EPA-600/2-78-004J
May 1978
Environmental Protection Technology Series
Industrial Invlrwitneiital Research Laboratory
ttfflce uf Rtsearch and
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U S Environmental
Protection Agency, have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields
The nine series are
1. Environmental Health Effects Research
2 Environmental Protection Technology
3. Ecological Research
4 Environmental Monitoring
5. Socioeconomic Environmental Studies
6 Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8. "Special" Reports
9 Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-78-0041
May 1978
SOURCE ASSESSMENT:
POLYVINYL CHLORIDE
by
Z. S. Khan and T. W. Hughes
Monsanto Research Corporation
Dayton, Ohio 45407
Contract No. 68-02-1874
Project Officer
Ronald J. Turner
Industrial Pollution Control Division
Industrial Environmental Research Laboratory
Cincinnati, Ohio 45268
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental
Research Laboratory-Cincinnati, U.S. Environmental Protection
Agency, and approved for publication. Approval does not signify
that the contents necessarily reflect the views and policies of
the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or
recommendation for use.
T;-.-S~r-~ •-:.-, ••- • *-
i-trVv.; ;.'. l^L, . _,
-------
FOREWORD
When energy and material resources are extracted, processed,
converted, and used, the related pollutional impacts on our
environment and even on our health often require that new and in-
creasingly more efficient pollution control methods be used. The
Industrial Environmental Research Laboratory - Cincinnati (IERL-
Ci) assists in developing and demonstrating new and improved
methodologies that will meet these needs both efficiently and
economically.
This report contains an assessment of air emissions from the
production of polyvinyl chloride. This study was conducted to
provide an overview of the information available on polyvinyl
chloride plants, including process technology, industry struc-
ture, control technology, and ambient concentrations. Further
information on this subject may be obtained from the Organic
Chemicals and Products Branch, Industrial Pollution Control
Division.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
m
-------
PREFACE
The Industrial Environmental Research Laboratory (IERL) of the
U.S. Environmental Protection Agency (EPA) has the responsibility
for insuring that pollution control technology is available for
stationary sources to meet the requirements of the Clean Air Act,
the Federal Water Pollution Control Act, and solid waste legisla-
tion. If control technology is unavailable, inadequate, or uneco-
nomical, then financial support is provided for the development
of the needed control techniques for chemical and extractive
process industries. Approaches considered include: process
modifications, feedstock modifications, add-on control devices,
and complete process substitution. The scale of the control
technology programs ranges from bench- to full-scale demonstra-
tion plants.
IERL has the responsibility for developing control technology for
a large number of operations (more than 500) in the chemical and
related industries. As in any technical program, the first step
is to identify the unsolved problems. Each of the industries is
to be examined in detail to determine if there is sufficient
potential environmental risk to justify the development of con-
trol technology by IERL. This report provides an overview of the
information available on polyvinyl chloride plants, including
process technology, industry structure, control technology, and
ambient concentrations.
Monsanto Research Corporation (MRC) has contracted with EPA to
investigate the environmental impact of various industries that
represent sources of emissions in accordance with EPA's respon-
sibility, as outlined above. Dr. Robert C. Binning serves as
Program Manager in this overall program, entitled "Source Assess-
ment," which includes the investigation of sources in each of
four categories: combustion, organic materials, inorganic mate-
rials, and open sources. Dr. Dale A. Denny of the Industrial
Processes Division at Research Triangle Park serves as EPA Pro-
ject Officer for this series. This study of polyvinyl chloride
plants was initiated by lERL-Research Triangle Park in March
1975; Mr. Kenneth Baker served as EPA Project Leader. The pro-
ject was transferred to the Industrial Pollution Control Divi-
sion, lERL-Cincinnati in October 1975; Mr. Ronald J. Turner
served as EPA Project Leader from that time through completion of
the study.
IV
-------
ABSTRACT
This report describes a study of air emissions from polyvinyl
chloride production. The study was completed to provide EPA with
sufficient information to determine whether additional control
technology needs to be developed for this emission source.
Polymers derived from vinyl chloride monomer (VCM) are called
polyvinyl chloride (PVC). The 1974 consumption of PVC was
2.2 x 106 metric tons/yr which represented the third largest
volume of plastic consumed. PVC is manufactured by 20 companies
at 35 plants. Each plant uses one or more of four possible poly-
merization processes: 1) suspension polymerization, 2) emulsion
polymerization, 3) bulk polymerization, and 4) solution poly-
merization, which account for 78%, 13%, 6% and 3% of the total
production capacity, respectively.
A representative PVC plant is defined as one using the suspension
process and having a nominal production capacity of 68 x 103 met-
ric tons, a population density surrounding the plant of 313
persons/km2, average emission heights for vinyl chloride monomer
and polyvinyl chloride resin dust (particulates) of 15.5 m and
21 m, respectively. The emission factors for vinyl chloride and
polyvinyl chloride from the representative plant are 35.5 g/kg
and 7.5 g/kg, respectively.
To assess the potential environmental effect of emissions from
this industry, the source severity (defined as the ratio of the
time-averaged maximum ground level concentration of a pollutant
to a hazard potential) was calculated for 16 chemical species
emitted from the representative plant. The two largest source
severities were for vinyl chloride (970) and polyvinyl chloride
(1.9) .
Polyvinyl chloride production in 1974 amounted to 2.22 x 106
metric tons and is expected to grow at a rate of 5.2%/yr through
1978 when production is estimated to be 2.86 x 106 metric tons.
If the 1978 level of emissions control is the same as the 1973
level, emissions from PVC manufacture will increase by 29% over
that period.
This report was submitted in partial fulfillment of Contract No.
68-02-1874 by Monsanto Research Corporation under the sponsorship
of the U.S. Environmental Protection Agency. This report covers
the period March 1975 to July 1977, and work was completed as of
August 1977.
-------
CONTENTS
Foreword iii
Preface iv
Abstract v
Figures viii
Tables x
Abbreviations and Symbols xii
Conversion Factors and Metric Prefixes xiv
1. Introduction 1
2. Summary 2
3. Source Description 8
Process description 8
Materials flow 29
Geographical distribution 29
4. Emissions 32
Locations and descriptions 32
Emission factors 35
Definition of representative source 39
Environmental effects 39
Growth factor 54
5. Control Technology 56
Control technology for hydrocarbons 56
Control technology for particulates 68
6. Growth and Nature of the Industry 70
Present technology 70
Emerging technology 71
Industry production trends 72
Outlook 77
References 78
Appendix 83
-------
FIGURES
Number Page
1 Suspension polymerization 9
2 Block flow diagram for production of polyvinyl
chloride by the suspension process 11
3 Flowsheet for production of polyvinyl chloride
by the suspension process 12
4 Effect of initiator on reaction rate during
production of PVC by suspension polymerization. 16
5 Block flow diagram for production of polyvinyl
chloride by the emulsion process 21
6 Flowsheet for production of polyvinyl chloride
by the emulsion process 22
7 Block flow diagram for production of polyvinyl
chloride by the bulk processes 24
8 Flowsheet for production of polyvinyl chloride
by the bulk process 25
9 Block flow diagram for the production of poly-
vinyl chloride by the solution process 27
10 Flowsheet for production of polyvinyl chloride
by the solution process 28
11 Simplified material balance for suspension
process 30
12 Polyvinyl chloride plant locations 31
13 Cumulative percent of PVC plants having an emis-
sion rate and a source severity less than or
equal to indicated value 48
14 Cumulative percent of PVC plants having an emis-
sion rate and a source severity less than or
equal to indicated value 49
-------
Number
FIGURES (continued)
15 Cumulative percent of samples having ground
level concentrations less than or equal to
indicated value 52
16 Stages of plant slurry stripping 66
17 Polyvinyl chloride production, 1946-1979 .... 73
18 U.S. consumption of polyvinyl chloride by
compounding process 75
19 U.S. consumption of polyvinyl chloride by
end use 75
IX
-------
TABLES
Number Page
1 Emission Factors for a Representative Polyvinyl
Chloride Plant (Suspension Process) 3
2 Source Severity by Compound for a Representative
Polyvinyl Chloride Plant 4
3 Polyvinyl Chloride Industry Contributions to
National Stationary Source Emissions of
Criteria Pollutants 5
4 Polyvinyl Chloride Industry Contributions to
State Emissions of Criteria Pollutants 6
5 Raw Materials for Suspension Polymerization of
Vinyl Chloride 14
6 Vinyl Chloride Monomer Composition 14
7 Locations and Capacities of Polyvinyl Chloride
Manufacturing Plants 29
8 Points of Emission at a Representative Polyvinyl
Chloride Plant 32
9 Vinyl Chloride Emission Factors for Polyvinyl
Chloride Processes 36
10 Polyvinyl Chloride Industry Contributions to
National Stationary Source Emissions of
Criteria Pollutants 36
11 Polyvinyl Chloride Industry Contributions to
State Emissions of Criteria Pollutants 37
12 Characteristics of Emissions from a Representa-
tive Polyvinyl Chloride Plant 38
13 Polyvinyl Chloride - Summary of Plant Data - I. . 40
-------
TABLES (continued)
Number Page
14 Summary of Criteria used to Define a Representa-
tive Polyvinyl Chloride Plant 41
15 Emission Factors for a Representative Polyvinyl
Chloride Plant 41
16 Polyvinyl Chloride Manufacturing Emission
Factors by Point of Emission for a Representa-
tive Plant (g/kg) 42
17 Polyvinyl Chloride - Summary of Plant Data - II. 44
18 Time-Averaged Maximum Ground Level Concentration
by Compound for a Representative Polyvinyl
Chloride Plant 45
19 Time-Averaged Maximum Ground Level Concentration
for Emissions from a Representative Polyvinyl
Chloride Plant by Point of Emission 45
20 Source Severity by Compound for a Representative
Polyvinyl Chloride Plant 46
21 Source Severity for a Representative Polyvinyl
Chloride Plant by Point of Emission 46
22 Input Data 47
23 Summary of Sampling Results 51
24 Comparison of Measured and Calculated Emission
Data 51
25 Controlled Vinyl Chloride Emissions 53
26 Affected Area and Affected Population 54
27 Control Technology for Polyvinyl Chloride
Manufacture 57
28 Losses of Monomer in Three Stages of Stripping of
Batch from Slurry of 4,540 kg Monomer Charge . 65
29 United States Consumption of Polyvinyl Chloride
Resin by Compounding Process "74
30 Consumption of Polyvinyl Chloride by Major
Markets 76
31 United States Consumption of Polyvinyl Chloride
Resins by End Use 77
XI
-------
ABBREVIATIONS AND SYMBOLS
A — affected area
AR — factor defined as Q/aciru
AAQS — ambient air quality standard
a,..f,v..z -- constants
BR -- factor defined as -H2/2c2
C^ -- production capacity of plant i
Dp — capacity-weighted mean population density
Dp. -- county population density for plant i
e — constant; 2.72
exp -- natural log base, e
F — hazard factor; for criteria pollutants, F is
the primary ambient air quality standard; for
noncriteria pollutants, F is a reduced TLV
value
H — effective emission height
P — total affected population
PVC — polyvinyl chloride
Q — mass emission rate
S — source severity
SHQ -- source severity for hydrocarbon emissions
Sp -- source severity for particulate emissions
SVCM -- source severity for vinyl chloride
t -- averaging time for ambient air quality standard
t — "instantaneous" averaging time
TLV — threshold limit value
u -- wind speed
u -- average wind speed
VCM — vinyl chloride monomer
x -- downwind distance from source of emission
xx, x2 — roots of equation for affected area calculation
xii
-------
ABBREVIATIONS AND SYMBOLS (continued)
y -- horizontal distance from centerline of
dispersion
z -- vertical distance from centerline of dispersion
TT -- constant; 3.14
o -- standard deviation of horizontal dispersion
a -- standard deviation of vertical dispersion
Z
X -- downwind ground level concentration of a
pollutant
X -- average downwind ground level concentration of
a pollutant
X -- maximum ground level concentration of a
max pollutant
X -- time-averaged maximum ground level concentration
of a pollutant
X(x) -- annual mean ground level concentration of a
pollutant at a specific distance (x) from the
source
-------
CONVERSION FACTORS AND METRIC PREFIXES
To convert from
Degree Celsius (°C)
Grams/second (g/s)
Joule (J)
Kilogram (kg)
Kilogram/meter3 (kg/m3)
Kilometer2 (km2)
Meter (m)
Meter3 (m3)
Metric ton
Metric ton
Metric ton
Pascal (Pa)
Pascal (Pa)
Pascal (Pa)
CONVERSION FACTORS
to
Degree Fahrenheit
Pounds/hr
Calorie
Pound-mass
(avoirdupois)
Pound/foot3
Mile2
Foot
Feet3
Kilogram
Pound-mass
Ton (short, 2,000 pound
mass)
Atmosphere
Torr (mm Hg, 0°C)
Pound-force/inch (psi)
Multiply by
t° = 1.8 t° + 32
7.936
2.388 x 10-1
2.204
6.243 x 10-2
3.860 x 10-1
3.281
3.531 x 101
1.000 x 103
2.205 x 103
1.102
9.869 x 10~6
7.501 x 10-3
1.450 x 10-1*
METRIC PREFIXES
Prefix Symbol Multiplication factor
Example
Kilo
Mega
Micro
k
M
v
103
106
ID'6
1 kPa = 1 x 103 pascals
1 MJ = 1 x 106 joules
1 g = 1 x 10"6 gram
xiv
-------
SECTION 1
INTRODUCTION
Products and components fabricated from polyvinyl chloride are
used in nearly every branch of industrial and commercial activity
— building and construction, home furnishings, consumer goods,
electrical devices and goods, packaging, and transportation.
Polyvinyl chloride resins are manufactured by four polymerization
processes: suspension polymerization (the largest production
method), emulsion polymerization, bulk polymerization, and
solution polymerization.
This document assesses the atmospheric emissions and potential
environmental effects of polyvinyl chloride polymerization proc-
esses, using data determined from literature references.
The major findings of this study are summarized in Section 2.
Section 3 provides detailed descriptions of the polyvinyl chlo-
ride polymerization processes, including the major processing
steps, flow diagrams, process chemistry, and material and energy
balances.
Section 4 discusses types of emissions, emission points, mass of
emissions, ground level concentrations, source severity and
affected population.
Section 5 considers the present and future aspects of pollution
control technology in the polyvinyl chloride industry. The
growth and nature of the industry are discussed in Section 6.
-------
SECTION 2
SUMMARY
Polymers derived from vinyl chloride monomer (VCM) are called
polyvinyl chloride (PVC). These polymers may be homopolymers,
which are made only from vinyl chloride monomer, or copolymers,
made from vinyl chloride monomer and another monomer such as
vinyl acetate, ethylene, propylene, vinylidene chloride, or an
acrylate. Polyvinyl chloride is used in the manufacture of
apparel, building and construction materials, wire and cable
insulation, home furnishings, packaging, recreation items,
transportation components, and other saleable commodities. The
1974 consumption of polyvinyl chloride was 2.2 x 106 metric
tons/yr;a it was the third largest volume plastic, trailing only
low density polyethylene and styrenics in consumption.
PVC is manufactured by 20 companies at 35 plants. With individ-
ual production capacities ranging from 6.8 x 103 metric tons/yr
to 136 x 103 metric tons/yr, these plants have a combined
capacity of 2.4 x 106 metric tons/yr. Each of the plants uses
one or more of four possible polymerization processes: 1) sus-
pension polymerization, 2) emulsion polymerization, 3) bulk
polymerization, and 4) solution polymerization, which account
for 78%, 13%, 6% and 3% of the total production capacity, respec-
tively. PVC production causes atmospheric emissions consisting
of criteria pollutants and chemical substances. Criteria pol-
lutants include hydrocarbons (volatile organic materials),
sulfur oxides (SOX), and particulates (airborne polyvinyl chlo-
ride resin dust). Chemical substances include vinyl chloride,
ethylene, propylene, acetylene, butadiene, ethylene dichloride,
vinylacetate, vinyl bromide, vinylidene chloride, acetaldehyde,
ethyl chloride, chloroprene, hydrogen chloride, and phenol.
Table 1 lists emission factors for atmospheric emissions from a
representative polyvinyl chloride plant using the suspension
process. These materials are emitted inside each plant from
reactor safety relief valves, reactor entry purges, stripper
jets, monomer recovery condenser vents, slurry blend tank vents,
centrifuge vents, dryer discharges, resin storage silos, bulk
loading facilities, bagger vents, storage tanks, and fugitive
emission points. In a typical PVC plant, there may be 600 or
more separate emission points.
al metric ton = 106 grams = 2,205 pounds; conversion factors and
metric system prefixes are presented in the prefatory pages.
-------
TABLE 1. EMISSION FACTORS FOR A REPRESENTATIVE POLYVINYL
CHLORIDE PLANT (SUSPENSION PROCESS)
Emission factor,
Material emitted g/kg
Vinyl chloride
Polyvinyl chloride
Stabilizer (phenol)
Ethyl chloride
Sulfur oxides
Butadiene
Hydrogen chloride
Vinylidene chloride
Acetaldehyde
Acetylene
Propylene
Vinylacetylene
Ethylene
Ethylene dichloride
Chloroprene
Vinyl bromide
35.5 ± 8.24a
7.5 ± 3.18a
204 x 10~5
92 x 10~5
23 x 10~5
21 x 10~5
21 x 1CT5
9 x 10~5
7 x 10" 5
7 x 10~5
7 x 10~ 5
5 x 10~5
5 x 10~5
4 x 10~5
<4 x 10~5
2 x 10-5
These values indicate the mean values
for the emission factor; the 95%
confidence limit is given in g/kg.
Atmospheric emissions from a representative PVC plant have been
determined and are presented in this document. A representative
PVC plant was defined as one using the suspension process and
having the following mean values for various plant parameters:
• Nominal production capacity of 68 x 103 metric tons.
• Population density surrounding the plant of 313 persons/km2
• Average emission height for vinyl chloride monomer of
15.5 m.
• Average emission height for polyvinyl chloride resin dust
(particulates) of 21 m.
• Total vinyl chloride emission factor of 35.5 g/kg.
• Total polyvinyl chloride emission factor of 7.5 g/kg.
The representative plant (defined on page 39) is typical of 78%
of the total production capacity of PVC in the United States.
The remaining 22% of production capacity consists of polyvinyl
chloride produced via the emulsion, bulk, and solution polymer-
ization process; these have vinyl chloride emission factors of
60.1 g/kg, 24.2 g/kg, and 17.8 g/kg, respectively.
-------
Atmospheric emissions from the representative source have been
characterized in this assessment by calculation of a source sever-
ity, determination of the national burden of criteria pollutants,
determination of state burdens of criteria pollutants, estimation
of the population affected by the atmospheric emissions, and
estimation of the rate of increase of emission with time.
Source severity is defined as the time-averaged maximum ground
level concentration divided by a hazard potential. The time-
averaged maximum ground level concentration is determined using
Gaussian plume dispersion methodology. The hazard potential is
equal to the primary ambient air quality standard for criteria
pollutants and to a reduced threshold limit value (TLV®) for
chemical substances. Table 2 lists source severities and
TLV's for atmospheric emissions from a representative polyvinyl
chloride plant.
TABLE 2. SOURCE SEVERITY BY COMPOUND FOR A
REPRESENTATIVE POLYVINYL CHLORIDE PLANT
Material emitted
Vinyl chloride
Polyvinyl chloride
Stabilizer (phenol)
Ethyl chloride
Sulfur oxides
Butadiene
Hydrogen chloride
Vinylidene chloride
Ace t aldehyde
Acetylene
Propylene
Vinylacetylene
Ethylene
Ethylene dichloride
Chloroprene
Vinyl bromide
TLV,
g/m3
0.
0.
0.
2.
0.
2.
0.
0.
0.
1.
1.
0.
1.
0.
0.
1.
0026
1090
02
60
013
20
007
004
18
16
88
048
25
20
09
10
Source
severity3
970
7.
2.
1.
6.
2.
1.
2.
4.
2.
7.
2.
1.
3.
1.
2
5
3
8
1
6
8
3
6
4
8
4
2
3
1
X
X
X
X
X
X
X
X
X
X
X
X
X
X
.9
10
10-
10~
10"
10-
10-
10-
10-
10-
10-
10~
10-
10-
10-
3
5
3
6
3
3
5
6
6
5
6
5
5
6
a_ ., Amax
Source severity = TLV x 8/24 x 1/100
The national mass of criteria pollutants emitted from PVC plants
and their percent contributions to national emissions are shown
in Table 3.
