------
O
§
u
o
I
I
5
"m
m
i
50
40
30
20
10 -
0
-10
-20
-30
-40
-50
-60
-70
Baseline
Baseline
rosin/Wow
ater
1 - 1 — !•-
Baseline LowVOC Styrene- Bl
In/Pressure- ie&M AAA exi suppressed re
tod roller " " " ' fnijt QUO rosin/ AAA oxt — —
mix aun
9O-cataly*
sin/ AAA i
mix gun
— j_
id B
xt
— e
•O-catalyz
esin (slow
all/AAA e:
mix gun
'
id BI'O-cala!yz
h resin (fast
t — ^lal^AAA-A;
mix aun
id Styrene-
suppressed-
resin/AAA ext
Condition
Figure 6-11. Emission reductions observed in the resin experiment:
"controlled spraying" technique baseline
(reductions based on total emissions, or emission factors in grams per square meter).
-------
9
&
O
|
1
.2
'5
.2
UJ
ou -
eo
40 -
30
2O -
10 -
-i
"10 -
in _
i
i
— ' i " i
resin/AAA ext resin/Flow resin/Pressure-
gun/Controlled
r
1...
-i • •
1 i i
Low vpc Styrene- BPfMntalyMd
resin/ AAA ext suppressed resin/ AAA ext
mix gun- — refin/ AAA *st mix gun
mix gun
B!
-fl
i
'O-catalyz
esin (slow
jIl/AAA e
mix gun
._IIMIII
1
*{j BPO*C9t8lvZ
resin (fast
t — g«l)/AAA-8)
mix gun
{
i
1 I
id Styrene>
suppressed +
resin/AAA ext
Condition
Figure 6-12 Emission reductions observed in the resin experiment:
"normal spraying" technique baseline
(reductions based on total emissions, or emission factors in grams per square meter).
-------
6.6 Comparison of Emission Measurement Method and Mass Balance Calculation
Method
6.6.1 Pure Styrene Evaporation Test
Three sets of pure styrene evaporation tests were conducted at Reichhold in June 1995.
One set was conducted on June 5, the second set on June 7, and a final set on June 29. Within
each set, there were several periods that different styrene evaporation rates were generated. The
results of these different test runs are shown in Table 6-11 and Figure 6-13. The agreement
between the emission measurements and the mass balance calculation method during the pure
styrene evaporation runs was good. Table 6-11 shows that the average ratio of mass balance
calculation method to emission measurements for these 15 runs was 0.99. Figure 6-13 indicates
that the agreement between the mass balance calculation method and emission measurement
method was within ±10 percent for 12 out of 15 runs.
6.6.2 Comparison of Emissions Measured by Two Test Methods
Mass balance measurements were compared with emission measurements during four
experimental test runs and 55 official test runs at the Reichhold Chemicals facility. These 59 test
runs can be separated into:
4 Experimental runs
12 Pilot test runs (involving gel coat spraying)
18 Gel coat application runs
25 Resin application runs
59 Total runs
These two methods for these 59 test runs are compared in Table 6-12 and Figure 6-14.
Prior to June 22, the pilot and gel coat application test runs were made with only the 5-gallon gel
coat supply container on the scale. The amount of material used in each test run was calculated
from the weight loss from the 5-gallon container. It was found that the pump system is a single-
action pump that withdraws material from the container only during the upstroke action.
Therefore, there was a potential error in estimating the exact amount of material dispensed from
the spray gun when the piston pump starts and ends at different positions. In this case, the
amount of material in the pump system could not be accounted for by weighing the container
only. Beginning in run G16, the project team member made sure that the pump started and ended
at the same position, so that the amount of material in the pump system remained the same at the
beginning and the end of a test run. This approach improved the accuracy of the mass balance
calculation method. Table 6-12 indicates that the ratio between the mass balance calculation
method and emission measurement method came close to 1.0 (indicating perfect agreement)
much more consistently between test runs G16 and R2.
