EPA-650/2-74-107

OCTOBER  1974
Environmental  Protection Technology Series
  »•:••:•.••••

                     •;»>:•:•:
                     W




                               8i?:§£^

-------
                              EPA-650/2-74-107
      CHARACTERIZATION
OF ATMOSPHERIC  EMISStONS
FROM  POLYURETHANE  RESIN
         MANUFACTURE
                  by

       Wayne E. Sinith and John R. LaShelle

          Midwest Research Institute
            425 Volker Boulevard
          Kansas City , Missouri 64110
           Contract No. 68-02-0228
                Task 38
            ROAP No. 21AXM-60
          Program Element No. 1AB015
         Task Officer: Belur N. Murthy

          Control Systems Laboratory
      National Environmental Research Center
    Research Triangle Park, North Carolina 27711
               Prepared for

     OFFICE OF RESEARCH AND DEVELOPMENT
    U.S. ENVIRONMENTAL PROTECTION AGENCY
          WASHINGTON, D.C.  20460

               October 1974

-------
This report has been reviewed by the Environmental Protection Agency
and approved for publication.  Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
                                  11

-------
                                CONTENTS

                                                                     Page

List of Figures	   v

List of Tables 	

Acknowledgements 	

Summary and Conclusions	   1

Sections

   I      Introduction 	   3

  II      Site Selection	   5

 III      Experimental Plan	   6

  IV      Process Data and Their Analysis	21

   V      Conclusions	37

  VI      Recommendations.	38

Appendix A - List of Companies Producing Urethane Coating Resins
               for Either Captive Use or for Sale or Both:
               Companies and Plant Locations 	  39

Appendix B - Analytical Methods for TDI	45

Appendix C - Reference Material on Gaussian Plume Dispersion
               Model from EPA Workbook AP-26	52

Appendix D - Mass Spectrometry Data for Number 5 Outlet Sample
               and Number 302 Mineral Spirits Sample 	  59

                                   iii

-------
                                 FIGURES

No.                                                                 Page

 1    Photograph of General Test Area and Mobile Laboratory and
        Schematic of the Test Site	    8
 2    Mobile Laboratory Interior	    9

 3    Flow Diagram for Polyurethane Reaction System 	   10

 4    Inlet and Outlet Sampling Manifolds 	   14

 5    Sampling System at Test Site	15

 6    Sampling Equipment	16

 7    Scrubber Outlet Manifold	18

 8    Test Schedule and Related Activities	19

 9    TDI + Amine Concentration (colorimetric)  Versus  Time	26

10    Infrared Spectrum of Liquid From No. 5 Outlet	28

11    Infrared Spectrum of No. 302 Mineral Spirits	29

12    Simplified View of Test Site and Vicinity (not to scale).  .  .   34

13    Schematic of Test Site (overhead view)	35
                                    IV

-------
                                 TABLES

No.

 1   Selected Characteristics of Chemicals Used in Polyurethane
       Resins Manufacture	12

 2   Sample Schedule 	  20

 3   Raw Data	22

 4   Test Results - TDI Analyses (Samples Nos. 0-3)	24

 5   Test Results - TDI Analyses (Samples Nos. 4-7)	25

 6   Gas Chromatograph Retention Times 	  30

 7   Results of Dispersion Calculations	32

-------
                               ACKNOWLEDGEMENT

 The work reported herein was conducted by Midwest Research Institute (MRI),
 pursuant to a task order issued by the Environmental Protection Agency
 (EPA) under the terms of Contract No. 68-02-0228.  Dr. Wayne E. Smith
 served as the project leader.  He was assisted by Dr. James Spigarelli,
 Messrs. Thurman Oliver and Fred Bergman in the chemical aspects of field
 testing and the analyses of the samples; by Dr. George Scheil and Mr. Fred
 Bergman in development of sampling systems; by Dr. George Scheil and
 Messrs. Thurman Oliver, George Cobb, Bruce DaRos, William Maxwell and
 Bob Kamerman in field testing; by Dr. Chatten Cowherd, Jr., assisted by
 Ms. Christine Guenther, in selecting the dispersion model and perform-
 ing the dispersion calculations; and by Mr. John LaShelle in compiling
 this report.
 Approved for:

 MIDWEST RESEARCH INSTITUTE

         /
       e-rf  i
.
•^Paul C. Constant, Jr.
 Program Manager
 29 August 1974
 s Iw                                 vi

-------
                         SUMMARY AND CONCLUSIONS

Midwest Research Institute arranged for and conducted tests at a poly-
urethane resin facility to help answer the question as to whether or
not this industry has an air pollution problem.  Requirements for con-
ducting meaningful tests include:

1.  Adequate analytical methods,

2.  Adequate sampling techniques,

3.  Accurately monitored process variables,

4.  Raw data for dispersion modeling, and

5.  Scrubber efficiency data.

After the necessary negotiations, literature and laboratory work were
completed, a mobile laboratory was taken to the test site where samples
were analyzed both on-site and on a delayed basis at MRI.  In the process
studied, nitrogen continuously sweeps through the reactor containing
alkyd resin, toluene diisocyanate (TDI) and mineral spirits.  Samples
were taken of the "sweep" gas as it left the reactor and also after it
had passed through a scrubber whose major function was to hydrolyze
TDI to amines.  On the inlet side of the scrubber, a manifold was in-
serted to take samples through an impinger for colorimetric measurement
of TDI; cryogenic traps and an MDA-tape analyzer for TDI were also in-
serted.  The outlet sampling manifold was the same as the inlet manifold,
except evacuated bulbs were used in place of the cryogenic traps.  Cryo-
genic traps could not be used in the outlet system because of the high
moisture content.

Emissions are greatest during the first part of the reaction when an
exotherm occurs (from 150 to 200°F) and TDI concentration is the highest.
Analytical data show both the inlet and outlet scrubber streams contain
mainly nitrogen, mineral spirits, and TDI.  The aqueous scrubber removed
very little TDI during these tests.  Scrubber outlet composition during
the peak 1-hr emission period averaged:

-------
               Nitrogen                 97.5% by vol
               Mineral spirits           2.5% by vol
               TDI                       1.2 ppm by vol

Applying the Gaussian plume dispersion model presented in EPA workbook
AP-26 to these analytical data, one can predict that the maximum ground-
level concentration of TDI and mineral spirits should be only about 0.1%
of the maximum allowable value for an 8-hr averaging period.  Total emis-
                              -8
sion of TDI was about l'B x 10	lb TDI  .
                        Pounds Product

Based on this study, one can draw the following major conclusions:

1.  The maximum ground-level concentration of both TDI and mineral  spirits
is about 0.1% of the maxima allowable for an 8-hr averaging period, thus
indicating that no hazard existed at the test site under the conditions
studied.

2.  Statement No. 1 is true in spite of the fact that the TDI scrubber
was operating poorly* at the time of sampling.  This would indicate that
scrubbers are not needed for TDI removal if the specific process and con-
ditions which were studied are kept under control.

3.  Should several such plants be clustered in or around a common com-
munity, as is frequently the case in the chemical processing industry, a
problem could result.

From the conclusions, it is recommended that:

1.  Other test sites be studied under a broader range of operating  con-
ditions to determine air emissions.

2.  At present, caution should be exercised in extrapolating the data
presented herein to the entire polyurethane coatings industry, since it
is specific to one plant.

3.  Priorities should be extended to other areas of the plastics industry
having potentially high hazardous/toxic emissions.

4.  A search for possible clustered polyurethane sites might be in  order;
this might also apply to other plastics processes.
*  Reported by plant site personnel subsequent to MRI sampling.

-------
                               SECTION I

                              INTRODUCTION

The Environmental Protection Agency (EPA) recognizes the plastics indus-
try as a potentially serious source of air pollution.  In a recent study,
Foster D. Snell, Inc., determined the most likely resin manufacturing
processes that contribute to hazardous/toxic objectionable emissions.
Through a series of semiquantitative calculations they used a "priority
decision model" to show that the four industries, listed in decreasing
order of priority below, were at the top of the list insofar as potential
air pollution is concerned:

1.  Polyurethanes

2,  Acrylics

3.  Alkyds

4.  Phenolica

The priority ratings were arrived at by including only resin manufacturing
considerations and not such factors as monomer synthesis, polymer fabrica-
tion, etc.  In the case of polyurethanes, foams were specifically excluded.

As the above-described findings were being made, MRI was contracted through
Task No. 38 of EPA Contract No. 68-02-0228 to conduct the field work asso-
ciated with the task under consideration.  On 8 November 1973, a joint
meeting of EPA, Foster D. Snell, Inc., and MRI personnel was held to initiate
the field-testing phase of the program.  Consistent with mutual agreement
Snell was to be contacted before making industrial contacts; it was also
agreed to limit early contacts to polyurethanes.  In compliance with these
agreements, MRI personnel proceeded to contact the following companies
suggested by Foster D. Snell, Inc.:  Ashland Chemical Company, Diamond
Shamrock Company, and Witco Chemical Company.  Ashland Chemical Company
management reported that it used only a low level of toluene diisocyanate
(TDI) as a viscosity builder and believed its operation to be atypical
for study.  Diamond Shamrock had sold its operations to Stephan Chemical
Company; only know-how, and not facilities, was sold.  Witco Chemical

-------
Company admitted to being a producer of polyurethane coatings and promised
some stack composition data as soon as it had a chance to make the proper
analyses.  Witco did not seem willing to admit anyone for tests, and its
own analyses have not been forthcoming.

As the next approach, MRI obtained permission from Foster D. Snell, Inc.,
and EPA to contact other companies for potential emission studies.  At
this point, MRI was allowed to guarantee complete anonymity to the com-
pany where the tests were to be run.  This approach was ultimately success-
ful and was the way admittance was obtained to the test site described in
this report.

Midwest Research Institute also asked EPA to contact Mobay Chemical Company
as a potential test site.  Mobay seemed willing to cooperate; however,
project funds did not permit testing at a second polyurethane plant at
this time.  The approach used with Mobay might well be the easiest way
to obtain admittance to test sites in the future.

As soon as MRI received word that studies migirt be permitted at an "anony-
mous" location, a meeting was held with the industrial concern involved
and the potential test site was visited.  While negotiations were uncer-
tain at this point, a formal proposal (in writing) by MRI solidified ad-
mittance to the final test site which must remain anonymous.  The actual
testing and necessary preparatory data were obtained on 20, 21 and 22 May
1974, during a typical run producing a urethane coating resin.

The following section gives several further comments on site selection,
before the overall schedule of the project is discussed in Section III.
All aspects of experimental details including a description of the test
site, technical considerations of both an equipment and analytical nature,
and test schedule are present in Section IV.  The process studied and
analyses are given in Section V, followed by Conclusions and Recommenda-
tions, which are given respectively in Sections VI and VII.