-------
TABLE 3.
POLYVINYL CHLORIDE INDUSTRY CONTRIBUTIONS TO NATIONAL
STATIONARY SOURCE EMISSIONS OF CRITERIA POLLUTANTS
Emissions from the
PVC industry
Total national Percent
emission (1), Of national
Material emitted 10 metric tons/yr 10 metric tons/yr emissions
Hydrocarbons
(vinyl chloride,
phenol stabilizer,
ethyl chloride,
butadiene, vinyl-
idene chloride,
acetaldehyde,
acetylene, propy-
lene, vinylacety-
lene, ethylene,
ethylene dichlo-
ride, chloroprene,
vinyl bromide)
Particulate
(polyvinyl
chloride)
Sulfur oxides
25
85
0.34
18
30
18
0.10
5.5 x 10-1* 2 x ID"6
Hydrocarbon emissions from PVC plants in New Jersey, Massachu-
setts, West Virginia, Delaware, Kentucky, Oklahoma, and Missis-
sippi range from 1.1% to 6.9% of each state's hydrocarbon emis-
sions. New Jersey has particulate emissions from PVC plants
which represent 1.6% of the state's total particulate emissions.
All other states with PVC plants have hydrocarbon and particu-
late emissions from polyvinyl chloride manufacture which are
less than 1% of the state totals. Table 4 gives a complete list
of polyvinyl chloride industry contributions to state emissions
of criteria pollutants.
PVC plants are located in counties with population densities
ranging from 9 to 1,900 persons/km2, with the population for the
representative source being 313 persons/km2. The area sur-
rounding the representative plant for which the source severity
for vinyl chloride monomer is greater than or equal to 0.1 was
calculated to be 2,780 km2. The affected population is thus
870,000 persons for the representative source.
Polyvinyl chloride production in 1973 amounted to 2.22 x 106
metric tons and is expected to grow at a rate of 5.2%/yr through
1978 when production is estimated to be 2.86 x 106 metric tons.
If the 1978 level of emissions control is the same as the 1973
level, emissions from PVC manufacture will increase by 29% over
that period.
(1) 1972 National Emissions Report. EPA-450/2-74-012, U.S.
Environmental Protection Agency, Research Triangle Park,
North Carolina, June 1974. 422 pp.
-------
TABLE 4. POLYVINYL CHLORIDE INDUSTRY CONTRIBUTIONS TO
STATE EMISSIONS OF CRITERIA POLLUTANTS (1)
State
New Jersey
Massachusetts
Ohio
California
West Virginia
Illinois
Texas
Delaware
Louisiana
New York
Kentucky
Florida
Maryland
Oklahoma
Mississippi
Pennsylvania
Material emitted
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
State emissions,
10 3 metric tons/yr
819.5
151.8
440.5
96.16
1,153
1,766
2,161
1,006
116.2
213.7
1,826
1,143
2,219
549.4
63.89
36.81
1,920
380.6
1,262
160
326.3
546.2
619.9
226.5
295.9
494.9
341.4
93.6
196
168.4
891.8
1,811
PVC emissions
metric tons/yr
9,450
2,370
5,845
855
10,620
3,030
2,145
1,505
8,030
380
7,090
1,775
6,980
1,820
2,770
2,870
6,915
715
2,055
150
7,380
1,295
1,075
490
1,860
510
3,660
65
4,185
885
4,150
365
Percent
1.15
1.56
1.33
0.09
0.92
0.17
0.10
0.15
6.91
0.18
0.39
0.16
0.31
0.33
4.34
0.78
0.36
0.19
0.16
0.09
2.26
0.24
0.17
0.22
0.63
0.10
1.07
0.07
2.14
0.52
0.47
0.02
Vinyl chloride hydrocarbon emission; PVC particulate emission.
Available emissions control technology is divided into hydrocar-
bon control and particulate control. Controls for hydrocarbons
include adsorption, absorption, refrigeration, incineration,
stripping, purging of equipment with inert gas or water, and con-
trol of fugitive emissions.
Fugitive emissions are being reduced through the use of a monitor-
ing program. Double mechanical seals and leakproof metal
discs are used to control fugitive emissions from leaking pumps,
-------
compressors, agitators, seals, and pressure relief valves. Proc-
ess modifications to control fugitive emissions include the use
of larger reactors in newer plants. Since large reactors require
fewer connections, they reduce fugitive emissions by reducing
the number of potential leaks. EPA standards on vinyl chloride
emissions require a 95% or better reduction in atmospheric vinyl
chloride emissions.
Particulate emissions are being controlled through the use of
fabric filters and cyclonic collectors.
When the link between VCM and cancer was established in 1974, the
PVC industry entered a new era. The emission and control data
used in this report were obtained in 1974 before the recent
dynamic changes in the industry. Many plants currently report
meeting emission standards established by EPA and OSHA.
-------
SECTION 3
SOURCE DESCRIPTION
Polyvinyl chloride is one of the three largest volume thermo-
plastics produced in the United States (2). In 1974 there were
20 companies at 35 locations in the United States (3) capable of
producing 2.45 x 106 metric tons/yr of PVC. PVC is produced by
VCM polymerization in one of four processes. Three of these are
batch processes - suspension polymerization (which accounts for
78% of capacity); emulsion polymerization (13%); and bulk polym-
erization (6%). The fourth process, solution polymerization, is
a continuous process and accounts for 3% of the PVC resin
capacity (4).
The suspension process for manufacturing PVC is described in de-
tail below. The other three processes are briefly described in
terms of their variations from the suspension process.
PROCESS DESCRIPTION
Suspension Polymerization
The suspension process accounts for 78% of the polyvinyl chloride
homopolymers and copolymers made in the United States (5) .
(2) Olivier, G. What's the Future for PVC? Hydrocarbon Process-
ing, 45(9):281-284, 1966.
(3) PVC Chemical Profile. Chemical Marketing Reporter.
205(20) :9, 1974.
(4) Evans, L. , C. Kleeberg, S. Wyatt, A. Basola, W. Hamilton and
W. Vatavuk. Standard Support - Environmental Impact Document,
An Investigation of Health Effects and Emission Reduction of
Vinyl Chloride in the Vinyl Chloride Monomer and Polyvinyl
Chloride Industries. Volume II. Draft copy of report.
U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, March 1975. 450 pp.
(5) Carpenter, B. H. Vinyl Chloride - An Assessment of Emissions
Control Techniques and Costs. EPA-650/2-74-097, U.S. Envi-
ronmental Protection Agency, Washington, D.C., September
1974. 84 pp.
8
-------
Suspension polymerization is applied to a system in which th
water-insoluble monomer is suspended as liquid droplets, and
the resultant polymer is obtained as a dispersed solid phase.
Initiators used in the process are soluble in the liquid mono
phase (6) .
Figure 1 depicts dispersed vinyl chloride droplets suspended ,
a water medium. Mechanical agitation causes the larger unstob
droplets to break into smaller ones. The smaller droplets
coalesce and reform into larger droplets until a dynamic equi!
rium between dispersion and coalescence is reached (7).
VINYL
CHLORIDE
MONOMER
MECHANICAL AGITATION
ADSORBED MOLECULAR FILM
OF PROTECTIVE COLLOIDS
1 INTERRACIAL TENSION
rP~000_
00° o§°
PROTECTIVE COLLOIDS
TO STABILIZE
DROPLETS
n^ 0°
0°o°o0o
o°o o
o
Figure 1. Suspension polymerization.
(6) Suspension Polymerization. In: Encyclopedia of Polymer
Science and Technology; Volume 13: Plastics, Resins, Rubb,
Fibers. John Wiley & Sons, Inc., New York, New York, 19/T
pp. 552-571.
(7) Albright, L. F. Polymerization of Vinyl Chloride. Cheioi,
Engineering, 74(10):151-158, 1967.
-------
Protective colloids, which are water-soluble, are added to stabi-
lize the vinyl chloride droplets and help prevent agglomeration
of PVC droplets. Colloids increase the viscosity of the water
layer and delay coalescence (7).
A water-soluble initiator starts the polymerization process. As
polymerization occurs, the viscosity of the organic phase in-
creases, and polymer molecules form throughout the droplets (7).
PVC produced in the dispersed vinyl chloride droplets forms a
solid phase, since polyvinyl chloride is insoluble in vinyl
chloride (7). At this stage, the reduction in vinyl chloride
concentration is accompanied by an increase in the polymerization
rate instead of the expected decrease. Autoacceleration of the
reacting medium (7) explains this phenomenon.
Polymerization occurs in both the VCM and PVC phases. In the
PVC phase, VCM diffuses to the active sites located throughout
the semisolid PVC phase, and polymerization occurs. Due to
limited mobility in the PVC phase, these active sites cannot
react with each other to cause coupling or disproportionation
reactions that destroy free radicals (7).
Figure 2 is a block flow diagram (8); Figures 3a and 3b are
detailed flowsheets (4, 5, 8-12) of the manufacture of PVC by
the suspension process. Figures 3a and 3b designate representa-
tive plant emission points which are described later in Table 8
in Section 4.
Raw Materials--
The raw materials required for the suspension polymerization of
vinyl chloride include vinyl chloride monomer, initiator, sus-
pending agent, emulsifier, and deionized water.
There are numerous patents for suspension polymerization recipes
(formulation) available in the literature. The basic formulation
is shown in Table 5.
(8) Kardos, L. A. Polyvinyl Chloride. Report No. 13 (a private
report by the Process Economics Program), Stanford Research
Institute, Menlo Park, California, June 1966. 224 pp.
Labine, R. A. Drying Tricks Tailor Resin Properties.
Chemical Engineering, 66(23):166-169, 1959.
(10) From Vinyl Chloride to...PVC by Suspension Polymerization.
Chemical Engineering, 62 (7):128, 130, 132, 1955.
(11) Ohta, K. Polyvinyl Chloride - Supplement-A. Report No.
13A (a private report by the Process Economics Program),
Stanford Research Institute, Menlo Park, California, May
1970. 170 pp.
(12) Albright, L. F. Vinyl Chloride Polymerization by Suspension
Process Yields Polyvinyl Chloride Resins. Chemical Engineer-
ing, 74(12):145-152, 1967.
10
-------
0)
T3
•H
M
c
•H
>
>(
rH
O
1
11
-------
0>
'-C
•rH
^
o
rH
A
O
>1
C
•H
>
>1
rH
O
a
m
o •
in
C tn
O o
-rH 0
-p o
u M
3 a
T3
o c
M O
dl-rH
tfi
M C
O 0)
MH a
en
-P 3
QJ cn
a)
,d a)
w ^
B^
r-l >,
fa ,Q
(0
ro
(1)
M
3
tn
-rH
PH
12
-------
— —I ^
CO
10
-------
TABLE 5. RAW MATERIALS FOR SUSPENSION
POLYMERIZATION OF VINYL CHLORIDE (8)
Component
Parts by weight
Vinyl chloride monomer
Deionized water
Initiator
Suspending agent
Emulsifier
100
200
0.025
0.04
0.02
Vinyl chloride monomer—Pure monomer is used in polymerizing
vinyl chloride by suspension polymerization. Specifications for
commercial VCM along with the major impurities present are listed
in Table 6 (9, 13, 14) .
Shipping of uninhibited monomer is permitted in the United
States. Some VCM is delivered to the PVC plant containing
100 ppm phenol which is used to prevent polymerization during
shipment. This inhibitor must be removed by scrubbing with
aqueous sodium hydroxide (9).
TABLE 6. VINYL CHLORIDE MONOMER COMPOSITION
Material
Maximum level, ppm
Vinyl chloride
Ethylene
Propylene
Acetylene
Butadiene
Ethylene dichloride
Vinylacetylene
Vinyl bromide
Vinylidene chloride
Acetaldehyde
Ethyl chloride
Chloroprene
Hydrogen chloride
Iron
Sulfur
Water
Nonvolatiles
Stabilizer (phenol)
<99.9%
1.3
2.0
2.0
6.0
0.1 to 2
0.01 to 3.0
0.05 to 1
0.1 to 5
2.0
2.0 to 50
2.0 to 10.0
0.4
3.0 to 10
15.0 to 200
10 to 200
25 to 90
(13) Lunde, K. E. Vinyl Chloride. Report No. 5 (a private re-
port by the Process Economics Program), Stanford Research
Institute, Menlo Park, California, 1965. 212 pp.
(14) Matheson Gas Data Book, Fourth Edition. The Matheson Com-
pany, Inc., East Rutherford, New Jersey, 1966. pp. 489-492
14
-------
Deionized water—Water used for suspension polymerization
is deionized, deaerated, and free of organic matter and sulfur
(10, 15, 16). Water serves three purposes; it provides heat
transfer, it is a medium for the suspending agent which controls
the surface properties of the particles and it also moderates
the bulk viscosity during processing (17).
Initiators--Initiators are compounds capable of forming free
radicals by thermal decomposition (8). Patent literature reports
the use of many initiators. Suspension polymerization of vinyl
chloride monomer is initiated by organic peroxides in industrial
practice, although azo compounds, boron derivatives, and redox
systems can be used (8).
Major producers report using isopropyl peroxide carbonate (IPP)
(11). Figure 4 shows conversion of VCM as a function of reaction
time for polymerization of vinyl chloride using two different
initiators (18). Advantages claimed for the use of IPP initiator
include: 1) reduced batch time; 2) little or no induction
period; 3) improved polymer quality because of fewer initiator
fragments; and 4) less chain branching during polymerization.
Suspending agents--Suspending agents are surface active compounds
that prevent agglomeration of PVC particles during polymerization
of vinyl chloride (2, 17). The suspending agent influences
particle size, porosity, and thus processing characteristics of
the product.
Conventional suspending systems such as natural gums and gelatin,
or synthetic polymers such as partially hydrolyzed polyvinyl
acetate (polyvinyl alcohol - polyvinyl acetate) and methyl
cellulose, efficiently promote transition from a monomer droplet
containing precipitated PVC to a polymer particle swollen with
monomer (17). However, these systems produce a resin that does
not readily absorb plasticizers, requiring high processing
(15) Ruebensaal, C. F. Vinyl Resins - How Vinyl Chloride is
made...How Vinyl Chloride is Polymerized. Chemical Engineer-
ing, 57(12):102-105, 1950.
(16) Meinhold, T. F., and W. M. Smith. Produces Dust Free PVC
Resins. Chemical Processing, 22(7):61-62, 1959.
(17) Manufacture of Plastics, Volume I. Chapter 7. Reinhold
Publishing Corporation. W. M. Smith, ed. New York,
New York, 1964. pp. 303-343.
(18) Marous, L. F., and C. D. McCleary. Polymerization Catalyst
for Vinyl Chloride. U.S. Patent 3,022,282 (to United States
Rubber Company), February 20, 1962.
15
-------
120
105
90
75
60
45
30
15
0
-•\'\ii
TEMPERATURE: 50°C
0.025 PART ISOPROPYL
PERCARBONATE
0.03 PART I SOPROPYL //
' PERCARBONATE
/
/-
CONTROL 0.25 PART
LAUROYL PEROXIDE
I
I
4 6 8 10
REACTION TIME, hr
12 14
Figure 4. Effect of initiator on reaction rate during
production of PVC by suspension polymerization.
temperatures or premastication in an intensive mixer or extruder
before the final processing to plastic products (11).
Industry is presently investigating all synthetic suspending
systems that produce "easy processing" or "fast blending" resins.
These resins have high plasticizer absorption capacity in dry
blending and greater ease of homogenization when subjected to
heat and mechanical shear of extrusion or calendering (17).
Emulsifier—Processing of the final product is improved by the
addition of small quantities of a secondary emulsifier to the
system (17). Such emulsifiers include sulfonated oils or esters,
ethylene oxide condensation products with polyols, and other
synthetic surfactants.
Raw material storage and purification—Vinyl chloride monomer
received at the plant contains phenol stabilizer to prevent
polymerization during shipping. The stabilizer is removed from
the monomer by caustic washing or by distilling the monomer from
an aqueous caustic mixture (15,17). Heavy impurities are removed
by distillation in a plate column, and the liquid monomer is
stored in a refrigerated stainless steel or glass lined tank at
less than 16°C and 345 kPa to retard hydrolysis and peroxide
formation (17).
16
-------
Water is deionized in a mixed-bed ion exchanger and then deaerated
by heat and vacuum in the deareator column before use in the
process (10,15,16).
The suspending agent and emulsifier are dissolved in the suspend-
ing agent solution makeup tank using deionized water. Separate
charge pots and storage tanks are used to mix and proportion the
initiator, hydrochloric acid, and caustic (8).
Polymerization--
Vinyl chloride polymerization is carried out in stainless steel,
glass-lined carbon steel, or glass-lined stainless steel reactors,
depending on raw materials used, corrosion resistance, and
desired lifetime of reactors (6). Reactor sizes vary between
11.3 m3 and 103.2 m3; each plant uses 4 to 18 such reactors (4).
Newer plants tend to have larger and fewer reactors (4). Each
reactor is equipped with an agitator, baffles, and temperature
controls (18) .
The reactor is charged first with deionized, deaerated water;
then the suspending agent solution is introduced. The tempera-
ture of the reactor is raised to 55°C by passing steam through
the reactor jacket. The initiator is placed in the charge pot
and dissolved by the liquid monomer as it is fed through the
batch meter (8).
Cooling water is circulated through the reactor jacket to keep
the temperature at 55°C during the polymerization (8).
An agitator located at the bottom of the vessel uses multiple
baffles and/or multiblade shafts to provide uniform agitation
(17), which is important for both efficient heat transfer and
control of polymer particle size (6, 17).
Reaction temperature is one of the primary control variables in
suspension polymerization (6). Temperature influences molecular
weight, molecular weight distribution, crystallicity of the pro-
duct, the particle size of the polymer and the solubility and
adsorptivity of the suspending agent (6). A master-slave cascade
instrument system is used for temperature control. Steam, cold
water, and refrigerated water or brine are circulated through the
reactor jacket as required. The polymerization temperature can
be controlled with 30°C cooling water up to 70% conversion.
Subsequently, the reaction rate increases more rapidly due to
autoacceleration. At this point, refrigerated water at 16°C is
required to control the temperature (8).
Polymerization takes place at a pressure of 517 kPa to 690 kPa
(6). Reactors are protected from overpressure by safety relief
valves and rupture discs. Completion of the reaction is indi-
cated by a drop in pressure. Prolongation of the cycle is
harmful to resin porosity and color (7). The cycle is termin-
ated at 88% conversion (276 kPa) by blowing the slurry to the
batch strippers (8).
17
-------
Unreacted VCM is sent by vacuum to the recovery system and recycled
Noncondensable gases accumulate in the recovery system and must be
vented.
Monomer Recovery and Slurry Blending--
In many plants, slurry from the reactor is transferred to a strip-
per for removal of unreacted vinyl chloride by the application of
heat and/or vacuum. Stripping can also be completed effectively
in the reactor, but most producers do not use reactors for the
time-consuming stripping operation. Vent gas from the stripper
is transferred to the vapor recovery system for recycling (4).
The monomer-free polymer slurry is transferred to the slurry
blend tank, where various batches are blended together to form a
uniform product. Slurry blending tanks also serve as a buffer
volume between the batch polymerization in the reactor and the
continuously operated equipment downstream (8). These tanks are
open and release residual VCM to the atmosphere.
Polymer Dewatering and Drying--
Slurry from the blend tank is pumped to a centrifuge for separa-
tion of the polymer and water. The centrifuge is conical; the
bowl rotates at 500 rpm while a plow mechanism rotates in the
same direction but at reduced speeds. Solids containing about
30% moisture are transported to the small end of the bowl, and
water is discharged from the larger end (8). Filtration may be
used to separate the suspension instead of centrifuging (7).
The wet PVC cake from the centrifuge is dropped to a dryer. Dry-
ing techniques used include spray drying, flash-rotary drying,
rotary drying, and two-stage flash drying. The polymer particle
size governs the choice of drying techniques (7). The polymer is
dried to 0.25 wt percent to 0.4 wt percent moisture content. The
maximum allowable product temperature is 55°C, because degradation
of the polymer occurs above 65°C (7,8) .
