United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
March 1980
Air
VOC Fugitive           Draft
Emissions in Synthetic  EIS
Organic Chemicals
Manufacturing Industry —
Background Information
for Proposed Standards
Preliminary
Draft

-------
                             NOTICE

This document has not been formally released by EPA and should not now be construed to represent
Agency policy. It is being circulated for comment on its technical accuracy and policy implications.
               VOC  Fugitive  Emissions
         in  Synthetic  Organic Chemicals
              Manufacturing  Industry —
               Background Information
               for  Proposed  Standards
                   Emission Standards and Engineering Division
                   U.S. ENVIRONMENTAL PROTECTION AGENCY
                      Office of Air, Noise, and Radiation
                    Office of Air Quality Planning and Standards
                   Research Triangle Park, North Carojina 27711

                             March 1980

-------
This report has been revised by the Emission Standards and
Engineering Division of the Office of Air Quality Planning and
Standards, EPA, and approved for publication.   Mention of trade
names or commercial products is not intended to constitute
endorsement or recommendation for use.   Copies of this report
are available through the Library Services Officer (MD-35),
U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina 27711, or from the National  Technical
Information Services, 5285 Port Royal Road, Springfield,
Virginia 22161.

-------
                         METRIC  CONVERSION  TABLE

    In keeping with  U.S.  Environmental  Protection  Agency  policy,  metric
units are used in  this  report.   These  units may be converted to common
English units by using  the  following conversion factors:
        Metric Unit
             m
             m
             m,
             m"
             Kg
             Mg
             Gg
             GJ
             GJ
            J/g
          Nm /sec
             m/s
        Metric Name

          LENGTH

          meter
          meter

          VOLUME

          1i ters
       cubic meters
       cubic meters

          HEIGHT

   kilogram (106 grams)
   megagram (10g grams)
   gigagram (10  grams)

          ENERGY

         giga joule
         gigajoule
       joule per gram

     VOLUMETRIC FLOW   -.

normal cubic meters per second

          SPEED

    meters per second
 Equivalent
English Unit
 39.3700 in.
  3.2810 ft.
  0.2.642 U.S. gal
   264.2 U.S. gal
    6.29 Barrels (bbl)
  2.2046 Ib.
  1.1023 tons
  1,102.3 tons
  9.48 X 10  Btu
    277.76 KWh
     0.430 Btu/lb.
   2242 SCFM (ft /min)
   196.86 ft/min
    Temperature in  degrees  Celcius  (°C)  can  be  converted to temperature
in degrees Farenheit (°F) by  the  following formula:
                           (°F)  =  1.8 (°C)  +  32

-------
                               TABLE OF CONTENTS
                                                                          Page
LIST OF FIGURES	x
LIST OF TABLES   .  .-	vi
ABBREVIATIONS AND  CONVERSION  FACTORS   	  	  .  .
2.  . INTRODUCTION	  ...  2-1
     2.1  Authority  for  the Standards	2-1
     2.2  Selection  of Categories of Stationary Sources  	  2-5
     2.3  Procedure  for  Development of Standards of
          Performance	2-7
     2.4  Consideration  of Costs	2-8
     2.5  Consideration  of Environmental  Impacts   	  2-10
     2.6  Impact on  Existing  Source  	  2-11
     2.7  Revision of  Standards through Experience  	  2-12
3.   DESCRIPTION OF  FUGITIVE  EMISSION  SOURCES  	  3-1
     3.1  Introduction and General  Industry  Information	  .  3-1
     3.2  Fugitive Emission Definition and Potential	3-3
     3.3  Baseline Control	  3-17
     3.4  References	3-21
'4.   EMISSION CONTROL  TECHNIQUES	4-1
     4.1  Leak  Detection and  Repair Methods	4-1
     4.2  Equipment Specifications	4-13
     4.3  References.	  4-24
5.   MODIFICATION  AND  RECONSTRUCTION	  .  5-1
     5.1  General  Discussion  of Modification
          and Reconstruction		5-1
     5.2  Applicability  of Modification and  Reconstruction
          Provisions to  the SOCMI	.5-3
6.   MODEL  PROCESS UNITS AND  REGULATORY ALTERNATIVES	6-1
     6.1  Model Units	6-1
     6.2  Regulatory Alternatives	  6-4
     6.3  References	6-7

-------
                           TABLE OF CONTENTS (cont.)
                                                                        Page
7.    ENVIRONMENTAL IMPACT 	  7-1
     7.1   Impact on Atmospheric Emissions 	  7-1
     7.2  Impact of Water Quality	7-8
     7.3  Impact on Solid Waste	 	  7-11
     7.4  Enerqy Impact	  7-12
     7.5  Other Environmental Concerns	7-14
     7.6  References	7-14
8.    COST ANALYSIS	8-1
     8.1   Cost Analysis of Regulatory Alternatives	8-1
     8.2  Other Cost Considerations .	8-24
     8.3  References	8-25
9.    ECONOMIC ANALYSIS	9-1
     9.1   Industry Profile	9-1
     9.2  Economic Impact Analysis	9-16
     9.3  Socio-Economic and Inflationary Impacts 	  9-40
     9.4  References	9-41
APPENDIX C	C-l
APPENDIX D	D-l
APPENDIX E	E-l
APPENDIX F	F-l

-------
                                 LIST  OF TABLES


                                                                         Page

 3-1  Approximate Level of Uncontrolled Fugitive Emission
      Factors in the Synthetic Organic Chemical Manufacturing
      Industry  (SOCMI)  .  .	3-18

 4-1  Fraction  of Total Mass  Emissions From Various Source
      Types That Would  be Controlled  by Different Action Levels	4-7

 4-2  Estimated Occurrence and Recurrence Rate for Various
      Monitoring Intervals	 4-10

. 4-3  Percent of Mass Emissions  Affected by Various Repair
      Intervals	4-11

 4-4  Average Emission  Rates  from Sources Above 10,000 ppmv
      and at 1000 ppmv	4-11

 4-5  Example of Control  Efficiency Calculation	 4-15

 4-6  Impact of Monitoring Interval on Correction Factor Accounting
      for Leak  Occurrence/Recurrence  (For Example Calculation) 	 4-15

 4-7  Effectiveness of  Equipment Modifications 	 4-23

 6-1  Fugitive  Emission Sources  for Three Model Units  	 6-3

 6-2  Regulatory Alternatives for NSPS for Fugitive Emission
      Sources in the SOCMI	6-5

 7-1  Emission  Factors  for Sources Controlled Under Regulatory
      Alternative II	7-3

 7-2  Emission  Factors  for Sources Controlled Under Regulatory
      Alternative III	7-4

 7-3  Emission  Factors  for Sources Controlled Under Regulatory
      Alternative IV	7-5  .

 7-4  Example Calculation of  VOC Fugitive Emissions from Model
      Unit A Under Regulatory Alternative II  	 7-6

 7-5  Estimated Emissions and Emission Reduction on A Model
      Unit Basis	7-7

 7-6  Total VOC Fugitive  Emissions from Affected Model Units
      for Regulatory Alternatives   	 7-10

-------
                           LIST OF TABLES  (cont.)

                                                                          Page

8-1   Capital Cost Data	8-2

8-2   Capital Cost Estimates for New Model Units	8-3

8-3   Annual Monitoring and Leak Repair Labor Requirements for
      Regulatory Alternative II 	  8-5

8-4   Annual Monitoring and Leak Repair Labor Requirements for
      Regulatory Alternative III  	  8-6

8-5   Annual Monitoring, and Leak Repair Labor Requirements for
      Regulatory Alternative IV 	  8-7

8-6   Derivation of Annualized Labor, Administrative, Maintenance,
      and Capital Charges	8-8

8-7   Labor-Hour Requirements for Initial  Leak Repair 	  8-11

8-8   Recovery Credits	8-12

8-9   Annualized Control  Cost Estimated for Model Unit A	8-13

8-10  Annualized Control  Cost Estimated for Model Unit B	8-14

8-11  Annualized Control  Cost Estimated for Model Unit C	8-15

8-12  Cost Effectiveness for Model Units	8-16

8-13  Capital Cost Estimates for Modified/Reconstructed
      Facilities	8-18

8-14  Annualized Control  Cost Estimated for Modified/
      Reconstructed Model Units Under Regulatory Alternative  IV  	  8-19

8-15  Nationwide Costs for the Industry Under Regulatory
      Alternative II	8-21

8-16  Nationwide Costs for the Industry Under Regulatory
      Alternative III	8-22

8-17  Nationwide Costs for the Industry Under Regulatory
      Alternative IV	8-23

8-18  Statutes That May be Applicable to SOCMI	8-25

9-1   Estimated  Plant Capacity By State  	  9-3

-------
                           LIST OF. TABLES (cont.)

                                                                          Page


9-2   Distribution of Plants by Capacity and Region	 . .  .9-4
                                                                              '
9-3   Distribution of Industry Capacity By Plant Size and Region  .... 9-6

9-4   Production and Sales of Synthetic Organic Chemicals 	 9-7

9-5   SOCMI Resource Use	9-9 -

9-6   Industrial Organic Chemicals:   U.S. Imports and Exports,         •
      1966-77	 9-10

9-7   Industrial Organic Chemicals:   U.S. Trade, By Principal
      Trading Partners, 1976 and 1977	9-13

9-8   Industrial Organic Chemicals:   U.S. Imports For Consumption
      By Principal Source, 1972-77	9-14

9-9   Industry Concentration  	 9-18

9-10  Estimated Cost of Capital for Firms in SOCMI	9-19

9-11  Average Rate of Return Impacts	9-28

9-12  Model Units Experiencing Significant Rate of Return Impacts
      Under Full Cost Absorption	9-30

9-13  Average Price Impacts of Regulatory Alternatives.	 9-32

9-14  Model Units Requiring Significant Price Increases  to
      Maintain Target Rates of Return 	 	 9-33

9-15  Investment Impacts of Regulatory Alternatives .  .  . ...  .  ... .  . 9-36

9-16  Employment Impacts of Regulatory Alternatives 	 9-38

9-17  Model Unit and Industry Annualized Control Costs	9-39

C-l   Frequency, of Leaks from Sources In Synthetic Organic
      Chemical Plants		 C-4

C-2   Sampled Process Units from Nine Refineries During
      Refinery Study	 .  . £--5 ,

C-3   Leak Frequencies and Emission Factors from Fugitive Sources
      in Petroleum Refineries  	 C-7
                                      vm

-------
                           LIST OF TABLES (cont.)

                                                                         Page

C-4   Comparison of Leak Frequencies for Fugitive Emission
      Sources in SOCMI Units and Petroleum Refineries 	 C-8

C-5   Summary of Maintenance Study Results from Union Oil Company
      Refinery in Rodeo, California 	 C-10

C-6   Summary of Maintenance Study Results from the Shell Oil
      Company Refinery in Martinez, California	C-12

C-7   Summary of EPA Refinery Maintenance Study Results  	 C-14

C-8   Unit D-Ethylene Unit Block Valve Repair 	 C-15

El-1  SOCMI Chemicals Included in ITC Category Ratios 	 E-3

El-2  Ratios Used To Weight ITC Data	E-5

E2-1  Projections of Replacement Investment 	 E-8

E3-1  Yields by Rating Class for Cost of Debt Funds, 1979	E-12

E3-2  Financial Data for 100 Firms in SOCMI	E-14

-------
                               LIST OF FIGURES
 3-1   General  schematic of process levels that make up the organic
      chemical  industry ...... .  .................  3-2

 3-2   Diagram  of a simple packed seal .................  3-4

 3-3   Diagram  of a basic single mechanical seal-.-., ....... ...  3-5

 3-4   Diagram  of a double mechanical  seal (back-to-back arrangement) .  3-6

 3-5   Diagram  of a double mechanical  seal (tandem arrangement) ....  3-6

 3-6   Chempump canned-motor pump  .................. .  3-7

 3-7   Shriver  mechanically actTFated-daap.hragm pump ......... t  3-8
                                          ""' "''"*' -- -f.^.
,3-8   Liquid-film compressor shaft seal. ...... . ; = \-., ,  .....  3-9

 3-9   Diagram  of a gate valve .....................  3-10::

 3-10 Example  of bellows seals .......  . .  ...........  3-11

 3-11 Diagrams of valves with diaphragm seals ..... . .......  3-12

.3-12 Diagram  of a spring-loaded relief valve .............  3-13

'3-13 Cooling  tower (cross-flow)  ...................  3-14

 3-14 Diagram  of hydraulic seal for agitators .......... ...  3-15

 3-15 Diagram  of agitator lip seal ............ ......  3-16

 4-1   Cumulative distribution  of total emissions by screening values -
      valves - gas/vapor streams  ...................  4-16

 4-2  Cumulative distribution of  sources by screening values -
      valves - gas/vapor streams  ...................  4-16

-------
                           2.   INTRODUCTION

     Standards of performance  are proposed following  a  detailed investi-
gation of air pollution control  methods  available to  the affected industry
and the impact of their costs  on the industry.   This  document summarizes
the information obtained from  such a study.   Its purpose is to explain in
detail the background and basis  of the proposed standards and to facilitate
analysis of the proposed standards by interested persons, including those
who may not be familiar with the many technical aspects of the industry.
To obtain additional  copies of this document or the Federal Register
notice of proposed standards,  write to EPA Library (MD-35), Research
Triangle Park, North  Carolina, 27711.  When  ordering, specify the Back-
ground Information Document (BID), Volume 1:  Proposed  Standards of
Performance for the Synthetic  Organic Chemical  Manufacturing Industry.
2.1  AUTHORITY FOR THE STANDARDS
     Standards of performance  for new stationary sources are established
under Section 111 of  the Clean Air Act (42 United States Code 7411), as
amended, hereafter referred to as the Act.   Section 111 directs the
Administrator to establish standards of  performance for any category of
new stationary source of air pollution which "...causes or contributes
significantly to health or welfare."
     The term new source is defined as "any  stationary  source, the
construction or modification of  which is commenced after the publication
of regulations (or, if earlier,  proposed regulations) prescribing a
standard of performance under  this section which will be applicable to
such source."
     The term stationary source  is further defined as "any building,
structure, facility or installation which emits or may  emit any air
pollutant."
                                  2-1

-------
     The term standard of performance as applied to stationary sources
(other than fossil-fuel-fired sources) is defined as an "allowable emis-
sion limitation for such category of sources."  The Act requires  that
standards of performance for stationary sources reflect, "...the  degree
of emission limitation achievable through the application of the  best
technological system of continuous emission reduction.  . .  the Administra-
tor determines has been adequately demonstrated."  In addition, for
stationary sources whose emissions result from fossil fuel  combustion,
the standard must also include a percentage reduction in emissions.  The
Act also provides that the cost of achieving the necessary emission
reduction, the non-air quality health and environmental impacts and the
energy requirements all be taken into account in establishing standards
of performance.
     The term technological system of continuous emission reduction is
interpreted as either:
     1)   "a technological process for production or operation by any
          source which is inherently low polluting or nonpolluting", or
     2)   "a technological system for continuous reduction of the
          pollution generated by a source before such pollution is
          emitted into the ambient air, including precombustion
          cleaning or treatment of fuels."
     If a standard of performance as defined above cannot be prescribed
or enforced, "the Administrator . . . may distinguish among classes,
types, and sizes within categories of new sources for the purposes of
establishing such standards."  This allows certain types and sizes of
facilities to be exempted from compliance with a general standard, or to
have a different standard of performance specified.  This might be done,
for example, to avoid extreme economic hardship on very small facilities.
     Section 111 prescribes three steps to follow in establishing
standards of performance.
     1)   The Administrator must identify those categories of stationary
          sources for which standards of performance will ultimately
          be promulgated  by listing them in the Federal Register.  For
          those categories of major stationary sources which have not
                                  2-2

