RESEARCH  TRIANGLE
   RTI/5522/042-02 FR
INSTITUTE
                            March 16, 1994
                                    FINAL REPORT
     DETERMINATION OF TEST METHODS FOR INTERIOR ARCHITECTURAL COATINGS
                                      Prepared for
                              ICF Work Assignment Manager
                                    ICF Incorporated
                                   9300 Lee Highway
                                 Fairfax, VA  22031-1207
                y 0
                                     Submitted by
                              Linda S. Sheldon, Task Leader
                            Dennis F. Naugle, Program Director

                            Under EPA Contract No. 68-D2-0131
                                Work Assignment No. 2-9
      >. Sheldon
   Task Leader
   Analytical and Chemical Sciences
          E.D. Pellizzari, Ph.D.
          Vice-President
          Analytical and Chemical Sciences
 POST OFFICE  BOX 12194  RESEARCH  TR I ANG LE PARK,  NORTH CAROLI NA  27709-2194

-------
TABLE OF CONTENTS
1.0
2.0




3.0
4.0

5.0



6.0



7.0








8.0




INTRODUCTION 	 	
SUMMARY 	
2.1 ASTM METHODS 	
2.2 BULK ANALYSIS OF PAINT SAMPLES 	
2.3 SMALL CHAMBER METHOD 	
2.4 METALS ANALYSIS 	
RECOMMENDATIONS 	
LITERATURE SEARCH ON ADDITIONAL
OR ALTERNATIVE METHODS 	
PAINT SAMPLES FOR TESTING 	
5.1 SELECTION AND PROCUREMENT OF PAINT SAMPLES 	
5.2 SAMPLE ALIQUOTING 	
5.3 SAMPLE UNIFORMITY AND STORAGE EFFECTS 	
ASTM METHOD 	
6.1 OVERVIEW 	
6.2 METHODS 	 	
6.3 RESULTS 	
BULK PRODUCT ANALYSIS 	
7.1 OVERVIEW 	
7.2 METHODS 	
7.2.1 SAMPLE PREPARATION 	
7.2.2 GC/MS ANALYSIS 	
7.3 RESULTS 	 	
7.3.1 QUALITATIVE IDENTIFICATION 	
7.3.2 QUANTITATIVE ANALYSES 	
7.3.3 STORAGE STABILITY 	
SMALL CHAMBER EMISSIONS TESTS 	
8.1 OVERVIEW AND STUDY DESIGN 	
8.2 METHODS 	
8.2.1 APPLICATION OF PAINT SAMPLE 	
8.2.2 TEST CHAMBERS 	
1-1
2-1
2-1
2-3
2-3
2-9
3-1

4-1
5-1
5-1
5-4
5-4
6-1
6-1
6-1
6-2
7-1
7-1
7-2
7-2
7-3
7-7
7-7
7-14
7-21
8-1
8-1
8-2
8-2
8-4

-------
           8.2.3  SAMPLING AND ANALYSIS METHOD 	    8-7

     8.3    RESULTS 	    8-12

           8.3.1  PERFORMANCE OF PAINT APPLICATION METHODS ...    8-12
           8.3.2  PERFORMANCE OF TEST METHODS	    8-15
           8.3.3  VOC EMISSIONS FROM ALKYD PAINT SAMPLES  	    8-28
           8.3.4  SVOC EMISSIONS FROM LATEX PAINT SAMPLES	    8-40
           8.3.5  ALDEHYDE EMISSIONS FROM PAINT SAMPLES 	    8-53
           8.3.6  EMISSION PARAMETERS 	    8-58

9.0   METALS ANALYSIS	    9-1

     9.1    STUDY DESIGN	    9-1
     9.2    METHOD	    9-1

           9.2.1  ICP METHOD	    9-1
           9.2.2  XRF ANALYSES	    9-4

     9.3    RESULTS	    9-4

10.0  QUALITY ASSURANCE/QUALITY CONTROL	    10-1

     10.1   OVERVIEW	    10-1
     10.2   QUALITY ASSURANCE PROJECT PLAN 	    10-1
     10.3   QUALITY CONTROL SAMPLES  	    10-1

           10.3.1  BLANKS	    10-1
           10.3.2  CONTROL SAMPLES	    10-5
           10.3.3  REPLICATE SAMPLES	    10-5

     10.4   QUALITY CONTROL PROCEDURES	    10-5

           10.4.1  VOLATILES CONTENT 	    10-5
           10.4.2  BULK PRODUCT ANALYSIS	    10-5
           10.4.3  SMALL CHAMBER EMISSIONS TESTING	    10-9

     10.5   QUALITY ASSURANCE	   10-10

11.0  REFERENCES	    11-1


APPENDIX. A      RESULTS OF LITERATURE REVIEW FOR TEST METHODS

APPENDIX B      PAINT SELECTION MEMOS

APPENDIX C      EXAMPLE CALIBRATION CURVE FOR QUANTITATING VOC EMISSIONS
                 FROM ALKYD PAINT SAMPLES

-------
APPENDIX D     EMISSION RATE DATA
               ALKYD PAINTS
               LATEX PAINTS

APPENDIX E      METALS DATA PROVIDED BY EPA/NERL
                                 m

-------
                                   LIST OF TABLES

2-1      Test Methods Evaluated in the Study for
        Interior Architectural Coatings 	     2-2

2-2     Comparison of Data for Chamber Emissions Tests to
        Results for Bulk Product Analysis for Alkyd Paints  	     2-7

2-3     Comparison of Data for Chamber Emissions Tests to
        Results for Bulk Product Analysis for Latex Paints	     2-8

2-4     Percentage of Paint Samples with Measurable
        Concentrations of Metals	    2-10

5-1      List of Paints Selected for Testing	     5-2

6-1      Results of ASTM Tests on Paint Samples  	     6-3

7-1'     Operating Parameters for the Capillary GC/MS System	     7-3

7-2     Calibration  Solutions for Bulk Product Analysis
        of Alkyd Paint	     7-5

7-3     Calibration  Solutions for Bulk Product Analysis
        of Latex Paint  	     7-6

7-4     Identification of Major Chromatographic Peaks in
        Alkyd Paint Samples	     7-8

7-5     Identification of Major Chromatographic Peaks in
        Latex Paint Samples  	     7-9

7-6     Quantitative Results for Target VOCs in Alkyd
        Paint Samples	    7-15

7-7     Quantitative Results for Target VOCS in Latex
        Paint Samples	    7-16

7-8     Precision of Bulk Analysis Method for Alkyd Paints	    7-17

7-9     Precision of Bulk Analysis Method for Latex Paints 	    7-18

7-10    Results of Method Controls for Bulk Product
        Analysis of Alkyd Paint 	    7-19

7-11    Results of Method Controls for Bulk Product
        Analysis of Latex Paint	    7-20

7-12    Results of the Storage Stability Test for
        Alkyd Paints	    7-22
                                            IV

-------
7-13    Results of the Storage Stability Test for
        Latex Paints 	    7-23

8-1      Proposed Chamber Tests for Quantification
        of VOC/SVOC Emissions	     8-3

8-2      GC/MS Operating Conditions for Analysis
        of SVOC Emissions from Latex Paint Samples  	     8-8

8-3      GC/FID Operating Conditions for Analysis
        of SVOC Emissions from Latex Paint Samples  	     8-9

8-4      HPLC Operating Conditions for the Analysis
        of Aldehyde Emissions from Paint Samples	    8-13

8-5      Performance of Paint Application Methods	    8-14

8-6      Analysis of VOC Emissions from Alkyd Paint -
  -'    % Recovery from Method Controls (MC)  	    8-16

8-7      Chamber Recovery Tests for Alkyd Paint Components	    8-19

8-8      Relative Response Factors for the Analysis of
        SVOC Emissions from Latex Paint Samples by GC/MS  	    8-21

8-9      Example Response Factors for the Analysis of
        Latex Paint Using Flame lonization Detection	    8-23

8-10    Estimated Linear Dynamic Range for SVOC Test Method	    8-24

8-11    Analysis of SVOC Emissions from Latex Paint
        Samples Method Controls	    8-25

8-12    Chamber Recovery Tests of Latex Paint Component	    8-27

8-13    Method Performance Data for Aldehyde Testing 	    8-29

8-14    Results of Range Finding Test (Test 2) for VOC
        Emissions from Alkyd Paint - Chamber Air
        Concentration	    8-30

8-15    Results of Range Finding Test (Test 2) for VOC
        Emissions in Alkyd Paint - Chamber Air
        Concentration per Gram of Paint  	    8-31

8-16    Results of Single Chamber Repeatability Tests
        (Tests 5 and 6) for VOC Emissions from Alkyd
        Paints - Chamber Air  Concentration	    8-33

-------
8-17    Results of Single Chamber Repeatability Tests
        (Tests 5 and 6) for VOC Emissions from Alkyd
        Paints - Chamber Air Concentration per Gram
        of Paint	   8-34

8-18    Results of Interchamber Variability Tests
        (Tests 11 and 12) for VOC Emissions from
        Alkyd  Paints - Chamber Air Concentration	   8-36

8-19    Results of Interchamber Variability Tests
        (Tests 11 and 12) for VOC Emissions from
        Alkyd  Paints - Chamber Air Concentration
        per Gram of Paint	   8-37

8-20    Results of Interchamber Variability Tests
        (Tests 13 and 14) for VOC Emissions from
        Alkyd  Paints - Chamber Air Concentration	   8-38

8-21    Results of Interchamber Variability Tests
        (Tests 13 and 14) for VOC Emissions from
        Alkyd  Paints - Chamber Air Concentration
        per Gram of Paint	   8-39

8-22    Results of Air Velocity Tests (Tests 21
        and 22) on VOC Emissions from Alkyd Paints -
        Chamber Air Concentration	   8-41

8-23    Results of air Velocity Tests (Tests 21
        and 22 on VOC Emissions from Alkyd Paints -
        Chamber Air Concentration per Gram of Paint	   8-42
                      \
8-24    Results of Range Finding Test (Test 1) for
        SVOC  Emissions from Latex Paint - Chamber
        Air Concentration	   8-43

8-25    Results of Range Finding Test (Test 1) for
        SVOC  Emissions from Latex Paint - Chamber
        Air Concentration per Gram of Paint  	   8-44

8-26    Results of Single Chamber Repeatability Tests
        (Tests 3 and 4) for SVOC Emissions from
        Latex Paint - Chamber Air Concentration	   8-45

8-27    Results of single Chamber Repeatability Tests
        (Tests 3 and 4) for SVOC Emissions from
        Latex Paint - Chamber Air Concentration
        per Gram of Paint	   8-46

-------
8-28    Results of Intel-chamber Variability Tests
        (Tests 15 and 16) for SVOC Emissions from
        Latex Paint - Chamber Air Concentration	    8-47

8-29    Results of Interchamber Variability Tests
        (Tests 15 and 16) for SVOC Emissions from
        Latex Paint - Chamber Air Concentration
        per Gram of Paint	 .    8-48

8-30    Results of Single Chamber Repeatability Tests
        (Tests 17 and 18) for SVOC Emissions from
        Latex Paint - Chamber Air Concentration	    8-49

8-31    Results of Single Chamber Repeatability Tests
        (Tests 17 and 18) for SVOC Emissions from
        Latex Paint - Chamber Air Concentration
        per Gram of Paint	    8-50

8-32'    Effects of Air Velocity on SVOC Emissions
        from Latex Paint (Tests 19 and 20) - Chamber
        Air Concentration	    8-52

8-33    Effects of Air Velocity on SVOC Emissions
        from Latex Paint (Tests 19 and 20) - Chamber
        Air Concentration per Gram of Paint  	    8-54

8-34    Results of Single Chamber Repeatability Tests
        (Tests 3 and 4) for Aldehyde Emissions from
        Latex Paint - Chamber Air Concentration	    8-55

8-35    Results of Interchamber Variability Tests
        (Test 15 and 16) for Aldehyde Emissions from
        Latex Paint - Chamber Air Concentration	    8-56

8-36    Results of Single Chamber Repeatability Tests
        (Tests 17 and 18) for Aldehyde Emissions from
        Latex Paint - Chamber Air Concentration	    8-57

8-37    Results of the Effects of Air Velocity on the
        Emissions of Aldehydes from  Latex Paint -
        Chamber Air Concentrations	    8-59

8-38    Results of Interchamber Variability Tests
        (Tests 11 and 12) for Aldehyde Emissions
        from Alkyd Paint - Chamber Air Concentration 	    8-60

8-39    Results of Interchamber Variability Tests
        (Tests 13 and 14) for Aldehyde Emissions
        from Alkyd Paint - Chamber Air Concentration	    8-61

-------
8-40     Emission Parameters Calculated from Small
         Chamber Emissions Tests for Alkyd Paints	    8-66

8-41     Emission Parameters Calculated from Small
         Chamber Emissions Tests for Latex Paints  	    8-67

8-42     Comparison of Data for Chamber Emissions
         Tests to Results for Bulk Product Analysis
         for Alkyd Paints  	    8-69

8-43     Comparison of Data for Chamber Emissions
         Tests to Results for Bulk Product Analysis
         for Latex Paints	    8-70

9-1      Target Metals for Liquid Paints by XRF and ICP  	     9-2

9-2      Paint Samples for Metal Analysis  	     9-3

9-3-''     Results of Metals Analysis for Alkyd Paints	     9-5

9-4      Results of Metals Analysis for Latex Paints	     9-8

9-5      Percentage of Paint Samples with Measurable
         Concentrations of Metals	    9-12

'9-6      Method Performance Results for ICP Analysis  	    9-13

10-1     Summary of Quality Assurance Objectives	    10-2

10-2     Summary of Test Methods   	    10-3

10-3     Blank Sample Summary  	    10-4

10-4     Spiked Control Sample Summary	    10-6

10-5     Summary of Precision Measurements	    10-7

-------
                                    SECTION 1.0
                                  INTRODUCTION

      Numerous building materials and consumer products are used in indoor
environments.  In recent years there has been increased awareness that exposure to
pollutants indoors may occur due to emissions of volatile organic compounds (VOCs) from
indoor materials and consumer products. Emissions of volatile organic compounds decrease
over time, but many products are used on a repeated basis indoors.  Some materials such as
architectural coatings/ for example, may be used periodically during remodeling and
renovation of interior spaces.
      The Office of Air and Radiation (OAIR) and the Office of Pollution Prevention and
Toxics (OPPT) of the U.S. Environmental Protection Agency (EPA) have jointly undertaken a
project to analyze the emissions of VOCs from various types of products used indoors.  The
ultimate goals of the project are to determine which types  of products result in the greatest
overall exposures to pollutants indoors and which specific chemical emissions from the
products may present a risk to the populations exposed. Critical to the performance  of this
project is the use of appropriate techniques to evaluate emissions of total and specific VOCs
from these sources.
      A preliminary study was performed by Research Triangle Institute (RTI) for the EPA
in 1992 (1). This preliminary study used an alkyd and a latex interior architectural coating
(IAC) to evaluate seven methods that might be used to determine product content or
emissions  of organic compounds from lACs.  This work assignment is a followup to the
preliminary work and is intended to provide a more detailed analysis of three of those
methods; ASTM standard methods, bulk product analysis, and small chamber testing.
      The purpose of this work assignment was the determination of definitive, valid
methods to test for VOCs, semivolatile organic compounds (SVOCs), total volatile organic
compounds (TVOCs), aldehydes, and metals in interior architectural coatings. The specific
objectives  of the work assignment were:
      •      To perform an evaluation of selected test methods that can be used to
             characterize the composition, and  for small chamber testing, the rate of
             VOC/SVOC and aldehyde emissions from architectural coatings used indoors;
                                         1-1

-------
       •     To perform an evaluation of test methods for the determination of the content
             of metals in these coatings;
       •     To compare the data obtained by the test methods for VOCs/SVOCs,
             aldehydes, or metals;
       •     To obtain definitive, valid test methods for the determination of VOC/SVOC
             and aldehyde emissions from and metal content in lACs; and
       •     Estimate the cost of performing subsequent analyses using each method.
       This report describes the study methods. Analytical results, quality assurance/quality
                                         x
control results, method evaluation results, and methods comparisons are also given. Costs
for routine testing of lACs by the finalized methods have been estimated and are provided.
                                         1-2

-------
                                 SECTION 2.0
                                 SUMMARY

       This study was conducted to provide an evaluation of selected methods to
characterize the composition and /or rate of emissions for VOCs, SVOCs, and aldehydes from
architectural coating used indoors.  It was also designed to evaluate methods for the
determination of metal content in these coatings. Test methods were evaluated using ten
latex paints and ten alkyd paints.  Paints were selected from commercially available brands
to provide a range of gloss types and colors.  Tests were replicated to provide information on
method precision.  Five different methods were evaluated. Table 2-1 briefly summarizes the
methods and their potential applicability to testing of alkyd and latex paints. Prior to testing
with paint samples, modifications to methods  used previously (1) were evaluated and
incorporated into the methods to improve performance.

2.1     ASTM METHODS
       The ASTM methods for gravimetric determination of the volatile content of the paint
and the determination of water by the Karl Fischer method were easily performed.  These
data were then used to estimate total volatile organic concentration in the paints. The
volatile content of the alkyd paints ranged from 30 to 55% by weight. The water content for
all of the six alkyd paints tested was less than 1 %. TVOC concentrations were therefore
nearly  equivalent to the volatile content, ranging from 29 to 54%.  For the six latex paints
tested,  the water content ranged from 45 to 55%, the volatiles ranged from 55 to 65%, and the
TVOC  ranged from 3.5 to 9.5%. The precision  of the method was excellent, with relative
standard deviations (RSDs) less than 0.6% for  the gravimetric determinations.  The precision
of the Karl Fisher method for water determinations was also very good. For the latex paints
RSDs were less than 2.5%. For the alkyd paints, RSD values were higher (1.4 to 17%) but
this reflects the very low water content found  in these samples. The costs for  these
determinations are relatively low (approximately $40 to $50 per sample).
                                         2-1

-------
                      TABLE 2-1.  TEST METHODS EVALUATED IN THE STUDY FOR INTERIOR ARCHITECTURAL COATINGS
            Method
   Collection/Preparation        Analysis8
          Method               Method
                   Information
                    Collected
                      Approximate
                    Cost per Analysis
                  Applicability
      ASTM D2369
      ASTM D4017
                            Gravimetric
                            Karl Fischer
                 Volatile content
                 Water
                          $20
                          $50
               ASTM standard
               method
      Direct analysis
Dilution in solvent
GC/MS
ro
i
      Small chamber
      ICP
      XRF
Collection on charcoal tubes,
solvent desorption with
                          Collection on Tenax TA,
                          thermal desorption
Acid digestion
GC/MS
                            GC/FID
ICP
                            Direct XRF
                            of paint
Compound
identification
and quantitation
Emission rates
over time for
TVOCand
individual VOCs
from alkyd paint
               Emission rates over
               time for TVOC and
               individual SVOCs
               from alkyd paints
Metals content of
paint


Metals content of
paint
 $100-250      Inexpensive - may
               be used to identify
               important
               components for
               emission

$2,000-3,500     Data can be used to
               estimate
               concentrations over
               time and decay in
               indoor
               environments

$2,000-3,500     Data can be used to
               estimate
               concentrations over
               time and decay in
               indoor
               environments

  $50-75        Suitable for
               quantitating metals
               in paint samples

  $30-60        May identify metals
               in paint samples
      aGC/MS - Gas chromatography/mass spectrometry.
      ICP - Inductively coupled plasma emission spectroscopy.
      XRF - X-ray fluorescence spectroscopy.

-------
2.2    BULK ANALYSIS OF PAINT SAMPLES
      Analysis of the liquid paint samples diluted in an appropriate solvent and analyzed
by gas chromatography/mass spectrometry (GC/MS) provided information on the individual
target analytes that could be expected to occur in the emissions from the paint and on TVOC
concentration.
      The alkyd paints contained hundreds of compounds that are predominately branched
chained hydrocarbons. For the gloss and semigloss alkyd paints tested, there was a higher
relative abundance  of the more volatile species compared to the flat finish paints. The most
abundant target VOCs for paints with the gloss and semigloss finish were m,r>xylene,
iv-nonane, ivdecane, and iv-undecane. For the flat alkyd paint the most abundant compounds
were n-undecane, n-dodecane, m^-xylene, and iv-decane. Concentrations for VOCs  in alkyd
paints ranged from 0.5 to 10 mg/g of paint. TVOC concentrations were in the range 300 to
500 mg/g of paint.
      In contrast, the latex paints contained relatively few volatile organic chemicals. The
SVOCs identified during analysis generally accounted for all of the components in the
GC/MS chromatogram. For the various paint samples tested, the presence and relative
abundance of individual VOCs varied between samples with no clear trend for gloss type.
For four of the six latex paints, ethylene glycol was the most abundant compound (19 to 40
mg/g).  1,2-Propanediol (-38 mg/g) and 2-(2-methoxyethoxy)ethanol (23 mg/g) had the
highest concentration in the other two paint samples.  Other compounds with relatively high
concentrations  (>10mg/g) included 2-(2-butoxy-ethoxy)ethanol, and Texanol (2,2,4-trimethyl-
1,3-pentanediol monoisobutyrate).
      Duplicate samples were analyzed for selected paint samples as a way to assess
uniformity of the sample aliquots as well as overall method precision.  Results showed low
%RSD values (generally less than 10%)  for both paint types. These results suggest that both
the replicate sample aliquots were uniform and the precision of the overall method was
good.

2.3    SMALL CHAMBER METHOD
      The small chamber test method  provided quantitative data that could be used to
estimate emission rates for VOCs, SVOCs and, aldehydes from paint samples.  A set of 22
chamber tests were performed, 11 each for the alkyd and latex paints.  Preliminary tests were

                                        2-3

-------
performed as range finding tests to determine the appropriate air sample collection volumes,
sample collection time points, and test durations for the two types of paints.  Duplicate tests
were performed  to determine the recovery of target VOCs/SVOCs from the chamber. Sets of
single chamber repeatability and interchamber variability tests were then performed to obtain
emissions data and results on the performance of the chamber emissions method using alkyd
and latex paints  of different gloss types.  Finally, chamber tests for an alkyd and a latex
paints were performed to evaluate the effect of air velocity on emission rates.  Results were
compared with and without a fan operating in the chamber.
      Data reported for each chamber test included chamber air concentration (mg/m3) and
chamber air concentration (mg/m3) per gram of paint applied in the chamber at each
sampling point. These data were then applied  to various models to estimate emission
parameters for each of the target VOCs, SVOCs, TVOC, and aldehydes.  Emission parameters
included the initial source strength (mg/h.g of paint) and emission decay constants (h*1).
The resulting parameters were then used to estimate the total mass of each target
VOC/SVOC and TVOC emitted per gram of paint during testing.
      For the chamber recovery  tests, good recoveries were calculated for all of the test
compounds under constant concentration conditions suggesting minimal losses of target
VOCs and SVOCs during emissions testing. Results generated between test chambers were
similar suggesting good reproducibilty for the overall chamber test method.
      Results of the chamber tests for the alkyd paint samples showed highest chamber air
concentrations for the n-alkanes (n-decane, n-nonane, and ivundecane).  For the more volatile
compounds, the  highest chamber air concentrations were seen for the earliest sample
collection points. In contrast, the highest chamber air concentrations for the less volatile
compounds were seen at the later time points.  For the flat and semigloss alkyd paints,
chamber air concentrations for the least volatile compounds  (i.e., 2-methyldecane, trans-
decahydronaphthalene, n-undecane, pentylcyclohexane, and ivdodecane) were still relatively
high at the end of the 24-hour test period.  This was in contrast to the concentration
measurements for these same compounds measured for the gloss paint. These chamber
concentration results are consistent with visual observations where the  flat and gloss paints
were still tacky when removed from the chamber after testing.  Variability between tests
using the same paint was evaluated as the %RSD between paired chamber air concentrations
for the two tests. For most cases, RSD values were less than 30% indicating acceptable

                                         2-4

-------
reproducibility. Highest %RSD values were calculated for the latter time points were
chamber air concentrations were low.
       For the alkyd paints, a final set of chamber emissions tests were performed to
evaluate the effect of air surface velocity on VOC emissions.  Results for the test with the fan
(surface air velocity -10 cm/sec) showed higher air concentrations at the earlier time points
with a more rapid decrease in air concentration when compared to results for the test
without the fan (surface air velocity < 2 cm/sec).  In the absence of a fan the less volatile
components still showed relatively high chamber air concentration at the end of the 24-hour
test period.    ,           \     -".'>!-•'•
       Results from applying the chamber air concentration data  to emission models showed
that a decaying source model represented most of the VOCs and TVOCs in the alkyd paint
quite well. Relatively poor fits were seen for compounds with the lowest chamber air
concentrations. Best fits were seen for the TVOC data.  Chamber air concentrations for the
more volatile VOCs appeared to peak at approximately 5 hours.  For the less volatile VOCs,
particularly rv-dodecane and pentylcyclohexane, the chamber air concentrations appeared to
peak well beyond five hours, but there were not sufficient data points beyond this time to
                  - w-- •    ..-.*..-     •  , -\  •• '.
define parameters for a slow buildup* model.  For these less volatile components, the present
models do not define the chamber air concentration very well.  When the chamber tests were
performed with the fan, the models did a better job of describing the chamber air
concentrations, especially for the less volatile species.
       Results of the single chamber repeatability and interchamber variability tests for the
latex paint samples showed several trends.
       •     Reproducibility between paired tests both within a single chamber and across
                                                     •30).
       •     Greatest variability between paired samples was generally found when air
             concentrations were low.
                         <^ "•
       •     For all paint ^types, measured air concentration for  ethylene glycol were high
             (>60 mg/m3).
       •     For the semigloss paint, air concentration  for 1,2-propanediol were very high
             (>200 mg/m3).
                                         2-5

-------
       •     Chamber air concentration for target SVOCs gradually increased over time
             with highest concentrations measured at either 24 or 48 hours. After that time,
             concentrations showed a gradual decrease.
       •     For the semigloss paint, all target SVOCs were at relatively low concentrations
             at the end of the 168-hour test period with only 2-(2-butoxyethoxy)ethanol at
             measurable levels. In contrast, relatively high concentrations of ethylene glycol
             were still present in the chamber air samples for the flat paint (Tables 8-26 and
             8-27) and the gloss paint (Tables 8-28 and 8-29).
       Tests performed to evaluate the effect of surface air velocity on SVOC emissions from
latex paint samples showed the same trends as seen for the alkyd paints.  Basically, the test
performed with the fan showed higher air concentrations at the earlier time points with a
more rapid decrease in air concentrations over time when compared to the paired test
without the fan.
       The chamber air concentrations for most of the SVOCs and TVOC  emitted from the
latex paint were best described using a slow buildup model.  This model did not perform as
well for these emissions as the decaying source model performed for the VOCs emitted from
the alkyd paints.  However it did appear to capture the general pattern of latex emissions.
The poorer agreement between measured and modeled concentrations may be due to the fact
that these relatively polar and less volatile SVOCs are more difficult to analyze and as a
result may have more variability associated with  the measured chamber air concentrations.
       Finally, the emission models were used to estimate the mass of each VOC/SVOC and
TVOC emitted per gram of paint during the small chamber tests.  This information is given
in Table 2-2 for the tests with the alkyd paints. The table also gives the mass of  each VOC
and TVOC  per gram of paint measured during bulk  product  analysis  of the same paint
samples. The difference between the estimated mass emitted and the mass measured in the
bulk paint samples is presented as the  %RSD calculated for the two measures. For TVOC,
the mass per gram of paint estimated from the ASTM methods is also given.  Similar data are
given for the latex paints in Table 2-3.  Results generally show good agreement between the
two measures suggesting that the chamber data and  models can be used to describe the
organic emissions from paint samples.  Poorest agreement is seen for some of the less volatile
VOCs in the alkyd paint where sufficient chamber data were  not available to adequately
                                         2-6

-------
TABLE 2-2. COMPARISON OF DATA FOR CHAMBER EMISSIONS TESTS TO RESULTS FOR BULK PRODUCT ANALYSIS FOR ALKYD PAINTS
GL"-Gloss
(Hyacinth)
Tests 5 and 6
Mec CV1
Compounds (mg/g) (mg/g)
m,£-Xylene 2.3 3.6
n-Nonane 6.9 9.8
o-Xylene 0.76 1.1
Propylcyclohexane 2.6 4.2
3- & 4-Ethyl toluene 1.9 1.8
1,3,5-Trimethylbenzene 0.82 0.79
n.-Decane 21 18
2-Ethyl toluene 0.51 0.62
1,2,4-Trimethylbenzene 2.6 2.8
T3 1,2,3-Trimethylbenzene 0.9 0.84
^ 2-Methyldecane 1.9 2.0
trans-Decahydronaphthalene 2.4 2.1
n-Undecane 10 9.1
Pentylcyclohexane 1.2 0.52
jn-Dodecane >1.7 2.6
TVOC 280 280
380*
1 Glidden.
b Sherwin Williams,
%RSD
31
25
26
33
3.8
2.8
11
13
4.2
7
3.7
8.5
7.3
56
-
0
21

GL-Flat
(Chim Cham)
Tests 11 and 12
Me
(mg/g)
1.3
0.90
0.41
0.26
0.26
0.10
5.0
0.090
0.31
0.12
2.2
3.5
5.5
1.7
5.9
310


cb
(mg/g)
1.6
0.68
0.37
0.20
NO*
ND
4.5
ND
ND
ND
3.2
3.2
19
3.0
12
180
299e

%RSD
13
20
8
18
-
-
7
-
-
-
2.7
5
78
39
46
38
56

GL - Semigloss
(Sea Foam)
Tests 13 and 14
Me
(mg/g)
2.7
4.6
0.98
1.1
1.5
0.61
5.9
0.42
2.0
0.81
3.3
3.1
3.6
2.7
>1.7
220


cb
(mg/g)
4.0
5.3
1.1
1.5
1.3
0.52
14
0.48
1.8
0.64
2.6
2.6
16
2.1
8.1
210
330e

%RSD
27
10
7
19
10
11
58
10
7
17.1
16
13
91
19
-
3.2
28

SW* - Gloss
(Bumbershoot)

Me
(mg/g)
3.4
13
0.65
3.0
039
0.15
25
0.13
0.46
0.19
3.9
4.2
19
2.0
9.7
560


Test 21
cb
(mg/g)
4.3
11
0.83
3.4
ND
ND
19
ND
0.53
ND
3.2
35
16
2.2
7.9
270
540*


7.RSD
16
9
17
9
-
-
18
-
10
-
13.1
13.7
10
5.4
15
49
2.6


Me
(mg/g)
3.1
11
0.65
3.0
0.34
0.18
20
0.12
0.44
0.13
4.3
33
28
>1.5
>1.9
430


Test 22
cb
(mg/g) %RSD
4.3 23
11.2 2.6
0.83 17
3.4 9
ND
ND
19 2.5
ND
0.53 14
ND
32 21
35 4.0
16 38
2.2
7.9
270 32
540* 16

c Estimated mass per gram of paint during chamber tests.
d Measured concentration measured during bulk product analysis.
e Below the method quantitation limit.
' Pctimafoff f*nn/»onfrafirtn frrtm A 
-------
     TABLE 2-3.  COMPARISON OF DATA FOR CHAMBER EMISSIONS TESTS TO RESULTS FOR BULK PRODUCT ANALYSIS FOR LATEX PAINTS
CO
SW'-Flat
(Marmalade)
Tests 3 and 4

Compounds
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
TVOC

Formaldehyde
Acetaldehyde
Mec
(mg/g)
2.6
37
1.2
14
49

0.08
0.53
chd
(mg/g)
ND
29
1.5
5.1
36
65e
NM'
MM

%RSD
-
17
18
66
22
20
-
-
SW - Gloss
(Rose Dawn)
Tests 15 and 16
Me
(mg/g)
NC
60
25
2.7
88

0.03
0.14
Cb
(mg/g)
ND
48
' 13
27
65
84e
NM
NM

%RSD
-
15
46
1
21
3
-
-
GLb - Semigloss
(Sea Foam)
Tests 17 and 18
Me
(mg/g)
79
45
11
89
140

0.02
0.06
Ct
(mg/g)
38
19
4.4
5.7
68
93e
NM
NM

%RSD
49
57
60
124
49
31
-
-
SW - Flat
(Marmalade)

Me
(mg/g)
1.9
25
1.1
14
34

0.12
NC«
Test 19
Cb
(mg/g)
ND
29
1.5
5.1
36
65«
NM
NM
Test 20

%RSD
-
10
24
66
3
44
-
-
Me
(mg/g)
2.1
30
1.3
16
41

0.11
0.21
Cb
(mg/g)
ND
29
15
5.1
36
65«
NM
NM

7.RSD
-
2
12
73
10
32
-
-
• Sherwin Williams.
b Glidden.
c Estimated mass emitted per gram of paint during chamber tests.
d Concentration measured during bulk product analysis.
e Estimated concentration from ASTM methods.
' Not measured.
8 Not calculated - curve did not adequately describe the data.

-------
model the emissions.  The reason for the very poor agreement between the two measures for
Texanol from the latex paints is unknown.

2.4    METALS ANALYSIS
      Two methods were evaluated for the analysis of metals in alkyd and latex paints. The
first method used X-ray fluorescence spectroscopy (XRF) on untreated liquid paint samples.
The second method used inductively coupled plasma emission spectroscopy (ICP) on
previously digested paint samples.  The XRF method was performed on single aliquots of
each of 20 paint samples.  The ICP method was performed on triplicate aliquots for the same
20 paint samples to provide data on metal concentrations and method precision.
      The ICP method showed acceptable performance in terms of low background
contamination in method blanks and good recovery for spiked blanks and spiked paint
samples. Variability evaluated as %RSD was generally low (<30%) for most metals measured
in replicate samples. For some samples where precision was poor for all metals, high
variability appeared to be a result of poor sample homogeneity.  For other samples,
incomplete digestion may have caused interferences with specific metals in the sample.
Performance of the XRF method was not evaluated on this work assignment.
      Table 2-4 gives the percentage of samples with measurable concentrations of the target
metals.  Percent measurable values are provided by manufacturer and paint type as well as
for all paint samples.  For the metals that were detected, aluminum gave the highest
concentrations ranging from 1.5 to 40 mg/g. Concentration results reported by the ICP and
XRF methods were not in good agreement and generally differed by a factor of two or more.
                                        2-9

-------
             TABLE 2-4. PERCENTAGE OF PAINT SAMPLES WITH MEASURABLE CONCENTRATIONS OF METALS
I
I—1
o
% Measurable
Latex
Quantitative
Limit (ug/g)

Aluminum
Selenium
Barium
Antimony
Cobalt
Cadmium
Arsenic
Chromium
Copper
Strontium
Lead
Manganese
Molybdenum
Nickel
Mercury
Tin
Zinc
Bismuth
Calcium
Titanium
Iron
ICP
50*
60
0.5
40
10
1.0
30
2.0
2.0
1.0
15
2.0
5.0
15
30
75
—
—
—
—
—
XRF
—
2
50
5
20
5
20
5
2
—
10
5
5
15
10
25
1
5
0.1%
NRb
NRb
Sherwin
Williams
ICP
100
0
80
0
100
0
0
60
40
80
20
80
0
0
80
0
—
—
—
—
—
XRF
—
0'
—
0
0
0
0
40
100
—
40
20
0
80
0
0
80
0
100
100
100
Glidden
ICP
100
20
80
20
100
0
60
100
20
100
80
60
0
0
60
0
—
—
—
—
—
XRF
—
0
0
0
20
0
0
20
100
—
80
60
0
40
0
0
100
0
100
100
100
Alkyd
Sherwin
Williams
ICP
100
20
100
40
100
0
40
100
20
100
80
80
0
40
40
20
—
—
—
—
—
XRF
—
0
0
0
0
0
0
20
100
—
60
20
0
20
80
0
100
0
80
100
100
Glidden
ICP
100
0
80
0
80
0
40
100
20
100
0
100
0
0
40
0
—
—
—
—
—
XRF
—
0
0
0
0
0
0
40
100
—
20
20
100
100
20
0
100
0
20
100
100
ALL
ICP
100
0
90
15
95
0
35
90
25
95
45
80
0
10
55
5
—
—
—
—
—
XRF
—
0
0
0
5
0
0
30
100
—
50
30
25
60
25
0
95
0
75
100
100
     a Not analyzed by test method.
     b Not reported.

-------
                                    SECTION 3.0
                                RECOMMENDATIONS

      Throughout this work assignment, the methods proposed for testing appeared to
work well and provided suitable data for assessing emissions from paint samples.  For the
small chamber tests, the model fitting process can adequately explain most of the features of
the paint emissions, characterizing the analytes in terms of emission factors and decay rates.
There are, however, some inadequacies that could be profitably addressed. These are
discussed as recommendations below.
      1.     Placing the painted plates into the chamber sometimes introduces a large
             quantity of paint vapors.  This can be accounted for in the fitting process, but
             introduces a variable into the measurement that would be better removed. A
             brief, high volume gas purge of the chamber during and immediately after the
             plate is set into  the chamber would lower any vapor concentration significantly
             and reduce its impact on the fitting process.
      2.     Some of the alkyd VOC analyte emissions are not well-defined after 24 hours,
             particularly those less volatile than jvundecane. If these need better
             characterization, sampling should be performed at roughly 6-8 hour intervals
             for the period from 12 hours after application to 48 hours after application.
      3.     The source models used for  the SVOCs and VOCs showing gradual  increases
             in chamber concentrations over time could be better justified. An investigation
             of different, diffusion-limited emission models might find a physically
             reasonable model that  would better represent the emission rates. It would
             then give more  reliable estimates  of the total emissions and the long-time
             concentrations.
      4.     Consideration should be given to models that combine a fast-peaking
             component with a slower-peaking component. The slow-peaking component
             may not contribute much to the maximum concentration, but does keep the
             long-time concentration much higher than expected from the fast-peaking
             model.
                                         3-1

-------
                                    SECTION 4.0
        LITERATURE SEARCH ON ADDITIONAL OR ALTERNATIVE METHODS

      Prior to initiating any experimental work, a review was performed to identify
additional or alternative methods for characterizing emissions from lACs that should be
evaluated as part of this work assignment. To accomplish this, a computerized literature
search was performed to identify published reports or methods for the determination of
VOCs, SVOCs, aldehydes, and metals in the liquid IAC products, as well as methods for the
measurement  of emissions of VOCs, SVOCs, and aldehydes from liquid  products.
Keywords for the search were determined by reviewing in-house literature on the subject.
Keywords included paints, coatings, metals, specific metals (e.g., lead, titanium), aldehydes,
formaldehyde, VOCs, SVOCs, and others. A hierarchical search  was performed to obtain a
manageable list of titles and abstracts to review.  Relevant reports were obtained at local
libraries or ordered. In-house literature was reviewed, including the ASTM Section 6
volumes on paints and coatings (2).  Contacts were also made with researchers in the U.S.
who have reported results of these types of tests.
      A letter report describing the results of this review was submitted to the EPA Work
Assignment Manager (WAM). A copy of this report is given in Appendix A.
                                        4-1

-------
                                    SECTION 5.0
                           PAINT SAMPLES FOR TESTING

5.1     SELECTION AND PROCUREMENT OF PAINT SAMPLES
       The lACs tested during the work assignment were limited to latex (water-borne) and
alkyd (solvent-borne) paints. Specific paints for testing were selected based on four factors:
       •      type of paint - alkyd versus latex; within the latex category, a further
             distinction was made between those based on vinyl acetate and acrylic
             polymers,
       •      type of gloss - flat, semi-gloss, or gloss,
       •      color - a random assortment of colors were obtained in an attempt to represent
             a wide range of metallic and organic pigments, and
   ,/   •      grade of paint - a medium to high grade, most typical of that purchased for
             residential applications.
A total of twenty paint samples  were selected and purchased for volatile emissions testing
and metals analysis. Twelve paints were selected for VOC testing. These twelve plus an
additional eight paints were used for metals testing.
       Paints for emissions testing were selected to provide a broad range of paint types in
order to assure that the tested methods are applicable to a wide variety of paints. The focus
of this work assignment was on comparisons between paint types.  In order to address
differences between alkyd  and latex paints and three major gloss types, six different paints
were tested. Those selected are  shown in Table 5-1. To address differences between
manufacturers, paints manufactured by Sherwin-Williams, which has 16% of  the market
share for architectural coatings (3) and Glidden (13% market share) were selected.  Since latex
paints based on vinyl acetate copolymers may contain aldehydes, an attempt  was made to
select vinyl latex paints from each manufacturer.  However, vinyl latex paints were available
only from Sherwin Williams.
       Twelve paints were selected for bulk product analysis by GC/MS, total volatile and
water content by the ASTM methods, and small chamber emission tests.  These twelve plus
an additional eight paints were analyzed by the two metals methods.  Paints  were selected
based on the following criteria:
                                         5-1

-------
             TABLES-!. LIST OF PAINTS SELECTED FOR TESTING
Paint
Type3
SHERWIN
Alkyd
Alkyd
Alkyd
Latex
Latex
Latex
Latex
Latex
Alkyd
Alkyd
GLIDDEN
\
Alkyd
Alkyd
Alkyd
Latex
Latex
Latex
Latex
Latex
Alkyd
Alkyd
Gloss
Type
WILLIAMS
Flat
Semi-gloss
Gloss
Flat
Semi-gloss
Gloss
Flat
Semi-gloss
Flat
Semi-gloss

Flat
Semi-gloss
Gloss
Flat
Semi-gloss
Gloss
Semi-gloss
Gloss
Flat
Gloss
Manufacturer
Series

ProMar 200
ProMar 200
ProMar 200
ProMar 200
ProMar 200
ProMar 200
ProMar 200
ProMar 200
ProMar 200
ProMar 200

5700
UH8000
4550
3480
UH6380
6918
UH6300
6987
5718
4550
Color
Group

Yellow
Blue
Green
Orange
Purple
Red
Green
Other1*
Other
Other

Yellow
Green
Purple
Red
Blue
Orange
Blue
Orange
Green
Other
Manufacturer's
ID No.

SW1352
SW1529
SW1435
SW1629
SW1545
SW1604
SW1734
SW1125
SW1003
SW1309

25312
46212
76262
01044
64984
16112
64542
20573
34722
20852
Color Name

Crescent Cream
Violet Veil
Bumbershoot
Marmalade
Vibrant Violet
Rose Dawn
Grass Roots
Praline
First Star
Coral Canyon

Chim Cham
Seafoam
Hyacinth
Tomahawk
Down Yonder
Orange Glaze
Ice Cap
Orange Ice
Antigua
Sheriffs Star
a  Paints 1 through 6 of each manufacturer were used for analysis by the ASTM
   methods, bulk product analysis, metal analysis, and small chamber tests; paints 7
   through 10 were used for metals analyses only.
b  Other refers to paint colors that could not be classified in the basic color groups
   (i.e., greens, browns).
                                      5-2

-------
       •      Selections included five latex and five alkyd paints manufactured by Glidden
             and five latex and five alkyd paints manufactured by Sherwin-Williams.
       •      Selections from Sherwin Williams included two flat, two semi-gloss, and one
             gloss latex paint and two flat, two semi-gloss, and one gloss alkyd paints.
       •      Selections from Glidden included one flat, two semi-gloss and two gloss latex
             paints and two flat, semi-gloss and two gloss alkyd paints.
       •      The Sherwin Williams latex contained vinyl acetate polymers.
       •      Paint colors were selected at random after stratification for the six primary
             color groups (red, orange, yellow, green, blue, red).  Manufacturer's color
             charts were obtained for the paints selected.  The color chart  was divided  into
             six strata based on primary color group. Each color available within that strata
  ''         were assigned a sequential number. A random number generator was used to
          •
             select one color from each strata. This procedure was performed first for the 6
             paints from each manufacturer selected for bulk product analysis. Then
             sequential numbers were assigned to all paint colors on the chart (no
             stratification by color group) for each manufacturer. Four additional paint
             colors were selected at random from each manufacturer to obtain the eight
             additional paints needed for metals analysis.
       Medium to high  grade paints were  selected for testing. The Glidden paints
represented "homeowner"  used paints.  These paints are typically sold in home improvement
stores and represent the medium grade paints. The Sherwin Williams paints represented a
"top of the line" contractor paint. Selection of this series of paint was based on the
availability of all gloss types in the same paint series.  The Sherwin Williams series also
contained vinyl polymers in the latex paints.  Selection of the paint series was made in
consultation with the ICF and EPA WAMs.
       A final list of the paints selected were provided to the ICF and EPA WAM for
approval prior to purchase. Paints were purchased in the Raleigh/Durham, NC area from
the company retail outlets. Two gallons of each paint were purchased, one for use in these
tests and one to archive.  Procedures for collection of product information, such as lot
numbers and Material Safety Data Sheets (MSDSs) were performed as in the previous study
                                          5-3

-------
on lACs performed by RTI (1).  Additional information was unavailable from the
manufacturer on product color and pigment content.
       Additional information pertaining to the paint selection are presented in Appendix B.