The time required to dry the batch of polymer in the blend tank
ranges from 5 hr to 8 hr. The exit end of the dryer is con-
stricted to raise the air velocity high enough to entrain dry PVC
particles. A cyclone separator removes the coarse particles
(99.93%) and fines (99.4%). Fabric filters are provided to clean
the exit air. Solid PVC recovered from the cyclone and baghouses
is sized by screens and oversize particles are recycled (7).
Bulk Polymer Handling--
Dry polymer is screened to separate oversize particles. Screened
PVC particles are then pneumatically conveyed to storage bins or
silos. The product can either be shipped, bagged, or sent to
the fabricating plant (8).
Recycle Purification—
Recovered monomer is accumulated in the recycle surge tank and
continuously fed to the purification section. The purified
monomer is recycled to the monomer plant.
18
-------
Emulsion Polymerization
In the United States, emulsion polymerization is carried out as
a batch process involving polymerization of vinyl chloride in an
emulsion system (19). Emulsifiers disperse monomer droplets in
water; polymerization proceeds in the aqueous phase surrounding
the monomer droplets (11).
Vinyl chloride, water, emulsifying agent and initiators make up
a typical recipe for emulsion polymerization (20). Many soaps
and surfactants are used as emulsifiers. Natural and synthetic
colloidal protective agents such as cellulose derivatives and
polyvinyl alcohol are also used (11).
Relatively large amounts of emulsifiers are utilized, usually in
pairs, where one agent is soluble in monomer and the other in
water (11) . Such system stability, combined with strong agita-
tion, prevents coalescence of polymer particles, resulting in
smaller particles than are obtained in the suspension process
(11). A study was made of the relationship between the chemical
nature of the emulsifier and the rate of polymerization; the
properties of the resultant polymer indicated that the effect of
emulsifiers on the polymerization reaction decreases as the
molecular weight of the emulsifier increases (20).
Since the emulsion polymerization reaction proceeds in the
aqueous phase surrounding the monomer droplets, initiators are
mostly water soluble (8). Important initiators include persul-
fates, hydrogen peroxide, and various oxidation/reduction systems
such as chlorate-bisulfite combinations.
The emulsifier and initiator (and possibly a buffer) are dissol-
ved in cold deionized water. Air is excluded from the system as
the water solution is added to the reactor (20). Measured
amounts of vinyl chloride are added and agitated to form a rela-
tively stable emulsion (20). The reaction starts when the emul-
sion is heated. Temperature control is important; it is achieved
by circulating cold water or brine in the reactor jacket (20).
The polymerization is terminated at 90% to 95% monomer conver-
sion. Polymerization rates decrease rapidly at higher conver-
sion, and pressure decreases to signify completion of the
reaction (20) . Unreastecl vinyl chloride is recycled after
purification (20).
(19) Odian, G. Principles of Polymerization. McGraw-Hill Book
Company, New York, New York, 1970. pp. 279-298.
(20) Albright, L. F. Vinyl Chloride Polymerization by Emulsion,
Bulk and Solution Processes. Chemical Engineering, 74(14):
145-152, 1967.
19
-------
Emulsion polymerization differs from suspension polymerization in
its drying operations. Emulsion resins, because of their smaller
polymer particle size, are spray dried (4).
Polyvinyl chloride resins produced by emulsion polymerization
retain about 2% to 5% of the emulsifier (20); hence, resins tend
to be hazy and have low water absorptivity (20). Emulsion resins
cost more than suspension resins, but they are used when liquid
form compounds are needed, as in organosols and plastisols (20).
Organisols and plastisols are used in dip coatings, slush mold-
ings, rotational moldings and foam applications (20).
Figure 5 gives a simplified block flow diagram and Figure 6 a de-
tailed flowsheet for the manufacture of polyvinyl chloride resins
by the emulsion process.
Bulk Polymerization
Bulk polymerization, a relatively new process introduced by
Produits Chimiques Pechiney-Saint-Gobain, is used to produce 6%
of the PVC in the United States.
In the batch process, vinyl chloride is polymerized without the
addition of other liquids (20). The process produces no major
by-products, uses negligible amounts of initiators, and yields a
pure product without drying. No solvents, emulsifiers or sus-
pending agents are needed (21).
Resins manufactured by bulk polymerization are used for molding,
extrusion and surfacing applications (21). Bulk resins resemble
suspension resins in appearance and are homogeneous with regard
to shape and size of the beads and porosity (22). They have good
heat stability, improved fusion properties, high purity, and
unrivaled clearness (22).
In the two-step bulk polymerization process, two reactors operate
batchwise and in series (20). In the first reactor, or prepoly-
merizer, 7% to 12% of the vinyl chloride is polymerized at tem-
peratures of 40°C to 70°C (23). High-speed agitation forms
particles of uniform size and promotes good heat transfer with
the cooled walls of the prepolymerizer (20).
(21) Krause, A. Mass Polymerization for PVC Resins. Chemical
Engineering, 72 (26) : 72-74 , 1968.
(22) Thomas, J. C. New Improved Bulk PVC Process. Hydrocarbon
Processing, 47 (11) : 192-196, 1968.
(23) Herbert, T. and S. Nagy. System Analysis of Air Pollutant
Emissions from the Chemical Plastics Industry. EPA-650/2-
74-106 PB 239 880. Environmental Research Center. U.S.
Environmental Protection Agency, Research Triangle Park,
North Carolina, October 1974. 281 pp.
20
-------
O W
•H W
4->
-------
iv I
CO
;|% "fa.
i-ig t-'8fc
t;;'fo §,4$
^.rj)oi .^,,^
^^tij '"\iv'
ffp'GZ ' .'[.{i.
0
e
^ -*"
a:
UJ
o
' rr
u_
CxL
1
a:
O
22
-------
0)
T3
•H
n
o
rH
X!
•H
a
c
0
•H
4J
U '
3 tn
-d W
o ^
w 2
0,0
(1)
0)
(U
tn
•H
23
-------
PVC particles are first observed at approximately 4% to 8% con-
version of vinyl chloride. The exact conditions at which precip-
itation starts depend on various operating parameters including
temperature (19).
The PVC precipitate solvates five to six parts of VCM per part of
PVC polymer. At 15% to 20% conversion, the liquid phase of vinyl
chloride essentially disappears (20). Temperature control up to
10% conversion is provided by heat transfer from vinyl chloride
liquid through the reactor walls (20). It has been reported that
10% of the vinyl chloride has polymerized after three hours, the
time depending upon the temperature and amount of initiator (4).
The suspension of PVC in VCM liquid is then transferred to the
large reactor, or polymerizer. The polymerizer is stirred with
ribbon blenders consisting of two or three ribbons wound on whorls
of different diameters (4) that turn in opposite directions.
As the reaction mixture changes from slurry, to stocky solid, to
dry particles, it is important to prevent undesired agglomeration
of PVC particles ands to control temperature. The speed of agita-
tion is reduced as polymerization proceeds. The second stage
requires 10 hr to 15 hr for completion of the reaction (4).
Figure 7 is a block flow diagram and Figure 8 a detailed flow-
sheet of the manufacture of PVC by the bulk process.
INITIATOR
VINYL CHLORIDE
VINYL CHLORIDE
VINYL CHLORIDE
• PVC DUST
POLYVINYITCHLORIDE
AND INITIATOR
PVC DUST
Figure 7. Block flow diagram for production of polyvinyl
chloride by the bulk processes.
Solution Polymerization
In the United States, only one company uses solution polymeriza-
tion for the manufacture of polyvinyl chloride. Solution polym-
erization is the only continuous process for producing PVC, and
most resins produced are copolymers of polyvinyl chloride (75% to
90%) and polyvinyl acetate (10% to 25%) (4).
-------
•0
•H
A
U
o
c
o
•H
4J
O •
3 W
•a w
o a)
n u
a o
o
4-1
-P
(U
OJ
oo
tn
•H
CM
25
-------
Solution polymerization of PVC is not a true solution polymeri-
zation, since the polymer precipitates. It is sometimes called
precipitation polymerization (20).
Polymerization is carried out at 40°C to 55°C with an initiator in
a liquid medium for 12 hr to 18 hr (22). Vinyl chloride and its
comonomers are soluble in the solvent, but the polymer is not
(24). Suitable solvents used for vinyl resins are aliphatic
alcohols, aliphatic hydrocarbons, aromatic hydrocarbons, alipha-
tic ketones, aliphatic esters, and chlorinated hydrocarbons (24,
25). The character of the resulting resin depends on the solvent
used (7).
As the reaction proceeds, the polymer appears as a powder sus-
pended in the solvent. As the polymer precipitates from solu-
tion, autoacceleration occurs because of monomer occluded in the
precipitates (11). The resin is removed by circulating the
slurry through a filter press into a settling tank. The filter
cake is dried by flash evaporation, and recovered monomer and
solvent are recycled (4).
Solution polymerization of vinyl chloride proceeds as in bulk
polymerization. With a granular precipitate and rate accelera-
tion from the start of the reaction9, solution polymerization
obviates many of the disadvantages of the bulk process (19).
Temperature control is easier and, because of a decreased viscos-
ity, stirring is efficient. Problem areas associated with solu-
tion polymerization include proper solvent selection to avoid
chain transfer, and careful removal of solvent from product to
avoid contamination.
PVC resins obtained by solution polymerization are relatively
pure, because emulsifiers or suspending agents are not required
(20). Another advantage of solution polymerization is simplified
product recovery, because water is not used in the process (20).
A block flow diagram and a flowsheet for the solution polymeriza-
tion process are shown in Figures 9 and 10, respectively.
Information obtained from EPA files concerning private communi
cation between E. M. Smith, Continental Oil Company, Ponca City,
Oklahoma, and D. Goodwin, EPA, Research Triangle Park, North
Carolina, 12 July 1974.
(24) Douglas, S. D. Process for Producing Vinyl Resins. U.S.
Patent 2,075,429 (to Union Carbide), March 30, 1937.
(25) Reid, E. W. Process for Producing Vinyl Resins. U.S. Patent
2,064,565 (to Union Carbide), December 15, 1936.
(26) Reid, E. W. Vinyl Resins. U.S. Patent 1,935,577 (to Union
Carbide), November 14, 1933.
26
-------
>m
"Z. d
> K
>- O
£5
t
£§£
sSl ^.
o o o
o
^f ^
n:
1
Q
< O
il
S 2
Q UJ
:= o
cc oo
0
S
a
=3 UJ
j* t za:
t
O
3= <
< 0
ni £
<:
s
- A
§53 T
UJ 0 1
o
1 1
oo
1 «
i t^
P 5
Z LL.
0
o
1
o
M
IX.
>:
0
Q_
o
uT UJ
a
Is
31 =g
II 11
4-1
O
O
•H
-P
U •
3 w
O 0)
M O
04 O
(1) QJ
-P C
O
O '-P
4-) 3
rH
e o
0)
•H -P
H 0)
m T3
•H
A; M
o o
O rH
rH .£
CQ O
0)
en
-H
CM
27
-------
0)
'd
•H
^
o
iH
JC
u
O en
-<-\ en
-P 0)
o o
3 o
T3 S-l
O DJ
5-1
a c
0
H -H
O -P
0)
-P O
<\) W
0)
£
Cfl
5 4->
o
•H >^
fe XI
0)
S-I
3
tn
-H
28
-------
MATERIALS FLOW
A simplified material balance for the suspension polymerization
process for a representative plant with a production rate of
68 x 103 metric tons/yr is shown in Figure 11.
GEOGRAPHICAL DISTRIBUTION
Table 7 provides the locations and capacities of PVC manufac-
turing plants. Based on capacities of PVC manufacturers, the
average plant capacity was calculated to be 68 x 10 3 metric
tons/yr. Figure 12 shows the locations of the U.S. facilities.
TABLE 7. LOCATIONS AND CAPACITIES OF POLYVINYL CHLORIDE
MANUFACTURING PLANTS
Producing company
Air Products, Inc.
American Chemical Corp.
Borden, Inc.
Continental Oil Co.
Diamond Shamrock Corp.
Ethyl Corp.
Firestone Tire Co.
General Tire Co.
B. F. Goodrich Co.
Goodyear Tire Co .
Great American Chemical
Corp.
Keysor-Century Corp.
Occidental Petroleum
Olin Corp.
Pantasote Co.
Robintech, Inc.
Stauffer Chemical Co.
Tenneco Chemicals , Inc.
Union Carbide Corp.
Uniroyal , Inc .
AVERAGE CAPACITY
Plant location
Calvert City, KY
Pensacola, FL
Long Beach, CA
Illiopolis, IL
Leominster, HA
Springfield, MA
Aberdeen, MS
Oklahoma City, OK
Delaware City, DE
Deer Park, TX
Baton Rouge , LA
Perryville, MD
Pottstown, PA
Ashtabula, OH
Avon Lake, OH
Henry , IL
Long Beach, CA
Louisville, KY
Pedricktown, NJ
Niagara Falls, NY
Plaquemine, LA
Fitchburg, MA
Saugus , CA
Burlington, NJ
Hicksville, NY
Assonet, MA
Passaic, NJ
Point Pleasant, WV
Painesville, OH
Delaware City, DE
Burlington, NJ
Flemington, NJ
South Charleston, WV
Texas City, TX
Painesville, OH
Capacity,
10 3 metric tons/yr
61.23
34.02
68.04
63.50
81.65
31.75
117.93
99.79
45.36
122.47
81.65
104.33
122.47
56.70
117.93
99.79
52.16
65.77
63.50
45.36
49.63
18.14
15.88
76.20
6.80
68.04
27.22
43.09
113.40
79.38
74.84
31.75
72.57
136.08
48.99
68.44
aPantasote'g Point Pleasant, West Virginia plant is 50% owned by
General Tire Company.
29
-------
s-M
•iigp
7: — £! K
.t
l
>=b£
^1
o.
s *
Ill
Rsjs;3
13
SSis
'
«»
Isll
B i
I is
o
i—
a)
n
3
tr«
-H
T 7 T
S s
2 •
30
-------
Figure 12. Polyvinyl chloride plant locations.
31
-------
SECTION 4
EMISSIONS
LOCATIONS AND DESCRIPTIONS
A typical polyvinyl chloride plant has 12 sources of vinyl chlo-
ride monomer and particulate polyvinyl chloride emissions. The
emission points are identified in Table 8. Information in this
section was obtained from EPA files of data reported by operating
companies during 1974. The data were obtained only partially
through actual field sampling; most came through material balance
and engineering estimates.
TABLE 8. POINTS OF EMISSION AT A REPRESENTATIVE
POLYVINYL CHLORIDE PLANT
Identification point
in Figure 3 Description
A Reactor safety relief valves
B Reactor entry purge
C Stripper jets
D Monomer recovery condenser vent
E Slurry blend tank vents
F Centrifuge vent
G Dryer discharge
H Storage silos
I Bulk loading
J Bagger vent
K Storage tanks
L Fugitive emissions
Reactor Safety Relief Valves
Reactor safety relief valves are an intermittent source of emis-
sions9 with reported discharges occurring from 3 to 20 times per
Information obtained from EPA files concerning private communi-
cation between E. M. Smith, Continental Oil Company, Ponca City,
Oklahoma, and D. Goodwin, EPA, Research Triangle Park, North
Carolina, 12 July 1974.
32
-------
year (27). Power failure, operator error and equipment failure
will cause venting by pressure relief valves and rupture discs
(27). Manual venting is also used to reduce pressure, thereby
preventing greater losses.3
Atmospheric emissions from venting of the polymerizer consist of
VCM, PVC, or some combination thereof, depending upon the stage
of VCM conversion.9 Emissions from this source range between
0.6 g/kg and 2.2 g/kg and discharge at a height of 15 ma (4).
Reactor Entry Purge
The reactor is purged with air (27) after PVC slurry has been
transferred to the slurry blend tanka and after most of the VCM
has been removed by vacuum or by water displacement to the mono-
mer recovery system (4). Until recently it was necessary to open
reactors after each batch to remove PVC buildup on the reactor
walls (4). Plants are now reducing VCM exposure as much as 90%
by reducing work forces through automation, and by improved
cleaning techniques (28). Emissions range from 0.8 g/kg to
5.0 g/kg and are emitted at a height of 15 m (4).
Stripper Jets
Emissions from stripper jets are intermittent. After polymer-
ization, unreacted VCM is removed by venting the reactor to a
recovery system. Some vinyl chloride remains in the water or
trapped in the PVC particles. This residual vinyl chloride is
stripped in the reactor or in a second vessel called the stripper
(28) where stripping is carried out in vacuum and/or by contact
with steam (27).
Information obtained from EPA files concerning private communi-
cation between H. C. Holbrook, B. F. Goodrich Chemical Company,
Cleveland, Ohio, and D. Goodwin, EPA, Research Triangle Park,
North Carolina, 17 June 1974.
(27) Evans, L. B., and L. L. Beck. The Vinyl Chloride and PVC
Industry Emissions and Control Techniques. Draft copy of
report. U.S. Environmental Protection Agency. Emission
Standards and Engineering Division, Industrial Studies
Branch, July 23, 1974.
(28) PVC Plants are Ready to Pass First Test. Chemical Week,
116(19):49-50, 1975.
33
-------
Stripping operations are important; control (28) of emissions
from the slurry blend tank, the centrifuge, the dryer and the
bulk storage silos is dependent upon effective removal of resid-
ual vinyl chloride trapped in the PVC granule (4).
Atmospheric emissions consist of inerts and vinyl chloride (27).
The emissions range between 0.5 g/kg and 12.3 g/kg (4).
Monomer Recovery Condenser Vent
The monomer recovery condenser vent is an intermittent emission
source by nature.3 Recycled vinyl chloride is treated in a two-
stage compression system where the monomer is dewatered and puri-
fied (12, 15). Inert gas, water vapor, and VCM are discharged to
the atmosphere. The emissions from this source range between
3.1 g/kg and 15.0 g/kg and exhaust at a height of 18 m (4).
Slurry Blend Tank Vents
Slurry blend tank vents are a continuous source of emissions
resulting from the continuous purging of the vapor space in the
atmospheric pressure slurry blend tanks with fresh air.a In
the slurry tank, vinyl chloride is released from the PVC granules
where it was trapped (27).
The emission rate for this source ranges from 2.5 g/kg to
5.7 g/kg (27).
Centrifuge Vent
Emissions from the centrifuge vent are continuous. The centri-
fuge separates the slurry into wet solids containing 75% to 77%
PVC in water (12). Some of the vinyl chloride trapped within the
PVC granules is released (27). The atmospheric emissions from
the centrifuge vent consist of water vapor, air, VCM and PVC
resin.
The rate of emission varies between 0.04 g/kg and 1.3 g/kg (4),
exhausted at a height of 17 m.a
Dryer Discharge
Dryer discharges are a continuous emission source. The wet
polymer contains about 20% to 25% moisture (15) and is dried
using air at temperatures ranging from 60°C to 66°C (15). Atmos-
pheric emissions from the dryer exhaust consist of air, water
vapor, vinyl chloride and polyvinyl chloride.9
Information obtained from EPA files concerning private communi-
cation between H. C. Holbrook, B. F. Goodrich Chemical Company-,
Cleveland, Ohio, and D. Goodwin, EPA, Research Triangle Park,
North Carolina, 17 June 1974.
34
-------
Emissions from this source range between 2.0 g/kg and 25.6 g/kg
(27) .
Storage Silos
Storage silos are a continuous source of emissions. The polymer
stored in the silos is frequently mixed by passing dry air
through the silos. This prevents moisture condensation and the
buildup of explosive concentrations of VCM (27).
Atmospheric emissions consist of air, VCM, and PVC. The rate of
emission varies between 0.2 g/kg and 1.7 g/kg (27) emitted at a
height of 21 m.
Bulk Loading
A continuous discharge of air and particulate PVC takes place
during loading operations. The emission rate is estimated to be
0.4 g/kg.
Bagger Vent
The vent from the bagging operations is a continuous source of
particulate emissions. This stream, containing PVC and air, is
ducted to a baghouse for recovery of the solid product.9 The
emission rate for this source is estimated to be 0.2 g/kg.
Storage Tanks
The emissions from storage tank vents have been estimated to be
0.6 g/kg.
Fugitive Emissions
Leaks occurring from pressure relief valves, pumps, compressors,
agitator seals, loading and unloading of monomer, valve stems,
flanges, unrepaired purging equipment and samples for laboratory
analysis are defined as fugitive emissions (4). There may be as
many as 600 points of fugitive emissions at a typical PVC plant
(4) .