-------
          already been  listed,  the  following  schedule  has  been  specified
          for the promulgation  of standards:   25  percent by  August  7,
          1980,  75 percent by August  7,  1981, and 100  percent by  August 7,
          1982.
     2)   The regulations  applicable  to  a  category so  listed must be
          proposed by publication in  the Federal  Register  within  120
          days of its listing.   This  proposal provides interested persons
          an opportunity to comment.
     3)   Within six months after proposal,  the standard must be
          promulgated,  incorporating  any alterations deemed  necessary  or
          desirable.
     It is further required that standards of performance  be reviewed
every four years.  If there has been  significant  change in the  industry or
control technology, then the standard must be revised  to reflect  the new
condition.
     Standards of performance do not  guarantee protection  of health or
welfare because they are not designed to achieve  any specific air quality
levels.  They are designed to reflect the degree  of emission limitation
achievable through application  of the best adequately  demonstrated  tech-
nological  system of continuous  emission  reduction.  In this  application
the cost of achieving such emission reduction, non-air quality  health  and
environmental impacts,  and energy requirements should  be considered.
     Congress had several  reasons for including these  requirements.  First,
standards with a degree of uniformity are needed to avoid  situations where
some states may attract industries by relaxing standards relative to other
states.  Second, stringent standards  enhance the potential for  long-term
growth.  Third, stringent standards may  help achieve  long-term  cost savings
by avoiding the need for more expensive  retrofitting when  pollution
ceilings may be reduced in the  future.  Fourth, the standard-setting
process should create incentives for improved technology.
     Promulgation of standards  of performance does not prevent  state  or
local agencies from adopting more stringent emission  limitations  for the
same sources.  States are free  under Section 116 of the Act to  establish
even more stringent emission limits than those established under  Section
                                  2-3

-------
Ill or those necessary to attain or maintain the National  Ambient Air
Quality Standards (NAAQPS) under Section 110.   Thus,  new sources may in
some cases be subject to limitations more stringent than standards of
performance under Section 111, and prospective owners and  operators of
new sources should be aware of this possibility in planning for such
facilities.
     A similar situation may arise when a major emitting facility is to
be constructed in a geographic area which falls under the  prevention of
significant deterioration of air quality provisions of Part C of the Act.
These provisions stipulate, among other things, that major emitting facil-
ities to be constructed in such areas are to be subject to best available
control technology.  The term "Best Available Control Technology (BACT),
as defined in the Act, means "...an emission limitation based on the
maximum degree of reduction of each pollutant subject to regulation under
this Act emitted from or which results from any major emitting facility,
which the permitting authority, on a case-by-case basis, taking into
account energy, environmental, and economic impacts and other costs,
determines is achievable for such facility through application of pro-
duction processes and available methods, systems, and techniques,
including fuel cleaning or treatment or innovative fuel combustion tech-
niques for control of each such pollutant.  In no event shall application
of  "Best Available Control Technology1 result in emissions of any pollu-
tants which will exceed the emissions allowed by any applicable standard
established pursuant to Section 111 or 112 of this Act."
     In addition, Section lll(h) authorizes the Administrator to grant
waivers of compliance to permit a source to use innovative continuous
emission control technology.  In order to grant the waiver, the Adminis-
trator must find:  (1) a substantial likelihood that the technology will
produce greater emission reductions than the standards require, or an
equivalent reduction at lower economic, energy or environmental costs;
(2) the proposed system has not been adequately demonstrated; (3) the
technology will not cause or contribute to an unreasonable risk to public
health, welfare or safety;  (4) the governor of the state where  the source
is  located consents; and that (5) the waiver will not  prevent the
                                  2-4

-------
attainment or maintenance of any ambient standard.   A waiver may have
conditions attached to assure the source will  not prevent attainment of
any NAAQPS.   Any such condition will  have the force of a performance
standard.  Finally, waivers have definite end dates and may be terminated
earlier if the system fails to perform as expected.  In such a case, the
source may be given up to three years to meet the standards, with a
mandatory progress schedule.
2.2  SELECTION OF CATEGORIES OF STATIONARY SOURCES
     Section 111 of the Act directs the Administrator to list categories
of stationary sources which have not been listed before.  The Administrator
"...shall include a category of sources in such a list if in his judgment
it causes, or contributes significantly to, air pollution which may
reasonably be anticipated to endanger public health or welfare."  Proposal
and promulgation of standards of performance are to follow while adhering
to the schedule referred to earlier.
     Since passage of the Clean Air Amendments of 1970, considerable
attention has been given to the development of a system for assigning
priorities to various source categories.  The approach specifies areas of
interest by considering the broad strategy of the Agency for implementing
the Clean Air Act.  Often, these "areas" are actually pollutants which
are emitted by stationary sources.   Source categories which emit these
pollutants were then evaluated and ranked by a process involving such
factors as:  (1) the level of emission control (if any) already required
by state regulations; (2) estimated levels of control that might be
required by standards of performance for the source category; (3) pro-
jections of growth and replacement of existing facilities for the source
category; and (4) the estimated incremental amount of air pollution that
could be prevented, in a preselected future year, by standards of per-
formance for the source category.  Sources for which new source performance
standards were promulgated or were under development during 1977 or earlier
were selected on these criteria.
     The Act amendments of August 1977 establish specific criteria to be
used in determining priorities for all source categories not yet listed
by EPA.  These are:   (1) the quantity of air pollutant emissions which
                                 2-5

-------
each such category will emit, or will  be designed to emit; (2) the extent
to which each pollutant may reasonably be anticipated to endanger public
health or welfare; and (3) the mobility and competitive nature of each
category of sources and the consequent need for nationally applicable new
source standards of performance.
     In some cases, it may not be feasible to immediately develop a
standard for a source category with a high priority.  This might happen
when a program of research is needed to develop control techniques or
because techniques for sampling and measuring emissions may require refine-
ment.  In the development of standards, differences in the time required
to complete the necessary investigation of different source categories
must also be considered.  For example, substantially more time may be
necessary if numerous pollutants must be investigated from a single source
category.  Further, even late in the development process the schedule for
completion of a standard may change.  For example, inability to obtain
emission data from well-controlled sources in time to pursue the develop-
ment process in a systematic fashion may force a change in scheduling.
Nevertheless, priority ranking is, and will continue to be, used to
establish the order in which projects are initiated and resources
assigned.
     After the source category has been chosen, determining the types of
facilities within the source category to which the standard will apply
must be decided.  A source category may have several facilities that
cause air pollution and emissions from some of these facilities may be
insignificant or very expensive to control.  Economic studies of the
source category and of applicable control technology may show that air
pollution control is better  served by applying standards to the more
severe pollution sources.  For this reason, and because there is no
adequately demonstrated system for controlling emissions from certain
facilities, standards often  do not apply to all air pollutants emitted.
Thus, although a source cateogry may be selected to be covered by a
standard of performance, not all pollutants or facilities within that
source category may be  covered  by the standards.
                                  2-6

-------
2.3  PROCEDURE FOR DEVELOPMENT  OF  STANDARDS OF PERFORMANCE
     Standards of performance must:   (1)  realistically reflect the best
demonstrated control  practice;  (2) adequately consider the cost, and the
non-air quality health and environmental  impacts  and energy requirements
of such control; (3)  be applicable to existing sources that are modified
or reconstructed as well  as new installations; and (4) meet these condi-
tions for all variations  of operating conditions  being considered anywhere
in the country.
     The objective of a program for  development of standards is to identify
the best technological system of continuous emission reduction which has "
been adequately demonstrated.   The legislative history of Section 111 on
what has been adequately demonstrated is  not limited to systems that are
in actual routine use.  The search may include a  technical assessment of
control systems which have been adequately demonstrated but for which
there is limited operational experience.   In most cases, determination of
the "...degree of emission reduction achievable..." is based on results
of tests of emissions from well controlled existing sources.  At times,
this has required the investigation  and measurement of emissions from
control systems found in other  industrialized countries that have developed
more effective systems of control  than those available in the United
States.
     Since the best demonstrated systems of emission reduction may not be
in widespread use, the data base upon which standards are developed may be
somewhat limited.  Test data on existing well-controlled sources are
obvious starting points in developing emission limits for new sources.
However, since the control of  existing sources generally represents
retrofit technology or was originally designed to meet an existing state
or local regulation, new sources may be able to meet more stringent
emission standards.  Accordingly, other infomation must be considered
before a judgment can be made  as to  the level at which the emission
standard should be set.
     A process for the development of a standard has evolved which
considers the  following:
                                  2-7

-------
     1.    Emissions  from  existing well-controlled sources as measured.
     2.    Data  on  emissions from such sources are assessed with consi-
          deration of  such factors as:   (a) how representative is the
          tested source with regard to feedstock, operation, size, age,
          etc.; (b)  age and maintenance  of the control equipment tested;
          (c) design uncertainties of control equipment being considered;
          and (d)  the  degree of uncertainty that new sources will be able
          to achieve similar levels of control.
     3.    Information  from pilot and prototype installations, guarantees
          by vendors of control equipment, unconstructed but contracted
          projects,  foreign technology,  and published  literature are also
          considered during the standard development process.  This is
          especially important for sources where "emerging" technology
          appears  to be a significant alternative.
     4.    Where possible, standards are  developed which permit the use
          of more  than one control technique  or licensed process.
     5.    Where possible, standards are  developed to encourage or permit
          the use  of process modifications or new processes as a method
          of control rather than  "add-on" systems of air pollution control.
     6.    In appropriate  cases, standards are developed to permit the
          use of systems  capable of controlling more than one pollutant.
     7.    Where appropriate,  standards for visible  emissions are developed
          in conjunction  with  concentration/mass emission standards.  The
          opacity standard  is  established at  a  level that will require
          proper operation  and maintenance of the emission control system
          installed to meet the concentration/mass  standard on a day-to-day
          basis.   In some cases,  however, it  is not possible  to develop
          concentration/mass  standards,  such  as with fugitive  sources of
          emissions.  In  these cases,  only opacity  standards may be
          developed to limit  emissions.
2.4  CONSIDERATION OF  COSTS
     Section 317 of the  Act requires,  among  other  things,  an  economic
impact assessment with respect to  any  standard  of  performance  established
                                  2-8

-------
under Section 111  of the Act.   The assessment  is  required to contain an
analysis of:
     1.    The costs of compliance with the regulation and standard
          including the extent to which the cost  of compliance varies
          depending on the effective date of the  standard or regulation
          and the  development  of less expensive or more efficient methods
          of compliance;
     2.    The potential inflationary or recessionary effects of the
          standard or regulation;
     3.    The effects of the standard or regulation on competition
        .  among small businesses;
     4.    The effects of the standard or regulation on consumer cost,
          and
     5.    The effects of the standard or regulation on energy use.
     Section 317 requires that the economic impact assessment be as
extensive as practical, taking into account the time and resources
available to EPA.
     The economic  impact of a  proposed standard upon an industry is
usually addressed  both in absolute terms and by comparison of the costs of
typical  existing state control regulations with the control costs that
would be incurred  as a result  of complying with the standard.  An incre-
mental approach is taken since both new and existing plants would be
required to comply with state  regulations in the absence of a federal
standard of performance.  This approach requires  a detailed analysis of
the impact upon industry resulting from the cost differential that exists
between a standard of performance and the typical state standard.
     The costs of controlling  air pollutants are not the only costs con-
sidered in analyzing the economic impacts of the proposed standard.  The
costs associated with the control of water pollutants and solid wastes
are also analyzed  wherever possible.
     A thorough study of the profitability and price-setting mechanism of
the industry is essential to the analysis so that an accurate estimate of
potential adverse economic impacts can be made.  It is also essential to
know the capital requirements  placed on plants in the absence of federal
standards of performance so that the additional capital requirements
                                  2-9

-------
necessitated by these standards can be placed in the proper perspective.
Finally, it is necessary to recognize any constraints on capital  availa-
bility with an industry, as this factor also influences the ability of
new plants to generate the capital required for installation of additional
control equipment needed to meet the standards of performance.
2.5  CONSIDERATION OF ENVIRONMENTAL IMPACTS
     Section 102(2)(C) of the National Environmental Policy Act (NEPA) of
1969 required federal agencies to prepare detailed environmental  impact
statements on proposals for legislation and other major federal actions
significantly affecting the quality of the human environment.  The
objective of NEPA is to build into the decision-making process  of federal
agencies a careful consideration of all environmental aspects of proposed
actions.
     In a number of legal challanges to standards of performance for
various industries, the Federal Court of Appeals have held that environ-
mental impact statements need not be prepared by the Agency for proposed
actions under Section 111 of the Clean Air Act.  Essentially, the Federal
Court of Appeals have determined that "...the best system of emission
reduction,...require(s) the Administrator to taken into account counter-
productive environmental effects of a proposed standard, as well  as
economic costs to the industry..."  On this basis, therefore, the Courts
"...established a narrow exemption from NEPA for EPA determination under
Section 111."
     In addition to these judicial determinations, the Energy Supply and
Environmental Coordination Act (ESECA) of 1974 (PL-93-319) specifically
exempted proposed actions under the Clean Air Act from NEPA requirements.
According to Section 7(c)(l),  "No action taken under the Clean Air Act
shall be deemed a major federal action significantly affecting the quality
of  the human environment within the meaning of the National Environmental
Policy Act of 1969."
     The Agency has concluded, however, that the preparation of environ-
mental impact statements could have beneficial effects on certain
regulatory actions.  Consequently, while not legally required to do so by
Section 102(2)(C) of NEPA, environmental impact statements will be
                                 2-10

-------
prepared for various regulatory actions,  including standards of performance
developed under Section 111  of the Act.   This voluntary preparation of
environmental impact statements, however, in no way legally subjects the
Agency to NEPA requirements.
     To implement this policy, a separate section is included in this
document which is devoted solely to an analysis of the potential environ-
mental impacts associated with the proposed standards.  Both adverse and
beneficial impacts in such areas as air and water pollution, increased
solid waste disposal, and increased energy consumption are identified and
discussed.
2.6  IMPACT ON EXISTING SOURCES
     Section 111  of the Act defines a new source as "...any stationary
source, the construction or modification of which is commenced..." after
the proposed standards are published.  An existing source becomes a new
source if the source is modified or reconstructed.  Both modification and
reconstruction are defined in amendments to the general provisions of
Subpart A of 40 CFR Part 60 which were promulgated in the Federal Register
on December 16, 1975 (40 FR 58416).  Any physical or operational change
to an existing facility which results in an increase in the emission rate
of any pollutant for which a standard applies is considered a modification.
Reconstruction, on the other hand, means the replacement of components of
an existing facility to the extent that the fixed capital cost exceeds 50
percent of the cost of constructing a comparable entirely new source and
that it be technically and economically feasible to meet the applicable
standards.  In such cases, reconstruction is equivalent to new construc-
tion.
     Promulgation of a standard of performance requires states to establish
standards of performance for existing sources in the same industry under
Section lll(d) of the Act if the standard for new sources limits emissions
of a designated pollutant (i.e., a pollutant; for which air quality criteria
have not been issued under Section 108 or which has not been listed as a
hazardous pollutant under Section 112).  If a state does not act, EPA
must establish such standards.  General provisions outlining procedures
for control of existing sources under Section lll(d) were promulgated on
November 17, 1975, as Subpart B of 40 CFR Part 60 140 FR 53340).