5.2     SAMPLE ALIQUOTING
       Each one gallon can of paint purchased for testing was divided into a number of
small aliquots to be used for all subsequent testing. Individual aliquots were prepared for
each test.  Generally, an aliquot was used for a single test or analysis and once opened was
not used again.  For ASTM method 2369 and 4017 the same aliquots were used.  By doing
this, fresh, unexposed aliquots were available for all tests.
       To aliquot each paint sample, a one gallon can of paint was mixed at the store after
addition of the color pigments using the store's paint shaker.  The paint was then delivered
to the laboratory. Immediately prior to aliquoting, the paint sample was again thoroughly
mixed by  placing it in a rotating mixer and tumbling it end-over-end for at least 1 hour (1620
revolutions). The can was then removed from the mixer, opened, and the paint gently
poured directly into a clean, acid washed 2 L glass separatory funnel for dispensing. Vials to
be used for aliquoting were cleaned and labeled.  Three 6 mm glass beads were placed in
each vial as an aid for mixing the sample immediately prior to use.  The labeled vials and
bottles were then quickly filled with paint and  sealed. After aliquoting, the vials were
weighed.  Aliquots were organized by paint code and aliquot number in boxes and stored
protected  from light at room temperature until tested. Chain of custody/aliquot tracking
forms were prepared for each sample. Examples of these forms are shown in Figures 5-1 and
5-2.

5.3     SAMPLE UNIFORMITY AND STORAGE EFFECTS
       The procedures for dispensing and storing  the aliquots were designed to provide
uniformity between aliquots and to minimize losses of volatile components during storage.
In order to assess the uniformity between aliquots, analysis of duplicate aliquots was
performed using the ASTM methods D2369 and D4017 for total volatile content and  water
content respectively. Additionally, the bulk product analysis by GC/MS was performed  on
duplicate  aliquots of one latex flat, latex semigloss, alkyd semigloss, and alkyd gloss paint.
Aliquots selected for these analyses were randomly selected and were not sequential aliquots.

                                          5-4

-------
                                           CHAIN OF CUSTODY RECORD -N
                                  ARCHITECTURAL COATINGS STUDY-PROJECT 5522
PAINT ALIQUOT CODE:
TESTMETHOD:
TESTDATA - NOTEBOOK NO:.
SW200-1629-LVFO-28
         PAGES;
SAMPLE CODE:
                            SAMPLE COLLECTION       STORAGE
MEDIA  ID    DATE        LOCATION
                                           RECEIVED
              SAMPLE ANALYSIS
ID   DATE   ID    DATE
FILE NO.
                                        RESEARCH TRIANGLE INSTITUTE, ANALYTICAL AND CHEMICAL SCIENCES
                                             P.O. BOX 12194. RESEARCH TRIANGLE PARK. NC 27709

                                    Figure 5-1.  Chain of Custody Record

-------
                             PAINT ALIQUOT TRACKING SHEET
                        ARCHITECTURAL COATINGS—PROJ. 5522-022
MFG. CODE:
MFG. SERIES:
MFG. ID NO:
PAINT TYPE:
GLOSS TYPE:
COLOR GROUP:
MFG. COLOR:

PAINT CODE:
        SW
        200  (ProMar)
DATE:
VOLUME:
       1435
        AD
Bumbershoot
SW200-1435-ADGG
  ALIQUOT  STORAGE   RELINQ.    REC'D             TEST
  VIAL NO.   LOCATION     ID        ID    DATE    METHOD
                                               COMMENTS
    -01
    -02
    -03
    -04
    -05
    -06
    -07
    -08
    -09
    -10
    -11
    -12
    -13
    -14
    -15
    -16
    -17
    -18
    -19
    -20
    -21
    -22
    -23
    -24
    -25
    -26
     -27
                           RESEARCH TRIANGLE INSTITUTE. ANALYTICAL AND CHEMICAL SCIENCES
                                P.O. BOX 12194. RESEARCH TRIANGLE PARK, NC 27709

                       Figure 5-2.  Paint Aliquot Tracking Sheet

                                          5-6

-------
       Effects of storage and sample uniformity after several months of storage were
evaluated by repeating the GC/MS bulk product analysis.  Also, the weight of each vial
measured at the time of aliquoting was compared to the weight at the time of analysis as an
indicator of overall aliquot integrity.  *
                                         5-7

-------
                                    SECTION 6.0
                                  ASTM METHOD

6.1    OVERVIEW
      The ASTM Methods D2369 and D4017 were performed on the 12 paint samples
designated in Table 5-1 to determine total volatile content and water content, respectively.
Method D2369 is a gravimetric method; D4017 is the Karl Fischer titration method for water
determination. The methods are described in the ASTM Volume 6.01 (4). As discussed in
Section 5, paints selected for testing on this work assignment were mixed, then aliquotted
into small glass jars and stored for all subsequent testing (ASTM methods, bulk analysis,
small chamber tests).  To evaluate sample uniformity between jars, measurements using the
ASTM methods were performed  on samples taken from duplicate jars.  In addition,
measurements were made on duplicate samples taken from each jar to obtain data on
method precision.

6.2    METHODS
      The ASTM Standard Practice for Determining Volatile Organic Compound (VOC)
Content of Paints and Related Coatings, ASTM D3960-91 (ASTM D3960-91 (2) describes the
tests used in this study for determination of VOC and water content in the paints.
      The ASTM Standard Test Method for Volatile Content of Coatings, ASTM D2369-90,
(4) was followed for the gravimetric determination of the volatile content.  To perform the
test, the paint was drawn into a disposable 3-mL syringe, which was then weighed.
Approximately 0.5 grams of paint was dispensed into a tared aluminum weighing pan. The
syringe was then re-weighed to determine the exact mass dispensed into the pan. This
technique minimized loss of volatile components during preparation of the test samples. The
samples were dried at 112°C for  60 minutes. The pan with the paint sample was re-weighed
to determine the mass of solids remaining. The percent volatile content was determined by
difference. Two blank samples consisting of empty aluminum weighing pans were also
weighed and dried.  The differences between the pre- and post-drying weights of the blanks
were less than 0.0003 g.
      The water content of the paint samples was determined according to  the ASTM
Standard Test Method for Water in Paints and Paint Materials by Karl Fischer Method,

                                        6-1

-------
ASTM D4017-90 (5).  The instrument used was a Fisher Coulomatic K-F Titrimeter (Model
447, Fisher Scientific, Pittsburgh, PA).  This model determines moisture content of samples by
automatic titration with coulometrically generated Karl Fischer reagent (Coulomat A and
Coulomat C - both pyridine free).

6.3    RESULTS
       Results for the ASTM test are given in Table 6-1 for % nonvolatile, % volatile, and %
water content. Calculated % TVOC values for each sample are also provided. This estimate
was made for each sample by subtracting the mean % water content from the mean %
volatiles content. For all measures, the mean paint content as well as the % RSD for replicate
analyses and replicate aliquots are provided. Results show excellent precision (i.e., low %
RSD values) for all measures both within and between samples, indicating both good method
precision and sample uniformity. Highest % RSD values (1.4 to 16) were calculated for %
water content of alkyd paints.  This is  expected since the water content of these paints is so
low (i.e., >1%).
                                          6-2

-------
                                     TABLE 6-1. RESULTS OF ASTM TESTS ON PAINT SAMPLES
o>
i
CO
% Nonvolatiles3
Paint Type
SHERWIN WILLIAMS
• Latex
• Flat (Marmalade)
- Gloss (Rose Dawn)
- Semigloss (Vibrant Violet)
• Alkyd
- Flat (Crescent Cream)
- Gloss (Bumbershoot)
- Semigloss (Violet Veil)
GLIDDEN
• Latex
- Flat (Tomahawk)
- Gloss (Orange Glaze)
- Semigloss (Down Yonder)
• Alkyd
- Flat (Chim Cham)
- Gloss (Hyacinth)
- Semigloss (Seafoam)
Mean8


44.20
44.14
41.53

69.20
45.75
61.03


42.89
45.64
36.71

69.70
61.02
66.41
Mean
% RSD for
Replicate
Analysis

'•
0.30*1
0.09
0.15

0.16
0.10
0.28


0.24
0.67
0.40

0.11
0.19
0.22
%RSD
for
Replicate
Aliquots


0.52
0.18
0.32

0.16
0.21



0.20
0.87
0.35

0.16
0.31
0.58
Volatiles3


55.8
55.9
58.5

30.8
54.3
39.0


57.1
54.4
63.3

30.3
39.0
33.6
Mean


49.23
47.51
51.60

0.26
0.60
0.15


53.52
45.32
54.02

0.33
0.85
0.50
% Water"
Mean
% RSD for
Replicate
Analysis


0.64
2.0
0.96

6.14
2.35
158


0.64
1.72
0.23

6.26
5.69
6.75

% RSD for
Replicate
Aliquots


0.93
2.22
0.79

6.01
952
1.45


1.31
1.66
0.68

8.18
5.32
16.28
%TVOCC


6.6
8.4
6.9

30.5
53.7
38.8


3.6
9.0
9.3

30.0
38.1
33.1
          "Measured using ASTM method D269-90
          bMeasured using ASTM method D4017-90
          'Calculated as % volatiles - % water

-------
                                    SECTION 7.0
                             BULK PRODUCT ANALYSIS

7.1    OVERVIEW
      Bulk product analysis is performed by diluting the paint sample with an appropriate
solvent then analyzing the resulting solution by GC/MS to identify and quantify organic
components.  The method should provide information on the identity and concentration of
individual VOCs/SVOCs as well as the TVOC concentration in paint samples.  Results of
analysis should also identify the VOCs/SVOCs that are likely to occur in emissions from
paint samples and may also predict emission rates for these compounds.
      In our preliminary study (1), the alkyd paint was diluted with n-hexane and the latex
paint was diluted with a solution of 20%  water/80% methanol. r»-Hexane was an
appropriate solvent for diluting the alkyd paint, except that the calculation of TVOC
concentrations could not be made by integrating the total ion chromatogram beginning with
hexane, as is done for air samples. For latex paints, the use of water/methanol for diluting
the latex paint may have resulted in some polymerization of the organic paint components.
To overcome  these problems, methods for preparing the samples for bulk product analysis
were further refined and evaluated on this study.  Modifications included a change in
dilution solvents and improved methods  for removing solids from the diluted samples.
      During testing, bulk product analysis was performed on six paints (latex flat, latex
semigloss, latex gloss, alkyd flat, alkyd semigloss,  and alkyd gloss) manufactured by Sherwin
Williams and a comparable set of six paints manufactured by Glidden.  The identity of these
paint samples is given in Table 5-1. The first analysis was performed to identify the eight
most abundant VOCs/SVOCs in the twelve paints. Reference standards were then procured,
calibration standards prepared at multiple concentration levels, and quantitative analysis
performed.
      For four of the paints (e.g., Sherwin Williams latex flat, Glidden latex semigloss,
Glidden alkyd semigloss, and Sherwin Williams alkyd gloss), duplicate paint aliquots (paint
samples taken from separate vials) were diluted and analyzed to determine method
precision. A single aliquot of the other eight paints was analyzed  to obtain data on the major
analytes in  the paints.
                                         7-1

-------
       Results of the proposed tests were intended to provide data on the appropriate
dilution solvents, the precision of the method, and the composition of six different types of
paints formulated by two different manufacturers.

7.2    METHODS
7.2.1   Sample Preparation
       Paint samples for the bulk product analysis were prepared by diluting a known
weight of paint, either alkyd or latex, to a fixed volume using a suitable solvent.  All alkyd
paints were diluted in jvpentane and all latex paints were diluted in acetone.  Prior to
aliquoting the paints for testing, sample vials containing the paint were first shaken then
vortex mixed for 60 seconds to assure the thorough mixing of the paint within the vial. A
pipette was used  to transfer paint directly into a clean, tared 15 mL graduated centrifuge
tube.  The centrifuge tube containing paint was then weighed. The amount of paint (PA)
transferred to the tube was calculated as
                                    PA = WTa - W-n,                       (7-1)
where WTa and W-^, are the weights of the tube after and before paint was added.  The
sample was immediately diluted to volume (10 mL) using the appropriate solvent and spiked
with a known amount of external quantitation standard. The tube containing the diluted
sample was sealed and vortex mixed for approximately 30 seconds. To facilitate  the removal
of particulates and polymers from the sample prior to GC/MS analysis, the samples were
centrifuged approximately five minutes. For analysis a  small aliquot of the supernate was
transferred to an autosampler  vial for subsequent analysis by GC/MS.  Aliquots for analysis
were diluted as necessary so the sample concentrations  fell within the calibration range.
7.2.2   GC/MS Analysis
       Samples of paints diluted in solvents were analyzed by direct injection (1  \iL) onto the
GC column.  Operating parameters for the GC/MS system are listed in Table 7-1. Target
analytes for quantitation  were identified based on an electronic database search of the
NIH/EPA/MSDC Mass Spectral Data Base and the Registry of Mass Spectral Data (7).
       During quantitative analyses, identification of target analytes was based on
chromatographic  retention times relative to the external standard and on relative
                                          7-2

-------
      TABLE 7-1.  OPERATING PARAMETERS FOR THE CAPILLARY GC/MS SYSTEM
 Parameter
                        Setting
 GAS CHROMATOGRAPH
 Instrument
 Column
 Temperature Program
 Carrier Gas Flow Rate
 Capillary injector
 Injector temperature
Hewlett-Packard 5890
30m x 0.32mm DB-624 widebore fused silica capillary column
35°C (5 min) to 250°C (5 min) @ 5°C/min
1.98 mL/min
1 min splitless
200°C
 MASS SPECTROMETER
'Instrument
 lonization Mode

 Emission Current
 Source Temperature
 Electron Multiplier
Hewlett Packard, Model 5988A
Electron lonization
Scan 25-350 m/z
0.3 mA
200°C
2000 volts3
 aTypical value.
                                         7-3

-------
abundances of the extracted ion fragments selected for quantitation. Fragment ions were
selected based on mass spectra acquired during qualitative analysis and historical data.
       Quantitation of VOCs/SVOCs was accomplished using chromatographic peak areas
derived from extracted ion profiles. Specifically, relative response factors (RRFs) for each
target compound were generated from the analysis of standard solutions (Tables 7-2 and 7-3)
prepared at five different concentrations. For each standard, RRFs were calculated as:
                                       AT • Cos(ng/uL)
                               RRFT  =   '    Wi>                                (7-2)
                                   T    AQS  • CT(ng/uL)

where AT is the peak area of the quantitation ion for the target VOC/SVOC and AQ§ is the
peak area of the external standard. CT is the concentration of target compound in the
calibration standard and Cgg is the concentration of the external standard in the calibration
standard sample.
       Because the calibration standards encompassed such a wide concentration range, the
instrumental response for many of the target analytes was not linear over the entire range.
For the alkyd paint standards both the response of the external quantitation standard and  the
target analytes decrease substantially at the higher concentrations.  Only the low level
standards (5 to 280 ng/uL) were used for quantitation and samples were diluted so that
concentrations did not exceed the concentrations of the highest quantitation standard. Where
instrumental linearity was not demonstrated, the standards used to quantitate a specific
analyte (shown in Table 7-2 and  7-3) were selected to bracket the concentration of that
analyte found in the diluted paint samples. Using these standards, mean values and
standard deviations of the RRFs  were calculated for each target analyte. For the calibration
to be considered acceptable, the mean RRF value had to be defined by at least three
calibration standards that bracketed the sample concentrations and the  RSD of the calculated
RRF had to be less than 30%. During each day of analysis, an additional standard was
analyzed.  If the RRF values for this standard were within ±25% of the RRFs for the same
concentration standard obtained  during calibration, the GC/MS system was considered "in
control" and the mean RRF values  from the calibration standards were  used to calculate the
concentration of the  target  VOCs in sample extracts (CEX) as:
                                          7-4

-------
   TABLE 7-2.  CALIBRATION SOLUTIONS FOR BULK PRODUCT ANALYSIS OF ALKYD PAINTS
Compound
ANALYTES
o-Xylene
m-Xylene

j>-Xylene
Propylcyclohexane
1,3,5-Trimethylbenzene
\
1,2,3-Trimethylbenzene
1 ,2,4-Trimethylbenzene
2-Methyldecane
n-Nonane
n-Decane
fniHS-Decahydronaphthalene
n-Undecane
Pentylcyclohexane
n-Dodecane
2-Ethyltoluene
3-Ethyltoluene
4-Ethyltoluene
1,1-Dimethylcyclohexane
Toluene
EXTERNAL QUANTITATION
Bromopentaflurobenzene
Concentration (ng/ul) in

*m:;m^;**, * a» „ '«,
, 4*9? " SEEP*, :?;,'/$w - ,m , - - ;:: 258
f -. fff f f *. S$f fff "• f f.
-w ^aw^ ;£»„" i«* ,_»
'>: v » ^ ;;^:
-;4^<^*^^1^'^?:' ^^ -T' - *H ..

4,92^'.;275rT,;to , '/102 ' - 252
4.89 27^ W 102 , 250
4^1 25.8 47,? 95,8 ' 236
4.49 25,1 46J 933 230
f 4^6 '25.f ' 475 94,? 234
5.00 ^28« ;/.52^ , 104 256^
j* jjjs "OC -D- ^ •• 4ff ^fc *QC "T 'J'X'jf
*. ST»*W^ -• 
-------
  TABLE 7-3. CALIBRATION SOLUTIONS FOR BULK PRODUCT ANALYSIS OF
                               LATEX PAINTS
Compound
        Concentration (ng/ul) in Acetone3
24.2
23.9
     rj
~Y. fSf& *.J:
24.0

24.0
24.0
ANALYTES
2-Methyl-2-propanoI
Ethylene glycol
1,2-Propanediol
p^-Xylene
m-Xylene
j>Xylene
n-Butyl ether (dibutyl ether)
Methyl sulfoxide
rt-Propylbenzene
1,2,3-Trimethylbenzene
1,2,4-Trimethylbenzene
1,3,5-Trimethylbenzene
2-(2-Methoxyethoxy)ethanol
2-Ethyltoluene
3-Ethyltoluene
4-Ethyltoluene
Diethylene glycol
Dipropylene glycol
2-{2-Butoxyethoxy)ethanol
2-(2-Butoxyethoxy)ethyl acetate 24-4
Dimethoxymethane
Texanol
INTERNAL STANDARD
n-Octanol
244
24.0
*'„„<,"                 "6 249°
50.0   99.9   251  <
492   985  248b   492
                     wv.
           ,,251   .4519   998 2495
           >;251   49f   997 2492
           r250   W,  995 2486
            24?   494   988 2471
      100   252   501  1003 ;2507
      9?*O   24?   495   990 2475
                            V  ••
      994   249   4?6   991 2478
      984   248   493   986 2464
      992   250   4?6   992 2481
50£  - 100   251  - 500 '1000 2500
      9fci   247   4?2   984 2459
      9?3   250   497   993 2483
49,4   98-9   249   494   989 2472
50.2   100   252   $02  1004 2509
      972   24?   486   972 2431


503   101   253   503  1007 2517
49.4   98#   248   494   988 2469
$5.0   110   277   550  1100 2749
4990
4984
4973
4942
5014
4950
4956
4929
4961
5000
4918
4967
4944
501*
4861
                                         5035
                                         4938
                                         5498
              1$$   1$$   19$  198   198  198
 aShaded standards used for instrumental calibration.
 bUsed for single point quantitation.
                                     7-6

-------
                            r  ,   ,  M
                            C£X(ng/uL) =        "     -                       (7-3)
                                                     "
The concentrations of the analytes in the paint sample were then adjusted for the dilution
factor and weight of paint diluted as:
                                   = CEX(ng/uL) • 10,000 uL
                              * *          W(g) -1000
where W is the weight (grams) of the paint.
       For alkyd paints, TVOCs were calculated from the reconstructed ion chromatogram
(Rip).  The total area of the RIC was integrated for the retention time window from n-hexane
through n-tetradecane.  The TVOC concentration was calculated based on the average total
ion response factor generated for toluene.  Since the only compounds in the latex paints were
target analytes, TVOCs  for these samples were calculated by summing the measured
concentrations of the individual targets.

7.3    RESULTS
7.3.1   Qualitative Identification
       Organic constituents identified during bulk analysis of alkyd and latex paints are
given in Tables 7-4 and 7-5 respectively. GC/MS total ion chromatograms for these analyses
are provided in Figures 7-1 to 7-4.  As seen in Figures 7-1 and 7-2, the alkyd paints contain
hundreds of compounds that are predominately branched chained hydrocarbons.  The
compounds identified in these paint samples represent the most abundant VOCs where
probable isomeric identification could be made. A visual comparison of the chromatograms
for the alkyd paints shows a higher relative abundance of the more volatile species for the
semigloss and  gloss finishes compared to the flat finishes.
       In contrast, the latex paints contain relatively few volatile organic chemicals.  The
VOCs identified during analysis generally account for all of the components in the GC/MS
chromatograms.  For the various paint samples tested, the presence and relative abundance
                                         7-7

-------
 TABLE 7-4. IDENTIFICATION OF MAJOR CHROMATOGRAPHIC PEAKS IN ALKYD PAINT SAMPLES
Paint Sample3
Gloss
Compound
Xytene isomers
Propylcyclohexane
3- or 4- Ethyltoluene
iv-Decane y
Trimethylbenzene isomers
2-Methyldecane
Decahydronaphthalene
n-Undecane
Pentylcyclohexane
iv-Dodecane
Sherman "
Williams
(Bumbershoot)
X
X

X

X
X
X

X
Glidden
(Hyacinth)
X
X
X
X
X
X
X

X
Semigloss
Sherman
Williams
(Violet
Veil)
X
X

X
X
X
X

X
Glidden
(Sea Foam)
X
X
X
X
X
X
X
X

X
Flat
Sherman
Williams
(Crescent
Cream)
X




X
X
X
X
X
Glidden
(Chim
Cham)
X


X

X
X
X
X
* Compounds indicated are those compounds that were at the highest abundance in
  each sample.
                                           7-8

-------
 TABLE 7-5.  IDENTIFICATION OF MAJOR CHROMATOGRAPHIC PEAKS IN LATEX PAINT SAMPLES
Compound
2-Methyl-2-propanol
Ethylene glycol
1,2-Propanediol (propylene glycol)
Xylene isomers
n-Butyl ether
Methyl sulfoxide
n-Propylbenzene
Trimethylbenzene isomers
2-(2-Methoxyethoxy)ethanol
Diethylene glycol
Dipropylene glycol
2-(2-Butoxyethoxy)ethanol
2-<2-Butoxyethoxy)ethyl acetate
Texanol

Gloss
Sherman
Williams Glidden
(Rose (Orange
Dawn) Glaze)

X X



X
X
X
X
X

X X

X X
Paint Sample3
Semigloss
Sherwin
Williams Glidden
(Vibrant (Down
Violet) Yonder)
X
X X
X X
X
X



X

X X
X
X X

Flat
Sherwin
Williams Glidden
(Marmalade) (Tomahawk)

X X
X





X
X
X X
X
X
"Major compounds identified in each sample are identified with an X.
                                        7-9

-------
              Rat:
              (Crescent Cream)
                                 X
                                 _&.,
                                   6|
              Semi-Gloss:
              (Violet Veil)
                         JL
             Gloss:
             (Bumbershoot)
         0
                           12     16     20     24
                                       Time (min)
28
32
36
40
Figure 7-1.    Total ion chromatograms of Sherwin Williams Alkyl Paint Bulk Samples
             (representative compounds indicated on chromatogram).
                                     7-10

-------
              Flat:
              (Chim Cham)
             Semi-Gloss:
             (Sea Foam)
             Gloss:
             (Hyacinth)
         0
^_A
            4      8      12     16     20     24     28     32     36     40
                                    Time (min)
Figure 7-2.   Total ion chromatograms of Glidden Alkyl Paint Bulk Samples (representative
            compounds indicated on chromatogram).
                                       7-11

-------
     Flat: (Tomahawk)
               ,v
8
>,
•eb
                          *
                                         X
                                        w

                                   o
                                   =2
     Semi-Gloss: (Down Yonder)




                      ^

                       "o ^
                       "§  *


                     "*(§"'
                                      n
                                                         3
                                                        CO

                                                        d
                                                        «N
     Gloss: (Orange Glaze)
                               >.
                               g

                               S,
                               X
                              Ci
                    Ol
                    I
                    ">,
0
                                         (2
                                          JL

                  12
     16
  20     24
Time (min)
28
32
36
40
       Figure 7-3. Total ion chromatograms of Glidden Latex Paint Bulk Samples.
                                       7-12

-------
Flat: (Marmalade)


"o
~5 §
2 TJ i 2
'•8 J f H
I ill
nl II !
"ab £ >• w to X
~C( If
LU
(5
1
Sv
1




Semi-Gloss: (Vibrant Violet) y. i
•2 "2 1
I ^5
1 8 1
•3 ^ :
"g^l c I
>> S W -5. t
60 C ?
g S. f
jj £ I S
I*1 -S I
Gloss: (Rose Dawn)

2-(2-Butoxyethoxy) et

"o
I "O T3 ^1
C C ;.,
1 1 If
C ^
i
>,
tu
1
1 Jl ^ A
A

1
S


\h


1
i

I


     0
               8
12
16
 20     24
Time (min)
28
32
36
40
Figure 7-4. Total ion chromatograms of Sherwin Williams Latex Paint Bulk Samples.
                                   '7-13

-------
of individual VOCs varied between samples with no clear trend for manufacturer or gloss
type.
7.3.2   Quantitative Analyses
       Results of quantitative analysis for the alkyd paint samples are given in Table 7-6.
Similar results are provided for the latex paints in Table 7-7. For the gloss and semigloss
alkyd paints, n-nonane, n-decane, n-undecane, ivdodecane, and propylcyclohexane were
among the most abundant compounds. For the flat alkyd paint, the most abundant
compounds are n_-undecane, ivdodecane, and n_-decane.  For four of the six latex paints,
ethylene glycol was the most abundant compound.  1,2-Propanediol and 2-(2-methoxy-
ethoxy)ethanol had the highest concentrations in the other two paint samples.  Other
compounds with relatively  high concentrations (>10 mg/g) included 2-(2-butoxy-
ethoxy)ethanol and Texanol.
       Duplicate sample aliquots were analyzed for selected paint samples as a way to assess
uniformity of the sample aliquots as well as overall method precision.  The %RSD values for
these duplicate samples are given in Tables 7-8 and 7-9 for alkyd and latex paint samples,
respectively.  Results show low %RSD values (generally less than 10%) for both paint types.
These data suggest that both the replicate sample aliquots are uniform and the precision of
the overall method is good.
       Method controls were prepared and analyzed as a further evaluation of method
performance. Method controls were dilution solvent spiked with target chemical and
external quantitation standards which were then handled and analyzed  in a manner identical
to paint samples. Nominal spiking levels were equivalent to 1 mg/g of paint. Results for
these analyses (Table 7-10) for the alkyd paints show good recovery  (82 to 99%) and precision
for the paint  samples extracted prior to storage.  Recovery of method controls  prepared and
analyzed for  the storage stability study (T=29 weeks) were uniformly low (60 to 62%).  This
could be due to a systematic error (i.e., external quantitation standard spiked to high).
However, no reason for this result could be documented. Precision evaluated  as the
standard deviation of replicate analysis was very good (generally less than 5%) for both sets
of method controls. For the latex paint controls (Table 7-11), the recoveries for the initial
study were reasonable although somewhat high.  1,2-Propanediol gave a low recovery  (45%).
However this compound was spiked at a level below the method quantitation limit. The reported
recovery is probably a result of a lower GC/MS response for lower concentrations of this

                                            7-14

-------
          TABLE 7-6. QUANTITATIVE RESULTS FOR TARGET VOCS IN ALKYD PAINT SAMPLES
Concentration
Gloss
Sherwin
Williams" Glidden
Compound (Bumbershoot) (Hyacinth)
mj>Xylene
n-Nonane
2-Xylene
Propylcydohexane
3- & 4-Ethyltoluene
1,3,5-Trimethylbenzene
n-Decane
2-Ethyltoluene
1,2,4-Trimethylbenzene
1,23-Trimethylbenzene
2-Methyldecane
frans-Decahydronaphthalene
n-Undecane
Pentylcyclohexane
ii-Dodecane
TVOC
4.29
11.2
0.828
3.40
NDb
ND
19.3
ND
0.533
ND
3.24
3.46
16.2
2.16
7.88
274
3.58
9.82
1.06
4.19
1.82
0.788
18.0
0.617
2.76
0.839
1.98
2.10
9.11
0.518
2.60
284
(mg/g)
Semigloss
Sherwin
Williams
(Violet
Veil)
5.14
9.76
0.781
3.24
0.694
0.322
15.9
ND
0.854
ND
2.49
2.54
14.0
1.54
6.89
314
Glidden8
(Sea
Foam)
3.97
5.32
1.08
1.45
1.30
0.520
14.1
0.483
1.81
0.635
2.63
2.58
16.2
2.06
8.10
207



Hat
Sherwin
Williams
(Crescent
Cream
2.78
0.164
0.565
ND
ND
ND
1.33
ND
ND
ND
2.97
3.49
183
3.73
11.0
202
Glidden
(Chim
Cham)
1.56
0.675
0.367
0.196
ND
ND
4.51
ND
ND
ND
3.24
3.24
19.1
3.01
11.5
179
MQLC
(mg/g)
0.558
0.279
0.280
0.278
0.556
0.278
0.256
0.273
0.274
0.275
0.258
0.280
0.259
0.270
0.263
—
a Mean of Duplicate Analysis
b Not detected, below MQL
c Method Quantitation Limit estimated as the paint concentration that would be equivalent to lowest concentration
  calibration standard and 1 gram of paint.
                                              7-15

-------
          TABLE 7-7. QUANTITATIVE RESULTS FOR TARGET SVOCS IN LATEX PAINT SAMPLES
Concentration (mg/g)
Gloss
Sherman
Williams
(Rose
Compound Dawn)
2-Methyl-2-propanol
Ethylene glycol
1,2-Propanediol
mj>-Xylene
iv-Butyl ether .
o-Xylene
Methyl sulfoxide
n-Propyl benzene
1,3,5-Trimethylbenzene
2-(2-Methoxyethoxy)ethanol
1 ,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
Diethylene glycol
Dipropylene glycol
2-{2-Butoxyethoxy)ethanol
2-<2-Butoxyethoxy)ethyl acetate
Texanol
TVOC
0.36
48
NDb
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.62
ND
13.0
ND
2.7
64.7
Glidden
(Orange
Glaze)
ND
5.61
ND
ND
ND
ND
0.889
ND
ND
22.9
0.569
ND
ND
0.993
13.0
ND
10.5e
545
Semigloss
Sherman
Williams
(Vibrant
Violet)
0.307
40.1
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
6.50
0.392
5.35
ND
4.22
56.9
Glidden3
(Down
Yonder)
ND
19.1
38.3
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
1.23
4.42
0.498
5.72
69.3
Hat
Sherman
Williams3
(Marmalade)
ND
29.0
1.64d
ND
ND
ND
ND
ND
ND
ND
ND
ND
5.14
0.347
1.54
ND
5.13
41.2
Glidden MQLC
(Tomahawk) (mg/g)
ND
20.8
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
3.40
4.43
ND
28.6
0.241
5.00
10.0
0.483
0.240
0.242
0.243
0.240
0.241
0.242
0.240
0.239
0.243
0.236
0.241
0.244
0.267
—
a Mean of Duplicate Aliquots.
b Not detected: below MQL.
c Method Quantitation Limit estimated as the paint concentration
  concentration calibration standard and 1 gram of paint.
d Curve was not linear at lower concentrations, quantitated using
e Greater than 10% above the highest concentration standard.
that would be equivalent to the lowest

a single point calibration.
                                                 7-16

-------
  TABLE 7-8. PRECISION OF BULK ANALYSIS METHOD FOR ALKYD PAINTS
Compounds
m,j>-Xylene
n-Nonane
p^Xylene
Propylcyclohexane
3- & 4-Ethyl toluene
1 ,3,5-Trimethy Ibenzene
iv-Decane
2-Ethyl toluene
1 ,2,4-Trimethy Ibenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
iv-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
%
Semigloss
Glidden
(Sea Foam)
5.3
2.7
4.0
0.093
5.2
1.7
2.2
4.6
0.36
3.0
0.73
1.2
1.2
0.56
0.18
12
> RSD of Duplicates3
Gloss
Sherwin Williams
(Bumbershoot)
5.2
1.2
1.7
0.62
NC
NC
1.8
NC
0.36
NC
2.9
4.4
1.4
6.4
4.7
6.7
a% Relative Standard Deviation.
''Not calculated, not found in sample above the MQL.
                                 7-17

-------
       TABLE 7-9. PRECISION OF BULK ANALYSIS METHOD FOR LATEX PAINT
Compound
2-Methyl-2-propanol
Ethylene glycol
1,2-Propanediol
m,p-Xylene
it-Butyl ether
^-Xylene
Methyl sulfoxide
n.-Propyl benzene
1,3,5-Trimethylbenzene
2-(2-Methoxyethoxy)ethanol
1,2,4-Trimethylbenzene
1 ,2,3-Trimethylbenzene
Diethylene glycol
Dipropylene glycol
2-(2-Butoxyethoxy)ethanol
2-(2-Butoxyethoxy)ethyl acetate
Texanol
TVOC
%RSDof
Semigloss
Glidden
(Down Yonder)
NCb
0.52
1.4
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
1.6
0.19
.0.28
0.93
0.60
Duplicates3
Flat
Sherwin Williams
(Marmalade)
NC
7.7
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
4.0
6.1
2.7
NC
1.2
6.0
a% Relative Standard Deviation.
bNot calculated, not found in sample above the MQL.
                                     7-18

-------
TABLE 7-10. RESULTS OF METHOD CONTROLS FOR BULK PRODUCT ANALYSIS
                        OF ALKYD PAINT
Compounds
m,j>-Xylene
n-Nonane
o-Xylene
Propylcyclohexane
3- & 4-Ethyl toluene-.
1 ,3,5-Trimethy Ibenzene
H-Decane
X
2-Ethyl toluene
1 ,2,4-Trimethy Ibenzene
1,23-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
ivUndecane
Pentylcyclohexane
n-Dodecane
Spike Level
(mg/g)
2.07
0.93
1.04
1.03
2.06
1.03
0.95
1.02
1.02
1.02
0.96
1.04
0.96
1.00
0.98
% Recovery
Initial Study
(n=2)
87 ± 2.1
93 ± 2.5
82 ± 0.54
92 ± 1.9
95 ± 1.2
92 ± 3.6
96 ± 3.0
93 ± 2.4
96 ± 1.7
96 ±0.21
102 ± 3.4
98 ± 1.0
97 ± 1.5
99 ±4.6
95 ± 3.0
±S.D.
Storage Study
(n=2)
64 ± 0.57
60 ± 0.11
61 ± 0.78
60 ± 0.42
62 ± 0.27
60 ± 0.51
60 ± 0.33
61 ± 0.10
63 ± 0.18
61 ± 0.26
63 ± 0.12
64 ±1.3
61 ± 0.60
60 ± 0.010
62 ± 1.0
                             7-19

-------
   TABLE 7-11. RESULTS OF METHOD CONTROLS FOR BULK PRODUCT
                     ANALYSIS OF LATEX PAINT
Compound
2-Methyl-2-propanol
Ethylene glycol
1 ,2-Propanediol
m,j>-Xyleneb
n-Butyl ethei* >,
o-Xyleneb
Methyl sulfoxide
Propyl benzeneb
l,3,5-TrimethylDenzeneb
2-(2-Methoxyethoxy)ethanol
1,2,4-Trimethylbenzene
l/2/3-Trimethylbenzeneb
Diethylene glycol
Dipropylene glycol
2-(2-Butoxyethoxy)ethanol
2-(2-Butoxyethoxy)ethyl acetate
Texanol
Spike Level
(mg/g)
1.00
l.OO3
0.99"
ZOO
0.99
1.00
1.00
0.99
0.99
1.00
0.99
0.99
1.00
0.97
1.00
1.01
1.10
% Recovery
Initial Study
(n = 2)
116 ± 3.6
89 ± 3.5
45 ±3.5
141 ± 2.8
.122 ± 2.9
128 ± 4.2
116. ±0.7
132 ± 4.3
136 ± 2.8
124 ±0
133 ± 1.4
130 ± 0.78
112 ± 3.5
117 ± 3.6
163 ± 3.6
128 ± 0.7
150 ± 2.6
±S.D.
Storage Study
(n = 2)
146 ±8
77 ±5.5
88 ± 4.5
171 ± 12
153 ± 15
160 ± 10
119 ± 11
160 ± 13
166 ± 11
128 ± 8.0
160 ± 8.5
156 ± 4.0
96 ± 6.5
106 ± 7.0
143 ± 13
123 ±10
146 ±10
Spike level below the method quantifiable limit.
                               7-20

-------
polar compound.  Precision was good (RSD values <10%). Recovery of the aromatic targets
(i.e., xylenes and trimethyl benzenes) were very high for controls prepared and analyzed for
the storage stability study. Although these recoveries were high, these compounds were not
measured at detectable levels during bulk product analysis.
7.3.3  Storage Stability
      In order to evaluate the stability of the paint samples over prolonged storage periods,
bulk product analysis of paint samples was performed at the time the first small chamber
emissions tests were performed (T=7 weeks after aliquoting) and at the time the last small
chamber emissions tests  were performed (T=29 weeks after aliquoting). As discussed in
Section 5 paint samples were stored in screw cap glass vials at room temperature in  the dark.
      A review of the results for the latex paints indicated the external quantitation
standard was compromised in the highest calibration standards used for the bulk product
analyses for both time periods. If these high standards were removed from the calibration
curve, many of the target VOCs exceeded the upper calibration limit. As a result, all alkyd
sample extracts prepared for the bulk product analyses were diluted and reanalyzed. For the
alkyd paints, the results  from the reanalyzed extracts are presented in this  report.  Results of
the analyses are given for the alkyd paints in Table  7-12 and the latex paint in Table 7-13.
The results for method controls associated with these paint samples are given in Tables 7-10
and 7-11.  In each table,  measured concentrations for target analytes and TVOC are given for
each time period.  The ratio of measured concentrations for paint samples stored for 29
weeks and 7 weeks of storage is also given.  If there were no sample losses during storage,
the ratio for these two measured concentrations should be 1.0.  Ratios less  than 1.0 indicate
losses during storage. Results for the alkyd paints show a uniform ratio of -0.70 suggesting
approximately 30% loss  of volatile constituents during storage. Although this is likely, the
magnitude of the losses are similar to those seen for method controls prepared and analyzed
at the same time (Table 7-10). Consequently, there could have been a uniform analytical bias
during the analyses of the paint samples after storage.  Unfortunately, it was not possible to
determine if this was the case or if decreases in measured paint concentrations over time
were due to volatility losses  during storage.  These are the same trends that were seen for the
analyses performed at the time the samples were prepared.
      For the latex paint samples, the ratio of measured analyte concentrations at 29 and 7
weeks of storage ranged from 0.50 to 1.30.  Lowest ratios are seen for the most polar

                                         7-21

-------
                      TABLE 7-12. RESULTS OF THE STORAGE STABILITY TESTS FOR ALKYD PAINTS
I
ro
ro
Glidden/Gloss
(Hyacinth)
Compounds
m,j>-Xylene
rv-Nonane
o-Xylene
Propylcyclohexane
3- & 4-Ethyl toluene
1 ,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1 ,2,4-Trimethylbenzene
1 ,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
n-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
T=7 Weeks
(mg/g)
3.58
9.82
1.06
4.19
1.82
0.788
18.0
0.617
2.76
0.839
1.98
2.10
9.11
0.518
2.60
284
T=29 Weeks
(mg/g)
2.66
7.66
0.817
3.03
1.38
0.623
14.0
0.474
2.08
0.640
1.48
1.56
6.96
0.483
1.96
274
Ratio3
0.74
0.78
0.77
0.72
0.76
0.79
0.78
0.77
0.75
0.76
0.75
0.74
0.76
0.93
0.75
0.96
Sherwin Williams/Semi-gloss
(Violet Veil)
T=7 Weeks
(mg/g)
5.14
9.76
0.781
3.24
0.694
0.322
15.9
ND
0.854
ND
2.49
2.58
14.0
1.54
6.90
314
T=29 Weeks
(mg/g)
3.64
6.54
0.554
2.23
ND*
0.209
11.4
ND
0.608
ND
1.68
1.85
10.4
1.17
4.65
276
Ratio
0.71
0.67
0.71
0.69
—
0.65
0.72
—
0.71
—
0.68
0.72
0.74
0.76
0.67
0.88
       3 Ratio of measured concentrations at T=29 weeks to T=7 weeks.

       b Below the method quantitation limit.

-------
                          TABLE 7-13. RESULTS OF THE STORAGE STABILITY TESTS FOR LATEX PAINTS
ro
co
Glidden Semigloss
(Down Yonder)
Compounds
2-Methyl-2-propanol
Ethylene glycol
1,2-Propanediol
m,r>Xylene
n-Butyl ether
o-Xylene
Methyl sulfoxide
Propyl benzene
1 ,3,5-Trimethy Ibenzene
2-(2-Methoxyethoxy)ethanol
1,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
Diethylene glycol
Dipropylene glycol
2-(2-Butoxyethoxy)ethanol
2-(2-Butoxyethoxy)ethyl acetate
Texanol
TVOC
T=7 Weeks
(mg/g)
NDb
19.1
38.3
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
1.23
4.42
0.498
5.72
69.3
T=29 Weeks
(mg/g)
ND
14.3
31.0
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.946
4.94
0.247
7.39
58.8
Ratio3
NCC
0.75
0.81
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
0.77
1.11
0.50
1.30
0.85
Sherwin Williams Flat
(Marmalade)
T=7 Weeks
(mg/g)
ND
29.0
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
5.14
0.347
1.54
ND
5.13
41.2
T=29 Weeks
(mg/g)
ND
26.8
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
3.44
0.285
1.42
ND
5.76
37.7
Ratio
NC
0.92
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
0.67
0.82
0.90
NC
1.12
0.92
       aRatio of measured concentration at T=29 weeks to T=7 weeks.
       ''Below the method quantifiable limit.
       cNot calculated.