-------
                               SECTION II

                             SITE SELECTION

Foster D. Snell, Inc., provided MRI with a list of companies producing
urethane coating resins for either captive use, or for sale, or both.
This information is given in Appendix A; locations of plants are also
listed.  The companies have been alphabetized for easy reference.  The
first thing one notes is that there is a large variety of companies—large
and small—producing urethane coating resins.  Obviously, there is ru>
typical recipe and no typical set of operating conditions, because each
producer has his own group of proprietary products made in his own equip-
ment.  This makes it very difficult to extrapolate data from one plant
and caution indeed needs to be exercised in this regard.  On the other
hand, it was felt that information from one carefully selected plant might
give an indication of whether there is a problem with polyurethanes in
general.

From conversations with producers of polyurethane resins and fromMRI's
own technical information (see Section IV), we know that the most dangerous
chemical used during the manufacture of polyurethane coating resins is
toluene diisocyanate.  Furthermore, no major producer of such resins is
known to operate without a closed reaction system.  The task's goal thus
becomes fairly simple:  to run tests at a site where major quantities of
TDI are handled in a closed system producing urethane resins.

-------
                               SECTION III

                            EXPERIMENTAL PLAN

TEST REQUIREMENTS

Several requirements were necessary to help answer the basic question:
"Does the polyurethane coatings industry have an air pollution problem?"
Among them were:

1.  Adequate analytical methods for quantifying the major constituents
of polyurethanes air emissions were needed.  This required some literature
searching as well as considerable laboratory work to check out the methods
to be used.

2.  Adequate sampling techniques needed to be worked out which would yield
representative samples.  Considerable experience at MR! on related pro-
grams assisted in this area which is described in detail later in this
section of the report.

3.  Temperature, pressure and stack velocity had to be monitored to define
process variables and calculate results.

4.  Sufficient chemical, process and geometric data were needed to deter-
mine the dispersion of air emissions.

5.  Some idea of the efficiency of the cleanup or scrubber system (if
used) was desired.

6.  Before any test of this type can be successful, the proper mix of
personnel and talents was needed.  Thus, a blend of chemical engineers,
mechanical engineers, analytical chemists, polymer chemists and technolo-
gists were used to adequately pursue the goals of the project.

TEST SITE

The analyses were conducted partially on-site and partially on a delayed
basis by collecting samples to bring back to MRI.  Delayed analyses were

-------
required for some analyses which could not be performed in the field.  A
photograph showing the general test area with the MRI mobile laboratory
parked at the test site, and a schematic of the test site are given in
Figure 1.  The schematic shows the location of the scrubber system, stack,
unloading station, etc.  Because of the complexity of the roof line, sev-
eral simplifications have been made.  Although not immediately obvious
from the drawing, the TDI unloading inlet is located near the entrance to
the reaction area where the scrubber is also located.  The interior of the
mobile laboratory is shown in Figure 2.

Figure 3 shows a block diagram of the process itself.  Charge components
for the urethane resin reactor consist of alkyd resin, mineral spirits
and TDI.  The alkyd resin (previously diluted with mineral spirits) is
charged to the reactor first and then preheated to about 150°F.  Nitrogen
pressure (about 2 oz) is maintained on the reactor at all times, with the
vented gases going through the scrubber system.  At 150°F, TDI is added
and reaction commenced; an exotherm to about 200°F usually occurs during
the period subsequent to TDI addition.  The reaction is followed by with-
drawal of aliquot portions directly from the reactor which are checked for
viscosity.  As soon as the proper viscosity is reached, the reaction is
considered ready to terminate through a quench of alcohol and water.  The
quench removes the last traces of TDI.  During the tests conducted in this
study, gas stream samples were taken at the points indicated in Figure 3
at Sample Points Nos. 1 and 2.  Samples from the former have no meaning
insofar as air emission is concerned, but are needed to calculate scrubber
efficiency.  Ambient samples were also taken near the scrubber.

TECHNICAL CONSIDERATIONS

Pertinent Variables

In the reaction under study, the major process variables influencing gas
emissions to the atmosphere are:

1.  Reaction temperature,

2.  Nitrogen purge rate,

3.  Reaction rate (uptake of TDI),

4.  Scrubber design and efficiency,

5.  Diluent type and amount,

6.  Alkyd type and amount, and

7.  TDI amount.

-------
35'
                      _ . North
                                                TDI Vent-

                                                s'1 Diq.
                                                 Entrance
                                               Scrubber
                                               System
                                              MDA
                                              Monitor

                                            EL	n
                                                  Unloading

                                                  Inlet
        Figure 1.   Photograph of general test area and mobile laboratory

                           and  schematic of the test site.
                                            8

-------
Photograph A.  Colorimetric
      test equipment
  Photograph B.  Gas chromatograph
Photograph C.  Colorimetric
      test equipment
Photograph D.  Combined colorimetric
   equipment and gas chromatograph
               Figure 2.  Mobile laboratory interior.

-------
                                                                                                      To Stock
N2-tx>-j
                                                                      Discharge
                         Figure 3.  Flow diagram for polyurethane  reaction system.

-------
It would appear that all of these factors are easy to establish and con-
trol.  In actuality, the exotherm that occurs after the addition of TDI
varies somewhat depending on how "old"* or "new"** the TDI is when it is
added.  Thus, a "hot" reaction normally results from "new" TDI, while some
coaxing (heat input) of the reaction is needed whenever "old" TDI is used.
The amount of TDI that is free to escape obviously varies between "old"
and "new" TDI.  Test data reported herein were obtained using "new" TDI
and, thus, the reaction period was somewhat shorter than the 24-hr period
obtained using the older TDI.

Constituents and Their Behavior

Of the major charge components, the alkyd resin can vary greatly in com-
position; however, the one used in this study was made using glycerine,
linseed, pentaerythrytol, phthalic anhydride, xylene, and mineral spirits.
It recently came to our attention that alkyd resin plant emissions were
studied in detail under Contract No. 68-02-0259.  While that information
is not readily available to MRI, data from a private source show that the
major component coming from an alkyd plant is xylene.  The behavior of
this solvent is well documented.

Turning more to the urethane resin process under study, it is informative
to look at TDI itself.  First of all, selected characteristics of chem-
icals used in polyurethane resin manufacture are given in Table 1.  TDI
has a relatively high boiling point (250°C), a low odor threshold (0.4
ppm), and an even lower maximum allowable exposure concentration (8-hr
weighted average) of 0.02 ppm.  It is obviously of far greater toxicity
at low concentrations than mineral spirits, which have a maximum allowable
exposure (8 hr) of 500 ppm.

Chemical Analyses

From a presurvey of the process under study, it was evident that TDI (plus
its hydrolysis products)  and mineral spirits would probably be the major
constituents in the streams analyzed.  Thus, the best and most reliable
means for determining low to moderate concentrations of these components
were sought.  Insofar as TDI is concerned, it can be detected at 0.0035-
0.070 ppm levels on 40-liter air samples using a method first reported by
Marcali—  and presently sanctioned by the Department of Health, Education,
and Welfare.  The only apparent shortcoming of this method is that it is
not specific to TDI but determines both TDI and amines.
*   Presumably part of the isocyanate has already reacted.
**  Recently received from supplier.
II  Marcali, Anal. Chem.. 29_, 552-558 (1957).

                                    11

-------
       Table 1.  SELECTED CHARACTERISTICS OF CHEMICALS USED  IN POLYURETHANE RESINS MANUFACTURE
Toluene diisocyanate (TDI)
  Odor
Threshold

 0.4 ppm
4,4'-Diphenylmethane diisocyanate (MDI)    N.A.


Hexamethylene diisocyanate                 N.A.

4,4",4"-Triphenylmethane triisocyanate     N.A.

Mineral spirits                            N.A.
 Maximum Allowable
     Exposure
(8-hr weighted avg)—'

     0.02 ppm
     0.14 mg/m3

     0.02 ppm
     0.2 mg/m3

       N.A.

       N.A.

     500 ppm
     Physical Datji

250°C boiling point
< 1 mm Hg vapor pressure

194-199°C boiling point
          N.A.

          N.A.

          N.A.

150-210°C boiling point
a/  Values reported in Threshold Limit Values of Airborne Contaminants and Intended Changes,
      adopted by American Conference of Governmental Industrial Hygienists (1970).

-------
A continuous tape apparatus manufactured by MDA Scientific of Park Ridge,
Illinois, apparently overcomes this shortcoming and detects TDI at levels
as low as 20 ppb while meeting OSHA criteria.  In the present study both
methods were used to complement one another.

Mineral spirits were analyzed by gas liquid chromatography, mass spectrom-
etry and infrared spectroscopy.  The goal was not to identify individual
components of the mineral spirits, but rather to identify the mixture of
hydrocarbons on a "fingerprint" basis.  A further goal was to make certain
that other major organics were not hidden in the complex mixture.

Details of the various analyses are given in Appendix B.

SAMPLING

Equipment

A block diagram showing the overall sampling scheme is shown in Figure 4.
On the inlet side of the scrubber, the manifold was connected to the MDA
monitor, a cryogenic trap and a bubbler.  Temperature and pressure in the
gas stream were also monitored.  A photograph of the inlet sampling sys-
tem is shown in Figure 5-D.  Outlet gases were sampled through a similar
manifold shown in Figures 5-A, 5-B, and 5-C.  The main difference between
the inlet and outlet manifolds was the substitution of a preevacuated
bulb for the cryogenic trap at the outlet; this was necessary because a
cryogenic trap could not be used on the water vapor-saturated outlet side.

Procedures

TDI-Colorimetric Bubbler Trains - The impinger or bubbler (Figure 6-B)
was rinsed with absorbing reagent and filled with 15 ml of the same re-
agent.  For analysis, the bottom of the first impinger only was removed.
A complete discussion of this procedure is given in Appendix B.

TDI-MDA Tape - Two MDA tape monitors were mounted in enclosed, air-purged
boxes.  Approximately 3-ft lengths of 1/4-in. Teflon tubing connected
each monitor to the manifold valve (Figure 4).  A complete discussion
of this procedure is given in Appendix B.

Cryogenic Traps - Stainless steel traps (Figure 6-C) were cleaned with
chloroform, the open ends were capped, and then the traps were pumped
down.  The traps were shipped to the test site evacuated.  At the test
site, each trap was numbered, uncapped and a reducing fitting (1/2 in.
x 1/4 in.) attached to the valve outlet.
                                   13

-------
Water
Manometer
From Inlet
Side of
Scrubber
 Vacuum
 Pumps
From Outlet
Side of
Scrubber
Water
Manometer
                                                                            Pre-Evacuated
                                                                            Bulb
                 Figure  4.  Inlet and outlet sampling manifolds.