52
-------
Table 6-11. Comparison of Mass Balance Method and Emission Measurements for Pure Styrene Evaporation Tests
Date
(m/d/y)
6/5/95
6/5/95
6/5/95
6/5/95
6/5/95
6/7/95
6/7/95
6/7/95
6/7/95
6/7/95
6/7/95
6/29/95
6/29/95
6/29/95
6/29/95
Start Time
(hrmin)
15:46
16:18
16:43
16:56
1 7:09
10:25
10:42
11:06
11:14
11:43
11:58
16:42
16:52
17:00
17:32
End Time
(hrmin)
15:59
16:30
16:55
1 7:06
17:19
10:40
10:57
11:11
11:35
11:55
1 2:05
16:49
16:59
17:07
17:45
Emission rate
by mass balance (MB)
(g/min)
9.2
9.9
17.4
11.8
12.3
40.1
15.3
43.8
9.7
3.0
34.1
48.7
21.3
50.3
29.5
Average concentration
by THC (FID) analyzer
(ppm)
8.4
9.5
14.1
10.6
11.8
39.0
14.8
39.0
9.4
2.6
34.0
49.1
21.5
49.9
31.4
Exhaust
flowrate
(cfm)
8961
8961
8961
8961
8961
8924
8924
8924
8924
8924
8924
8646
8646
8646
8646
Emission rate
by THC (FID) analyzer
(g/min)
9.3
10.4
15.5
11.6
13.0
42.3
15.8
42.4
9.9
2.5
36.9
52.1
22.8
52.9
33.3
Average
Emission
ratio
(MB/THC)
O.99
O.95
1.12
1.02
O.95
O.95
O.97
1.03
O.98
1.20
0.92
O.93
O.94
O.95
O.89
O.99
Ol
U)
10OSSUM.XLS
-------
Ui
1.5
1.4
1.3
o m to
« N '-2 '
ll 1.1
•a o>
m
9
«
a
« 0.9
^ 8 0.8
at 5
OC S
0.7
0.6
0.5
o>
in
o
to
09
(ii
m
«O
n
(O
(O
o
CO
in
o>
§
f-LJ-H-
-+-
• June 5, 1995
D June 7, 1995
E June 29, 1995
r^
m
in
n
in
in
in
O
in
-------
Table 6-1 2. Comoar
Data
Time
EXPERIMENTAL RUNS
6/6/95
6/6/95
7/7/95
7/7/95
10:22
14:55
12:19
14:12
PILOT EXPERIMENT
6/7/95
6/8/95
6/8/95
6/8/95
6/9/95
6/9/95
6/12/95
6/12/95
6/12/95
6/13/95
6/13/95
6/13/95
14:50
10:01
12:11
14:40
10:20
14:45
10:36
13:47
15:44
10:37
12:57
15:21
GELCOAT EXPERIMENT
6/14/95
6/14/95
6/14/95
6/1 5/95
6/1 5/95
6/15/95
6/16/95
6/16/95
6/16/95
6/19/95
6/19/95
6/19/95
6/21/95
6/21/95
6/21/95
6/22/95
6/22/95
6/22/95
10:11
13:46
16:06
11:38
14:03
15:59
10:23
12:49
15:33
10:36
12:47
15:23
10:46
13:18
16:27
11:14
13:41
15:55
RESIN EXPERIMENT
6/23/95
6/23/95
6/23/95
6/26/95
6/26/95
6/26/95
6/27/95
6/27/95
6/27/95
6/28/95
6/28/95
6/2B/95
6/29/95
6/29/95
6/29/95
6/30/95
6/30/95
6/30/95
7/5/95
7/5/95
7/5/95
7/6/95
7/6/95
7/6/95
7/7/95
11:06
14:30
16:35
10:42
13:27
16:05
10:46
13:14
15:51
10:21
12:45
15:20
10:22
12:20
14:36
10:14
13:13
15:51
11:53
14:02
16:19
10:12
14:34
16:41
10:10
Teat run ff
HF1-EXP
GF1-EXP
EXP1
EXP2
P1
P2
P3
P4
P5
P6
P7
P8
P9
P10
P11
PI 2
G1
G2
G3
G4
G5
G6 •
G7
G8
G9
G10
Gil
G12
G13
G14
G15
G16
G17
G18
R1
R2
R3
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
820
R21
R22
R23
R24
R25
son of IV
Exhaust flow
rate, dm
9124
9054
8S10
8681
8909
8980
8953
8681
8818
8704
8864
8796
8681
9034
8987
8909
8764
8680
•8563
8610
8609
8633
8916
8887
8819
8752
8866
8700
8804
• 8739
8660
8775
8820
8722
8710
8683
8399
8563
8495
8495
8613
8543
8358
8501
8471
8501
8566
8495
8633
8495
8521
8457
8637
8471
8419
8589
8661
8492
8658
ass Balance Method and Emission Measurements
Avg. net
cone., ppm
6.35
7.30
0.58
4.48
3.85
7.14
5.05
7.39
4.72
6.21
4.99
7.53
5.92
5.11
6.04
6.02
5.2S
5.30
6.46
3.41
5.41
5.13
3.07
4.97
5.07
5.14
3.18
2.74
3.22
2.85
5.25
3.00
3.21
3.19
2.86
3.52
3.70
3.17
6.27
3.65
5.05
5.16
4.76
7.49
2.81
2.18
5.01
5.04
5.59
4.03
7.10
4.07
3.63
3.51
5.87
5.74
5.55
3.96
S.S3
Test run Total emission*
duration, mini by THC. g
88.8
56.2
102.2
31.2
76.3
68.2
68.8
66.8
77.8
76.3
78.7
68.2
66.9
70.0
71.9
72.3
70.0
70.0
60.3
75.7
70.4
71.2
80.7
70.5
70.2
69.6
85.1
85.0
83.7
87.5
60.7