-------
compounds (diethylene glycol, dipropylene glycol, and ethylene glycol) or compounds that
have relatively low concentrations (2-(2-butoxyethoxy)ethyl acetate). These compounds
would be expected to give the poorest analytical performance.  Thus although there appear
to be some losses during storage, a low ratio could also reflect analytical performance.
                                         7-24

-------
                                   SECTION 8.0
                        SMALL CHAMBER EMISSIONS TESTS

8.1    OVERVIEW AND STUDY DESIGN
      Small chamber tests are intended to measure emissions from paint samples over time
for individual VOCs/SVOCs and TVOC using carefully controlled conditions. During
testing, a paint sample is applied to a glass plate which is immediately placed in a 52.7 L
stainless steel chamber.  The chamber is sealed and air is passed through the chamber at a
rate of one air change per hour (ACH).  Air samples are collected from the chamber outlet at
specified time points for measuring VOC/SVOC and aldehyde emissions.
      During the previous work, VOCs/SVOCs in chamber air samples were collected on
Tenax GC cartridges with subsequent thermal desorption and GC/MS analysis (1).
Unfortunately, problems were encountered when applying this method to emissions testing
for both alkyd and latex paints.  For alkyd paints, high concentrations of VOCs in air
samples from the small chambers caused several problems.  For  example,
      •      The mass of VOCs analyzed from the sorbent cartridges exceeded the linear
             dynamic range of GC/MS analysis, unless very small sample volumes were
             collected.
      •      The collection of small sample volumes resulted in poor precision.
      •      The preparation of vapor phase standards at very high VOC concentrations
             was difficult.
For the latex paints, the polar SVOCs such as Texanol and 2-(2-butoxyethoxy)ethanol were
not recovered from large Tenax GC cartridges during thermal desorption.
      To overcome these problems, a first step on this work assignment was to modify and
evaluate sampling and analysis methods for measuring VOC/SVOC emissions from paint
samples.  Activated charcoal-based sampling tubes with solvent  extraction and GC/MS
analysis of sample extracts was evaluated for estimating VOC emissions from alkyd paint
samples.  The charcoal method was considered advantageous in that large sample volumes
(e.g., 5 to 20 L) can be collected even when the concentrations of the VOCs are high.  Solvent
extracts can then be diluted as required to obtain analyte concentrations  within the linear
range of the GC/MS system. Sample collection on small Tenax TA sample cartridges
followed by thermal desorption and analysis by GC/MS or gas chromatography/flame

                                        8-1

-------
ionization detection (GC/FID) was proposed for measuring SVOC emissions from latex paint
samples. This method has been reported previously for the sampling and analysis of polar
SVOCs, although performance results had not been reported (6 and Appendix A). It was felt
that more efficient desorption of SVOCs might be achieved compared to the Tenax GC
method used previously, since higher desorption temperatures and smaller diameter tubes
are used.  For both of the proposed sampling and analysis methods, recovery tests were
performed to determine the precision and accuracy of the method.  Tests were also
performed to determine the dynamic range of the methods.
       After acceptable performance was demonstrated for the proposed sampling and
analysis methods, a series of small chamber tests were performed to evaluate overall method
performance for the determination of VOC/SVOC and aldehyde emissions from latex and
alkyd paints. Table 8-1 outlines the tests that were performed.  As shown in the table, Tests
1 and 2 were range finding tests for the latex and alkyd paints. These range rinding tests
were performed to determine the appropriate air sample collection volumes, sample
collection time points, and test durations for the two type of paints. Tests 3 to 6 were single
chamber repeatability tests designed to determine method precision for the same  test
chamber for both an alkyd and a latex paint. Once acceptable precision was demonstrated,
Tests 7 to 10 were performed to determine the recovery of target VOCs/SVOCs from the
chambers. Single chamber repeatability and interchamber variability tests were then
performed (Tests 11 to 18).  These eight additional chamber tests were performed to obtain
data on the performance ;of the method for all six types of paints, with the variability being
determined for tests performed in the same chamber for three paints and in different
chambers for the other three paints.  Finally, a  set of four chamber tests (Tests 19 to 22) were
performed to evaluate the effect of air velocity on emission rates. Results were compared
with and without the fan for both an alkyd and a latex paint.  VOC/SVOC emissions were
measured during all chamber tests. Aldehyde  emissions were measured based on
preliminary screening tests.

8.2    METHODS
8.2.1   Application of Paint Sample
       In order to properly evaluate the small  chambers, a reproducible method for applying
the wet paint to a glass panel was necessary. In the previous study (1), a brush was used

                                         8-2

-------
           TABLE 8-1.  PERFORMED CHAMBER TESTS FOR QUANTIFICATION OF VOC/SVOC EMISSIONS
Test
Number
1

2

3

4

5

6

7
8
9
10
11

12

13

14

15

16

17

18

19

20

21

22

Description of Test
Range finding

Range finding

Single chamber repeatability

Single chamber repeatability-Duplicate

Single chamber repeatability

Single chamber repeatability-Duplicate

Analyte Recovery Tests
Analyte Recovery Tests-Duplicate
Analyte Recovery Tests
Analyte Recovery Tests-Duplicate
Inter-chamber variability

Inter-chamber variability-Duplicate

Inter-chamber variability

Inter-chamber variability-Duplicate

Inter-chamber variability

Inter-chamber variability-Duplicate

Single chamber repeatability

Single chamber repeatability-Duplicate

Effect of air velocity (w/fan)

Effect of air velocity (w/o fan)

Effect of air velocity (w/fan)

Effect of air velocity (w/o fan)

Paint Type
Latex (vinyl)

, Alkyd

Latex (vinyl)

Latex (vinyl)

Alkyd

Alkyd

VOCsb
VOCs
SVOCsd
SVOCs
Alkyd

Alkyd

Alkyd

Alkyd

Latex

Latex

Latex (vinyl)

Latex

Latex

Latex

Alkyd

Alkyd

Gloss Type
Flat
(Marmalade)
Gloss
(Hyacinth)
Flat
(Marmalade)
Flat
(Marmalade)
Gloss
(Hyacinth)
Gloss
(Hyacinth)
NAC
NA
NA
NA
Flat
(Chim Cham)
Flat
(Chim Cham)
Semigloss
(Sea Foam)
Semigloss
(Sea Foam)
Gloss
(Rose Dawn)
Gloss
(Rose Dawn)
Semigloss
(Down Yonder)
Semigloss
(Down Yonder)
Flat
(Marmalade)
Flat
(Marmalade)
Gloss
(Bumbershoot)
Gloss
(Bumbershoot
Test Duration'
(Days)
5

3

7

7

1

1

2
2
2
2
1

1

1

1

7

7

7

7

7

7

1

1

Sampling Tune*
Points (Hrs.)
8,24,72,96,120

4,8,12,24,72

1,12,24,48,96,120,168

1,12,24,48,96,120,168

0.5,1,2,3,4,8,12,24

0.5,1,2,3,4,8,12,24

0,1,2,4,8,12,24
0,1,2,3,4
0,l,-2,-3,4,6
0,1,2,3,4,6
0.5,1,2,3,4,8,12,24

0.5,1,2,3,4,8,12,24

0.5,1,2,3,4,8,12,24

0.5,1,2,3,4,8,12,24

1,12,24,48,96,120,168

1,12,24,48,96,120,168

1,12,24.48,96,120,168

1,12,24,48,96,120.168

1,12,24,48,96,120,168

1,12,24,48,96,120,168

0.5,1,2,3,4,8,12,24

0.5,1,2,3,4,8,12.24

* Test duration and sampling times determined in range-finding tests.
b Alkyd paint target analytes.
c Not applicable.
d Latex paint target analytes and aldehydes.
                                                      8-3

-------
but was considered unacceptable since it did not apply a uniform thickness and consistent
amount of paint.
       For this study a "drawdown" method utilizing an 11 inch adjustable Microm Film
Applicator (Paul M. Gardner Co., Pompano Beach, FL) was used. A standard pane of
window glass (12 in. x 14 in.) that had been cleaned, dried and weighed served as the
substrate for application. A flat surface with a straight edge secured on one side served as a
guide for the applicator and a support for the test glass  panel. It is important that this
support/surface be as rigid and flat as possible since differences of only several thousandths
of an inch caused problems with uniform coatings of the paint samples. The glass plate was
secured to the support surface with masking tape to prevent movement during coating. The
tape was placed as single pieces stretching across the top and bottom ends of the glass plate.
The tape was carefully removed where the runners  of the applicator contacted the glass plate.
In addition to securing the glass plate to the support surface, the tape served as a resist to
the paint. The combination of the tape and the width of the applicator provided an area on
the glass 279 x 283 mm for coating with paint. This area allowed a loading of 1.5 m2/m3
(paint area to chamber volume).
       During application, the paint vials were weighed, then vigorously shaken both
manually and with a vortex mixer for approximately 2 minutes before opening. The
applicator gate opening was adjusted to 7 mil and placed at the top of the  prepared glass
pane. A pool of paint was poured onto the top strip of masking tape between the sides of
the applicator. With a steady motion, the applicator was pulled through the paint using the
straight edge as a guide. When the applicator and excess paint had cleared the glass and
were on the bottom strip of tape, the applicator was removed and the strips of masking tape
carefully removed.  The coated glass was then weighed  to determine the wet paint mass.
The approximate thickness of the wet paint was 4 mil.
8.2.2   Test Chambers
       The small chamber test system used on this work assignment consisted of two
electropolished stainless steel chambers housed in a temperature controlled incubator.  The
chambers have a volume of 52.7 L and are of an identical design to those used at the Air and
Energy Engineering Laboratory at the US. EPA facility in Research Triangle Park, NC.
Nominal dimensions of the chambers are 51 cm (width) by 25 cm (height)  by 41 cm (depth).
A stainless steel plate, fitted with an O-ring, is used to seal the one open side. The chambers

                                          8-4

-------
are fitted with inlet and outlet manifolds designed to ensure adequate mixing in the
chamber. Air supplied to the chamber is first passed through a series of filtration devices to
minimize background contamination.  Air flow rates are controlled by mass flow
controllers/meters. A water vapor generator is used to control the relative humidity of the
air stream.  A diagram of this system is given in Figure 8-1. Performance of the chambers
has been previously validated [81. Operating conditions for tests conducted during this work
assignment were:
       •     23 ± 1°C temperature,
       •     50 ± 5% relative humidity (input air),
       •     1.0 ± 0.05 air exchange per hour,
Airflow rates, temperature, and relative humidity were monitored continuously during each
test; average hourly values were recorded.
   _y                                                   "
       During tests 19 to 22 (Table 8-1) small chamber tests were performed with both alkyd
and latex paints to determine the effect of air velocity on emission rates. A small fan was
installed in one of the small chambers for comparison against another small chamber without
a fan. The fan size and configuration was based on the design currently being used by Dr.
Bruce Tichenor, Air and Energy Engineering Research  Laboratory (AEERL), USEPA, Research
Triangle Park, NC. A 1 9/16 inch diameter 12 VDC brushless micro fan (Model No. 273-
244A, 3.5 cfm airflow, Radio Shack Division of Tandy Corp., Fort Worth, TX 76102), was
suspended by springs between the inlet and  exhaust manifolds inside the chamber. The
springs were simply hooked in the holes on both ends of the manifold pipes and the
mounting holes in the fan. This  placed the fan approximately in the center of the chamber 5
inches above the paint sample. The fan  was oriented such that the airflow was in the
direction of the top of the chamber.  Power to the fan was supplied by a variable DC power
supply. Power leads to the fan passed through a Teflon faced silicone septum in an unused
chamber port. Based on multi-point air  velocity measurements inside the chamber  using a
constant temperature anemometer (CTA), an operating voltage of 9.00 VDC was determined
to provide a suitable fan speed.  This generated air velocities across the sample plate ranging
from 5.9 to 16.2 cm/s depending on the point of measurement. The highest  velocities were
observed in a diagonal line from the left rear to the right front of the test chamber.  The
average velocity of the test points was 11.2 cm/s.  The velocities measured in the chamber
                                         8-5

-------
                  Dreyfus House Air
               Coarse Filter / Regulator
                 |  Balston Dx Filter
                  Balston Ex Filter
Purekol
i
r
                       Purafil
)
r
Balston Membrane Filter
                        I
     I

Precision Regulator
i
r

                                                 Sampling Ports
Figure 8-1. Diagram of the Chamber Emissions Test System.




                             8-6

-------
without the fan were below the detection limit of 2 cm/s.  A complete description of the air
velocity measurements can be found in the report prepared by Peters and Rodes [9].
8.2.3  Sampling and Analysis Methods
      8.2.3.1 VOCs from Alkvd Paints
      For chamber tests with alkyd paints, VOCs in air samples were collected by passing
air from the chamber outlet through sampling tubes (7 cm x 6 mm o.d.) containing two
sections of activated coconut shell charcoal (charcoal tubes, No. 226-01 GWS, SKC, Inc.,
Eighty-four, PA). Sample flow rates ranged from approximately 200 to 265 mL/min.
Sampling times varied from 10 to 75 minutes to give nominal sample volumes ranging from
2 to 20 L.
      VOCs were extracted from the sorbent material by combining then extracting both
portions of the sample tube charcoal beds with 2 mL of carbon disulfide. Samples were
spiked with the external quantitation standard, Orxylene-d10/ during extraction.  Aliquots
(1 uL) of the sample extracts were immediately analyzed by GC/MS using the conditions
given in Table 7-1.  Instrument calibration and quantitation of VOCs in sample extracts was
performed using relative response factor (RRFs) as described in Section 7.2.2.  Calibration
standards of target analytes in carbon disulfide were prepared  at six levels ranging in
concentration from 0.5 ng/pL to 500 ng/uL for each of the target VOCs.
      Concentrations of target VOCs in sample extracts (CEX)  were converted to chamber air
concentrations (C) as
                                       ,  C£v(ng/uL) • VWuL)
                      CCA(ng/L or ug/m3)= EX    *     EX*                   (8-1)
where VEX is the volume of the extract and VCA is the volume of the collected air sample.
      8.2.3.2 SVOCs from Latex Paints
      For chamber tests with latex paints, SVOCs in air samples were collected by passing
air from the chamber outlet through Tenax TA cartridges (200 mm x 6 mm o.d., Envirochem,
Kimblesville, PA).  Sample flow rates ranged from approximately 30 to 45 mL/min.
Sampling times varied from 20 to 80 minutes to give nominal sample volumes ranging from
0.5 to 5 L.
      Exposed cartridges were analyzed by thermal desorption followed by GC/MS or
GC/FID using the conditions shown in Table 8-2 and 8-3.
                                         8-7

-------
 TABLE 8-2.  GC/MS OPERATING CONDITIONS FOR ANALYSIS OF SVOC EMISSIONS
                         FROM LATEX PAINT SAMPLES
Parameter
Setting
THERMAL DESORPTION

Thermal desorption temperature
Valve and fitting temperature
Cryo trap temperature
      - minimum
      - maximum
Purge Flow Rate

GAS CHROMATOGRAPH

Instrument
Column
Temperature Program
Carrier gas flow rate

MASS SPECTROMETER

Instrument
lonization Mode
Emission Current
Source Temperature
Electron Multiplier
275°C (max 320°C)
220°C

-190°C
255°C
69 mL/min
Hewlett-Packard 5890
DB624-30M widebore fused silica capillary column
35°C(5 min) to 250°C(2 min) at 5°C/min
2.3 mL/min
Hewlett Packard, Model 5988A
Electron lonization Scan 25-350 m/z
0.3mA
200°C
2000 voltsa
^Typical value
                                     8-8

-------
   TABLE 8-3. GC/FID OPERATING CONDITIONS FOR ANALYSIS OF
	SVOC EMISSIONS FROM LATEX PAINT SAMPLES	

 Parameter                        Setting

 THERMAL DESORPTION

 Thermal desorption temperature      275°C
 Valve and fitting temperature        220°C
 Cryo trap temperature
       -minimum                  -200°C
       -maximum                  255°C
 Air flow rate                      300 mL/min
 Hydrogen flow rate                50 mL/min
 Makeup flow rate                  30 mL/min
 Purge flow rate                    30 mL/min
 GAS CHROMATOGRAPH

 Instrument                       Varian 3700
 Column                          DBWAX - 30 m x 0.32 mm
 Temperature Program              35°C (5 min) to 185°C at 5°C/min
 Carrier gas flow rate               2 mL/min
 Detector                          Flame ionization
                              8-9

-------
       During GC/MS analysis, identification of target analytes was based on
chromatographic retention times relative to standards and the relative abundances of
extracted ion fragments selected for quantitation.  Quantitation was accomplished using
chromatographic peak areas derived from extracted ion profiles. Calibration standards
containing the target analytes were prepared on Tenax TA cartridges at masses ranging from
20 to 5000 ng /cartridge. Each calibration standard and sample contained a known mass of
the quantitation standard, bromopentafluorobenzene.
       Relative response factors (RRF) were calculated as

                                        =         £                             (8-2)
where MT is the mass of target analyte (ng/cartridge), MQS *s ^e mass °f quantitation
standard (ng/cartridge), AT is the peak area of the target analyte, and MQS is the mass of the
quantitation standard (ng/cartridge).
       Because the instrumental response was not linear over the entire calibration range for
many of the SVOC targets, the standards used for quantitation were determined by the
amount of target analyte found on the samples. In most cases, the analyte amount in the
samples was greater than the highest amount in the calibration standard. Where this
occurred, the RRF from the highest standard was used for quantitation.  This approach was
considered acceptable, since GC/MS analysis was only performed during range finding tests.
During  each day of analysis, an additional standard was analyzed.  If the RRF values for this
standard were within ±25% of the RRFs obtained for the same concentration standard during
the instrument calibration, the GC/MS system was considered "in control" and the
appropriate RRF values from the calibration standards were used to calculate the mass of the
target SVOCs on sample cartridges (MT) as

                                        AQS • RRFT
                                         8-10

-------
      For GC/FID analysis, identification of target analytes was based on chromatographic
retention times relative to standards. Quantitation was accomplished using response factors
(RFs) generated from chromatographic peak areas. Calibration standards containing target
analytes were prepared on Tenax TA.  Response factors for each standard were calculated as

                                 .    RFT = ^1                                (8-4)
                                       T   MT

where MT is the mass of target analyte (ug/cartridge) and AT is peak area of target analyte.
      Mean values and  standard deviations of the RFs were calculated for each target
SVOC. The calibration curve was considered acceptable if the standard deviation for each
response factor was less than 30%. During each day of analysis, an additional standard was
analyzed. If the RF values for this standard were within ±25% of the RFs obtained for the
same concentration standard during calibration, the GC/FID system was considered "in
control".  The mean RF values generated during calibration were used to calculate the
amount of the target SVOCs on the sample cartridge as
                                                  AT
                                M-j^ug/cartridge)  = 	                           (8-5)
                                                  KT

For all analyses, chamber air concentrations were calculated by dividing My by the sample
volume in liters. Since the target analytes  were the only compounds in the air samples,
TVOC concentrations were calculated by summing individual analyte concentrations.
      8.2.3.3 Aldehydes
      During chamber tests, aldehydes in air samples were collected by passing air from the
chamber outlet through silica gel cartridges impregnated with 2,4-dinitrophenylhydrazine
(DNPH) (Waters Assoc,  Medford, Ma). Sample flow rates  were approximately 400 mL/
minute.  Sampling times  varied from 25 to 75 minutes to give nominal sample volumes of 10,
20, and 30 L.
      DNPH/aldehyde  derivatives on the sample cartridges were extracted by eluting each
cartridge with 5 mL of HPLC grade acetonitrile into a 5 mL volumetric flasks. The final
volume was adjusted to 5.0 mL and the samples aliquoted for analysis. Blank cartridges
were eluted with each sample set to identify  background contaminants.  Additional blank
                                        8-11

-------
cartridges were spiked with known amounts of DNPH/aldehyde standards as a means of
assessing recovery.
       DNPH/aldehyde derivatives in sample extracts were analyzed by HPLC with UV
detection using the conditions shown in Table 8-4. Purified and certified DNPH derivatives
of the target aldehydes were purchased for the preparation of calibration solutions.  Target
aldehydes were identified by comparison of their chromatographic retention times with those
of the purified standards. Quantitation of the target compounds was accomplished by the
external standard method using calibration standards  prepared in  the range 0.02 to 15 ng/uL
of the DNPH/aldehyde derivatives. Standards were analyzed singly for the aldehyde DNPH
derivatives and a calibration curve (through zero) was calculated by linear regression of the
concentration and chromatographic response data. All calibration curves had r2 £0.998.
       To demonstrate on-going instrumental performance, a calibration standard was
analyzed each day prior to the analysis of any samples. The calibration was considered "in
control" if the measured concentration of the aldehyde/DNPH derivatives in the standard
were 85 to 115% at the real value.
       The concentration of each target analyte in chamber air samples was calculated as:
                                     _Cy,VyiPF
                                            v,

where Cx =  Concentration of aldehyde in the sample  (ug/m3)
       C, =  Concentration of DNPH/analyte derivative in the sample extract (ng/uL)
       Vy =  Total volume of sample extract (i.e., 5000 pL)
       Vs =  Sample volume in liters
       Dp = Molecular weight of analyte -r molecular  weight of analyte/DNPH derivative

8.3    RESULTS
8.3.1   Performance of Paint Application Methods
       The performance of the paint application method was evaluated by comparing the
wet and dry masses of the paint applied to duplicate  small chamber test plates. A summary
of these results are shown in Table 8-5.  In addition to the masses, the thickness of the dried
paint samples was determined by gently removing 0.5 in. square chips of the dried paint film
                                         8-12

-------
  TABLE 8-4. HPLC OPERATING CONDITIONS FOR THE ANALYSIS OF ALDEHYDE
	EMISSIONS FROM PAINT SAMPLES	
 Parameter                                        Setting
 Instrument
 Column
 Solvent System

 Gradient

 Mobile Phase Row Rate
 Injection Size
 UV Wavelength
Waters Series 510
NOVA-PAK CIS, 3.9 x 150 mm
A: Water/Acetonitrile/Tetrahydrofuran 60/30/10 v/v
B: Acetonitrile/Water 40/60/v/v
100% A for 3 min; then a linear gradient to 100% B in
10 min. Hold 15 min at 100% B
1.5 mL/min
20 uL
360 nm
                                     8-13

-------
                                      TABLE 8-5.  PERFORMANCE OF PAINT APPLICATION METHODS
00
I
Mass Applied (g)
Paint Type Finish Mfg.
Latex Flat SW1
SW
SW
SW
SW


Semigloss GLb
GL


Gloss SW
SW


Alkyd Flat GL
GL


Semigloss GL



Gloss GL




SW



Test
1
3
4
19
20


17
18


15
16


llc
12C


13
14


2
5
6


21
22


Paint Color
Marmalade
Marmalade
Marmalade
Marmalade
Marmalade
Mean:
% RSD:
Down Yonder
Down Yonder
Mean:
% RSD:
Rose Dawn
Rose Dawn
Mean:
% RSD:
Chim Cham
Chim Cham
Mean:
% RSD:
Seafoam
Seafoam
Mean:
% RSD:
Hyacinth
Hyacinth
Hyacinth
Mean:
% RSD:
Bumbershoot
Bumbershoot
Mean:
% RSD:
Wet
11.18
12.74
11.80
9.13
10.60
11.09
12
7.12
8.29
7.71
11
10.79
10.48
11
2.1
13.53
14.83
14.18
6.5
11.19
12.23
11.71
6.3
9.72
9.50
8.78
9.33
5.3
7.09
7.56
7.33
4.5
Dry
5.67
6.18
5.75
4.46
5.13
5.44
12
3.16
3.48
3.32
6.8
5.23
5.16
5.2
0.95
9.76
12.11
10.94
15
7.97
8.54
8.26
4.9
6.39
6.39
5.88
6.22
4.7
4.08
4.31
4.20
3.9
Point f
1.9
2.0
1.8
1.9
1.8


1.3
1.4


1.7
1.6


2.7
2.6


2.3
23


2.0
1.5
1.8


1.6
1.7


Point 2
1.8
1.9
1.9
1.8
1.7


1.1
1.3


1.5
15


2.4
2.6


2.2
2.5


1.4
1.9
1.8


1.3
1.7


Dry Film Thickness
Point 3
1.8
1.9
1.8
1.4
1.6


0.9
1.0


1.6
1.3


NMd
MM"1


2.0
2.1


1.4
1.8
1.5


1.5
1.2


Point 4
1.8
2.0
1.9
15
1.6


12
1.3


1.7
1.5


25
2.7


2.2
2.3


2.0
1.8
2.0


1.4
1.7


(mils)
Points
2.6
1.8
1.9
1.4
1.6


0.9
1.1


1.6
15


1.5
2.4


2.0
25


25
3.0
2.9


1.1
15



Mean
2.0
1.9
1.9
1.6
1.7


1.1
1.2


1.6
1.5


2.3
2.6


2.1
2.3


1.9
2.0
2.0


1.4
1.6



7. RSD
17.6
4.4
2.9
14.9
5.2


16.4
13.1


5.2
7.0


23.4
4.9


5.4
6.1


25.1
28.9
26.7


13.9
14.0


    •SW - Sherwin Williams
    bGL - Glidden
    Taint could not be removed as a chip for measurement.  Thickness determined using a micrometer.
               , micrometer could not reach middle test point.

-------
from five representative locations on each test plate. The thickness of these chips was then
measured with a dial indicator.
8.3.2   Performance of Test Methods
       8.3.2.1 VOC Emissions from Alkvd Paints
       Sampling and Analysis Methods
       Evaluations of the sampling and analysis methods for measuring VOCs from alkyd
paint samples were performed to determine the accuracy, precision, background
contamination, and linear dynamic range of the proposed method for quantitating VOCs in
chamber air.
       Accuracy was evaluated as % recovery for target VOCs spiked directly onto the
sampling cartridge. Preliminary experiments were performed to demonstrate that target
VOCs could be recovered from the charcoal sampling tubes.  Since it became important to
decrease method detection limits, it was also important to demonstrate that low levels of
targets were recovered well from the charcoal tubes.  During testing, target chemicals were
spiked directly onto the tubes at several different levels (2 to 100 ug/sample).  Tubes were
extracted and extracts analyzed by GC/MS using procedures described for sample analysis
above. Percent recovery was calculated as
                         _  '          Amount Measured x 100%                   /o T\
                         % Recovery =	.	                   vo-/;
                                          Amount Spiked
Results for these recovery tests are given in Table 8-6. Percent recoveries at all spiking levels
were considered acceptable and ranged from 104 to 133%.  For the preliminary tests,
recoveries were also relatively uniform across spiking levels.
       Method controls were prepared and analyzed throughout the chamber emissions
testing. Spiking levels and % recovery values for these controls are also given in Table 8-6.
Mean recoveries ranged from 104 to 113% for these controls indicating acceptable
performance.
       Method precision was evaluated as the % RSD of replicate recovery measurements.
Data for this parameter are also given in Table 8-6 for both the preliminary recovery tests
and the method controls. Results show excellent method precision for replicate samples with
RSD values ranging from 0 to 13%.
                                         8-15

-------
            TABLE 8-6. ANALYSIS OF VOC EMISSIONS FROM ALKYD PAINT - % RECOVERY FROM METHOD CONTROLS (MC)
00
I
CTi
MC-High (n=3)'
Spike
Level
Compound (|ig/sample)
nyj-Xylene
o-Nonane
o-Xylene
Propylcyclohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
in-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
iv-Undecane
Pentylcyclohexane
n,-Dodecane
100
45
50
50
99
50
46
49
49
49
46
50
46
48
47
%
Recovery
123
128
114
127
123
118
132
118
114
112
133
130
129
120
133
%RSD
2
3
1
4 -
3
3
2
2
0
2
3
4
5
1
4
MC-Medium (n=3)'
Spike
Level %
(ug/sample) Recovery
10
'' '5
5
5
10
5
5
5
5
5
5
5
5
5
5
119
123
112
131
120
117
132
117
113
105
130
125
125
121
122
- %RSD
1
5
4
6
4
1
9
3
3
2
7
4
4
4
2
MC-Low(n=3)*
Spike
Level %
(ug/sample) Recovery %RSD
4
2
2
2
4
2
2
2
2
2
2
2
2
2
2
118
118
-no
124
117
111
119
113
109
104
123
117
115
110
104
0
3
1
1
1
1
4
1
1
2
8
6
2
2
13
MC
Test (n=6)b
Spike
Level %
(tig/sample) Recovery %RSD
13
6
6
6
13
6
6
6
6
6
6
6
6
6
6
107
108
104
110
117
111
114
106
106
107
109
112
109
113
112
8
5
9
5
8
8
8
7
12
9
6
10
9
7
8
     'Results of preliminary testing
     bResults of method controls prepared and analyzed as part of small chamber testing

-------
      Unspiked sampling cartridges and air samples collected from blank chambers were
extracted and analyzed to assess background contamination. None of the target VOCs were
detected above the quantitation limit in any of these blank samples.
      The linear dynamic range of the method for quantitating VOCs in chamber air was
determined based on the dynamic range of the calibration curve and the range of sample
volumes collected during emissions testing.  During instrumental calibration, linearity for  all
VOC target analytes was demonstrated for standards in the range 0.5 to 500 ng/uL. The
instrument was considered linear if the RRF values did not change substantially as the
concentration changed. An example of the relative response factors generated over the
calibration range during instrumental calibration is given in Appendix C.  Standard
concentrations (C9td) were then converted to air concentrations (Cair) as
                                   3.   C>a(ng/vL) * 2000 liL
                                                   1000
where Vair is the volume of the chamber air sample collected for analysis.  The calculated
linear dynamic range using this approach was 0.05 to 2000 mg/m3 for all target VOCs.
During small chamber tests, this range was found acceptable for quantitating emissions over
a 24-hour test period.
      Chamber Test Method
      Chamber recovery tests were performed to demonstrate that the chamber test method
could be used to accurately measure VOC emissions from alkyd paint samples.  These tests
were designed to measure recovery of selected target VOCs in the chamber outlet air in the
presence of a constant  concentration source. The decay in chamber concentrations over time
after the source had been removed from the chamber was also monitored to evaluate sink
effects within the chamber. This was considered important for compounds that gave poor
recoveries.  VOCs were selected for the recovery tests that represented the range of
volatilities found in the chamber air samples during emission tests.  The concentrations of
VOCs selected for study were similar to those measured during the first hour of emissions
testing with actual paint samples.
      During testing,  a constant concentration of VOCs was generated using a syringe
pump that injected a small, constant volume of a neat mixture of the target compounds into
the gas stream entering the chamber. A Brownlee Labs Micro Gradient SFC system was used
                                        8-17

-------
to deliver a constant volume (0.1 to 1 pL/minute) of the neat mixture to the heated inlet air.
The air was passed though a heated glass mixing bulb and a heated transfer line into the
chamber. This vapor mixture was then introduced directly into the chamber for a minimum
of 24 hours.
       During the first test (chamber A, Test 7 in Table 8-1), samples of the inlet air mixture
were collected at three time points during the 24-hour equilibration period. This was done to
assure that a constant concentration mixture was generated.  For chamber B (Test 8 in Table
8-1), only a single inlet sample was collected at the end of the equilibration period. After the
equilibration period (approximately 24 hours), air samples were also collected at the chamber
outlet. Once sample collection under constant concentration conditions was complete, the
                      i
vapor generator was turned off thus removing the source from the chamber. Additional
samples were then collected from the chamber outlet at selected time points after the source
had'been removed. All collected samples were extracted and analyzed for target VOCs as
described above.  Percent recovery for the target VOCs from the test chamber were then
calculated as
                                            C
                               % Recovery = -JL x 100%                          (8-9)
                                            ("out

where Cin was the air concentration measured in the air mixture prior to introduction into
the test chamber.  Cout was the air concentration measured in the chamber outlet air.
       Data for the chamber recovery study are provided in Table 8-7 which gives the
measured VOC concentrations in the chamber inlet  air and the % recovery of the target
VOCs  in the chamber outlet air.  For the samples collected after the source had been
removed, the % recovery value that would occur in the absence of sink effects is also given
as the  theoretical % recovery. Several important observations can be made from the data
provided in Table  8-7.
       •      The low variability in the VOC air concentrations at the chamber
              inlet suggests that constant concentration conditions were
              achieved during the chamber equilibration period.
       •      Good recoveries were calculated for all of the test compounds
              under constant concentration conditions (T=O) suggesting
              minimal losses of target VOCs during emissions testing.

                                         8-18

-------
                               TABLE 8-7. CHAMBER RECOVERY TESTS FOR ALKYD PAINT COMPONENTS
oo
I

Compound
TEST 7 (CHAMBER A)
m,j>-Xylene
n.-Nonane
o-Xylene
Propylcyclohexane
n-Decane
1 ,2,4-Trimethylbenzene
trans-Decahydronaphthalene
iv-Undecane
Pentylcyclohexane
iv-Dodecane
TEST 8 (CHAMBER B)
m,j>-Xylene
rv-Nonane
o-Xylene
Propylcyclohexane
ivDecane
1 ,2,4-Trimethy Ibenzene
trans-Decahydronaphthalene
n.-Undecane
Pentylcyclohexane
iv-Dodecane
THEORETICAL % RECOVERY*1
aFor test 7, n=3; for test 8, n=2.
kNot tested.
Mean Inlet
Concentration
±SD (mg/m3)

100±12
210±22
32±3
130±18
280±31
68±5
99±9
140±23
16±2
43±4

100
180
33
130
240
58
91
120
17
40



% Recovery

TM)3

101±2
104±7
107±1
109±2
104±4
96±4
105±7
108±2
104±8
106±3

100±15
103±18
99±11
93±7
114±5
110±17
109±10
109±3
99±17
108±9
100



T=l.l

38
42
37
41
59
34
36
45
34
35

38
57
43
42
69
43
45
48
34
43
33.3



T=2.1

13
18
13
16
26
13
15
18
14
14

16
21
14
16
28
16
17
19
14
17
12.2



T=3.1

NTb
NT
NT
NT
NT
NT
NT
NT
NT
NT

6.9
8.4
5.7
7.2
14
6.5
8.3
8.2
6.2
7.7
4.5



T=4.1

2.0
2.6
2.0
2.4
4.4
2.0
2.2
3.0
ND
2.4

2.7
3.5
2.6
3.3
5.7
3.0
3.1
4.2
ND
3.6
1.7



T=8.3

ND*
0.057
ND
0.054
0.12
ND
0.061
0.13
ND
0.18

NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
0.025



T=116

ND
ND
ND
ND
0.047
ND
ND
0.068
ND
ND

NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
3.3X10"1



T=24.6

ND
ND
ND
ND
0.017
ND
ND
ND
ND
ND

NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
0


         cBelow the method quantitation limit of 0.05 mg/m3.

         dCalculated as Ct = C0 • e'rt

           where r is the air exchange rate (h"1), t is time after source removal, Ct is air concentration at t, and C0 is air concentration before

           source removal.

-------
       •     Measured recoveries after the source was removed from the
             chamber inlet (T=l.l to 24.6) agreed well with the theoretical
             recoveries indicating that the test chamber contains few sinks for
             the target VOCs. Greatest deviations appeared to occur for the
             n-alkanes whose chamber air concentrations appeared to decay
             more slowly than predicted in the absence of sink effects.
       •     Results for the two chamber tests were similar suggesting good
             reproducibility of the overall small chamber method.
       8.3.2.2. SVOC Emissions from Latex Paint Samples
       Sampling and Analysis Method
       Evaluations of the method for quantitating SVOC emissions in chamber air samples
are similar to those described above. Tests were performed to determine the accuracy,
precision, background contamination, and linear dynamic range for the sampling and
analysis method.
       Several of the SVOCs that are emitted from latex paint samples are very polar
compounds which present problems during chromatographic analysis. Most importantly,
when a small mass is injected into a GC column, adsorptive losses may occur. As a result, a
linear calibration curve cannot be generated over a large calibration range and response
factors or relative response factors will increase with increasing sample mass. This is
illustrated in Table 8-8 which gives  the relative response factors (RRFs) for target SVOCs
generated during GC/MS analysis.  As shown in the table, the relatively nonpolar targets
such as o^xylene, Texanol, 2-(2-butoxyethoxy)ethyl acetate, and  2-(2-butoxyethoxy)ethanol
have relatively constant RRFs over the entire calibration range (25 to 2500 ng).  In contrast,
the more polar target chemicals (i.e., ethylene glycol, 1,2-propanediol, 2-(2-
methoxyethoxy)ethanol, diethylene glycol, and dipropylene glycol) have very low RRFs for
the low injected masses with increasing RRF values as the injected mass increases.
       GC/MS analysis was only performed for the range find test (Test 1, Table 8-1).
Analysis of all other air samples collected for emissions testing was performed by GC/FID.
GC/FID analysis was selected since it gives a linear response for high masses injected (i.e.,
>1 ug). This was considered important since the SVOC levels measured during the range
finding tests were so high.  Fortunately, only a few SVOCs are  emitted from paint samples,
thus the use of chromatographic retention times for compound identification was possible.

                                          8-20

-------
       TABLE 8-8. RELATIVE RESPONSE FACTORS FOR THE ANALYSIS OF SVOC EMISSIONS FROM LATEX PAINT SAMPLES BY GC/MS
00
I
ro
Relative Response Factor per Spiking Level (ng)
Compound
Ethylene glycol
Ethylene glycol
Ethylene glycol
1,2-Propanediol
1,2-Propanediol
2-(2-Methoxyethoxy)ethanol
2-(2-Methoxyethoxy)ethanol
Diethylene glycol
Diethylene glycol
Dipropylene glycol
Dipropylene glycol
2-(2-Butoxyethoxy)ethanol
2-(2-Butoxyethoxy)ethanol
2-(2-Butoxyethoxy)ethyl acetate
2-(2-Butoxyethoxy)ethyl acetate
Texanol
Texanol
o-Xylene
o-Xylene
Ion
62
43
31
61
76
89
59
75
45
89
59
132
89
87
101
173
143
91
106
25
-
-
-
-
-
0.35
0.48
0.42
0.17
0.48
0.17
0.098
0.37
2.2
0.32
0.37
0.33
4.9
2.5
50
-
-
-
-
-
0.54
0.85
0.42
0.16
0.76
0.24
0.13
0.45
3.5
0.48
0.46
0.40
4.5
2.4
100
-
-
-
-
.
0.37
0.52
0.29
0.12
0.45
0.16
0.10
0.37
2.4
0.31
0.38
0.31
4.8
2.4
250
0.050
0.014
0.017
0.071
0.012
0.55
0.90
0.59
0.20
0.82
0.27
0.12
0.46
3.5
0.44
0.44
0.39
4.5
2.4
500
0.098
0.030
0.029
0.11
0.021
0.62
1.2
1.1
0.42
0.96
0.33
0.10
0.49
3.8
0.41
0.41
0.36
4.4
2.1
750
0.11
0.037
0.039
0.15
0.027
0.66
1.4
1.3
0.62
1.2
0.44
0.087
0.48
3.5
0.35
0.33
0.32
4.3
1.9
1000
0.15
0.050
0.050
0.23
0.040
0.81
1.7
1.8
0.88
1.5
0.59
0.094
0.55
3.9
0.40
0.37
0.37
4.9
2.0
1500
0.18
0.056
0.055
0.26
0.048
0.84
2.1
2.1
0.89
1.7
0.69
0.094
0.54
3.8
0.40
0.34
0.33
4.8
1.8
2500
0.18
0.056
0.057
0.25
0.047
0.78
2.1
2.3
1.0
1.7
0.68
0.097
0.52
3.4
0.37
0.32
0.29
4.2
1.7

-------
       The range of the injected mass for calibration was limited during emissions testing
and compound amounts were selected for calibration that were in the range expected in the
chamber air samples.  This was done in order to generate a linear calibration.  An example of
response factors (RFs) generated by GC/FID for the analysis of chamber air tests is given in
Table 8-9. Generally, these response factors do not show a trend of increasing RFs with
increasing mass.  Exceptions are 1,2-propanediol where the lowest calibration standard shows
low RF values compared to the other standards and diethylene glycol where RFs increased
with increasing mass over the entire calibration range.  In addition to poor linearity over the
calibration range, the RFs for diethylene glycol  were very erratic.  Throughout the sample
analysis period, the RF for the daily calibration check was not "in control" (greater  than 25%
deviation from the mean RF). As a result, quantitative data could not be generated for this
compound.  This poor performance is probably due to the very polar nature of the
compound.
       During emissions testing, the linear dynamic range of the test method was estimated
based on the air sample volume and the calibration range as described above. The linear
dynamic ranges for target SVOCs estimated for the small chamber tests are given in
Table 8-10. This range is considerably smaller than that reported for the VOCs from alkyd
paints due to the much smaller range for the calibration standards.
       Accuracy and precision of the sampling and analysis method were evaluated by
determining % recovery and %RSD for target SVOCs spiked onto  sampling cartridges.
Results of these analyses are given in Table 8-11.  As indicated in the table, six method
controls were prepared and analyzed throughout emissions testing.  Results for one of the
controls gave very high % recovery values.  It appeared as though this control had been
spiked at twice the specified amount.  Mean % recovery and %RSD values have been
calculated both with and without data from this control. When the high control is deleted,
the precision and accuracy of the method appears to be acceptable for all compounds except
diethylene glycol. Poor precision for diethylene glycol is consistent with the erratic response
found during the daily calibration checks.  This result again suggests that quantitative results
cannot be generated for diethylene glycol.
       Unspiked sampling cartridges and air samples collected from blank chambers were
analyzed to assess background contamination.  None of the target SVOCs were detected
above the quantitation limit in any of the blank samples.