                                 2-11

-------
2.7  REVISION OF STANDARDS THROUGH EXPERIENCE
     Congress was aware that the level  of air pollution control  achievable
by any industry may improve with technological advances.  Accordingly,
Section 111 of the Act provides that the Administrator "...shall, at
least every four years, review and, if appropriate, revise..." the
standards.  Revisions are made to assure that the standards continue to
reflect the best systems that become available in the future.   Such
revisions will not be retroactive but will apply to stationary sources
constructed or modified after the proposal of the revised standards.
                                 2-12

-------
               3.   DESCRIPTION  OF  FUGITIVE  EMISSION SOURCES

3.1  INTRODUCTION  AND GENERAL  INDUSTRY  INFORMATION
3.1.1   Introduction
     The primary purposes  of this  chapter are  to define the synthetic
organic chemical manufacturing  industry (SOCMI)  and describe the potential
fugitive emission  sources  that  are typically found  in this industry.   Where
possible, the leak rates of uncontrolled emissions  from the various poten-
tial  fugitive emission sources  are quantified.   Industrial practices  and
state or local  regulations, that currently reduce fugitive emissions from
the SOCMI are also briefly discussed in this chapter.
3.1.2  General  Information
     Organic chemicals are manufactured in  a multi-leveled system of
chemical processes that is based on about ten  feedstock chemicals which
are principally produced  in petroleum refineries.  These feedstocks then
proceed through one or more of  the process  levels and result in literally
thousands of intermediate  or finished chemicals  (see Figure 3-1).
Generally, each process level  contains  more chemicals than the preceding
level; the plants manufacturing the products are smaller than the plants
supplying the feedstock;  and the volatilities  of the products are lower
than the volatilities of the feedstocks.  Because of the number and
diverse nature of the organic  chemicals included in the multi-leveled
system, the synthetic organic  chemical  manufacturing industry (SOCMI)
is defined, for this study, to consist of 378 of the higher volume, higher
volatility intermediate and finished products.   A list of the 378
chemicals is presented in  Appendix F.
                                   3-1

-------
                         RAW MATERIALS
               (CRUDE OIL. CRUDE NATURAL GAS. ETC)
                 REFINERIES
 CHEMICAL
FEEDSTOCK
  PLANTS
                                                 CHEMICAL
                                                FEEDSTOCKS
                                                 CHEMICAL
                                                  PLANTS
                                                 CHEMICAL
                                                 PRODUCTS
Figure 3-1.   General schematic of process levels that make up
             the organic chemical  industry.
                             3-2

-------
     Although there are organic chemical  manufacturing plants in most
industrialized areas of the country,  about 60 percent of the SOCMI  volume
is produced in Texas and Louisiana.   Each plant site may manufacture from
one to several organic chemicals using one or more processes.  Although
most processes result in one basic product, some produce a family of
chemicals.  Conversely, many chemicals are produced by more than one
process.  Yearly, production quantities at each plant can range from a
few million to several billion kilograms.
3.2  FUGITIVE EMISSION DEFINITION AND POTENTIAL SOURCE DESCRIPTION
3.2.1  Definition
     In this study, fugitive emissions in the SOCMI are considered to be
those volatile organic compound (VOC) emissions that result when process
fluid (either liquid or gaseous) leaks from plant equipment.  Those VOC
emissions resulting from the transfer, storage, treatment, and/or disposal
of process wastes will be covered by other standards.
3.2.2  Potential Source Characterization and Description
     There are many potential sources of fugitive emissions in a typical
synthetic organic chemical plant.  The following sources will be con-
sidered in this chapter:  pumps, compressors, in-line process valves,
pressure relief devices, open-ended  valves, sampling connections, flanges,
agitators, and cooling towers.  These potential sources are described below.
     3.2.2.1  Pumps.  Pumps are used extensively in the SOCMI for the
movement of organic liquids.1  The centrifugal pump is the most widely
used pump in the SOCMI; however, other types, such as the positive-
displacement, reciprocating and rotary action, and special canned and
diaphragm pumps, are also used in this industry.  Chemicals transferred
by pumps can leak at the point of contact between the moving shaft and
stationary casing.  Consequently, all pumps except the shaftless type
(canned-motor and diaphragm) require a seal at the point where the shaft
penetrates the housing in order to isolate the pump's interior from the
atmosphere.
     Two generic types of seals, packed and mechanical, are currently in
use on pumps in the SOCMI.  Packed seals can be used on both reciprocating
                                    3-3

-------
and rotary action types of pumps.  As Figure 3-2 shows, a packed seal
consists of a cavity ("stuffing box") in the pump casing filled with
special packing material that is compressed with a packing gland to form a
seal around the shaft.   Lubrication is required to prevent the buildup of
frictional heat between the seal and shaft.  The necessary lubrication is
provided by a lubricant that flows between the packing and the shaft.2
Deterioration of the packing will result in process liquid leaks.
                               box.
               FLUID
               E-kJO
1



_j

xixixjXTxixixr
^ SEAL FA
-teS
\

,/^C*\.**JQ
) 5HAfT
                          'tXIXIXIXJXiXiXI

                               /
                                               0
                                             ^^-
POiMBLE.
LC.AK.
             Figure 3-2.  Diagram of a simple packed seal.3

     Mechanical seals are limited in application to pumps with rotating
shafts and can be further categorized as single and double mechanical
seals.  There are many variations to the basic design of mechanical seals,
but all have a lapped seal face between a stationary element and a
rotating seal ring.  In a single mechanical seal application (Figure 3-3),
the rotating-seal ring and stationary element faces are lapped to a very
high degree of flatness to maintain contact throughout their entire
mutual surface area.  As with a packed seal, the seal faces must be
lubricated to remove frictional heat; however, because of its construc-
tion, much less lubricant is needed.
     A mechanical seal is not a leak-proof device.  Depending on the
condition and flatness of the seal faces, the leakage rate can be quite
low (as small as a drop per minute) and the'flow is often not visually
                                   3-4

-------
detectable.  In order to minimize fugitive emissions due to seal leakage,
an auxiliary sealing device such as packing can be employed.**
               PUMP
                                                        UA.VJD
                                                       RltOO,
                                                        KJSER.T  PXCKJUfi
   FCUID
   EXlO
                                                        S.TA.TIOUARV
                                                        EUE.MEJJT
POSil&LE.
L.C.A.K.
      Figure 3-3.  Diagram of a basic single mechanical seal. 5

     In a double mechanical seal application, two seals can be arranged
back-to-back or in tandem.  In the back-to-back arrangement (Figure 3-4),
the two seals provide a closed cavity between them.  A seal liquid, such
as water or seal oil, is circulated through the cavity.  Because the seal
liquid surrounds the double seal and lubricates both sets of seal faces
in this arrangement, the heat transfer and seal life characteristics are
much better than those of the single seal.  In order for the seal to
function, the seal liquid must be at a pressure greater than the operating
pressure of the stuffing box.  As a result some seal liquid will leak
across the seal faces.  Liquid leaking across the inboard face will enter
the stuffing box and mix with the process liquid.  Seal liquid going
across the outboard face will exit to the atmosphere.  Therefore, the seal
liquid must be compatible with the process liquid as well as with the
environment.6
     In a tandem double mechanical seal arrangement  (Figure 3-5), the
seals face the same direction.  The secondary seal provides a backup for
the primary seal.  A seal flush is used in the stuffing box to remove the
heat generated by friction.  The cavity between the two seals is filled
                                   3-5

-------
 Stuffing-box
  housing'^-,,.'
         • Pumped liquid

Inner mating ring-
Inner primary—
   ring
                                                                                ___-Gland plate
                                               Shaft
                                                                Outer mating ring .
                                                               T  Outer primary
                                                            .../_      ring
                 Figure  3-4.   Diagram  of a  double mechanical seal
                                 (back-to-back arrangement).7
Stuffing-box   Bypass
  housing     / flush
                                                Buffer liquid,
                                               out
                                               (top)
                                     in
                                  (bottom)
                   Pumped
                    liquid
Gland
plate
                                                              Outer
                                                     primary   mating
                                                      ring      ring
                                                            Shaft
                  Figure  3-5.
              Diagram of a  double mechanical seal
              (tandem arrangement).8
                                              3-6

-------
with a buffer or barrier  liquid.   However,  the  barrier  liquid  is  at  a
pressure lower than that  in  the  stuffing  box.   Therefore,  any  leakage
will be from the stuffing box  into the  seal  cavity  containing  the barrier
liquid.  Since this liquid is  routed  to a closed  reservoir,  process  liquid
that has leaked into the  seal  cavity  will also  be transferred  to  the
reservoir.  At the reservoir,  the process liquid  could  vaporize and  be
emitted to the atmosphere.  To ensure that VOC's  do not leak from the
reservoir, the reservoir  can be  vented  to a control device.9
     Another type of pump that has been used in the chemical  industry
is the shaftless pump which  includes  canned-motor and diaphragm pumps.
In canned-motor pumps (Figure  3-6) the  cavity housing the  motor rotor
and the pump casing are interconnected.  As a result, the  motor bearings
run in the process liquid and  all seals are eliminated. Because  the
process liquid is the bearing  lubricant,  abrasive solids cannot be
tolerated.  Canned-motor  pumps are being  widely used for handling
organic solvents, organic heat transfer liquids,  light  oils, as well as
many toxic or hazardous liquids, or where leakage is an economic
problem.10
               Figure 3-6.  Chempump canned-motor pump.11
                                  3-7

-------
     Diaphragm pumps (see Figure 3-7) perform similarly to piston and
plunger pumps.  However, the driving member is a flexible diaphragm
fabricated of metal, rubber, or plastic.   The primary advantage of this
arrangement is the elimination of all packing and seals exposed to the
process liquid.  This is an important asset when hazardous or toxic
liquids are handled.12
                                     Suction
       Figure 3-7.  Shriver mechanically actuated diaphragm pump.
                                                                 13
     3.2.2.2  Compressors.  Gas compressors used in the SOCMI are similar
to pumps in that they can be driven by rotary or reciprocating shafts.
Correspondingly, the same types of seals that are used on pumps are
used on compressors to  isolate the process gas from the atmosphere.
As with pumps, these seals are likely to be the source of fugitive
emissions from compressors.
     In addition to the mechanical seals that can be used on compressors,
centrifugal compressors can be equipped with liquid film seals (Figure
3-8).  This seal is formed by a film of oil between the rotating shaft
and stationary gland.   The seal oil exits the compressor from chambers
on both sides of the gland.  The oil leaving the chamber on the process
                                  3-8

-------
 side  is  under  pressure and contaminated with process gas.   When  the
 contaminated oil is returned to the oil reservoir, process  gas can be
 released and emitted to the atmosphere.11*  To eliminate the release  of
 VOC emissions  from the oil reservoir, the reservoir can be  vented to a
 control  device.
                                          OIL.  »M
     IKJTERWAL,
GAS  PRE.S5URE-
                   COM TAMI KJ ATE.D
                       OIL.  OUT
                   TO  R.£.SEJ=^VOIR
OIL. OUT
            Figure 3-8.  Liquid-film compressor shaft seal.15

      3.2.2.3   Process  Valves.  One of the most common pieces  of equipment
 in  organic chemical plants is the valve.  The types of  valves  commonly
 used  are control, globe, gate, plug, ball, relief, and  check  valves.  All
 except the relief valve  (to be discussed further below) and check  valve
 are activated  by a valve stem, which may have either a  rotational  or
 linear motion, depending on the specific design.  This  stem requires  a
 seal  to isolate the process fluid inside the valve from the atmosphere as
 illustrated by the diagram of a gate valve in Figure 3-9.  The possibility
 of  a  leak through this seal makes it a potential source of fugitive
                                  3-9

-------
emissions.  Since a check valve has no stem or subsequent packing gland,
it is not considered to be a potential source of fugitive emissions.
     Sealing of the stem to prevent leakage can be achieved by packing
inside a packing gland or 0-ring seals.  Valves that require the stem to
move in and out with or without rotation must utilize a packing gland.
Conventional packing glands are suited for a wide variety of packing
material; the most Common are various types of braided asbestos that
contain lubricants.  Other packing materials include graphite, graphite-
impregnated fibers, and tetrafluorethylene; the packing material used
depends on the valve application and configuration.16   These conventional
packing glands can be used over a wide range of operating temperatures.
At high pressures these glands must be quite tight to attain a good seal.17
             PACK-IMG  GUAMD
               VAUVE1
POiSlSUE.
(-2.AK. AFLEAS
                  Figure 3-9.  Diagram of a gate valve.18

     Elastomeric 0-rings are also used for sealing process valves.  These
0-rings provide good sealing but are not suitable where there is sliding
motion through the packing gland.  Those seals are rarely used in high
pressure service and operating temperatures are limited by the seal
material ,19
                                  3-10

-------
     Bellows seals  are more effective  for preventing process fluid leaks
than the conventional  packing  gland  or any other  gland-seal  arrangement.20
This type of seal  incorporates a  formed metal  bellows that makes  a barrier
between the disc and body bonnet  joint.  An example of this  seal  is
presented in Figure 3-10.   The bellows is the  weak point of the system
and service life can be quite  variable.  Consequently, this  type  of seal
is normally backed  up with a conventional packing gland and  is often fitted
with a leak detector in case of failure.21
               Disc
                  Bellows
                                              Body
                                                Bonnet
               Figure 3-10.  Example of bellows seals.
                                                      22
     A diaphragm may be used to isolate the working parts of the valve and
the environment from the process liquid.  Two types of valves which utilize
diaphragms are illustrated in Figures 3-11(a) and (b).  As Figure 3-11(b)
shows, the diaphragm may also be used to control the flow of the process
fluid.  In this design, a compressor component pushes the diaphragm toward
the valve bottom, throttling the flow.  The diaphragm and compressor are
connected in a manner so that it is impossible for them to be separated
under normal working conditions.  When the diaphragm reaches the valve
                                   3-11

-------
 bottom, it seats firmly against the bottom, forming a leak-proof seal.
 This configuration is recommended for fluids containing solid particles
 and for medium-pressure service.  Depending on the diaphragm material,
 this type of valve can be used at temperatures up to 205°C and in severe
 acid solutions.  If failure of the seal occurs, a valve employing a dia-
 phragm seal can become a source of fugitive emissions.23
Diaphragm-'
                                                Stem
Diaphragm
                 (a)
          Figure  3-11.   Diagrams  of  valves with  diaphragm  seals.21*

       3.2.2.4   Pressure Relief Devices.   Engineering  codes  require  that
  pressure-relieving  devices  or systems  be used  in  applications  where  the
  process  pressure may exceed the maximum  allowable working  pressure of the
  vessel.   The  most common type of pressure-relieving  device used  in the
  SOCMI  is the  pressure relief valve (Figure  3-12).  Typically,  relief valves
  are  spring-loaded and designed to  open when the process  pressure exceeds a
  set  pressure, allowing the  release of  vapors or liquids  until  the  system
  pressure is  reduced to its  normal  operating level.  When the normal
                                     3-12

-------
pressure is re-attained,  the valve  reseats,  and a seal  is again formed.25
The seal is a disk on a seat, and the possibility of a  leak through this
seal makes the pressure relief valve a potential  source of VOC fugitive
emissions.  Two potential causes of leakage  from relief valves are:
"simmering or popping", a condition due to the system pressure being
close to the set pressure of the valve, and  improper reseating of the
valve after a relieving operation.26
     Rupture disks are also common in the SOCMI.  These disks are made of
a material that ruptures when a set pressure is exceeded, thus allowing
the system to depressurize.  The advantage of a rupture disk is that the
disk seals tightly and does not allow any VOC's to escape from the system
under normal operation.  However, when the disk does rupture, the system
depressurizes until atmospheric conditions are obtained; this could result
in an excessive loss of product or correspondingly an excessive release
of fugitive emissions.
              SEAT
                                             SPRI-UG
                                               DISK.
                                               WOZ.Z.UE
                              PROCESS  SIDE.