-------
Photograph A
Photograph B
Photograph C                                   Photograph D




        Figure 5.  Sampling system at test site.




                          15

-------
                                                                                     Inlet
                                                                                     Manifold
               Needle Adapter
               and Syringe Needle
Septum
Rubber
 Stainless Steel
 Shutoff Valve
 and 1/2 "x 1/4"
 Reducer
                                                                                          1/2 "x 1/4"
                                                                                         Adapter
                                                                                      Liquid Nitrogen"
                                                                                          l/2"Stainless
                                                                                          Steel
                                                                                         Dewar  Flask
 Pre-evacuated Bulb

       (A)
Midget
Impingers

  (B)
Cryogenic Trap

   (C)
                                          Figure 6.   Sampling  equipment,

-------
The same needle flow control orifice--standardized at 1 liter/min--was
used for all tests.  The uncapped end was fitted to the manifold via a
1/2 in. x 1/4 in. reducing union attached to the trap.  A short section
of Teflon tubing between the manifold valve and the reducing union pro-
vided thermal isolation to prevent excessive condensation.

For each test, the manifold and trap valves were opened, a rotameter in-
serted on the outlet side of the trap, and the needle orifice inserted
through the wall of 1/2 in. I.D. rubber tube which was connected to a
vacuum pump.  After obtaining the proper flow reading, the rotameter was
removed and the sample period started by immersing the trap into liquid
nitrogen.  At the end of sampling the valves were closed and the trap re-
moved and capped.

Evacuated Bulbs - Prior to shipment, new septa were installed on the bulbs
(Figure 6-A) and the bulbs were pumped down to a stable high vacuum.  Each
bulb was again pumped down immediately before sampling.  For sampling, a
22-gauge hypodermic needle connected to the manifold valve (Figure 7) was
inserted through the septum.  The valve was opened for 30 sec, then closed
and the bulb removed for analysis.

TEST SCHEDULE

The test schedule in relation to related activities is shown in Figure 8.
Before the test actually began, it was necessary to install monitoring
equipment including manifolds and to make a "dry run" to insure that all
equipment was functioning properly.  TDI was delivered to the plant just
prior to the run, and, since the TDI storage tank is vented through the
scrubbers, it was possible to check out the concentration of TDI.  During
this time it was discovered that the nitrogen plus TDI coming from the
storage tank wa§ not moving at the desired velocity.  In addition, water
was not flowing properly through the scrubber system.  Once these and
several other preliminaries were taken care of, the actual tests were
begun without further difficulties.

A copy of the sampling schedule is shown in Table 2.  The time scale was
compressed when compared to the one presented in the "work plan" because
reaction time was about 5 hr rather than the anticipated 24 hr.  As men-
tioned previously, shorter reaction times are frequently observed with
fresh or "new" TDI, as was used in this particular run.  As a result of
the decrease in reaction time, far fewer samples were taken (or needed)
than originally thought necessary to define the system.
                                    17

-------
                          WATER MANOMETER
MDA MONITOR
                                                                               BUBBLERS
                                              PRE-EVACUATED BULB
                 OUTLET MANIFOLD




                         Figure  7.   Scrubber outlet manifold,




                                        18

-------
ACTIVITY
Preparation
at MRI
Travel & Set-Up
at Test Site
Analysis
Test *
Tear Down &
Travel After Test
W








T








F








S








S








M








T





-


w








T








F








Tl
S








ME
S








IN
M








DX
T








WS
W








T








F








S








S








M








T








W








T








F








S








S








M








*5 hrs. , 18 min
                           Figure 8.  Test  schedule  and  related activities,

-------
                                           Table 2.  SAMPLE SCHEDULE
     Sample Point    Components Sought
                          Time
                          (min)
                   Total
                    No.
                                                                 Sample Train    Analytical Method
Inlet to
  scrubber
                     Major organics
TDI and related
  amines (through
  hydrolysis)

TDI
+9,-' +32, +71,
+161, +173

-12, -1 +5, +18,
+73, +118, +223
                                          Continuous
                                                 Cryogenic
                            Impinging
                              solution
                                                   GC
                                                                                       Colorimetric
                            Heated line     MDA tape analyzer
ro
o
Outlet to
  scrubber
                     Major organics
-12, +3, +16,
+62, +107, +211
                                          6-/
Evacuated
  bulb
                                                   GC
TDI and related
  amines (through
  hydrolysis)
-9,  +3,  +23,  +73,
  +118,  +223
                                                                 Impinging
                                                                   solution
                                                                                       Colorimetric
                     TDI
                     Continuous
                            Heated line     MDA tape analyzer
     a/  Time is relative to onset of TDI charging at 1747; times are medians  of  sampling periods.
     b/  Includes duplicates.

-------
                               SECTION IV

                     PROCESS DATA AND THEIR ANALYSIS

PROCESS DATA

A compilation of raw process data taken during the run is given in Table 3.
Samples were labeled from 0 to 7 in order of increasing time.  Pressures,
temperatures, times and flowrates are given for outlet and inlet samples,
as required.

MATHEMATICAL FORMULAE

TDI^Colorimetric

Concentration of TDI (plus amines) in the effluent stream were computed
from the equation:


TDI (plus amines), ppm (V/V) = TDI (plus amines), ug   22.4 fl/mole
                                  Sample Volume, H     174 g/mole
                               760 mm Hg  x  273 + T
                               738 mm Hg      273

Micrograms (ug) of TDI were obtained colorimetrically according to the
accepted procedure in Appendix B; sample  volume was measured by  flowmeter.
Grams of TDI (plus amines) emissions for  each sample period were calculated
from:
 TDI (plus amines), g = TDI (plus amines), ug ^ Effluent pl    £/min
                           Sample Volume, i
                        x Sample Time, min x 10
                                    21

-------
                                 Table 3.  RAW DATA

Process Parameters
Kettle temperature, °F
Time
Inlet pressure, mm 1^0
Outlet pressure, mm 1^0
Inlet temperature, °C
Outlet temperature, °C
Room temperature, °C
Barometric pressure, mm Hg
Cryogenic Samples
Time on
Time off
Flow at start, cm-^/min
Flow at stop, cm^/min
Evacuated Bulb Samples
Time on
Pump down pressure, monthly

0
146
1735
7
2
28
28
37
738

—
--
--
--

1735
707

1
145
1750
11
3
43
28
34
738

1752
1801
940
--

1750
707

2
150
1800
12
8
44
29
34
738

1810
1848
940
--

1803
707
Sample
3
176
1850
13
8
55
34
32
738

1813
1934
940
--

1849
707
No.
4
190
1930
18
8
43
34
30
738

1939
2118
940
900

1934
707

5
202
2115
18
13
44
18
19
738

2123
2157
940
900

2118
707

6i/ 7i/
..
2140 2240
..
-.
--
._
_.
--

-_
__
__
--

..
._
MDA Tape Monitor

Inlet TDI ppm
Flow, inlet, cm-Vmin
Outlet TDI, ppm
Flow, outlet, cm-3/min

Impinger Samples
 0.038  0.027  > 0.08  > 0.08  > 0.08  > 0.08  0.02  0.008
   400    400     400     400     400     400
0.0055  0.000   0.000  > 0.08    --    > 0.08
   400    450     450     450    --       400
Time start, inlet
Time end, inlet
                      o
Flow, start, inlet, cm /min
Volume, inlet, SL
Volume, outlet, £
Time start, outlet
Time end, outlet
Flow start, outlet, cnr/min
1738
1738
940
4.7
11.6
1733
1744
1050
1747
1752
940
4.7
8.4
1747
1755
1050
1801
1807
940
5.6
21.0
1801
1821
1050
1848
1908
940
18.8
23.1
1848
1910
1050
1933
1953
940
18.8
21.0
1933
1953
1050
2117
2137
940
18.8
21.0
2117
2137
1050
a_/  Only inlet TDI levels were recorded.
                                        22

-------
Sample time is the time from the beginning of the sample period until the
beginning of the next period.

TDI-MDA Tape

No special calculations are needed to obtain parts per million of TDI
using the MDA tape apparatus; levels are recorded on a continuous tape
and "spot" checked by reading a precalibrated card.

Cryogenic Traps

No special calculations required.

Evacuated Bulbs

No special calculations required.

Dispersion Model

A Gaussian plume dispersion model, presented in EPA workbook AP-26, was
used.  Pertinent information from this workbook is included in Appendix
C for reference purposes.

RESULTS

TDI-Colorimetric

Colorimetric data for both outlet and inlet samples as a function of
sample time are given in Tables 4 and 5.  Note that TDI plus amine con-
centration was low, between 0.070 (inlet) and 0.033 (outlet) ppm, before
the actual run began.  TDI was in the system as a result of vapor vented
during unloading of a tank truck of TDI just prior to run time.  As reac-
tion time and temperature increased, the concentration of TDI initially
increased, as would be expected.  Smoothed plots showing the TDI (plus
amines) in inlet and outlet scrubber streams are shown in Figure 9.  The
TDI concentration built to a maximum during the first 30 min of the run;
experimental maxima were 2.43 ppm (inlet) and 1.45 ppm (outlet) at medium
times of 1805 and 1848, respectively.  The run began at 1747 and it took
less than 10 min to add about 3,000 Ib of TDI.
                                   23

-------
                            Table 4.  TEST RESULTS - TDI ANALYSES
                                      (Samples Nos. 0-3)
Sample Number
0
Sample Point
TDI and amines, ppm
TDI, colorimetric, ug
TDI, MDA tape, ppmfL/
Time, startk/
Time, median
Total sample time, min
Gas flowrate, cnr/min— '
Gas volume, i
Temperature, °C
TDI and amines, g
Inlet
0.070
2.24
0.038
1733
1735
5
940
4.7
28
0.002
Outlet
0.033
2.61
0.005
1733
1738
11
1050
11.5
28
0.001
1
Inlet
1.62
49.6
0.02
1747
1750
5
940
4.7
43
0.041
Outlet
1.07
61.3
0.0
1747
1750
8
1050
8.4
28
0.028
2
Inlet
2.43
89.0
Limit
1801
1805
6
940
5.64
44
0.20
Outlet
0.831.
119
Limit
1801
1810
20
1050
21.0
29
0.073
3
Inlet
1.19
141
Limit
1848
1900
20
940
18.8
55
0.093
Outlet
1.45
225
Limit
1848
1900
22
1050
23.1
34
0.120
a/  15-20 min delay on MDA monitor readings;  range of the MDA instrument was 0.0-0.08 ppm.
b_/  With tests 6 and 7, TDI concentration was read visually at time indicated.
c_/  Flowrate  from exhaust was 160-240 ft/min from 3-in.  I.D. pipe; 200 ft/min rate was used.  Total
      flow = 9.8 ft3/min = 275 4/min.
Notes:  Cumulative process discharge of TDI and amines for entire process cycle (1747-2305 hr) :
          inlet  - 0.36 g
          outlet - 0.25 g
          TDI charging began at 1747 hr.  Test 0 was prior to reactor startup.
          Outlet TDI concentration on monitor was > 0.08  ppm from 1800 to 2130 (corrected for time
          delay).