                                         8-22

-------
  TABLE 8-9. EXAMPLE RESPONSE FACTORS FOR THE ANALYSIS OF LATEX PAINT USING
                            FLAME IONIZATION DETECTION
Compound
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
Standard
Mass

0.186
0.226
0.174
1.2
0.067
Response Factor (X104)
Xla
89.7
161
184
282
35.1
X3
118
84.8
203
291
45.5
X5
189
142
285
402
55.8
X10
17.7
143
263
354
80.2
X30
141
122
205
270
89.6
X60
NAb
132
NA
NA
NA
Mean
143
131
228
319
61
% RSD
29
21
19
18
38
a  Mass injected for the standard is equal to the standard mass times the number indicated.
b  Not analyzed; high standard for ethylene glycol added to allow quantitation of very high
   concentration air samples.
                                     8-23

-------
  TABLE 8-10. ESTIMATED LINEAR DYNAMIC RANGE FOR SVOC TEST METHOD
Air Concentration

Compound
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
Lowest Quantifiable3
mg/m3 mg/m3/gb
0.37 0.037
0.53 0.053
0.35 0.035
2.4 0.24
0.13 0.013

Maximum Quantifiable
11
32
10
72
3.9
a  Defined as method quantitation limit.
b  Method quantitation limit reported as mg/m3 per gram of paint; calculated based on a
   10 gram paint sample used for chamber testing.
                                         8-24

-------
   TABLE 8-11.  ANALYSIS OF SVOC EMISSIONS FROM LATEX PAINT SAMPLES -
                              METHOD CONTROLS
Compound
1,2-Propanediol
Ethylene glycol
2-(2-butoxyethoxy)ethanol
Texanol
Diethylene glycol
Spike
Level
(ug/m3)3
1.8
4.6
1.8
12
0.68
% Recoveryb
88(107)
78(93)
92(109)
94(107)
84(133)
%RSD
4.2(36)
9.4(34)
21(33)
16(28)
87(86)
a  Assuming a 0.5 L sample volume.
b  A total of six method controls were analyzed throughout emissions testing; one control
   appeared to be spiked at twice the level. Results from this control were deleted from
   calculations.  Values in parentheses were calculated with sixth method control included.
                                     8-25

-------
       Chamber Test Method
       Chamber recovery tests (Tests 9 and 10 on Table 8-1) were performed to demonstrate
that the chamber test method could be used to accurately measure SVOC emissions from
latex paints. These tests were designed to measure recovery of selected target SVOCs in the
presence of a constant concentration source.  The decay in chamber air concentrations over
time after the source had been removed from the chamber was also measured to evaluate
sink effects.  The procedures described in Section 8.3.2.1 were used for these tests.  Tests
were performed in both chambers.  SVOCs for testing were selected to represent the most
abundant compounds measured during the chamber emission tests.  Concentrations
generated for testing were selected to be similar to those found during the emissions tests.
       Results for the chamber recovery tests are given in Table 8-12.  Data are provided on
the measured SVOC air concentrations in the chamber air inlet and the % recovery of the
target SVOCs in the chamber outlet air.  For the samples collected after the source had been
removed, the % recovery value that would occur in the absence of sink effects is also given
as the Theoretical % recovery.  The following observations can be made from the results
provided in Table 8-12.
       •     The low variability in the SVOC air concentrations at the chamber inlet
             indicate that constant concentration conditions were achieved during the
             chamber equilibration period.
       •     Good recoveries were calculated for the test compounds under constant
             concentration conditions (T=0)  suggesting minimal losses of target
             VOCs during emissions testing.
       •     Results for the two chamber tests were similar suggesting  good
             reproducibility for the overall small chamber method for measuring
             SVOC emissions from latex paint samples.
       •     For all of the test compounds, measured recoveries after the source had
             been removed (T = 1.2 to 6.2) were higher than  the theoretical value in
             the absence of sink effects.  This result suggests that there may be sinks
             for the SVOCs within the chambers.  The trend  is most noticeable at the
             latter time points. Recoveries for 2-(2-butoxyethoxy)ethanol are most
             similar  to the theoretical values suggesting the weakest sink effects. In

                                         8-26

-------
      TABLE 8-12. CHAMBER RECOVERY TESTS OF LATEX PAINT COMPONENTS
Mean Inlet
Concentration
Compound ± SD (mg/m3)
TEST9
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
TEST 10
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
THEORETICAL % RECOVERY1*

5.3 ± 0.14
90 ± 0.57
3.0 ± 0.07
18 ± 0.91

4.2 ± 0.21
70 ± 5.6
3.0 ± 0.71
16 ± 1.2

% Recovery
T=0a T=1.2

94 ±2
92 ±2
88±3
100 ±1

131 ±4
122 ±2
100 ±18
135 ± 14
100

35
38
27
35

32
38
23
32
30
T=2.2

14
19
11
15

15
17
9.1
15
11
T=3.2

6.0
9.4
5.6
7.4

5.5
9.1
4.0
7.5
4.0
T=4.2

2.7
4.7
3.6
4.4

5.2
5.0
2.1
5.3
1.5
T=6.2

1.1
1.8
1.6
1.9

0.19
1.3
1.2
1.9
0.2
aMean recovery and SD, n=3.
bCalculated as Ct = C0 • e'rt
   where r is the air exchange rate (h"1), T is time (hours) after source removal, CT is air
   concentration at T, and C0 is air concentration before source removal.
                                     8-27

-------
       contrast, recoveries for ethylene glycol show the greatest deviation from the
       theoretical recovery suggesting the strongest sink effects.
       8.3.2.1 Aldehyde Emissions
       Aldehyde emissions from paint samples were collected on silica gel/DNPH cartridges.
The DNPH/aldehyde derivatives formed during sample collection were eluted with
acetonitrile.  They were analyzed by HPLC/UV. Because this is a previously validated
method (9), performance evaluation studies were not performed; rather quality control
samples were analyzed to assess method performance throughout the study. QC samples
included method controls, method blanks, and chamber blanks. Method controls were
sampling cartridges spiked directly with a solution of aldehyde/DNPH derivatives. Method
controls were extracted and analyzed along with chamber air samples. Method blanks were
unspiked sampling cartridges that were extracted and analyzed along with chamber air
samples. Chamber blanks  were 30 L air samples collected from the chamber immediately
prior to emissions testing.
       Results of these analyses along with information on method quantitation limits are
given in Table 8-13. Highest blank and chamber background levels were found for
formaldehyde and acetaldehyde. The amounts reported for the method blanks are typical of
those found on the DNPH cartridges (9). For these two compounds, the method quantitation
limit was determined by the concentration found in the chamber blanks. The recovery data
for the method controls indicate acceptable  accuracy (% recovery = 80 to 119%) and precision
(S.D. = 8.3 to 25%) for all the target aldehydes except acrolein. The reason for poor recovery
for this compound is unknown. Since acrolein was not recovered from the method controls,
quantitative results are not reported  for this compound.
8.3.3   VOC Emissions from Alkyd Paint Samples
       Results of the range finding test (Test 2 in Table 8-1) performed with the Glidden
gloss alkyd paint (Hyacinth) are given in Table 8-14 and 8-15. Table 8-14 represents
measured chamber air concentrates (mg/m3) for each of the target VOC at each sampling
paint.  Table 8-15 provides similar data reported as measured air concentration per gram of
paint.  Results for these tests show a very rapid increase in chamber air content ratios for the
target VOCs with a corresponding rapid decay.  For most of the target VOCs the highest air
concentrations were measured at the first sampling point (t=4.6 hours). Only the least
volatile components, rv-undecane, pentylcyclohexane and n.-dodecane give maximum

                                         8-28

-------
                              TABLE 8-13. METHOD PERFORMANCE DATA FOR ALDEHYDE TESTING
CO
I
ro
UD
                       Parameter             Formaldehyde  Acetaldehyde  Acrolein     Propionaldehyde    Benzaldehyde

            Amount in method blank              25 ± 17        59 ± 31       3.0 ±7.9         1.0 ± 3.2          0 ± 0
            ng/sample ± S.D. (n=10)

            ConcentraHon in chamber            0.85 ± 0.57      2.2 ± 0.40    0.07 ±0.16      0.02 ± 0.06       0.07 ±0.17
            background ug/m3 ± S.D.a (n=9)

            % Recovery of method controls ±      105 ± 8.3       9 ± 25        NRe          89 ± 16          120 ± 17
            S.D.a (n=10)
Method Quantitation Limit
- Calibration Curveb
ng/sample
ug/m3 (10L)C
ug/m3 (20L)
ug/m3 (SOL)
- Chamber blanksd - (ug/m3)


43
4.3
2.2
1.4
2.6*


40
4.0
2.0
1.3
3.4*


24
2.4*f
1.2*
*
0.80
0.56


25
2.5*
1.3*
0.83*
0.26


36
3.6*
1.9*
1.2*
0.58
            "Spiked from 71 to 215 ng/sample. Calculated as:

                                                w D         MMC ~ MMB   innw
                                                % Recovery = - x 100%
            where [A]MC and [A]MB are the amount measured in the method control and method blank respectively. [A]s is the
            amount spiked onto the method control.
            W«.
            ^ased on the lowest calibration standard analyzed.
            °Air concentration calculated based on air sample volume in parenthesis.
            dEqual to the mean plus 3 x S.D. of concentration found in chamber background.
            eNot recovered.
            f* indicates method quantitation limit value that was used.

-------
            TABLE 8-14. RESULTS OF RANGE FINDING TEST (TEST 2) FOR VOC EMISSIONS FROM ALKYD PAINT
                                         CHAMBER AIR CONCENTRATION
00
I
CO
o
Compound
TEST 2
m,r>Xylene
ivNonane
p_-Xylene
Propylcyclohexane
3- & 4-Ethyl toluene
1 ,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1 ,2,4-Trimethylbenzene
1 ,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
ri-Undecane
Pentylcyclohexane
ivDodecane
Chamber Air Concentrations (mg/m )
T=4.6b
29
160
10
71
33
15
450
11
46
13
35
39
110
5.9
10
T=8.6
ND0
21
ND
12
12
6.9
310
ND
25
8.9
33
30
150
8.1
17
' T=12.6
ND
ND
ND
ND
ND
ND
110
ND
9.0
ND
22
14
120
7.3
22
T=24.6
ND
ND
tib
ND
ND
ND
ND
ND
ND
ND
ND
ND
11
ND
14
T=48.8
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
T=716
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
      "Test 2 on Table 8-1, performed using Glidden 4550-76262 Gloss: (Hyacinth); sample wt. - 9.72 g.
      ''Sampling time in hours.
      °Below the method quantitation limit (5 mg/m3).

-------
           TABLE 8-15. RESULTS OF RANGE FINDING TEST (TEST 2) FOR VOC EMISSIONS FROM ALKYD PAINT* -
                               CHAMBER AIR CONCENTRATION PER GRAM OF PAINT
00
oo
Compound
TEST 2
m,j>Xylene
rvNonane
o-Xylene
Propylcyclohexane
3- & 4-Ethyl toluene
1 ,3/5-Trimethylbenzene
ri-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1 ,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
rj-Undecane
Pentylcyclohexane
ivDodecane
Chamber Air Concentrations (mg/m3)
T=4.6b
3.0
17
1:1
7.3
3.4
1.5
46
1.1
4.8
1.3
3.6
4.1
11 .
0.60
1.0
T=8.6
ND0
2.1
ND
1.3
1.2
0.72
32
ND
2.6
0.91
3.4
3.1
15
0.83
1.7
T=12.6
ND
ND
ND
ND
ND
ND
12
ND
0.92
ND
2.2
1.5
12
0.76
2.3
T=24.6
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
1.1
ND
1.4
T=48.8
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
T=72.6
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
      aTest 2 on Table 8-1, performed using Glidden 4550-76262 Gloss:(Hyacinth); sample wt. - 9.72 g
      ''Sampling time in hours
      'Below the method quantitation limit

-------
concentrations at later time points. Because of the rapid decay in chamber air concentrations,
VOC emissions could not be quantitated at the later time points.  Based on the results of this
test, sampling points were selected as 0.5,1, 2,3, 4, 8,12, and 24 hours to allow better
characterization during the period of highest emissions. In order to reduce the method
quantitation limit for sampling of the later time points, larger sample volumes (20 L) were
collected and lower concentration standards (0.5 ng/uL) were added for instrument
calibration.
       Once test conditions were defined, a set of single chamber repeatabilty tests were
performed to evaluate the variability of the small chamber emission test for alkyd paint
samples when identical conditions (including test chamber and paint type) were used. These
two tests, Tests 5 and 6, are described in Table 8-1 and were performed using a Glidden
gloss alkyd paint (Hyacinth). Results for the single chamber  repeatability tests are given in
Tables 8-16 and 8-17. Similar to the range finding tests, results are given for both chamber
air concentrations (Table 8-16) and chamber air concentrations per gram of paint (Table 8-17).
Each table gives concentration results for both tests at each time point. The variability of the
measured concentrations between tests is presented as the %RSD between concentration
values for samples collected at the same time point for each test.  Results show highest air
concentrations for the n-alkanes (n-decane, tv-nonane, and ivundecane).  For the more volatile
compounds, the highest chamber air concentrations are  seen for the earliest sample collection
points.  For.example, the highest chamber air concentrations for rn,rj-xylene are seen at 1.2
hours. In contrast, the highest chamber air concentrations for the less volatile compounds are
                       s
seen at later time points. For ivundecane,  highest air concentrations are seen at 4.1 hours.
For ivdodecane, highest air concentrations are seen at 12 hours. These trends are seen for
both chamber air concentrations and chamber air concentrations per gram of paint.
       Variability between the two tests has been evaluated as the %RSD between paired
chamber air concentrations for the two tests. Data in Tables 8-16 and 8-17 show reasonably
low %RSD values between the two tests.  Slightly better agreement (lower %RSD values) is
found when chamber air concentrations are expressed per gram of paint (Table 8-17).  For
these single chamber repeatability tests, the highest %RSD values were calculated for the
most volatile components at the earliest time points. This result may be due to the fact that
the samples were placed into the chamber at different times (2 vs. 6 minutes) after the paint
application.  It  is feasible that under these  conditions, a substantial and varying fraction of

                                          8-32

-------
    TABLE 8-16. RESULTS OF SINGLE CHAMBER REPEATABILITY TESTS (TESTS 5 AND 6) FOR VOC
              EMISSIONS FROM ALKYD PAINTS3 - CHAMBER AIR CONCENTRATION
Compound
TESTS
mj>-Xylene
tt-Nonane
oO(ylene
Propylcydohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1 ,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
iv-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
TEST6
m.p-Xvlene
rf-Nonane
o-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1 ,2,4-Trimethylbenzene
1 ,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
jn-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
%RSD
m^g-Xylene
iv-Nonane
o-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1 ,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1 ,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
n-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
Chamber
T=0.65b
120
250
38
89
48
16
380
15
51
14
20
33
59
3.6
5.1
5600
T=0.70
78
190
26
78
39
14
330
12
42
11
17
31
53
3.4
4.3
4420

27
18
28
8.7
15
10
10
13
13
14
9.0
5.9
7.2
4.8
12
17
T=1.2
130
300
45
100
65
22
510
19
70
20
30
52
100
7.8
9.0
7760
T=1.2
77
210
25
69
41
14
360
12
43
12
20
32
60
3.8
5.5
4730

36
25
40
26
33
31
25
31
34
38
29
34
35
48
34
34
T=2.1
92
280
32
100
59
22
570
18
68
20
35
54
110
8.8
12
7300
T=2.2
54
220
20
80
47
18
530
14
58
17
32
46
110
7.1
11
5950

37
17
33
17
16
13
4.3
17
11
10
5.6
11
0.80
16
1.3
14
Air Concentrations (mg/m3)
T=3.1
69
250
25
110
58
21
600
18
70
21
42
57
140
10
13
7260
T=3.2
72
250
26
84
49
18
480
15
56
16
27
43
92
5.8
8.6
5900

3.7
0.37
1.4
16
12
11
16
13
16
19
29
19
30
37
31
15
T=4.1
40
180
16
71
44
18
540
15
58
18
41
52
140
10
16
5940
T=4.2
35
190
14
81
44
18
550
14
57
17
40
55
130
9.3
13
5820

9.5
1.9
8.8
9.3
0.20
2.7
1.6
3.0
1.6
4.0
1.2
4.0
4.4
2.4
11
1.4
T=7.8
4.5
16
2.6
6.7
13
6.0
140
1.5
19
9.0
17
20
78
11
17
1950
T=7.9
3.5
15
2.1
5.2
13
6.0
150
2.2
20
9.0
14
18
82
10
16
1780

17
4.3
13
18
4.2
0.7
2.4
27
2.4
0.1
12.6
6.9
3.3
3.7
1.0
6.4
T=11.8
0.45
3.3
0.30
2.2
4.2
2.7
89
1.5
10
5.0
13
13
74
9.9
19
1320
T=11.9
0.34
3.0
0.25
1.9
4.0
2.4
95
1.6
11
5
14
14
82
10
21
1300

20
7.0
14
10
3.8
5.8
4.6
5.7
1.4
0.2
8.0
2.3
7.0
0.7
6.1
1.1
T=24.8
ND0
ND
ND
ND
ND
ND
0.58
ND
0.19
0.24
0.87
0.34
15
1.6
16
176
T=24.9
ND
ND
ND
ND
ND
ND
0.48
ND
0.21
0.23
0.73
0.29
13
1.5
16
163

-
-
-
-
-
-
13
-
4.6
5.7
13
10
6.8
4.7
1.4
5.4
  Tests 5 and 6 on Table 8-1, performed using Glidden 4550-76262 Gloss:(Hyacinth). Test 5 - sample wt. 95
  g; Test 6 - Sample wt. 8.8 g.
b Sampling times in hours.
c Below the method quantitation limit of 0.05 mg/m3.

                                           8-33

-------
TABLE 8-17. RESULTS OF SINGLE CHAMBER REPEATABILITY TESTS (TESTS 5 AND 6) FOR VOC
EMISSIONS FROM ALKYD PAINTS" - CHAMBER AIR CONCENTRATION PER GRAM OF PAINT
Compound
TESTS
















rn,p-Xylene
n-Nonane
g-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
n-Undecane
Pentylcyclohexane
n-Dodecane >
TVOC
TEST 6
















rnj2rXylene
"'n-Nonane
2-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1 ,35-Trimethylbenzene
n-Decan6
2-Ethyl toluene
1,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
n-Undecane
Pentylcyclohexane
n-Dodecane ;
TVOC
Chamber Air Concentrations (mg/m3) per gram of paint
T=0.65b 7=1.2 T=2.1
12
26
4.0
9.3
5.1
1.7
40
15
5.3
1.4
2.1
35
6.2
0.38
0.53
589
14
32
4.7
11
6.9
2.3
54
2.0
7.4
2.1
3.1
5.5
11
0.82
0.94
817
10
30
3.4
11
6.2
2.3
59
1.9
7.2
2.1
3.7
5.7
12
0.93
1.2
768
T=0.70 T=1.2 T=2.2
8.9
22
2.9
8.9
4.4
1.6
37
1.4
4.8
1.3
2.0
3.5
6.1
0.38
0.49
503
8.8
24
2.9
7.9
4.6
1.6
41
1.4
4.9
1.3
2.2
3.7
6.8
0.44
0.62
538
6.1
26
2.3
9.1
5.4
2.0
60
1.6
6.6
1.9
3.7
5.3
13
0.81
1.3
677
T=3.1
7.2
26
2.7
11
6.1
2.2
63
1.9
7.4
2.2
4.4
6.0
14.9
1.0
1.4
764
T=3.2
8.2
28
2.9
9.6
5.6
2.1
55
1.7
6.4
1.8
3.1
4.9
105
0.66
1.0
671
T=4.1
4.2
19
1.7
7.4
4.7
1.9
57
1.5
6.1
1.9
4.3
5.4
15
1.0
1.6
625
T=4.2
4.0
22
1.6
9.2
5.0
2.0
63
1.6
65
1.9
4.6
6.2
15
1.1
1.5
662
T=7.8
0.47
1.7
0.27
0.71
1.4
0.63
15
0.16
2.1
0.94
1.8
2.1
8.2
1.2
1.7
205
T=7.9
0.40
1.7
0.24
0.59
1.4
0.68
17
0.25
2.3
1.0
1.6
2.1
9.3
1.2
1.9
203
T=11.8 T=24.8
0.047
0.35
0.032
0.23
0.44
0.28
9.3
0.15
1.1
0.52
1.3
1.4
7.8
1.0
2.0
139
ND°
ND
ND
ND
ND
ND
0.061
ND
0.020
0.026
0.091
0.035
15
0.16
1.7
185
T=11.9 T=24.9
0.038
0.34
0.028
0.21
0.45
0.28
11
0.18
1.2
0.56
1.6
1.6
9.3
1.1
2.4
148
ND
ND
ND
ND
ND
ND
0.055
ND
0.023
0.026
0.083
0.033
1.5
0.17
1.8
185
%RSD
















a

b
c
m.p-Xylene
n-Nonane
c^-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1 ,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
ri-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
22.1
125
22.1
3.1
9.8
4.9
4.6
7.7
7.3
8.2
3.4
0.3
1.6
0.7
6.7
11
Tests 5 and 6 on Table 8-1, performed using
g; Test 6 - Sample wt. 8.78 g.
Sampling times in hours.
Below the method quantitation


30.9
19.7
34.6
20.4
27.6
255
19.4
26.0
28.3
32.6
23.6
285
30.1
43.3
29.1
29
Glidden


32.1
11.1
27.8
115
10.6
7.6
1.2
11.4
5.7
4.5
0.0
5.2
4.8
10.1
4.2
8.9
9.3
5.2
7.0
10.9
6.6
5.5
105
7.1
10.4
13.7
23.7
13.8
24.6
31.2
25.7
9.2
3.9
7.5
3.3
14.9
5.4
2.9
7.2
2.6
4.0
1.6
4.4
9.6
1.2
3.2
5.9
4.1
11.8
1.3
7.3
12.4
1.4
4.9
8.0
32.0
7.9
5.6
7.0
1.4
8.9
1.9
4.6
0.69
4550-76262 Gloss:(Hyacinth). Test


limit of -0.005 mg/m3 per


gram of


paint.



14.1
1.4
8.6
45
1.8
0.3
10.2
11.2
7.0
5.4
135
7.9
12.6
6.3
11.6
4.4
5 - sample



-
-
-
-
-
-
7.3
-
10.2
0.1
7.1
4.5
1.4
8.6
7.0
0
wt. 95



                                8-34

-------
the more volatile components could have been emitted from the paint sample before it was
placed into the chamber.  To avoid this problem for additional tests, paint samples were
placed into the chambers 5 minutes after application.  Five minutes was selected as
reasonable, since it should allow all application activities to take place without excessive
delays in placing the paint sample into the chamber.
       Two sets of interchamber variability tests were then performed to evaluate
reproducibility of the test method across chambers.  Tests 11 and 12 as shown in Table 8-1
were performed using a Glidden flat alkyd paint (Chim Cham). Air concentration data for
these tests are provided in Table 8-18; air concentration data normalized per gram of paint
are given in Table 8-19. .Both tables give calculated  %RSD values for measured air
concentrations for paired samples collected during the two tests.  Similar results for Test 13
and 14 are given in Tables 8-20 and 8-21.  Tests 13 and 14 were performed using a Glidden
serriigloss alkyd paint (Sea Foam).
       Data for these two sets of interchamber variability tests generally show the same
trends as discussed for the single chamber repeatability  tests. Variability between tests
performed in different chambers is low as indicated by the low %RSD values calculated for
air concentrations for paired samples.  For these tests, high %RSD values are only reported at
the latter time points when chamber air concentrations are very low.  Measured air
concentrations for all four of these tests were still relatively high (approximately 10 mg/m3)
at the end of  the 24-hour test period for the least volatile components (i.e., 2-methyldecane,
trans-decahydronaphthalene, ivundecane, pentylcyclohexane, and ivdodecane). This is in
contrast to the concentration measurements for these same compounds measured during  the
single chamber repeatability tests using the gloss paint type. It is interesting to note that
when the paint samples were removed from the chamber at  the end of the 24-hour test
period, the gloss paint samples (tests 3 and 4) were  dry, whereas the flat (tests 11 and 12)
and the semigloss (tests 13 and 14) paint samples were still tacky.  This visual observation for
the flat and semigloss paint samples is consistent with continued emission of the less volatile
components reported in Tables 8-18 to 8-21.
       Tests 21 and 22  (Table 8-1)  were performed to evaluate the effect of surface air
velocity on VOC emissions for the alkyd paints .  A Sherwin Williams gloss alkyd paint
(Bumbershoot) was used for these  tests.  Test 21 was performed using a fan inside the
chamber to generate air velocities of approximately  10 cm/s across the surface of the paint

                                         8-35

-------
TABLE 8-18. RESULTS OF INTERCHAMBER VARIABILITY TESTS (TESTS 11 AND 12)
FROM ALKYD PAINTS3 - CHAMBER AIR CONCENTRATION
Compound
TEST 11
mjj-Xylene
iv-Nonane
o-Xylene
Propylcyclohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1 ,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
n-Undecane
Pentylcyclohexane \
n-Dodecane
TVOC
TESJ12
m,r>-Xylene
n.-Nonane
pj-Xylene
Propylcyclohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1 ,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
ri-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
%RSD
m.p-Xylene
iv-Nonane
o-Xylene
Propylcyclohexane
3- & 4-Ethyl toluene
1 ,3,5-Trimethylbenzene
iv-Decane
2-Ethyl toluene
1 ,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
n-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
FORVOC
EMISSIONS
Chamber Air Concentrations (mg/m3)
T=0.63b
78
37
20
9.6
7.1
2.4
110.0
2.3
7.0
2.1
40
61
180
27
31
3300
T=0.65
74
28
20
7.5
6.3
2.0
86
2.0
5.9
1.8
32
53
150
18
26
2400

3.7
19
1.3
17
7.6
13
18
11
12
12
17
9.3
12
28
14
22
T=l.l
68
31
17
8.5
6.8
2.4
110.0
2.2
7.2
2.2
45
67
180.0
31
37
3700
T=1.2
72
30
16
7.0
6.0
2.0
78
2.0
6.3
1.9
34
59
160
23
30
2600

4.1
3.2
1.7
14
8.6
13
21
6.8
10
7.7
19
9.1
9.3
20
15
25
T=2.1
40
25
11
6.3
5.5
2.1
98
1.9
6.6
2.0
44
66
180.0
29
38
3400
T=2.2
49
24
12
6.3
5.3
2.0
86
1.7
5.5
1.8
36
53
160
25
30
2500

15
3.0
8.4
0.40
3.2
4.2
10
8.1
12
6.6
15
15
9.1
10
16
22
T=3.1
28
21
7.2
5.0
4.3
1.7
95
1.5
5.6
1.7
44
60
190
27
41
3200
T=3.2
30
18
8.3
5.2
4.1
1.5
72
1.5
4.6
1.5
30
47
140
20
28
2100

4.7
10
10
1.9
3.4
9.1
19
1.6
14
10
26
17
17
20
27
29
T=4.1
18
15
4.9
4.3
3.6
1.5
79
1.3
4.9
1.7
43
59
190
32
43
2900
T=4.2
19
14
5.4
3.6
3.4
1.3
64
1.2
3.7
1.2
28
40
140
20
24
1800

4.6
6.5
6.5
12
5.1
12
16
7.2
19
23
30
28
19
34
40
33
T=8.7
1.9
3.2
0.71
1.0
1.2
0.6
26
0.52
2.2
0.79
11
18
36
8.6
18
1400
T=8.7
4.2
6.0
1.6
1.9
1.9
0.78
26
0.68
2.6
0.86
12
20
29
8.6
16
1100

53
43
56
43
29
14
0.93
19
13
5.4
6.3
7.1
16
0.17
7.4
17
T=12.7
0.21
0.84
0.10
0.29
0.42
0.25
20.0
0.19
1.0
0.42
11
15
28
8.4
17
1200
T=12.7
0.85
2.98
0.40
0.86
1.0
0.52
22
0.39
1.7
0.63
11
18
27
8.8
15
1000

86
79
84
71
58
49
7.4
48
38
28
0.91
15
3.3
2.9
8.5
13
T=24.7
ND°
ND
ND
ND
ND
ND
5.1
ND
0.058
0.056
13
5.2
36
9.3
25
840
T=24.7
ND
0.13
ND
0.054
0.11
0.090
13
0.068
0.42
0.22
12
12
26
7.7
16
790

-
-
-
-
-
-
60
-
110
85
2.3
57
22
13
29
4
  Tests 11 and 12 on Table 8-1, performed using Glidden 5700-25312 Flat: (Chim Cham). Test 11 - Sample wt.
  1353 g; Test 12 - Sample wt. 14.83 g.
b Sampling times in hours.
c Below the method quantitation limit of 0.05 mg/m3.
                                          8-36

-------
    TABLE 8-19. RESULTS OF INTERCHAMBER VARIABILITY TESTS (TESTS 11 AND 12) FOR VOC
    EMISSIONS FROM ALKYD PAINTS8 - CHAMBER AER CONCENTRATION PER GRAM OF PAINT
Compound
TEST 11
mj>-Xylene
n-Nonane
oOCylene
Propylcydohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
n-Undecane
Pentylcyclohexane !
n-Dodecane
TVOC
TEST 12
m,p-Xylene
n-Nonane
o-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n-Decane
I-Ethyl toluene
1,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
n-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
%RSD
m.p-Xylene
iv-Nonane
o-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1 ,3,5-Trimethylbenzene
ii-Decane
2-Ethyl toluene
1 ,2,4-Trimethylbenzene
l,2,3-7rimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
ivUndecane
Pentylcyclohexane
n-Dodecane
TVOC
Chamber Air Concentrations (mg/m3) per gram of paint
T=0.63b
5.8
2.7
1.5
0.71
0.52
0.18
8.1
0.17
0.52
0.16
3.0
4.5
13
2.0
2.3
240
T=0.65
5.0
1.9
1.4
0.50
0.43
0.14
5.8
0.13
0.40
0.12
2.1
3.6
10
1.2
1.7
160

10
25
5.2
23
14
19
24
18
19
18
23
16
18
34
21
28
T=l.l
5.0
2.3
1.2
0.63
0.50
0.18
7.8
0.16
0.53
0.16
3.3
4.9
14
2.3
2.8
270
7=1.2
4.9
2.0
1.1
0.47
0.40
0.13
5.3
0.13
0.42
0.13
2.3
4.0
11
1.6
2.0
170

2.4
10
8
20
15
19
27
13
16
14
25
16
16
26
22
32
T=2.1
2.9
1.8
0.81
0.47
0.41
0.15
7.2
0.14
0.49
0.15
3.3
4.9
14
2.1
2.8
250
T=2.2
3.3
1.6
0.83
0.43
0.36
0.13
5.8
0.11
0.37
0.12
2.4
3.6
11
1.7
2.1
170

8.7
9.5
1.9
6.1 .
10
11
16
15
19
13
22
22
16
17
23
27
T=3.1
2.1
1.5
0.53
0.37
0.32
0.13
7.0
0.11
0.41
0.13
3.2
4.4
14
2.0
3.0
240
T=3.2
2.0
1.2
0.56
0.35
0.28
0.10
4.9
0.10
0.31
0.10
2.0
3.1
9.8
1.4
1.9
140

1.8
16
3.9
4.6
10
16
25
8
21
16
32
24
23
27
33
37
T=4.1
1.3
1.1
0.37
0.31
0.27
0.11
5.9
0.10
0.36
0.12
3.2
4.4
14
2.4
3.2
220
T=4.2
1.3
0.95
0.37
0.24
0.23
0.08
4.3
0.079
0.25
0.082
1.9
2.7
9.6
1.3
1.6
120

1.9
13
0.067
18.7
12
19
22
14
26
29
36
34
26
40
46
42
T=8.7
0.14
0.24
0.05
0.076
0.092
0.047
2.0
0.038
0.16
0.06
0.81
1.3
2.7
0.64
1.3
100
T=8.7
0.28
0.41
0.11
0.13
0.13
0.05
1.8
0.046
0.17
0.058
0.81
1.3
1.9
0.58
1.1
71

47
37
51
37
22
8
7
13
6
1.1
0.19
0.6
23
6.3
14
24
T=12.7
0.015
0.062
0.007
0.021
0.031
0.019
1.5
0.014
0.072
0.031
0.79
1.1
2.1
0.62
1.3
89
T=12.7
0.057
0.20
0.027
0.058
0.067
0.035
1.5
0.026
0.11
0.04
0.73
1.2
1.8
0.59
1.0
68

82
74
80
66
53
43
0.88
42
32
21
5.6
8.2
10
3.6
15
18
T=24.7
ND*
ND
ND
ND
ND
0.38
ND
0.0043
0.0
0.0041
0.93
0.39
2.7
0.69
1.8
62
T=24.7
0.0051
0.009
ND
0.0037
0.0072
0.0061
0.85
0.0046
0.028
0.015
0.82
0.82
1.8
0.52
1.1
53

-
-
-
-
-
-
55
-
104
81
8.8
51
28
20
36
11
a Tests 11 and 12 on Table 8-1, performed using Glidden 5700-25312 Flafc(Chim Cham). Sample wt -1353 g;
  sample wt - 14.83 g.
b Sampling time in hours.
e Below the method quantitation limit of -0.005 mg/m3 per gram of paint.

                                        8-37

-------
TABLE 8-20.  RESULTS OF INTERCHAMBER VARIABILITY TESTS (TESTS 13 AND 14) FOR VOC EMISSIONS
                    FROM ALKYD PAINTS" - CHAMBER AIR CONCENTRATION
Compound
TEST 13
m,p-Xylene
n_-Nonane
Oj-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1 ,3,5-Trimethylbenzene
n_-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
ivUndecane
Pentylcyclohexane
n-Dodecane -
TVOC
TEST 14
injj-Xylene
n-Nonane
o^-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
n.-Undecane
Pentylcyclohexane ;
n-Dodecane
TVOC
%RSD
mjj-Xylene
ivNonane
o-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1 ,3,5-Trimelhylbenzene
n-Decane
2-Ethyl toluene
1 ,2,4-Trimethylbenzene
1 ,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
n-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
Chamber Air Concentrations
T=0.65b
100
130
31
31
27
9.1
130
8.7
26
7.5
15
25
73
7.1
10
2200
T=0.65
100
120
31
28
24
7.7
140
7.5
23
6.0
14
21
72
7.6
8.6
2100

0.68
4.9
0.54
7.5
8.4
12
6.1
10
10
16
7.6
13
0.84
4.9
11
2.3
T=1.2
110
140
32
32
31
11
150
10
31
8.9
21
30
95
10
13
2700
T=1.2
120
140
37
34
30
10
160
9.5
29
8.0
19
29
94
9.3
12
2400

4.0
1.4
9.1
4.2
2.9
7.8
3.7
3.3
3.5
7.8
9.0
4.0
1.0
3.7
7.5
5.9
T=22
93
120
28
34
32
11
150
10
31
10
22
32
93
11
15
2600
T=2.2
100
140
31
32
31
10
170
9.5
29
8.6
21
30
103
9.3
13
2500

5.8
6.6
6.1
3.6.
3.0
3.8
6.1
4.4
4.6
7.9
6.1
4.6
7.0
8.7
12
2.00
T=3.2
67
120
23
32
30
11
150
10
32
10
26
36
110
11
17
2700
T=3.2
77
120
23
29
27
9
160
9.3
27
8.1
21
29
97
9.1
13
2400

9.2
1.4
0.8
6.9
7.3
9.3
4.7
6.1
11
13
15
14
9.1
16
18
5.9
T=4.2
48
91
16
23
26
8.9
140
9
28
9
24
33
100
11
16
2500
T=4.2
60
110
18
27
25
9
150
8.2
26
7.6
19
28
91
9.4
13
2000

16
13
5
11
1.6
3.3
4.9
3.5
5.8
8.6
18
11
8.5
8.9
15
11
(mg/m3)
T=8.7
9.2
25
4.5
5.8
12
4.3
34
2.5
13
5.4
11
16
24
7.0
13
1100
T=8.7
15
33
6.4
7.9
14
4.6
36
3.1
13
5.5
12
17
26
7.6
13
1090

32
19
25
21
8.6
4.5
3.4
16
3.8
1.5
2.3
2.1
5.6
5.3
4.4
0.46

T=12.7
1.7
11
1.0
3.8
6.9
3.1
34
2.1
10
4.4
12
16
25
8.4
16
1060
T=12.7
4.3
17
2.2
5-0
10
3.8
36
2.7
12
4.9
12
17
28
7.3
15
808

59
32
54
20
25
14
3.1
17
9.7
7.8
0.66
4.0
6.4
9.4
3.4
13

T=24.7
ND*
ND
ND
ND
ND
0.39
18
0.28
2.0
1.2
11
7.2
27
6.6
18
620
T=24.7
ND
0.9
0.062
0.38
1.60
1.0
23
0.83
4.4
2.2
10
10
23
5.6
13
550

-
-
-
-
84
64
16
70
53
38
1.7
22
12
12
22
6.00
   Tests 13 and 14 on Table 8-1, performed using Clidden 8000-46212 Semigloss:(Sea Foam). Test 13 - sample
   wt. 11.19 g; Test 14 - Sample wt. 12.23 g.
   Sampling times in hours.
   Below the method quantitation limit of -0.005 mg/m3 per gram of paint.

                                       8-38

-------
TABLE 8-21. RESULTS OF INTERCHAMBER VARIABILITY TESTS (TESTS 13 AND 14) FOR VOC
EMISSIONS FROM ALKYD PAINTS3 - CHAMBER AIR CONCENTRATION PER GRAM OF PAINT
Compound
Chamber Air Concentrations (mg/m3) per Gram of Paint
TEST 13 T=0.65b
















m^Xylene
iv-Nonane
o-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
ri-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
ivUndecane
Pentylcyclohexane
n-Dodecane
TVOC
TEST 14
•f















rnjj-Xylene
iv-Nonane
S-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1 ,3,5-Trimethylbenzene
iv-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1 ,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
rt-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
9.1
11
2.8
2.8
2.4
0.82
12
0.78
2.4
0.67
1.4
2.2
6.5
0.64
0.89
200.9
T=0.65
8.4
9.7
2.5
2.3
2.0
0.63
12
0.62
1.9
0.49
1.1
1.7
5.9
0.62
0.70
170
7=12 T=2.2
10.0
12
2.9
2.9
2.8
1.0
14
0.89
2.7
0.80
1.9
2.7
8.5
0.87
1.2
240
8.3
11
2.5
3.0
2.8
0.94
14
0.90
2.8
0.86
2.0
2.9
8.4
0.94
1.4
240
7=12 T=2.2
9.7
12
3.0
2.8
2.5
0.79
13
0.77
2.4
0.66
1.5
2.3
7.7
0.76
1.0
200
8.2
11
2.5
2.6
2.5
0.81
14
0.77
2.4
0.70
1.7
2.5
8.4
0.76
1.1
200
T=3.2
6.0
10
2.0
2.8
2.7
1.0
14
0.90
2.8
0.87
2.4
3.2
9.8
1.0
1.5
240
T=3.2
6.3
9.8
1.9
2.4
2.2
0.78
13
0.76
2.2
0.66
1.7
2.4
7.9
0.74
1.1
190
T=4.2
4.3
8.2
1.5
2.1
2.3
0.80
12
0.77
2.5
0.76
2.2
2.9
9.1
0.95
1.5
220
T=4.2
4.9
8.9
1.4
2.2
2.1
0.70
12
0.67
2.1
0.62
1.5
2.3
7.4
0.77
1.1
160
T=8.7
0.82
2.2
0.40
0.52
1.1
0.39
3.0
0.22
1.1
0.48
1.0
1.5
2.2
0.63
1.2
103
T=8,7
1.2
2.7
0.52
0.64
1.1
0.38
2.9
0.25
1.1
0.45
1.0
1.4
2.1
0.62
1.0
89
T=12.7 T=24.7
0.16
1.0
0.09
0.34
0.62
0.28
3.05
0.19
0.90
0.39
1.1
1.4
2.3
0.75
1.4
95
ND0
ND
ND
ND
0.037
0.035
1.6
0.025
0.177
0.11
1.0
0.639
2.5
0.59
1.7
55
T=12.7 T=24.7
0.35
1.4
0.18
0.41
0.80
0.31
2.9
0.22
0.9
0.40
1.0
1.4
2.3
0.60
1.2
66
ND
0.08
0.006
0.034
0.14
0.093
2.1
0.074
0.39
0.19
0.94
0.88
2.1
0.50
1.2
45
%RSD
















a

b
c
mjj-Xylene
iv-Nonane
p^Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
iv-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
5.6
11
6.8
14
15
18
0.23
17
16
22
14
19
7.1
1.4
17
8.6
Tests 13 and 14 on Table 8-1, performed
11.19 g; Sample wt. (12.23 g)
Sampling times in hours.
Below the method quantitation


2.3
4.9
2.8
2.1
9.2
14
2.5
10
10
14
15
10
7.2
10
14
9.1
using


limit of -0.005
0.51
0.29
0.21
10
9.3
10
0.23
11
11
14
12
11
0.68
15
18
8.6
3.0
4.8
5.5
13
14
16
1.6
12
17
19
21
20
12
22
25
12
Glidden 8000-46212


mg/m3 per


gram of
9.4
6.4
1.3
4.6
7.9
10
1.4
10
12
15
24
18
15
15
21
16
26
13
19
15
2.3
1.7
2.9
10
2.5
4.7
4.0
4.2
0.73
1.0
11
7.3
54
26
48
14
18
7.5
3.1
11
3.4
1.5
5.6
2.3
0.12
16
10
18
Semigloss:(Sea Foam). Sample


paint.