       Figure 3-12.  Diagram of a spring-loaded relief valve.27
                                 3-13

-------
     3.2.2.5  Cooling Towers.   Cooling towers (Figure 3-13)  are found
in most SOCMI plants.  The purpose of these towers is to cool  the
plant's process cooling waters which have been heated while  removing
heat from various process equipment (reactors, condensers, heat
exchangers).  This cooling process is achieved by evaporation  when  the
process cooling water and air are contacted.   Under normal operating
conditions, a cooling tower would not be considered a fugitive emission
source.  However, if a leak occurs in the process equipment  and if  this
equipment is operating at a pressure greater  than that of the  cooling
water, organic chemicals can leak into the water.  When the  process water
is recirculated to the cooling tower, these chemicals can be released to
the atmosphere.
               Figure 3-13.  Cooling tower (cross-flow).28

     3.2.2.6  Agitators.  Agitators are commonly used in the SOCMI  to
stir or blend chemicals,  Like pumps and compressors, agitators  may leak
organic chemicals at the point where the shaft penetrates the casing.
Consequently, seals are required to minimize fugitive emissions  from
agitators.  Four seal arrangements are commonly used with agitators; they
include:  compression packing (packed seal), mechanical  seals, hydraulic
                                  3-14

-------
seals, and lip seals.29  Packed seals for agitators are very similar in
design and application to the packed seals for pumps (Section 3.2.2.1).
     Although mechanical seals are more costly than the other three seal
arrangements, they offer a greatly reduced leakage rate to offset their
higher cost.   The maintenance frequency of mechanical seals is, also, one-
half to one-fourth that of packed seals.30  In fact, at pressures greater
than 1135.8 kPa (150 psig), the leakage rate and maintenance frequency are
so superior that the use of packed seals on agitators is rare.31  As with
packed seals, the mechanical seals for agitators are similar to the design
and application of mechanical seals for pumps (Section 3.2.2.1).
     The hydraulic seal (Figure 3-14) is the simplest and least used
agitator shaft-seal.  In this type of seal, an annular cup attached to the
process vessel contains a liquid that is in contact with an inverted cup
attached to the rotating agitator shaft.  The primary advantage of this
seal is that it is a non-contact seal.  However, this seal is limited to
low temperatures and pressures and can only handle very small pressure
fluctuations.  Organic chemicals may contaminate the seal liquid and then
be released into the atmosphere as fugitive emissions.32
                    Inverted cup.^

                Annular cup „  ——


                   a. Hydraulic seal
       Figure 3-14.  Diagram of hydraulic seal for agitators.33

     A lip seal (Figure 3-15) can be used on a top-entering agitator as a
dust or vapor seal.  The sealing element is a spring-loaded elastomer;
subsequently, lip seals are relatively inexpensive and easy to install.
Once the seal has been installed the agitator shaft rotates in continuous
                                  3-15

-------
contact with the lip seal.  Pressure limits  of the  seal  are  2  to  3  psi
because it operates without lubrication.   Operating temperatures  are
limited by the characteristics of the elastomer.   Fugitive VOC emissions
could be released through this seal  when  this  seal  wears excessively  or
the operating pressure surpasses the pressure  limits of the  seal.3k
                                              Lip Seal
             Figure 3-15.  Diagram of agitator lip seal.
                                                        35
     3.2.2.7  Open-Ended Valves or Lines.  Some valves are installed in a
system so that they function with the downstream line open to the atmos-
phere.  Examples are purge valves, drain valves, and vent valves.  A
faulty valve seat, or incompletely closed valve would result in leakage
through the valve and fugitive VOC emissions to the atmosphere.
     3.2.2.8  Sampling Connections.   The operation of a process unit is
checked periodically by routine analyses of feedstocks and products.   To
obtain representative samples for these analyses, sampling lines must
first be purged prior to sampling.  The purged liquid or vapor is
sometimes drained onto the ground or into a sewer drain, where it can
evaporate and release VOC emissions  to the atmosphere.
     3.2.2.9  Flanges.  Flanges are  bolted, gasket-sealed junctions used
wherever pipe or other equipment such as vessels, pumps, valves, and heat
exchangers may require isolation or  removal.  Normally, flanges are
employed for pipe diameters of 50 mm or greater and are classified by
pressure and face type.
                                  3-16

-------
     Flanges may become fugitive emission sources when leakage occurs due
to improperly chosen gaskets  or a poorly assembled flange.   The primary
cause of flange leakage is  due to thermal  stress that piping or flanges in
some services undergo;  this results in the deformation of the seal  between
the flange faces.36
3.3  BASELINE CONTROL
     There are presently no federal regulations that specifically reduce
emissions from synthetic organic chemical  manufacturing plants.  However,
some fugitive emission  reduction is achieved by operating practices
currently followed by industry and applicable state or local regulations.
Because these practices and regulations only "incidentally" control
fugitive emissions,  they are considered, in this study, to be the baseline
control level.  The  procedures, specific control techniques, and regula-
tions that make up the  baseline control level are discussed below.
     Fugitive emissions occurring under the baseline control level  are
subsequently considered in  this report to be uncontrolled emissions.   Data
characterizing the uncontrolled levels of fugitive emissions in the SOCMI
are presently unavailable.   However, data of this type have been obtained
for the refining industry.   These data are presented in Table 3-1.
Because the operation of the various process equipment in the SOCMI is not
expected to differ greatly  from the operation of the same equipment in
the refining industry,  it is felt that the refinery fugitive emission data
can be used to approximate  the levels of fugitive emissions in SOCMI.
3.3.1  Industrial Practices
     The organic chemical industry has been primarily interested in leaks
that are large enough to be physically evident (leaks that can be seen,
heard, or smelled);  such leaks are normally repaired to minimize the loss
of product and are,  consequently, termed "easily detectable leaks".  Fugi-
tive emissions, as they are considered in this report, are also the results
of leaks from process equipment but have considerablj^_small^r_emission
rates than "easily detectable leaks".  In the past, SOCMI has generally not
monitored equipment for fugitive emissions nor repaired equipment on the
basis of reducing the level of fugitive emissions.  Processes which have
emitted toxic or hazardous  compounds have been .exceptions to this rule.
                                   3-17

-------
 TABLE 3-1.   UNCONTROLLED FUGITIVE EMISSION  FACTORS  IN  THE  SYNTHETIC
             ORGANIC CHEMICAL MANUFACTURING  INDUSTRY (SOCMI)
        Fugitive omission  source
                              Uncontrolled emission
                                  factor.9 kg/hr
 Pumps
      Light  liquids
          With packed
          With
          With
          With
       seals
single mechanical
double mechanical
no seals
   d
     Heavy Liquids
          With packed seals
          With single mechanical
          With double mechanical
          With no seals

Valves  (in-line)

     Gas         .
     Light liquid^
     Heavy liquid

Safety/relief valves

     Gas         .
     Light liquid.
     Heavy liquid

Open-ended valves

     Gas         .
     Light liquidjj
     Heavy liquid

Flanges
Sampling connections
Compressors
Cooling towers
Agitators
seals
seals
                  seals
                  seals
0.12
0.12
0.12C
0.0


0.020
0.020.
0.020
0.0
                                        0.021
                                        0.010
                                        0.0003
                                        0.16
                                        0.006
                                        0.009
                                        0.025
                                        0.014
                                        0.003

                                        0.0003
                                        0.015
                                        0.44
                                        13.6-1107*
                                        NAf
aThese uncontrolled emission levels are based upon  the  refinery data  presented
 in reference 37.
bLight liquid is defined as a fluid with vapor pressure greater than  0.3
 kPa at 20°C.  This vapor pressure represents the split between kerosene and
 naphtha and is based on data presented in reference 37.   The averaqe vapor
 pressure between these components is approximately 0.04  psi  at 68 °F.
GAssumes the inner seal leaks at the same rate as single  seal and that the
 VOC is emitted from the seal oil degassing vent.
dHeavy liquid is defined as a fluid with vapor pressure less  than 0.3 kPa at
 20°C.  This vapor pressure represents the split between  kerosene and
 naphtha and is based on data presented in reference 37.   The average vapor
 pressure between these components is approximately 0.04  psi  at 68 °F.
eThese levels are based on cooling tower circulation rates that range from
 0.05-3.66 m /sec (714-58,000 GPM). /Ref.  38.
 NA = no data available.

                                    3-18

-------
     While SOCMI has been primarily concerned with easily detectable
 leaks, certain equipment and procedures used in many organic chemical
 plants may help to reduce fugitive VOC emissions.  For instance, some
 plants cap-off or use double block valves on the end of process lines;
 either of these procedures will reduce fugitive emissions.  In some plants,
 relief valves are checked to see if the valve has reseated properly after
          39
 relieving.   As previously mentioned,  an  improperly  seated relief  valve  may
 allow fugitive VOC emissions to occur.  Rupture discs, which are commonly
.used in the SOCMI, also prevent fugitive VOC emissions.  Some organic
 chemical plants employ closed-loop sampling which may help to reduce fugi-
 tive emissions.
     The flaring of vapors vented from various vessels or equipment is
 another technique which is used by some plants (particularly those produc-
 ing toxic or hazardous chemicals) that will reduce fugitive emissions.
 3.3.2  Existing Regulations
     There are, presently, two types of regulations  that impact fugitive
 VOC emissions from organic chemical plants.  The first type is to  regulate
 industrial operating practices on the  basis of worker health and safety.
 Because some aspects of these regulations deal with  worker exposure to
 process emissions, they may have some  impact on fugitive VOC emissions.
 The second type of regulations is regulations that were specifically de-
 veloped to limit fugitive emissions.
     3.3.2.1  Health and Safety Regulations.  Several regulations  have
 been established under the direction of the Occupational Safety and Health
 Administration and National Institute  for Occupational Safety and  Health
 to:  1) limit the level of process emissions and 2)  limit worker exposure
 to process emissions.  These regulations may result  in a reduction in the
 levels of fugitive VOC emissions.
      In  the  vinyl  chloride monomer  and benzene  industries,  the  safety
 and health  regulations  are designed  to limit  the  ambient  VOC  levels
 to which  workers may  be  exposed.  Since these  standards do  not  stipulate
 how the  allowable  ambient  levels  should be  achieved, workers  can  be
 protected from  high ambient VOC  levels by:  1)  a  reduction  in  the  fugitive
                                    3-19

-------
VOC emissions or 2) the use of special  equipment (such as personal
respirators) to isolate the worker from the emissions.  This example
illustrates that the present health and safety regulations do not
mandate a reduction in fugitive VOC emissions, and any reduction in
fugitive emissions resulting from these regulations can be considered
to be "incidental".  By contrast, fugitive emission regulations do
require the fugitive emissions to be reduced.             »
     3.3.2.2  Fugitive Emissions Regulations.   Currently, there are no
federal fugitive emission regulations for the SOCMI.  However, California
has established such regulations, and organic chemical plants in this
state must comply with the appropriate regulations.
     California presently requires open-ended process lines to be capped-off
in order to minimize fugitive VOC emissions.  This state also requires
relief valves to be vented to a flare system, monitored and maintained, or
a rupture disk to  be used.  In addition to these regulations, the South
Coast Air Quality  Management District requires organic chemical plants
to vent fugitive emissions from compressor seals to a fired-heater or
flare system.  The South Coast and Bay Area AQMD also require periodic
inspection of valves in the chemical and refining  industries.
                                   3-20

-------
3.4  REFERENCES

1.   Erikson, D.  G.,  and  V.  Kalcevic.   Emissions Control  Options  for  the
     Synthetic Organic  Chemicals  Manufacturing  Industry,  Fugitive Emis-
     sions Report,  Draft  Final.   Hydroscience,  Inc.,  1979.   p.  II-2.

2.   Ref.  1.

3.   Ref.  1,  p. II-3.

4.   Ramsden, J.  H.   How  to  Choose  and  Install  Mechanical  Seals.   Chem.
     E., 85J22):97-102.   1978.

5.   Ref.  1,  p. II-3.

6.   Ref.  4,  p. 99.

7.   Ref.  4,  p. 100.

8.   Ref.  4,  p. 101.

9.   Ref.  4,  p. 99.

10.  Perry, R. H.,  and  C. H.  Chilton.   Chemical  Engineers'  Handbook,  5th
     Ed.  New York,  McGraw-Hill  Book Company,  1973.   p.  6-8.

11.  Ref.  10, p.  6-12.

12.  Ref.  10, p.  6-13.

13.  Ref.  10, p.  6-13.

14.  Ref.  1,  p. II-7.

15.  Ref.  1 ,  p. II-8.

16.  Lyons, J. L.,  and  C. L.  Ashland, Jr.   Lyons'  Encyclopedia of Valves,
     New York, Van Nostrand Reinhold Co.,  1975.  290 p.

17.  Templeton, H.  C.  Valve Installation, Operation and Maintenance.
     Chem. E., 78^(23)141-149, 1971.

18.  Ref.  1, p. II-5.

19.  Ref.  17, p.  147-148.

20.  Ref.  17, p.  148.

21.  Ref.  17, p.  148.

22.  Ref. 17, p.  148.
                                  3-21

-------
23.  Pikulik, A.  Manually Operated Valves.   Chem.  E.,  April  3,  1979.
     p. 121.

24.  Ref. 23, p. 121.                                           -

25.  Steigerwald, B. J.  Emissions of Hydrocarbons  to the Atmosphere  from
     Seals on Pumps and Compressors.  Report No.  6, PB  216 582,  Joint
     District, Federal and State Project for the  Evaluation of Refinery
     Emissions.  Air Pollution Control District,  County of Los Angeles,
     California.  April 1958.   37 p.

26.  Ref. 1, p. II-7.

27.  Ref. 1, p. II-6.

28.  Cooling Tower Fundamentals and Application Principles.  Kansas City,
     Missouri, The Marley Company, 1969.  p. 4.