The vinyl chloride emission rate from this source ranges from 6.2
6.2 g/kg to 17.5 g/kg (27).
EMISSION FACTORS
VCM emission factors for the four processes used to produce poly-
vinyl chloride are given in Table 9.
Information obtained from EPA files concerning private communi-
cation between H. C. Holbrook, B. F. Goodrich Chemical Company,
Cleveland, Ohio, and D. Goodwin, EPA, Research Triangle Park,
North Carolina, 17 June 1974.
35
-------
TABLE 9. VINYL CHLORIDE EMISSION FACTORS FOR
POLYVINYL CHLORIDE PROCESSES (4)
VCM emission factors,
Process type
Suspension process
Emulsion process
Bulk process
Solution process
35.5
60.1
24.2
17.8
Table 10 lists the contributions of PVC-producing plants in the
United States to national point source emissions (1) of criteria
pollutants. Table 11 lists contributions from PVC production
to state emissions of criteria pollutants. Since production
data by state was not readily available, plant capacities were
used. Table 12 lists TLV®, atmospheric reactivity and health
effects of each species emitted from a PVC manufacturing plant.
TABLE 10. POLYVINYL CHLORIDE INDUSTRY CONTRIBUTIONS TO NATIONAL
STATIONARY SOURCE EMISSIONS OF CRITERIA POLLUTANTS
Emissions from the
m . . .. n PVC industry
Total national A
emissions (1), 103 metric Percent of national
Material emitted 106 metric tons/yr tons/yr emissions
Hydrocarbons 25 85 0.34
(vinyl chloride, phenol
stabilizer, ethyl chlo-
ride, butadiene, vinyl-
idene chloride, acetal-
dehyde, acetylene, pro-
pylene, vinylacetylene,
ethylene, ethylene di-
chloride, chloroprene,
vinyl bromide)
Particulate 18 18 0.10
(polyvinyl chloride)
Sulfur oxides 30 5.5 x 10~4 2 x 10~6
36
-------
TABLE 11. POLYVINYL CHLORIDE INDUSTRY CONTRIBUTIONS
TO STATE EMISSIONS OF CRITERIA POLLUTANTS
State
New Jersey
Massachusetts
Ohio
California
West Virginia
Illinois
Texas
Delaware
Louisiana
New York
Kentucky
Florida
Maryland
Oklahoma
Mississippi
Pennsylvania
Material emitted
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
Vinyl chloride
PVC
State
emissions (i) ,
10 3 metric tons/yr
819.5
151.8
440.5
96.16
1,153
1,766
2,161
1,006
116.2
213.7
1,826
1,143
2,219
549.4
63.89
36.81
1,920
380.6
1,262
160
326.3
546.2
619.9
226.5
295.9
494.9
341.4
93.6
196
168.4
891.8
1,811
PVC emissions
metric tons/yr
9,450
2,370
5,845
855
10,620
3,030
2,145
1,505
8,030
380
7,090
1,775
6,980
1,820
2,770
2,870
6,915
715
2,055
150
7,380
1,295
1,075
490
1,860
510
3,660
65
4,185
885
4,150
365
Percent
1.15
1.56
1.33
0.09
0.92
0.17
0.10
0.15
6.91
0.18
0.39
0.16
0.31
0.33
4.34
0.78
0.36
0.19
0.16
0.09
2.26
0.24
0.17
0.22
0.63
0.10
1.07
0.07
2.14
0.52
0.47
0.02
Vinyl chloride hydrocarbon emission; PVC particulate emission.
37
-------
TABLE 12. CHARACTERISTICS OF EMISSIONS FROM A
REPRESENTATIVE POLYVINYL CHLORIDE PLANT
Compound
TLV, (29)
g/m3 Atmospheric reactivity
Health effects
Vinyl chloride 0.0026
Polyvinyl chloride 0.1090
Ethylene 1.25
Propylene 1.88
Acetylene 1.16
Butadiene 2.20
Ethylene dichloride 0.20
Vinylacetylene 0.0480
Vinyl bromide 1.10
Vinylidene chloride 0.004
Acetaldehyde 0.18
Ethyl chloride 2.60
Chloroprene 0.09
Hydrogen chloride 0.007
Sulfur oxides 0.013
Phenol (stabilizer) 0.02
Contributes to photo-
chemical smog
Stable
Contributes to photo-
chemical smog
Contributes to photo-
chemical smog
Contributes to photo-
chemical smog
Contributes to photo-
chemical smog
Contributes to photo-
chemical smog
Contributes to photo-
chemical smog
Contributes to photo-
chemical smog
Contributes to photo-
chemical smog
Contributes to photo-
chemical smog
Contributes to photo-
chemical smog
Contributes to photo-
chemical smog
Contributes to chlo-
ride formation
Contributes to sulfate
formation
Contributes to photo-
chemical smog
A recognized
carcinogen
Suspected
carcinogen
Moderate irritant
Moderate
asphyxiant
Moderate
asphyxiant
Moderate irritant
and asphyxiant
Sharp irritant
and asphyxiant
Simple irritant
and asphyxiant
Moderate irritant
and asphyxiant
Details unknown
Sharp irritant
and asphyxiant
Simple irritant
and moderate
asphyxiant
Sharp irritant
and asphyxiant
Sharp irritant
and asphyxiant
Sharp irritant
and asphyxiant
Sharp irritant
and asphyxiant
(29) TLVs® Threshold Limit Values for Chemical Substances in
Workroom Air Adopted by ACGIH for 1976. American Conference
of Governmental Industrail Hygienists. Cincinnati, Ohio,
1976. 94 pp.
38
-------
DEFINITION OF REPRESENTATIVE SOURCE
A representative plant for polyvinyl chloride manufacture was
defined in order to determine source severity. Factors consid-
ered include polymerization process, plant capacity, polymer
produced, vinyl chloride emission factor, PVC emission factor,
emission height for VCM emissions, emission height for PVC emis-
sions, maximum ground level concentration, and source severity.
Table 13 gives a summary of data used to determine a represen-
tative plant. Table 14 summarizes the data for a representative
plant .
Table 15 gives the representative PVC manufacturing plant emis-
sion factors for vinyl chloride and polyvinyl chloride. Table 16
lists the emissions from the 12 major sources at a representative
PVC plant. One of these major sources, fugitive emissions, is
further broken down into seven categories.
ENVIRONMENTAL EFFECTS
Maximum Ground Level Concentration
The maximum ground level concentration, Xmax' ror materials
emitted from each of 12 major points of emission for a polyvinyl
chloride plant were estimated by a Gaussian plume dispersion
method. Xmax' -*-n 9/m3 ' was calculated using the equation:
=
where Q = emission rate, g/s
h = effective emission height, m
e = 2.72
IT = 3.14
u = average wind speed = 4.47 m/s
Time-Averaged Maximum Ground Level Concentration
Xmax -"-s tne maximum ground level concentration averaged over a
given period of time. The averaging time is 24 hr for noncriteria
pollutants (chemical substances). For criteria pollutants,
averaging times are the same as those used in the primary ambient
air quality standard; (i.e., 3 hr for hydrocarbons_and 24 hr for
particulates) . The relationship between xm=^ and
-.
expressed as:
/ t ° • 1 7
= v
Amax xmax I t I
where t = "instantaneous" averaging time = 3 min
t - averaging time for ambient air quality standard
39
-------
EH
<
Q
o
>H
I
w
Q
H
PS
o
U
w
i-q
OQ
-JJ -H 1C PJ O O »
**- -r f Ji CCCCC
r- PI » o r- o M
3 s sssas
T, +.*.*. 0.* ~- "^ -- B CCCCC
TJ im_ «- oijc ^--^ -^ j<^ E ccccc
M 01--. *1D *rH • U10 >B OOOOO
p- r--r-ij "icu^u tor~
m M n ID
J J ^"-
o
-i
a- i. tout
S
o" " ^°'
own
B B -
X ffl
I 3 ||
| | ii
ff S S3
1 3 S5
sis
fr
1 i 1
u 8 1
1 F '>
i j
M S
01 >i U
S 2 «
o o o a
r- 1^1* PI
u» nr- MJ
wo r-
•* O CO
33 >
•1J Ul 1^ W
SS S
4J 3 3 >
U UO
a a a:
0 tin Corp ABsonet, Hassachusets 307.
Pantasote Co. Passaic, New Jersey 1,646.
Point Pleasant, Hest Virginia 21.
o oo
a a o
*"> "i
* X
3
X XX
X xx
S Sf^
" »-«-
U H Id
1 II
S ii
° o o
« +* +j
! IS
III
g «
3 a
n
« 8
!!
s's
ot m
^
y»
=,
^ M
-H»
a x
"•"•
1
U U
I!
a,
8
?
2
3
|
i
o
o
o
*% ^
t-- o
. .
as
j
-
H
„
>0
*
h
1 i
!'
a r « .
B
O O
• o
£ 2 £
!Q[M, rf
M "^< o c " 6 * -g Q
15- "3 ^I|«1«S«
p, -i •« «o se O O *
•3!" ^S ™«°J.dSi
Q.IJ41 UO - • « an
io§ £^ SO^Q"-^,^
K w.^ ing ~'oci«°
-------
TABLE 14. SUMMARY OF CRITERIA USED TO DEFINE A
REPRESENTATIVE POLYVINYL CHLORIDE PLANT
Criteria
Representative plant9
Process
Polymer
Density, persons/km2
Capacity, metric tons/yr
VCM emission factor, g/kg
PVC emission factor, g/kg
VCM emission height, m
PVC emission height, m
VCM ground level cone., g/m3
Suspension process
Homopolymer
310 ± 45%
68,000 ± 17%
36 ± 23%
7.5 ± 42%
16 ± 13%
21 ± 16%
0.018 ± 29%
Numbers indicate the mean values; the 95% confidence
limit is given as percent of the mean value.
TABLE 15. EMISSION FACTORS FOR A REPRESENTATIVE
POLYVINYL CHLORIDE PLANT
Material emitted
Emission factor,
g/kg
Vinyl chloride
Polyvinyl chloride
Stabilizer (phenol)
Ethyl chloride
Sulfur oxides
Butadiene
Hydrogen chloride
Vinylidene chloride
Acetaldehyde
Acetylene
Propylene
Vinylacetylene
Ethylene
Ethylene dichloride
Chloroprene
Vinyl bromide
35.5 ± 8.24a
7.5 ± 3.18a
204 x lO-5
92 x 10"5
23 x ID'5
21 x ID"5
21 x ID"5
9 x 10"5
7 x lO-5
7 x lO-5
7 x lO"5
5 x lO-5
5 x lO-5
4 x ID"5
<4 x lO-5
2 x lO-5
Values indicate the mean values
for the emission factor; the 95%
confidence limit is given in g/kg.
41
-------
m ^
Cn
CO Ai
PH —
O Cn
EH — '
u
<; EH
fa 2
^*H
^•"•^ J_^
O OH
H
CO W
CO >
H M
S EH
W i<
EH
2 W
H CO
PH W
HI) PH
EH OH
u w
fa
^H .
S O
fa
fa"
Q IS
H 0
PH H
O CO
J CO
E M
u s
W
rJ
>H fa
rS O
H
> EH
h-H M
0 0
p 1 p j
VD
r— {
W
OH
EH
0)
Cn in
IS .X
^H C
O 10
4-> 4J
CO
01 4-1
Ol C
O> 0)
IS >
cq
C
^ -H
rH TJ
CQ O
^
0)
IS O
H rH
0 -H
•p in
to
01
Oi
IH rH
01 IS
^1 ^
Q in
•H
•a
i
•H 10
M 11 4-1
4J O> C
c 3 o)
0) m >
u
w 13 ... in
^ £« *^« -P
3 01 JJ3 c
rH -H ™ 0)
M
>H >1 Cl
6 01 C 4J
O > Q) C
So 13 o>
0 C >
S ill O
>H U
r-l
ft to
O< 4J
•H 0)
jj 'r~l
to
S-l
o >, El Cn
O 4J In
18 C 3
(U (U ft
^
O >i >H fl)
4J 4J 01 >
O 01 -H H
is IH rH us
0) IS 01 >
B! to VH
•a 13
c ai
9 4J
o i
u
(O
VD
• ra
O O
ts
m
o o
rO
•q-
o d
ro m
^"^
d d
^ CTi
• •
rH M1
S
(O —
rH rH
5.
• <0
^* o
-
o
• (0
in o
IO
00
rH O
00
• IO
(S 0
ro
rH
• ro
cs o
•S
•r-l
0) O
•O rH
•H fi
tH U
0
rH rH
•H
c1 1-1
•H O
> e.
4J
IS
10
«
g
3
U
rd
jj
c S
-rH S
c p.
a> -H
Q, ^
o^ cr*
d)
o>
c
•H
H
m
«.
Oi
>( IH
4J
IH
m
§
rH
in
0)
Oi
§
iH
IH
in
I
g
0)
4->
C
•H
B
^i
Q
»w
^1
IH
O
to
to
^
ft
g
Q
O
111
>
rH
g
to
c
0
4->
2
0>
o
k.
to
IS
01
to
IS
(1)
4-1
to
T3
S
§
Q}
U
(3
^i
n
o
4->
IS
O
•a
H
O
IS
4J
•H
Cn
IS
US
Is
to
"fcj
01
rH
i^j
*
o
4J
in
4J
0)
•r-l
g
•H
JJ
O
01
a
to
.,_{
to
•H
in
>1
IS
i
in
rH
IS
0)
to
10
111
IS
0
to
•rH
•0
D1
c
Cn -H
a TJ
•H
•O
IS
,5
IS
o
H
§
1
1 0
1
1
1
1
*— o
^^
ro
m
rH
to
•H
in
01 0
o
IH
3
0
in
to
fi
0
•H
to
to
•H
g O
0)
^
•H
4-1
•H
1
rH O
rH
IS
0
rH
IS
O
EH 0
1
1
1
1
1
1 O
1
1
01
•jj
01 O
13 rH
•H X!
rH U
0
iH rH
0 C
rH "^
C H
•5 £ .
•
IH
0)
C
Oi
u
a
01
01
IH
01
OJ
4-)
IS
u
•H
13
C
-H
ro
13
§
^^
rH
to
J->
ft
•H
M
O
in
0)
ft
3
to
rH
IS
O
•H
g
•Z
S
• 53
H
01
g
O
to
01
ft
i?
in
3
c
•H
§
H
u-i
13
•H
0)
O
rH
C
O
•H
4J
(d
£3
fi
o
IH
C
•H
g
O
IH
IH
01
jJ
g
*H
4J
in
01
Cn
c
•rH
)H
0)
01
G
•H
cf
01
in
S
IS
0
•H
•a
"iS
5
4J
•53
rj
10
»H
ft
W
•
ro
01
rH
1
f*
•H
01
O
fl
01
13
C
a
10
01
IH
O
O
Cn
•S
C
rH
01
O
o
10
(U
4J
O
•g
Q
M-)
0)
01
in
42
-------
Using the plant capacity and emission factor data shown in
Tables 13 and 15, the emission rates shown in Table 17 were cal-
culated. Using the emission heights for VCM emissions shown in
Table 14 and Equations 1 and 2, values of Xmax' Xhigh and XT
were calculated as shown in Table 17. The mean values plus W
the 95% confidence limits for each value are shown at the bottom
of the table.
Table 18 gives the time-averaged maximum ground level concentra-
tion by compound for a representative PVC plant. Table 19 gives
the time-averaged maximum ground level concentration for each
major point of emission.
Source Severity
To obtain a quantitative measure of the hazard potential of poly-
vinyl chloride manufacture, a source severity, S, is defined as
the ratio of time-averaged maximum ground level concentration to
F, the hazard exposure level for that pollutant; i.e., S = x /F
F is the primary ambient air quality standard for criteria max
pollutants3 and is a corrected threshold limit value (i.e., TLV •
8/24 • 1/100) for noncriteria pollutants.
Table 20 lists the source severity factor for each material
emitted. Table 21 lists severity factors for each point of emis-
sion for the materials emitted.v Table 22 contains the data
(obtained from Table 18) used to prepare Figure 13, which shows
the emission rate and the source severity for individual plants
as a function of the cumulative percent of PVC plants. In
Figure 13, TLV values for VCM were used to calculate source
severity. Figure 14, which was developed from 1974 data,
shows changes in plant emission rate and in plant source severity
as a function of the cumulative percent of PVC plants. The
primary ambient air quality standard for hydrocarbons was used
to calculate the source severity of vinyl chloride. Figure 14
indicates that in 1974 the time-averaged maximum ground level
concentrations of vinyl chloride emissions from all vinyl chlo-
ride plants exceeded the primary ambient air quality standard
for hydrocarbons.
EPA conducted an ambient monitoring program around two plants
which manufacture polyvinyl chloride. The facilities chosen were
the Continental Oil Company's plant at Aberdeen, Mississippi, and
the B. F. Goodrich Company's plant at Louisville, Kentucky (30).
(30) EPA Programs of Monitoring Vinyl Chloride in Ambient Air.
Environmental Protection Agency Office of Air Quality
Planning and Standards. Research Triangle Park, North
Carolina, February 2, 1976. 14 pp.
Criteria pollutants are those emissions for which ambient air
quality standards have been established.
43
-------
TABLE 17. POLYVINYL CHLORIDE - SUMMARY OF PLANT DATA - II
Plant location
Calvert City, KY
Pensacola, FL
Long Beach, CA
Illipolis, IL
Leominster, MA
Springfield, MA
Aberdeen, MS
Oklahoma City, OK
Delaware City, DE
Deer Park, TX
Baton Rouge, LA
Perryville, MD
Pottstown, PA
Ashtabula, OH
Avon Lake, OH
Henry, IL
Long Beach, CA
Louisville, KY
Pedricktown, NJ
Niagara Falls, NY
Plaquemine, LA
Fitchburg, MA
Saugus, CA
Burlington, NJ
Hicksville, NY
Assonet, MA
Passaic, NJ
Point Pleasant, WV
Painesville, OH
Delaware City, DE
Burlington, NJ
Flemington, NJ
South Charleston, WV
Texas City, TX
Painesville, OH
Mean values
95% confidence limit
Q for VCM,
g/s
72.5
34.1
10.2
51.7
66.7
35. 7C
39. 2C
11.6
51. 0C
13.2
91. 8C
59.0
132.0
17.9
171.0
173.0
48.0
161.0
73.1
57.5
122.0
6.3
9.8
31.9
7.6C
76. 5C
42.1
113.0
126.0
37.0
110.0
42.0
141.0
208.0
21.6
70.43
±18.86
*max
-------
TABLE 18. TIME-AVERAGED MAXIMUM GROUND LEVEL CONCENTRATION BY
COMPOUND FOR A REPRESENTATIVE POLYVINYL CHLORIDE PLANT
Material emitted
Vinyl chloride
Polyvinyl chloride
Stabilizer (phenol)
Ethyl chloride
Sulfur oxides
Butadiene
Hydrogen chloride
Vinylidene chloride
Acetaldehyde
Acetylene
Propylene
Vinylacetylene
Ethylene
Ethylene dichloride
Chloroprene
Vinyl bromide
Emission
height , m
15.5
20.95
15
15
15
15
15
15
15
15
15
15
15
15
15
15
± 1.96
± 3.41
.5
.5
.5
.5
.5
.5
.5
.5
.5
.5
.5
.5
.5
.5
Q ' ^max '
g/s g/m
76
16
4.4
2.0
5.0
4.6
4.6
2.0
1.5
1.5
1.5
1.1
1.1
8.7
8.7
4.3
.9
.2
x
X
X
X
X
X
X
X
X
X
X
X
X
X
lO-3
10-3
10-"
10-"
10-"
IO-5
10-"
10-"
10-"
10-"
10-"
10-5
ID'5
10-5
1.7
1.9
9.7
4.4
1.1
9.9
9.9
4.3
3.3
3.3
3.3
2.4
2.4
1.9
1.9
9.5
x
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
IO-2
lO-3
10- 7
ID'7
io-7
io-8
io-8
io-8
10"8
10~8
io-8
io-8
io-8
10~8
io-8
io-9
g/m3
8.4
6.8
4.8
2.2
5.4
5.0
5.0
2.1
1.7
1.7
1.7
1.2
1.2
9.5
9.5
4.7
x
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
lO-3
10-"
io-7
io-7
10"8
10- 8
io-8
10" 8
10- 8
io-8
10- 8
io-8
10"8
ID'9
10" 9
io-9
TABLE 19. TIME-AVERAGED MAXIMUM GROUND LEVEL CONCENTRATION
FOR EMISSIONS FROM A REPRESENTATIVE POLYVINYL
CHLORIDE PLANT BY POINT OF EMISSION
capacity = 68,440 metric tons/yr
emission height = 15.49 m for VCM and 20.95 m for PVC
Material emitted
Vinyl chloride
Point of emission
Q,
g/s
Polyvinyl chloride
g/m3
g/m3
Q,
g/s
Xmax'
g/m*
Reactor safety
relief valve 4.6 1.0 x 10~3
Reactor entry
purge 6.1 1.3 x 10~3
Stripper jets 3.9 8.5 x 10~3
Monomer recovery
condenser vents 10.8 2.4 x 10"3
Slurry blend tank
vents 9.5
Centrifuge vents 2.8
Dryer discharge 3.0
Storage silos 1.5
Bulk loading 0
Bagger vents 0
Storage tanks 1.3
Fugitive
emissions 33.2
.1 x IO-3
.2 x IO-1*
.6 x 10-"
2.