NC
NC
NC
NC
80
59
10
65
48
32
8
16
18
18
28
10.0
wt-



                                  8-39

-------
samples. Test 22 was performed without the fan. Air velocities across the paint surface for
this test were less than 2 on/s.  Results for these two tests are given in Table 8-22 for
chamber air concentrations and  in Table 8-23 for chamber air concentrations normalized per
gram of paint Results for Test  21 with the fan show higher air concentrations at the earlier
time points with a more rapid decrease in air concentrations over time when compared to
results for Test 22 without the fan.  In the absence of a fan the least volatile components still
show relatively high chamber air concentration at the end of the 24-hour test period.
8.3.4   SVOC Emissions from Latex Paint Samples
       As with the alkyd paints, a range finding test (Test 1, Table 8-1) was  performed for
tests with a latex paint sample to determine the experimental conditions for  all subsequent
                      \
small chamber tests.  This test was  performed using a Sherwin Williams flat paint
(Marmalade).  Results for this test are given in Tables 8-24 and 8-25 as chamber air
concentrations and chamber air  concentrations normalized per gram of paint.  Results show a
slow increase in  air concentrations over time followed by a gradual decrease.  For most of the
SVOCs, highest chamber air concentrations were measured between 24 and 96 hours after
application.  Very high concentrations (>150 mg/m3) were measured for ethylene glycol.
Based on the results  of this test  several modifications were made in the test procedure.
       •      Smaller samples (0.5 L) were collected in order to keep analyte
              masses within the dynamic range of the test method.
       •      Samples were collected at 1,12, 24, 48, 96,120, and 168 hours to
              more fully characterize the SVOC emissions over time.
All subsequent emissions tests were performed using these modifications.
       Single chamber repeatability and interchamber variability tests were performed to
evaluate the reproducibility of the small chamber emissions test for latex paint samples under
carefully controlled conditions.  For the latex paints, single  chamber repeatability tests were
Tests 3 and 4 and Tests 17 and 18 as  shown in Table 8-1.  Tests 3 and 4 were performed
using a Sherwin Williams flat paint (Marmalade). Tests 17 and 18 were performed using a
Glidden semigloss paint (Down Yonder). Interchamber variability tests are Tests 15 and 16
in Table 8-1. Tests 15 and 16 were performed using a Sherwin Williams gloss paint (Rose
Dawn).  Results for  these six test are presented in Tables 8-26 to 8-31. Data for diethylene
glycol has been included in these tables to provide some information on its behavior during
emissions tests.  However, these data should only be considered estimates since the method

                                         8-40

-------
   TABLE 8-22. RESULTS OF AIR VELOCITY TESTS (TESTS 21 AND 22) ON VOC EMISSIONS
               FROM ALKYD PAINTS3 - CHAMBER AIR CONCENTRATION
Compound
Chamber Air Concentrations (mg/m3)
TEST 21 (FAN)
m.p-Xylene
rt-Nonane
oOCylene
Propylcyclohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n-Decane
I-Ethyl toluene
1,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane '
trans-Decahydronaphthalene
n-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
TEST 22 (NO FAN)
m,j>Xylene
rt-Nonane

-------
TABLE 8-23.  RESULTS OF AIR VELOCITY TESTS (TESTS 21 AND 22) ON VOC EMISSIONS FROM ALKYD
                PAINTS" - CHAMBER AIR CONCENTRATION PER GRAM OF PAINT
Compound
TEST 21 (FAN)
m,j>Xylene
n-Nonane
o-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1,3/5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
rt-Undecane
"'Pentylcyclohexane
n-Dodecane
TVOC
TEST 22 (NO FAN)
mjj-Xylene
jv-Nonane
o-Xylene
Propylcydohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n-Decane
2-Ethyl toluene
1 ,2,4-Trimethylbenzene
1,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
n-Undecane
Pentylcyclohexane
n-Dodecane
TVOC
Chamber Air Concentrations (mg/m3) per Gram of Paint
T=.67b
20 ,
59
3.5
16
1.3
0.42
76
0.39
1.7
0.42
7.1
10
30
2.3
3.8
1100
T=.67
12
28
2.1
6.9
0.53
0.23
29
0.15
0.62
0.15
1.7
3.2
8.0
0.50
0.82
370
7=1.2
17
58
3.5
16
1.2
0.65
75
0.37
1.7
0.46
8.1
11
39
2.4
4.6
1000
T=1.2
13
34
2.4
9.0
0.66
0.32
41
0.17
0.81
0.19
2.5
4.4
11
0.85
1.2
480
T=2.2
9.6
39
2.0
11
1.0
059
80
0.29
1.6
0.43
9.1
11
42
3.2
5.3
900
T=2.2
11
33
2.0
9.2
0.66
0.33
40
0.17
0.89
0.21
3.2
4.9
13
1.0
1.4
470
T=3.2
5.6
28
1.3
8.3
0.78
0.43
70
0.26
1.4
0.38
9.1
10
41
3.6
6.5
800
T=3.2
7.4
27
1.5
7.3
0.57
0.29
38
0.16
0.87
0.21
3.0
4.6
16
1.0
1.7
450
T=4.2
3.3
18
0.76
5.4
0.58
0.35
66
0.19
1.1
0.31
9.7
8.9
52
3.9
7.1
700
T=4.2
5.3
23
1.2
6.2
0.53
0.28
38
0.15
0.78
0.19
3.4
4.8
17
1.2
1.8
410
T=8.2
0.19
1.2
0.065
0.47
0.10
0.080
13
0.041
0.31
0.12
4.6
3.2
16
3.4
8.1
260
T=8.2
0.83
6.2
0.24
2.0
0.27
0.16
14
0.083
0.50
0.14
3.2
3.3
9.7
1.3
2.2
240
T=12.7
ND0
0.066
ND
0.031
0.013
0.013
3.0
ND
0.059
0.034
1.7
0.85
11
1.8
7.8
120
T=12.7
0.12
1.8
0.045
0.70
0.12
0.087
13
0.046
0.31
0.11
3.6
2.9
11
1.7
3.1
190
T=24.7
ND
ND
ND
ND
ND
ND
0.035
ND
ND
ND
0.021
0.0095
0.38
0.064
2.0
10
T=24.7
ND
ND
ND
ND
ND
ND
0.62
ND
0.012
0.0093
0.93
0.31
8.9
1.4
6.1
70
 Tests 21 and 22 on Table 8-1, performed using Sherwin Williams-1435 Gloss:(Bumbershoot). Test 21 Sample
 wt - 7.09 g; Test 22 sample wt. (7.56 g).
 Sampling time in hours.
 Below the method quantitation limit of -0.05 mg/m3 per g of paint.
                                      8-42

-------
   TABLE 8-24. RESULTS OF RANGE FINDING TEST (TEST 1) FOR SVOC EMISSIONS FROM
                   LATEX PAINT3 - CHAMBER AIR CONCENTRATION
Compound
TEST1
Ethylene glycol
1,2-Propanediol
o-Xylene
2-(2-Methoxyethoxy)ethanol
Diethylene glycol
Dipropylene glycol •
2-(2-Butoxyethoxy)ethanol
2-(2-Butoxyethoxy)ethyl acetate
Texanol
TVOC
Chamber Air Concentrations (mg/m3)
T=8.7b
37
3.9
ND0
ND
0.052
0.043
2.5
0.0060
9.3
53
T=24.7
176
13
ND
ND
0.70
' 0.30
5.4
ND
16
210
T=48.7
164
10
ND
ND
0.068
0.19
4.2
ND
11
190
T=72.7
169
9.0
ND
ND
1.7
0.40
- 4.4
ND
12
190
T=96.7
164
7.1
ND
ND
2.2
0.48
4.1
ND
10
190
T=120.7
57
1.9
ND
ND
1.4
0.29
1.6
ND
4.3
66
a Test 1 on Table 8-1 performed using Sherwin Williams 1629 Flat: (Marmalade). Sample Wt. -
b Sampling times in hours.
c Below the method quantitation limit.
                                  8-43

-------
  TABLE 8-25.  RESULTS OF RANGE FINDING TEST (TEST 1) FOR SVOC EMISSIONS FROM
        LATEX PAINT3 - CHAMBER AIR CONCENTRATION PER GRAM OF PAINT
Compound
TEST!
Ethylene glycol
1 ,2-Propanediol
o-Xylene
2-(2-Methoxyethoxy)ethanol
Diethylene glycol
Dipropylene glycol >.
2-(2-Butoxyethoxy)ethanol
2-(2-Butoxyethoxy)ethyl acetate
Texanol
TVOC
Chamber Air Concentration (mg/m3)
T=8.7b
3.3
0.35
ND*
ND
0.0046
0.0038
0.22
0.00053
0.83
4.7
T=24.7
16
1.1
ND
ND
0.062
0.027
0.48
ND
1.4
19
T=48.7
15
0.94
ND
ND
0.0061
0.017
0.38
ND
1.0
17
T=72.7
15
0.80
ND
ND
0.15
0.036
0.40
ND
1.1
17
per Gram
T=96.7
15
0.63
ND
ND
0.20
0.043
0.37
ND
0.93
17
of Paint
T=120.7
5.1
0.17
ND
ND
0.13
0.026
0.14
ND
0.38
5.9
a Test 1 on Table 8-1 performed using Sherwin Williams 1629 Flat:(Marmalade). Sample Wt. -
  Ol.lSg).
b Sampling times in hours.
c Below the method quantitation limit.
                                    8-44

-------
 TABLE 8-26. RESULTS OF SINGLE CHAMBER REPEATABILITY TESTS (TESTS 3 AND 4) FOR
	SVOC EMISSIONS FROM LATEX PAINT8 - CHAMBER AIR CONCENTRATION	
 Compound                             Chamber Air Concentration (mg/m3)


 TESTS                      T=1.2b  T=12.2  T=24.2  T=49.0   T=96.5   T=120   T=168
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol6
TVOC i
ND0
ND
ND
13
ND
13
2.0
9.2
1.5
14
ND
27
8.1
64
2.7
12
ND
87
SL
SL
SL
SL
SL
-
4.2
73
1.8
8.1
0.79
88
2.0
31
0.65
3.9
0.71
38
1.1
29
0.89
5.8
1.6
38
 TEST 4                      T=1.3   T=12.2  T=24.2  T=48.3   T=96.2   T=120   T=170
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
%RSD
1,2-PTopanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
ND
ND
ND
16
ND
16

-
-
-
18
-
18
2.6
12
1.6
14
ND
30

20
20
3.6
2.3
-
7.4
SL
SL
SL
SL
SL
-

-
-
-
-
-
-
7.5
88
2.6
13
0.76
111

-
-
-
-
-
-
3.3
44
1.4
6.8
0.65
56

17
35
17
12
14
31
2.3
40
1.3
5.9
1.1
51

10
20
45
29
30
21
0.72
20
0.77
4.3
1.2
27

27
26
11
22
21
24
 Tests 3 and 4 on Table 8-1, performed using Sherwin Williams 1629 Flat (Marmalade); Test 3
 sample wt (12.74 g); Test 4 - sample wt (11.8 g).
 ^Sampling time in hours.
 °Below the method quantitation limit - see Table 8-10.
 •Cample lost.
 Results reported for information, due to poor method performance results should only be
 considered semiquantitative.
                                     8-45

-------
 TABLE 8-27. RESULTS OF SINGLE CHAMBER REPEATABILITY TESTS (TESTS 3 AND 4) FOR
 SVOC EMISSIONS FROM LATEX PAINT3 - CHAMBER AIR CONCENTRATION PER GRAM OF
                                       PAINT
Compound
TEST3
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol6
TVOC
TEST 4
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
%RSD
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
Chamber Air Concentration (mg/m3) per Gram of Paint
T=1.2b
ND*
ND
ND
0.99
ND
0.99
T=1.3
ND
ND
ND
1.4
ND
1.4

-
-
-
23
-
23
T=12.2
0.15
0.72
0.12
1.1
ND
2.1
T=12
0.22
1.0
0.14
1.2
ND
2.6

25.0
25.4
9.0
3.1
-
15
T=24.2
0.63
5.0
0.21
0.94
ND
' 6.8
T=24
SL
SL
SL
SL
SL
SL

-
-
-
-
-
-
T=49.0
SLd
SL
SL
SL
SL
SL
T=48
0.64
7.5
0.22
1.1
0.065
9.5

-
-
•
-
-
-
T=96.5
0.33
5.7
0.14
0.63
0.062
6.9
T=96
0.28
3.7
0.12
0.58
0.055
4.7

11.5
29.7
11.2
6.9
8.1
27
T=120
0.16
2.4
0.05
0.31
0.056
3.0
T=120
0.20
3.4
0.11
0.50
0.092
4.3

15.4
24.9
50.1
34.3
35
25
T=168
0.084
2.3
0.07
0.46
0.13
3.0
T=170
0.061
1.7
0.065
0.36
0.010
2.2

22.2
20.4
5.1
16.3
16
23
^Tests 3 and 4 on Table 8-1, performed using Sherwin Williams 1629 Flat (Marmalade); Test 3 -
sample wt (12.74 g); Test 4 - sample wt (11.8 g).
bSampling time in hours.
cBelow the method quantitation limit - see Table 8-10.
dSample lost.
Results reported for information, due to poor method performance results should only be
considered semiquantitative.
                                    8-46

-------
  TABLE 8-28. RESULTS OF INTERCHAMBER VARIABILITY TESTS (TESTS 15 AND 16) FOR
       SVOC EMISSIONS FROM LATEX PAINT3 - CHAMBER AIR CONCENTRATION
Compound
TEST 15
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol6
TVOC
TEST 16
. 1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
%RSD
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
Chamber Air Concentration (mg/m3)
T=l hb
ND°
ND
14
2.2
ND
16
T=l h
ND
ND
13
2.1
ND
15

-
-
6.7
3.2
-
4.6
T=12h
ND
30
27
1.9
ND
59
T=12h
0.50
34
28
1.9
ND
64

-
8.0
3.5
2.6
-
5.7
T=24h
ND
74
29
1.6
0.14
100
T=24h
SL
SL
SL
SL
SL
SL

-
-
-
-
-
-
T=48h
SLd
SL
SL
SL
SL
SL
T=48h
ND
150
57
3.3
0.35
210

-
-
-
-
-
-
T=96h
ND
0
23
1.2
ND
24
T=96h
ND
60
23
1.2
ND
84

-
0.95
0.29
2.2
-
79
T=120 h
ND
53
20
1.1
ND
74
T=120 h
ND
59
23
1.3
ND
83

-
8.0
9.3
10
-
7.6
T=168 h
ND
31
16
0.78
0.39
48
T=168 h
ND
32
17
0.81
0.21
50

-
2.4
2.2
2.7
42
2.9
^herwin Williams 200-1604 Gloss:(Rose Dawn); Test 15 - sample wt. (10.79 g); Test 16 - sample
wt. (10.48 g).
^Nominal sampling times.
°Below the method quantitation limit - see Table 8-10.
dSample lost.
eResults  reported for information, due to poor method performance results should only be
considered semiquantitative.
                                   8-47

-------
 Table 8-29. RESULTS OF INTERCHAMBER VARIABILITY TESTS (TESTS 15 AND 16) FOR SVOC
 EMISSIONS FROM LATEX PAINT3 - CHAMBER AIR CONCENTRATION PER GRAM OF PAINT
Compound
TEST 15
1 ,2-propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol6
TVOC
V
TEST 16
1 ,2-propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
%RSD
1 ,2-propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
Chamber Air Concentration (mg/m3) per Gram of Paint
T=lhb
ND°
ND
1.3
0.20
ND
1.5

T=l h
ND
ND
1.2
0.20
ND
1.4 e

-
-
4.6
1.7
-
4.9
T=12h
ND
2.8
2.5
0.17
ND
5.5

T=12h
0.048
3.2
2.7
0.18
ND
6.1

-
10
5.6
4.7
-
7.3
T=24h
ND
6.9
2.7
0.15
0.013
9.8

T=24h
SL
SL
SL
SL
SL
SL

-
-
-
-
-
-
T=48h
SLd
SL
SL
SL
SL
SL

T=48h
ND
14
5.5
0.31
0.033
20

-
-
-
-
-
-
T=96h
ND
5.5
2.1
0.12
ND
7.7

T=96h
ND
5.8
2.2
ND
ND
8.0

-
3.0
1.8
0
-
2.7
T=120 h
ND
4.9
1.9
0.10
ND
6.9

T=120 h
ND
5.6
2.2
ND
ND
7.8

-
10
11
12
-
8.7
T=168 h
ND
2.9
1.5
0.072
0.036
4.5

T=168 h
ND
3.1
1.6
ND
ND
4.7

-
4.5
4.2
5.2
40
3.1
"Sherwin Williams 200-1604 Gloss:(Rose Dawn); Test 15 - sample wt. (10.79 g); Test 16 - sample wt.
(10.48 g).
Nominal sampling times.
°Below the method quantitation limit - see Table 8-10.
dSample lost.
eResults reported for information, due to poor method performance results should only be
considered semiquantitative.
                                    8-48

-------
  TABLE 8-30. RESULTS OF SINGLE CHAMBER REPEATABILITY TESTS (TESTS 17 AND 18) FOR
         SVOC EMISSIONS FROM LATEX PAINT3 - CHAMBER AIR CONCENTRATION
Compound
TEST 17
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
TEST 18
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
%RSD
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
Chamber Air Concentration (mg/m3)
T=lhb
32
6.3
8.8
31
ND
78
T=1.3 h
21
3.9
6.5
25
ND
56

30
33
21
15
-
23
T=13h
180
57
11
17
ND
270
T=12.3 h
180
53
12
19
ND
260

1
5
5
6
-
2.7
T=24h
220
79
15
23
ND
340
T=24.3 h
210
69
12
18
ND
310

4
9
16
19
-
6.8
T=48h
190
81
13
18
ND
300
T=48.3 h
150
61
10
13
ND
230

19
19
22
22
-
19
T=96h
39
39
11
14
ND
100
T=96.3 h
47
40
8.9
11
ND
110

14
3
18
18
-
2.7
T=120 h
3
4.9
6.3
10
ND
19
T=120 h
10
4.9
10
13
ND
33

79
71
32
19
-
38
T=168 h
ND6
ND
2.4
7.4
ND
2.4
T=168.3 h
0.15
ND
5.7
ND
ND
2.5

-
-
3
-
-
2.9
"Glidden 64984 Semigloss:(Down Yonder); Test 17 - sample wt (7.12 g); Test 18 - sample Wt (8.29 g).
Nominal sampling times. Actual times for Test 17 were 1,13, 24, 48, 96,120,168.
 Actual times for Test 18 were 1.3,12.3, 24.3, 48.3, 96.3, 120.3,168.3.
cBelow the method quantitation limit - see Table 8-10.
                                     8-49

-------
  TABLE 8-31. RESULTS OF SINGLE CHAMBER REPEATABILITY TESTS (TESTS 17 AND 18)
  FOR SVOC EMISSIONS FROM LATEX PAINT3 - CHAMBER AIR CONCENTRATION PER
                                GRAM OF PAINT
Compound
TEST 17
1,2-propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC •>
TEST 18
1 ,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
%RSD
1 ,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycol
TVOC
Chamber Air Concentration (mg/m3) per Gram of
T=l hb
4.5
0.88
1.2
4.3
ND
11
T=lh
2.5
0.47
0.79
3.0
ND
6.8

40
43
31
26
-
33
T=13h
26
8.0
1.6
2.4
ND
38
T=12h
22
6.4
1.4
2.3
ND
32

12
16
5.7
4.6
-
12
T=24h
31
11
2.2
3.3
ND
48
T=24h
25
8.4
1.5
2.1
ND
37

14
20
26
29
-
18
T=48h
27
11
1.8
2.5
ND
42
T=48h
18
7.4
1.2
1.5
ND
28

29
30
32
32
-
28
T=96h
5.5
5.4
1.6
1.9
ND
14
T=96h
5.7
4.8
1.1
1.3
ND
13

3.0
8.2
28
29
-
5.2
T=120h
0.41
0.69
0.89
1.4
ND
2.7
T=120 h
1.2
1.8
1.2
1.6
ND
5.8

69
63
22
7.9
-
52
Paint
T=168 h
ND6
ND
0.34
1.0
ND
0.34
T=168 h
0.018
ND
0.28
0.69
ND
0.30

-
-
14
28
-
8.3
aGlidden 64984 Semigloss: (Down Yonder); Test 17 - sample wt (7.12 g); Test 18 - sample wt
(8.29 g).
"Nominal sampling times.
cBelow the method quantitation limit - see Table 8-10.
                                 8-50

-------
performance data for diethylene glycol was poor.  For each test, results are presented both as
chamber air concentrations and chamber air concentrations per gram of paint.  To evaluate
the variability between test methods, %RSD values have been calculated for measured air
concentrations for paired samples from each test. Several observations can be made from the
data in Tables 8-26 to 8-31.
       •     Reproducibility between paired tests both within a single
             chamber and across chambers was generally good (%RSD values
             >30).
       •     Greatest variability between paired samples was generally found
             when air concentrations were low.
                      V
       •     For all paint types, measured air concentration for ethylene
             glycol were high (>60 mg/m3).
   -•"'   •     For the Glidden semi gloss paint (Tests 17 and 18), air
             concentration for 1,2-propane diol were very high (>200
             mg/m3).
       •     Chamber air concentration for target SVOCs gradually increased
             over time with highest concentrations measured at either 24 or
             48 hours. After  that time, concentrations showed a gradual
             decrease.
       •     For the Glidden semigloss paint (Tables 8-30 and 8-31), all target
             SVOCs were at relatively low concentrations at the end of the
             168-hour test period with only 2-(2-butoxyethoxy)ethanol at
             measurable levels.  In contrast, relatively high concentrations of
             ethylene glycol were still present in the chamber air samples for
             the Sherwin Williams flat paint (Tables 8-26 and 8-27) and the
             Sherwin Williams gloss paint (Tables 8-28 and 8-29).
       A final set of comparison tests were performed (Test 19 and 20 on Table 8-1) to
evaluate the effect of air surface velocity on SVOC emissions from  paint samples.  A Sherwin
Williams flat paint (Marmalade) was used  for these tests.  Test 19 was performed using a fan
inside the chamber to generate air velocities of approximately  10 cm/s  across the surface of
the paint samples. Test 20 was performed without the fan. Air velocities across the paint
surface for this test were less than 2 cm/s. Results for these two tests are given in Table 8-32

                                         8-51

-------
     TABLE 8-32. EFFECTS OF AIR VELOCITY ON SVOC EMISSIONS FROM LATEX PAINT
                 (TESTS 19 AND 20)a - CHAMBER AIR CONCENTRATION
Compound
TEST 19
1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycold
TVOC
TEST 20
,/ 1,2-Propanediol
Ethylene glycol
2-(2-Butoxyethoxy)ethanol
Texanol
Diethylene glycold
TVOC
Chamber Air Concentration (mg/m3)
T=1.3b
0.38
ND
0.89
14.6
0.031
18
T=1.3
ND
ND
0.46
11.8
0.033
14
T=12.3
7.7
61
2.7
12
0.12
83
T=12.3
5.1
38
2.2
10
0.093
55
T=24.3
7.1
67
2.5
10
0.24
87
T=24.3
6.2
57
2.2
9.1
0.26
75
T=48.3
3.1
41
1.4
6.4
0.28
52
T=48.3
4.0
44
1.6
6.9
0.35
57
T=96.3
0.57
20
0.75
4.0
1.2
26
T=96.3
1.8
33
1.1
4.9
0.68
41
T=120
ND°
8.2
0.43
2.9
1.1
13
T=120
1.0
27
0.84
4.6
0.7
34
T=168
ND
ND
ND
ND
1.3
4.1
T=168
ND
8.5
0.49
3.3
1.1
14
Tests 19 and 20 on Table 8-1, performed using Sherwin Williams 1629 Flat (Marmalade); Test 19 -
sample wt (9.13 g); Test 20 - sample wt (10.6 g).
Sampling time in hours.
'Below the method quantitation limit.
^Results reported for information, due to poor method performance results should only be
considered semiquantitative.
                                   8-52

-------
for chamber air concentrations and in Table 8-33 for chamber air concentrations normalized
per gram of paint.  Results for Test 19 with the fan showed higher air concentrations at the
earlier time points with a more rapid decrease in air concentrations over time when
compared to results for Test 20 without the fan.
8.3.5  Aldehyde Emissions from Paint Samples
      Based on information provided by the EPA, acetaldehyde and formaldehyde have
been measured in emissions from polyvinyl acetate latex paints (10). Thus in order to
address all of the important emissions components, a method for measuring aldehyde
emissions from paints during small chamber tests was also evaluated.  Testing for aldehyde
emissions was performed at the same time as the emissions testing for VOCs and SVOCs
from alkyd and latex paints.  For Tests 1 to 20 shown in Table 8-1, samples were also
collected for the analysis of aldehydes in chamber air.  Samples from selected chamber tests
were then analyzed based on the results of the range finding tests.  Although only
formaldehyde and acetaldehyde have been  reported in paint emissions, acrolein,
propionaldehyde, and benzaldehyde were added to the list of target analytes since the
analytical method allows  simultaneous determinations of a range of aldehydes. As shown in
Table 8-14, acrolein was not recovered from method controls thus data for this chemical has
not been reported here.
      During the range finding tests for the latex paint (Test 1) and the alkyd paint (Test 2)
screening analysis was performed for the aldehydes target in collected chamber air samples.
Although these analyses did  not allow the quantitation of aldehyde emissions, they did
indicate the  presence of formaldehyde and acetaldehyde in the chamber air samples during
testing.  It also appeared that the emissions from the latex paint sample were higher than the
emissions from the alkyd paint sample.  Based on these results, quantitative analysis of
aldehyde emissions were performed  for all  of the tests for the latex paints and for both sets
of interchamber variabilty tests (Tests 11 to 14) for the alkyd paints.
      Data for aldehyde emissions from the single chamber repeatability tests (Tests 3 and
4, Tests  15 and 16) and the interchamber variability tests (Tests 17 and 18) for the latex paints
are provided in Tables 8-34 to 8-36.  Each table gives measured chamber air concentrations
for samples collected at each time point for the paired tests.  The variability of the measured
concentrations between tests is presented as the %RSD for air samples collected at the same
time point for each of the paired tests. Results of all tests show relatively high emissions for

                                        8-53

-------
TABLE 8-33. EFFECTS OF AIR VELOCITY ON SVOC EMISSIONS FROM LATEX PAINT (TESTS
          19 AND 20)a - CHAMBER AIR CONCENTRATION PER GRAM OF PAINT
Compound
TEST 19
1,2-propanediol
Ethylene glycol
2-(2-butoxyethoxy)ethanol
Texanol
Diethylene glycold
TVOC
TEST 20
-'' 1,2-propanediol
Ethylene glycol
2-(2-butoxyethoxy)ethanol
Texanol
Diethylene glycold
TVOC
Chamber Air Concentration (mg/m3) per Gram of Paint
T=1.3b
0.042
ND
0.10
1.60
0.003
1.9
T=1.3
ND
ND
0.043
1.1
0.003
1.3
T=12.3
0.84
6.6
0.29
1.30
0.013
9.0
T=12.3
0.48
3.6
0.21
0.96
0.088
5.3
T=24.3
0.77
7.2
0.27
1.1
0.026
9.4
T=24.3
0.59
5.4
0.20
0.86
0.024
7.1
T=48.3
0.34
4.5
0.15
0.70
0.031
5.7
T=48.3
0.38
4.1
0.15
0.65
0.033
5.3
T=96.3
0.062
2.2
0.082
0.44
0.13
2.9
T=96.3
0.17
3.1
0.10
0.46
0.064
3.9
T=120
ND0
0.90
0.047
0.32
0.12
1.4
T=120
0.094
2.6
0.080
0.43
0.068
3.3
T=168
ND
ND
ND
ND
0.14
0.44
T=168
ND
0.80
0.046
0.31
0.10
1.3
"Tests 19 and 20 on Table 8-1 performed using Sherwin Williams 1629 Flat (Marmalade); Test 19 -
sample wt (9.13 g); Test 20 - sample wt (10.6 g).
^Sampling time in hours
C3elow the method quantitation limit.
dResults reported for information, due to poor method performance results should only be
considered semiquantitative.
                                   S--54

-------
 TABLE 8-34.  RESULTS OF SINGLE CHAMBER REPEATABILITY TESTS (TESTS 3 AND 4) FOR
    ALDEHYDE EMISSIONS FROM LATEX PAINT8 - CHAMBER AIR CONCENTRATION

Compound                               Chamber Air Concentration (ug/m3)
TEST 3
Formaldehyde
Acetaldehyde
Propionaldehyde
Benzaldehyde
T=1.2b
59
542
NDC
ND
T=12.2
67
26
ND
ND
T=24.2
15
21
ND
INT1
T=49.0
6.5
14
ND
ND
T=96.5
5.0
7.4
ND
ND
T=120
3.9
7.4
ND
ND
T=168
4.4
7.5
ND
ND
TEST 4            ,          T=1.2    T=12.2   T=24.2  T=49.0  T=96.5  T=120   T=168
Formaldehyde
, Acetaldehyde
Propionaldehyde
Benzaldehyde
%RSD
Formaldehyde
Acetaldehyde
Propionaldehyde
Benzaldehyde
64
329
ND
32

5.3
35
.-
-
70
25
ND
ND

3.6
1.1
-
-
10
16
ND
ND

26
18
-
-
4.8
9.0
ND
ND

21
30
-
-
3.7
7.5
ND
ND

21
0.95
-
-
3.2
6.5
ND
ND

14
9.1
-
-
2.4
4.8
ND
2.2

41
31
-
-
"Tests 3 and 4 on Table 8-1, performed using Sherwin Williams 1629 Flat (Marmalade); Test 3 -
sample wt (12.74*g); Test 4 - sample wt (11.8 g).
''Sampling time in hours.
'Below the method quantitation limit - see Table 8-13.
dlnterferent in extract prevented quantitation.
                                 8-55

-------
  TABLE 8-35. RESULTS OF INTERCHAMBER VARIABILITY TESTS (TESTS 15 AND 16) FOR
    ALDEHYDE EMISSIONS FROM LATEX PAINT3 - CHAMBER AIR CONCENTRATION

Compound                              Chamber Air Concentration (pg/m3)
TEST 15
Formaldehyde
Acetaldehyde
Propionaldehyde
Benzaldehyde
T=1.3b
25
439
2.1
23
T=12.3
31
132
ND*1
6.5
T=24.3
NC°
45
ND
5.1
T=48.3
38
23
ND
2.7
T=96.3
19
9.0
ND
ND
T=120
19
7.4
ND
1.8
T=168
14
6.5
ND
1.5
TEST 16           v         T=1.3     T=12.3(   T=24.3  T=48.3   T=96.3  T=120   T=168
Formaldehyde
/ Acetaldehyde
Propionaldehyde
Benzaldehyde
%RSD
Formaldehyde
Acetaldehyde
Propionaldehyde
Benzaldehyde
28
431
ND
21

8.7
1.2
-
7.3
35
129
ND
2.9

8.5
1.9
-
54
31
46
ND
2.7

-
0.6
-
44
33
22
ND
ND

10
3.4
-
-
26
10
ND
ND

21
8.1
-
-
23
8.6
ND
ND

14
11.1
-
-
16
7.4
ND
ND

10
9.1
-
-
aTests 15 and 16 on Table 8-1, performed using Sherwin Williams 1604 Gloss:  (Rose Dawn);
Test 15 sample wt (10.79 g); Test 16 - sample wt (10.48 g).
^Sampling time in hours.
Interference prevented quantitation.
dBelow the method quantitation limit - see Table 8-13.
                                    8-56

-------
TABLE 8-36.  RESULTS OF SINGLE CHAMBER REPEATABILITY TESTS (TESTS 17 AND 18) FOR
    ALDEHYDE EMISSIONS FROM LATEX PAINT3 - CHAMBER AIR CONCENTRATION

Compound                             Chamber Air Concentration (ug/m3)
TEST 17
Formaldehyde
Acetaldehyde
Propionaldehyde
Benzaldehyde
T=1.3b
SLC
SL
SL
SL
T=12.7
22
17
ND
3.2
T=24.3
15
8.8
ND
ND
T=48.3
8.1
4.1
ND
ND
T=96.3
3.5
ND
ND
ND
T=120
ND*1
ND
ND
ND
T=168
ND
ND
ND
ND
TEST 18                    T=1.3    T=12.7  T=24.3  T=48.3   T=96.3  T=120   T=168
Formaldehyde
, Acetaldehyde
Propionaldehyde
Benzaldehyde
%RSD
Formaldehyde
Acetaldehyde
Propionaldehyde
Benzaldehyde
36
127
ND
63

-
-
-
-
25
20
ND
3.5

10
10
-
6.3
17
11
ND
ND

6.6
12
-
-
8.5
4.0
ND
ND

3.4
1.7
-
-
3.5
ND
ND
ND

0.0
-
-
-
ND
ND
ND
ND

-
-
-
-
ND
ND
ND
ND

-
-
-
-
"Tests 17 and 18 on Table 8-1, performed using Glidden-64984 Semigloss: (Down Yonder) Test 17
sample wt (7.12 g); Test 18 sample wt (8.29 g).
Sampling time in hours.
cSample lost.
dBelow the method quantitation limit - see Table 8-13.
                                   8-57

-------
formaldehyde and acetaldehyde. Highest concentrations in all tests were found for
acetaldehyde at the earliest time points. Low concentrations of benzaldehyde were also
measured in samples collected at the early time points during all the emissions tests.
Propionaldehyde was not detected in any of the air samples.  For both the single chamber
repeatability and the interchamber variability test, reprodutibility of measured air
concentrations was generally acceptable (<30% RSD).
       Table 8-37 provides results for Tests  19 and 20 which were designed to evaluate the
effect of air surface velocities on emissions from paint samples.  Results are presented as
chamber air concentrations at each sampling point. Comparison of the air concentration data
between the two tests, shows lower air concentrations for formaldehyde and acetaldehyde
when the fan is used to mix the chamber air. This result shows the opposite trend that was
found for VOX! and SVOC emissions where  higher air concentrations were found at the
earlier time points when the fan was used.
       Data for aldehyde emissions from the interchamber variability tests (Tests 11 and 12,
Tests 13 and 14) for the alkyd paints are provided in Tables 8-38 and 8-39.  Each table gives
measured chamber air concentrations for samples collected at  each time point for the paired
tests. The variability of the measured concentrations between tests is presented as the %RSD
for air samples collected at the same time point for each of the paired tests. Results of both
set of tests show measurable emissions for formaldehyde and  acetaldehyde although at lower
levels than reported from the latex paints. Highest concentrations in all cases are found for
acetaldehyde at the earliest time points. Low concentrations of propionaldehyde were also
measured during emissions testing with the Glidden semigloss paint (Sea Foam).
Benzaldehyde was not detected during emissions testing for the alkyd paints.  For Test 13
and  14 (Table 8-39), reproducibility of measured air concentrations was generally acceptable
(<30% RSD).  In contrast, reproducibility for measured air concentrations between tests 11
and  12 (Table 8-39) was poor; highest %RSD values were calculated for acetaldehyde.  The
reason for the poor reproducibility is unknown and is not consistent with data generated for
VOC air concentrations for the same tests.
8.3.6  Emission Parameters
       Emission parameters for VOCs, SVOCs, and aldehydes in paint samples were
estimated by fitting the chamber air concentration data from emissions testing to specific
source models.  For the aldehydes, emission parameters were  generated only for

                                         8-58

-------
     TABLE 8-37. RESULTS OF THE EFFECTS OF AIR VELOCITY ON THE EMISSIONS OF
	ALDEHYDES FROM LATEX PAINT - CHAMBER AIR CONCENTRATIONS

 Compound                             Chamber Air Concentration (ug/m3)
TEST 19 (FAN)
Formaldehyde
Acetaldehyde
Acrolein
Propionaldehyde
Benzaldehyde
V
TEST 20 (NO FAN)
Formaldehyde
s Acetaldehyde
Acrolein
Propionaldehyde
Benzaldehyde
T=1.3b
25
140
ND
ND
14

T=1.3
29
300
ND
ND
18
T=12.3
14
14
ND
ND
6.3

T=12.3
25
17
ND
ND
3.2
T=24.3
5.7
14
ND
ND
0.90

T=24.3
3.7
13
ND
ND
ND
T=48.3
SLC
SL
SL
SL
SL

T=48.3
2.9
NC
ND
ND
ND
T=96.3
6.8
NC6
ND
ND
ND

T=96.3
2.7
NC
ND
ND
ND
T=120
3.3
NC
ND
ND
ND

T=120
ND
NC
ND
ND
ND
T=168
ND*
NC
ND
ND
ND

T=168
ND
NC
ND
ND
ND
 aTests 19 and 20 on Table 8-1, performed using Sherwin Williams 1629 Flat (Marmalade); Test 19 -
 sample wt (9.13 g); Test 20 - sample wt (10.6 g).
 ^Sampling time in hours.
 Sample lost.
 dBelow the method quantitation limit - see Table 8-13.
 "Interference prevented quantitation.
                                  8-59

-------
   TABLE 8-38. RESULTS OF INTERCHAMBER VARIABILITY TESTS (TESTS 11 AND 12) FOR
     ALDEHYDE EMISSIONS FROM ALKYD PAINT3 - CHAMBER AIR CONCENTRATION
Compound
TEST 11
Formaldehyde
Acetaldehyde
Propionaldehyde
Benzaldehyde
TEST 12
Formaldehyde
Acetaldehyde
.yPropionaldehyde
Benzaldehyde
%RSD
Formaldehyde
Acetaldehyde
Propionaldehyde
Benzaldehyde
Chamber Air Concentration (ug/m3)
T=0.63b
12.6
22.1
ND
ND
, T=0.65
19.6
53.4
ND
ND

31
59
-.
-
T.1.1
8.3
11.8
ND
ND
T=1.2
11.7
26.4
ND
ND

24
54
-
-
T=2.1
5.8
7.6
ND
ND
T=2.2
8.4
23.8
ND
ND

26
73
-
-
T=3.1
3.9
7.1
ND
ND
T=3.2
7.3
28.5
ND
ND

43
85
-
-
T=4.1
2.9
4.4
ND
ND
T=4.2
5.1
20.6
ND
ND

39
92
-
-
T=8.7
N^
5.0
ND
ND
T=8.7
7.3
39.3
ND
ND

-
109
-
-
T=12.7
ND
7.8
ND
ND
T=12.7
8.1
45.7
ND
ND

-
110
-
-
T=24.7
ND
5.7
4.1
ND
T=24.7
4.6
25.1
ND
ND

-
55
-
-
Tests 11 and 12 on Table 8-1, performed using Glidden 5700-25312 Flat: (Chim Cham); Test 11
sample wt (13.53 g); Test 12 - sample wt (14.83 g).
''Sampling time in hours.
°Below the method quantisation limit - see Table 8-13.
                                    8-60

-------
   TABLE 8-39.  RESULTS OF INTERCHAMBER VARIABILITY TESTS (TESTS 13 AND 14) FOR
     ALDEHYDE EMISSIONS FROM ALKYD PAINT3 - CHAMBER AIR CONCENTRATION

Compound                              Chamber Air Concentration (ug/m3)
TEST 13
Formaldehyde
Acetaldehyde
Propionaldehyde
Benzaldehyde
TEST 14
\
Formaldehyde
Acetaldehyde
^Propionaldehyde
Benzaldehyde
%RSD
Formaldehyde
Acetaldehyde
Propionaldehyde
Benzaldehyde
T=.65b
4.7
19.3
7.8
ND
T=.65b

6.2
21.9
8.1
ND

19
8.9
2.7
-
T=1.2
ND6
11.4
5.1
ND
T=1.2

ND
13.7
4.4
ND

-
13
10
-
T=2.2
ND
6.9
5.2
ND
T=2.2

ND
10.3
3.8
ND

-
28
22
-
T=3.2
SLd
SL
SL
SL
T=3.2

ND
9.8
3.3
ND

-
-
-
-
T=4.2
ND
3.9
2.1
3.1
T=4.2

ND
4.4
2.1
ND

-
8.5
0
-
T=8.7
ND
4.3
1.7
2.3
T=8.7

3.4
5.6
1.6
ND

-
18
4.3
-
T=12.7
ND
ND
1.8
2.9
T=12.7

ND
ND
1.9
ND

-
-
3.8
-
T=24.7
5.0
8.4
6.9
ND
T=24.7

3.2
6.1
NC
ND

31
22
-
-
'Tests 13 and 14 on Table 8-1, performed using Glidden 8000-46212 Semigloss; (Sea Foam); Test 13 -
sample wt (11.19 g); Test 14 - sample wt (12.23 g).
^Sampling time in hours.
cBelow the method quantitation limit - see Table 8-13.
•^Sample lost.
                                    8-61

-------
formaldehyde and acetaldehyde emissions from latex paints.  Two models were used, as
determined by the character of the data.  The first is a simple, exponentially decaying source:

                                             e-kt                               (8-10)
where
       t is time after placing the paint sample in the chamber, h;
       S(t) is the source strength at time t, mg/h-g of paint;
       S0 is the initial source strength, mg/h-g of paint; and
       k is the decay constant, h"1.

       This model can be combined with a well-mixed chamber model to predict the
concentration in the chamber as a function of time:
                                      7 C  /„-*<   *-N*\
                               CO) =   °     ~
                                           N - k

where
       C(t) is the analyte concentration at time t, mg/m3;
       N is the air change rate, h"1;
       L is the product loading, g/m3; and
       all other terms are as before.
       This model works well for some paints but not for others. The data indicate that the
chamber concentration at time t = 0 may not be zero, but instead may be almost equal to the
peak concentration measured. This may happen if paint emissions are carried into the
chamber as the plate is inserted.  In such cases, a three-parameter fit to the data is made,
using:
                                         8-62

-------
                                 r c fa~t   a
                                     V-V

where
       CQ is a fitting parameter representing concentration in the chamber at time t = 0,
mg/m3 .
       In fact, this equation was used in preference to Equation 8-11 to allow the fitting
procedure to evaluate the initial chamber concentration, whether or not it appears to be zero.
       The latex paints contained some components that were well-described by this
equation, but others showed a maximum concentration at times as long as 50 hours after
application. Fitting  this data requires a source whose strength builds up with time, as if
there is a slow diffusion of the components to the surface of the paint, from where it is
      zed.
       The concentration equation used for this case is:

                                 LS(l-e
volatilized.
                                           N - k

where
       k2 presents a time constant for emission buildup, h'1.
       This corresponds approximately to a source model given by:
       The model fitting criterion was the minimization of the quantity R:
                                            - c(o)2
                               R _
                                               exp
where
                                                      ,3.
       C(t) is the predicted concentration at time t, mg / m;
       Ce  (t) is the measured concentration at time t; and
       n is the number of data points in the set being fitted.
                                         8-63

-------
       The denominator in Equation 8-15 is chosen for the following reason.  If the power of
CeXp *s 0 (or Cex^ is removed from the denominator), the concentrations of highest
magnitude, near time t = 0, determine the fitting parameters; if the power of Cexp is 2, then
all concentrations affect the fitting parameters equally, especially those of lower precision at
the lowest concentration. Using the power 1 for Cexp gives the higher concentration data
more weight, but does  not completely eliminate the influence of the lower concentration
measurements.
       The overall goodness of fit is described by the relative standard deviation of the fit
(STD), given by:
                                                                               (8-16)
where
       R is defined by Equation 8-15; and
       MAX(CexJ is the maximum concentration measured for the analyte under
consideration.
       STD is not strictly an accurate standard deviation because it does not take into
account the data weighting in Equation 8-15, nor the number of fitting parameters. It is
adequate for ranking the fits and setting a cutoff value for which the fit does not represent
the data very well.
       For the chamber test here, STD values less than 0.10 were considered good and values
between 10 and 20 were considered acceptable.  Values above 0.20 indicate that the model
does not give a very good fit to the measured concentrations; however, the poor fit is often
only at one or two times out of six or more. Moreover, the fits to the aldehyde data give
high STD primarily because the maximum concentrations are much lower than the other
analyte concentrations.
       The fitted models can be integrated analytically over time to compute the total
emission from the paint for each analyte.  For those analytes with nonzero concentrations at
time t = 0,  the mass of vapor in the chamber is also included in the emissions for that
analyte.  In general, the computed emissions agree well with the analyses of the analyte
fractions in the bulk paint, as discussed below.

                                         8-64

-------
       Estimated emission parameters (S^ kv kj, and STD) are given in Table 8-40 for alkyd
paints and in Table 8-41 for the latex paints.  Appendix D gives all of the calculated emission
data plots of the theoretical and measured chamber air concentrations plotted over time for
all target analytes and emission tests.
       When estimating emission parameters, results from paired chamber tests were
combined (i.e., single chamber repeatability and interchamber variability tests for each paint).
A review of the plots in Appendix D shows that in most cases, the differences in measured
air concentration for paired chamber tests were smaller than the differences between the
measured and theoretical air concentration.  When test results are combined, the STD reflects
both the variability between air concentrations measured in paired tests and between the
measured and the theoretical air concentrations.
       The results  in Table 8-40 and 8-41 and the plots in Appendix D show that the
decaying source model, Equation 8-10, represents most of the components in both types of
paint quite well.  The slow buildup model, Equation 8-13, does not fit the latex paint analytes
as well as the decaying source model fits the alkyd analytes, but it does capture the general
pattern of the latex emissions.  As a result, it appears that the alkyd paints produce their
peak concentration about 5 hours after application, but the latex paints  do not peak until
about 50 hours after application.
       For a few analytes, particularly n-dodecane and pentylcyclohexane, the concentrations
appear to  peak at times well beyond five hours, but there are not enough long time data to
define the parameters of the slow buildup model.  For these components, the present models
do not define the concentration behavior very well, even though the STDs are not extreme.
       In addition, many of the alkyd analytes do not agree very well with the model fits at
24 hours after application. Most of the measured 24 hour concentrations are much higher
than the model predicts.  This indicates that there is a slow decay process in addition to the
rather rapidly decaying source determined by the fit.  The impact of the slowly decaying
source on the total emissions is small, but not completely negligible.
       Although the presence of a slowly decaying source can be modeled for the alkyd
paints, the present  data would not define the parameters adequately. Concentrations
measurement would be needed at 16, 30, 36, 42, and 48 hours after application to provide
                                         8-65

-------
          TABLE 8-40.  EMISSION PARAMETERS8 CALCULATED FROM SMALL CHAMBER EMISSIONS TESTS FOR ALKYD PAINTS
00
I
CT>
GLb-Gloss
(Hyacinth)
Tests 5 and 6
Compounds
m,j>-Xylene
n-Nonane
o-Xylene
Propylcyclohexane
3- & 4-Ethyl toluene
1 ,3,5-Trimethylbenzene
ivDecane
2-Ethyl toluene
1,2,4-Trimethylbenzene
1 ,23-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
.n-Undecane
Pentylcydohexane
ivDodecane
TVOC
S0
1.1
3.2
.34
1.1
.55
.19
4.8
.18
.58
.15
.26
.43
.80
.055
.069
60
k
55
.46
.47
.44
.28
.23
.23
.35
.22
.16
.14
.18
.076
.046
-.02
.22
STD
.059
.033
.096
.064
.050
.076
.020
.15
.043
.080
.082
.058
.045
.26
.11
.005
GL-Flat
(Chim Cham)
Tests 11 and 12
S0
.26
.12
' .057
.031
.024
.009
.40
.008
.027
.008
.17
.25
1.0
.11
.12
10
k
.46
.27
38
.25
.20
.18
.15
.17
.16
.13
.13
.13
.23
.11
.039
.070
STD
.026
.084
.045
.19
.15
.35
.062
.30
.19
.34
.14
.079
.061
.16
.13
.014
GLa - Semigloss
(Sea Foam)
Tests 13 and 14
so
.086
1.0
.23
.26
.20
.064
1.2
.068
.18
.050
.11
.17
.74
.049
.070
13
k
.42
.28
.34
.27
.17
.14
.23
.18
.12
.083
.050
.073
.21
.026
.002
.096
STD
.022
.024
.033
.057
.047
.095
.050
.14
.064
.097
.12
.068
.077
.10
.087
.011
SW0 - Gloss
(Bumbershoot)
Test 21d
S0
1.1
5.2
.26
1.4
.091
.058
7.8
.026
.14
.035
.80
.99
3.6
.26
.41
67
k
.69
.62
.66
59
.40
.38
.33
.34
.31
.25
.20
.26
.19
.13
0.042
0.21
STD
.006
.005
.028
.010
.042
.070
.011
.082
.038
.059
.047
.019
.019
.13
.075
.001
Test22e
S0
1.2
3.0
.19
.79
.050
.022
3.2
.012
.070
.015
.21
.36
.84
.61
.08
29
k
.48
31
39
.28
.18
.14
.17
.14
.16
.12
.047
.11
.03
-.01
-.06
.098
STD
.007
.004
.018
.010
.033
.063
.018
.034
.12
.27
.090
.058
.029
.084
.006
.003
      a S0 = initial source strength (mg/h • g of paint).
        k = decay rate (h"1).
        STD = % Relative standard deviation of fix.
      b Glidden.
      c Sherwin Williams.
      d Test with fan.
      e Test without fan.