-------
                                     Table 5.  TEST RESULTS  - TDI ANALYSES
                                               (Samples Nos.  4-7)
SJ
Ui
Sample Point

TDI and amines, ppm
TDI colorimetric, ug
TDI, MDA tape, ppm3-'
Time, start-'
Time, median
Total sample time, min
Gas  flowrate,  cnrVmin£'
Gas volume, £,
Temperature, °C
TDI and amines, g
                                                            Sample Number
Inlet
0.31
38.0
Limit
1933
1945
20
940
18.8
43
0.024
Outlet
0.32
44.4
Limit
1933
1945
20
1050
21.0
34
0.026
    0.0
    0.0
  Limit
   2117
   2130
     20
    940
   18.8
     44
< 0.001
0.023
  3.4
Limit
 2117
 2130
   20
 1050
 21.0
   18
0.005
                                                                                      Inlet   Outlet
                                                                     0.02    0.004
                                                                      2140
2140
        0.008
2240
          0.0
2240
     a/   15-20 min delay on MDA monitor readings; range of  the MDA instrument was 0.0-0.08 ppm.
     b/   With tests  6 and  7, TDI concentration was read visually at  time indicated.
     c_/   Flowrate  from exhaust was  160-240  ft/min from  3-in. I.D. pipe; 200 ft/min  rate was  used.  Total
            flow = 9.8 ft3/min = 275  jj/min.
     Notes:  Cumulative process discharge of TDI and amines  for entire process cycle  (1747-2305  hr):
               inlet - 0.36 g
               outlet - 0.25 g
               TDI charging began at 1747 hr.  Test 0 was prior to reactor  startup.
               Outlet TDI  concentration on monitor was > 0.08 ppm from 1800 to 2130  (corrected for  time
               delay).

-------
    3.0,
    2.5
    2.0
>

>

 S


<  1.5
+
0.
Q.
    1.0
    0.5
      0
                               50          100          150

                                Minutes After TDI Charging
250
       Figure  9.   TDI + amine concentration  (colorimetric)  versus time,
                                    26

-------
TDI-MDA Tape

The MDA tape apparatus reportedly determines only TDI.  Measurements taken
with this apparatus are given in Tables 4 and 5.  MDA data are limited be-
cause the instrument can detect a maximum level of 0.08 ppm; there is also
an 18-min delay in the printout.  Nevertheless, the MDA data parallel
colorimetric data in that a "limit" (> 0.08 ppm) reading was obtained
about 18 min after the colorimetric method indicated TDI readings above
0.08 ppm.  Also, MDA readings dropped below 0.08 ppm at 2140, or about 10
min after the colorimetric method showed 0.0 and 0.023 ppm in the inlet
and outlet streams, respectively.

One of the MDA tape instruments was used to monitor ambient air in front
of the scrubbers at 1920; it showed no detectable TDI.

Cryogenic Traps

Only two of the cryogenic traps from the inlet to the scrubber (Nos. 4
and 5 of Table 5) contained any significant amount of condensibles.
Chromatographic data on these samples were typical of mineral spirits
No.  302 as covered in the next section on "Evacuated Bulbs."  Because
of the greater relative importance of the outlet stream with respect to
air emissions, no more work was done to analyze inlet components.

Evacuated Bulbs

The evacuated bulbs contained gas and a small amount of liquid which con-
densed in the line leading from the sampling manifold.  The liquid was
compared by infrared spectroscopy, gas chromatography and mass spectrom-
etry with the mineral spirits added to the alkyd resin as a diluent.  The
results showed:

Infrared Spectroscopy:   No discernible difference is detected between the
spectrum of the liquid phase of the No. 5 outlet sample (Figure 10) and
that of the No. 302 mineral spirits (Figure 11).  The spectra indicate a
preponderance of saturated hydrocarbons similar to those in mineral spirits,
Levels of extraneous organic materials of about 17» or greater would have
shown up, if present.

Gas Chromatography:  Retention times for 23 peaks obtained from No. 302
spirits matched retention times for 23 peaks from the No. 5 outlet sample
(see Table 6).  Peak No. 2 in the outlet sample is not present in the
mineral spirits, but represents less than 0.1% of the sample volume.
                                   27

-------
ro
oo
        100 r~
         4000    3000
2000
1500
1200
1000
900
800
                                                                                                            700
                                                            CM
                                                               -1
                             Figure 10.  Infrared  spectrum of liquid from No.  5  outlet

                                   Conditions:  Perkin-Elmer Infracord NaCl cell.

-------
VO
         4000   3000
2000
1500
1200
1000
800
700
                                                           CM
                                                              -1
                             Figure 11.  Infrared spectrum of No.  302 mineral spirits,
                                   Conditions:  Perkin-Elmer  Infracord NaCl cell.

-------
              Table 6.  GAS CHROMATOGRAPH RETENTION TIMES
Conditions:  Varian Aerograph Gas Chromatograph, Model 1420, equipped with
               dual columns, dual high-sensitivity thermal conductivity
               detectors and 6-port gas sample valve.
             Sample volume:  gas, 5 ml; liquid, 3 ul.
             Column:   6.5 ft x 1/8 in. Porapak Q.
             Flow rate:  60 ml/min.
             Column temperature:  50-250°C at A t20°C/min.
             Sensitivity:  50 ug/ml.
     Peak Number

          1
          2
          3
          4
          5
          6
          7
          8
          9
         10
         11
         12
         13
         14
         15
         16
         17
         18
         19
         20
         21
         22
         23
No. 5 Outlet
   (min)	
      75
      63
      88
    3.69
    4.75
    5.56
    6.19
    6.44
    7.34
    7.85
    8.54
    8.69
    9.0
    9.19
    9.60
   10.38
   10.92
   11.53
   11.86
   12.43
   13.25
   14.43
   15.75
No. 302 Mineral Spirits
	(min)	

          1.75
      Not present
          2
          3,
          4.
          5,
  .88
  .74
  .69
  .61
 6.16
 6.38
 7.38
 7.88
 8.38
 8.73
 9.0
 9.23
 9.51
10.39
10.95
11.53
11.86
12.50
13.19
14.06
15.73
                                   30

-------
Mass Spectrometry:  Mass spectrometry was utilized to examine the 10 major
peaks of the gas chromatogram.  These peaks represented over 99% of the
sample.  These mass spectra, included as Appendix D, show that there were
probably no unresolved materials in these peaks of a nature different from
mineral spirits.

The gas phase of the evacuated bulb samples was saturated with mineral
spirits which condensed both in the bulbs and in the line connected to
the bulbs.  Vapor pressure measurements at experimental conditions showed
2.5 vol % mineral spirits present at saturation.

Dispersion Model

The data that have been presented describe gas emissions (vol %), emission
rates, stack configuration, stack diameter, and stack surfaces.  To make
the information meaningful from an environmental standpoint, the ground-
level concentrations of the toxic gases must be determined.  In order to
do this, the following source characteristics were used:

Internal diameter of vent = 3 in.

Volumetric discharge rate =275 £/min

Composition of discharge during the peak 1-hr period:

     Nitrogen            97.5% (vol)
     Mineral spirits      2.5% (vol)
     TDI                  1.2 ppm (vol)

For the wind speeds greater than about 10 mph, aerodynamic downwash, which
would be greatest on the leeward side of the building (from which emissions
emanate), is the dominant factor in the atmospheric dilution process.
Emissions would be swept into the region of induced turbulence and quickly
mixed over the cross-section of the building normal to the wind direction.
In this case, the dilution factor (DF) is simply:

          (Downwash)
        Dilution factor =   Volume rate of pollutant emission
            (DF)          Wind speed x building cross section

For wind speeds less than about 10 mph, the TDI vent may be treated as an
elevated point source which is unaffected by the presence of the building.
The location and magnitude of the maximum ground-level pollutant concentra-
tion vary with the atmospheric stability and wind speed, as shown in Table 7.
                                    31

-------
                                  Table 7.  RESULTS OF DISPERSION CALCULATIONS
U)
ro

Point of Max. Ground-
Level Concentration
Atmospheric Condition Distance From
Stability Wind Speed Source DF*/
—
Extremely
Unstable
Unstable

Slightly
Unstable
Neutral
Slightly
Stable
Stable
a/ Dilution
> 10 mph < 100 m 1.1 x 105
5 mph 50 m 2.8 x 105
5 mph 70 m 3.2 x 105

5 mph 110 m 3.4 x 105

5 mph 160 m 3.7 x 105
5 mph 220 m 3.9 x 105

5 mph 390 m 4.2 x 105
factor - Concentration in Flue Gases
Residential Area
Ground Level
DF£/

1.4 x
2.5 x
2.5 x
1.8 x
1.0 x

3.4 x
1.8 x

9.2 x

108
107W
108
I07b/
107

106
106

105
Plume Centerline
DF3/

1.4 x 108
2.5 x 108

1.0 x 107

3.6 x 106
2.0 x 106

1.1 x 106
       ~~                     Concentration in Ambient Air

       b/  With  inversion at  100 m;  for more stable atmosphere,  DF  is  unaffected by 100 m inversion.

-------
Dilution factors are given for the points of maximum ground-level concen-
tration and for the residential receptor distance of approximately 3,000 ft.
The evaluation was made by considering the location of the residential
area as shown in Figure 12, and the general plant area shown in Figure 13.

If one assumes the worst possible case (a DF of 10^), the following holds
true:

                    Maximum ground-level     Maximum allowable 8-hr
     Component      concentrations, ppm       concentrations, ppm

       TDI               1.2 x 10'5                  0.02
     Mineral spirits        0.25                   500

The ground-level concentrations of both TDI and mineral spirits are only
about 0.1% of the maximum allowable values for an 8-hr averaging period
under such adverse conditions.  One should remember that these numbers
assume that none of the TDI is converted to amines; in all likelihood,
most of the TDI is hydrolyzed, thus increasing the safety factor further.

Each run contains about 30,000 Ib of resin and emits about 0.25 g (5.5 x
10~^ Ib) of TDI as a maximum.  Based on these numbers,


                     5.5 x IP"4 x 1.8 x 10"8  Ib TDI
                     3.0 x 104       Ib product
is emitted into the atmosphere.