6
6.
3.3 x
0
0
2.8 x 10-"
10
-"
4.5 x 10
-"
7 x 10-"
3 x 10-"
1.2 x
x
x
1.1
3.1
3.3 x
1.7 x
0
0
1.4 x 10-"
10-"
10-"
10-"
7.3 x 10-3 3.6 x 10-3
0
2.8
10.6
1.5
0.9
0.4
0
0
3.3 x 10-"
1.3 x IO-3
1.8 x 10-"
1.1 x 10-"
4.8 x IO-5
0
0
2 x 10-"
4 x 10-"
3 x IO-5
8 x IO-5
7 x IO-5
0
45
-------
TABLE 20. SOURCE SEVERITY BY COMPOUND FOR A
REPRESENTATIVE POLYVINYL CHLORIDE PLANT
Material emitted
Vinyl chloride
Polyvinyl chloride
Stabilizer (phenol)
Ethyl chloride
Sulfur oxides
Butadiene
Hydrogen chloride
Vinylidene chloride
Acetaldehyde
Acety lene
Propylene
Vinylacetylene
E thy lene
Ethylene dichloride
Chloroprene
Vinyl bromide
TLV,
g/m3
0.
0.
0.
2.
0.
2.
0.
0.
0.
1.
1.
0.
1.
0.
0.
1.
0026
1090
02
60
013
20
007
004
18
16
88
048
25
20
09
10
Y
Amax'
g/m3
8.4
6.8
4.8
2.2
5.4
5.0
5,0
2.1
1.7
1.7
1.7
1.2
1.2
9.5
9.5
4.7
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
lO-3
io-4
10- 7
io-7
IO-8
io-8
10~8
10- 8
io-8
10- 8
10- 8
10- 8
io-8
IO-9
io-9
io-9
F,a
g/m3
8.7
3.6
6.7
8.7
4.3
7.3
2.3
1.3
6.0
3.9
6.3
1.6
4.2
6.7
3.0
3.7
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
10~6
io-4
10-5
lO-3
10-5
ID'3
IO-5
IO-5
io-4
IO-3
ID'3
io-4
10-3
io-4
io-4
io-3
Source ,
severity
970
7.2
2.5
1.3
6.8
2.1
1.6
2.8
4.3
2.6
7.4
2.8
1.4
3.2
1.3
1
X
X
X
X
X
X
X
X
X
X
X
X
X
X
.9
lO-3
10~5
lO-3
10~6
10- 3
lO-3
10- 5
10- 6
io-6
lO-5
10- 6
10- 5
10- 5
io-6
F = hazard factor = TLV -_8/24 • 1/100.
b . Xmax
Source severity = TLV . 8/24 . 1/100
TABLE 21. SOURCE SEVERITY FOR A REPRESENTATIVE POLYVINYL
CHLORIDE PLANT BY POINT OF EMISSION
Point of emission
Source severity for
polyvinyl chloride
emissions
Source severity for
vinyl chloride
emissions
Reactor safety
relief valve
Reactor entry purge
Stripper jets
Monomer recovery
condenser vent
Slurry blend tank
vents
Centrifuge vents
Dryer discharge
Storage silos
Bulk loading
Bagger vent
Storage tanks
Fugitive emissions
0
0
0
0
0
0.
1.
0.
0.
0.
0
0
32
22
17
10
05
51.5
76.7
49.1
136.0
121.0
35.5
37.7
19.2
0
0
16.3
417.0
46
-------
TABLE 22. INPUT DATA
Vinyl chloride
emission rate,
10 3 metric tons/yr
0.10
0.20
0.24
0.31
0.32
0.42
0.56
0.68
1.08
1.13
1.16
1.33
1.33
1.51
1.61
1.63
1.81
1.86
2.10
2.29
2.31
2.41
2.89
3.49
3.58
3.66
3.86
3.98
4.15
4.18
4.45
5.09
5.39
5.46
6.56
Cumulative
percent of
PVC plants
2.9
5.7
8.6
11.4
14.3
17.1
20.0
22.9
25.7
28.6
31.4
34.3
37.1
40.0
42.9
45.7
48.6
51.4
54>3
57.1
60.0
62.9
65.7
68.6
71.4
74.3
77.1
80.0
82.9
85.7
88.6
91.4
94.3
97.1
100.0
VCM source
severity9
8
10
123
128
131
225
304
360
424
439
448
492
513
519
552
838
961
1,020
1,110
1,150
1,150
1,200
1,360
1,370
1,380
1,380
1,450
1,460
1,650
1,650
1,720
2,090
3,260
3,300
3,310
Hydrocarbon
source
severity
0.43
0.54
6.7
6.9
7.1
12
16
20
23
24
24
27
28
28
30
45
52
55
60
62
62
65
74
74
75
75
79
79
89
89
93
113
177
179
199
aq _ xmax
VCM TLV • 8/24 • 1/100
hydrocarbon
where
AAQS
, , ,
hydrocarbon
= PrimarY ambient quality standard
for hydrocarbons.
47
-------
-a 10-°
c 9.0
3 8.0
« 7.0
J= 6.0
0)
e 5.0
2 4.0
S 3.0
2 2.0
GO
00
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
SOURCE SEVERITY, S, FOR VINYL CHLORIDE EMISSIONS
where S = ^ max _
8/24 -1/100
EMISSION RATE FOR VINYL CHLORIDE EMISSIONS
J L
J I L
I I L
10
20 30 40 50 60 70 80
CUMULATIVE PERCENT
90
9,000
8,000
7,000
6,000
5,000
4,000
3,000
2,000
1,000
900
800
700 >_
600 t
500 |
400 £
300
200
100
90
80
70
60
50
40
30
20
o
a:
O
GO
10
98
Figure 13.
Cumulative percent of PVC plants having
an emission rate and a source severity
less than or equal to indicated value.
48
-------
LTl
I/)
10.0
9.0
c 7.
-2 6.0
£ 5.0
1 4.0
°2 3.0
LJ-T
< 2.0
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
SOURCE SEVERITY, S, FOR VINYL CHLORIDE EMISSIONS
where S
max
AAQS hydrocarbon
EMISSION RATE FOR VINYL CHLORIDE EMISSIONS
I
I I I I
I
I
900
800
700
600
500
400
300
200
100
90
80
50
40
30
20
10
9
8
7
6
5
4
3
di
o
CO
10 20 30 40 50 60 70 80
CUMULATIVE PERCENT
CUMULATIVE PERCENT OF PLANTS
90
98
Figure 14.
Cumulative percent of PVC plants having an
emission rate and a source severity less
than or equal to indicated value.
49
-------
Integrated samplers were set up at various locations around the
two plants. Each week two samples were collected at each sam-
pling site and returned to the laboratory for gas chromatographic
analysis (30).
At the Aberdeen, Mississippi plant, 15 24-hour integrated samplers
were set up; 530 samples were collected between November 6, 1974,
and March 27, 1975. At the Louisville, Kentucky plant, 17
24-hour integrated samplers were set up between November 6, 1974,
and May 15, 1975. When it became clear that prevailing meteoro-
logical patterns were not as predicted, 38 24-hour integrated
samplers were set up and a second set of samples collected bet-
ween May 15, 1975, and June 12, 1975. A total of 1,155 samples
were collected at the Louisville plant (30).
Information on wind direction and wind speed was recorded at each
plant during the sampling effort. Also the plants recorded
unusual occurrences which would be expected to affect their emis-
sion rates, and submitted these records to EPA (30).
Data obtained from the sampling program are summarized in
Table 23. A histogram of the cumulative percent of samples
having ground level concentration less than or equal to the
indicated value is shown in Figure 15.
The ground level concentrations obtained from the sampling
results are compared with the time averaged maximum ground level
concentrations (from Table 17) calculated using the Gaussian
plume dispersion methodology. These results are summarized in
Table 24 (30) .
~X calculated for the 35 polyvinyl chloride manufacturing
facilities using dispersion modeling ranged between 688 yg/m3
and 28,800 yg/m3; 30 plants have
976 yg/m3 < XTTiav <23,430 yg/m3,
ITlclX
3 plants have
and 2 plants have
X > 23,430 yg/m3;
in 9.x
Xmax < 976
50
-------
^•— •»
o
ro
' •*
CO
E~t
-]
D
CO
W
py*!
CJ
^
H
rH1
OH
^
l<
W
pT j
0
«
s]
D
CO
ro
CM
W
m
S
j>,
c
id
ft
g^
O >,
CJ X
o
rH 3
(fl Jj
U C
•H HI
g r**!
0)
U 01
O -H
•H >
>H 01
13 -H
O O
U i-5
ft)
pa
C -H
ft ft
O in
U 01
rH 01
•rH 01
° g
ontinental
Aberdeen,
CJ
c
" K
o a
r-
01
01 t
§ £
Pn (
>•
C
01
•rH 4->
•P C
rH C
3 SH
E o)
3 ft
u
in
IH C
0 C
-P -i-
C(Tl
1U
0) >
U SH
SH 01
01 o:
01
C
IH O
0 -H
4-)
SH id
A -H
g (1)
3 01
2"§
>
•H 4-1
4-1 C
frt fli
r-H O
3 SH
E 0)
3 ft
U
01
IH C
o o
-rH
c id
01 >
0 SH
SH 0>
0) W
«,.§
01
C
IH O
O -H
4->
in (0
0) >
•9 *H
5 01
3 U]
23 •§
*.
01
H C
J 0
> -H
J 4_>ro
H Id g
SH X
3 -P 01
5 C 3.
3 01
3 CJ
^ s
n 0
CJ
rHinrHmcNr-rHoor-oo o
m-tfcocriroinroM'r-oo o "O
rH rH rH (d
id
c
-rH
rH
ro O
g Ju
X (d
tji U
r- ( uo LO o^ ro LO ^j1 r^- o^ ^ O o ~^
intTlfOrHrOCNr-rHCNCNO O ^ -P
rH rH f— 1 rH rH (H O Is*- M
i— 1 O^ O
K
y
id
0)
rH
tn
rHtNrHrHrHrH rH -rH
SH
rH EH
X
O
id
0)
M
(NO^IMI- (N^O S
O ^ O CO CO vD ^D l/l ro O O O *•
i — I (N ro ro ^ LO {\Q r^ CO o^ o o CQ •
i— t rH Q ^f
x. n4^
< rH
W 0)
°e ~ §
^ -P r?
01 -P
rMOOM'COM'rHcOHlDCMkD O 3- idCTi
rH rH rH O rO O
*. p4 ,q
ro O
IN
W '
•H
C -P
o) us
01 C.
"? C
4J -rH
01 O
ft C
in 0
•rH - '
~ id 2
tn CJ W
> -H H
3 rt]
01 43 B W
3 1 O
rH ^T U *
OOOOCJi idrs] VH
OOOOOOOCT* > — r— 10)
o g 13 C, SH
OOOOOOOOOOO 3g O3
4J4J4J4J4J4J4J4J4-)4->- 1-1 gP OlEn
rH < -H O SH
OrOVOrHrHrHrHrHrHrHAI EH XMH 01-
rHcNinoooo O id (irX
rH CN ro in EH S id
<
EH
Q
*f^l
O
H
ro
UJ
CO
H
S
W
Q
W
-i
D
U
1 — 1
U
Q
<
Q
W
PH
I — ^
CO
^i
§
t
5 >
U Ai
CJ
H P
id -P
o c
•rj 0]
01
O 01
rH
O -H
•H >
SH 01
O 3
O O
U J
•
h
•
0]
>1
C -H
frt n.
L Oil Comp;
Mississip]
td -
-P d
d QJ
a> o)
c -a
-P 0)
*y
O rt]
CJ
T)
O)
SH
en
01
2
•a
0)
3
03
id
H
CU
^
i a
g
•H -a
X £
s c
SH
U
rrj
OJ
4-1
id
rH
O
r-J
rt)
O
Maximum
•ound level
M
en
13
0)
Iti
rH
3
CJ
rH
10
O
>
0) -
O -H
SH SH
3 OJ
O >
C/l 01
01
fj
o
•H
4->
Id ro
iJ-S
C 0
0) 3
O
C
o
u
01 4-1
O -H
SH SH
3 01
O >
W 01
in
S
£2-
X
01 4-1
U -H
SH SH
3 01
Ul 0)
01
c"
mcentratic
yg/m3
8
01 4->
O -H
SH SH
3 01
O >
CO 01
01
•.ro
X g
e^
X 3.
i-H
rH
ID
r-
cn
o
o
ro
ro
O
O
00
CN
o
ff\
ID
IN
O
ro
ro
CN
(N
01
^*
O
00
CM
«*
51
-------
10
10J
o
•z.
o
=3
O
102
10
TIME AVERAGED MAXIUM GROUND LEVEL CONCENTRATION
CALCULATED FOR A REPRESENTATIVE PLANT USING
DISPERSION MODELING
SAMPLES FROM
CONTINENTAL 01 LCD
SAMPLES FROM
B.F. GOODRICH CHEMICAL CO,
GROUND LEVEL CONCENTRATION
CORRESPONDING TO A SOURCE
SEVERITY OF "ONE "FOR VINYL
CHLORIDE! BASED ON TLV =
0.0026 g/m3)
12 5 10 20 30 40 50 60 70 80
CUMULATIVE PERCENT
90 95 98 99
Figure 15.
Cumulative percent of samples having
ground level concentrations less than
or equal to indicated value.
52
-------
From Figure 15 it is seen that 30% of the samples for the Aberdeen
plant and 46% of the samples for the Louisville plant have ground
level concentrations less than 8.7 yg/m3 which corresponds to a
source severity of "one" based on a TLV for vinyl chloride equal
to 0.0026 g/m3.
The EPA standard would result in a 95% or greater reduction in
vinyl chloride emissions from polyvinyl chloride plants. Table 25
summarizes ambient ground level concentrations of controlled vinyl
chloride emissions.
TABLE 25. CONTROLLED VINYL CHLORIDE EMISSIONS
95% Controlled vinyl chloride emissions
Calculated
Measured
Plant
y
Amax'
yg/m
Source
severity
/Basis: \ Xm,
\TLV = 0.0026 g/m3/ yg
Source
severity
ax' /Basis: \
/mj VTLV = 0.0026 g/mv
Continental Oil
Company
B. F. Goodrich
Chemical Co.
Representative
plant
214
1,435
420
24.6
165
48.3
1,171.5 134.7
48.8 5.6
NOTE: Blanks indicate data not applicable.
Data used in the calculation of X^x was reported to EPA in early
1^74. The sampling data was obtained in mid-1975. The glaring
difference between measured and calculated ground level concen-
tration for the B. F. Goodrich Chemical Company plant can be
attributed in pa-rt to installation of control equipment during
the crucial period when industry was involved in reducing emis-
sions to comply with the temporary emission standard. Also at
the time of sampling, the plant was operating at approximately
50 percent capacity.
All data used in this report were obtained prior to 1975. The
polyvinyl chloride industry has undergone considerable modifica-
tion since that time, and these results may not be currently
representative .
A_f ffJr-t6. d Population
The population affected by emissions from a typical PVC plant
was obtained as described below.
53
-------
The area exposed to the time-averaged ground level concentration,
X, for which x/F > 1_!_^ was obtained by determining the area with-
in the isopleth for x (3-'< ) • The number of persons within the
exposed area was then calculated, using the population density
for a plant whose production and emission criteria closely match
those used to define a representative plant. Table 26 shows the
affected population for a representative PVC plant.
TABLE 26. AFFECTED AREA AND AFFECTED POPULATION
Parameter for Particulates Hydrocarbon
representative plant (polyvinyl chloride) (vinyl chloride)
Population density,
persons/km ?-
Height of emission, m
TLV, g/m3
Q, g/s
F (primary ambient air
quality standard) ,
g/m3
Xmax , g/m3
Maximum source severity
Affected area, A, km2
Affected population, P,
persons
313
21
0.11
16
3.6 x 10-4
210
2.4
11
3,400
313
15
0.003
77
8.7 x 10~6
150
900
2,800
870,000
GROWTH FACTOR
In 1974, 2.2 x 106 metric tons of polyvinyl chloride were pro-
duced in the United States; 1979 production is expected to total
2.62 x 106 metric tons. Vinyl chloride emissions from polyvinyl
chloride manufacturing facilities have been estimated to be
35.5 g/kg of product as seen in Table 15. Therefore, total vinyl
chloride emissions from polyvinyl chloride plants in 1974 are
determined to have been 7.81 x 104 metric tons. The EPA requires
vinyl chloride plants to reduce vinyl chloride emissions by 95%.
Assuming proportional emission growth from 1974 through 1979,
total 1979 vinyl chloride emissions to the atmosphere would be
estimated as 9.26 x 10^ metric tons. If all PVC manufacturing
facilities reduce vinyl chloride emissions by 95%, total vinyl
(31) Turner, D. B. Workbook of Atmospheric Dispersion Estimates.
Public Health Service Publication No. 999-AP-26, U.S. Depart-
ment of Health, Education, and Welfare, Cincinnati, Ohio,
May 1970. 84 pp.
54
-------
chloride emissions to the atmosphere would be 4,640 metric tons,
Therefore, vinyl chloride emissions from the polyvinyl chloride
industry are expected to decrease by 94% from 1974-1979.
Emissions in 1979 _ 4,640 metric tons _
Emissions in 1974 78,100 metric tons
55
-------
SECTION 5
CONTROL TECHNOLOGY
Emissions from the manufacture of polyvinyl chloride consist of
hydrocarbons and particulates. Table 27 shows the control equip-
ment used for each emission point at a PVC plant.
CONTROL TECHNOLOGY FOR HYDROCARBONS
Activated Carbon Adsorption
Adsorption is a highly selective, three-step phenomenon in which
molecules become attached to the surface of a solid. A given
adsorbent or adsorbing agent will adsorb only certain types of
materials, or adsorbates (32). First, the adsorbent comes in
contact with the stream containing the adsorbate, and adsorption
occurs. Next, the unadsorbed portion of the stream is separated
from the adsorbent. Finally, removal of the adsorbate regenerates
the adsorbent (33).
Activated carbon, the most suitable adsorbent for removing organic
vapors (32), adsorbs 95% to 98% of all organic vapor from air at
ambient temperature regardless of variations in concentration and
humidity conditions (34).
When a gas stream is passed over an activated carbon bed, the
carbon adsorbs the organic vapor or gas, and the purified stream
passes through. Initially, adsorption is rapid and complete (32).
(32) Air Pollution Engineering Manual, Second Edition. J. A.
Danielson, ed. Publication No. AP-40, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina,
May 1973. 987 pp.
(33) Hughes, T. W., D. A. Horn, C. W. Sandy and R. W. Serth.
Source Assessment: Prioritization of Air Pollution from
Industrial Surface Coating Operations. EPA-650/2-75-019-a,
U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, February 1975. 303 pp.
(34) Hydrocarbon Pollutant Systems Study. Volume I. Stationary
Sources, Effects and Control. APTD-1499 (PB 219 073), U.S.
Environmental Protection Agency, Research Triangle Park,
North Carolina, October 20, 1972. 379 pp.
56
-------
X— •*
in
^— ^
W
j"y^
D
EH
U
r
X
O
I-H
O
fa
P>H
o
o
t_^l
0
2
ffi
U
w
EH
0
fVJ
EH
2
O
U
r~.