29.  Ramsey, W. D., and G. C.  Zoller.  How the Design of Shafts, Seals
     and Impellers Affects Agitator Performance.   Chem. E., 83(18):
     101-108.  1976.

30.  Ref. 29, p. 105.

31.  Ref. 29, p. 105.

32.  Ref. 29, p. 105.

33.  Ref. 29, p. 106.

34.  Ref. 29, p. 106.

35.  Ref. 29, p. 106.

36.  McFarland, I.  Preventing Flange Fires.  Chem. E.  Prog., 65(8):
     59-61.  1969.

37.  Wetherold, R. G., et al.   Emission Factors and Frequency of Leak
     Occurrence for Fittings in Refinery Process  Units, interim report.
     EPA Contract No.  68-02-2665.  Austin, Texas, Radian Corporation,
     February 1979.   p. 22.

38.  Radian Corporation.  The  Assessment of Environmental Emissions  From
     Oil Refining.  Draft Report, Appendix B.  EPA Contract No.  68-02-2147,
     Exhibit B.  Austin, Texas.  August, 1979.

39.  Letter with Attachments from J. M. Johnson,  Exxon Company, U.S.A.,
     to.Robert T. Walsh, U.S.  EPA.  July 28, 1977.
                                   3-22

-------
                  4.   EMISSION  CONTROL  TECHNIQUES

     Sources of fugitive VOC  emissions  from  SOCMI  plants  were  identified
in Chapter 3 of this  document.   The potential  emission  control  techniques
that can be applied to SOCMI  fugitive emission sources  are  discussed  in
this chapter.   The applicability and estimated control  effectiveness  of
each technique are also presented.   The quantitative  control effective-
ness for many of the control  techniques is  not known.   Qualitative
discussions of effectiveness  and references  to technology, transfer  from
similar industries are presented wherever applicable.
4.1  LEAK DETECTION AND REPAIR  METHODS
     Leak detection and repair  methods  can  be  applied in  order to reduce
fugitive emissions from any source. Leak detection methods are used  to
identify equipment components that  are  emitting significant amounts of
VOC.  Emissions from leaking  sources may be  reduced by  three general
methods:  repair, modification, or  replacement of  the source.
4.1.1  Leak Detection Methods
     Leak detection methods include individual component  surveys, area
(walk-through) surveys, and fixed point monitors.   They are described in
this order since the first method is also included as part  of  the other
methods.
     4.1.1.1  Individual Component  Survey.   Each fugitive emission  source
(pump, valve, compressor, etc.) is  checked  for VOC leakage  in  an individ-
ual component survey.  The source may  be checked for  leakage by visual,
audible, olfactory, soap bubble, or instrument techniques.   Visual  methods
are good for locating liquid leaks, especially pump seal  failures.
Observation of a visible leak does  not  necessarily indicate VOC emissions,
since the leak may be composed of non-VOC compounds.   High  pressure leaks
                                 4-1

-------
may be detected by hearing the escaping vapors, and leaks  of odorous
materials may be detected by smelling the odor.  Current industry
practices include leak detection by visual, audible, and olfactory
methods.  However, in many instances, even very large VOC leaks are not
detected by these methods.
     Spraying soap bubbles on equipment components is another individual
survey method.  If the soap bubbles expand or are blown away, it is an
indication that something is leaking from the component.  A disadvantage
of this method is that it does not distinguish leaks of non-VOC compounds
from VOC leaks.  Consequently, air or steam leaks would produce the same
observed effect as VOC leaks.  This method is only semtqiiantitattve since  it
requires that the observer subjectively determine the rate of leakage
based on behavior of the soap bubbles.  This method is limited to "cool"
sources, since temperatures above 100°C would cause the water in the soap
solution to boil away.  This method is also not suited for moving shafts  on
pumps or compressors, since the motion of the shaft may interfere with  the
motion of the bubbles caused by a leak.
     Portable hydrocarbon detection instruments are the best method for
identifying leaks of VOC from equipment components.  The instrument is
used to sample and analyze the air in close proximity to the potential
leak surface by traversing the sampling probe tip over the entire area
where leaks may occur.  This sampling traverse is called "monitoring"  in
subsequent descriptions.  The hydrocarbon concentration of the sampled air
is displayed  on  the  instrument meter.  The  performance  criteria for moni-
toring  instruments and a  description  of  instrument  survey methods are
included  in Appendix  D.   The  hydrocarbon concentration  observed during
monitoring of a  component is  proportional to the  VOC emission rate from
the  component.   Data  from petroleum  refineries have  been used to develop
relationships  between monitoring  concentration and  mass emission rates.
The  hydrocarbon  concentration which  defines a  component needing mainte-
nance must be chosen. Components  which  have  indicated  concentrations
higher  than  this "action  level"  are  marked  for repair.  Data  from
petroleum refineries indicate that large variations in  mass  emission rate
                                 4-2

-------
may occur over short time  periods  for  an  individual  equipment  component.
More frequent monitoring intervals  tend to  reduce  the  chance of missing
"large leaks" because of their  variable leak  rates.
     4.1.1.2  Area Survey.   An  area survey  (also known as a walk-through
survey) requires the use of a portable hydrocarbon detector and a  strip
chart recorder.  The procedure  involves carrying the instrument within one
meter of the upwind and downwind  sides of process  equipment and associated
fugitive emission sources.   An  increase in  observed  concentration  indi-
cates leaking fugitive emission sources.  The instrument is then used  for
an individual component survey  in  the  suspected leak area.   The efficiency
of this method for locating leaks  is not  well  established.   It has been
estimated that the walk-through survey combined with selected  individual
surveys will detect about 50 percent of the number of leaks identified in  a
complete individual  survey.   The  time and  labor requirements  for  the
walk-through are much lower.  This  method will  not detect leaks from
sources such as elevated valves or relief valves.   Leaks from  adjacent
units and adverse meteorological  conditions can also interfere with the
walk-through survey.   Consequently, the walk-through survey is best for
locating only  large leaks with  a  small resource expenditure.
     4.1.1.3  Fixed Point Monitors. This method consists of placing
several automatic hydrocarbon sampling and  analysis  instruments at
various locations in the process  unit. The instruments may sample the
ambient air intermittently or continuously.  Elevated hydrocarbon  concen-
trations indicate a leaking component. As  in the  walk-through method, an
individual  component survey is  required to  identify the specific leaking
component in the area.  For this  method,  the  portable hydrocarbon  detec-
tor is also required.  Leaks from adjacent  units and adverse meteorologi-
cal conditions may interfere with the  method.  The efficiency  of this
method is not well established, but it has  been estimated that 33  percent
of the number of leaks identified by a complete individual  component
                                                2
survey could be located by fixed-point monitors.   Fixed-point monitors
are more expensive, multiple units may be required,  and the portable
instrument is also required to  locate  the specific leaking component.
                                4-3

-------
Calibration and maintenance costs may be higher.  Fixed-point monitors have
been used to detect emissions of hazardous or toxic substances (such as vinyl
chloride) as well as potentially explosive conditions.  Fixed-point monitors
have an advantage for these cases, since a particular compound can be  selected
as the sampling criterion.
4.1.2  Repair Methods
     The following descriptions of repair methods include only those
features of each fugitive emission source (pump, valve, etc.) which need
to be considered in assessing the applicability and effectiveness of each
method.  They are not intended to be complete repair procedures.  The
effectiveness of repairs in reducing fugitive emissions has not been
well documented; however, data for valve repairs have been collected in
various petroleum refineries.  In many cases, perfect repair will not be
achieved, but whenever repairs are performed, the portable hydrocarbon
detector should be used to identify the lowest achievable emission rate.
     4.1.2.1  Pumps.  Many pumps have spares which can be operated while
the  leaking  pump  is  being  repaired.  Leaks from packed seals may  be reduced
by  tightening the  packing  gland.  At some point, the  packing may  deteriorate
to  the  point where  further tightening would  have no effect or  possibly even
increase  fugitive  emissions  from  the seal.   The packing can  be replaced with
the  pump  out of  service.   When mechanical seals are utilized,  the pump must
be  dismantled so  the leaking  seal can be  repaired or  replaced.   Dismantling
pumps  will  result  in spillage of  some process  fluid and evaporate emissions  of
VOC.   These  temporary emissions may  be  greater  than the continued leak from  the
seal,  if the seal  leak is  small.
      4.1.2.2  Compressors.   Leaks from  packed  seals may be reduced by  the
same repair  procedure that was  described  for pumps.   Other types  of seals
require that the  compressor  be  out of service  for repair.  Since  most  compressors
do  not  have  spares,  repair or replacement of the seal would  require a  shut-
down of the  process.   Temporary emissions resulting from  a shutdown may be
greater than the  emissions from the  seal  if  it  was allowed to  leak until the
next scheduled  shutdown.
                                  4-4

-------
     4.1.2.3  Relief Valves.   In  general,  relief valves  which leak must
be removed in order to repair  the leak.   In  some cases of improper reseat-
ing, manual  release of the valve  may improve the seat seal.   In order to
remove the relief valve without shutting down the process,  a block valve
is required  upstream of the relief valve.  A spare relief valve should be
attached while the faulty valve is repaired  and tested.   After a relief
valve has been repaired and replaced,  there  is no guarantee that the
next over-pressure relief will not result in another leak.
     4.1.2.4  Valves.   Most valves have  a  packing gland  which can be
tightened while in service. Although  this procedure should decrease the
emissions from the valve, in some cases  it may actually  increase the
emission rate if the packing is old and  brittle or has  been overtightened.
Plug type valves can be lubricated with  grease to reduce emissions around
the plug.  Some types of valves have no  means of in-service repair and
must be isolated from the process and removed for repair or replacement.
Other valves, such as control  valves,  may be excluded from in-service
repair by operating or safety  procedures.   In many cases, valves cannot
be isolated from the process for  removal.  Most control  valves have a
manual bypass loop which allows them to  be isolated and  removed.  Most
block valves cannot be isolated easily although temporary changes in
process operation may allow isolation  in some cases.  If a process unit
must be shut down in order to  isolate a  leaking valve,  the emissions
resulting from the shutdown will  probably be greater than the emissions
from the valve if allowed to leak until  the  next process change which
permits  isolation for repair.
     Depending on site specific  factors, it  may be possible to repair process
valves by injection of a sealing  fluid into  the source.   This type of repair
may affect the operability of the valve  such that replacement of the source
might be necessary within a short time after its repair, and the emissions
that could result due to the replacement of  the source should be evaluated
when considering this type of repair.   It should be noted that injection of
sealing  fluid  has  been  successfully used to  repair  leaks from  valves  in.
petroleum refineries  in  California.
                                    4-5

-------
     4.1.2.5  Flanges.  In some cases, leaks from flanges can be reduced
by replacing the flange gaskets.  Most flanges cannot be isolated to
permit replacement of the gasket.  Data from petroleum refineries show
that flanges emit very small amounts of VOC.
4.1.3  Control Effectiveness of Leak Detection and Repair Methods
     The instrument survey of individual  components is the only type of leak
detection method for which control effectiveness has been quantified.   Therefore,
the following estimations of control effectiveness do not pertain to the
soap bubble leak detection method, area surveys, or fixed-point monitoring
methods.
     There are several factors which determine the control  effectiveness of
individual component surveys; these include

          Action level or leak definition,
          Inspection interval or monitoring frequency,
          Achievable emission reduction of maintenance, and
          Interval  between detection and  repair of the leak.
Some of these factors can be estimated by using data collected from
                     *t
petroleum refineries.
     4.1.3.1  Action Level.  The action level  is the minimum hydrocarbon
concentration observed during monitoring  which defines a leaking component
which requires repair.  The choice of the action level for defining a
leak is influenced by a number of important considerations.  First, the
percent of total mass emissions which can potentially be controlled by
the monitoring and repair program can be  affected by varying the leak
definition, or action level.  Table 4-1 gives the percent of total mass
emissions affected by various action levels for a number of equipment
types.  The data in this  table, indicate that, in general, a low action
level results in larger potential emission reductions.  However, the
choice of an appropriate  leak definition  is most importantly limited by
the ability to repair leaking components.  Test data indicate that about
50 percent  of valve leaks with  initial screening values equal to or greater
  than  10,000  ppmv can be  successfully repaired.  Similar data indicate
  that attempted  repair of valve leaks with  initial screening values

                                      4-6

-------
TABLE 4-1.   FRACTION OF TOTAL MASS  EMISSIONS FROM VARIOUS SOURCE TYPES THAT
            WOULD BE.CONTROLLED BY  DIFFERENT ACTION LEVELS
 Action  level9  (ppmv)
                               Fraction of mass  emissions  (as
100.000
50.000
10.000
1.000
 Source  type
   Pump  seals
     Light  liquid  service
     Heavy  liquid  service
    56
     0
   68
    0
   87
   21
  97
  66
   In-line  valves
     Vapor  service              85
     Light  liquid service       49
     Heavy  liquid service        0
                 92
                 62
                  0
               98
               84
                0
                99
                96
                23
 Safety/relief valves
 Compressor  seals
 Flanges
    20
    28
   33

   48

    0
   69

   84

    0
  92

  98

  48
  Level  of emission at which repair of the source is  required.
  These  data show the fraction  of the total  emissions from a given source
  type that is attributable  to  sources with  leaks above the various action
  levels.
                                  4-7

-------
  of  less  than  10,000  ppmv  can  increase instead of decrease emissions
  from  these  values.   From  these data it is concluded that repairing leaks
  with  screening  values  in  the  1,000-10,000 ppmv range may not result in a
  net reduction in  mass  emissions.   The nature of repair techniques for
  pipeline valves,  for instance, are such that          to repair leaks below
  a certain level by tightening valve packing may actually result in an
  increase in emissions.   In  practice, valve packing material becomes hard
  and brittle after extended  use.  As the packing loses its resiliency, the
  valve packing gland  must  be tightened to prevent loss of product to the
  atmosphere.   Excessive tightening, however, may cause cracks in the packing,
  thus  increasing the  leak  rate.  Unbalanced tightening of the packing gland
  may also cause  the packing  material to be positioned improperly in the valve
  and allow leakage.   Valves  which are not often used can build up a "static"
  seal  of  paint or  hardened lubricant which could be broken by tightening
  the packing gland.   Therefore, it may be important not to cause small
  leaks to become large  leaks by requiring tightening of valves to meet a
  very  low leak repair action level.
     4.1.3.2  Inspection Interval.  A monitoring plan may include annual,
quarterly, monthly, or even weekly  inspections.  The length of time
between inspections should depend  on the  expected occurrence  and
recurrence of leaks after a piece  of equipment  has been checked and/or
repaired.   This interval can be related  to the  type of equipment and
service conditions, and different  intervals can be specified  for different
pieces  of equipment after appropriate  equipment histories have been
developed.  In  the refinery VOC leak Control  Techniques Guideline  (CTG)
document,7 the  recommended monitoring  intervals are:  annual—pump seals,
pipeline valves in liquid service,  and  process  drains; quarterly--
compressor seals, pipeline valves  in gas  service, and pressure relief
valves  in gas service; weekly—visual  inspection of pump seals; and no
individual monitoring—pipeline flanges  and other connections, and
pressure relief valves in liquid service.  The  choice of the  interval
affects the emission reduction  achievable since more  frequent inspection
will  result in leaking sources  being  found and  fixed  sooner.  In order
to evaluate the effectiveness of different  inspection  intervals, it  is
                                      4-8