DISCUSSION OF RESULTS

Considerable caution is necessary in evaluating the graph of TDI-plus-
amines versus time (Figure 9).  Since the MDA monitor was off scale for
most of the process run, no continuous measure of TDI levels is available.
The colorimetric method is not sensitive enough for shorter sample periods
than those used during testing.  Due to the rapid reaction rate during
this process run, sufficient colorimetric analyses were not obtainable to
properly describe the curve.

Thus, the exact shape of the inlet and outlet curves is not determinable
with the present information.  Since the sample periods are fairly long,
the actual maxima may be higher than shown in the graph and the error
limits for the calculated totals of TDI from the process are estimated to
be approximately -50% to +100%.  The lag in the outlet concentration is
expected, since the internal volume of each scrubber is about 200 liters
with turbulent flow.  Several minutes will therefore be required to sweep
the "old" gas from the scrubber.
                                   33

-------
LO
•P-
                                                                               N
                                             Bluffs ~200' High
                                                                            : Residential:
/" Stack O [
650' Tall A-^. ^^^

-------
            I
          North
U>
                                                                       Connecting Piping
                                                                                                TDI Vent
                                                                                                on 38' High Roof
                                                                                                             n
                                                                                                   MRI  Laboratory Truck
                                         Entranceway
                                         10' High-12 to 15' Wide
                                         Opening
                          Interior Hallway
TDI Scrubber
System
                                       :::: Plant Area:::::::::::::
                                                                                                                             Storage
                                                                                                                             Area
                   Inside-Plant
                      TDI
                    Storage
                     Area
                                      Figure  13.   Schematic of test  site  (overhead view)

-------
With the information gained from this test, the response of the MDA moni-
tors could be modified by a gas dilution technique to allow the instruments
to measure the levels of TDI present.

From the analytical results for total organics, it is reasonably certain
that the only significant toxic materials from this particular process are
TDI and mineral spirits.  It is interesting to note that, while the permis-
sible exposure level for mineral spirits is 25,000 times higher than TDI,
the concentration of the mineral spirits is also 25,000 times greater than
that of the TDI.  The TDI concentration rises rapidly to a peak and then
falls to zero by the end of the reaction.  The concentration of the min-
eral spirits is at a constant level throughout the process, since conden-
sation in the lines of mineral spirits was observed from a few minutes
after the start of the reaction until the end.  Thus, the mineral spirits
emissions could far exceed the levels found during this test if the tem-
perature of the exiting gas stream were higher than 25°C.  At higher tem-
peratures the mineral spirits would be emitted instead of condensing out
in the piping.

Thus, if the process conditions of the plant studies are representative of
the industry and if other plants exist with less favorable diffusion con-
ditions from the stack or much higher total flowrates, the toxic level of
mineral spirits could be reached well before that for TDI.
                                   36

-------
                                SECTION V

                               CONCLUSIONS

The major conclusions to be drawn from the results of this study are:

1.  The maximum ground-level concentration of both TDI and mineral spirits
is about 0.1% of the maxima allowable for an 8-hr averaging period, thus
indicating that no hazard existed at the test site under the conditions
studied.

2.  Statement No. 1 is true in spite of the fact that the TDI scrubber
was operating poorly* at the time of sampling.  This would indicate that
scrubbers are not needed for TDI removal if the specific process and condi-
tions which were studied are kept under control.

3.  Should several such plants be clustered in or around a common community,
as is frequently the case in the chemical processing industry, a problem
could result.
*  Reported by plant site personnel subsequent to MRI sampling.
                                   37

-------
                               SECTION VI

                             RECOMMENDATIONS

Based on the conclusions drawn in this report, it is recommended that:

1.  Other test sites be studied under a broader range of operating condi-
tions to determine air emissions.

2.  At present, caution should be exercised in extrapolating the data
presented herein to the entire polyurethane coatings industry, since it
is specific to one plant.

3.  Priorities should be extended to other areas of the plastics industry
having potentially high hazardous/toxic emissions.

4.  A search for possible clustered polyurethane sites might be in order;
this might also apply to other plastics processes.
                                    38

-------
                      APPENDIX A

LIST OF COMPANIES PRODUCING URETHANE COATING RESINS FOR
        EITHER CAPTIVE USE OR FOR SALE OR BOTH:
             COMPANIES AND PLANT LOCATIONS
                           39

-------
A&H Paint Company
Columbus, Ohio
Absolute Coatings, Inc.

Acron Chemical Company

Aero Chemical Products Corporation

Adelphi Paint & Color Works, Inc.


Adhesive Products Corporation

Admiral Paint Company

Advance Process Supply Company


American Herberts Corporation

Amercoat Corporation

John L. Armitage & Company



Ashland Oil, Inc.


The Baker Castor Oil Company

Ball Chemical Company

Beatrice Foods Company

The Biggs Company

Cargill, Inc.



Celanese Corporation
Bronx, New York
Long Valley, New Jersey

Carlstadt, New Jersey
Ozone Park, New York

Bronx, New York
Brooklyn, New York
Chicago, Illinois

Woodside, New York

Brea, California

Newark, New Jersey
Elk Grove, Illinois
Richmond, California

Los Angeles, California
Newark, New Jersey

Bayonne, New Jersey

Glenshaw, Pennsylvania

Wilmington, Massachusetts

Santa Monica, California

Carpentersville, Illinois
Lynwood, California
Philadelphia, Pennsylvania

Los Angeles, California
Louisville, Kentucky
Newark, New Jersey
                                   40

-------
Chem-Seal Corporation of America
Los Angeles, California
Chemical Coatings & Engineering Company

Commercial Solvents Corporation


CONCHEMCO, Inc.


Continental Polymers Corporation

Cook Paint & Varnish Company



Desoto, Inc.



The Dexter Corporation
Diamond Shamrock Corporation

E. I. du Pont de Nemours & Company, Inc.
Emerson & Cuming, Inc.

The Epoxylite Corporation

Farnow, Inc.
Media, Pennsylvania

Carpentersville, Illinois
Chicago, Illinois

Baltimore, Maryland
Kansas City, Missouri

Santa Ana, California

Detroit, Michigan
Houston, Texas
North Kansas City, Missouri

Berkeley, California
Chicago Heights, Illinois
Garland, Texas

El Monte, California
Clean, New York
Cleveland, Ohio
Hayward, California
Los Angeles, California
Rocky Hill, Connecticut
Waukegan, Illinois

Harrison, New Jersey

Chicago, Illinois
Fort Madison, Iowa
Parlin, New Jersey
Philadelphia, Pennsylvania
South San Francisco, California
Toledo, Ohio
Tucker, Georgia

Canton, Massachusetts

El Monte, California

South Kearny, New Jersey
                                   41

-------
France, Campbell & Darling, Inc.
Kenilworth, New Jersey
Freeman Chemical Corporation


Furane Plastics, Inc.

General Latex and Chemical Corporation


The P. D.  George Company

The Goodyear Tire & Rubber Company

Hoover Ball and Bearing Company

Interplastic Corporation

Isochem Resins Company

Jewel Paint & Varnish Company

Kohler-McLister Paint Company

Lu-Sol Corporation

The Master Mechanics Company

McCloskey Varnish Company



Midwest Manufacturing Corporation

Minnesota Mining and Manufacturing
  Company

Minnesota Paints, Inc.

Mobay Chemical Company/Naftone, Inc.

National Lead Company

Norris Paint & Varnish Company, Inc.
Ambridge, Pennsylvania
Saukville, Wisconsin

Los Angeles, California

Ashland, Ohio
Cambridge, Massachusetts

St. Louis, Missouri

Akron, Ohio

Ann Arbor, Michigan

Minneapolis, Minnesota

Lincoln, Rhode Island

Chicago, Illinois

Denver, Colorado

El Monte, California

Cleveland, Ohio

Los Angeles, California
Philadelphia, Pennsylvania
Portland, Oregon

Burlington, Iowa

Decatur, Alabama
Fort Wayne, Indiana

New Martinsville, West Virginia

Philadelphia, Pennsylvania

Salem, Oregon
                                   42

-------
Northeastern Laboratories Company, Inc,
Melville, New York
C. J. Osborn Chemicals, Inc,

Poly Resins, Inc.

PPG Industries, Inc.
Preservative Paint Company

Prime Leather Finishes Company

Raffi and Swanson, Inc.

Reichhold Chemicals, Inc.
Schenectady Chemicals, Inc.
SCM Corporation
The Sherwin-Williams Company
Sta-Crete, Inc.
A. E. Stanley Manufacturing Company
Pennsauken, New Jersey

Sunn Valley, California

Circleville, Ohio
Houston, Texas
Milwaukee, Wisconsin
Springdale, Pennsylvania
Torrance, California
Cleveland, Ohio

Seattle, Washington

Milwaukee, Wisconsin

Wilmington, Massachusetts

Azusa, California
Detroit, Michigan
Elizabeth, New Jersey
Houston, Texas
South San Francisco, California
Tacoma, Washington

Schenectady, New York

Chicago, Illinois
Cleveland, Ohio
Reading, Pennsylvania
San Francisco, California

Chicago, Illinois
Cleveland, Ohio
Emeryville, California
Garland, Texas
Los Angeles, California
Newark, New Jersey

San Francisco, California

Marlboro, Massachusetts
                                   43

-------
Textron Inc.,
  Spencer Kellogg Division

Thiokol Chemical Corporation

Francoa Chemical Corporation

Union Carbide Corporation

Westinghouse Electric Corporation

Wilmington Chemical Corporation

Witco Chemical Corporation
Woburn Chemical Corporation
Wyandotte Chemical Corporation
Bellevue, Ohio
Trenton, New Jersey

Reading, Massachusetts

Institute, West Virginia

Manor, Pennsylvania

Wilmington, Delaware

Chicago, Illinois
Lynwood, California
New Castle, Delaware

Harrison, New Jersey

Wyandotte, Michigan
                                   44

-------
        APPENDIX B




ANALYTICAL METHODS FOR TDI
             45

-------
A.  Analysis of TDI in Air (Modified Marcal:  Method)

Principle

TDI is hydrolyzed by the absorbing solution to the corresponding toluene
diamine derivative.

The diamine is diazotized by the sodium nitrite-sodium bromide solution.

The diazo compound is coupled with N-(l-Naphthyl)ethylenediamine to form
a colored complex.

The amount of colored complex formed is in direct proportion to the amount
of TDI present.  The amount of colored complex is determined by reading the
absorbance of the solution at 550 nm.

Toluene diamine is formed on a mole for mole basis from TDI.  This amine
is used in place of TDI for standards.  This accomplishes two things.  First,
the amine is not as toxic as TDI.  Second, TDI is semisolid at room tempera-
ture.  Weighing the semisolid is more difficult than weighing the dry amine.
Both compounds have been tested by this method and the results compare
favorably.