85.0
85.7
82.4
103.1
81.3
75.0
83.8
67.3
78.0
75.7
71.8
152.4
81.5
91.3
95.2
76.5
75.1
76.5
79.1
102.6
69.3
75.6
74.7
77.2
75.9
88.9
78.0
85.1
631
456
62
149
322
538
382
526
397
506
427
554
422
396
479
476
395
395
409
273
403
387
271
382
386
384
294
•249
291
267
339
274
298
282
316
305
286
279
440
296
404
389
743
636
267
217
403
394
453
332
762
293
291
272
468
459
524
322
500
Total emissions
by MB, g
888
478
65
152
380
607
442
415
412
506
468
499
389
457
535
442
348
406
404
227
355
410
273
472
398
360
327
346
374
332
387
• 302
310
286
284
335
267
280
434
302
424
361
694
607
247
203
374
366
442
.336
732
265
262
273
428
431
48a
289
488
Average
Emission ratio
tMB/THC)
1.40
1.05
1.05
1.02
1.18
1.13
1.16
0.79
1.04
1.00
1.09
0.90
0.92
1.15
1.12
0.93
0.88
1.03
0.99
0.83
0.88
1.06
1.01
1.24
1.03
0.94
1.11
1.39
1.28
1.24
1.14
1.10
1.04
1.01
0.90
1.10
0.93
1.00
0.99
1.02
1.05
0.93
0.93
0.95
0.93
O.S4
0.93
0.93
0.98
1.01
0.96
0.90
0.90
1.00
O.S1
0.94
0.93
O.SO
0.98
1.02
-------
0\
1 .3 '
14.-
W
C -1 0 .
0 > •*
.2 <
' c u> 10 .
. I" '-Z
II 1, -
11 1 '
S-a oa
g « u.a •
i «
2 n P
•g «o O.o -
.6 -
ft R -
•
-4-4
4
4
4
4
4
4
4
4
4
4
4
4-4-
4-
4-
4-
4-
4-
•
4-
I-
4-
4-
4-
4-
4-
4-
4-
4-
4-
4-
4-4-
a. a. .- CM ^-csiM'tmior^oo
-------
Beginning in run R3, resin usage was measured by placing both the 5-gallon container
and the entire piece of application equipment on the scale. This procedure avoided the need to
make sure that the pump was returned to its initial position at the end of the run. With the pump
and the gun on the scale, all material exiting the gun was directly measured. Figure 6-14 shows
that the ratio between mass balance measurements was within +5 percent and -10 percent of 1.0
for runs R3 through R25.
Overall, the mass balance method and emission measurement method were within ±10
percent for 43 out of 59 runs (i.e., 73 percent of the runs) and the two methods agreed to within
±10 percent for all 25 resin runs (the last 25 runs). Table 6-12 shows that the average ratio
between mass balance calculation measurements and emission measurements was 1.02. These
results show that, on average, the two methods agreed to within 2 percent.
Table 6-13 also shows that the mass balance/emission measurement ratio for the resin
runs had a smaller standard deviation than the previous pilot and gel coat test runs. This
comparison shows that correct measurement of the amount of material used in a test run
improved the accuracy of the mass balance calculation method. These test results indicate that,
when proper procedures are carefully followed, the mass balance calculation method can provide
calculated emissions that are in good agreement with emission measurement using EPA Method
25A and a total enclosure.
Table 6-13. Comparison of Mass .Balance and Emission Measurement Test Results
(Testing at Reichhold, June-July 1995)
Test
designation
Experimental
Pilot
Gel coat
Resin .
Total tuns
Weighted average
Number of runs
4
12
18
25
59
Average ratio,
(mass
balance/emission
measurement)
1.13
1.03
1.07
0.96
1.02
Standard
deviation
0.16
0.12
0.14
0.05
0.10
57
-------
Chapter?
Data Quality Issues
Overall data quality met or exceeded the objectives outlined in the Category HI quality
assurance project plan (QAPP). Specific QA activities are presented in the following sections.