-------
necessary to estimate  the  rate at which  new  leaks will occur  and  repaired
leaks will  recur.   The estimates which have  been used to  evaluate yearly,
quarterly,  and monthly inspections  are shown in Table 4-2.
     4.1.3.3  Allowable Interval Before  Repair.  If  a leak  is  detected,
the equipment should be repaired within  a  certain time period.  The
allowable repair time  should  reflect  an  interest in  eliminating a source
of VOC emissions but should also allow the plant operator sufficient  time
to obtain necessary repair parts and  maintain some degree of  flexibility
in overall  plant maintenance  scheduling.   The determination of this
allowable repair time  will affect emission reductions by  influencing  the
length of time that leaking sources are  allowed to continue to emit
pollutants.  Some of the components with concentrations  in  excess of the
leak definition action level  may not be  able to be  repaired until the
next scheduled  unit shutdown, e.g., a unit turnaround.
     The effects of different allowable repair intervals are shown in
Table 4-3.  The percentages  shown  in the table are the percent of emis-
sions from  the  component which will be affected by the repair.  The
emissions which occur  between the  time the leak is  detected and repair is
attempted are  increased with increasing allowable repair intervals.
     4.1.3.4  Achievable Emission  Reduction.  Repair of leaking components
will not always result in complete emission  reduction.  The repair of
components  which have  initial monitoring levels below 1,000 ppm  has  not
been adequately demonstrated.  Repair of those components with low initial
leak rates, mav  actually result in  an emission rate increase.  In order to
 estimate repair effectiveness, it  was assumed that  emission  would be
 reduced to  a  level equal  to  components  with screening values  of  1,000
 ppm.   The  average  emission rates of  components above 10,000  ppm  and  at
 1,000 ppm  are shown in Table 4-4.
      4.1.3.5   Development of Controlled Emission Factors.  The uncon-
trolled emission levels for  the emission  sources that are  typically  found
in the  model  plants were  previously  presented in Chapter 3 (Table 3-1).
Controlled  VOC  emission levels can be calculated by a "controlled emission"
factor.  This  factor  can  be  developed for each type of emission  source by
using the general  expression:
                                     4-9

-------
                TABLE 4-2.   ESTIMATED OCCURRENCE  AND  RECURRENCE RATE  FOR VARIOUS MONITORING INTERVALS
 I

o
Estimated perci
of sources leal
at above 10,000
Source type initially3
Pump seals
Light liquid service 23
Heavy liquid service 2
In-line valves
Vapor service 10
... Light liquid service 12
Heavy liquid service 0
Estimated percent of
*nt initial leaks which
^in9 are found leaking at
PP™ subsequent inspections'3
Annual

20
20

20
-.20
20
Quarterly

10
10

10
10
10
Monthly

5
5

5
.5 .
5
Estimated. percent of
sources which are
found leaking at
subsequent inspections c
Annual

4.6
0.4

2.0
2.4
0.0
Quarterly

2.3
0.2

1.0
' 1.2
0.0
Monthly

1.2
0.1

0.5
. 0.6
0.0
        Safety/relief valves
 8
20
10
1.6
0.8
0.4
         Compressor seals
33
20
10
6.3
3.3
1.7
         Flanges
               20
          10
                  0.0
          0.0
          0.0
        ^Approximate  fraction  of  sources having leaks  equal to or greater than 10,000 ppm prior to repair.
         Approximate  fraction  of  leaking sources that were repaired but found to leak during subsequent
         inspections.   These approximations are based on engineering judgment.
        Approximate  fraction  of  sources that were repaired but found to leak during a subsequent inspection.
         These  approximations  are the  product of the information presented in footnotes a and b.

-------
TABLE 4-3.   PERCENT OF  MASS  EMISSIONS AFFECTED  BY  VARIOUS  REPAIR  INTERVALS
Allowable repair interval  (days)
 30
15
Percent of emissions affected
95.9     97.9     99.3     99.9
           TABLE 4-4.  AVERAGE EMISSION RATES FROM SOURCES
                       ABOVE 10,000 PPMV AND AT 1000 PPMV9
Source type
Pump seals
Light liquid service
Heavy liquid service
In-line valves
Vapor service
Light liquid service
Heavy liquid service
Safety/relief valves
Compressor seals
Flanges
(Y)
Emission rate
from sources above
10,000 ppmva
(kg/hr)

0.45
0.21

0.21
0.07
0.005
1.4
l.'l
0.003
Emission rate
from sources .at
1000 ppmvD
(kg/hr)

0.035
0.035

0.001
0.004
0.004
0.035
0.035
0.002
/Xx
Percentage
reduction

92.0
83.0

99.5
94.0
20.0
97.5
97.0
33.0
a. . . ,, .... , , .
  screening values above 10,000 ppmv.

 'Emission rate  of all  sources, within a source type,.having screening
  values  of 1000 ppmv.
                                 4-11

-------
     Controlled emission factor = Uncontrolled factor - uncontrolled
                                  factor x emission reduction efficiency
The reduction efficiency can be developed by the following expression and
correction factors:
                Reduction efficiency =AxBxCxD
Where:
     A = Theoretical Maximum Control Efficiency = fraction of total  mass
         emissions for each source type with VOC concentrations greater
         than the action level (Table 4-1, Figure 4-1).
     B = Leak Occurrence and Recurrence Correction Factor = correction
         factor to account for sources which start to leak between
         inspections (occurrence) and for sources which are found to
         be leaking, are repaired and start to leak again before the
         next inspection (recurrence) (Table 4-2, 4-6).
     C = Non-Instantaneous Repair Correction Factor = correction factor
         to account for emissions which occur between detection of a leak
         and subsequent repair; that is, repair is not instantaneous
         (Table 4-3).
     D = Imperfect  Repair Correction Factor = correction factor to
         account  for the fact that  some sources which are repaired are
         not reduced to zero emission levels.  For computational pur-
         poses, all sources which are repaired are assumed to be reduced
         to a 1000  ppm emission level (Table 4-4).
These correction  factors can, in turn, be determined from the following
expressions:
                                 rf
                     0)  B = 1 - /

                     ty\  r - 365 -  t
                     u;  L "   365

                     (3)  D = 1 - I
                                     4-12

-------
Where:
     n  = Average number of leaks occurring and recurring over the
      m                                                               .
          monitoring interval.
      N = Total number of sources at or above the action level (Figure
          4-2).
      t = Average time before repairs are made (with a 15-day repair limit,
          7.5  is the average used).
      f = Average emission factor for sources at the average screening
          value achieved by repair.
      F = Average emission factor for all sources at or above the  action
          level.

An example of a control  effectiveness  calculation  is  presentee!  in  Table  4-5.
Support data for this  calculation are  presented  in Tables  4-1,  4-2, 4-3,
4-4, and 4-6, as well  as  in Figures  4-1  and  4-2.
4.2  EQUIPMENT SPECIFICATIONS
     Equipment specifications  for each emission  source  are  described
below.  Some of the  specifications may be  applicable  to more than  one
type of source.  In  these cases,  references  are  made  to the preceding
description  with any differences  in  applicability  or  effectiveness  notes.
4.2.1  Pumps
     Fugitive emissions  from  pumps occur at  the  junction  of a moving shaft
and a stationary casing.   Equipment  specifications that may be  implemented
for pumps  include elimination  of  this  junction,  improvement of  the  seal  at •
the junction, or collection and control  of the  emissions  from the  junction.
     4.2.1.1  Seal!ess Pumps.  Pumps such  as diaphragm  type pumps  or
"canned" pumps do not  have a  shaft/casing  junction and  therefore  do not
leak the pumped fluid  as  a normal course of  operation.  Failure of the
diaphragm may result in  temporary emissions  of  VOC.   Sealless pumps are
used primarily in SOCMI  processes where  the  pumped fluid  is hazardous  or
toxic, and every effort  must  be made to  prevent  leaks of  the fluid.
                                    4-13

-------
       TABLE 4-5.  EXAMPLE OF CONTROL EFFICIENCY CALCULATION
Assume:
    1)  A leak detection and repair program to reduce emissions from
        valves in gas/vapor source.
    2)  Action level = 10,000 ppm.
    3)  Average screening value after directed repair = 1,000 ppm.
    4)  Leak detection monitoring interval = 3 months.
    5)  Allowable repair interval =15 days.
    6)  Number of valves having new or recurring leaks between repair
        intervals, n  = 0.2N (see Table 4-6).
                    m

Calculations:
    A = 0.98 (from Figure 4-1 for a screening value of 10,000 ppmv)
    B = 0.9 (from Table 4-6)
    C = 0.979  (from Table 4-3 for 15-day interval)
    where:
        c -    A(Avg. uncontrolled emission factor)9
            Fraction of sources screening > 10,000 ppmD
          = (0.98)(0.021:kg/hr)/0.10 = 0.206 kg/hr
        f = Emission factor at 1000 ppme
          = 0.001 kg/hr
    and D = (1 - jj^-) = 0.995

    Overall percentage reduction  =AxBxCxD
                                  = (0.98) x  (0.9) x  (0.979) x  (0.995)
                                  = 86 Percent
    Therefore:
        Control  effectiveness factor = 0.021 kg/hr -  (0.86)(0.021 kg/hr)
                                     = 0.003 kg/hr
?   Reference  10.  '
    From Figure 4-2.
c   Reference  11.
                                  4-14

-------
TABLE 4-6.   IMPACT  OF  MONITORING  INTERVAL ON  CORRECTION FACTOR ACCOUNTING
            FOR LEAK OCCURRENCE/RECURRENCE  (FOR  EXAMPLE CALCULATION)
      Monitoring                                 -
       interval                  m                   m
       1  month                0.1N                0.05N               0.95
                                       x


       3 months               0.2N                0.1N                0.90


       1  year                 0.4N                0.2N                0.80
a   n  = Total number of leaks which occur, recur, and remain between
     m   monitoring intervals.

    n"  = Average number of leaks over the monitoring interval.
     m

c   B  = Correction factor accounting for leak occurrence/recurrence.
                                 4-15

-------
 O
 C/J
 CO
 O
 EC
100

 90

 80

 70

 60

 50

 40

 30

 20

 10

  0
        11
i 111  i  i ni  i i_i 11 j—rrn  |  rm r
	  '""" W-^.
                 -i_>iU<
                       VS
                                 I
               I I I 1  I  I J II  I 1  t 11  I  111
                                                   UPPER LIMIT OF 90%
                                                   CONFIDENCE INTERVAL
                                                   ESTIMATED PERCENT OF
                                                   TOTAL MASS EMISSIONS

                                                   LOWER LIMIT OF 90%
                                                   CONFIDENCE INTERVAL
                                         PERCENT OF TOTAL MASS
                                                   EMISSIONS -
                                       INDICATES THE PERCENT OF
                                       TOTAL EMISSIONS ATTRIBU-
                                         TABLE TO SOURCES WITH
                                      SCREENING VALUES GREATER
                                       THAN THE SELECTED VALUE
             10    100    1000  10,000 100.000  1.000.000

           SCREENING VALUE (ppmv) 
-------
     4.2.1.2  Double Mechanical  Seals.   Double mechanical  seals consist
of two mechanical  sealing elements  with a barrier fluid in the chamber
between the seals.   This  chamber is either flushed with circulating
barrier fluid or is flooded with static barrier fluid.   The pressure of
the static barrier fluid  can be  monitored to de.tect failure of the inner
seal.14  Any leaks through the inner seal may be dissolved or suspended in
the barrier fluid, and subsequent degassing of the sealing fluid may
result in emission of VOC.  Therefore,  barrier fluid degassing vents must
be controlled in order to provide maximum control effectiveness of double
mechanical seals.   After  extended periods of use, double seals may also
develop leaks at the outer seal/shaft junction.
     Emissions of VOC from degassing vents can be controlled by a closed
vent system which consists of piping and, if necessary, flow inducing
devices to transport the  degassing emissions to a control  device such as
a process heater, or vapor recovery system.   Control  effectiveness of a
double mechanical seal  and closed vent system is  dependent on the effective-
ness of the heater, or  vapor recovery system, and the frequency of seal
failure.  Failure of both the inner and outer seals can  result in relatively
large VOC emissions at  the seal  area of the  pump.  As noted,  the pressure
monitoring of the static barrier fluid may be used in order to detect failure
of the seals.  In addition, visual  inspection of  the  seal  area also can  be
effective.for detecting failure  of the outer seals.   Upon  seal failure,  the
leaking pump would have to be shut down for  repair.
      Double  mechanical  seals are used  in many SOCMI   process applications;
 however,  there are  some conditions that preclude use of double mechanical
 seals.   Their maximum service temperature is usually limited  to  less than
 260°C,  and mechanical seals cannot be  used on pumps  with reciprocating
 shaft motion.  Process fluids containing catalyst fines or other  abrasive
 materials may not  be suitable for use  with mechanical seals.
      4.2.1.3 Closed Vent  Systems.   The system described above  for con-
 trolling degassing  vent emissions could also be applied to control
 emissions from the  seal area of pumps.  This application would  require
 the  use of some  type of flow inducing  device to transport the emissions
 from the seal area  to the  control device.  The seal   area would  be enclosed
                                    4-17

-------
in order to collect the emissions and a vacuum eductor or a compressor
could be used to remove vapors from the seal area.  However, normal pump
operating practices may require frequent visual inspection or mechanical
adjustments in the seal area.  This would not be possible with a closed
vent system at the seal area.  A potential  problem with this approach is
that explosive mixtures may be created by enclosing the pump seal area,
and therefore safety and operating practices may limit the use of closed
vent systems for pump seal areas.
4.2.2  Compressors
     Fugitive emissions from compressors occur at the junction of a moving
shaft and a stationary casing.  Emission reductions from this source type
may be achieved by improving the seal at the junction, or collecting and
controlling the emissions from the junction.
     4.2.2.1  Double Mechanical Seals.  Double mechanical seals for
compressors are similar to those described above for pump applications,
and reciprocating shafts cannot be fitted with mechanical seals.
Labyrinth type seals may also have barrier fluid systems.  Existing
compressors may have double mechanical seals and seal  oil flush systems,
but seal oil reservoir degassing vents must be controlled with closed
vent systems as described above.  Control efficiency is dependent on the
control device efficiency and the frequency of seal failures.
     4.2.2.2  Closed Vent Systems.  The seal area of a compressor may
be enclosed, and the VOC emissions routed to a control device through
a closed vent system.  However, flow inducing*, devices may be required to
transport vapors to the control device.  Although the formation of
explosive mixtures in the enclosed seal area may prohibit application
of this equipment modification, closed vent systems have been applied to
compressor seal areas in petroleum refineries.
4.2.3  Pressure Relief Devices
     Pressure relief devices include rupture disks and safety/relief
valves.  Fugitive emissions from these devices occur because of improper
seating or partial failure of the device.  These fugitive emissions do
not include emissions which result from normal operation of the devices
                                   4-18