CN
W
i "I
CQ
rf*1
EH
fO
•d
01
CO
3
4-1
C
a
•H
3
s
1 — 1
0
C
0
0
01
10 rH
rH 0
33
•H G
4-1 O
^4 CJ
id
A,
01 C rH
> 0 0
•H -H M
4-1 CO 4-1
•H CO C
C" -H O
360
Cn 0)
O >i CD O
U 4J H 4->
ra c 3 G
01 0) d O
fti CJ
H M
CD 0)
'O 4-1 0>
rH 10 tJV
W "O Q!
* c
en
c
•H
Ol
a
•H
^
4J
W
O
-H
4J
2
0)
C
•H
O
c
H
g
0
4J
^
CU
cr
^4
c1 c? <=>
X
X x
o o o o o o
CD 00 03 O3 03 O3
O O 0 O O 0
4-1 4-1 4-1 4J 4J 4J
O O O O O O
in in in in in in
CTi 0^ CT* CT*
cyi o°i o> CTI
0 O O O
4-14-14-1 4-1
o o o o
in in in in
o o o
O O O
4J 4J 4-1
O O O
" ^ ^
O O O O O
o o o o o
4J 4J 4-1 4J 4J
O O O O O
in tn in in m
CJl CTt CTt i -P CO
V4 C 4J CU CU -H
>iO) cucu GCn >co
4-!>>iCO>> OIS4U) CO
CUrH^J 4JQ t3 ^nJOCT* ^J'H
lwra4J 3-H id> cu
rlCUHOIOlrtd) M-I-OOIO CU>
O-H OcT'ftcy'CJ >i,V-H tPrH H CP-H
EQjECMC^inra cuiTj4-i
•H OO WI04-1 01 MX C"M-H
>4CU94-IG>iOrHcPOCT'
OJ 01 4JO i-{ CUH4J3l54J3
a; a! ens co uQcoiatnwfc
CO
•H
tr
o
•H
o
^
0
01
4J
rH
O
M
4J
G
O
o
4->
id
4J
CO
Ot
4J
10
O
-H
•O
•H
X
CU
01
>
fc
T3
01
CO
3
en
O
rH
O
c
X!
O
4)
4-1
0
£ §
0 0
0 G
^
O 4J
>U O
C
O CO
C -H
CU
•H C
14-1 0)
lt-1 -H
CU O
•H
O M-l •
rj 01 CU
±> rH
§5 ^
O-H o
•H
O 3 Qj
0) nj
C G 4J
10 -H o
M O c
ft
ti r. <"
(fl C JJ
ij 0 ra
4J -H o
CO -H
0) co TI
4J -H c
o S
•H m
•H 4J §
CD 4J m
n id
a>
'i "8 "
z s o
«3 2
57
-------
As the carbon bed approaches its vapor-retaining capacity or
breakpoint, traces of vapor appear in the exit air. If gas flow
is continued, organic material is adsorbed, but at a decreasing
rate.
The adsorption of a mixture of adsorbable organic vapors in air
is not uniform. The more easily adsorbed components are those
which have higher boiling points. When air containing a mixture
of organic vapors passes over activated carbon, the vapors are
equally adsorbed at the start. However, as the amount of the
higher boiling constituents retained in the carbon bed increases,
the more volatile vapors revaporize. After the breakpoint is
reached, the exit vapor consists largely of the more volatile
material (32). At this stage, the higher boiling component has
displaced the lower boiling compound; the procedure is repeated
for each additional component.
The quantity of an organic vapor adsorbed by activated carbon
is a function of the nature of the vapor, the adsorbent type and
temperature, and the vapor concentration. Removal of gaseous
vapors by physical adsorption is practical for gases having mole-
cular weights over 45 (35).
Each type of activated carbon has its own adsorbent properties
for a given vapor. The quantity of vapor adsorbed for a par-
ticular vapor concentration and temperature is best determined
experimentally. The quantity of vapor adsorbed increases when
the vapor concentration increases and the adsorbent temperature
decreases (33).
After reaching its breakpoint, the adsorbent is regenerated by
heating the solids until the adsorbate is released. A carrier gas
removes the vapors. Low pressure saturated steam is used as the
heat source for activated carbon and also acts as the carrier gas.
When high boiling compounds have reduced the carbon's adsorbing
capacity to the point where complete regeneration is necessary,
they may have to be removed with superheated steam at 350°C (33).
Steam requirements for regeneration are a function of external
heat losses and the nature of the organic material. The amount of
steam adsorbed per kilogram of adsorbate as a function of elapsed
time passes through a minimum; the carbon should be regenerated
for this length of time to permit the minimum use of steam (35).
After regeneration, the carbon is hot and water-saturated.
Organic-free air blown through the carbon bed evaporates the water
and thus cools and dries the carbon. If high temperature steam
has been used, other means of cooling the carbon are required.
(35) Chemical Engineers Handbook, Fifth Edition. J. H. Perry
and C. H. Chilton, eds. McGraw-Hill Book Company, New York,
New York, 1973.
58
-------
Fixed bed adsorbers arrayed in two or more parallel bed arrange-
ments are used to remove organic vapors from air. These are
batch-type arrangements, where a bed is used until breakthrough
occurs and is then regenerated. The simplest adsorber design of
this type is a two-bed system where one carbon bed is being
regenerated as the other is adsorbing organic vapors. In a
three-bed arrangement, a greater quantity of material can be
adsorbed per unit of carbon because the effluent passes through
two beds in series while the third bed is being regenerated.
This permits the activated carbon to be used after breakthrough
since the second bed in the series removes organic vapors in the
exit gas from the first bed. When the first bed is saturated, it
is removed from the stream for regeneration; the bed which was
used to remove the final traces of organic vapors from the
effluent then becomes the new first bed and the bed which has
been regenerated becomes the new second bed (33).
Heat is released in the adsorption process, which causes the tem-
perature of the adsorbent to increase. If the concentration of
organic vapors is not high, as in the case of room ventilators,
the temperature rise is typically 10°C (32, 33).
The pressure drop through a carbon bed is a function of the gas
velocity, bed depth, and carbon particle size. Activated carbon
manufacturers supply empirical correlations for pressure drop in
terms of these quantities. These correlations usually include
pressure drop resulting from directional change of the gas stream
at the inlet and outlet (33).
Activated carbon systems are not economical when large volumes of
gases containing low concentrations of organic compounds have to
be treated. This technique for emission control has only
recently been tried on high concentration streams in the PVC
industry (4) .
Carbon adsorption units are used to collect vinyl chloride from
the monomer recovery system vent, centrifugal vent, and the
slurry tank vent (4). The installation of carbon adsorption
units at dryers and bulk storage silos depends on the life of the
carbon bed.
Solvent Absorption
Absorption is a process for removing one or more soluble compon-
ents from a gas mixture by dissolving them in a solvent.
Absorption equipment is designed to insure maximum contact bet-
ween the gas and the liquid solvent, allowing interphase diffu-
sion between the materials (32). Absorption rate is affected by
factors such as the solubility of gas in the particular solvent
and the degree of chemical reaction; however, the most important
factor is the solvent surface exposed (32).
59
-------
Equipment that disperses liquid solvent into the gas stream con-
sists of packed towers, spray towers and venturi gas absorbers.
Absorbers that use gas dispersion include tray towers and vessels
with sparging equipment (32) .
A packed tower is filled with one of many packing materials
designed to expose a large surface area. When the solvent wets
the packing surface, a large area of liquid film for contacting
the solute gas is attained (32).
In spray-type absorbers, interphase contact is achieved by dis-
persing the liquid in a spray and passing the gas through it.
In venturi gas absorbers, interphase contact is provided through
the different velocities of the gas and liquid, and by turbulence
created in the venturi throat.
Tray towers induce contact by means of a number of trays arranged
so that the gas is dispersed through a layer of solvent on each
tray.
Solvents used to control vinyl chloride emissions include ethy-
lene dichloride, acetone, "Carnea oil" (a petroleum based hydro-
carbon) , and trichloroethane (4).
The absorbed material, regenerated from the solvent by applying
heat and vacuum, is then transferred to the monomer recovery
system.
Solvent absorption units are installed to collect VCM from the
monomer recovery condenser vent, storage area and slurry blend
tank (4) .
Refrigeration
Organic compounds can be removed from an air stream by condensa-
tion. A vapor will condense when, at a given temperature, the
partial pressure of the compound is equal to or greater than its
vapor pressure. Similarly, if the temperature of a gaseous mix-
ture is reduced to the saturation temperature (i.e., the tempera-
ture at which the vapor pressure equals the partial pressure of
one of the constituents), the material will condense. Thus,
either increasing the system pressure or lowering the temperature
can cause condensation (36).
(36) Control Techniques for Hydrocarbons and Organic Solvent
Emissions from Stationary Sources. Publication No. AP-68,
U.S. Department of Health, Education, and Welfare, Washington,
D.C., March 1970. pp. 3-1 through 3-26.
60
-------
The equilibrium partial pressure limits the control of organic
emissions by condensation. As condensation occurs, the partial
pressure of material remaining in the gas decreases rapidly,
preventing complete condensation. For example, at 0°C and atmo-
sphere pressure, a gas stream saturated with toluene would still
contain about 8,000 ppm of that gas. Thus a condenser is not
very successful in reducing VCM emissions and must usually be
followed by a secondary air pollution control device such as a
carbon adsorber or solvent absorber (36).
Surface condensers are used in the PVC industry. VC vapor con-
denses on the outside surface of tubes while the cooling medium
(water, freon, propane or propylene) flows within (4).
In the PVC plant, refrigeration is used on condenser vents, the
slurry blend tank vent, and the centrifuge vent.
Incineration
On combustion, vinyl chloride forms hydrogen chloride, carbon
dioxide and water:
2 CH2=CHC1 + 5 02 -> 4 C02 + 2 H20 + 2 HC1 (3)
Little free chlorine should be formed since there is sufficient
hydrogen in the VC molecule to combine with the chlorine and form
hydrogen chloride. Chlorine inhibits oxidation reactions.
Higher temperatures and longer residence times are needed for the
complete destruction of pollutants if chlorine is present, even
at low concentrations.
Table 27 identifies the emission points controlled by the types
of incineration equipment described below.
Flares--
Flares are used for the combustion of low concentration vinyl
chloride streams and intermittent emissions caused by plant upset
(4). They are not an ideal form of control, because vinyl chlo-
ride oxidation produces hydrogen chloride which is itself a pol-
lutant. Another disadvantage — dilute gas streams cannot sup-
port combustion. Fuel must be added to achieve combustion, and
the heat produced is wasted.
Direct-Flame Afterburners—
Direct-flame afterburners depend upon flame contact and high tem-
peratures to burn combustible materials (37). The combustible
(37) Rolkes, R. W., R. D. Hawthorne, C. R. Garbett, E. R. Slater,
T. T. Phillips and G. D. Towel1. Afterburner Systems Study.
EPA-R2-72-062 (PB 212 560), U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, August 1972.
512 pp.
61
-------
materials may be gases, vapors, or entrained particulates which
contribute opacity, odor, irritants, photochemical reactivity,
and toxicity to the effluent. A direct-flame afterburner con-
sists of a refractory-lined chamber, one or more burner tempera-
ture indicator-controllers, safety equipment, and, sometimes,
heat recovery equipment (37).
The afterburner chamber consists of a mixing section and a combus-
tion section. The mixing section provides contact between the
contaminated gases and the burner flame. Good mixing is provided
by high velocity flow which creates turbulence. The combustion
section is designed to provide a retention time of 0.3 s to 0.5 s
for completion of the combustion process. Afterburner discharge
temperatures range from 540°C to 800°C, depending on the air
pollution problem. Higher temperatures result in higher after-
burner efficiencies (37).
The gas burners used in afterburners are of the nozzle-mixing,
premixing, multiport, or mixing plate type. Burner placement
varies depending on burner type and on the design objective of
providing intimate contact of the contaminated air with the
burner flames. When all the contaminated air passes through the
burner, maximum afterburner efficiency is obtained (37).
Nozzle-mixing and premixing burners are arranged to fire tangen-
tially into a cylindrical afterburner. Several burners or
nozzles are required to ensure complete flame coverage, and
additional burners or nozzles may be arranged to fire along the
length of the burner. Air for fuel combustion is taken from the
outside air or from the contaminated air stream which is intro-
duced tangentially or along the major axis of the cylinder (37).
Multiport burners are installed across a section of the after-
burner separate from the main chamber. Although all air for
combustion is taken from the contaminated air stream, multiport
burners are not capable of handling all of the contaminated air
stream. Contaminated air in excess of that used for fuel com-
bustion must be passed around the burner and mixed with the
burner flames in a restricted and baffled area (37).
Mixing plate burners were developed for afterburner applications,
and are placed across the inlet section of the afterburner. The
contaminated air and the burner flames are mixed by profile
plates installed around the burner between the burner and after-
burner walls. The high velocities (1 m/s) provided by the burner
and profile plate design ensure mixing of the burner flames and
the contaminated air not flowing through the burner. The con-
taminated air stream provides air for fuel combustion (37).
62
-------
The efficiency of an afterburner is a function of retention time,
operating temperatures, flame contact, and gas velocity. No
quantitative mathematical relationship between these variables
exists because the kinetics of the combustion process are complex
and flow inside afterburners is not defined. However, for good
design, the following observations can be made with respect to
afterburner efficiency (37).
• Efficiency increases with increasing after
burner operating temperature.
• Efficiency decreases if the contaminated gases
entering the afterburner are excessively preheated.
• Efficiency increases with increasing contact
between the contaminated gases and the burner
flame.
• Efficiency increases with increasing retention
time for retention times less than 1 second.
• Efficiency is a function of the afterburner
design and the inlet concentration of organic
materials.
• Ninety percent afterburner efficiency is
difficult to reach below a 700°C operating
temperature if the generation of carbon
monoxide in the afterburner is included.
Afterburners are designed to recover heat present in the com-
bustion gases. When large volumes of dilute gases have to be
burned, supplemental fuel is needed for combustion. Heat
exchange can be used to reduce the amount of fuel required. Hot
exit gases can also be used to generate steam in a boiler. The
combustion gases leaving the heat exchanger or boiler rray be
scrubbed with water or caustic solution to remove hydrogen
chloride.
Catalytic Afterburners—
A catalytic afterburner contains a preheat burner section, a
chamber containing a catalyst, temperature indicators and con-
trollers, safety equipment, and heat recovery equipment. The
catalyst in such an afterburner promotes combustion by increasing
the rate of the oxidation reactions without itself appearing to
change chemically (33).
The contaminated air entering a catalytic afterburner is heated
to the temperature necessary for carrying out the catalytic com-
bustion. The preheat zone temperature, in the range of 340°C to
600°C, varies with the combustion and type of contaminants.
Because of thermal incineration in the preheat zone, the preheat
burner can contribute to the efficiency of a catalytic after-
burner (33).
63
-------
Catalysts used for catalytic afterburners may be platinum-family
metals supported on metal or matrix elements made of ceramic
honeycombs. Catalyst supports should have high geometric surface
area, low pressure drop, structural integrity and durability, and
should permit uniform distribution of the flow of the waste
stream through the catalyst. Catalysts can be poisoned by phos-
phorus, bismuth, arsenic, antimony, mercury, lead, zinc, and tin,
which are thought to form alloys with the metal catalyst. Cata-
lysts are deactivated by materials which form coatings on them,
such as particulate material, resins, and carbon formed during
organic material breakdown. High temperatures will also deacti-
vate catalysts. Because the combustion reaction is exothermic,
the catalyst bed temperature is above the inlet temperature. The
temperature increase depends on the concentration of organic mate-
rial burned and the heat of combustion of that material. Com-
pensation for decreased catalyst activity can be made by: 1) ini-
tial overdesign in specifying the quantity of catalyst required
to attain required performance; 2) increasing preheat temperature
as chemical activity decreases; 3) regenerating the catalyst; and
4) replacing the catalyst (33).
The quantity of catalyst required for 85% to 95% conversion of
hydrocarbons ranges from 0.5 m3 to 2 m3 of catalyst per 1,000 m3/
min of waste stream. Although the catalyst temperature depends
on the hydrocarbon burned and the condition of the catalyst, the
operating temperature of catalytic afterburners ranges from 260°C
to 540°C (37).
Steam Boilers--
Gaseous streams containing vinyl chloride can be incinerated in
the fireboxes of steam boilers. Such a process requires appro-
priate instrumentation, however, because hydrogen chloride formed
by combustion causes corrosion at temperatures above 316°C or
below 204°C.
Concentrated hydrocarbon streams containing 1% to 18% vinyl chlo-
ride are used with supplemental natural gas and air to generate
steam. The boiler is modified to burn chlorinated hydrocarbons.
The exit of the boiler is scrubbed in a packed column with a
wastewater stream having a pH of 11.
Slurry Stripping
After polymerization, unreacted VCM is removed from the reactor
by venting. The vinyl chloride remaining in the water or trapped
within PVC granules amounting to approximately 0.1% (38) must be
effectively removed to control emissions from the slurry blend
tank, the centrifuge, the dryer and the bulk storage silos (4).
(38) Mantell, G. J., J. T. Barr and R. K. S. Chan. Vinyl Chlor-
ide Emission Control: Stripping VCM from PVC Resin. Chem-
ical Engineering Progress 71(9):54-62, 1975.
64
-------
The VCM remaining in the water or in the PVC granules is recov-
ered by a process known as stripping, in which heat, pressure,
and vacuum are used to drive off the volatile VCM from the
reactor contents. The monomer so obtained is compressed and con-
densed for reuse (38).
The amount of VCM left after stripping depends on particle size
and porosity, the temperature and vacuum used, and the retention
time in the stripper. Increasing the temperature, reducing the
pressure (i.e., increasing the vacuum) and increasing the dura-
tion of the stripping operation (residence) favor removal of the
vinyl chloride from the resin (4).
Steam stripping involves increasing the temperature by introduc-
ing steam into the outside jacket of the reactor or into the
vessel directly.
Countercurrent multistage column stripping is being investigated.
The rate of VCM stripping is proportional to the difference be-
tween the amount of VCM in the resin and the amount of VCM in the
water surrounding the resin. The primary advantage of counter-
current multistage column stripping is maximization of force
because the resin leaving the column contacts water containing no
vinyl chloride (4). Table 28 and Figure 16 show changes in mono-
mer content, pressure, and temperature that occur when a batch of
slurry is dropped from a reactor into the stripper. As seen from
the figure, pressure increases rapidly, in the first few minutes
followed by a slower fall to the operating pressure. The operat-
ing temperature is reached at the same time. This is defined as
TABLE 28. LOSSES OF MONOMER IN THREE STAGES OF
STRIPPING OF BATCH FROM SLURRY OF
4,540 kg (10,000 Ib) MONOMER CHARGE (38)
Stages
II III
VCM in resin, ppm:
Initial 180,000 30,000 500
Final 30,000 500 1
Amount lost:
kg
Ib
580
1,275
114
251
1.8
4
Approximate conversion is 85%.
Approximate weight ratios of monomer in vapor,
liquid, and solid phases are 1:100:1000
during stripping.
65
-------
STAGES
-207
-0
L-103
10
-310 S W
-103 H)'
PRESSURE
30 45
TIME, min
60 75
60
49
38
Figure 16. Stages of plant slurry stripping (38).
Stage I and generally results in an approximately 80% unreacted
monomer reduction from approximately 18% in the solid to about
2% to 4%. In Stage II, the residual monomer drops to about
500 ppm. In Stage III where monomer residual is below 500 ppm
and rates of VCM removal are at their lowest (38).
Product Stripping
Residual vinyl chloride retained by granular PVC can be removed
in the storage silos by sweeping inert gas through the silos.
The removal rate is low, but no major change in equipment is
needed and the resin blending requirement is met (4). VCM
recovered by the inert gas is removed by a carbon adsorber, sol-
vent absorber, or incinerator. The inert gas can then be
recycled back to the storage silos (4).
Gasholder and Water Purge System
A gasholder and reactor water purge system can reduce emissions
from reactor entry purging, from reactor safety valve discharges,
from vinyl chloride recovery condenser vents, and from fugitive
emission sources. Emissions are reduced by purging vinyl chlo-
ride from the reactor after the batch is discharged to the
gasholder. Emissions from the recovery condenser vent are
66
-------
reduced because there is less air left in the system before the
reactor is charged (4). This increases the volume of gas that
must be vented from the monomer recovery system.