-------
TABLE 8-41. EMISSION PARAMETERS" CALCULATED FROM SMALL CHAMBER EMISSIONS
                          TESTS FOR LATEX PAINTS
SWbFlat
(Marmalade)
Compound Tests 3 and 4
1,2-Propanediol
S0
kz
STD
Ethylene glycol
ki
kz
STD
2(2-Butoxyethoxy)ethanol
k°
kz
STD
Texanol
S0
k.
kz
STD
Formaldehyde
S0
ki
STD
Acetaldehyde
S0
ki
kz
STD
TVOC
S0
kz
STD
0.51
0.023
0.003
0.27
5.8
0.014
0.0014
0.11
0.40
0.024
0.0019
0.37

0.066
0.0081
191
0.12
5.1E"5
l.OE'2
1.4E'2
0.95
l.OE-4
LIE'2
ME'1
0.368

20
0.019
0.0009
0.069
a S0 = initial source strength (mg/h • g of
k| = decay rate (h"1), ]^ = time constant
STD = Standard Deviation of fit.
b Sherwin Williams.
c Glidden.
d Test with fan.
e Test without fan.


SW-Gloss GLc-Semi-Gloss SW-Flat
(Rose Dawn) (Down Yonder) (Marmalade)
Tests 15 and 16 Tests 17 and 18
0.51
0.023
0.0030
10.3
0.018
0.0020
0.045
4.1
0.022
0.0033
0.094

0.012
0.0062
0.10
0.30
1.8E-4
5.6E'3
0
2.12
8.6E-*
4.4E'2
l.OE'1
0.466

30
0.020
0.0012
0.033
paint).
for emission


80
0.048
0.0024
0.03
45
0.031
0.0009
0.041 .
2.9
0.024
0.0023
0.21

0.137
0.0050
5.9
0.068
2.0E"4
.025
.0058
0.080
2.0E-4
4.4E'2
4.4E'2
0.17

305
.042
.00084
0.023
buildup (h'1).


Test 19s1
3.3
0.054
0.0016
0.15
40
0.035
0.0008
0.069
0.020
0.017
0.035
0.095

0.079
0.013
1.9
.033
3.9E'5
3.9E'3
8.0E'3
1.9
8.2E'5
5.0E-4
4.9E'2
0.001

0.82
0.021
0.12
.026



Test20e
2.1
0.036
0.0015
0.126
24
0.023
0.0007
.069
0.013
0.010
0.23
0.061

0.054
0.008
1.4
0.035
1.9E'5
4.1 E'3
8.5E'3
0.360
LIE"4
2.5E"2
9.5E'2
>0.000

0.51
0.011
0.095
0.046



                                8-67

-------
enough experimental data for good parameter determination.  Although the concentrations
for these compounds were relatively low, chamber air concentrations at these later time
points may fall below the method quantitation limit.
       Table 8-42 gives the estimated mass of each VOC and TVOC emitted per gram of
paint during the small chamber tests.  These masses were estimated as the area under the
theoretical concentration as time curve from t = o to t = infinity. Table 8-43 also gives the
mass of each VOC and TVOC per gram of paint measured during bulk product analysis of
the same paint samples.  The difference between the estimated mass emitted and the mass
measured in the bulk paint samples is presented as the % RSD calculated for the two
measures.  For TVOC, the mass per gram of paint estimated from the ASTM methods is also
given. Similar data are given for the latex paints in Table 8-43.  Results generally show good
agreement between the two measures  suggesting that the emissions data and modeling
parameters are describing the paint emissions adequately.
                                        8-68

-------
    TABLE 8-42. COMPARISON OF DATA FOR CHAMBER EMISSIONS TESTS TO RESULTS FOR BULK PRODUCT ANALYSIS FOR ALKYD PAINTS
00
GLa-Gloss
(Hyacinth)
Tests 5 and 6
Compounds
m,£-Xylene
ji-Nonane
o-Xylene
Propylcyclohexane
3- & 4-Ethyl toluene
1,3,5-Trimethylbenzene
n,-Decane
2-Ethyl toluene
1 ,2,4-Trimethylbenzene
1 ,2,3-Trimethylbenzene
2-Methyldecane
trans-Decahydronaphthalene
ii-Undecane
Pentylcyclohexane
ii-Dodecane
TVOC

Mec
(mg/g)
2.3
6.9
0.76
2.6
1.9
0.82
21
0.51
2.6
0.9
1.9
2.4
10
1.2
>1.7
280

(mg/g)
3.6
9.8
1.1
4.2
1.8
0.79
18
0.62
2.8
0.84
2.0
2.1
9.1
0.52
2.6
280
380*
%RSD
31
25
26
33
3.8
2.8
11
13
4.2
7
3.7
8.5
7.3
56
-
0
21
GL - Flat
(Chim Cham)
Tests 11 and 12
Me
(mg/g)
1.3
0.90
0.41
0.26
0.26
0.10
5.0
0.090
0.31
0.12
2.2
3.5
5.5
1.7
5.9
310

(mg/g)
1.6
0.68
0.37
0.20
ND*
ND
4.5
ND
ND
ND
3.2
3.2
19
3.0
12
180
299*
%RSD
13
20
8
18
-
-
7
-
-
-
2.7
5
78
39
46
38
56
GL - Semigloss
(Sea Foam)
Tests 13 and 14
Me
(mg/g)
2.7
4.6
0.98
1.1
1.5
0.61
5.9
0.42
2.0
0.81
3.3
3.1
3.6
2.7
>1.7
220

(mg/g)
4.0
5.3
1.1
1.5
1.3
0.52
14
0.48
1.8
0.64
2.6
2.6
16
2.1
8.1
210
330*
%RSD
27
10
7
19
10
11
58
10
7
17.1
16
13
91
19
-
3.2
28
SW1' - Gloss
(Bumbershoot)

Me
(mg/g)
3.4
13
0.65
3.0
0.39
0.15
25
0.13
0.46
0.19
3.9
4.2
19
2.0
9.7
560

Test 21
Cb
(mg/g)
4.3
11
0.83
3.4
ND
ND
19
ND
0.53
ND
3.2
3.5
16
2.2
7.9
270
540*
Test 22
%RSD
16
9
17
9
-
-
18
-
10
-
13.1
13.7
10
5.4
15
49
2.6
Me
(mg/g)
3.1
11
0.65
3.0
0.34
0.18
20
0.12
0.44
0.13
4.3
3.3
28
>1.5
>1.9
430

Cb
(mg/g)
4.3
11.2
0.83
3.4
ND
ND
19
ND
0.53
ND
3.2
3.5
16
2.2
7.9
270
540*
%RSD
23
2.6
17
9
-
-
25
-
14
-
21
4.0
38
-
-
32
16
   * Glidden.
   b Sherwin Williams.
   c Estimated mass per gram of paint during chamber tests.
   d Measured concentration measured during bulk product analysis.
   e- Below the method quantitation limit.
   ' Estimated concentration from ASTM methods.

-------
    TABLE &43.  COMPARISON OF DATA FOR CHAMBER EMISSIONS TESTS TO RESULTS FOR BULK PRODUCT ANALYSIS FOR LATEX PAINTS
CO
i
SWa-Flat
(Marmalade)
Tests 3 and 4
Compounds
1,2-Propanediol
Ethylene glycol
2-<2-Butoxyethoxy)ethanol
Texanol
TVOC

Formaldehyde
Acetaldehyde
Mec
(mg/g)
2.6
37
U
14
49

0.08
0.53
cbd
(mg/g)
ND
29
1.5
5.1
36
65e
NMf
NM
%RSD
-
17
18
66
22
20
-
-
SW - Gloss
(Rose Dawn) ->
Tests 15 and 16
Me
(mg/g)
NC
60
25
2.7
88

0.03
0.14
(mg/g)
,ND
48
' 13
-27-'- 1
65
84e
NM
NM
%RSD
-
15
46
1
21
3
-
-
GLb - Semigloss
*..'• -f^Ur (6oa Foam)
Tests 17 and 18
Me
(mg/g)
79
45
11
89
140

0.02
0.06
Cb
(mg/g)
38
19
4.4
5.7
68
93e
NM
NM
7oRSD
49
57
60
124
49
31
-
-
SW - Flat
(Marmalade)

Me
(mg/g)
1.9
25
1.1
14
34

0.12
NC»
Test 19
(mg/g)
ND
29
1.5
5.1
36
65e
NM
NM
Test 20
7.RSD
-
10
24
66
3
44
-
-
Me
(mg/g)
2.1
30
1.3
16
41

0.11
0.21
q,
(mg/g)
ND
29
1.5
5.1
36
65e
NM
NM
7.RSD
-
2
12
73
10
32
-
-
* Sherwin Williams.
b Glidden.
c Estimated mass emitted per gram of paint during chamber tests.
d Concentration measured during bulk product analysis.
e Estimated concentration from ASTM methods (Table 6-1).
' Not measured.
8 Not calculated - curve did not adequately describe the data.

-------
                                    SECTION 9.0
                                METALS ANALYSIS

9.1    STUDY DESIGN
      Two methods were evaluated for the analysis of metals in alkyd and latex paints. The
target list of metals provided by EPA is given in Table 9-1.  The first method used X-ray
fluorescence spectroscopy (XRF) on untreated liquid paint samples. The second method used
inductively coupled plasma emission spectroscopy (ICP) on digested paint samples.  The XRF
method was performed on single aliquots of each of the 20 paint samples shown in Table 9-2
to provide data on metal concentrations in each sample. The ICP method was performed on
triplicate aliquots (from the same vials) for each of the 20 paint samples to provide data on
metal concentrations and method precision.  For the ICP method, additional QC samples
including method blanks (unspiked reagents), method controls (spiked reagents), and matrix
spikes (spiked paint samples) were prepared, digested, and analyzed to provide additional
information on method performance.

9.2    METHOD
9.2.1  ICP Method
      9.2.1.1  Sample Preparation
      Homogenized paint samples (0.25 g) were placed in an acid washed Teflon digestion
vessel (CEM Corp, Mathews, NC).  Concentrated nitric acid (15 mL) (Baker Instranalyzed)
and 2 mL of AES hydrofluoric acid (Baker Instranalyzed) were added.  The vessels were
capped according to the manufacturers directions and placed in a CEM MDS810 microwave
digestion system and heated according to the following program:
      •      10 minutes at 600 Watts (100% power)
      •      8 minutes at 480 watts
      •      10 minutes at 600 watts
      All 12  positions in the MDS81D sample carousel were filled with sample vessels
containing an equivalent volume of sample mixture or water in order to distribute the
microwave energy evenly.
      After the heating program was completed, the sample  vessels were allowed to cool to
room temperature.  The digestion residues were then dissolved by adding 50 mL of

                                        9-1

-------
       TABLE 9-1. TARGET METALS FOR LIQUID PAINTS BY XRF AND ICP

Aluminum                Manganese                Tin
Chromium                Cobalt                    Barium
Molybdenum              Nickel                    Mercury
Cadmium                 Copper                   Arsenic
Antimony                 Selenium
Lead                     Strontium
                                    9-2

-------
TABLE 9-2. PAINT SAMPLES FOR METAL ANALYSIS
Paint Type
Gloss Type
Manufacturer
Series
Color
Group
Manufacturer's
ID No.
Color Name
SHERWIN WILLIAMS
(1) Alkyd
(2) Alkyd
(3) Alkyd
(4) Latex
(5) Latex
(6) Latex
(7) Latex
(8) Latex
(9) Alkyd
(10) Alkyd
GLIDDEN
(1) Alkyd
(2) Alkyd
(3) Alkyd
(4) Latex
(5) Latex
(6) Latex
(7) Latex
(8) Latex
(9) Alkyd
(10) Alkyd
Flat
Semi-gloss
Gloss
Flat
Semi-gloss
Gloss
Flat
Semi-gloss
Flat
Semi-gloss

Flat
Semi-gloss
Gloss
Flat
Semi-gloss
Gloss
Semi-gloss
Gloss
Flat
Gloss
ProMar200
ProMar200
ProMar200
ProMar 200
ProMar 200
ProMar 200
ProMar 200
ProMar 200
ProMar 200
ProMar 200

5700
UH8000
4550
3480
UH6380
6918
UH6300
6987
5718
4550
Yellow
Blue
Green
Orange
Purple
Red
Green
Other
Other
Other

Yellow
Green
Purple
Red
Blue
Orange
Blue
Orange
Green
Other
SW1352
SW1529
SW1435
SW1629
SW1545
SW1604
SW1734
SW1125
SW1003
SW1309

25312
46212
76262
01044
64984
16112
64542
20573
34722
20852
Crescent Cream
Violet Veil
Bumbershoot
Marmalade
Vibrant Violet
Rose Dawn
Grass Roots
Praline
First Star
Coral Canyon

Chim Cham
Sea foam
Hyacinth
Tomahawk
Down Yonder
Orange Glaze
Ice Cap
Orange Ice
Antigua
Sheriffs Star
                   9-3

-------
laboratory pure water (ASTM Type II) and the vessels recapped and microwaved for an
additional 20 minutes at 300 watts (50% power). The resulting digest was then transferred to
a 100 mL volumetric flask and diluted to volume with laboratory pure water.  Matrix spikes,
method blanks and method controls were similarly prepared, each at a 5% frequency.
       9.2.1.2 Sample Measurement   •
       The aqueous digests were analyzed using a Leeman Labs (Lowell, MA) Plasmaspec I
sequential inductively coupled plasma emission spectrometer (ICP).  Calibration is performed
quarterly using a four or five point calibration curve. Calibration updates are performed
using a blank and a mid level standard every 7-10 samples.  The accuracy of the
calibration /update standards is verified daily through the analysis of calibration check
standards prepared or purchased from a different source than those of the calibration/update
standards.  This analysis should be within 5% of the expected value or the calibration is
repeated.  Update sample analyses should be within 10% of expected value or samples
analyzed since the last successful update are reanalyzed.
9.2.2   XRF Analyses
       XRF screening of paint samples was performed by Dr. T. M. Spittler of EPA Region 2.
A description of the method used is provided in Appendix E.

9.3    RESULTS
       Results for the analysis of paint samples by both ICP and XRF method are
summarized for the alkyd paints in Table 9-3 and for the latex paints in Table 9-4. Each table
provides information on measured concentrations of target metals in paint samples.  For the
ICP method where triplicate aliquots were measured, concentration data are provided as the
mean and the %RSD of the triplicate determinations. For the XRF method, only a single
aliquot was analyzed for each paint.  Several additional elements were analyzed by the XRF
method. For the paint samples analyzed by ICP, results with high variability (%RSD > 30)
are also highlighted on the tables. Generally, precision of the ICP test method was
acceptable. For some samples the measured %RSD values for all metals with measurable
concentrations were high (>30%).  It is suspected that this was a result of using sample
aliquots that were poorly mixed. In other samples, unacceptable precision was
                                          9-4

-------
                                TABLE 9-3. RESULTS OF METALS ANALYSIS FOF ALKYD PAINTS
ID
cn
Mean Measured Concentration in ug/g (%RSD)a
Flat
Crescent Green
Metal
SHERWIN WILLIAMS
Aluminum
Selenium
Barium
Antimony
Cobalt
Cadmium
Arsenic
Chromium
Copper
Strontium
Lead
Manganese
Molybdenum
Nickel
Mercury
Tin
Zinc
Bismuth
Calcium
Titanium
Iron
ICPb

33,000(53)
<60e
17**(53)
<40
190(59)
<1.0
39(13)
26**(59)
<2.0
220*"(58)
35*
14**(52)
<5.0
50*
86(52)
119*
NA
NA
NA
NA
NA
XRF=
NAd
<2
<50
<5
<20
<5
<20
10
6
NA
10
<5
<5
7
<10
<25
9
<5
10.5%
14.0%
2500
Flat
First Star
ICP XRF3
25,000(9.0) NA
<60 <2
11(14) <50
<40 <5
160**(10) <20
<1.0 <5
<30 <20
18**(14) <5
<2.0 9
140**(12) NA
26* <10
37(25) 15
<5.0 <5
<15 10
76~(3.8) <10
<75 <25
NA 6
NA <5
NA 10.5%
NA 14.6%
NA 540
Semigloss
Coral Canyon
ICP
4,200(9.0)
<60
0.68*
<40
97"(82)
<1.0
37*
8.3(28)
31*
36**(39)
18" (6.7)
6.5(4.2)
<15
64*
53**(39)
<75
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
<20
<5
<20
<5
9
NA
30
<5
<5
5
<10
<25
6
<5
5.0%
29.8%
3800
Semigloss
Violet Veil
ICP
4,000(9.0)
" <60
0.7*
55
16002)
<1.0
<30
4.9(6.4)
<2.0
45(12)
21(22)
8.8(29)
<5.0
<15
83(44)
<75
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
<20
<5
<20
<5
10
NA
33
<5
<5
5
<10
<25
5
<5
4.6%
28.5%
470
Gloss
Bumbershoot
ICP
5,50009)
113*
2.5 (57)
59(19)
170(15)
<1.0
<30
4.7(24)
<10
5.4(75)
<15
3.9(24)
<5.0
<15
62(42)
112*
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
<20
<5
<20
<5
24
NA
<10
<5
• <5
<15
<10
<25
<1
<5
<0.1%
40.0%
500

-------
                     TABLE 9-3. RESULTS OF METALS ANALYSIS FOR ALKYD PAINTS (CONTINUED)
cr>
Mean Measured
Flat
Tomahawk
Metal
GLIDDEN
Aluminum
Selenium
Barium
Antimony
Cobalt
Cadmium
Arsenic
Chromium
Copper
Strontium
Lead
Manganese
Molybdenum
Nickel
Mercury
Tin
Zinc
Bismuth
Calcium
Titanium
Iron
ICP
11,000(26)
<60
13(27)
<40
<30
7.3(25)
<2
190(27)
14(26)
<5
<15
<30
<75
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
<20
<5
<20
13
7
NA
<5
<5
<5
4
<10
<25
4
<5
5.4%
1.2%
6800
Semigloss
Down Yonder
ICP
25,000(9)
<60
14(11)
<40
16(14)
<30
15(13)
170(7.2)
23(3.9)
2.2(6.4)
<5.0
<15
<30
<75
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
<20
<5
<20
7
172
NA
9
<5
<5
8
10
<25
8
<5
<0.1%
1.5%
640
Concentration in ug/g (%RSD)a
Semigloss
Ice Cap
ICP
13,000(15)
<60
7.3(22)
<40
105(19)
46**(14)
4.3(8.3)
<2.0
9.9(13)
2.6(5.7)
<5.0
<15
33(9.2)
<75
NA
NA
NA
XRF
NA
<2
<50
<5
<20
<5
<20
<5
34
NA
<5
<5
5
<10
<25
4
<5
=0.1%
NA 16.1%
NA
1060
Gloss
Orange Glaze
ICP

1600(43)
70*
<0.5
<40
52(101)
<1.0 4
<30
4.6(39)
<2.0
1.3~(22)
3.4(31)
<5.0
<15
<30
<75
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
<20
» <5
<20
<5
8
NA
<5
<5
5
<10
<25
7
<5
<0.1%
34.0%
6840
Gloss
Orange Ice
ICP

3600(30)
<60
14(35)
<40
64(30)
<1.0
33~(12)
4.9"(15)
<2.0
2.7~(31)
2.5"(2.3)
<5
<15
31*
<75
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
<20
<5
<20
<5
7
NA
12
<5
5
<10
<25
5
<5
^f\ 1 (ff*
^v» A fv
15.8
308

-------
 FOOTNOTES:
                                                                                 V.
 a  Numbers with squares around them indicate where RSD of ICP analysis was greater than 30%.
     A value above the QL was found in only one of three replicate samples; % RSD value was not calculated.
   ** Values above the QL were found in two of the three replicate samples. Mean and % RSD calculated for only samples above the QL.
 b  Triplicate sample analyzed.
 c  Only single replicate analyzed by XRF.
 d  Not analyzed by method.
 e  Less than the quantitation limit (QL) as indicated.
UD

-------
                         TABLE 9-4. RESULTS OF METALS ANALYSIS FOR LATEX PAINTS
I
00
Mean Measured Concentration in ug/g (%RSD)a
Flat
Marmalade
Metal
ICPb
SHERWIN WILLIAMS
Aluminum 12,000(4)
Selenium <60C
Barium
Antimony
Cobalt
Cadmium
Arsenic
Chromium
Copper
Strontium
Lead
Manganese
Molydenum
Nickel
Mercury
Tin
Zinc
Bismuth
Calcium
Titanium
Iron
6.7(18)
<40
51(13)
<1.0
<30
7.5(44)
<2.0
10(16)
<15
5.4(10)
<5.0
<15
<30
<75
NA
NA
NA
NA
NA
XRF*
NAd
<2
<50
<5
<20
<5
<20
9
5
NA
<5
<5
<5
7
<10
<25
34
<5
<0.1%
10%
760
Flat
Grass Root
ICP
13,000(15)
-<60
15(75)
<40
55(14)
<1.0
<30
10(14)
570(17)
13(17)
29*
7.4(28)
<5.0
<15
72(64)
<75
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
<20
<5.0
<20
<5.0
440
NA
<5
<5
<5
<1
<10
<25
<1
<5
<0.1%
6.8%
2300
Semigloss
Vibrant Violet
ICP
4600(5)
<60
3.505)
<40
55(3.2)
<1.0
<30
<2.0
12(58)
29(7.0)
<15
3.2(26)
<5.0
<15
35(3.4)
<75
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
<20
<5
<20
<5
49
NA
12
7
<5
7
<10
<25
69
<5
1.9%
9.6%
500
Semigloss Gloss
Praline Rose Dawn
ICP

1500(35)
<60
1.5(7.3)
<40
50(22)
<1.0
<30
3.4(16)

-------
                      TABLE 9-4.  RESULTS OF METALS ANALYSIS FOR LATEX PAINTS (CONTINUED)
IO
10
Mean Measured Concentration in ug/g (%RSD)
Flat
Antigua
Metal ICP XRF
GLIDDEN
Aluminum 12,000(12) NA
Selenium <60 <2
Barium 30(5.8) <50
Antimony <40 <5
Cobalt 170(8.8) <20
Cadmium <1.0 <5
Arsenic 47* <20
Chromium 11(12) <5
Copper <2.0 15
Strontium 290(7.0) NA
Lead 23* 12
Manganese 18(7.9) 15
Molybdenum <5.0 <5
Nickel <15 7
Mercury 80(9.1) <10
Tin 147* <25
Zinc NA 8
Bismuth NA <5
Calcium NA 15.5%
Titanium NA 15.0%
Iron NA 630
Flat
ChimCham
ICP
34,000(5)
<60
20(16)
<40
170(4.3)
<1.0
52~(10)
24(7.2)
<2.0
190(17)
30(55)
16(41)
<5.0
<15
7.7(14)
<75
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
<20
<5
<20
15
8
NA
10
12
<5
7
<10
<25
7
<5
10.0%
11.2%
1500
Semigloss
Sea Foam
ICP
6,200(10)
<60
1.4(36)
<40
300(21)
<1.0
<30
3.8(19)
<10
210(23)
<15
<2.0
20"(25)
<15
<30
<75
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
<20
<5
<20
<5
5
NA
<10
<5
<5
<1
<10
<25
2680
<5
8.7%
26.0%
950
Gloss
Sheriffs Star
ICP
8,200(13)
<60
1.7(29)
40*
370(17)
<1.0
39*
9.1(15)
<10
3.2(51)
19*
8.5(114)
<5.0
<15
63(18)
<75
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
<20
<5
<20
<5
6
NA
30
15
<5
<1
<10
<25
5
10
<0.1%
50.0%
3050
Gloss
Hyacinth
ICP

12,000(43)
<60
7.3(40)
<40
460(38)
<1.0
<30
8.0(56)
33(26)
4.0(43)
<15
<2.0
<5.0
<15
56"(52)
<75
NA
NA
NA
NA
NA
XRF
NA
<2
<50
<5
125
<5
<20
<5
66
NA
21
<5
<5
<1
<10
<25
5
<5
<0.1%
45.0%
150

-------
 FOOTNOTES:
                                                                                 \
 a Numbers with squares around them indicate where RSD of ICP analysis was greater than 30%.
   *  A value above the QL was found in only one of three replicate samples; % RSD value was not calculated.
   ** Values above the QL were found in two of the three replicate samples. Mean and % RSD calculated for only samples above the QL.
 b Triplicate sample analyzed.
 c Only single replicate analyzed by XRF.
 d Not analyzed by method.
 e Less than the quantitation limit (QL) as indicated.
I
t—•
o

-------
seen for only a few of the metals. This is most likely due to problems associated with the
digestion that did not adequately remove all samples prior to analysis interferences.  A
comparison between the results for analysis by the ICP and XRF methods shows that
measured concentrations often differed by greater than a factor of two.
      Table 9-5 summarizes information in Tables 9-3 and 9-4 and gives the percentage of
samples with measurable concentration of metals. Percent measurable values are provided
by manufacturer and paint type as well as for all paint samples.
      Method performance results for the ICP method are provided in Table 9-6.  Generally,
method  blanks were uncontaminated. Method controls showed acceptable recovery (>80%)
and reproducibility (%RSD <15). Exceptions to this were strontium and molybdenum which
                      v
give high %RSD values. Spiked paint samples generally showed good recovery. The single
exception was a very low recovery for aluminum in one of the spiked samples. However,
the analyses of the three unspiked samples -of that paint gave very high and variable
concentrations for aluminum.  Under these conditions, it was felt that meaningful recovery
data could not be determined.
                                        9-11

-------
             TABLE 9-5. PERCENTAGE OF PAINT SAMPLES WITH MEASURABLE CONCENTRATIONS OF METALS
I
I—'
ro
% Measurable
Latex
Quantitative
Limit (pg/g)

Aluminum
Selenium
Barium
Antimony
Cobalt
Cadmium
Arsenic
Chromium
Copper
Strontium
Lead
Manganese
Molybdenum
Nickel
Mercury
Tin
Zinc
Bismuth
Calcium
Titanium
Iron
ICP
50
60
0.5
40
10
1.0
30
2.0
2.0
1.0
15
2.0
5.0
15
30
75
—
—
. —
—
—
XRF
_a
2
50
5
20
5
20
5
2
—
10
5
5
15
10
25
1
5
0.1%
NRb
NRb
Sherwin
Williams
ICP
100
0
80
0
100
0
0
60
40
80
20
80
0
0
80
0
—
—
— •
—
—
XRF
_a
0'
—
0
0
0
0
40
100
—
40
20
0
80
0
0
80
0
100
100
100
Glidden
ICP
100
20
80
20
100
0
60
100
20
100
80
60
0
0
60
0
—
—
—
—
—
XRF
—
0
0
0
20
0
0
20
100
—
80
60
0
40
0
0
100
0
100
100
100
Alkyd
Sherwin
Williams
ICP
100
20
100
40
100
0
40
100
20
100
80
80
0
40
40
20
—
—
—
—
—
XRF
—
0
0
0
0
0
0
20
100
—
60
20
0
20
80
0
100
0
80
100
100
Glidden
ICP
100
0
80
0
80
0
40
100
20
100
0
100
0
0
40
0
—
' —
—
—
—
XRF
—
0
0
0
0
0
0
40
100
—
20
20
100
100
20
0
100
0
20
100
100
ALL
ICP
100
0
90
15
95
0
35
90
25
95
45
80
0
10
55
5
—
—
—
—
—
XRF
—
0
0
0
5
0
0
30
100
—
50
30
25
60
25
0
95
0
75
100
100
     a Not analyzed by test method.
     b Not reported.

-------
                                        TABLE 9-6.  METHOD PERFORMANCE RESULTS FOR ICP ANALYSIS
UD
 I
Metal
Aluminum
Selenium
Barium
Antimony
Cobalt
Cadmium
Arsenic
Chromium
Copper
Strontium
Lead
Manganese
Molybdenum
Nickel
Mercury
Tin
a
Calculated
Method Blank
(ug/Samples ± S.D.)
n = 3
0
0
8.8 ± 12
0
0
0
0
0.19 ± 0.87
0
0.58 ± 0.43
22 ±30
0.04 ± 0.35
0
0
0
0.10 ± 1.2
Measured Spiked (ug) - Measured
Method Control
Spiked
Amount ug '
2000
210
201
200
200
201
210
203
210
200
1003
201
200
1002
202
200
Unspiked (ug)
To Recovery*
86
105
94
98
105
101
95
107
97
128
108
112
98
89
97
102
x 100%
%RSD
2
5
13
3
4
3
3
10
4
49
4
41
2
5
3
4

V Spiked
Amount (ug)
8295
871
834
830
830
834
871
842
834
830
4160
834
830
4156
838
830

Spiked Samples
%Rec
86
86
83
85
87
96
87
89
80
83
94
96
83
90
86
77

Spiked
Amount (ug)
7994
840
803
799
799
803
840
811
803
799
4009
803
799
4005
807
799

%Rec
NRb
93
86
96
75
88
89
85
85
64
92
91
89
84
94
85

                                 Amount Spiked
        b  All reported values less than zero.
        c  Not reported levels in unspiked triplicate samples were high and variable.

-------
                                   SECTION 10.0
                     QUALITY ASSURANCEXQUALITY CONTROL

10.1   OVERVIEW
      Quality control (QC) and quality assurance (QA) activities were an integral part of
this research program. The work was carried out following the guidelines and procedures
detailed in the Work Assignment Revised Work Plan (10) and Revised Quality Assurance
Project Plan (QAPP)Ol).  Much of the research conducted was a follow-up to previous work
conducted on coatings (1).  Quality assurance objectives, as outlined in the QAPP, are shown
in Table 10-1.
                     \
      Quality assurance activities that were conducted in support of this study included:
      •   Preparation of a Quality Assurance Project Plan,
   -••  •   Meetings with work assignment staff on matters affecting data quality, and
      •   Systems audits of major study components.
      Quality control samples (blanks, spiked  controls, replicates) were analyzed as a part of
this study. In addition, quality control procedures were included in the sampling and
analysis phases of this study.

10.2   QUALITY ASSURANCE PROJECT PLAN
      A (revised) QAPP(ll) was prepared for this work assignment and covered all aspects
of this work.  The approved QAPP was used as a guide throughout the study to monitor QC
procedures and adherence to study objectives.
      Established test methods were utilized for sampling and analysis as summarized in
Table 10-2.  These procedures were supplemented with RTI standard operating procedures.

10.3   QUALITY CONTROL SAMPLES
10.3.1 Blanks
      Blanks were prepared and analyzed along with  samples to provide a measure of
background contamination  associated with sampling, handling and analysis.  Banks were
planned for each method as shown in Table 10-3. The table also includes completion rate
and a cross-reference for results.
                                        10-1

-------
       TABLE 10-1. SUMMARY OF QUALITY ASSURANCE OBJECTIVES
Precision
Parameter (%RSD)
Volatiles Content
Total Volatiles
Total Water
Bulk Product Analysis
VOC /SVOCb
TVOC
Metals (ICP)
Metals (XRF)
Small Chamber Emissions Testing
voc/svocb
TVOC
Aldehydes6
Emission Factors
mg/m2»h
mg/g»h

SI .5
S10

<20
£25
<20
<20
NS
NS
NS

NS
NS
Accuracy
(%Recovery)

ND*
ND

80-100
75-125
80-120
80-120
NS
NS
NS

NAe
NA
Completeness
(%)

295
295

295
295
295
295
295
295
295

295
295
a Bias has not been determined for this method.
b For each of the 8 most abundant compounds identified.
c For five target analytes.
d Specific objectives were not set.
e No accuracy assessments will be made.
                                   10-2

-------
               TABLE 10-2.  SUMMARY OF TEST METHODS
Method Description     Method Reference             Analysis Method
Volatile Content of Paints
Total Volatiles          ASTM D2369 [4]              Gravimetric
Water Content          ASTM D4017 [4]              Karl Fischer Titration
Bulk Product Analysis
VOC/SVOC            Report [1]                    GC/MS
Metals                NAa                        XRF
Metals                Que Hee and Boyle [12]        ICP
Small Chamber Test for Emissions
VOC/SVOC (latex)      Tichenor [13]                 Tenax TA-
                                                  GC/MS; GC/FID
VOC/SVOC (alkyd)     Tichenor [13]                 Charcoal- GC/MS
Aldehydes             Winberry et al. [14]           HPLC

a Work Plan modified - XRF analysis performed by EPA using their procedures
  (Appendix E).
                                 10-3

-------
                       TABLE 10-3. BLANK SAMPLE SUMMARY
Method
Type
    Number
Planned/Reported Results
Volatiles Content
   Total Volatiles
   Water Content
Bulk Product Analysis
   VOC/SVOC (Qual.)

   VOC/SVOC      ;
   (Quant.)
   Metals (ICP)
  -Metals (XRF)
Weigh unused trays
Reagent

Dilution Solvent

Dilution Solvent

Reagent Blank
Thin film substrate
      2/2
      2/0

      3/4

      3/4

      3/3
      3/0
Weight <0.0003 g
No VOCs/SVOCs detected
above trace levels
No targets detected above
quantitation limit
Result given in Table 9-6
Samples analyzed
voluntarily by EPA using
procedure given in
Appendix E
Small Chamber
Latex (SVOCs)
Alkyd (VOCs)

Chamber air/Tenax
Chamber
air/Charcoal

11/6
11/6

No targets detected above
the quantitation limit
No targets detected above
the quantitation limit
                                        10-4

-------
10.3.2  Control Samples
       Spiked samples were prepared, processed and analyzed along with samples to monitor
losses associated with sampling and analysis. A summary of sample types planned and results
is shown in Table 10-4.
10.3.3  Replicate Samples
       Replicate samples are scheduled to provide precision estimates for the overall sampling
and analysis process. For this study replicate tests were scheduled to provide precision
estimates for the overall  method.  For Volatiles Content and Bulk Product Analysis replicate
analysis of paint aliquots provides a measure of the analytical precision.  A summary of
precision evaluations is shown in Table 10-5.

10.4    QUALITY CONTROL PROCEDURES
10.4:1  Volatiles Content
       Total Volatiles
                    QC Procedure                           Result
        Balance daily check weight; acceptance    Daily check not performed;
        £1% deviation                           analysis based on difference
                                               rather than absolute weight
        Oven temperature; acceptance ± 5°C       Temperature verified
       Total Water

                    QC Procedure                           Result
        System performance standard; Recovery   Performance specifications met
        must be 98-102%

10.4.2  Bulk Product Analysis
       Qualitative Analysis (GC/MS)

                    QC Procedure                           Result
        Verify mass calibration using FC-43       Calibration verified
        Manual review of GC/MS data           Completed by MS laboratory
                                               supervisor
                                          10-5

-------
                  TABLE 10-4. SPIKED CONTROL SAMPLE SUMMARY
Method
Type
    Number     Results
Planned / Reported
Volatiles Content
   Total Volatiles       None
   Water Content       None
Bulk Product Analysis
   VOC/SVOC (Qual.)  None
   VOC/SVOC (Quant.) Dilution Solvent
                     i  containing standards
   Metals (ICP)         Latex paint spiked
                       with target metals
   Metals (XRF)
Small Chamber
   Latex (SVOCs)
   Alkyd (VOCs)
Latex paint spiked
with target metals
Spiked reagent
blank
Spiked Tenax tubes
Spiked Charcoal
tubes
                          0/0
                          0/0

                          0/0
                          3/8

                          3/0
      3/2

      0/3
      9/6
      9/6
Not applicable
Not applicable

Not applicable
Results given in Table 7-10
and 7-11
Sample analyzed
voluntarily by EPA using
procedure given in
Appendix E
Results given in Table 9-6

Results given in Table 9-6
Results given in Table 8-11
Results give in Table 8-6
                                         10-6

-------
              TABLE 10-5. SUMMARY OF PRECISION MEASUREMENTS
Method
Volatiles Content
Total Volatiles
Water Content
Bulk Product Analysis
VOC/SVOC (Qual.)
VOC/SVOC (Quant.)
i
Metals (ICP)
Metals (XRF)
Latex (SVOCs)
Alkyd (VOCs)
Type

Duplicates
Duplicates

None
Duplicates
Triplicate aliquot
Triplicate aliquot
Duplicate test
Duplicate test
Number
Planned/Reported

12/12
12/12

0/0
4/4
20/20
0/0*
4/4
4/4
Results

Reported in Table 6-1
Reported in Table 6-1

Not applicable
Reported in Tables 7-8
and 7-9
Reported in Tables 9-3
and 9-4
Not applicable3
Reported in Tables 8-12,
8-26 to 8-31
Reported in Tables 8-7,
8-16 to 8-22
a  As modified, single analyses were performed by EPA.
                                     10-7

-------
       Quantitative Analysis
       Quality control procedures were carried out as described in the QAPP. These are
summarized below. Calibration curve levels and number of points varied depending upon the
estimated range of the specific analytes.
                    QC Procedure
                Result
        LATEX PAINTS (GC/MS)
         -  3-point calibration: %RSD for
            average RF < 30%

         -  Verify mass calibration daily with
            FC-43
                      ,t
         -  Chromatographic performance
            check acceptance R > 1.5, TF > 0.5
         -  Daily calibration check: acceptance
            < 25% difference
        ALKYD PAINTS (GC/MS)
         -  3-point calibration: %RSD for
            average RF < 30%

         -  Verify mass calibration daily with
            FC-43

         -  Chromatographic performance
            check acceptance R > 1.5, TF > 0.5

         -  Daily calibration check acceptance
            < 25% difference
- 3-6 point calibration curve with < 30%
  RSD for average RF

- Calibration verified
- R = (2.1) 1,2-propanediol: ethylene
  glycol
  TF = (3.6) 1,2-propanediol

- < 25% difference in daily check
- 4-point calibration curve with < 30%
  RSD for average RF

- Calibration verified
- R = (5.8) o-xylene:propylcyclohexane
  TF = (1) o-xylene

- < 25% difference in daily check
                                          10-8

-------
  Metals Analysis
                   QC Procedure
                         Result
       XRF Analysis

          Yearly multi-point calibration




       ICP Analysis

          Quarterly multi-point calibration -
          Verify linear range

          Calibration check standard after
          every 10th sample; acceptance £10%
          Difference   \
         Analysis was performed voluntarily by
         EPA (Appendix E) following their
         normal procedures. QC procedures not
         reported.


         3 point calibration verification
         performed during analysis

         Calibration check standards (blank and
         standard) every fifth sample within
         ±10% except ±15% for As, Se, Sn, Hg
         (poor ICP elements)	
      10.4.3  Small Chamber Emissions Testing


      Emission Factors (Chambers)
       Parameter
Requirement
Result
       Temperature

       Relative Humidity

       Air Exchange Rate
23 ± 1 °C           Verified for all tests

50 ± 5% RH        Verified for all tests

1.0 ± 0.05 per hour  Verified for all tests
      Sample Analysis (GC/MS)

      Quality control procedures were carried out as described in the QAPP. These are

summarized below.  Calibration curves levels and number of points varied depending upon

the estimated range of the specific analytes.
                                        10-9

-------
                   QC Procedure
              Result
        LATEX PAINTS (GC/FID)

         -  5-point calibration: %RSD for
            average RF < 30%
         -  Verify mass calibration daily
            with FC-43

         -  Chromatographic performance
            check acceptance R > 1.5, TF >
            0.5

         -  Daily calibration check
            acceptance < 25% difference

        ALKYD PAINTS (GC/MS)

         -  5-point calibration: % RSD for
            average RF < 30%

         -  Verify mass calibration daily
            with FD-43

         -  Chromatographic performance
            check acceptance R > 1.5,
            TF>0.5

         -  Daily calibration check
            acceptance < 25% difference
- 5-point calibration curve, 6-point
  for ethylene glycol with < 30%
  RSD for average RF except
  diethylene glycol

- Calibration verified
  R = (4.1) 1,2-propanediol: ethylene
  glycol
  TF = (5.6) 1,2-propanediol

  < 25% difference in daily check
  except diethylene glycol
- 9-point calibration curve with
  < 30% RSD for average RF

- Calibration verified
- R = (4.5) o-xylene:
  proplycyclohexane
- TF = (1.4)
< 25% difference in daily check
  Sample Analysis (HPLC)
                    QC Procedure
               Result
        5-point calibration; acceptance, r2 for
        curve £0.98
        Daily calibration check; acceptance
        £25 % Difference
  3-point calibration for selected
  analyses, all calibration curves
  met r2 criteria

  25% difference criteria met
10.5 QUALITY ASSURANCE
  A summary of quality assurance activities, including systems audits, is shown in the

Quality Assurance Statement (last page of this section).
                                         10-10

-------
                                 Draft QA Statement
                            EPA Contract No. 68-D2-0131
                          ICF Work Assignment Number 1-18

      Quality Assurance activities undertaken by the Analytical and Chemical Sciences
(ACS) Quality Assurance Office in support of this program (RTI Project 5522-22) included:

      •   meetings with the Work Assignment Leader on matters affecting data quality,
      •   conducting periodic reviews and audits of the data measurement systems, and
      •   monitoring situations requiring corrective action.

      The ACS QA Office conducts systems audits of current ACS studies to ascertain that
data are being recorded properly, SOPs are being implemented, and that the results reported
reflect the raw data of the study. Written reports of all reviews and audits are maintained by
the ACS QA Officer, and results have been reported to the program management.
               Inspection/Audit
   Conducted
Reported
 Instrument Log Notebook Inspection
 (ACS-SOP-815-003)

 Notebook Inspection
 (ACS-SOP-815-002)

 SOP Review
 (ACS-SOP-110001)

 Training Files Inspection
 (ACS-SOP-110-002)
   June 1993
   August 1993
July 19, 1993
   September 1993     Sept. 23,1993


   September 1993     Sept. 21,1993
Aug. 26,1993
      The ACS QA Officer conducted systems audits of this study as specified in the QAPP.
Written reports of all audits are maintained by the ACS QA Officer, and the results have
been reported to the Work Assignment Leader. These systems audits, and other reviews
conducted in support of this study are listed below.
              Inspection /Audit
Conducted
Reported
 Preparation of testing materials/supplies

 Analytical measurement systems

 Data entry and processing

 Data validation

 Document Review
 (ACS-SOP-130-003)
Not done

March 3, 1994

Feb. 10-Mar. 7, 1994

Feb. 10-Mar. 7, 1994

March 3-11,1994
IP

IP

March 7, 1994

March 11, 1994
                                            Doris Smith
                                            ACS QA Officer
                             Date
                                       10-11

-------
                                   SECTION 11.0

                                   REFERENCES
1.     Fortmann, R.C, LS. Sheldon, M.R. Peterson, and D.J. Smith, 1992. Methods Comparison
      for Analyzing Emissions from«Architectural Coatings Used Indoors. Final Report for
      EPA Contract No. 68-02-4544, Work Assignment No. IV-120.