-------
caused by overpressure  of the  process  or vessel  which  the device protects.
Fugitive emissions  from rupture disks  may be caused by pinhole leaks in
the disk itself caused  by corrosion or fatigue.   Fugitive emission from
relief valves may be caused by failure of the valve seating surfaces,
improper reseating  after overpressure  relieving, or process operation
near the relief valve set pressure which may cause "simmering".
     4.2.3.1  Rupture Disks.   Although they are also pressure relief devices,
rupture disks can be installed upstream of a safety/relief valve in order
to prevent fugitive emissions  through  the relief valve seat.  This procedure
may require use of a larger size relief valve because of operating codes.  The
disk/valve combination  may also require appropriate piping changes to prevent
disk fragments from lodging in and damaging the relief valve when relieving
overpressure.  A block valve upstream of the rupture disk is also required
in order to permit in-service replacement of the disk after overpressuring.
If the disk could not be replaced, the first overpressure would result in
the relief valve being the same as an uncontrolled relief valve, and it may
actually be worse since disk fragments may prevent proper reseating of the
relief valve.  In some chemical plants, installation of a block valve up-
stream of a pressure relief device may be a common practice.  In others, it
may be forbidden by operating or safety procedures.  Tandem pressure
relief devices with a three-way valve can be used to avoid operation without
overpressure protection.  Rupture disk/relief valve combinations must have
some provision for testing the integrity of the disk.  The area between the
rupture disk and relief valve must be connected to a pressure indicator,
recorder, or alarm.  If the process fluid is not hazardous or toxic, a
simple bubbler apparatus could be used to test disk integrity by connecting
the bubbler to the disk/valve area.  The control efficiency of the disk
valve combination is assumed to be 100 percent for fugitive emissions.  If
the disk integrity is not maintained or if the disk is not replaced after
overpressure relief, the control efficiency would be lowered.  The disk/valve
combination has no effect on emissions which result from overpressure
relieving.
                                    4-19

-------
     4.2.3.2  Resilient Seat Relief Valves.  Manufacturers of relief
valves state that resilient seat or "0-ring" relief valves provide better
reseat qualities compared to standard relief valves.  No test data are
available to verify these statements.  These improvements would have no
effect on overpressure emissions or fugitive emissions due to seal
failure or "simmering'.'..
     4.2.3.3  Closed Vent Systems.  A closed vent system can be used to
transport the discharge or leakage of pressure relief devices to a
control device such as a flare.  Since overpressure discharges as well as
fugitive emissions are routed to the control device, it must be sized
appropriately.  A larger pressure relief device may be required for use
with a closed vent system.  The control efficiency of a closed vent
system is dependent on the effectiveness of the control device.  Typical
flare systems may be only 60 percent effective for fugitive emission
            15
destruction.     This efficiency reflects the fact that many flare systems
are not of optimum design.  As a result flares that are designed to handle
large volumes of vapors associated with overpressure releases are used to
handle low volumes of fugitive emissions.  With such designs, optimum mixing
is not achieved because the vent gas exit velocity is low and large
                                                                      16
flares generally cannot properly inject steam into low volume streams.    A
properly designed flare system typically exhibits a 99 percent hydrocarbon
                        17
destruction efficiency.    Closed vent systems for pressure relief devices
are used in existing SOCMI processes especially where the emissions may be •
hazardous or toxic.
4.2.4  Open-Ended Valves
     Fugitive emissions from open-ended valves are caused by leakage  through
the seat of the valve.  Emissions may also  occur  through the stem and gland
of the valve, and these emissions may be controlled by methods described
for valves in Section 4.1.2.  Approximately 28 percent of SOCMI valves
                                                                         l!8
(excluding safety/relief  and check valves)  in VOC service are open-ended.
They include drain, purge, sample, and vent valves.   Fugitive emissions from
open-ended valves can be  controlled  by installing a cap, plug, flange,  or
second valve to the open  end of the  valve.  In the case of a second valve, the
                                      4-20

-------
upstream valve should always  be closed first after use of the valves.   Each
time the cap,  plug,  flange,  or second valve is opened, any VOC which has
leaked through the first valve seat will  be released.   These emissions have not
been quantified.   The control  efficiency  of these devices is assumed to be
100 percent.   The actual efficiency will  be dependent  on the frequency of
removal of the cap or plug.   Caps,  plugs, etc. for open-ended valves do not
affect emissions  which may occur during use of the valve.  These emissions may
be caused by line purging for sampling, draining, or venting through the
open-ended valve.  Caps, plugs, flanges,  or second valves for open-ended
                                              19
valves are required  by California regulations..
4.2.5  Sampling Connections
     Fugitive  emissions from sampling connections occur as a result of
purging the sampling line in  order to obtain a representative sample of the
process fluid.  Approximately 25 percent  of open-ended valves are used for
                     20
sampling connections.    Fugitive emissions from sampling connections  can be
reduced by using  a closed loop sampling system.   The closed loop system is
designed so that  the purged  fluid is returned to the process at a point of
lower pressure.  A throttle  valve or other device is required to induce the
pressure drop  across the sample loop.  The purged fluid could also be  directed
to a control  device  such as  a flare.  In  this case the control efficiency would
be dependent on the  flare efficiency for  hydrocarbon destruction.  Closed
loop sampling  is  assumed to  be 100 percent effective for controlling fugitive
emissions.  Since some pressure drop is required to purge sample through the
loop, low pressure processes  or tankage may not be amenable to closed  loop
sampling.  Safety requirements may prohibit closed loop sampling in some
instances.
4.2.6  In-Line Valves
     Fugitive  emissions from valves occur at the stem  or gland area of the
valve body.  Diaphragm and bellows seal valves do not  have a stem or gland
and therefore  are not prone to fugitive emissions.  Diaphragm valves are
generally used where hazardous or toxic process fluids are present and
fugitive emissions must be eliminated.  Their control  effectiveness is
approximately 100 percent, although failure of the diaphragm may cause
                                   4-21

-------
large temporary emissions.  The applicability of diaphragm valves  is
limited by the strength of the diaphragm.   Diaphragm valves may not be
suitable for many applications because of process conditions or cost
considerations.
4.2.7  Effectiveness of Equipment Specifications
     In order to quantify the environmental  and economic impacts of apply-
ing controls, the control efficiency must be determined.  In some  cases,
there are many complicating factors which make it difficult to accurately
estimate control efficiency.  For example, the efficiency of caps  or  plugs
for open-ended valves is dependent on 1) the frequency of removal  of the
cap or plug, since this removal will result in emission of fluids  trapped
by the cap or plug, and 2) the emission rate through the valve seat.   The
estimated control efficiencies for various equipment modifications are
shown in Table 4-7.  These estimates represent the maximum emission re-
duction possible for the equipment modifications.  In some instances, the
actual emission reduction will depend on other factors such as the effi-
ciency of control devices attached to closed vent systems.  Carbon absorp-
tion or vapor recovery systems would approach the 100 percent efficiency,
but flares may be only 60 percent effective for hydrocarbon destruction.
These estimates of effectiveness are used to calculate environmental  and
economic impacts of regulatory alternatives  in Chapters 7 and 8 of this
document.
                                     4-22

-------
         TABLE 4-7.   EFFECTIVENESS  OF EQUIPMENT MODIFICATIONS
Source type/                                   Control  efficiency
  equipment modification	(%)	
Pumps
  Sealless pumps                                      100
  Double mechanical  seals/closed vent system         ^100
  Closed vent system on seal  area                    ^100
  Double mechanical  seals/closed vent system         %100a
Compressors
  Double me
  Closed vent system on  seal  area

Safety/relief valves
  Closed vent system
  Rupture disks                                       100
  Closed vent system                                   60
Open-ended lines
  Caps, plugs, blinds, second valves                  100

Sampling connections
  Closed loop sampling                                100

In-line valves
  Diaphragm valves                                    100
aAlthough a control efficiency is not attained in all cases, it is
 achievable in some cases.
 This control effectiveness reflects the fact that a closed vent system is
 normally sized for emergency relief.21
This is the control efficiency reflects the.use  of these devices  downstream
of an irntial  valve with  VOC  on one  side and.atmosphere  on  the  other
                                  4-23

-------
4.3  REFERENCES

1.   Hustvedt, K. C., and R. C. Weber.  Detection of Volatile Organic
     Compound Emissions from Equipment Leaks.  Presented at 71st Annual
     Air Pollution Control Association Meeting, Houston, Texas, June 25-30,
     1978.

2.   Ref. 1.

3.   Wetherold, R. G., and L. P. Provost.  Emission Factors and Frequency
     of Leak Occurrence for Fittings in Refinery Process Units.  Interim
     Report.  EPA/600/2-79-044.  Radian Corporation.  February 1979.  p. 2.

4.   Ref. 3.

5.   Ref. 3.

6.   Valve Repair Summary and Memo from F. R. Bottomley, Union Oil Company.
     Rodeo, California.  To Milton Feldstein, Bay Area Quality Management
     District.  April 10, 1979.

7.   Environmental Protection Agency.  Control of Volatile Organic Com-
     pound Leaks from Petroleum Refinery Equipment.  EPA-450/2-78-036.
     OAQPS No. 1.2-111.  Research Triangle Park, North Carolina. June 1978.

8.   Ref. 3.

9.   Ref. 3.

10.  Ref. 3.

11.  Ref. 3.

12.  Ref. 3.           '

13.  Ref. 3.

14.  Erikson, D. G., and V.:  Kalcevic.  Emission Control Options for the
     Synthetic Organic Chemicals Manufacturing Industry, Fugitive Emissions
     Report, Draft Final.  Hydroscience, Inc.  1979.  p. III-l.

15.  Draft-of Background  Information Document  (BID), Chapter 4, for
     a National  Emission Standard for Hazardous Air Pollutants from the
     Ethylbenzene/Styrene Industry.  U.S. Environmental Protection
     Agency, RTP, North Carolina, October 1979.
                                   4-24

-------
16.    Ref.  15.

17.    Environmental  Protection Agency.   Control  of Volatile Organic
      Emissions from Existing Stationary Sources.  Volume 1:  Control
      Methods for Surface Coating Operations.  EPA-450/2-76-028.
      Research Triangle Park, North Carolina.  November 1976, p. 42.

18.    Ref.  14, p. III-5.

19.    Ref.  14, p. III-5.

20.    Ref.  14.

21.    Ref.  15.
                                  4-25

-------
                     5.   MODIFICATION  AND  RECONSTRUCTION

     In accordance with  the  provisions of  40 Code  of Federal  Regulation  (CFR),
Sections 60.14 and 60.15, an "existing facility" can become an  affected
facility and,  subsequently,  subject to the standards of performance if it
is modified or reconstructed.   An  existing facility, as defined in  40 CFR
60.2 (aa),  is  a facility of  the type for which  standards  of performance  have
been promulgated and the construction  or modification of  which  was  begun
prior to the proposal  date of the  applicable standards.
     The applicability of provisions 40 CFR 60.14  and 60.15 to  the  SOCMI,
and the conditions, as outlined in these provisions, under which existing
facilities  could become  subject to standards of performance are discussed
below.
5.1  GENERAL DISCUSSION OF MODIFICATION AND RECONSTRUCTION PROVISIONS
5.1.1  Modification
     "Modification" is defined in  40 CFR 60.14.(a) as any physical  or
operational change of an existing  facility which increases the  emission rate
of any pollutant to which a  standard applies.  Exceptions to this definition
are presented in paragraphs  (d), (e),  and  (f) of Section  60.14.  These
exceptions  are as follows:
          Paragraph (d)  - In accordance with the paragraph, an
          existing facility  may undergo a  physical or operational
          change, which increases  the emission rate of any pollurant
          to which standards of performance apply, but not judged to
          be a modification, if the owner  or operator can demonstrate
          to the Administrator's satisfaction (by  any of  the pro-
          cedures prescribed in paragraph  (b) of this section)  that
                                     5-1

-------
          the total  emission rate of that pollutant has  not increased
          from the facility.
          Paragraph (e) - Physical or operational  changes  to an
          existing facility which will  not be considered modifica-
          tions are specified in this portion of Section 60.14.
          These changes include:
               a.   Routine maintenance, repair,  and replacement.
               b.   An increase in the production rate not  requiring
                   a capital expenditure as defined in Section
                   60.2(bb).
               c.   An increase in the hours of operation.
               d.   Use of an alternative fuel or raw material if
                   prior to the standard, the existing facility
                   was designed to accommodate that alternate fuel
                   or raw material.
               e.   The addition or use of any system or device
                   whose primary function is the reduction of
                   air pollutants, except when an emission control
                   system is removed or replaced by a system con-
                   sidered to be less efficient.
          Paragraph (f) - This paragraph provides for superceding
          any conflicting provisions of this section.
     Upon modification, an existing facility becomes an affected facility
for each pollutant to which a standard applies and for which there  is an
increase in the emission rate to the atmosphere.  Paragraph (c)  also declares
that the addition of an affected facility to a stantionary source through
any mechanism—new construction, modification, or reconstruction—does not
make any other facility within the stationary source subject to  the applicable
standards.
5.1.2  Reconstruction
     Under the provisions of Section 60.15, an existing facility becomes
an affected facility upon reconstruction, irrespective of  any change in
emission rate.  Generally, reconstruction is considered to occur upon the
replacement of components if the fixed capital cost of the new components
exceeds 50 percent of the fixed capital cost that would be required to
                                      5-2

-------
construct a comparable entirely new facility,  and it is  economically and
technically feasible for the facility to comply with the applicable
standards of performance.   The final  judgments on what replacement con-
stitutes reconstruction and when it is technologically and economically
feasible to comply with the applicable standards of performance is made
by the Administrator.   The  Administrator's  final determinations are made
on the following bases:
     (1) comparison of the  fixed capital costs of the replacement
         components and a newly constructed comparable facility,
     (2) the estimated life of the facility after the replacements
         compared to the life of a comparable  entirely new facility,
     (3) the extent to which the components being replaced cause
         or contribute to the emissions from the facility, and
     (4) any economic or technical limitations on compliance with
         applicable standards  of performance  which are inherent in
         the proposed replacements.
     The purpose of this provision is to ensure that an owner or operator
does not perpetuate an existing facility by replacing all  but vestigial
components, support structures, frames, housing, etc., rather than
totally replacing it in order to avoid subjugation to applicable standards
of performance.   In accordance with Section 60.5, EPA will, upon request,
determine if the action taken constitutes construction (including reconstruction)
5.2  APPLICABILITY OF MODIFICATION AND RECONSTRUCTION PROVISIONS TO THE
     SOCMI
5.2.1  Modification
     Several operating conditions that could be encountered in an organic
chemical plant are presented below.   These  conditions may or may not result
in an increase in emissions.
     The replacement of a potential  fugitive emission source such as a pump
or valve commonly occurs in an organic chemical plant.  If such a source
is replaced with an equivalent source (such as is done during routine
repair and replacement), the fugitive emissions from the facility should not
increase because the number of potential sources in the same vapor pressure
service (handling the same  organic chemical) remains unchanged.