More detailed analyses or evaluations are provided in Appendixes D, G, and H (Volume n).
7.1 Summation of Project QA Activities
A quality assurance project plan, Pollution Prevention Technology Demonstration,
Evaluation of Pollution Prevention Techniques to Reduce Styrene Emissions from Open-Contact
Molding Processes, Category III Quality Assurance Project Plan, was prepared by the RTI
project team and submitted to EPA on April 28, 1995, for approval prior to the start of the
proposed testing. Responses to comments from EPA were summarized in a memorandum from
Emery Kong to Carlos Nunez dated May 25, 1995, which is considered an addendum to the
QAPP. Both the QAPP and the addendum are included in Appendix A (Volume 13).
As described in the QAPP, the following QA activities were conducted as part of the project.
The RTI project QA manager, Cynthia Salmons, and William Yeager conducted an internal
technical system audit (TSA) of the project on June 9, 1995.
No formal corrective action requests were necessary for this project. Minor deviations from
the QAPP were documented in laboratory notebooks and data sheets, when necessary. Two
major deviations were (1) the modification of BPO resin after it was found to have a long gel
time and low viscosity and (2) the withdrawal of a water-emulsified resin from the test by the
manufacturer.
Other QA and QC activities during the course of this project included daily calibration of the
high-precision scale with standard weights ranging from 1 g to 2,000 g (described in Section 5.2),
periodic checks for scale drift by leaving a standard weight on the scale overnight, styrene
evaporation experiments (described in Section 6.6.1), comparison of the total hydrocarbon
analyzer's response to styrene cylinders with its response to the propane calibration cylinders
(described in Appendix C), and a comparison of direct injection to the THC analyzer with
delivery through the sampling line, as described in Appendix I (Volume n).
7.2 RTI Internal Technical System Audit (TSA) Results
The internal TSA found that the project activities were generally conducted in accordance
with the QAPP and that results were carefully documented. More extensive calibrations of the
THC analyzer we're performed than were described in the QAPP or the EPA method. The
measurement point for the exhaust flow rate was approximately five or six diameters
downstream of a bend, instead of the eight diameters recommended by the EPA method, but
there was no reasonable way of correcting this. The maximum number of traverse points
58
-------
suggested in EPA Method 1 for this type of situation was used. Checks for off-axis flow did not
indicate a problem. The records for the total enclosure test indicated that there was considerable
fluctuation in the hot wire anemometer readings and that a few of the flow velocity readings at
the natural draft opening were slightly less than 200 ft/min, but this did not seem to present a
problem, judging from the results of the styrene evaporation experiments. Due to the audit
schedule, several aspects of the project were not observed during the TS A. These included
sampling and analysis of the gel coat and resin, the styrene evaporation experiment, the
demonstration that the spray booth meets the criteria for a total enclosure, the weekly traverse
measurements, and the measurement of equipment delivery rate. Records of these activities were
reviewed when possible. A memorandum documenting the TSA activities is included in
Appendix H (Volume n).
7.3 EPA Performance Evaluation
EPA supplied RTI with a performance evaluation styrene standard gas cylinder, which RTI
analyzed on July 7,1995. The results of this EPA performance evaluation are presented in
Appendix I (Volume II). Using the THC analyzer and calibration standards, RTI predicted the
styrene concentration in the EPA performance evaluation standard to be 30.4 ppm. The certified
value of the styrene standard was 31.0 ppm by the Scott Specialty Gases. There was only a 2
percent difference between the predicted and certified values. Therefore, the data quality
objective for emission concentration measurement was met.
59
-------
Chapters
Conclusions and Recommendations
8.1 Conclusions
The results from the pilot experiment indicated that:
• Over the velocity range examined, 12 vs. 30 m/min (40 vs. 100 ft/min) linear air velocity
had no significant effect on styrene emissions.
• Controlled gel coat spraying technique reduced total styrene emissions by 24 percent
compared to normal spraying technique.
• Controlled spraying on the male mold reduced gel coat usage by 12 percent due to less
overspray.
• Under normal spraying, 48 percent of total emissions was emitted during gel coat
spraying; the remainder was emitted during curing.
• Under controlled spraying, 38 percent of total emissions was emitted during gel coat
spraying; the remainder was emitted during curing.
The results of the gel coat experiment indicated that:
• The low-VOC gel coat reduced total emissions by 28 percent when compared to the
regular gel coat.
• The low-VOC gel coat required a higher air supply pressure and larger spray tip to
achieve the same spray fan as the regular gel coat.