2.     ASTM, 1992, D 3960-91, Standard Practice for Determining Volatile Organic Compound
      (VOC) Content of Paints and Related Coatings, 1992 Annual Book of ASTM Standards,
      Vol 06.01.

3.     Mathtech, Inc., 1992. Interior Architectural Coatings.  Draft Report prepared for the U.S.
      Environmental Protection Agency.

4.     ASTM, 1992, D 2369-91, Standard Test Method for Volatile Content of Coatings, 1992
      Annual Book of ASTM Standards, Vol. 06.01.
   ./
5.     ASTM, 1992, D 4017-90, Standard Test Method for Water in Paints and  Paint Materials
      by Karl Fischer Method, 1992 Annual Book of ASTM Standards, Vol. 06.01.

6.     Clausen, P.A., Wolkoff, P., Hoist, E., and Nielsen, P.A., 1991.   Long-term Emission of
      Volatile Organic Compounds from Waterborne Paints - Methods of Comparison. Indoor
      Air 4:562-576.

7.     Fortman, R.C., November 22,1991.  Report on the Design, Configuration, and Validation
      of an Environmental Chamber System for Emissions Measurements. Miniproposal-Action
      Plan prepared for US. Environmental Protection Agency Air & Energy Engineering
      Research Laboratory in Research Triangle Park, NC.

8.     Peters, T.M. and Todes, C.E., 1994. Preliminary Velocity Field Characterization of AEERL
      Environmental Test Emissions  Chamber with CAT/AEERL  Constant Temperature
      Anemometer System.  Project Report prepared under RTI Project No. 95U-5114-018 for
      Center for Analytical Chemistry and Sciences, Research Triangle Institute and Air and
      Energy  Engineering  Research Laboratory,  U.S.  Environmental Protection  Agency,
      Research Triangle Park, NC.

9.     Sheldon, LS., 1993. Analysis of VOCs and Formaldehyde Samples Report - Analytical
      Support for the Pilot Study of Three Large Buildings.  Final Report for EPA Contract No.
      68-D2-0131, Work Assignment No. 2-9.

10.    Fortmann, R. C, 1993. Analytical Support Relating to Indoor Air Pollution.  Revised
      Work Plan for EPA Contract No. 68-D2-0131, Work Assignment No. 1-18.

11.    Smith, D.J., 1993. Analytical Support Relating to Indoor Air Pollution. Revised Quality
      Assurance Project Plan for Indoor EPA Contract Number 68-D2-0131, Work Assignment
      No. 1-18.

                                        11-1

-------
12.    Que Hee, S.S. and J.R. Boyle, 1988.  Simultaneous Multielemental Analysis of Some
      Environmental and Biological Samples by Inductively Coupled Plasma Atomic Emission
      Spectrometry. Anal. Chem. 60:1033-1042.

13.    Tichenor, B.A., August, 1989. Indoor Air Sources: Using Small Environmental Test
      Chambers to Characterize Organic Emissions from Indoor Materials and Products.
      EPA 600/8-89-074, US. Environmental Protection Agency, RTP.

14.    Winberry, W.T., N.T. Murphy and R.M. Riggin.  Compendium of Methods for the
      Determination of Toxic Organic Compounds in Ambient Air. EPA/600/4-89/017, U.S.
      EPA, Research Triangle Park, NC.
                                        11-2

-------
                 APPENDIX A




RESULTS OF LITERATURE REVIEW FOR TEST METHODS

-------
RESEARCH  TRIANGLE   INSTITUTE


Analytical and Chemical Sciences
                                                                 January 29,1993


       Dr. Niren L. Nagda
       ICF Work Assignment Manager
       ICF Incorporated
       9300 Lee Highway
       Fairfax, VA 22031-1207

       Dear Niren:

             This letter reports the results of the literature search performed to identify
       additional/alternative methods for measuring emissions of aldehydes, volatile organic
       compounds (VOCs), or semi-volatile organic compounds (SVOCs) from liquid products and
       methods for determining the content of aldehydes, VOCs/SVOCs, or metals in liquid
       products. This work was performed under Task 1 of Work Assignment Number 1-18,
       "Determination of Test Methods for Interior Architectural Coatings," under EPA Contract No.
       68-D2-0131,

             To identify potential sources  of information on additional or alternative methods, a
       computerized literature search was conducted using the DIALOG One Search System.  Files
       that were searched included the following:

                    Enviroline,
                    Pollution Abstracts,
                    NTIS,
                    Inspec 2,
                    Analytical Abstracts Online,
                    CA Search,
                    Energy Science & Technology, and
                    Compendex Plus

             A hierarchical searching method was used to limit the number of citations to be
       reviewed.  Key words, or permutations of the words (e.g., paint*) that were used for metals
       included paint, coating, ICP, XRF, inductively, coupled, plasma, and x-ray fluorescence For
       aldehydes, the keywords included paint, coating, aldehyde, ketone, formaldehyde, analysis,
       measurement, and detection.  Keywords to search for methods for emissions testing included
       paint, coating, VOC, volatile, organic, emission, test, measure, chamber, and chambre.  The
       search was not limited by year of publication or by country of publication.  Potentially
       relevant citations were printed and reviewed.  Copies of selected publications and reports
       were ordered, reviewed at local libraries, or collected from reports and publications available
       at Research Triangle Institute  (RTI).

             In addition to the computer search, reports were reviewed in proceedings from
       meetings and symposia such as the International Indoor Air Quality and Climate conferences,


Post Office Box 12194   Research Triangle Park,  North Carolina 27709-2194
Telephone  919 541-6507   Fax: 919 541-7208

-------
Dr. Niren L Nagda
January 29,1993
Page 2


annual American Society of Heating and Refrigerating Engineers (ASHRAE) indoor air
conferences, and American Waste Management Association (AWMA) meetings. A review of
analytical techniques applicable to the examination of coatings is published every two years
by the journal Analytical Chemistry. The reviews for the years from 1981 to 1991 were
examined to identify relevant citations.

      Telephone calls were also made to selected researchers in the US. who have
published papers related to emissions testing or characterization of indoor air contaminant
sources.

      The literature search has not identified any additional or alternative methods that we
would recommend for inclusion in this work assignment  As discussed below, there is an
alternative emissions chamber, the FLEC, that may be potentially useful for measuring
emissions from paints and other liquid products. However, we do not recommend
evaluating it in this work assignment because initial evaluations are being planned by Dr.
Tichenor at EPA. Information was not identified in the literature search on sampling or
analysis methods that should be substituted for those that we have selected, as outlined in
the Work Plan and Quality Assurance Project Plan, for VOCs/SVOCs, aldehydes,  or metals.
The following discussion summarizes the results of our literature search in four areas: (1)
methods for determining emissions of VOCs/SVOCs and aldehydes, (2) methods for
determining the VOC/SVOC content of paints, (3) methods for determining the aldehyde
content of paints, and (4) methods for determining the content of metals in liquid  paints.

Methods For Determining Emissions of VOCs/SVOCs and Aldehydes

      During the last decade, researchers in Europe and the United States have been active
in the development of methods for determining emissions of volatile organic compounds
from both solid and liquid materials that are used indoors. Much of the early work was
done to characterize formaldehyde emissions from products such as plywood, pressed wood
products, and urea-formaldehyde resins using large room-size chambers. More recently,
development work has been directed toward use of small chambers for VOC emissions
measurements.

      There are numerous reports in the literature of emissions measurements obtained with
chambers of various sizes and construction materials. The diversity of the construction
characteristics of emissions test chambers is best summarized in the recent report by
DeBortoli and Colombo  (1992) that describes the results of an international comparison
experiment on the determination of VOCs  emitted from indoor materials through  small test
chambers. Twenty-three chambers used in the experiment ranged in capacity from 0.004 to
1.475 m3. Seventeen chambers were constructed of stainless steel, one of plated steel, and the
others were glass.  For PVC tile, the inter-laboratory  variability, expressed as the relative
standard deviation for concentrations of the emissions of the target compounds measured
two hours after t^  was less than 45% for all of the participating laboratories.  But  for a wax
sample, the RSDs ranged from 45 to 160%  for the target compound concentrations. The
results showed acceptable method precision for the solid material, but large variability

-------
Dr. Niren L Nagda
January 29,1993
Page 3


between the laboratories for measurements of emissions from the liquid product However,
analysis of the data indicated that the chamber capacity did not introduce any systematic
difference in the results. Nor could any significant differences in the results be attributed to
the chamber wall material This suggests that, although there are differences between
emission chambers, instrumentation, and test methods, there may not be substantial
difference in emissions test results with different chambers.

       Wolkoff and his co-workers (1991) have reported development of the Reid and
Laboratory Emission Cell (FLEQ for emissions measurements. This is a dramatically
different design for an emission chamber.  It has a capacity of only 35 cm3 and a maximum
test surface area of 0.0177 m2.  The chamber is operated at 171 air exchanges per hour
compared to the 0.5 to 2 air exchanges per hour typically used in small chamber testing.
This micro emission cell showed satisfactory correlation with a 234 L chamber for tests with
vinyl floor carpet.  It is potentially useful for determining emissions of liquid products.  It is
our understanding that Dr. Bruce Tichenor's group at the EPA Air and Energy Engineering
Research Laboratory (AEERL) has ordered the FLEC for evaluation. We do not recommend
that it be included  for evaluation in this work assignment because preliminary evaluations
have not yet been initiated by Dr. Tichenor.

       With the exception of the FLEC, we did not identify any new emissions test methods.
Researchers use small chambers and large  (room-size) chambers for quantitating emissions.
Headspace analysis or direct analysis of products are used to identify the target analytes for
quantitation.

       Small chamber test methods have been used for measuring emissions of VOCs,
SVOCs, and aldehydes.  VOCs have been measured most frequently (Tichenor and  Mason,
1988).  Reports of emissions of less volatile organic compounds from household products
have been reported (e.g., Clausen et aL, 1991 and Colombo et al., 1990). Tenax has  been
widely used for collection of emissions from solid and wet products, although Carbotrap and
other charcoal-based sorbents have been used (Clausen et al., 1991; Volkl et aL, 1990; Black
et al., 1991; and others).  Work has been performed to evaluate the performance of various
sorbent materials.  The charcoal based sorbents are well-suited for the collection of VOCs.
Recovery by solvent extraction is good for a wide range of compounds. Collection  of air
samples on carbon-based sorbents with subsequent thermal desorption for analysis  has been
shown to be a suitable method  for many volatile compounds (Mason et aL, 1992), but not all
compounds can be recovered from the sorbent by thermal desorption. Data were not found
on the performance of different sorbents for the less volatile analytes that have been
identified in latex paints (e.g., Texanol).  Although Clausen and his coworkers (1991) used
Tenax TA to collect emissions from  latex paint, they did not report either  the percent
recovery of the analytes from spiked control cartridges nor the recovery of the higher
molecular weight, less volatile, compounds from the chamber. In summary, the literature
review did not identify performance data for sorbents that would suggest selection  of
alternatives to the Tenax TA (for VOCs/SVOCs emitted from latex  paint)  and  activated
carbon (for VOCs emitted from alkyd paint) proposed for evaluation in this study.

-------
Dr. Niren L. Nagda
January 29,1993
Page 4


       Formaldehyde has been collected in numerous studies using impingers containing
sodium bisulfite for subsequent analysis by the chromotropic add method (NIO6H, 1977).
Although this is an excellent method for determination of formaldehyde, collection of air
samples on silica gel coated with 2,4-dinitrophylhydrazine (DNPH) with subsequent analysis
by high performance liquid chromatography (HPLC) is now widely used. The method has
the advantage of lower detection limits, sample collection and analysis are easier to perform,
and the same method can be used to determine a number of different carbonyl compounds.
The fact that the method can be used for determination of a number of different aldehydes is
one of the most important advantages of the method. The method is widely used for
ambient air sampling,and has been used for emissions testing, including tests to determine
emissions from adhesives, floor cleaning products, waxes, and deodorizers (Person et al.,
1990).  No alternative methods that were more suitable for sampling and analysis of
emissions of aldehydes from paints were identified in this literature search.

Methods For Determining the VOC/SVOC Content of Paints

       Volume 06.01 of the Annual Book of ASTM Standards (ASTM, 1992) contains the
following methods and practices:

       •     D4457 - Determination of Dichloromethane and 1,1,1-Trichloroethane in Paints
             and Coatings by Direct Injection into a Gas Chromatograph.
       •     D3271 - Direct Injection of Solvent-Reducible Paints Into a Gas Chromatograph
             for Solvent Analysis.
       •     D3272 - Vacuum Distillation of Solvents From Solvent-Reducible Paints for
             Analysis.
       •     D3168 - Qualitative Identification of Polymers in Emulsion Paints.

Methods D4457 and D3271 are packed column methods, which are not as suitable for
analysis of paints as the capillary column methods that we used in the previous study and
that will be  used in this work assignment. Method D 3168  is a pyrolysis method intended to
identify monomers in paint and is not intended to identify volatile emissions from coatings.
Method D3272 is an alternative to direct injection. Results from our previous evaluation of
methods demonstrated that direct injection was an appropriate method for identifying the
VOCs that would be emitted from paint. Therefore,  distillation to separate the solvents from
the solids is not necessary.  Olson and co-workers (1987)  reported a method for determining
solvent formulations of paints using  a Unacon 810A  with a pyroprobe.  An aliquot of paint
was introduced into the tube furnace and heated to 150 °C  The vaporized solvents were
collected.on the Unacon concentrator traps, then analyzed by GC/FID.  The precision of the
method was generally better than ± 10% (%RSD) and recoveries ranged from 76 to 120%.
This method represents an alternative method that could be employed for certain liquids if
they are not amenable  to analysis by direct injection.

       Another method that has been proposed for the determination of VOCs in paints is
the method  developed  at RTI (Petersen et al., 1991) as an alternative to ASTM  methods
D2369 and D4017. The method was  evaluated in our previous study.  The performance of

-------
Dr. Niren L Nagda
January 29,1993
Page 5


the method was excellent for determining total volatile organic compound and water
concentrations. The charcoal sorbent was extracted for qualitative determination of the VOCs
emitted during drying of the paint Some of the most abundant VOCs in the alkyd paint
were identified. But the method was not suitable for determination of the less volatile
compounds emitted from the latex paint The method, however, is still under development
for the purpose of identifying the individual VOCs in the paints. It was not recommended
for evaluation in this work assignment

Methods for Determining the Content of Aldehydes in Liquid Paints
                  \
       Carbonyl compounds have been analyzed by gas chromatography as their
phenylhydrazones, 2,4-dinitrophenylhydrazones, and oximes (summarized by Peltonen, et al.,
1984).  A method of analysis of formaldehyde by GC/MS using deuterated internal standards
has also been reported (McGuire et aL, 1991).  Peltonen and his coworkers (1984) reported a
method for the separation and determination of dimethone adducts of aldehydes by GC

       HPLC has been used for the separation and detection of a number of carbonyl
compounds derivatized with DNPH (as described in Method TO-11, Winberry et aL, 1988).
HPLC analysis of air samples collected on DNPH coated silica gel will be used in this work
assignment. Selim (1977) reported on the quantitative conversion of propionaldehyde to its
2,4-dinitrophenylhydrazone. The quantitative conversion of the aldehydes and the low
detection limits with this method make it advantageous. Methods for derivatization of
aldehydes  for the preparation  of calibration standards are described in TO-11. Waters
Associates also has published a method for preparing derivatives using their DNPH reagent.
The method proposed for this work plan is based on direct derivatization of the aldehydes in
paints with the Waters reagent or derivatization of a solution of the paint in an appropriate
solvent.

       The literature search did not identify methods for determining aldehydes in paints
whkh are more appropriate than the HPLC method.

Methods for Determining Metals in Liquid Paints

       Numerous  methods have been published for the determination of metals in solid and
liquid matrices. A number of standardized methods for digestion of samples with
subsequent analysis by atomic absorption spectroscopy (AAS) have been  published as ASTM
methods and as EPA methods (e.g., EPA Methods for Chemical Analysis of Water and
Wastes, U.S. EPA, 1983).  AAS methods, however, are for determination of a single element.
To meet the objectives of EPA/ORIA and OPPT, multi-element methods are desired. The
selection of x-ray fluorescence spectroscopy (XRF) and inductively couple plasma (ICP)
spectroscopy are appropriate methods for evaluation in this work assignment because both
are applicable to multi-element analyses of paints.  Que Hee and Boyle (1988), for example,
have evaluated ICP for determination of metals  in a variety of matrices, including dry paint
samples. They reported good precision and accuracy for a microwave digestion/ICP method
for most of the 22  elements  that were determined by the method.  Binstock and coworkers

-------
Dr. Niren L. Nagda
January 29,1993
Page 6


(1991) reported similar results for the microwave method developed by RTL The method
should be suitable for liquid paint samples. ASTM method D4764 is a method for
determination of titanium by XRF. The method involves determination of titanium in liquid
paints by analysis of a sample prepared as a thin film. The method will serve as the basis for
the XRF method to be used in this study for multi-element determinations. Because XRF and
ICP are the most appropriate methods to be evaluated in this work assignment, a
comprehensive review of AAS methods is not included in this report
       The literature review has not identified additional methods that we would
recommend for further evaluation in this work assignment  Because of the tight time
schedule for reporting on the literature search, a few of the relevant papers/ reports have not
•yet been received for comprehensive review.  However, based on the titles of the citations
and our current information on available sampling, analysis, and emission test methods, we
do not believe that these citations will identify additional methods that should be
incorporated for evaluation in the current work assignment We will be reviewing additional
information to determine if any alternative methods should be considered, particularly for
the determination of aldehydes in the liquid paints.

       References cited in this letter report are included as an Attachment.


Submitted by:                                       Approved by:
    T
    t?.
        0-
Roy c. Fortmann                                    Edo D.iPellizzari
RT1 Work Assignment Leader                        Vice-president
                                                   Analytical and Chemical Sciences
cc:    Dr. D. Naugle, Program Manager
       5522-22 File

-------
                                     References
ASTM, 1992. Annual Book of ASTM Standards, Section 6, Volume 06.01, Paints, Related
Coatings, and Aromatics, American Society For Testing and Materials, Philadelphia, PA.

Binstock, D.A.,  Hardison, D.L., White,}., Grohse, P.M., and Gutknecht, W.F., 1991.
Evaluation of Atomic Spectroscopic Methods for Lead Analysis. Final Report prepared under
EPA Contract No. 68-02-4550 for US. Environmental Protection Agency, Research Triangle
Park, NC.

Black, M., Pearson, W., and Work, L, 1991. A Methodology for Determining VOC  Emissions
from New SBR Latex Backed Carpet, Adhesives, Cushions, and Installed Systems and
Predicting Their Impact on Indoor Air Quality. Measurement of Toxic and Related Air
Pollutants, Proceedings of the U.S. EPA/A&WMA International Symposium, Durham, NC

Clausen, P.A., Wolkoff, P., Hoist, E., and Nielsen, P.A., 1991. Long-term Emission of Volatile
Organic Compounds From Waterborne Paints - Methods of Comparison.  Indoor Air 4:562-576.
  ,/
Colombo, A., De Bortoli, M., Knoppel, H., Schauenburg, H., and Vissers H., 1990.
Determination of Volatile Organic Compounds Emitted from Household Products in Small
Test Chambers  and Comparison with Headspace Analysis. Indoor Air '90, Proceedings of
the 5th International  Conference on Indoor Air Quality and Climate, Toronto, Canada.

DeBortoli, M. and Colombo, A.,  1992.  International Comparison Experiment on the
Determination of VOC  Emitted from Indoor Materials through Small Test Chambers. Draft
Report for CEC, Joint Research Centre, Environment Institute, Ispra,  Italy.

Mason, M.A., Krebs,  K., Roache, N., and Dorsey, J.A., 1992. Practical Limitations of
Multisorbent Traps and Concentrators for Characterization of Organic Contaminants of
Indoor Air. Measurement of Toxic and Related Air Pollutants, EPA/A&WMA Symposium
Proceedings, Durham, NC.

McGuire, J.M., and Nahm, S.H.,  1991.  Determination, by GC-MS Using Deuterated Internal
Standards, of Formaldehyde and Methanol Evolved During Curing of Coatings.  Journal of
High Resolution Chromatography 14(4):241-244.

National  Institute of  Occupational Safety and Health: Manual of Analytical Methods, 2nd
Ed., 1977. Vol. II DHEW (NIOSH), Publication No. 77-175A Method No. P&CAM 125,
Cincinnati, OH.

Olson, K.L., Wong, C.A., and Fleck, L.L., 1987. Qualitative and Quantitative Determination
of Solvent Formulations in Automotive Paints. Journal of Chromatographic Science, 25:418-
423.

Peltonen, K., Pfaffli, P., and Itkonen, A., 1984.  Determination of Aldehydes in Air as
Dimethone Derivatives by Gas Chromatography with Electron-Capture Detection. Journal of
Chromatography 315:412-416.

-------
Person, A., Laurent, A.M. Louis-Gavet, M.C., Aigueperse, J., and Anguenot, R, 1990.
Characterization of Volatile Organic Compounds Emitted by Liquid and Household Products
Via Small Test Chamber.  Indoor Air '90, Proceedings of the 5th International Conference on
Indoor Air Quality and Climate, Toronto, Canada.

Peterson, M.R., Jayanty, R.K.M., McAlister, G.D., and Knoll, J.E., 1991.  Direct Measurement
of VOC in Water-Based Coatings. Measurement of Toxic and Related Air Pollutants,
Proceedings of the 1991 U.S. EPA/A&WMA International Symposium, Durham, NC

Que Hee, SS. and Boyle, J.R., 1988.  Simultaneous Multielemental Analysis of Some
Environmental and Biological Samples by Inductively Coupled Plasma Atomic Emission
Spectrometry.  Analytical Chemistry 60:1033-1041

Selim, S., 1977. Separation and Quantitative Determination of Traces of Carbonyl
Compounds as Their 2,4-Dinitrophenylhydrazones by .High-Pressure Liquid Chromatography.
Journal of Chromatography, 136:271-277.

Tichenor, B.A., and Mason, M.A., 1988. Organic Emissions from Consumer Products and
Building Materials to the Indoor Environment.  Journal of the Air Pollution Control
Association 38:264-268.

US. EPA, 1983. Methods for Chemical Analysis of Water and Wastes.  EPA-60074-79-020,
Office of Research and Development, VS. Environmental Protection Agency, Cincinnati, OH.

Volkl, S., Gebefugi, I.L., and Korte, F., 1990. Emission of Volatile Organic Compounds from
Coatings into Indoor Air. Indoor Air '90, Proceedings of the 5th International Conference on
Indoor Air Quality and Climate, Toronto, Canada.

Winbeny, W.T., Murphy, N.T., and Riggan, R.M., 1988. Compendium  of Methods for the
Determination of Toxic Organic Compounds in Ambient Air.  EPA/600/4-89/017, US.
Environmental Protection Agency, Research Triangle Park, NC

Wolkoff,  P., Clausen, P.A., Nielsen, P.A., Gustafsson, H., Jonsson, B., and Rasmusen, E., 1991.
Field and Laboratory Emission Cell: FLEC.  IAQ '91, Healthy Buildings, Washington, DC.

-------
RESEARCH  TRIANGLE  INSTITUTE

Analytical and Chemical Sciences

                                                                             March 5/1993
      Dr. Niren L. Nagda
      ICF Work Assignment Manager
      ICF Incorporated       •
      9300 Lee Highway
      Fairfax, VA 22-31-1207

      Dear Niren:

            As  suggested  by John Girman, I have made some additional telephone contacts to
      determine if any relevant data are available on recovery of Texanol and other latex paint analytes
      from Tenax TA or other sorbent media.  This work was performed under Task 1 of Work
      Assignment Number 1-18, "Determination of Test Methods for Interior Architectural Coatings,"
      under EPA Contract NO. 68-D2-0131.

            Al Hodgson of Lawrence Berkeley Laboratory  is about to begin some work  on the
      recovery of Texanol from the multi-sorbent traps, but does not have any data at this time.  He
      uses the Envirochem tubes and a Unacon desorber/concentrator similar to the unit we use.  Al
      indicated that  he has experienced some problems obtaining  a good calibration curve for
      butoxyethanol using the multi-sorbent tube/thermal desorption system, but has not identified
      the cause of the problem.

            Charlie Weschler and Helen Shields have experience with the measurement of Texanol
      using the 3M badge.  The collection media  in the badge is a filter impregnated with charcoal
      The carbon disulfide extraction apparently provides good recovery. They do not have experience
      with Tenax.

            Marilyn Black uses the Carbosieve/Carbopack multisorbents. She hasn't worked with
      Texanol. But she has performed recovery tests for 2-(2-butoxyethoxyethanol), one of the latex
      paint analytes.  She says recovery is approximately 75% with the thermal desorption method.

            I also talked  to  Peder Wolkoff at  the National Institute of Occupational Health in
      Denmark.  He was one of the authors of the paper on long-term emissions of VOCs from latex
      paint (Indoor Air, 4:562-576,1991).  They used sampling tubes containing 200 mg of Tenax TA
      and a Perkin-Elmer system for thermal desorption.  Peder advised me that they have not done
      tests to determine the recovery of individual VOCs from the sampling cartridges. Instead, they
      do what he referred to as "total recovery" for the method by performing their calibration using
      Tenax TA cartridges  spiked with the analytes. The standards are prepared in methanol, then
      loaded onto the front of the Tenax bed of the cartridge.  Helium is then passed through the


Post Office Box 12194   Research Triangle Park, North Carolina 27709-2194
Telephone 919541-6507   Fax:919541-7208

-------
Niren L. Nagda                           -2-                             March 5,1993
cartridge at 50 mL/min for 3 minutes. With this approach, if the recovery from the cartridge
is not dependent of the mass of analyte loaded onto the sorbent bed, you will obtain a linear
calibration curve over your dynamic range. Recovery of Texanol from air samples collected on
the cartridge is assumed to be the same as that from the spiked cartridges used for calibration.
He indicated that they get a good linear calibration and have had no problems with the major
compounds in the latex paint emissions.

       Preparation of calibration curves using standards loaded on the sampling cartridges is
our standard practice when performing thermal desorption/GC/MS with Tenax or multi-sorbent
tubes. We do not inject liquid solutions of the standards directly onto the GC column to develop
our calibration curves.  Our procedure, therefore, is the same as that of Wolkoff.  For this Work
Assignment we will perform initial recovery tests to determine  that the recovery of the target
analytes is adequate. This is done by comparing  the response of the instrument for standards
in solution injected onto the column against the response for standards thermally desorbed from
spiked Tenax cartridges.

       If you have any questions about our procedures, we can discuss then on March 11,1993
when you and Pauline Johnston visit our facility.


                                                          Sincerely,
                                                          Roy C Fortmann
                                                          RTI Work Assignment Leader
cc: Dr. D. Naugle, Project Office

-------
      APPENDIX B




PAINT SELECTION MEMOS

-------
RESEARCH   TRIANGLE  INSTITUTE
                                                                                /RTI
Analytical and Chemical Sciences
                                                          April 23,1993
      Dr. Niren L. Nagda
      ICF Work Assignment Manager
      ICF Incorporated
      9300 Lee Highway
      Fairfax, VA 22-31-1207
                                                  9.
      Dear Niren:

            As I discussed with you this morning, I have collected additional information about
      the Glidden and Sherwin-Williams paints that will be used for testing in the Work
      Assignment.  I would like to proceed with purchase of the paints after Pauline Johnston has
      a chance to review my proposed selection of paints. The procedure for selection will be as
      outlined in the Work Plan with one minor modification, as will be described below.

            The paints that will be purchased will be a medium to high grade, based on cost
      Both the Glidden and Sherwin-Williams paints that I have selected  are the "most popular"
      choice of homeowners according to staff at the retail outlet.  The paints that have Won
      selected are not "contractor" painU (axcopt for the gless paint) nui  aiu lUl-y lIUURJed fur-
                  «.  The Sherwin-Williams latex flat, latex semi-gloss, alkyd flat, and alkyd semi-
                           series.  The Sherwin-Williams latex gloss and alkyd gloss paints are
gloss are tne
the Pro Mar
                     series, a contractors paint, because the gloss paints are not manufactured in
              kr99 series or other "homeowner" paints. All Sherwin-Williams latex paints contain
      the vinyl polymer.  Glidden has all six types of paints in a medium to high grade, but the
      names vary for each type of paint.  The latex flat paint will be the Spread Satin series and
      will be the same paint as used in the previous testing.  Glidden has the vinyl polymer only
      in the latex flat paint; it is not used in the semi-gloss or gloss.

            All of the six types of  paints can be purchased in any of the manufacturer's colors. I
      have obtained paint "chips" from each manufacturer. Glidden has about 600 colors.
      Sherwin-Williams has over 800 colors.  The Sherwin-Williams paint chips use a letter code to
      indicate  the base paint that is tinted. An "X" indicates that a neutral colored base is used.
      Each paint chip usually has five colors of varying tint-strength that use the "X" base and two
      darker colors that are produced by adding tint to a colored base. The Glidden paint chips
      each show four similar colors (with varying amounts of tint), but they do not indicate the
      base that is used.

            As outlined  in Section 3.1 of the QAPP, paints for bulk product and emissions testing
      will be randomly selected after stratification into six major color groups (green, yellow, blue,
      red, purple, and orange).  Each color on the paint chips will be numbered within each color
Post Office Box 12194   Research Triangle Park  North Carolina  27709-2194
Telephone 919541-6507   FAX 919 541-7208

-------
Or. Niren L. Nagda
April 23,1993
Page Two
group.  A random number generator will be used to select the color for each of the 12 paint
types listed in Table 1 of the Work Plan.  I have also randomly assigned each of the twelve
paints to a color group using a random number generator. The paint type assigned to each
color group is as follows:

Color Group                  Sherwin-Williams                     Glidden
Green                        Alkyd gloss                          Alkyd semi-gloss
Yellow                        Alkyd flat                            Alkyd flat
Blue                          Alkyd semi-gloss                     Latex semi-gloss
Red                          Latex gloss                           Latex flat
Purple                        Latex semi-gloss                      Alkyd gloss
Orange                       Latex flat                            Latex gloss

       The eight additional paints to be used only for metals analyses (in addition to the
twelve paints listed above) will be colors that are randomly selected from all available colors,
without stratification by color group.  I have randomly selected the following additional
paints for metals analysis:

fthfrvyin-Williamf                                                   Glidden
Alkyd flat                                                         Alkyd flat
Alkyd semi-gloss                                                   Alkyd gloss
Latex flat         ,,                                                 Latex gloss
Latex semi-gloss                                                   Latex semi-gloss

       As I discussed with you, I believe all of the available colors on the paint chips should
be eligible for selection. However, I would like to add the criterion that no more than one
paint can be selected from any paint chip. I believe this is necessary because paints with
colors on the same chip may vary only in the amount of tint that was added. This would
reduce the diversity of paints included in the study.

       As you suggested, I will telefax a copy of this letter to Pauline Johnston for her
review.

                                                     Sincerely,
                                                         \J
                                                     Roy C. Fbrtmann
                                                     RTI Work Assignment Leader
cc Ms. Pauline Johnston, EPA
   Dr. D. Naugle, RTI Project Office

-------
 : IAI. tf » » ft V 4 WM
                              4VC  4.
                                              Iff
                                                           819 Sll  SH5:f 2
                                                          .ji  nu.uua r.v
                   UNITfO $TATI9 ENVIRONMENTAL PROTECTION AGENCY
                                WASHINGTON DC. 20460
IBCHNICAL DIRECTIVE  CTD# 3 for WA» i-10)

Date         ApHl 29, 1993

From:
Pauline Johnston
ETA Work Awl
To:
                   it Manager

Nlren Nagda
tCP Work Assignment Manager
This technical dir«ctty«. requests work under the cenfmcC work AUle;tunent/ and task Ibtcd
below.
      Contract No.:

      Work
      Task Number:
      (if appropriate)

      Description of Work:
                   68-D2-01S1

                   1-18

                   Task 3, Pirt 1
                   The ictcction of painU will be made In accordance willi
                   the attached April 23,1993, letter from Roy Fortmann to
                   you, with three changes as discuued in our conference
                   call earlier today. The changes are as follows:

                   •  Imtvad of choosing the Shcrwin Wilt Urns Classic 99
                      and Pro MAT 200 ieri«s painto, the Shirwin-Wllliams
                      paint* wfll be from th« Pro Mar 400 sfiri«. bithU
                      way/ h«J/ the paints chosen will be "homeowner
                      paints and half will be "contractor1' pjlntx. in
                      addition,  the contmricvr paint will be from UM
                      Intermediate frade series that Is used in the greatest
                      quantities by contractor*.  All of these paints contain
                      tnc vinyl polymer.

                   •  When seUctlng colon for the Pro Mar 400 8*rl*», th«
                      darkest color on a rotor swatch will not be used
                      since th«M colors are usually not available in thla
                      paint series.  If the darkest color on a swatch is
                      chosen/ the color will be rese1ect*d.

-------
                                                Iff  Incorporated        919  5419ffiJ 9
                                                          '..(••  fcrf'^V  AW'W4.  MV iVVV I  i V
                                           that ar* loo difficult to put Into a particular
                                    color group (e.g., white, gray, pitch* promw tonvc
                                    teals, etc.), rtprttoenting about 20% of th« total colon,
                                    will be excluded wlwn choosing from the six major
                                    color group for the ftnt 12 paints samples.
                                    However, the jvrnalnlng eight utnple colors win br
                                    choMm from all Ux: available color*.
cc:     Project Officer
                   Officer

-------
RESEARCH  TRIANGLE  INSTITUTE
/RTI
Analytical and Chemical Sciences                                               May 10, 1993


      Dr. Niren L. Nagda
      ICF Work Assignment Manager
      ICF Incorporated
      9300 Lee Highway
      Fairfax, VA 22-31-1207

      Dear Niren:

             As I advised you in our telephone conversation today, the selection method described
      in my letter dated April 23, 1993 for the eight additional paints to be used for metals analysis
      only differs from the method described in Section 3.1 of the Quality Assurance Project Plan
      (QAPP).  In the QAPP we state, in the second bullet, the numbers of each type of paint that will
      be selected for each manufacturer. This statement indicates that the paints for metals analysis
      only will include a flat latex, a semi-gloss latex, a semi-gloss alkyd and a gloss alkyd from each
      manufacturer.  The original intent of this selection method was to obtain the types of paints that
      are most widely used  (Le., alkyd flat and latex gloss  paints represent less market share).
      However, in my letter dated April 23, 1993, 1 proposed a simple random selection.  When I did
      the random selection, the types of paints that were selected were different than those stated in
      the QAPP. Based on currently available information, we do not know if the metals content will
      differ for different gloss types. Therefore, I believe the random selection process proposed in
      my letter is the most appropriate method to select the paints.

             We have also learned that the Sherwin-Williams Pro-Mar 400 series is limited to the latex
      flat, latex semi-gloss, and alkyd semi-gloss  types of paints. The Pro-Mar 200 series, however,
      includes all of the types of paints (latex and alkyd in flat, semi-gloss, and gloss). The Pro-Mar
      200 series is the "top-of-the-line" (i.e., most expensive) paint. I suggest that we use paints from
      the Pro-Mar 200 series for testing in this Work Assignment.

             Please advise me of your response to these issues and provide a technical directive to me
      regarding these matters. Feel free to call me if you would like to discuss these issues. We want
      to procure these paints as soon as possible.

                                                                 Sincerely,
                                                                 Roy C Fortmann
                                                                 RTI Work Assignment Leader
      cc:     Ms. Pauline Johnston, EPA
             Dr. D. Naugle, Project Office
Post Office Box 12194   Research Triangle Park. North Carolina 27709-2194
Telephone 919 541-6507   FAX: 919 541-7208

-------
Lil • ILr  i
                              A U' »J<_'
                      UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                                   WASHINGTON, D.C. 20460
                                                                                OFFICE Qf
                                                                             AIR AND RADIATION
  TECHNICAL DIRECTIVE (TD#4 for WA# 1-18)
  Date:

  From:


  To:
             May 12,1993

             Pauline Johnston
             EPA Work Assignment Manager

             Niren Nagda
             ICF Work Assignment Manager
  This technical directive requests work under the contract, work assignment, and task listed


        Contract No.:             68-D2-0131
belo,w.
        Work Assignment:

        Task Number:
        (if appropriate)

        Description of Work:
                                 1-18

                                 Task 3, Part 1
                                 The proposed changes to the paint selection criteria, as
                                 detailed in the attached letter from Dr. Roy Fortmann to
                                 you dated May 10,1993/ are reasonable. Therefore,
                                 proceed with a random selection of the eight additional
                                 types of paint for metals analysis as detailed in the first
                                 paragraph of the letter. Also, use the Pro-Mar series 200,
                                 rather than the Pro-Mar series 400, paints.
  cc:    Project Officer
        Contracting Officer
                                                                         T£  Punted on Recycled Paper

-------
                           APPENDIX C






EXAMPLE CALIBRATION CURVE FOR QUANTITATING VOC EMISSIONS FROM




                      ALKYD PAINT SAMPLES

-------
                       Calibration Report

            Title:  ALKYO  CHARCOAL  TEST  SAMPLES  IN  CS2 WITH O-XYLENE-010 AS  I.S.
       Calibrated:  940215 09:29
Files: >09U9
RF
Compound 25.00
TOLUENE
TOLUENE
H.P-XYLENE
M.P-XYLENE
NONANE
NONANE
0-XYLENE
0-XYLENE
PROPYL CYCLOHEXANE
PROPYL CYCLOHEXANE
3- & 4-ETHYL TOLUENE
3- & 4-ETHYL TOLUENE
1,3.5-TRIMETHYLBENZENE
1,3.5-TRIMETHYLBENZENE
DECANE
DECANE
2-ETHYL TOLUENE
2-ETHYL TOLUENE
1,2.4-TRIMETHYLBENZENE
..2,4-TRIHETHYLBENZENE
1,2.3-TRIMETHYLBENZENE
1,2.3-TRIMETHYLBENZENE
2-HETHYLOECANE
2-METHYLDECANE
TRANS-OECAHYDRONAPHTHALENE
TRANS-DECAHYDRONAPHTHALENE
UNOECANE
UNDECANE
PENTYLCYCLOHEXANE
PENTYLCYCLOHEXANE
N-DODECANE
N-DOOECANE
91 2.75012
92 1.56655
91 2.21536
106 1.10386
85 .50291
57 1.30552
91 2.24419
106 1.04300
83 1.72280
55 1.26416
105 2.55083
120 .75571
105 2.20976
120 1.09853
85 .41299
57 1.49838
105 2.53453
120 .76988
105 2.09620
120 1.00693
105 2.03628
120 .87246
85 .40501
57 1.23429
96 .52426
138 .69044
85 .43409
57 1.49023
83 1.76190
55 1.19079
85 .48383
57 1.57881
>09150
RF
50.00
2.97379
1.64079
2.37113
1.17638
.54304
1 .43077
2.45938
1.10876
1 .83355
1.32851
2.72021
.80854
2.32881
1.20292
.44614
1 .67285
2.81254
.82038
2.48200
1.04396
2.20477
.90738
.45417
1 .48697
.57921
.70982
.46426
1.80033
2.03120
1.24935
.53019
1.75955
>09151
RF
100.00
2.93922
1.62106
2.24327
1.12645
.51466
1.35413
2.28358
1.07079
1.77372
1.26818
2.68029
.76697
2.17905
1.11939
.42072
1.57173
2.56128
.76132
2.19192
.96681
2.03710
.87670
.40790
1.48508
.52049
.65876
.43377
1.67006
1.81377
1.24656
.47260
1.61296
>09152
RF
250.00
3.15177
1 .83786
2.06377
1.17927
.56059
1.58107
2.50995
1.14680
1 .92949
1 .48299
2.58439
.81668
2.34385
1.24368
.45610
1.66841
2.91797
.81169
2.33263
1.14329
2.36083
.93248
.44824
1 .45632
.54536
.72864
.47647
1.72297
1.99828
1.33068
.51768
1.73332
>09158
RF RF
500.00 1000.0
2.57010
1.71545
1.48318
.90759
.51626
1.33727
2.02376
1.08005
1.64848
1.21087
1.91664
.73657
1.94173
1.03099
.41519
1.49240
2.16690
.76645
1.93068
.98692
1.81962
.86952
.39128
1.26208
.48089
.66605
.42351
1.52251
1.63831
1.14177
.43332
1.42031
>09154
RF
.500
2.67490
1.51460
2.08347
1.02837
.35980
1.10040
2.17484
.96047
1 .60465
1.01048
2.40332
.65861
1.93152
.98853
.26006
1.26053
2.31754
.67309
1.93132
.80211
1.79380
.70337
.23553
1.11868
.38988
.54864
.30580
1.31378
1.51588
.98292
.27336
1.13389
>09155
RF
1.00
2.98360
1.64529
2.27202
1.10068
.43122
1.24209
2.32676
1.04090
1.66856
1.27686
2.60660
.75260
2.19024
1.08187
.37910
1.47266
2.58368
.76841
2.14747
.96062
1.98137
.83461
.33747
1.25837
.50712
.61065
.36924
1.45823
1.73408
1 .09024
.39066
1.32019
>09156
RF
2.50
3.12013
1.73975
2.33790
1.20206
.49261
1.31554
2.42462
1.14392
1.80461
1.26399
2.69237
.82206
2.29272
1.18217
.40056
1.43744
2.77387
.80473
2.26486
1.07964
2.14026
.93667
.41598
1.31895
.57660
.72349
.42743
1.59334
1.90022
1.24530
.48007
1.55480
>09157
RF
10.00
2.75025
1.56598
2.21399
1.11731
.50317
1.28225
2.28880
1.09754
1.76266
1.25499
2.57891
.78345
2.22023
1 . 12781
.41740
1.44283
2.61728
.80213
2.18633
1.02046
2.02584
.89043
.40364
1.21956
.54379
.67660
.43006
1.48587
1.83396
1.13374
.47718
1.55386
RF
2.87932
1.64970
2.14268
1.10466
.49158
1.32767
2.30399
1.07691
1.74984
1.26234
2.52595
.76680
2.18199
1.11954
.40092
1.50191
2.58729
.77534
2.17371
1.00119
2.04443
.86929
.38880
1.31559
.51862
.66812
.41829
1.56192
1.80307
1.17904
.45099
1.51863
X RSO
7.015 (C
6.097 (C
12.482 (C
8.191 
-------
    APPENDIX D




EMISSION RATE DATA

-------
ALKYD PAINTS

-------
Chamber
Time
hr
0.65
1.2
2.1
3.1
4.1
7.8
11.8
24.6
0.7
1.2
2.2
3.2
4.2
7.9
11.9
24.9
Sumsq
*^PdBN^^^^^^^^
SO (mg/hr-g)
K(/hr)
CO
Emission
(mg/g)
VOC Recovery Study
TEST 5 & 6 Combined
m,p-Xylene n-Nonane o-Xylene Propyteycloh 3&4-Ethyltol