The gasholder helps minimize reactor safety valve discharges. If
upset conditions are noted, the reactor can be manually dumped to
the gasholder which is sized to hold all of the VCM present in
one batch.
The gasholder also acts as a surge tank between the plant and the
vinyl chloride recovery system by holding a short-term, high
volume surge of vinyl chloride which would normally overload the
recovery system (4). The gasholder is also used to prevent fugi-
tive emissions from other sources in the plant (4).
Reactor Entry Purge Control
Each time a reactor is opened for maintenance, cleaning or
inspection, vinyl chloride emissions occur. These can be con-
trolled by reducing the number of reactor openings. The use of
high pressure water sprays inserted through a gland in the
reactor to clean its walls reduces manual cleaning requirements
to once every 12 batches.
The cleaning agent can be introduced as part of the reaction
recipe, and reactors can be redesigned to minimize scale forma-
tion. These two procedures have reportedly reduced rranual clean-
ing to once every 80 to 90 batches and in one case to one opening
per 200 batches (4) .
Heated organic solvent can be introduced into the reactor and
agitated until the solid scales of PVC which line the reactor are
broken up and dissolved. The mixture can then be distilled to
separate the solvent, VCM and PVC. The solids are reclaimed or
discarded, the monomer is recovered, and the solvent is recycled
(4). In this case, the frequency of opening is reduced from once
for each batch to once each 40 to 60 batches (4).
Control of Fugitive Emissions
Rapid detection and quick repair of a leak are necessary for
reducing fugitive emissions. Leaks may be detected by several
methods, and these are described below.
A fixed multipoint gas chromatograph, including analyzer and
recorder, may be used to sample vinyl chloride periodically at
points within the plant (4). The exact location of the leak in a
section where a high concentration has been detected is deter-
mined by a portable flame ionization-type hydrocarbon sensing
device (4).
67
-------
Another method of detecting fugitive leaks is to periodically
check each possible leak point with a portable detector (4). A
third method is to hydrostatically test piping, flanges, vessels,
manholes, and other process equipment after construction, main-
tenance, or inspection (4).
Control of pump, compressor, and agitator seals is provided by
the use of double mechanical seals between which a liquid is
maintained at a pressure greater than that which exists in the
pump, compressor or agitator. Any leakage that occurs will thus
leak into the pump, not out of it (4). All flanged pipe points
are potential leak sources for which welded connections can be
used.
Emissions resulting from sampling for laboratory analysis can be
eliminated by letting the gas to be sampled flow through the
sample flask to a lower pressure point in the process. The
sample flask is then blocked off and any vinyl chloride that
remains in the sample lines can be purged with inert gas to a
monomer recovery system or a control device (4).
Two hoses are connected to a railroad car or barge for loading or
unloading of VCM. The bottom hose transfers the liquid VCM
while the other, located at the top, maintains pressure. Mate-
rial left in the hoses may be lost to the atmosphere on discon-
nection. This can be controlled by purging the lines to a
control device with inert gas (4).
Emissions resulting from excessive pressure are controlled by
connecting the relief valve discharge to a flare or another
control device. Vinyl chloride present in equipment that is
opened for maintenance or inspection can be controlled by purging
the equipment with inert gas or displacing the contents with
water before opening (4).
CONTROL TECHNOLOGY FOR PARTICULATES
Fabric Filters
Fabric filters use a filter medium to separate particulate matter
from a gas stream. Two types of fabric filters are in use:
high energy cleaned collectors and low energy cleaned collectors
(39) .
(39) Frey, R. E. Types of Fabric Filter Installations. Journal
of the Air Pollution Control Association, 24 (12) :1148-1149,
1974.
68
-------
High Energy Collectors--
In high energy collectors, pulse jets clean the filter medium,
which is a felt fabric kept as clean as possible (39). The
principle of the pulse jet is based on the use of an air ejector
for dislodging dust from the bags. The ejector products a short
pulse of compressed air in the direction opposite to that of the
gas being filtered. The jet must accomplish three things (40):
1) stop normal filtering flow; 2) transmit a burst of air to the
filtration medium to give it a vibratory shock; and 3) create
enough pressure in the bag to assure a flow reversal from the
clean side to the dirty side of the bag.
Low Energy Collectors--
Low energy collectors use shaking or reverse air flow methods of
cleaning. The filter uses a woven cloth. However, the cloth
itself is not the true filter medium, but rather acts as a site
on which the true filter medium, dust cake, can build up (39).
Cyclones
In PVC production, cyclones are used to reduce the PVC dust emit-
ted. Centrifugal and gravitational forces to the dust particles
which are to be removed. This force is produced by directing the
gas in a circular path or causing an abrupt change in direction.
High density particles are forced against the wall of the cone
in a spinning motion. The smaller the diameter of the cone, the
faster the particles travel. Thus the particles become increas-
ingly heavy through centrifugal force as they travel downwards
in a spinning motion towards the bottom of the collector (32).
Meanwhile the carrier gas spirals downward at the outside (with
the dust particles) and upward at the inside of the cyclone,
leaving the dust at the cone bottom.
(40) Bakke, E. Optimizing Filter Parameters. Journal of the Air
Pollution Control Association, 24 (12) : 1150-1154, 1974.
69
-------
SECTION 6
GROWTH AND NATURE OF THE INDUSTRY
PRESENT TECHNOLOGY
On January 22, 1974, the B. F. Goodrich Company notified NIOSH
that four workers from its PVC polymerization plant in Louis-
ville, Kentucky had died from a rare cancer, angiosarcoma (41).
Since then the entire VCM and PVC industry has undergone con-
siderable change as companies have tightened manufacturing
processes to reduce exposures (42).
Worker exposure to VCM is being decreased by automating proc-
esses, reducing work forces, improving reactor cleaning methods
(27), using larger reactors, improving stripping operations,
and, in general, installing control equipment.
Polyvinyl chloride plants have reduced manual cleaning of reac-
tors from once per batch to once every 12 batches (27). Between
manual cleanings the reactor is cleaned with high pressure
(68,950 kPa) water or solvent (27). The number of manual clean-
ings required can also be reduced by "clean wall" polymerization
in which vessel walls are sprayed with a special material to
prevent the polymer from sticking (27).
The use of large reactors is considered a major step forward in
polyvinyl chloride technology (27). At present, 30% of the
industry capacity is in reactors smaller than 10 m3, and 70% of
capacity is in reactors smaller than 20 m3. Large reactors,
with capacities over 70 m3, will account for 16% to 20% of
industry capacity by the end of 1975. Reactors with capacities
up to 190 m3 are currently in use. Larger reactors have fewer
connections and therefore fewer potential leaks,
(41) Preliminary Assessment of the Environmental Problems Asso-
ciated with Vinyl Chloride and Polyvinyl Chloride. A
Report on the Activities and Findings of the Vinyl Chloride
Task Force. Compiled by the Office of Toxic Substances.
Environmental Protection Agency, Washington, D.C.
September 1974. 67 pp.
(42) Plastics Industry Developing Technology for VC Standard.
Chemocology. Published by the Manufacturing Chemists Asso-
ciation, Washington, D.C., July 1975. p. 7.
70
-------
and they require less manual cleaning. Entry for manual clean-
ing has been reduced by 90%, and only 25% of the personnel pre-
viously required are now needed for plant operation (27).
After polymerization, approximately 10% to 15% of the vinyl
chloride monomer remains unreacted. Stripping is essential to
control emissions from the slurry blend tank, centrifuge, dryer,
and bulk storage silos (27). The effectiveness of stripping
depends on the type of resin and the design of the stripping
system. Other factors affecting stripping efficiency include
particle size, porosity, temperature, vacuum used, and retention
time in the stripper. Polyvinyl chloride manufacturers are
phasing out grades of resin which are difficult to strip well.
Stripping studies represent the area with the biggest potential
payoff (27).
Currently available controls for VCM emissions are a basic part
of the processing system and serves to recover reactant and pro-
duct. These controls include: recycling of vent streams,
condensation with refrigeration, adsorption to carbon, and
absorption (scrubbing). Monomer loading and unloading involve
special controls: vapor collection adapters with recycling,
thermal level detectors with recycling, and magnetic gauges.
Polymer controls include vacuum stripping, steam stripping,
silos stripping, and recycling of carrier air streams (4).
EMERGING TECHNOLOGY
Vinyl chloride monomer and polymer manufacturing and processing
industries have entered a new era since January 1974 when the
link between vinyl chloride and cancer of the liver was noted
and subsequently reported (43).
PVC manufacturing plants have controlled fugitive emissions by
designing and installing new equipment to prevent leaks in the
hundreds of pumps, valves and flanges used in their operations.
As a result of these steps, VCM levels in PVC plants now average
between 1 ppm and 3 ppm (44).
Polymerizer cleaning was a major emission source of VCM. Many
companies have been involved in research to develop a completely
closed, automated and essentially leak-proof system. One company
reports using a "Clean Reactor Technology" which involves a com-
bination of new processes, techniques and materials for treating
the interiors of reactors. This system, plus others already
(43) Researchers See Progress on VCM; Study New Hazards. Chem-
ical Marketing Reporter. 207(13), 1975.
(44) Vercalin, C. H. Curtail Vinyl Chloride Exposure. Hydro-
carbon Processing, 55(2):182, 184, 186, 1976.
71
-------
developed, virtually elimijate the possibility of operator expo-
sure to VCM in reactor cleaning and emptying operations (44).
Another source of VCM emissions is the VCM residual in PVC resins
and compounds. The industry has developed new stripping columns
which collect slurry from reactors and remove all but traces of
the VCM remaining in the slurry.
The recovered VCM is cleaned in a closed system and recycled.
The resulting resin contains less than 1 ppm of residual monomer,
which may be further reduced in compounding or processing (44).
The stripping technology reduces the already low air and water
emissions. Current reports indicate that VCM levels in outside
air at the plant fence line are on the order of a few hundredths
of a ppm. Further downwind from the plant, concentrations are
further reduced by dilution with ambient air and also because of
the actual breakdown of VCM molecules in the atmosphere (44).
INDUSTRY PRODUCTION TRENDS
In 1974, polyvinyl chloride homopolymer resins were produced by
20 companies at 35 plants. Four basic processes were used to
produce the polymer: suspension, emulsion, bulk, and solution
polymerization (4).
Suspension polymerization accounted for 78% of all PVC resin
produced in the United States, while emulsion polymerization
accounted for 13%. Bulk polymerization, a relatively new proc-
ess, was used for 6% of the United States PVC production in
1974. Three percent of the resins produced in the United States
were made by the solution polymerization process (2).
Production of polyvinyl chloride resins totaled 2.2 x 106 metric
tons in 1974 (45). PVC production in the United States grew at
an average rate of 14%/yr from 1963 to 1972 (3). In 1973, resin
sales were 12% higher than the 1972 level and production was up
10% (46). After subtracting for exports, PVC sales in 1974 de-
clined 3% from 1973 (45). OSHA regulations on the level of
vinyl chloride emissions, the short supply of VCM, and the de-
crease in the supply of chlorine and plastizers have all been
responsible for the decline in PVC sales (45).
(45) Goodbye, Resin Storage? Don't You Believe It! Modern
Plastics, 52(l):44-58, 1975.
(46) Now There's a Lot of Resin Around, But Economic Upturn
Could Resume the Pinch. Modern Plastics, 51(12):18, 1974
72
-------
Figure 17 shows PVC production for the past 30 years. The short-
age of resin supply in early 1974 created essentially a seller's
market. By the end of 1974, there was a looseness in the resin
market (46), even though supply had not changed, because of the
recession of 1974-1975 (46). High interest rates, rising prices,
inflation-deflated consumer spending power, and the slump in the
automotive and housing industries were all responsible for the
availability of PVC resins (46).
4,000
3,000
2,000
j£. 1.000
2 900
* 800
B 700
OJ
e 600
"o
500
^
o
f= 400
300
200
100
90
80
70
60
50
1945
1955 1960 1965 1970 1975 1980
YEAR
Figure 17. Polyvinyl chloride production, 1946-1979
73
-------
Table 29 summarizes U.S. consumption of polyvinyl chloride
(47-52), and Figure 18 depicts this information graphically.
Major PVC markets and their consumption of PVC resins over the
past 3 years are shown in Table 30. Table 31 shows United States
resin consumption by end use, and this is graphically illustrated
in Figure 19.
TABLE 29. UNITED STATES CONSUMPTION OF POLYVINYL
CHLORIDE RESIN BY COMPOUNDING PROCESS
(103 metric tons)
Market
Calendering
Flooring
Textile coating
Other (includes film and sheet)
Coating
Flooring
Textile and paper coating
Protective coatings and
adhesives
Other
Extrusion
Wire and cable
Film and sheet
Pipe and conduit
Other
Molding
Bottles
Records
Pipe fittings
Other
Paste processes
Plastisol
Other
Export
All other uses
TOTAL
1968 (47)
111
NAa
238
26
49
38
NA
131
61
155
NA
NA
55
NA
38
49
NA
52
77
1,080
1969 (48)
127
NA
272
30
49
41
NA
177
68
191
NA
NA
53
NA
40
57
NA
55
88
1,247
1970 (49)
113
NA
259
34
43
39
NA
186
82
223
NA
NA
64
NA
70
52
NA
86
102
1,371 .
Year
1971 (50)
131
29
218
52
66
21
21
161
92
265
143
33
60
34
38
47
23
75
72
1,571
1972 (51)
156
32
283
58
80
31
26
195
103
404
170
32
68
39
68
50
28
73
79
1,975
1973 (52)
133
33
249
69
87
32
32
188
93
570
193
39
66
41
87
39
32
66
102
2,151
1974 (45)
92
39
263
64
91
34
40
161
98
555
188
34
65
44
70
39
35
145
123
2,180
Not available.
(47) The Plastics Industry in 1968, Materials and Markets.
Modern Plastics, 46(l):27-47, 1969.
(48) The Statistics: 1969. Modern Plastics, 47(1):69-80, 1970.
(49) The Statistics for 1970. Modern Plastics, 48(1) -.65-78, 1971.
(50) The Statistics for 1971. Modern Plastics, 49 (1) :41-48, 1972.
(51) Everything's Coming Up Roses, Thorns and All. Modern
Plastics, 50(l):53-63, 1973.
(52) We Produced Over 13 Million Tons of Resins in '73? Well,
Where Is It? And How About '74? Modern Plastics,
51(1):36-47, 1974.
74
-------
<;
LLj
>-
r-l
O
ft
M-H OJ
O W
O 13
•H C
en
C 0)
O Ti
O -H
M
• O
rH
(U
8
oi 'NOIldWnSNOO
w
en
0
O
O
j^/suo}DU)eui Oi 'NOIidWnSNOO
-rH
>
0 C
ft-H
O 3
O
d ft
o e
•rH O
-P O
ft
g >1
3 X!
en
C 0)
O t3
O -H
M
• O
Cfl rH
• &
D U
00
rH
0)
•rH
Cu
75
-------
TABLE 30. CONSUMPTION OF POLYVINYL
CHLORIDE BY MAJOR MARKETS
(103 metric tons)
Market
Apparel:
Baby pants
Footwear
Outerwear
Building and construction:
Extruded foam moldings
Flooring
Lighting
Panels and siding
Pipe and conduit
Pipe fittings
Rainwater systems, soffits,
f ascias
Swimming pool liners
Weather stripping
Windows, other profiles
Electrical:
Wire and cable
Home furnishings:
Appliances
Furniture
Garden hose
Wall coverings and wood
surfacing films
Packaging:
Blow molded bottles
Closure liners and gaskets
Coatings
Film
Sheet
Recreation:
Records
Sporting goods
Toys
Transportation:
Auto mats
Auto tops
Upholstery and seat covers
Miscellaneous;
Agriculture (including pipe)
Credit cards
Laminates
Medical tubing
Novelties
Stationery supplies
Tools and hardware
Export
Other
TOTAL
1972 (51)
11
64
30
23
214
5
32
365
39
14
20
18
25
195
16
135
20
58
32
8
8
62
40
68
23
34
18
16
82
53
7
22
21
6
16
6
73
96
1,975
1973 (52)
12
66
31
26
202
5
39
520
41
16
18
16
26
188
20
145
18
54
39
9
9
59
35
66
25
38
18
15
83
66
8
23
23
7
18
8
66
93
2,151
1974 (45)
11
63
30
22
156
6
44
505
44
15
19
16
24
161
21
144
17
58
34
10
9
57
37
65
28
37
19
13
84
72
10
24
23
8
20
10
145
119
2,180
76
-------
TABLE 31. UNITED STATES CONSUMPTION OF POLYVINYL
CHLORIDE RESINS BY END USE
(103 metric tons)
End use catetory
Building and
construction
Home furnishing
Consumer good
Electrical uses
Packaging
Transportation
Miscellaneous
uses and other
unspecified uses
TOTAL
1970 (49) 1971 (50) 1972 (51) 1973 (52) 1974 (45)
456
219
188
186
123
98
101
1,371
532
225
188
161
122
109
234
1,571
737
286
229
195
150
116
262
1,975
909
219
238
188
151
116
330
2,151
851
384
234
161
147
116
287
2,180
OUTLOOK
PVC resin consumption in the United States is expected to grow
at an average rate of 8% between 1977 and 1981, leading to a
consumption level of about 2.69 x 106 metric tons/yr by 1979 (53)
Production expansion programs are cautious compared to recent
years and could lag demand by 1981 according to current expecta-
tions. VCM expansions are seen as adequate (53).
Good growth is forseen for rigid extrusion products especially
in the construction industry where pipe fittings and conduit
already take one-third of production. Flooring, window compon-
ents, and siding are being touted as areas open to extensive
penetration by PVC. The automobile sector is expected to show
moderate growth, due to shrinking cars (53).
Export markets will decline as Canadian production comes on
stream, and world capacity increases faster than demand. Some
smaller end uses, such as records, apparel and sporting goods,
are not expected to show significant growth. Packaging has suf-
fered since VCM was declared a carcinogen (53).
On the whole, the PVC industry has solved many of the technical
and regulatory problems of recent years and can now look forward
to good growth and increasing demand from many largest and use
sectors (53).
(53) Chemical Profile: PVC.
211(22):9, 35, 1977.
Chemical Marketing Reporter,
77
-------
REFERENCES
1. 1972 National Emissions Report. EPA-450/2-74-012, U.S.
Environmental Prtoection Agency, Research Triangle Park,
North Carolina, June 1974. 422 pp.
2. Oliver, G. Vvhat's the Future for PVC? Hydrocarbon Process-
ing, 45(9):281-284, 1966.
3. PVC Chemical Profile. Chemical Marketing Reporter, 205
(2):9, 1974.
4. Evans, L., C. Kleeberg, S. Wyatt, A. Basola, W. Hamilton,
and W. Vatavuk. Standard Support - Environmental Impact
Document. An Investigation of Health Effects and Emission
Reduction of Vinyl Chloride in the Vinyl Chloride Monomer
and Polyvinyl Chloride Industries. Volume II. Draft copy
of report. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, March 1975. 450 pp.
5. Carpenter, B. H. Vinyl Chloride - An Assessment of Emis-
sions Control Techniques and Costs. EPA-650/2-74-097, U.S.
Environmental Protection Agency, Washington, D.C.,
September 1974. 84 pp.
6. Suspension Polymerization. In: Encyclopedia of Polymer
Science and Technology; Volume 13: Plastics, Resins, Rubber,
Fibers. John Wiley & Sons, Inc., New York, New York, 1970.
pp. 552-571.
7. Albright, L. F. Polymerization of Vinyl Chloride. Chemical
Engineering, 74 (10) :151-158, 1967.
8. Kardos, L. A. Polyvinyl Chloride. Report No. 13 (a private
report by the Process Economics Program), Stanford Research
Institute, Menlo Park, California, June 1966. 224 pp.
9. Labine, R. A. Drying Tricks Tailor Resin Properties.
Chemical Engineering, 66 (23) : 166-169, 1959.
10. From Vinyl Chloride to...PVC by Suspension Polymerization.
Chemical Engineering, 62(7) :128, 130, 132, 1955.
78
-------
11. Ohta, K. Polyvinyl Chloride - Supplement-A. Report No. 13A
(a private report by the Process Economics Program), Stan-
ford Research Institute, Menlo Park, California, May 1970.
170 pp.
12. Albright, L. F. Vinyl Chloride Polymerization by Suspension
Process Yields Polyvinyl Chloride Resins. Chemical
Engineering, 74 (12) : 145-152, 1967.