Range and Sensitivity

The range of the standards used is equivalent to 1.0-20.0 ug TDI.  In a
40-liter air sample, this range converts to 0.0035-0.070 ppm.  The sensi-
tivity can be increased by using longer path length spectrophotometer cells.

If the sample is so concentrated its absorbance is greater than the limits
of the standard curve, it can be diluted with absorber solution and the
absorbance reread.  This extends the upper limit of the range.  The upper
limit can also be extended by taking a smaller air sample.

A single bubbler absorbs 95% of the diisocyanate if the concentration is
below 2 ppm.  Above 2 ppm, about 907» is recovered.

Interferences

Any free organic amine will interfere, including any that may be present
in detergents.

Methylene-di-(4-phenylisocyanate) (MDI) will form a colored complex in this
reaction.  However, its color development time is about 1-2 hr compared
with 5 min for TDI.  Therefore MDI is not a serious problem, if color den-
sity is determined within 10 min of the addition of coupling reagent.

                                    46

-------
Apparatus

Beckman Model B spectrophotometer or equivalent

Cells, 1-cm, 4-cm, 5-cm, or 10-cm matched cells

Several (each) volumetric flasks:  50 ml, 100 ml, 1-liter, glass-stoppered

Balance capable of weighing to at least three decimal places

Pipettes:   0.5 ml, 1 ml, 15 ml

Graduated cylinders:  25 ml, 50 ml

Reagents

All reagents must be ACS reagent grade or better.

     Double distilled water
     2, 4-diaminotoluene
     Hydrochloric acid, concentrated, 11.7 N
     Glacial acetic acid, concentrated, 17.6 N
     Sodium bromide
Sodium nitrite solution:  Dissolve 3.0 g sodium nitrite and 5.0 g sodium
bromide in about 80 ml double distilled water.  Adjust volume to 100 ml
with double distilled water.

     Sulfamic acid

Sulfamic acid solution, 10% w/v:  dissolve 10 g sulfamic acid in 100 ml
double distilled water.

     N-(l-Naphthyl)ethylenediamine dihydrochloride

N-(l-Naphthyl)ethylenediamine solution:  Dissolve 50 mg in about 25 ml
double distilled water.  Add 1 ml concentrated hydrochlorid acid and
dilute to 50 ml with double distilled water.  Solution should be clear;
any coloring is due to contamination by free amines, and if the solution
is colored it should not be used.

Absorber solution:  Add 35 ml concentrated hydrochloric acid and 22 ml
glacial acetic acid to approximately 600 ml double distilled water.
Dilute the solution to 1 liter with double distilled water.  Use 15 ml
in each sample-collecting impinger.

                                   47

-------
Standard solution A:  Weigh out 140 mg of 2,4-toluenediamine (equivalent
to 200 mg of 2,4-toluene diisocyanate).  Dissolve in 660 ml of glacial
acetic acid, transfer to a 1-liter glass-stoppered volumetric flask, and
make up to volume with double distilled water.

Standard solution B:  Transfer 10 ml of standard solution A to a glass-
stoppered 1-liter volumetric flask.  Add 27.8 ml of glacial acetic acid
so that when solution B is diluted to 1 liter with double distilled water,
it will be 0.6N with respect to acetic acid.

Procedure

Cleaning Equipment:  Wash all glassware in a hot amine-free detergent solu-
tion, or soak in a 1% aqueous trisodium phosphate (analytical reagent)
solution at room temperature, preferably overnight, to remove any oil.

Rinse well with hot tap water.

Rinse well with double distilled water.  Repeat this rinse several times.
Any amines from organic detergents must be removed to prevent interferences.

Analysis of Samples:  Remove impinger tube taking care not to lose any
absorber solution.

Start blank at this point by adding 15 ml fresh absorber solution to a
clean impinger cylinder.  To each bubbler add 0.5 ml of 3% sodium nitrite
solution, gently agitate, and allow solution to stand for 2 min.

Add 1 ml of 107» sulfamic acid solution, agitate and allow solution to stand
about 2 min to destroy all the excess nitrous acid present.

Add 1 ml of 0.1% N-(l-Naphthyl)ethylenediamine solution.  Agitate and allow
color to develop.  Color will be developed in 5 min.  A reddish-blue color
indicates the presence of TDI.

Add double distilled water to adjust the final volume to 20 ml in the
cylinder.  Mix.

Transfer each sample and blank to 1-cm or longer spectrophotometer cell.

Using the blank, adjust the spectrophotometer to 0 absorbance.

Determine the absorbance of each sample at 550 nm.
                                    48

-------
From the previously prepared calibration curve (see below) read the micro-
grams TDI corresponding to the absorbance of the sample and calculate the
parts per million TDI.

Calibration and Standards

To each of a series of eight 25-ml graduated cylinders add 5 ml of 1.2N
hydrochloric acid.

To these cylinders add the following amounts of 0.6N acetic acid:   10.0,
9.5, 9.0, 8.0, 7.0, 6.0, 5.0, and 0.0 ml, respectively.

To these cylinders add standard solution B in the same order as the acetic
acid is added:  0.0, 0.5, 1.0, 2.0, 3.0, 4.0, 5.0, and 10.0 ml, so that
the final volume is 10 ml.  None (0.0 ml) of the standard is added to the
10 ml acetic acid; 0.5 ml of the standard is added to the 9.5 ml acid; and
so on.  The cylinders now contain the equivalent of 0.0, 1.0, 2.0, 4.0, 6.0,
8.0, 10.0, and 20.0 ug TDI, respectively.  The standard containing none of
the standard solution is the blank.

Add 0.5 ml of the 3.0% sodium nitrite reagent to each cylinder.  Mix.  Allow
to stand for 2 min.

Add 1 ml of the 10% sulfamic acid solution.  Mix.  Allow to stand for 2 min.

Add 1 ml of the N-(l-Naphthyl)ethylenediamine solution.  Mix.  Let stand
for 15 min.

Make up to exactly 20 ml with double distilled water.

Transfer each solution to 1 cm or longer spectrophotometer cell.  (At the
lower end of the calibration curve, 5 cm cells give an 11% relative in-
strumental error for the 1.0 ug TDI standard.  For smaller path lengths,
the error is greater.)

Using the blank, adjust the spectrophotometer to 0 absorbance at 550 nm.

Determine the absorbance of each standard at 550 nm.

A standard curve is constructed by plotting the absorbance against micro-
grams TDI.
                                    49

-------
Calculations


               ppm = micrograms x 24.45 = mi    rams x 0.00351
                      174.15 x 40

                     micrograms - micrograms TDI taken from standard curve

                     mol wt = wt TDI = 174.15

                          V = volume of air sample in liters (40 liters)


B.  Analysis of the TDI in Air (MDA TDI Monitor. Model 7000)

Principle

The MDA apparatus is based on a chemical reaction which takes place on a
special test paper which has been impregnated with suitable chemicals.
The result of this reaction is to produce a colored stain which is moni-
tored photoelectrically.  The test paper is in the form of a continuous
reel of paper tape contained in a cassette.  A motor drive pulls the tape
past a sensing head at a slow, constant speed.  A metered air sample is
drawn into the instrument by a small pump.  This air sample passes through
a portion of the tape.  An equivalent section of the tape width is shielded
from the gas and is used as a reference.  The TDI concentration of the gas
sample is proportional to the density of the color stain.  The outputs of
the photocell detectors are amplified to drive the meter, recorder and
alarm system.

Range and Sensitivity

The instrument range is from 0-0.08 ppm with logarithmic response.  The
sensitivity limit is approximately 0.002 ppm.

Interferences

Methylene-di-(4-phenylisocyanate) will also form a colored complex, but
the development time is 1-2 hr; the sensor reads the stain after an 18
min delay.

Apparatus

MDA Model 7000 monitor was used in this study.  Each tape cassette will run
for 168 hr.  Shelf life is 3 months for the unopened tape.  Specially
printed recorder charts, available from MDA, were used.
                                    50

-------
Calibration and Standards

The instrument is supplied with a calibration card with zero and 0.02 ppm
level stains.
                                    51

-------
                  APPENDIX C

REFERENCE MATERIAL ON GAUSSIAN PLUME DISPERSION
         MODEL FROM EPA WORKBOOK AP-26
                       52

-------
                                 Chapter 1 — INTRODUCTION
   During recent years methods of estimating at-
 mospheric dispersion have undergone considerable
 revision, primarily due  to results of experimental
 measurements.  In most dispersion  problems the
 relevant atmospheric layer is that  nearest the
 ground, varying in thickness from several hundred
 to a  few  thousand  meters.   Variations in  both
 thermal and  mechanical turbulence and in  wind
 velocity are greatest in the layer in contact with
 the surface. Turbulence induced by buoyancy forces
in the atmosphere is closely related to the vertical
                                    temperature structure. When temperature decreases
                                    with height at a rate higher than 5.4 °F per 1000 ft
                                    (1°C per 100 meters), the atmosphere is  in  un-
                                    stable equilibrium  and  vertical motions are  en-
                                    hanced.  When temperature decreases at a  lower
                                    rate or increases with height (inversion), vertical
                                    motions are damped or reduced.  Examples of typ-
                                    ical variations in  temperature and wind speed with
                                    height for daytime and nighttime conditions are
                                    illustrated in Figure 1-1.
  600r
 500
 400
 300
 200
  100
   0
   -1
234567
  TEMPERATURE, °C
                                      8   9  10  11   12
3   4   5   6   7   8
 WIND SPEED,  m/i.c
10  II
       Figure 1-1.  Examples of variation of temperature and wind speed with height (after Smith, 1963).
   The transfer of momentum upward or  down-
ward in the atmosphere is also related to stability;
when  the atmosphere is  unstable, usually in the
daytime, upward motions transfer the momentum
"deficiency" due to eddy friction losses near the
earth's  surface through  a relatively deep  layer,
causing the wind  speed  to increase  more  slowly
with height than at night (except  in the lowest few
meters). In addition to thermal turbulence, rough-
ness elements  on the ground engender mechanical
 turbulence, which  affects both the  dispersion of
 material in the atmosphere and  the wind  profile
 (variation of wind with height). Examples of these
 effects on the  resulting wind profile  are shown in
 Figure 1-2.
                                       As wind  speed increases, the effluent from  a
                                    continuous source is introduced into a greater vol-
                                    ume of air per unit time interval.  In addition to
                                    this  dilution  by wind speed, the spreading of  the
                                    material (normal  to the mean direction of trans-
                                    port) by turbulence is  a major factor in the dis-
                                    persion process.