• The AAA and HVLP (internal and external catalyst mixing) gel coat spray guns made no
difference in terms of total emissions.
The results of the resin experiment indicated that:
• Controlled resin spraying emitted 30 percent less styrene than normal spraying
technique.
• Flow coater and pressure-fed roller equipment resulted in 31 to 33 percent less styrene
than controlled resin sprayup.
• Flow coater .and pressure-fed roller equipment resulted in 52 to 53 percent less styrene
than normal resin sprayup.
• Thirty-eight to 63 percent (average 50 percent) of total emissions was emitted during the
resin application stage; the remainder was emitted during the wet-out rolling and curing
stages.
• The low-styrene resin emitted 11 percent less styrene than the low-profile resin.
• The styrene-suppressed resin emitted 36 percent less styrene than the low-profile resin.
• The styrene-suppressed resin with 0.1 percent additional wax emitted 40 percent less
styrene than the low-profile resin.
60
-------
• The BPO-catalyzed resin emitted more styrene than the low-profile resin because of
higher styrene content and/or longer gel time.
• For the BPO-catalyzed resin, a shorter gel time reduced total emissions.
Other observations made from this testing were:
• On an average of 55 official test runs and 4 experimental test runs, total emissions
determined by emission measurements and the mass balance calculation method are in
good agreement within 5 percent.
• The mass balance calculation method could potentially be used to determine emissions
from open molding processes.
• Emission factors derived from the test results are 1.6 to 2.5 times the respective mid-
range EPA AP-42 emission factors; this implies that AP-42 emission factors for resin and
gel coat sprayup may underrepresent actual emissions for these processes.
8.2 Recommendations
Based on the findings of this testing, the following recommendations are made to
facilities using the open molding process:
• Train operators to improve their spraying technique to reduce overspray, material wasted,
and emissions.
» Use nonspraying equipment when feasible to reduce emissions. -.
• Use low-styrene or styrene-suppressed materials when feasible.
» Reduce gel time when feasible to curtail emissions. .
• Combine the effects of operator technique, materials, and application equipment to
achieve the maximum emission reduction.
• The mass balance calculation method and in-house personnel can be used to determine
emission factors for materials and equipment.
61
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/R-97-018a
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Evaluation of Pollution Prevention Techniques to
Reduce Styrene Emissions from Open Contact
Molding Processes; Volume I, Final Report
5. REPORT DATE
March 1997
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Emery Kong, Mark Banner, Robert Wright, and
Andrew Clayton
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Research Triangle Institute
P. 0. Box 12194 .
Research Triangle Park, North Carolina 27709
11. CONTRACT/GRANT NO.
CR 818419-03
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 4/94 -.9/95
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES APPCD project officer is Geddes H. Ramsey, Mail Drop 61. 919/
541-7963. Volume II contains appendices.
is. ABSTRACT The repOrt gives results of a study to evaluate several pollution prevention
techniques that could be used to reduce styrene emissions from open molding pro-
cesses in the fiberglass-reinforced plastics/composites (FRP/C) arid fiberglass boat
building industries. Styrene emissions using standard industry techniques, mater-
ials, and equipment were evaluated in a controlled environment and compared to a
basline condition to determine the effects of these, pollution prevention techniques on
styrene emissions. The study found that using controlled spraying (i. e., reducing
overspray), low-styrene and styrene-suppressed materials, and nonatomizing appli-
cation equipment can reduce styrene emissions by from 11 to 52%. Facilities should •'•
investigate the applicability and feasibility of these pollution prevention options to
reduce their styrene emissions. The calculated emission factors were from 1.6 to
2. 5 times the mid-range AP-42 emission factors for the corresponding gel coat and
resin application. These results indicate that facilities using AP-42 emission factors
to estimate emissions in open molding processes are likely to underestimate actual
emissions.
17.
KEY WORDS AND-DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Styrene Resins
Emission
Molding Techniques
Fiberglass-reinforced Plastics
Boats
Pollution Prevention
Stationary Sources
Boat Building
13B
111. 11J
14G
13H
11D
13 J
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
71
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
62
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
National Risk Management Research Laboratory
Technology Transfer and Support Division
Cincinnati, Ohio 45268
OFFICIAL BUSINESS
PENALTY FOR PRIVATE USE S3OO
AN EQUAL OPPORTUNITY EMPLOYER
If your address is incorrect, please change on the above label
tear off; and return to the above address.
If you do not desire to continue receiving these technical
reports, CHECK Wf/ff CD; tear off label, and return it to the
above address.
Publication No. EPA- 600/11-97-018
------- |