13. Lunde, K. E. Vinyl Chloride. Report No. 5 (a private
report by the Process Economics Program), Stanford Research
Institute, Menlo Park, California, 1965. 212 pp.
14. Matheson Gas Data Book, Fourth Edition. The Matheson Com-
pany, Inc., East Rutherford, New Jersey, 1966. pp. 489-492.
15. Ruebensaal, C. F. Vinyl Resins - How Vinyl Chloride is
Made...How Vinyl Chloride is Polymerized. Chemical
Engineering, 57 (12) :102-105, 1950.
16. Meinhold, T. F., and W. M. Smith. Produces Dust Free PVC
Resins. Chemical Processing, 22(7):61-62, 1959.
17. Manufacture of Plastics, Volume I. Chapter 7. Reinhold
Publishing Corporation. W. M. Smith, ed. New York, New
York, 1964. pp. 303-343.
18. Marous, L. F., and C. D. McCleary. Polymerization Catalyst
for Vinyl Chloride. U.S. Patent 3,022,282 (to United States
Rubber Company), February 20, 1962.
19. Odian, G. Principles of Polymerization. McGraw-Hill Book
Company, New York, New York, 1970. pp. 279-298.
20. Albright, L. F. Vinyl Chloride Polymerization by Emulsion,
Bulk and Solution Processes. Chemical Engineering, 74(14):
145-152, 1967.
21. Krause, A. Mass Polymerization for PVC Resins. Chemical
Engineering, 72 (26) : 72-74, 1965.
22. Thomas, J. C. New Improved Bulk PVC Process. Hydrocarbon
Processing, 47(11):192-196, 1968.
23. Herbert, T. and S. Nagy. System Analysis of Air Pollutant
Emissions from the Chemical Plastics Industry. EPA-650/2-
74-106, (PB 239 880). Environmental Research Center. U.S.
Environmental Protection Agency, Research Triangle Park,
North Carolina, October 1974. 281 pp.
24. Douglas, S. D. Process for Producing Vinyl Resins. U.S.
Patent 2,075,429 (to Union Carbide), March 30, 1937.
79
-------
25. Reid, E. W. Process for Producing Vinyl Resins. U.S.
Patent 2,064,565 (to Union Carbide), December 15, 1936.
26. Reid, E. W. Vinyl Resins. U.S. Patent 1,935,577 (to Union
Carbide), November 14, 1933.
27. Evans, L. B., and L. L. Beck. The Vinyl Chloride and PVC
Industry Emissions and Control Techniques. Draft copy of
report. U.S. Environmental Protection Agency. Emission
Standards and Engineering Division Industrial Studies
Branch, July 23, 1974.
28. PVC Plants are Ready to Pass First Test. Chemical Week.
116(19):49-50, 1975.
29. TLVs® Threshold Limit Values for Chemical Substances in
Workroom Air Adopted by ACGIH for 1976. American Conference
of Governmental Industrial Hygienists. Cincinnati, Ohio,
1976. 94 pp.
30. EPA Programs of Monitoring Vinyl Chloride in Ambient Air.
Environmental Protection Agency Office of Air Quality
Planning and Standards. Research Triangle Park, North
Carolina, February 2, 1976. 14 pp.
31. Turner, D. B. Workbook of Atmospheric Dispersion Estimates.
Public Health Service Publication No. 999-AP-26, U.S. Depart-
ment of Health, Education, and Welfare, Cincinnati, Ohio,
May 1970. 84 pp.
32. Air Pollution Engineering Manual, Second Edition. J. A.
Danielson, ed. Publication No. AP-40, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina,
May 1973. 987 pp.
33. Hughes, T. W., D. A. Horn, C. W. Sandy, and R. W. Serth.
Source Assessment: Prioritization of Air Pollution from
Industrial Surface Coating Operations. EPA-650/2-75-019-a,
U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, February 1975. 303 pp.
34. Hydrocarbon Pollutant Systems Study. Volume I. Stationary
Sources, Effects and Control. APTD-1499 (PB 219 073), U.S.
Environmental Protection Agency, Research Triangle Park,
North Carolina, October 20, 1972. 379 pp.
35. Chemical Engineers Handbook, Fifth Edition. J. H. Perry and
C. H. Chi1ton, eds. McGraw-Hill Book Company, New York, New
York, 1973.
80
-------
36. Control Techniques for Hydrocarbons and Organic Solvent
Emissions from Stationary Sources. Publication No. AP-68,
U.S. Department of Health, Education, and Welfare, Washing-
ton, D.C., March 1970. pp. 3-1 through 3-26.
37. Rolkes, R. W., R. D. Hawthorne, C. R. Garbett, E. R. Slater,
T. T. Phillips, and G. D. Towell. Afterburner Systems
Study. EPA-R2-72-062 (PB 212 560), U.S. Environmental Pro-
tection Agency, Research Triangle Park, North Carolina,
August 1972. 512 pp.
38. Mantell, G. J., J. T. Barr and R. K. S. Chan. Vinyl Chlor-
ide Emission Control: Stripping VCM from PVC Resin.
Chemical Engineering Progress, 71(9) :54-62, 1975.
39. Frey, R. E. Types of Fabric Filter Installations. Journal
of the Air Pollution Control Association, 24 (12) :1148-1149,
1974.
40. Bakke, E. Optimizing Filter Parameters. Journal of the Air
Pollution Control Association, 24 (12) : 1150-1154, 1974.
41. Preliminary Assessment of the Environmental Problems Asso-
ciated with Vinyl Chloride and Polyvinyl Chloride. A
Report on the Activities and Findings of the Vinyl Chloride
Task Force. Compiled by the Office of Toxic Substances.
Environmental Protection Agency, Washington, D.C., September
1974. 67 pp.
42. Plastics Industry Developing Technology for VC Standard.
Chemocology. Published by the Manufacturing Chemicals
Association, Washington, D.C., July 1975. p. 7.
43. Researchers See Progress on VCM; Study New Hazards. Chemi-
cal Marketing Reporter. 207 (13) :1975.
44. Vercalin, C. H. Curtail Vinyl Chloride Exposure. Hydrocar-
bon Processing, 55(2) :182, 184, 186, 1976.
45. Goodbye, Resin Storage? Don't You Believe It! Modern
Plastics, 52(l):44-58, 1975.
46. Now There's a Lot of Resin Around, But Economic Upturn
Could REsume the Pinch. Modern Plastics, 51(12):18, 1974.
47. The Plastics Industry in 1968, Materials and Markets. Mod-
ern Plastics, 46(l):27-47, 1969.
48. The Statistics: 1969. Modern Plastics, 47(1):69-80, 1970.
49. The Statistics for 1970. Modern Plastics, 48(l):65-78, 1971.
81
-------
50. The Statistics for 1971. Modern Plastics, 49(l):41-48, 1972.
51. Everything's Coming Up Roses, Thorns and All. Modern
Plastics, 50(l):53-63, 1973.
52. We Produced Over 13 Million Tons of Resins in '73? Well,
Where Is It? And How About '74? Modern Plastics, 51(1):
36-47, 1974.
53. Chemical Profile: PVC. Chemical Marketing Reporter, 211
(22) :9,35, 1977.
54. Martin, D. 0., and J. A. Tikvart. A General Atmospheric
Diffusion Model for Estimating the Effects on Air Quality of
One or More Sources. Presented at the 61st Annual Meeting
of the Air Pollution Control Association, St. Paul,
Minnesota, June 23-27, 1968. 18 pp.
55. Eimutis, E. C., and M. G. Konicek. Derivations of Contin-
uous Functions for the Lateral and Vertical Atmospheric
Dispersion Coefficients. Atmospheric Environment, 6(11):
859-863, 1972.
56. Gifford, F. A., Jr. An Outline of Theories of Diffusion in
the Lower Layers of the Atmosphere. In: Meteorology and
Atomic Energy 1968, Chapter 3, D. A. Slade, ed. Publica-
tion No. TID-24190, U.S. Atomic Energy Commission Technical
Information Center, Oak Ridge, Tennessee, July 1968. p. 113.
57. Code of Federal Regulations, Title 42 - Public Health,
Chapter IV - Environmental Protection Agency, Part 410 -
National Privacy and Secondary Ambient Air Qualtiy Stand-
ards, April 28, 1971. 16 pp.
82
-------
APPENDIX
DERIVATION OF SOURCE SEVERITY EQUATIONS3
SUMMARY OF MAXIMUM SEVERITY EQUATIONS
The maximum severity of pollutants may be calculated using the
mass emission rate, Q, the height of the emissions, H, and the
ambient air quality standard, AAQS. The equations summarized in
Table A-l are developed in detail in this appendix.
TABLE A-l. POLLUTANT SEVERITY EQUATIONS FOR ELEVATED SOURCES
Pollutant Severity equation
Particulate Sp = ~^2~
Hydrocarbons Sur, = 16^2Q
DERIVATION OF xm-,v FOR USE WITH U.S. AVERAGE CONDITIONS
JlldX
The most widely accepted formula for predicting downwind ground
level concentrations from a point source is (28):
X = „„'„ „ exp|- \
vv
where x - downwind ground level concentration at reference
coordinate x and y with emission height of H, g/m3
Q = mass emission rate, g/s
°y = standard deviation of horizontal dispersion, m
°z = standard deviation of vertical dispersion, m
u = wind speed, m/s
y = horizontal distance from centerline of dispersion, m
This appredix was prepared by T. R. Blackwood and E. C. Eimutis,
Monsanto Research Corporation, Dayton Laboratory, Dayton, Ohio.
83
-------
H = height of emission release, m
x = downwind emission dispersion distance from source of
emission release, m
TT = 3.14
We assume that Xmax occurs when x »0 and y = 0. For a given
stability class, standard deviations of horizontal and vertical
dispersion have often been expressed as a function of downwind
distance by power law relationships as follows (54) :
oy = axb (A-2)
a = cxd + f (A-3)
z
Values for a, b, c, d, and f are given in Tables A-2 and A-3.
Substituting these general equations into Equation A-l yields:
x = vrr^ h exp[ ?r 1 (A~4)
ac^uxD+a + avufx L 2 (cxa + f) 2J
Assuming that Xmax occurs at x <100 m or the stability class is
C, then f = 0 and Equation A-4 becomes:
r -H
|_2c2x
(A-5)
aCTTUX" ~ ' "--v2Cl
A^ = — and B_ =
For convenience, let:
so that Equation A-5 reduces to:
, -(b+d)
X = ARx exp
B
R
_x2d
Taking the first derivative of Equation A-6
Q*. = a I x~b~d exo/B -x-2d^ M-2dB x"2^"1
dx AR ) X [ p\ R /J\ R
(A-6)
+ expfi x-^U-b-dW'13"^1 (A-7)
(54) Martin, D. 0., and J. A. Tikvart. A General Atmospheric
Diffusion Model for Estimating the Effects on Air Quality of
One or More Sources. Presented at the 61st Annual Meeting
of the Air Pollution Control Association, St. Paul,
Minnesota, June 23-27, 1968. 18 pp.
84
-------
TABLE A-2. VALUES OF a FOR THE
COMPUTATION OF a a (54)
Stability class
A
B
C
D
E
F
0.3658
0.2751
0.2089
0.1471
0.1046
0.0722
-j V"!
For Equation A-2: a = ax
where x = downwind distance
b = 0.9031 (Reference 55)
TABLE A-3.
VALUES OF THE CONSTANTS USED TO
ESTIMATE VERTICAL DISPERSION3 (54)
Stability
Usable range, m class Coefficient
>1,000 A
B
C
D
E
F
100 to 1,000 A
B
C
D
E
F
<100 A
B
C
D
E
F
0.00024
0.055
0.113
1.26
6.73
18.05
C2
0.0015
0.028
0.113
0.222
0.211
0.086
= 3
0.192
0.156
0.116
0.079
0.063
0.053
2.094
1.098
0.911
0.516
0.305
0.18
d2
1.941
1.149
0.911
0.725
0.678
0.74
d3
0.936
0.922
0.905
0.881
0.871
0.814
-9.6
2.0
0.0
-13
-34
-48.6
£2
9.27
3.3
0.0
-1.7
-1.3
-0.35
£3
0
0
0
0
0
0
For Equation A-3:
°z = cx
+ f
(55) Eimutis, E. C., and M. G. Konicek. Derivations of Contin-
uous Functions for the Lateral and Vertical Atmospheric
Dispersion Coefficients. Atmospheric Environment, 6(11):
859-863, 1972.
85
-------
and setting this equal to zero (to determine the roots which give
the minimum and maximum conditions of x with respect to x) yields
x-2d -b _ d\ (A_8)
Since we define that x ^ 0 or °° at x / the following expression
must be equal to 0 : max
Therefore
-2dBrx-2d -d - b = 0 (A-9)
(b + d) x2d = -2dB0 (A-10)
K.
or
-i "D *) f^ TT £- f\ TT ^
__?rl _ K _ ^uil __ an
b + d 2c2 (b + d) c2 (b + d)
Hence
x = —-—V at x (A-12)
d)/ max
Thus Equations A-2 and A-3 (at f = 0) become:
a = a( d H b/2d (A-13)
2(d + b)
d H2 \d/2d / d
a = i
z
The maximum will be determined for U.S. average conditions of
stability. According to Gifford (56) , this is when a., = az.
Since b = 0.9031, and upon inspection of Table A-2 unaer U.S.
(56) Gifford, F. A., Jr. An Outline of Theories of Diffusion in
the Lower Layers of the Atmosphere. In: Meteorology and
Atomic Energy 1968, Chapter 3, D. A. Slade, ed. Publication
No. TID-24190, U.S. Atomic Energy Commission Technical Infor-
mation Center, Oak Ridge, Tennessee, July 1968. p. 113.
86
-------
average conditions, oy = oz, it can be seen that 0.881
<_ d £ 0.905 (class C stability3). Thus, it can be assumed that
b is nearly equal to d in Equations A-13 and A-14 or:
a = — (A-15)
Z /2
and
a = - — (A-16)
Y c /2
Under U.S. average conditions, ay = az and a = c if b = d and
f = 0 (between class C and D, bur closer to belonging in class C).
Then
a = — (A-17)
Y /2
Substituting for ay from Equation A-17 and for oz from Equa-
tion A-15 into Equation A-l and letting y = 0:
X = — exp
Am n v ~ "
max
H
(A-18)
or
X = 2 Q (A-19)
Amax
DEVELOPMENT OF SOURCE SEVERITY EQUATIONS
Source severity, S, has been defined as follows
~ AAQS
where x = time-averaged maximum ground level concentration
max
AAQS = ambient air quality standard
The values given in Table A-3 are mean values for stability
class. Class C stability describes these coefficients and
exponents, only within about a factor of two.
87
-------
Values of x are found from the following equation:
in 9.x
/t°-17
xmax = xmax \t
where to is the "instantaneous" (i.e., 3-min) averaging time and
t is the averaging time used for the ambient air quality standard
as shown in Table A-4.
TABLE A-4. SUMMARY OF NATIONAL AMBIENT AIR
QUALITY STANDARDS (57)
Pollutant
Particulate
Hydrocarbons
(nonme thane)
Averaging
time
Annual
(geometric mean)
24 hrb
3 hr
(6 to 9 a.m.)
Primary
standards
75 y g/m 3
260 yg/m3
160 yg/m3
(0.24 ppm)
Secondary
standards
60a yg/m3
150 yg/m3
(Same as
primary)
The secondary annual standard (60 yg/m3) is a guide for assess-
ing implementation plans to achieve the 24-hr secondary standard.
Not to be exceeded more than once per year.
Hydrocarbon Severity
The primary standard for hydrocarbon is reported for a 3-hr
averaging time. Therefore, t = 180 min. Hence, from
Equation A-21:
*max = xmax (ifo) ' = °'5xmax (A~22)
Substituting for x -, from Equation A-19 yields:
max
(0.5)(0.052) Q = 0.026 Q (A 23)
H2 H2
For hydrocarbons, AAQS = 1.6 x lO"4 g/m3. Therefore
S = ^ = °'026 Q (A-24)
AAQS 1.6 x 10-l+ H2
(57) Code of Federal Regulations, Title 42 - Public Health, Chap-
ter IV - Environmental Protection Agency, Part 410 - National
Privacy and Secondary Ambient Air Quality Standards, April 28,
1971. 16 pp.
88
-------
or
SHC = •*•"'•••' ^ (A-25)
Particulate Severity
The primary standard for particulate is reported for a 24-hr
averaging time. Therefore, t = 1,440 minutes. Hence, for
Equation A-21:
- / 3 \°-17
xmax = xmax ( TTTTT ) (A-26)
Substituting for x from Equation A-19 yields:
- = ^ (Q_35) =
max H2 R2
For particulates, AAQS = 2.6 x 10"^ g/m3. Therefore
_ xmax _ 0.0182 Q
~ C ~~ , ^ '
2.6 x lO-4 H2
or
S = (A-29)
H2
AFFECTED POPULATION CALCULATION
Another form of the plume dispersion equation is needed to calcu-
late the affected population since the population is assumed to
be distributed uniformly around the source. If the wind direc-
tions are taken to 16 points and it is assumed that the wind
directions within each sector are distributed randomly over a
period of a month or a season, it can be assumed that the efflu-
ent is uniformly distributed in the horizontal within, the sector.
The appropriate equation for average concentration, \, in g/m3
is then (57) :
- 2.03 Q f I/ H\2 ._ _n.
* = -VU5T SXP - 2 — (A-30)
z L \
To find the distances at which x/AAQS - 1-0, roots are determined
for the following equation:
2'03 Q expl- y(-^l I = 1.0 (A-31)
(AAQS) a ux
["- V^!21 -
L 4°J \ ~
89
-------
keeping in mind that:
= ax + c
where a, b, and c are functions of atmospheric stability and are
assumed to be selected for stability Class C. Since Equation
A-28 is a transcendental equation, the roots are found by an
iterative technique using the computer.
For a specified emission from a typical source, x/AAQS as a
function of distance might look as follows:
x
AAQS
1.0
Figure A-l.
X
DISTANCE FROM SOURCE
as a function of distance from source.
AAQS
The affected population is contained in the area
A = TT(X22 - Xl2)
(A-32)
If the affected population density is Dp, the total affected popu-
lation, P, is
P = DA (persons)
(A-33)
90
-------
TECHNICAL REPORT DATA
(Please read Instructions on lite reverse before completing)
1 REPORT NO.
EPA-600/2-78-0041
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
SOURCE ASSESSMENT:
POLYVINYL CHLORIDE
6 REPORT DATE
May 1978 issuincr date
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
Z. S. Khan and T. W. Hughes
8. PERFORMING ORGANIZATION REPORT NO.
MRC-DA-700
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Monsanto Research Corporation
1515 Nicholas Road
Dayton, OH 45407
»O. PROGRAM ELEMENT NO.
1AB604
11. CONTRACT/GRANT NO
68-02-1874
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Lab., Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Task Final 3/75-8/77
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOTES
IERL-Ci project leader for this report is Ronald J. Turner,
513-684-4481.
16. ABSTRACT
This report summarizes data on air emissions from the polyvinyl
chloride (PVC) industry. PVC is manufactured by 20 companies at 35
plants. Each plant uses one or more of four possible polymerization
processes: (1) suspension polymerization, (2) emulsion polymerization
(3) bulk polymerization, and (4) solution polymerization. A repre-
sentative PVC plant was defined to assess the severity of emissions
from this industry. Source severity, defined as the ratio of the
time-averaged maximum ground level concentration of a pollutant to
a hazard potential, was calculated for 16 chemical species emitted
from a representative plant. The two largest severities were 970 for
vinyl chloride and 1.9 for PVC. Control technology for hydrocarbons
includes adsorption, absorption, refrigeration, incineration,
stripping, purging of equipment with inert gas or water, and control
of fugitive emissions.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air Pollution
Assessments
b. IDENTIFIERS/OPEN ENDED TERMS
Air Pollution Contro
Source Assessment
Source Severity
COSATl Field/Group
68A
8 DISTRIBUTION STATEMENT
Release to Public
19 SECURITY CLASS (This Report)
Unclassified
21 NO. OF PAGES
105
2O SECURITY CLASS (Thlt pagel
Unclassified
23. PRICE
EPA Form 222O-I (»-73)
91
4US GOVERNMENT PRINTING OFFICE 1978— 757-140/68a8
------- |