                                       The procedures presented here  to estimate at-
                                    mospheric dispersion are applicable when mean wind
                                    speed and direction can be determined, but meas-
                                    urements of  turbulence, such as the standard de-
                                    viation of wind direction fluctuations, are not avail-
                                    able.  If such measurements are at hand, techniques
                                    such as those outlined by Pasquill (1961) are likely
                                    to give more accurate results. The diffusion param-
                                                53

-------
eters presented here are most applicable to ground-
level or low-level releases (from the surface to about
20 meters), although they are commonly applied at
higher elevations without full experimental valida-
tion.   It  is assumed  that  stability  is  the same
throughout the diffusing layer, and  no turbulent
transfer occurs through layers of dissimilar stability
characteristics. Because mean values for wind direc-
tions and speeds are required, neither the variation
of wind speed  nor  the variation of wind  direction
with height in  the mixing layer are taken into ac-
count. This usually is not a problem in neutral or
unstable  (e.g.,  daytime) situations, but can cause
over-estimations  of downwind  concentrations  in
stable conditions.
                                                              REFERENCES

                                              Davenport, A. G., 1963:  The relationship of wind
                                                 structure  to  wind  loading.  Presented  at Int.
                                                 Conf. on  The Wind Effects on Buildings and
                                                 Structures, 26-28 June 63, Natl. Physical Lab-
                                                 oratory, Teddington, Middlesex, Eng.

                                              Pasquill, F., 1961:  The estimation of the dispersion
                                                 of wind  borne material.   Meteorol.  Mag.  90,
                                                 1063, 33-49.

                                              Smith, M. E., 1963: The use  and misuse of the at-
                                                 mosphere, 15  pp., Brookhaven Lecture Series,
                                                 No.  24, 13 Feb 63, BNL 784  (T-298)  Brook-
                                               ,  haven National Laboratory.
  600t—
  500
 ..400

 0>
 
-------
                                  Chapter  2 —BACKGROUND
     For a number of years estimates  of concentra-
  tions were calculated either from  the equations of
  Sutton  (1932)  with  the  atmospheric  dispersion
  pnrameters C,, C«, and n, or from the equations of
  Bosanquet (1936) with  the dispersion  parameters
  p and q.

    Hay and Pasquill (1957) have presented experi-
  mental  evidence  that  the  vertical distribution of
 spreading particles from an elevated point is re-
 lated to  the standard deviation  of the wind eleva-
 tion angle, 
-------
                Chapter 3 — ESTIMATES  OF ATMOSPHERIC DISPERSION
    Phis chapter outlines the basic procedures  to
  '  used in  making dispersion  estimates as  sug-
 .•••-tcrl by Pasquill (1961) and modified by Gifford
 INORDINATE SYSTEM

    In the system considered  here  the  origin is at
 around level at or beneath the point of emission,
 with the x-axis extending horizontally in the direc-
 tion of the mean wind.  The y-axis is in the hori-
 zontal plane perpendicular  to  the  x-axis,  and the
 x-axis extends vertically.  The  plume travels along
 or parallel to the x-axis.  Figure 3-1 illustrates the
 coordinate system.

 DIFFUSION EQUATIONS

   The concentration, x,  of gas or  aerosols (parti-
cles less than about 20 microns diameter) at x,y,z
from a continuous source with an effective emission
height, H, is given by equation 3.1.  The notation
used to  depict this concentration  is x  (*,y,z;H).
H is the  height  of the  plume centerline  when it

                     I
becomes essentially level, and  is the  sum  of  the
physical stack  height, h, and the plume rise, AH.
The following  resumptions are made:   the  plume
spread has a Gaussian distribution  (see Appendix
2) in both the  horizontal and vertical planes, with
standard deviations of plume concentration  distri-
bution in the horizontal and vertical of ar and (
respectively;  the mean  wind speed affecting the
plume is u; the uniform emission rate of pollutants
is Q; and total reflection of the plume takes place
at the earth's  surface, i.e., there is no deposition
or reaction at the surface (see problem  9).
                                                    X (x,y,z;H)
                                                                                                (3.1)
                                                    'Note: exp —a/b — e~°/b where e is the base of natural logarithms
                                                         and is approximately equal to 2.7183.
                                                                                   (x,-y,Z)
                                                                                    (x,-y,0)
Estimates
         Figure 3-1.  Coordinate system showing  Gaussian distributions in the horizontal and vertical.

                                                      56

-------
 Any consistent set of units may be used.  The most
 common is:

    x (g m~a) or, for radioactivity (curies m~:i)
    Q (g sec"1) or  (curies sec"')
    u (msec"1)
    ay, a., H.x.y, and z (m)

 This equation is the same as equation (8.35) p. 293
 of Sutton  (1953)  when a's are substituted for But-
 ton's parameters  through  equations  like (8.27)  p.
 286.  For evaluations of the exponentials found  in
 Eq. (3.1)  and those that follow, see Appendix  3.
 X is a mean over the same  time interval as the time
 interval for which the a's and u are representative.
 The values of both ay and   6
1
Day

' , Incoming Solar Radiation
III,

Strong
A
A-B
B
C
C

Moderate
A-B
B
B-C
C-D
D

Slight
B
C
C
D
D
Night
Thinly Overcast

3=4/8 Low Cloud

E
D
D
D


J/ 0
Cloud

F
E
0
D
The neutral class, D, should be assumed for overcast conditions during
day or night.

   "Strong" incoming solar radiation corresponds
to a solar altitude greater than 60° with clear skies;
"slight" insolation  corresponds to a solar altitude
from  15° to 35° with clear skies.  Table 170, Solar
Altitude and  Azimuth, in the Smithsonian Mete-
orological Tables (List, 1951) can he used in deter-
mining the solar altitude. Cloudiness will decrease
incoming solar radiation and should be  considered
along with solar altitude in determining solar radia-
tion.   Incoming  radiation that  would  be  strong
with clear skies can be expected  to be reduced  to
moderate with broken (% to % cloud cover) mid-
dle clouds and to  slight  with broken low  clouds.
An objective  system of classifying  stability from
hourly  meteorological observations  based  on the
above  method has been suggested (Turner,  1961).

   These methods  will give  representative indica-
tions  of stability over open country  or rural areas,
but are less reliable  for urban areas. This differ-
ence is due primarily to the  influence of the city's
larger  surface  roughness  and heat  island effects
upon  the stability  regime over urban areas. The
greatest difference occurs on calm clear  nights; on
such nights conditions over rural areas are very
stable, but over urban areas they are slightly un-
stable  or near neutral to a height  several times the
average building height, with a stable layer above
(Duckworth and Sandberg, 1954; DeMarrais, 1961).
                                            57
                                                            ATMOSPHERIC DISPERSION ESTIMATES

-------
   Some preliminary results of a dispersion experi-
 ment in St.  Louis  (Pooler, 1965) showed that the
 dispersion over the city during the daytime behaved
 somewhat like types B and C; for one night experi-
 ment a, varied with distance between types D and E.

 ESTIMATION OF VERTICAL AND
 HORIZONTAL  DISPERSION

   Having determined the  stability  class  from
 Table 3-1, one can evaluate the estimates of  a, and
 ",. as  a  function  of downwind distance  from  the
 source, x,  using Figures 3-2 and  3-3. These  values
 of «,. and 2 XT,; xr. is where a,.
                                0.47 L
     for any z from 0 to L
     for x >2Ni.; x,. is where 
-------
                   APPENDIX D

MASS SPECTROMETRY DATA FOR NUMBER 5 OUTLET SAMPLE
      AND NUMBER 302 MINERAL SPIRITS SAMPLE
                       59

-------
Instrument:  (1) Atlas CH-AB medium resolution mass spectrometer equipped





                 with a Varian 620/i computer;





             (2) GC/MS system consisting of a Microtek 220 Gas Chromato-





                 graph interfaced to the Atlas CH-4B mass spectrometer via





                 a two-stage helium separator.





     The mass spectra in this appendix show the strong similarity between





a major gas chromatographic peak from the No. 5 outlet sample and a com-





parable peak from the No. 302 mineral spirits.  No attempt was made to





identify the specific compound.





     Ten peaks representing over 99 vol 7» of the sample were chosen for





comparison.  The mass spectrometer was run at higher than normal sensitiv-





ity so that minor variations would be magnified.





     Two spectra are paired for comparison under each figure.  Thus, Figure




D-1A is to be compared to Figure D-1B, etc.
                                   60

-------
1 OQ-.
90-j
80-
60-
SO-
f 0-
30-1
eo-
10-j
o —










0








*J







ill J
'(' "T ""I"







1

50








i|ll









|









•HI 1
"1








•1
1







1,





1
1,1
1 00








, ,1,,,,l,,,,l







,
1.
150 200 250 3






„

f\
00
ECU 32004 Lh «5 OUTLET STEP MflSS=l.- I-'B^S = 1 Y.
Figure D-1A

-------
CTI
ro
100-q
t
90H
80-
70^
50-
eo-i
1 0-j
fi . "







•mrmnrntttJ
irii|iM>|iiil|M
0






1 1


,
T1"'1 '.' '
S P E C tt 31930 Lfl













1. '!ll
	 r"Ti1


i



,41






hJi!
itiiji!it|nii ..IIJI.M






||






, II
ttfl ur| ,,|,
1






U


i







T1 "'i 11|linr"'j
00 <;
3 OS MIMERRL SPIRITS






i

-


-
*~

1 1 	 |-T".|-.|""| 	 -.|».. 	 | 	 | 	 i[[[ 	 |.,,,|,,..|.,,,,..,, !„„,....„
cj 150 200 250 300
7-3-74 STEP H«SS=1, I/'B/'S -

-------
CJ
1 00-q
904
804
704
604
so4
H04
304
804
i o4
0."