12
14
10
7.2
4.2
0.47
0.047

8.9
8.8
6.1
8.2
4
0.4
0.038
.•'
1.955
0.059
1 .050E+00
5.534E-01
4.186E+00
2.30


26
32
30
26
19
1.7
0.35

22
24
26
28
22
1.7
0.34

3.248
0.033
3.189E+00
4.594E-01
O.OOOE+00
6.94


4
4.7
3.4
2.7
1.7
0.27
0.032

2.9
2.9
2.3
; 2.9
1.6
0.24
0.028

1.078
0.096
3416E-01
4.744E-01
1 .234E+00
0.86


9.3
11
11
11
7.4
0.71
0.23

8.9
7.9
9.1
9.6
9.2
0.59
0.21

2.336
0.064
1.127E+00
4.366E-01
O.OOOE+00
2.58


5.1
6.9
6.2
6.1
4.7
1.4
0.44

4.4
4.6
5.4
5.6
5
1.4
0.45

1.027
0.050
5.484E-01
2.844E-01
O.OOOE+00
1.93

Vol
1,3,5-Trimet n-Oecane 2-Ethyltoluen 1 ,2.4-Trimet

1.7
2.3
2.3
2.2
1.9
0.63
0.28

1.6
1.6
2
2.1
2
0.68
0.28

0.582
0.076
1 .855E-01
2.327E-01
O.OOOE+00
0.80


40
54
59
63
57
15
9.3
0.065
37
41
60
55
63
17
11
0.055
4.543
0.020
4.866E+00
2.299E-01
O.OOOE+00
21.17


1.5
2
1.9
1.9
1.5
0.16
0.15

.4
.4
.6
.7
.6
0.25
0.18

0.911
0.145
1.783E-O1
3.498E-01
O.OOOE+00
0.51


5.3
7.4
7.2
7.4
6.1
2.1
1.1
0.02
4.8
4.9
6.6
6.4
6.5
2.3
1.2
0.023
1.160
0.043
5.790E-01
2.213E-01
O.OOOE+00
2.62

0.05276
1,2,3-Trimet

1.4
2.1
2.1
2.2
1.9
0.94
0.52
0.026
1.3
13
1.9
1.8
1.9
1
0.56
0.026
0.633
0.080
1.528E-01
1.640E-01
O.OOOE+00
0.93

S(t) - SO exp(-M) with CO - Concentration at t*0

-------
cu m
            ACH
2-MettiykJec  trans-Decah n-Undecane  Pentyteycloh n-Dodecane TVOC
2.1
3.1
3.7
4.4
4.3
1.8
1.3
0.091
2
2.2
3.7
3.1
4.6
1.6
1.6
0.083/
1.357
0.082
2.597E-01
1.350E-01
O.OOOE+00
3.5
5.5
5.7
6
5.4
2.1
1.4
0.035
3.5
3.7
5.3
4.9
6.2
2.1
1.6
0.033
1.303
0.058
4.282E-01
1.804E-01
O.OOOE+00
6.2
11
12
14.9
15
8.2
7.8
1.5
6.1
6.8
13
10.5
15
9.3
9.3
1.5
2.452
0.045
8.024E-01
7.637E-02
O.OOOE+00
0.38
0.82
0.93
1
1 .
1.2
1
0.17
0.38
0.44
0.81
. 0.66
1.1
1.2
1.1
0.17
1.125
0.260
5.540E-02
4.609E-02
O.OOOE+00
0.53
0.94
1.2
1.4
1.6
1.7
2
1.7
0.49
0.62
1.3
1
1.5
1.9
2.4
1.8
0.953
0.110
6.950E-02
-1.995E-02
O.OOOE+00
589
817
768
764
625
205
139
18.5
503
538
677
671
662
203
148
18.5
12.958
0.005
6.048E+01
2.174E-01
2.457E-05
       1.92
2.37
10.51
1.20
278.16
                                                                                             /
                                                                                           ,^J/l-t*TAj-> —
                                                                                                            
-------
   100
                           Chamber Results
                         Test 5 & 6 Combined
f
o>
O
O
§  0.1
O
  0.01
                           Time (hr)
                                      10
15
                                                              m,p-Xy!ene
                                                               v
                                                              n-Nonane
                                                               o
                                                              o-Xylene
         Fit

-------
  100
                             Chamber Results
                           Test 5 & 6 Combined
*o>

o
I
0)
o
o
o
    10
Propylcyclohexane
 V
3&4-Ethyltoluene
 ©
1,3,5-Trimethylbenzene
Fit
                             Time (hr)

-------
                            Chamber Results
                          Test 5 & 6 Combined
   100 -
%  «>
O>
o
*s
1

-------
                              Chamber Results
                            Test 5 & 6 Combined
    10
•5*
eo
1
    0.1
 u
 o
o
              1,2,3-Trimethylbenzene
               V
              2-Methyldecane
               ©
              trans-Oecahydronaphthalin
              Fit
   0.01
                            10         15
                              Time (hr)
20
25

-------
  100
                           Chamber Results

                         Test 5 & 6 Combined
-5*
CO
c
o
0)
o

o
o
   10
1 -
                                                         n-Undecane

                                                          v


                                                         Pentylcyclohexane

                                                          ©


                                                         n-Dodecane
                                                         Fit
   0.1
                         10        15

                           Time (hr)
                                        20
25

-------
                         Chamber Results

                        Test 5 & 6 Combined
   1E3
•5*
<*>

•i 1E2
O)




o


1
*rf

§ 1E1
O
c
o
o
                  TVOC
                  Fit
    1EO
                         i        i

                        10       15

                         Time (hr)
20
25

-------
Chamber VOC Paint Study
TEST 11 & 12 Combined
Time m,p-Xylene n-Nonane o-Xylene Propylcycloh 3&4-Ethyltol 1 ,
hr
0.63
1.1
2.1
3.1
4.1
8.7
12.7
24.7
0.63
1.1
2.1
3.1
4.1
8.7
12.7
24.7
Sumsq
Std
SO
K
CO
Emission
(mg/g)

5.8
5
2.9
2.1
1.3
0.14
0.015

5
4.9
3.3
2
1.3
0.28
0.057
-' 0.00051
0.4582
0.026
0.259
0.462
6.688
1.32


2.7
2.3
1.8
1.5
1.1
0.24
0.062

1.9
2
1.6
1.2
0.95
0.41
0.2
0.009
0.5829
0.084
0.123
0.274
2.371
0.90


1.5
1.2
0.81
0.53
0.37
0.052
0.0074

1.4
1.1
0.83
• 0.56
0.37
0.11
0.027
0.0007
0.2171
0.045
0.057
0.379
1.839
0.41


0.71
0.63
0.47
0.37
0.31
0.076
0.021

0.5
0.47
0.43
0.35
0.24
0.13
0.058
0.0037
0.3249
0.188
0.031
0.247
0.628
0.26


0.52
0.5
0.41
0.32
0.27
0.092
0.031

0.43
0.4
0.36
0.28
0.23
0.13
0.067
0.0072
0.2293
0.154
0.024
0.199
0.515
0.26

3,5-Trlmet

0.18
0.18
0.15
0.13
0.11
0.047
0.019
0.00081
0.14
0.13
0.13
0.1
0.085
0.053
0.035
0.0061
0.1878
0.347
0.009
0.177
0.154
0.10

                                                                            Vol
                        0.05276
                                                               2-Ethyltoluen 1.2.4-Trimet  1,2,3-Trimet
                                                           8.1
                                                           7.8
                                                           7.2
                                                            7
                                                           5.9
                                                            2
                                                           1.5
                                                          0.38

                                                           5.8
                                                           5.3
                                                           5.8
                                                           4.9
                                                           4.3
                                                           1.8
                                                           1.5
                                                          0.85

                                                        1.6128
                                                         0.062

                                                         0.395
                                                         0.145
                                                         6.250

                                                          4.98
  0.17
  0.16
  0.14
  0.11
   0.1
 0.038
 0.014
  0.13
  0.13
  0.11
   0.1
 0.079
 0.046
 0.026
0.0046

0.1445
 0.298

 0.008
 0.174
 0.155

  0.09
  0.52
  0.53
  0.49
  0.41
  0.36
  0.16
 0.072
0.0043

   0.4
  0.42
  0.37
  0.31
  0.25
  0.17
  0.11
 0.028

0.3417
 0.193

 0.027
 0.162
 0.439

  0.31
  0.16
  0.16
  0.15
  0.13
  0.12
 0.059
 0.031
0.0041

  0.12
  0.13
  0.12
    0.1
 0.082
 0.058
 0.042
 0.015

0.1815
 0.336

 0.008
 0.130
 0.135

  0.12
S(t) - SO exp(-M) with CO « Concentration at  t»0

-------
cum
           ACH
                    1
2-MethyUec trans-Oecah n-Undecano  Pentyteyctoh rvDodacane  TVOC
3
3.3
3.3
3.2
3.2
0.81
0.79
0.93
2.1
2.3
2.4
2
1.9
0.81
0.73
0.82''
1.6552
0.139
0.168
0.127
2.117
4.5
4.9
4.9
4.4
4.4
1.3
1.1
0.39
3.6
4
3.6
3.1
2.7
1.3
1.2
0.82
1.3985
0.079
0.249
0.128
3.880
13.3
13.5
13.5
13.7
13.8
2.7
2.1
2.7
10.3
10.8
10.8
9.8 '•
9.6
1.9
1.8
1.8
3.0289
0.061
0.992
0.231
5.644
2
2.3
2.1
2
2.4
0.64 .
0.62
0.69
1.2
1.6
1.7
1.4
1.3
0.58
0.59
0.52
1.3706
0.158
0.109
0.106
1.283
2.3
2.8
2.8
3
3.2
1.3
1.3
1.8
1.7
2
2.1
1.9
1.6
1.1
1
1.1
1.5465
0.134
0.117
0.039
2.098
240
270
250
240
220
100
89
62
160
170
170
140
120
71
68
53
12.8174
0.014
10.007
0.070
224.077
       2.20
3.54
5.58
1.68
5.85
311.82

-------
                           Chamber Results
                        Test 11 & 12 Combined
     10
•5*
w

"B)
    0.1 -.
o 0.01
I
g 0.001
o
o
  0.0001
                                                                 m,p-Xylene
                                                                  V  •
                                                                 n-Nonane
                                                                  ©
                                                                 o-Xylene


                                                                 Fit
  1E-05
                            10         15
                               Time (hr)
                                                  20
25

-------
                             Chamber Results
                          Test 11 & 12 Combined
      1 -
o 0.01
0)
o
o 0.001
o
              Propylcyclohexane
               y
              3&4-Ethyltoluene
               o
              1,3,5-Trlmethylbenzene
              Fit
  0.0001
                           10        15
                             Time (hr)
20
25

-------
                            Chamber Results
                         Test 11  & 12 Combined
     10
•5?
CO
"3)
c
o
1
£
0)
o
o
o
 0.1
0.01
    n-Decane
     V
    2-Ethyttoluene
     o
    1,2,4-Trimethylbenzene
                                                         Fit
   0.001
                           10       15
                             Time (hr)
                                          20
25

-------
                             Chamber Results
                          Test 11 & 12 Combined
     10
eo
O)

O
1
    0.1

-------
  100
                           Chamber Results
                        Test 11 & 12 Combined
O)
CO
£ 10
D)
S
§   i
o
O
o
0.1 -r

   0
                         —i	1—
                          10         15
                           Time (hr)
20
                                                        25
              n-Undecane
               T

              Pentylcyclohexane
               ©

              n-Dodecane
                                                            Fit

-------
                     Chamber Results
                  Test 11 & 12 Combined
1E3
                                                        TVOC
                                                        Fit
1E1
                       10         15
                         Time (hr)
20
25

-------
Chamber VOC Recovery Study
TEST 13 & 14 Combined
Time m,p-Xylene n-Nonane o-Xylene Propyteycloh 3&4-Ethyltol '
hr
0.65
1.2
2.1
3.1
4.1
7.6
11.8
24.8
0.7
1.2
2.2
3.2
4.2
7.9
11.9
24.9
Sumsq
9BBBiqk3&
9^T
SO
K
CO

9.1
10
8.3
6
4.3
0.82
0.16

9.2
10.6
9
6.8
5.4
1.3
0.38

0.7714
0.022
8.663E-01
4.183E-01
4.631 E+ 00

11.4
12.4
11
10.5
8.2
2.24
0.97

10.6
12.6
12.1
10.7
9.8
2.94
1.54
0.08
1.0457
0.024
1 .022E+00
2.773E-01
4.555E+00

2.8
2.9
2.5
2
1.5
0.4
0.09

2.8
3.3
2.7
2
1.6
0.57
0.2
0.006
0.3815
0.033
2.266E-01
3.351 E-01
1.949E+00

2.8
2.9
3
2.8
2.1
0.52
0.34

2.5
3.1
2.9
2.6
2.4
0.7
0.45
0.034
0.6164
0.057
2.565E-01
2.741 E-01
8.834E-01

2.4
2.8
2.8
2.7
2.3
1.08
0.62
0.037
2.1
2.7
2.7
2.4
2.3
1.22
0.88
0.14
0.4705
0.047
2.028E-01
1 .682E-01
9.780E-01
Vol
1,3,5-Trimet n-Decane 2-Ethyltoluen 1 ,2.4-Trimet

0.82
1
0.94
1
0.8
0.39
0.28
0.035
0.69
0.87
0.89
0.85
0.76
0.41
0.34
0.093
0.3436
0.095
6.430E-02
1.397E-01
3.976E-01

11.8
13.6
13.6
13.5
12.3
3.03
3.05
1.6
12.9
14.4
14.8
14.5
13.2
3.18
3.19
2.1
2.6649
0.050
1.200E+00
2.355E-O1
3.525E+00

0.78
0.89
0.9
0.9
0.77
0.22
0.19
0.025
0.67
0.85
0.85
0.83
0.73
0.28
0.24
0.074
0.4621
0.142
6.813E-02
1.971 E-01
2.678E-01

2.4
2.7
2.8
2.8
2.5
1.12
0.9
0.177
2
2.6
2.6
2.4
2.3
1.19
1.03
0.39
0.6482
0.064
1 .786E-01
1 .246E-01
1.301E+00
0.05276
1,2,3-Trimet

0.67
0.8
0.86
0.87
0.76
0.48
0.39
0.11
0.54
0.72
0.77
0.72
0.68
0.49
0.44
0.19
0.3053
0.097
5.047E-02
8.314E-02
3.219E-01
Emission
(mg/g)
2.66
                                     S(t) = SO exp(-M)  with CO « Concentration at t=0

                               4.55         0.98        1.11        1.51        0.61
                                                                          5.88
0.42
1.S
0.81

-------
cum
            ACH
                     1
2-MethyWec  trans-Decah n-Undecane  Pentylcycloh n-Dodecane TVOC
1.4
1.9
2
2.4
2.2
1.01
1.06
1
1.2
1.7
1.8
1.9
1.7
1.04
1.07
0.94
1.0036
0.116
1.061E-01
4.594E-02
8.669E-O1
2
2.7
2.9
3.2
2.9
1.47
1.43
0.639
1.9
2.6
2.7
2.6
2.5
1.52
1 .51
0.88
0.7867
0.068
1.704E-01
7.291 E-02
1.033E+00
6.5
8.5
8.4
9.4
9.1
2.17
2.26
2.5
6.4
6.4
9.2
8.6
8.1
2.35
2.48
2.1
2.6110
0.077
7.384E-01
2.079E-01
1.237E-01
0.64
0.87
0.94
1
0.95
0.63
0.75
0.59
0.68
0.83
0.83
0.81
0.84
0.68
0.65
0.5
0.3592
0.100
4.913E-02
2.586E-02
4828E-01
0.89
1.2
1.4
1.5
1.5
1.19
1.42
1.7
0.77
1
1.2
1.2
1.2
1.12
1.36
1.2
0.5341
0.087
6.965E-02
-1.718E-03
4.163E-01
200.9
240
240
240
220
103
95
55
170
200
202
190
160
89
66
45
9.2442
0.011
1.279E+01
9.594E-02
1.S80E+02
       3.31
3.08
3.58
2.89
1.67
220.19

-------
   100
                           Chamber Results
                        Test 13 & 14 Combined
-£>  10

₯   1
Q>
O
O 0.1
O
    m,p-Xylene
     V
    n-Nonane
     ©
    o-Xylene

    Fit
0.01 -,-
    0
                              Time (hr)
                                          10
15

-------
10
                         Chamber Results
                     Test 13 & 14 Combined
                       10        15
                        Time (hr)
20
25
                                                        Propylcyclohexane
                                                         T
                                                        3&4-Ethyltoluene
                                                         ©
                                                        1,3,5-Trimethylbenzene
                                                    ©   Fit

-------
                           Chamber Results

                        Test 13 & 14 Combined
   100
•2>  10
eo
c
o

1
0)
u
o
o
0.1
   0.01
                                                       n-Decane

                                                        v


                                                       2-Ethyltoluene

                                                        ©


                                                       1,2,4-Trimethy I benzene
                                                       Fit
                         10        15

                           Time (hr)
                                         20
25

-------
   10 -
                             Chamber Results

                          Test 13 & 14 Combined
"o>


c
o
0)
o

o
o
   0.1
                         10        15

                           Time (hr)
20
              1,2,3-Trimethyl benzene
              2-Methyldecane

          ©    ©


              trans-Decahydronaphthaline
25
              Fit

-------
                          Chamber Results
                        Test 13 & 14 Combined
    10
CO
§
c
o
o
                                                           n-Undecane
                                                            T
                                                           Pentylcyclohexane
                                                            ©
                                                           n-Dodecane
Fit
  0.01
                          10        15
                           Time (hr)

-------
                        Chamber Results

                     Test 13 & 14 Combined
   1E3
•5?
«
§  1E2
c
0)
o

o
o
                                  TVOC
                                                            Fit
                                                        25
10         15

  Time (hr)
20

-------
Chamber VOC Recovery Study TEST 21 & 22 w & w/o tan Vol 0.05276
w Fan w/o Fan
Time m.p-Xylene n-Nonane o-Xylene Propylcycloh 3&4-Ethyltol 1.3.5-Trimet n-Decane 2-Ethyltoluen 1 ,2,4-Tnmet 1 ,2,3-Trimet
hr
0.67
1.2
2.2
3.2
4.2
8.2
12.7
24.7

0.67
1.2
2.2
3.2
4.2
8.2
12.7
24.7 -•'
Sumsq
Std
SO
K
CO
Emission
Sumsq
Std
SO
K
CO
Emission
(mg/g)
7.09
19.75
16.93
9.59
5.64'
3.39
0.20


7.56
11.90
13.49
10.71
7.41
5.29
0.83
0.12

0.2676
0.006
1.137
0.693
22.990
3.39
0.2174
0.007
1.177
0.480
5.827
3.09


59.24
57.83
39.49
28.21
17.77
1.23
0.07


27.78
34.39
33.07
27.78
22.49
6.22
1.85

0.7316
0.005
5.244
0.619
39.380
11.82
0.3021
0.004
3.037
0.307
5.356
10.80


3.53
3.53
1.97
1.31
0.76
0.06



2.12
2.38
1.98
\ 1.46
1.16
0.24
0.04

0.2200
0.028
0.261
0.664
3.395
0.66
0.1062
0.018
0.193
0.389
1.167
0.65


16.93
16.93
11.28
8.32
5.36
0.47
0.03


6.88
8.99
9.26
7.28
6.22
1.98
0.70

0.4119
0.010
1.427
0.577
12.267
3.59
0.2374
0.010
0.786
0.278
0.810
2.98


1.41
1.21
0.96
0.78
0.58
0.10
0.01


0.53
0.66
0.66
0.57
0.53
0.26
0.12

0.1452
0.042
0.091
0.396
1.227
0.39
0.0539
0.033
0.049
0.177
0.232
0.34


0.42
0.65
0.59
0.44
0.35
0.08
0.01


0.22
0.26
0.33
0.29
0.28
0.16
0.09

0.1109
0.070
0.058
0.377
0.000
0.15
0.0508
0.063
0.022
0.136
0.042
0.18


76.16
76.16
80.39
70.52
66.29
12.55
2.96
0.04

29.10
41.01
39.68
38.36
38.36
14.55
13.23
0.62
2.3119
0.011
7.821
0.328
11.906
25.73
1.9158
0.018
3.198
0.173
3.785
19.63


0.39
0.38
0.28
0.25
0.18
0.04



0.15
0.17
0.17
0.16
0.15
0.08
0.05

0.0723
0.082
0.026
0.342
0.350
0.13
0.0143
0.034
0.012
0.137
0.084
0.12


1.69
1.69
1.55
1.41
1.07
0.31
0.06


0.62
0.81
0.90
0.87
0.78
0.49
0.32
0.01
0.1585
0.038
0.142
0.311
0.894
0.61
0.2827
0.119
0.070
0.161
0.000
0.44


0.42
0.42
0.44
0.38
0.31
0.12
0.03


0.15
0.19
0.21
0.21
0.19
0.13
0.11
0.01
0.0629
0.059
0.035
0.248
0.217
0.19
0.1530
0.273
0.015
0.121
0.010
0.13

S(t) o SO exp(-M) with CO » Concentration at t=0

-------
cum
           ACH
2-Methyldtc trans-Decah  n-Undecane Pentylcycloh n-Oodecane TVOC
7.05
8.04
9.03
9.17
9.73
4.65
1.69
0.02
1.72
2.51
3.17
3.04
3.31
3.17
3.57
0.93 -''
1.2009
0.047
0.803
0.208
0.000
3.87
0.8511
0.090
0.206
0.047
0.000
9.87
11.00
10.86
10.30
8.89
3.24
0.85
0.01
3.17
4.37
4.89
4.63
4.89
3.31
2.91
0.30
0.5582
0.019
0.992
0.264
2.183
4.20
0.7513
0.058
0.357
0.109
0.000
31.03
38.08
42.31
42.31
52.19
15.51
11.00
0.38
7.94
10.71
13.10
15.74
17.20
9.79
10.98
8.86
2.6503
0.019
3.583
0.185
0.000
19.39
1.3244
0.029
0.841
0.030
0.000
2.26
2.40
3.10
3.67
3.95
3.39
1.83
0.06
0.50
0.85
0.97
1.04
1.16
1.32
1.72
1.32
1.3435
0.129
0.260
0.133
0.000
1.96
0.3826
; 0.084
0.061
-0.010
0.000
3.81
4.65
5.36
6.49
7.05
8.04
7.76
1.97
0.82
1.18
1.46
1.72
1.72
2.25
3.04
6.08
1.6055
0.075
0.413
0.042
0.000
9.74
0.0922
0.006
0.079
-0.059
0.176
1086.04
1043.72
902.68
803.95
705.22
253.88
121.30
9.59
370.37
476.19
476.19
451.19
410.05
238.10
185.19
68.78
3.2797
0.001
66.556
0.210
982.508
564.20
3.7837
0.003
29.839
0.098
240.414
       4.34
3.27
28.07
1.46
1.89
433.81

-------
   100 -
                         Chamber Results
                       Fan Test (m,p-Xylene)
CO
   10
0)
o
o  0.1
o
  0.01
                       5                 10
                             Time (hr)
                                                              wFan
                                                               T
                                                              w/o Fan
                                                              Fit
15

-------
                       Chamber Results

                      Fan Test (n-Nonane)
  100
W 10
CO


1

§  1
0>
u


8  0.1
w Fan

 T:-

w/o Fan
               Fit
  0.01
                                                      —i

                                                       15
                            Time (hr)
                        10

-------
   10
                         Chamber Results
                        Fan Test (o-Xylene)
.O)
CO


"o>



o
   n
   U.
O
O
  0.01
                             Time (hr)
                                        10
w Fan

 V

w/o Fan
Fit
          15

-------
  100
                      Chamber Results

                Fan Test (Propylcyclohexane)
"3
?5
E
C»
E,
c
o
:p
CO
10 - 	 	 Jt
: v^
-
'
-
1 -
I
-
0)
o
  0.1
  0.01
                          Time (hr)
                                    10
15

-------
   10
                         Chamber Results

                    Fan Test (3 & 4-Ethyltoluene)
O)
o>


o

1
*
8 0.1
c
o
o
  0.01
                             Time (hr)
                                         10
                                                 w Fan

                                                  T


                                                 w/o Fan
Fit
          15

-------
O)
eo
E  0.1
"B)


O

1
•4-1

8 0.01
c
O
O
                           Chamber Results
                   Fan Test (1,3,5-Trimethylbenzene)
w Fan

 V

w/o Fan
                     Fit
  0.001
                         5                10
                               Time (hr)
                                       15

-------
  100
CO
E  10

o>



.0
33
CO
0)
o

o
o
                       Chamber Results

                     Fan Test (n-Decane)
w Fan

 V

w/o Fan
                  Fit
   0.1
                     5               10
                          Time (hr)
                                    15

-------
   10
                       Chamber Results
                    Fan Test (2-Ethyltoluene)
•S?
n

"o>
w Fan
 v

w/o Fan
                                                  Fit
                           Time (hr)

-------
    10
                        Chamber Results
                Fan Test (1,2,4-Trimethylbenzene)
o>    1
E
c
o
33
o
O
   0.1 -
  0.001
                          10         15
                            Time (hr)
                                                  w Fan
                                                   V
                                                  w/o Fan
                                                  Fit
                                             20
25

-------
                         Chamber Results
                Fan Test (1,2,3-TrimethyIbenzene)
CO
E   0.1
O)


o

1
g 0.01
c
o
o
  0.001
                          10        15
                             Time (hr)
     wFan

      V

     w/o Fan
     Fit
20
25

-------
100
                     Chamber Results
                 Fan Test (2-Methyldecane)
0.01
                       10        15
                         Time (hr)
20
25

-------
   100
                       Chamber Results

             Fan Test (trans-Decahydronaphthalene)
o
o
   0.01
  0.001 -r
                         10        15

                           Time (hr)
20
25

-------
100
                   Chamber Results
                 Fan Test (n-Undecane)
 0.1
                     10        15
                       Time (hr)
20
25

-------
                       Chamber Results
                 Fan Test (Pentylcyclohexane)
   10
CO
E
CD


o
O
O
   n i
   U.I
w Fan

 V

w/o Fan
                 Fit
  0.01 -r

      0
         10        15
           Time (hr)
20        25

-------
  10
                      Chamber Results
                   Fan Test (n-Dodecane)
•5?
o

f

g" 1
o
o
o
w Fan
 V
w/o Fan
                                      Fit
  0.1 -r
                        10        15
                          Time (hr)
     20
25

-------
1E4
                     Chamber Results
                     Fan Test (TVOC)
                                         wFan
                                          T

                                         w/o Fan
                                         Fit
                      10        15
                        Time (hr)
20
25

-------
LATEX PAINTS

-------
Chamber    VOC       Paint       Study
            TEST 3 & 4 Combined
Time
hr














1 ,2-propane Ethylene gly 2-(2-butoxye Texanol Formaldehyd Acetaldehyd TVOC
1.22
12.2
24.2
49.0
96.5
120.2
168.2
1.25
12.2
24.23
48.3
96.2
120.2
168.2
Sumsq
Std
SO
K
K2





0.154
0.634

0.329
0.159
0.084

0.220

0.64
0.279
0.198
0.0609
•s 0.5210
0.271
5.05E-01
2.26E-02
3.00E-03

0.718
5.046

5.73
2.40
2.29

1.032

7.50
3.74
3.43
1.71
2.3742
0.106
5.78E+00
1.41E-02
1.40E-03

0.121
0.209

0.140
0.0509
0.0700

0.138.

0.224
\ 0.119
0.107
0.0651
0.2494
0.372
3.95E-01
2.38E-02
1.90E-03
0.990
1.125
0.939

0.634
0.305
0.456
1.37
1.175

1.13
0.575
0.501
0.362
0.5334
0.117
6.62E-02
8.02E-03
9.07E-01
0.005
0.005
0.001
0.001
0.000
0.000
0.000
0.005
0.006
0.001
0.000
0.000
0.000
0.000
0.0209
0.976
5.13E-05
1.02E-02
1 .48E-02
0.043
0.002
0.002
0.001
0.001
0.001
0.001
0.028
0.002
0.001
0.001
0.001
0.001
0.000
0.0564
0.368
1.04E-04
1.14E-02
1.12E-01
1.037
2.125
6.831

6.834
2.919
2.900
1.408
2.572

9.489
4.718
4.237
2.200
2.1567
0.069
2.21 E+01
1 .87E-02
8.09E-04
Emission          2.62
(mg/g)
for texanol - acetaldehyde:
otherwise:
36.99
1.23
14.22
0.08
0.53
      S(t)» SO exp(-kt) with K2 - Concentration at t-0
      S(t)- SO exp(-kt) [1 - exp(-Wt)]
48.83

-------
100
                             Chamber Results
                           Test 3 & 4 Combined
0.01
                 50
   100         150
Time (hr)
200
                                                           1,2-propanediol
                                                            v
                                                           Ethylene glycol
                                                            o
                                                           2-(2-butoxyethoxy)-ethanol
                                                           Fit

-------
     10
                           Chamber Results
                          Test 3 & 4 Combined
c
o
I
0>
o
    0.1
0.01
O 0.001
  0.0001
                    so
                             100
                           Time (hr)
150
                                                       Texanol
                                                        T
                                                       Formaldehyde
                                                        o
                                                       Acetaldehyde
                                                    Fit
200

-------
                     Chamber Results
                    Test 3 & 4 Combined
10
                                             TVOC
                                             Fit
0.1 -,-
   0
50
  100
Time (hr)
150
200

-------
Chamber   VOC        Paint       Study
           TEST 15 & 16 Combined
Time
hr





















1,2-propane Ethylenegly 2-(2-butoxye Texanol Formaldehyd Acetaldehyd TVOC
1.3
12.3
24.3
48.3
96.3
120.3
166.3
1.3
12.3 0.048
24.3
48.3
96.3
120.3
168.3
Sumsq
Std
SO
K
K2

5.05E-01
2.26E-02
2.98E-03

2.801
6.875

5.517
4.875
2.911

3.230

13.991
5.757
5.618
3.101
1.8992
0.045
1 .03E+01
1.77E-02
2.02E-03
1.283
2.511
2.697

2.112
1.897
1.484
1.201
2.717

5.477
2.166
2.228
1.575
1.7089
0.094
4.10E+00
2.17E-O2
3.27E-03
0.200
0.170
0.150

0.120
0.100
0.072
0.200
0.180

0.310
0.120
0.120
0.078
0.3101
0.302
1.17E-02
6.17E-03
1.00E-01
0.002
0.003

0.004
0.002
0.002
0.001



0.003


0.002
0.0198
2.122
1.84E-04
5.64E-03
O.OOE+00
0.041
0.012
0.004
0.002
0.001
0.001
0.001
0.041
0.012
0.004
0.002
0.001
0.001
0.001
0.0691
0.466
8.59E-04
4.39E-02
1.04E-01
1.526
5.497
9.726

7.752
6.874
4.469
1.443
6.187

19.783
8.044
7.966
4.756
2.1820
0.033
3.12E+00
1.54E-02
1.20E-O2
Emission
(mg/g)
for texanol - acetaldehyde:
otherwise:
                             59.79
      24.66
                                                    2.76
                             0.03
0.14
S(t) - SO exp(-M) with K2 * Concentration at t=0
S(t)-SOexp(-W)[1-exp(*2t)]
88.24

-------
100
                          Chamber Results
                       Test 15 & 16 Combined
                                                        1,2-propanediol
                                                         T
                                                        Ethylene glycol
                                                         e>
                                                        2-(2-butoxyethoxy)-ethanol
                                                        Fit
0.01
                50
   100
Time (hr)
150
200

-------
     1 -
                         Chamber Results
                      Test 15 & 16 Combined
5 0.001
o
  0.0001 -T
                   50
   100
Time (hr)
150
                                                        Texanol

                                                         v

                                                        Formaldehyde
                                                         o

                                                        Acetaldehyde
                                                        Fit
200

-------
  100 -
                      Chamber Results
                    Test 15 & 16 Combined
•S?
co
   10
o

I
0>
o
o
o
                                                    TVOC
                           FK
   0.1
                 50
  100
Time (hr)
150
200

-------
Chamber    VOC        Paint       Study
            TEST 17 & 18 Combined
Time
hr















1
1.3
12.7
24.3
48.3
96.3
120.3
168.3
1.3
12.3
24.3
48.3
96.3
120.3
168.3
Sumsq
Std
SO
K
K2











i ,2-propane Etnylene gly 2-(2-butoxye Texanol Formaldehyd Acetaldehyd TVOC
4.51
25.6
30.6
27.1
5.45
0.405

2.53
21.7
25.0
17.9
5.69
1.23
0.02
3.2144
0.030
8.01 E+01
4.81 E-02
2.39E-03
0.88
8.01
11.1
11.3
5.41


0.469
6.42
8.38
7.38
4.82
1.79

1.4651
0.041
4.57E*01
3.01 E-02
9.25E-04
1.23
1.57
2.16
.1.84
1.61
0.889
0.340
0.787
1.44
1.48
1.16
1.07
1.21
0.280
1.6195
0.208
2.88E+00
2.36E-02
2.33E-03
4.33
2.42
325
2.46
1.93
142

2.99
2.27
2.13
1.54
1.28
1.59

0.9745
0.068
1.37E-01
5.04E-03
5.92E+00

3.090E-03
2.107E-03
1.138E-03
4.916E-O4


4.343E-03
3.016E-03
2.051 E-03
1 .025E-03
4.222E-04


0.0099
0.803
2.02E-04
2.46E-02
5.79E-03

2.388E-03
1 236E-03
5.758E-04



1 .532E-02
2.413E-03
1 .327E-03
4.825E-04
•


0.0064
0.169
2.02E-04
4.38E-02
4.42E-02
10.946
37.627
47.115

14.408
2.714
0.340
6.793
• 31.811

27.941
12.864
5.814
0.298
35644
0023
3.05E+02
4.15E-02
8.37E-04
Emission          79.02
(mg/g)
for texanol - acetaldehyde:
otherwise:
45.34
10.93
89.08
                                    0.02
                                    0.06
      S(t) - SO exp(-kt) with K2 » Concentration at t»0
      S(t) - SO exp(*t) [1 - exp(-k2t)l
                                                          145.07

-------
 100
                          Chamber Results
                       Test 17 & 18 Combined
                                                       1,2-propanediol
                                                        v
                                                       Ethylene glycol
                                                        o
                                                       2-(2-butoxyethoxy)-ethanol
                                                       Fit
0.01
               —I—
                50
   100
Time (hr)
150
200

-------
                         Chamber Results
                      Test 17 & 18 Combined
   10
0.0001
                  50
   100
Time (hr)
150
200
                                                           Texanol
                                                            v
                                                           Formaldehyde
                                                            ©
                                                           Acetaldehyde
                                                           Fit

-------
                       Chamber Results

                    Test 17 & 18 Combined
  100
I10
c
o
0)
o
c
o
o
                          TVOC
                                                   Fit
   0.1
                50
  100

Time (hr)
150
200

-------
Chamber
Time
hr
1.3
12.7
24.3
48.3
96.3
120.3
168.3
1.3
12.3
24.3
48.3
96.3
120.3
168.3
Sumsq
Std
SO
K
K2
Emission
(mg/g)
Sumsq
Std
SO
K
K2
Emission
(mg/g)
VOC Paint Study
TEST 1 9 & 20 Combined 1 9 w fan 20 w/o fan
w Fan w/o Fan
1 ,2-propane Ethylene gly 2-(2-butoxye Texanol Formaldehyd Acetaldehyd TVOC
0.0421
0.840
0.773
0.342
0.062
0.015
0.005
0.030
0.48
0.59
0.38
0.17
0.094
0.015
-'' 0.3076
0.150
3.33E+00
5.38E-02
1.67E-03
1.86

0.1819
0.126
1.94E+00
3.63E-02
1.46E-03
2.06

0.174
6.632
7.200
4.499
2.221
0.900
0.102
0.10
3.6
5.4
4.1
3.1
2.6
0.80
1.2110
0.069
4.01 E+01
3.49E-02
7.82E-04
25.14

0.8359
0.069
2.40E+01
2.32E-02
6.90E-O4
29.96

0.097
0.291
0.270
0.152
0.082
0.047
0.018
0.043
0.21
0.20
0.15
0.10
0.080
0.046
0.0680
0.095
1.98E-02
1.74E-02
3.50E-01
1.08

0.0309
0.061
1.31E-02
9.85E-03
2.25E-01
1.27

1.604
1.298
1.142
0.698
0.437
0.316
0.182
1.1
0.96
0.86
0.65
0.46
0.43
0.31
0.1304
0.033
7.86E-02
1.34E-02
1.94E+00
13.51

0.0951
0.035
5.36E-02
7.91 E-03
1.37E+00
15.90

2.738E-03
1.533E-03
6.243E-04

7.448E-04
3.614E-04

2.689E-03
2.358E-03
3.491 E-04
2.736E-04
2.547E-04


0.0103
1.890
3.82E-05
3.85E-03
7.97E-03
0.12

0.0019
0.360
1.91E-05
4.11 E-03
8.48E-03
0.11

1.479E-02
1.544E-03
1.533E-03




2.866E-02
1.613E-03
1.198E-03




0.0000
0.001
8.20E-05
5.67E-04
4.91 E-02
4.72

0.0000
0.000
1.12E-04
2.48E-02
9.48E-02
0.21

1.934
9.064
9.388
5.691
2.803
1.278
0.306
1.314
5.239
7.058
5.319
3.861
3.189
1.170
0.5950
0.026
8.24E-01
2.10E-02
1 .24E-01
33.64

07908
0.046
5.08E-01
1.10E-02
9.54E-02
41.47

for texanol - acetaldehyde:
otherwise:
S(t) • SO exp(-M) with K2 « Concentration at t=0
S(t) = SOexp(-W)I1-exp(-k2t)]

-------
                          Chamber Results

                     Fan Test (Ethylene Glycol)
    10

en
8
C
o
o
                                                               w Fan

                                                                T


                                                               w/o Fan




                                                               Fit
  0.01 -r
                   50
   100

Time (hr)
150
200

-------
  10
o


O)


c
O
0)
O

O
O
                    Chamber Results

                     Fan Test (TVOC)
                         w Fan

                         v

                         w/o Fan
                                                 Fit
  0.1
               50
  100
Time (hr)
150
200

-------
           APPENDIX E




METALS DATA PROVIDED BY EPA/NERL

-------
SEP-29-93 WED 10:39
NtKL
                                         ui
        date:   9/28/93                                          XRF-222

        re:  XRF  Screening  Report  of  HQ Pain£  Sample*

        from:   Dr.  T. M.

        to:  Donald H. Whitaker

        Twenty  samples  were  .submitted  for   screening for heavy metals
        using the Kevex XRF analyser.    Samples  were  homogenized ar.d ar.
        aliquot was analysed  using   target  14 and target $2 conditions.
        Quantitation  was   performed  using  a low  level  AA  reference
        solution  (6-120  ppm).    This  standard  contained PB,Cd,Se,As,
        Cr,Ag,Tl,Mn,Zn,Cu,  Ni. Co, V  and Sb).   Quantitation of Ca, Fe, :,
        Zr and  Br  was  done using standards  prepared  in our lab.  Only
        elements  found above instrument background are reported here.
Sample f  Field ID       Cr*  Mn   Ni   Cu   Zn

21221     First Star     <5   15   10   9    6
21222     Bumbershoot                   24
21223     Sherrif's Star      15        65
21224     Violet Veil              5    10   5
21225     Hyacinth                      66   5

21226     Crescent Green 10        769
21227     Coral Canyon             596
21228     Grass Roots                   435
21229     Praline        9         4    10   9
21230     Marmalade      9         7    5    34

21231     Tomahawk       13        474
21232     Dawn Yonder    7         8    172  8
21233     Vibrant Violet      7    7    49   69
21234     Orange Ice               5    75
21235     Ice Cap                  5    34   4

21236     Orange Glaze             587    <10
21237     Rose Dawn                10   7    48   <5
21238     Chim Cham      15   12   7    8    7    10
21239     Seafoam                      [5]   2680 <10
21240     Antigua             15   7    15   8    12
                                                          Pb   Bi
                                               Co
                                   30
                                   33
                                   21

                                   10
                                   30

                                   12
                                   <5***

                                   <5
                                   9
                                   12
                                   12
                                                               10

-------
SEP-29-93 WED 10:40
  EPA NERL
                 NU.  DUOOU40SI
                               r. uo
        Lab ID

        21221
        21222
        21223
        21224
        21225

        21226
        21227
        21228
        21229
        21230

        21231
        21232
        21233
        21234
        21235

        21236
        21237
        21238
        21239
        21240
Ca%

10.5


4.6
10.5
5.0

2.0
5.4

1.9
10.0
8.7
15.5
14.6
40.0
50.0
28.5
45.0

14.0
29.8
6.8
10.0
6.7

1.2
1.5
9.6
15.8
16.1

34.0
27.0
11.2
26.0
15.0
Pe

540
500
3050
470
150

2500
3800
2280
6080
760

6800
640
504
308
1060

6840
240
1500
950
630
                    Br
700
          Zr   Hg**
          250

          310
1.7%
34             10

     1400
                              16
                                   1600
          140
          340
          100
        * All values in ppm by weight unless noted otherwise.

        **   Hg values are possibly  low because of evaporation of organic
        mercury compounds while  samples are in the analytical chamber.
        However, Sample 21226  (reported to have  high Eg)  was rerun vith
        no  time   delay  for    evaporative  loss  and  still  showed  nc
        ir.easureable level of Kg  above 5 ppm.

        No detectable levels of  V, As, Se, Sn, ( see note below)  Mo, Cd,
        Sb or  Ba.   Al not quantifiable on Kevex instrument.  No attempt
        made to measure Cl or  S.

        Sample 21226 was reported to have high Sn.  This sample was rerun
        for five  times as  long to  recheck Sn level.  No Sn was detected
        at the 5 ppm level.

        *** In some samples detection limit is lower  than others because
        of absence of interfering elements.

-------
SEP-29-93 WED 10:40     EPA KERL
fHA MU.  01(ODUH03I
                                i • ui
          Listing of samples for XRF Analysis:
                                  GL6987-20573-LAGO-05  (ORANGE  ICE)
                                  GL6918-16112-LAGO-23  (ORANGE  GLAZE)
                                  GL3480-010U-LCFR-27  (TOMAHAWK)
                                  GL6300-64542-LASB-04  (ICE CAP)
                                  GL6380-6A98A-LASB-24  (DOWN YONDER)
                                  GL8000-46212-ADSG-25  (SEAFOAM)
                                  GU550-76262-ADGP-27  (HYACINTH)
                                  GL5700-25312-ADFT-26  (CHIH CHAR)
                                  GL*550-20852-ADGX-06  (SHERIFF'S STAR)
                                  GL5718-34722-ADFG-09  (ANTIGUA)

                                  SW200-1734-LVFG-08    (GRASS ROOTS)
                                  SW200-1604-LCG8-28    (ROSE DAWN)
                                  SW200-1545-LCSP-23    (VIBRANT VIOLET)
                                  SW200-1629-LVFO-24    (MARMALADE)
                                  SW200-1175-LCSX-03    (PRALINE)
                                  SW200-1435-ADGG-28    (BDMBERSHOOT)
                                  SW200-1529-ADSB-28    (VIOLET  VEIL)
                                  SW200-1003-ADFX-03    (FIRST STAR)
                                  SW200-1352-ADFT-21    (CRESCENT CREAM)
                                  SW200-1309-ADSX-03    (CORAL CANTON)

-------