	 ,|,|
mi|i!ii|iiii|ll
0





















liLj











so
1









hi











i

1









II
ty*'











hh







1













id








In
100










1,1 II,











ll 1 1

EC« 32008 LH «5 OUTLET










1,
. ..,.,, j....,...^...,....!....,,,^..,.,....!....,....!....,....! 	 j....,....|.,,,,...,| 	 ,..|....,,.,.|....,....|..,, 	 |....,..I.j....l..c
ISO 200 250 3










00
STEP M fl S S = 1 , I/B-'S
                                                                                                                       IJi
                                                    Figure D-2A

-------
   1 00-=
    90-:

    80-

    ?0H
    t>0-

    50-

    HO-

    30-1
    SO-I
        0              50         .100
SPECtt  31934 LM   302  MINERflL  SPIRITS
    150
7-3-74
200            250             300
        STEP  HflSS=l, I/'B/'S    =  IZ
                                              Figure D-2B

-------
0>
                     70-
                     60-E
                        c
                     so4
                     eo-
SPECtt 32010  LM
                                       50
                                     «5  OUTLET
                                    1 00
IT
  150
200            250            300
       STEP M fl S S = 1 ,  I x B .-• S   =  \'f.
                                                   Figure D-3A

-------
1 00-a
i
90-4
i
80-j
60-f
50-E
30-|
20-{
1 0-f
0:











I
0










,|























ECU 31936 LM










ill
"I"1










tUrl
I1"1 '
50
II









ul i 1
JIHI .
1









'HI























,











In
















Ll









I
Ml








1 1
111, I,!,
100 150 200 250 3
302 MINERflL SPIRITS 7-3-74 STEP nflSS=l, I/B/S
Figure D-3B

-------
a\
t 00-:
90-
80-
70-j
s
50-{
to-i
eo-f
1 0-1
O.r











ill
1 i
0









(
i '










III III
1 •'










	 1

5









|

0








i
IJll











H'Jl
!•••!










'ill
.1










t







i



ill
1
10









ll 1

0








1
il Lull HI liiii

150 200 £50 3
         SPEC if  32013 L M
STEP MflSS=l,
IV.
                                                     Figure D-4A

-------
00
80-[
60-1
50-E
tO-E
304
£0-|
1 0-|
0-^







Ihl
..,.(,,,,j,,,l|ll
0















	 ' i1

Ctt 31939 LM







ill
"''







nil
i-'-'|







l,,,li|l








iilill
II|IMI| 1






I,
l.lj
50







h








III





L








•kill ill lh '•! 11 i
l|llll l.ll III.,.., ,.l 1 III 	 |.l. ,...,,.„. 	 , 	 , 	 .....|,,,,|,,.,,.,,.j 	 1,...|. ...j....^...,™.!.!!!^!.^.!.!!!!!!!
100 150 £00 250
30eniNERflL SPIRITS 7-3-74 STEP rtfiSS=l,
                                                                                                           300
                                                     Figure  D-4B

-------
VO
1 00-3
904
80-1
704
504
c
30-=
204
i o4
0-









0















,
,i,,| ,.
ECtt 32015 LM







J

1





1
ill
|m,|ll
50







Jl







I









$ 5 OUTLET




1


111



1
















IU
1










,ll
00










1,1
jllll








„









,ll









I.,
1 r
15








ll
1 I""1 I""' •'••"'•-' 	 I--.....,...,.-[[[ 	
0 200 £50 3


-

-



00
STEP M fl S S = 1 , I/BXS

-------
1 00-q
H
9 0 ^«
80-1
E
70-f
604
50-E
40-f
30-i
20-|
O.J:











' 1
0




















1 1 I

E C if 31941 L M










1










l-'l









llll
1










1
50









,,
1





















III









i
Mllpll



11






. 1 II, ..n ..!: 1 1 Ij^
'!" 	 r • 	 l....|»..,....|....|....| 	 ,....,....,....,....1....,....,....,,...,....,....! 	 |....,..,.j,,.,|,..,|...,1.,,.( 	 p..,,,...!...,.,,,.
100 150 200 250 3
308 MIMERflL SPIRITS 7-3-74 STEP f1flSS=l, I/B^S
Figure D-5B

-------
1 00-g

 80J
 po-i

 so-i
 40-E
 304

 i o4
                      rm|H
        0              50
SPECtt  32020  LM   05  OUTLET
1 00
                                                  150
200            350             300
       STEP  Mfl S S =1,  IxB/S   =  IX
                                        Figure D-6A

-------
   1 00-q
    80-E
    70-=
       c
    60-?




    50-j









    30-i









    1 0~
                                           ii y
        0               50              100             150


SPEC!*  31917  LM    302  MINERflL  SPIRITS      7-3-74
200            250              300


        STEP  f 1 R S S = 1 ,  I / B •••• S    =  :
                                              Figure  D-6B

-------
1 00-T
90-j
80-
60-
50-
40-
30-
eo-
0-
—









inl i i<
	 "i"1'1" T"t








I) i,-!,]









	 1,
'"i"









1,..








,,,,llll









ii III ill III! Hill nil h, II Win! i In III 1 il
IH|IIII nil llll|l IJIII III |lll>|1liqilll|!iii|mi llll| lii|iiii|Iili|iilljilii|nil|liil|lli!|UIIIMil|lin|ilii|iiii|ilii|MiYiii|iiii|i!.:Miiinnl|l.,ljl!.. .iiiijiin. .ni|!..i|ii.. ( ...i.al
        0              50
SPECtt  3S025  LM   «5 OUTLET
100
150
200            250             300
       STEP  fl fl S S = 1 , I/B/S   =  iJi
    Figure D-7A

-------
   100-3
       t
    eo
       :

    60-E
    50-r





    HO-j





    30-f


       *


    £04





    i o-i
        0              50              100



SPECtt  31954 U1   302  MINERfiL  SPIRITS      7-3-74
     I'  I' I  I"'T"T"'|	|....,.M,|,..,|M..|M,.R.-,,.M,,,,,,|MM,,,.,p,,,,n.,jm.p,npm,m,j,TTT




150             200            £50             300




                        STEP  n n s s = i,  i/E xs   =  i
                                           Figure D-7B

-------
100-
90-!
80-|
70-1
604
50-1
HO-i
3o4
£0-^
0-











""'""I"1
0










r





















.|.i..,i ,.

ECtt 32030 LM









J|




1,1,










1 '


i Jin) !

50 1
«5 OUTLET









nlnl

00










III,












I












|,l|l!nl!ll .111

1












•
-
'.
-1 0
t
-
-
.
-

50 200 250 300
STEP MRSS=1, I/B/S = 1
Figure D-8A

-------
1 00-:
90-J
80-j
e,o-i
504
I
40-|
304
eo-j
1 O-j
0~










in
	 -Hi1
0










.{..nil































1 Jl
''""i
50









	 ,
'ni|ii!l|r









|
1









Jll|









-
i
i mi 1 il iii HI! llni iili'liiiill !| III li |ii '|inl|i iiiii
100 150 200 250
SPECtt  31959  LM    30£  MINERflL   SPIRITS
S T E P   M fl S S = 1 .,  I x E.-- S
-..-p-.r^ mpm -  (J



     300



       =  i;
                                                     Figure  D-8B

-------
1 00-:

 90 4

 80-j

 PO-j


 50-j

 H0-|

 30-1

 20-E
        0              50
S PEC«  32040 L M   «5 OUTLET
                                    1 00
150
                                                   30
                                                  -20
                                                  -1 0
200            250             300
        STEP  M fl s s = i,  i / B -' s    =  i :-•:
                                       Figure  D-9A

-------
00
            100-|
             90-i
             BOH
             70-
             40-j

             30-

             20-

             10-
                    _L
                                                                         1 0
J_ta
                 0              50             100
         SPECtt  31969 LM   302 HINERflL SPIRITS
                       150
                   7-3-74
200
       250
STEP MfiSS=l,  I/B/S
300
 =  IK
                                                      Figure D-9B

-------
v£>
100-
90-|
80-f
H
ri
30-j
SO -|
O..T






0




1 ll
IJIUIIII




'l1





,1





1,.
''""I""
50










1 i ill
'"'••'•I'"





100 150 200 250 3


30
SO
-10
ft
00
               SPEQtt  32052 LM   US  OUTLET
                                                        Figure D-10A

-------
00
o
                 1 00-
                   80-E
                   50-

                   40-|

                   30-1

                   20-E
                                                                                                                                                           •1 0
                                             T
                                             50                  100                 150
              SPECtt  31981   LN     302  MINERflL  SPIRITS       7-3-74
                                                                                                                                                              0
'"""I	|...T."|""j""|""p""i""r"'jTT"<-TTTmTi--r"i'"T"1i""l	
 200                £50                   300
           STEP  M fl S 5 = 1 ,   I / B / S    =   I'/.
                                                                              Figure D-10B

-------
                                 •TECHNICAL REPORT DATA
                          (Please read IntlTuctions on the reverse before completing)
 1. REPORT NO.
  EPA-650/2-74-107
                                                       3. RECIPIENT'S ACCESSION-NO.
 4
      AND SUBTITLE char acterization of Atmospheric
 Emissions from Polyurethane Resin Manufacture
             5. REPORT DATE
             October 1974
                                                       6. PERFORMING ORGANIZATION CODE
 7. AUTHOR(S)

 Wayne E.  Smith and John R. LaShelle
                                                       8. PERFORMING ORGANIZATION REPORT NO
 9. PERFORMING ORGANIZATION NAME AND ADDRESS
 Midwest Research Institute
 425 Volker Boulevard
 Kansas  City, Missouri  64110
             10. PROGRAM ELEMENT NO.
             1AB015; ROAP 21AXM-060
             11. CONTRACT/GRANT NO.

             68-02-0228 (Task 38)
 12. SPONSORING AGENCY NAME AND ADDRESS

 EPA, Office of Research and Development
 NERC-RTP, Control Systems Laboratory
 Research Triangle Park, NC 27711
             13. TYPE OF REPORT AND PERIOD COVERED
             Final: 10/73-10/74     	
             14. SPONSORING AGENCY CODE
 15. SUPPLEMENTARY NOTES
 16. ABSTRACT
           The report describes the characterization of air pollutant emissions from
 a polyurethane resin manufacturing plant. Samples were taken before and after the
 air pollution control device (scrubber). Analysis for toluene di-isocyanate (TDI) and
 amines was conducted both on site and on a delayed basis.  The sampling train incl-
 uded an impinger for colorimetric measurement of TDI.  Cryogenic traps and a tape
 sampler for TDI were also used for the scrubber inlet samples. The outlet sampling
 manifold consisted of evacuated bulbs in place of the cryogenic traps because of the
 high moisture content. TDI emissions were found to be maximum in the first part of
 the resin formation reaction.  Total emissions consisted  of 0.000 000 018 Ib TDI per
 Ib of product. The scrubber had negligible effect on the TDI emissions.
 7.
                             KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
                                           b.IDENTIFIERS/OPEN ENDED TERMS
                         c. COSATl Field/Group
Air Pollution
Polyurethane Resins
Manufacturing
Analyzing
  yanates
Amines
Scrubbers
Air Pollution Control
Stationary Sources
Characterization
Toluene Di-Isocyanate
13B
111, 11J
13H
14B
07C
 8. DISTRIBUTION STATEMENT
                                           19. SECURITY CLASS (This Report)
                                           Unclassified
                                                                    21. NC
                                                                          PAGES
Unlimited
20. SECURITY CLASS (This page)
 Unclassified
                                                                    22. PRICE
EPA Form 2220-1 (9-73)
                                        81

-------