                                     5-3

-------
     Process equipment pieces such as heat exchangers, reactors,  distilla-
tion columns, reboilers, filters and separators,  or new control  loops  are
commonly added to existing facilities in the organic chemical  industry to
increase the capacity of or to optimize a process.   The addition  of this
equipment would normally increase fugitive emissions from a facility due to
the increased number of potential emission sources  (pumps, valves,  sampling
connections, etc.) that are associated with the process equipment.   However,
in those cases where some sources are physically removed from service, the
addition of new fugitive emission sources would not necessarily increase the
level of fugitive emissions from the stationary source.
     In some cases a facility in the organic chemical industry can  be
converted from the production of one chemical to the production of  a second
chemical.  This normally occurs when production of  the second chemical results
in greater profits.  In such a case, whenever either the number of  fugitive
emission sources or the vapor pressure of the second chemical  increases during
this conversion, the level of VOC emissions from the facility could be
expected to increase.
     Changes may be made to a process, although the chemical being  produced
remains the same.  One such case would be a change  in catalyst for  producing
a given chemical.  In such a case the level of fugitive emissions would not
be expected to change because the number of sources nor the vapor pressure of
the chemical would change.
     In many cases, there may be a desire to increase the capacity  of an
existing facility.  This may be achieved by replacing certain process equip-
ment (pumps, heat exchangers, reactors, etc.) with  similar equipment but of
larger capacity or addition of process equipment.  If this replacement or
addition does not increase the number of fugitive emission sources  handling
the given organic chemical, the level of fugitive emissions would not be
expected to increase.  However, if the number of sources were to increase due
to this replacement or addition, then VOC emissions could be expected to
increase.
                                      5-4

-------
5.2.2  Reconstruction
     When an owner or operator replaces several  components of an existing
facility, that facility may or may not become subject to applicable
standards of performance under the provisions of Section 60.15.   For
example, if an owner or operator replaces several  reactors in an existing
facility, reconstruction is considered to have occurred if the fixed
capital costs for these reactors exceeds 50 percent of the costs that
would be required to construct an entirely new facility.  Replacement
of other major equipment components such as heat exchangers, and distillation
columns may also be considered as reconstruction if the fixed capital
costs for the replaced equipment exceeds 50 percent of the costs of construct-
ing an entirely new facility.
                                     5-5

-------
             6.   MODEL  PROCESS  UNITS  AND REGULATORY  ALTERNATIVES

     This chapter presents  model  process unit  parameters  and alternative
emission controls considered  for  reduction  of  fugitive emissions from SOCMI
sources.  The model  units were  selected  to  represent the  range of processing
complexity in the industry  to provide a  basis  for comparing environmental
and economic impacts of the regulatory alternatives.   The selected regulatory
alternatives provide varying  levels of emission  control.
6.1  MODEL UNITS
     Available data  show that fugitive emissions are proportional to the
number of potential  sources,  but  are  not related to  capacity, throughput,
age, temperature, or pressure.1  SOCMI model  units therefore represent
different levels  of  process complexity (number of sources) rather than
unit size.
6.1.1  Sources of Fugitive  Emissions
     The various  potential  fugitive emission  sources in a SOCMI process
unit were described  in  Chapter  3.   Data  from  petroleum refineries indicate
that cooling towers  are very  small  sources  of  VOC emissions.2  Differences
in SOCMI operating procedures,  such as recirculation of process water, might
result in cooling tower VOC emissions, but  no  data are available to verify
this.  The number of agitator seals in SOCMI  is  not  known.  Furthermore, the
emission rate from SOCMI agitator seals  has not  been measured.  Since there
are no data from  similar sources  in other industries, no  estimates of emission
rate can be made.  Because  of these uncertainties, cooling towers and agitator
seals are not included  in  the Model Units.
                                   6-1

-------
6.1.2  Model Unit Parameters
     SOCMI process units vary considerably in size,  complexity,  age,  and
types of products manufactured.   In order to estimate emissions, control
costs, and environmental impacts on a unit specific  basis,  three model
units were developed.  The technical  parameters for  the model  units  are
shown in Table 6-1.  These three model  units represent the  range of
emission source populations that may exist in .SOCMI  process units.   The
technical parameters were developed from a data base compiled  by Hydro-
             3
science, Inc.   The data base included  equipment source counts from  62  SOCMI
plants which produce 35 different chemicals.  These  plant sites  represent
approximately 5 percent of the total  existing SOCMI  plants  and include  large
and small capacities, batch and continuous production methods, and varying
levels of process complexity.  Hydroscience estimates that 52  percent of
existing SOCMI plants are similar to Model Unit A, 33 percent  are similar
to B, and 15 percent are similar to C.   The source counts for  the 35 chemicals
include pumps, valves, and compressors.  These counts were  used in com-
bination with the number of sites which produce each chemical  in order  to
                                                 4
determine the average number of sources per site.
     Data from petroleum refineries indicate that emission rates of  sources
decrease as the vapor pressure  (volatility) of the process fluid decreases.
Three classes of volatility have been established based on the petroleum
refinery data.  These include gas/vapor service, light liquid  service,  and
heavy liquid service.5  The split between light and heavy liquids for the
refinery data is between streams called naphtha and kerosene.   Since simi-
lar stream names may have different vapor pressures, depending on site
specific factors,  it is difficult to quantify the light-heavy  split.  The
break point is approximately at a vapor pressure of 0.3 kPa at 20°C.
The data  collected  by Hydroscience were used to estimate the split between
                                                  6
gas/vap.or and liquid service for each source type.   In order to apply
emission  factors for light and  heavy liquid service, it is assumed that
one half  of SOCMI  liquid service sources are in light liquid service.  There
are no data available on the actual distribution of sources in volatility
ranges.   It  is assumed  that all SOCMI packed seal pumps are in heavy liquid
service.  This assumption is reasonable, since more volatile liquids are
                                  6-2

-------
     TABLE 6-1.  FUGITIVE EMISSION SOURCES  FOR  THREE  MODEL UNITS
Number of components in model unitc

Equipment component3
Pump seals
Light liquid service
Single mechanical
Double mechanical
Seal less
Heavy liquid service
Single mechanical
Packed
In-line valves
Vapor service
Light liquid service
Heavy liquid service
Safety/ relief valves
Vapor service
Light liquid service
Heavy liquid service
Open-ended valves and lines
Vapor service
Light liquid service
Heavy liquid service
Compressor seals
Sampling connections0
Flanges

Cooling towers
Agitator seals
^Equipment components in VOC
Sample, drain, purge valves
^Based on 25% of open-ended
^-* C O "/ A f j-i**^*»+-T»-*j-t i i n •? 4- «— rt v* rt
Model unit
A


5
3
0

5
2

90
84
84

11
1
1

9
47
48
1
26
600
c

600e
service only.
•
valves. From Ref.
*» T m 4 1 a w» -t- f\ Mrtrltt 1 1!
Model unit
B


19
10
1

24
6

365
335
335

42
4
4

37
189
189
2
1.04
2400
c

2400e


3, pg. IV-3.
U44. A
Model unit
C


60
31
1

73
20

1117
1037
1037 ;
;
130 1
13
14

115
581
581
8
320 ;
7400 <
e :'

7400e



3,3% of existing units
15% of existing units
Ref. 3, pg. IV-1.
Data not available.
are similar to Model Unit B.
are similar to Model Unit C.
                                 6-3

-------
more suitable for mechanical seal applications,  and newer process units tend
to use fewer packed seals.  Sampling connections are a subset of the open-
ended valve category.  Approximately 25 percent  of open-ended valves are used
for sampling connections.7  Emissions which occur through the valve stem,
gland, and open-end are included in the open-ended valve category.   The
emission factor for sampling connections applies only to emissions  which
result from sample purging.
6.2  REGULATORY ALTERNATIVES
     Regulatory alternatives represent comprehensive programs for reduction
of emissions by combining the individual control techniques described in
Chapter 4.  The regulatory alternatives described in this section contain
feasible control techniques for reducing fugitive emissions of VOC from
SOCMI sources.
     The purpose of developing different regulatory alternatives is to
provide a basis, along with model unit parameters, for determing the air-
quality and non air-quality environmental impacts, energy requirements, and
the costs associated with varying degrees of VOC fugitive emissions reduction.
The regulatory alternatives selected for analysis include a "status quo of
fugitive emission control" case and three increasingly restrictive levels of
emission control requirements.  The "status quo" case allows for the analysis
of not implementing standards of performance.  The three increasingly restric-
tive control requirements allow for analysis of the impacts of different
systems with varying degrees of emission reduction.  The requirements for
each of these regulatory alternatives are summarized in Table 6-2 and are
described below.
6.2.1  Regulatory Alternative I
     Alternative I represents the general level  of control that would exist
in the absence of establishing any VOC fugitive emission control requirement.
For this case, SOCMI facilities  located in oxidant National Ambient Air
Quality Standard (NAAQS) attainment areas, in general, would not be subject
to any requirements.   However, some states may require leak detection and
repair programs to control  fugitive emissions of VOC through prevention of
significant deterioration  (PSD)  statutes.  SOCMI facilities located in
                                     6-4

-------
               TABLE  6-2.   REGULATORY ALTERNATIVES  FOR  FUGITIVE  EMISSION  SOURCES IN  SOCMI



Source typed
Pumos
Light liquids
with single mechanical seals



with double mechanical seals


with no seals
Heavy liquids
with packed seals
with single mechanical seals
Valves (in-line)
Gas
Light liquid
Heavy liquid
Safety/relief valves
Gas

Light liquid
Heavy liquid
Open-ended valves and lines
Gas
Light liquid
Heavy liquid
Flanges
Sampling connections

Compressor seals



^Sources In VOC service.


Monitoring
interval

None



None


None

None
None

None
None
None

None

None
None

None
None
None
None
None

None





I
Equipment
specification

None



None


None

None
None

None
None
None

None

None
None

None
None
None
None
None

None






Monitoring
interval

Annually



Annual ly°


None

None
Hone

Quarterly
Annually
None

Quarterly0

None
None

Quarterly
Annually
None
None
Noned

Quarterly




Regulatory <
I!
Equipment
specification

None



Hone


None

None
None
N.
None
None
None

None

None
None

Capsf
Capsf
Caps
None
None

None




ilternative

Monitoring
interval

Monthly6



Monthly0


None

None
None

Monthly
Monthly
None

Monthly0

None
None

Monthly
Monthly
None
None
Noned

Monthly





111
Equipment
specification

None



None


None

None
None

None
None
None

None

None
None

Capsf
Capsf
Capsf
None
None

None






Monitoring
interval

Noneb



None6


None

None
None

Monthly
Monthly
None

Nonec

None
None

Monthly
Monthly
None
None
Noned

None





IV
Equipment
specification

Double seals;
degassing vents
connected to
control device
Degassing vents
connected to
control device6
None :

None
None

None
None
None

Upstream
rupture disks
None
None

Capsf
Capsf
Capsf
None
Closed loop
sampling
Seal area or
degassing vents
connected to
control device

"Plus weekly visual inspection.   If liquid leak is observed,  instrument monitoring  is required to determine if action level  is being exceeded.
 Monitoring is required after each over pressure  release.   If it  Is found to be  leaking, the valve will  be repaired
 Included  1n open-ended valves.
eSealless  pumps nay also be used.
'Or blinds, plugs', second valves.             "       	

-------
non-attainment areas would be subject to the applicable SIP regulations  and
other permitting requirements.  In some areas control  of fugitive VOC emissions
may be used to achieve hydrocarbon emission offsets.   However,  no present or
anticipated SIP regulations would be generally applicable to SOCMI.   This
baseline control alternative merely presents a generalized fugitive  emission
control level that can be used to compare the impacts  of the more stringent
alternatives.  As such, this alternative does not consider the  levels of
control for specific facilities.
6.2.2  ReguaTatory Alternative II
     This alternative would require leak detection and repair methods as
in the petroleum refinery control techniques guideline (CTG), EPA-450/2-78-036.
Leak detection would be accomplished by checking equipment components for
emissions of VOC using a portable VOC detection instrument to sample and
analyze the air in close proximity to the potential leak area.   A measured
VOC concentration greater than some predetermined level, known  as an "action
level", would be defined as a leak that would require  equipment repair.   A
measured VOC concentration less than the action level  would not require  equip-
ment repair.  The action level is defined as 10,000 ppmv VOC concentration
for all cases.
     Quarterly monitoring of compressors, gas service  relief valves, inline
valves, and open-ended valves would be required.  Annual monitoring  of light
liquid service pumps and valves would be required.  Weekly visual inspections
of light liquid pump seals would also be required.  Leaks detected visually
would require instrument monitoring to determine if the action  level is
exceeded.  Relief valve monitoring after over pressure relieving would be
required.  Open-ended valves would be required to be sealed with a cap,  blind,
plug, or another valve.
6.2.3  Regulatory Alternative III
     Regulatory Alternative III would provide for more restrictive control
than Alternative II by increasing the inspections for  all applicable equipment
to monthly.  Increasing the inspections would result in a reduction  of
emissions from residual leaking sources; i.e., those sources which are found
                                   6-6

-------
leaking and are repaired  and  recur  before  the  next inspection and those
sources that begin leaking  between  inspection.   Thus,  although this alterna-
tive is similar in approach to Alternative II,  it provides  for more emissions
reduction.   The requirements  for weekly visual  pump seal  inspections, relief
valve monitoring.after over pressure,  and  caps  for open-ended valves are the
same as Alternative II.
6.2.4  Regulatory  Alternative IV
     Alternative IV would require equipment specifications  instead of more
frequent equipment inspections.  This  alternative would provide a more
restrictive level  of control  than the  other alternatives.   Several equipment
specifications would be required, including caps for open-ended valves as in
Alternatives II and III.   Closed loop  sampling  techniques would be required
and rupture disks  would be  required on gas service relief valves venting to
atmosphere.  The integrity  of the disk would be required and replacement of
the disk would be  required  whenever a  failure  is detected.   No monitoring
would be required  for relief  valves which  have rupture disks upstream or
which vent to a control  device header.  Compressor seal areas and degassing
vents from seal oil reservoirs, or  both, would be required  to be connected
to a control device with a  closed vent system.   Pumps in light liquid
service would be required to  have double mechanical seals with a seal oil
flushing system.  The degassing vent from  the  seal oil reservoir would be
required to be connected to a control  device with a closed  vent system.
6.3  REFERENCES
1.  Wetherold, R.G., L.P. Provost,  D.D. Rosebrook, and C.D. Smith.
    Emissions Factors and Frequency of Leak Occurrence for  Fittings in
    Refinery Process Units.  Interim Report.  Radian Corporation.
    Austin, Texas.  EPA Contract Number 68-02-2665.  p. 11-49.
2.  The Assessment of Environmental Emissions  from Oil Refining.  Draft
    Report, Appendix B.   Radian Corporation.  Austin, Texas.  August 1979.
    pp. 3-5 through 3-16.
3.  Erikson, D.G., and V. Kalcevic.  Emissions Control Options for the
    Synthetic Organic Chemicals Manufacturing  Industry.  Draft Report.
    Hydroscience,  Inc.  Knoxville,  Tennessee,   EPA Contract Number 68-02-2577.
    p. IV-1,2.
                                     6-7

-------
4.  Ref. 3, p. II-9-13.
5.  Ref. 1, p. 11-23.
6.  Ref. 3, p. 11-10.
7.  Ref. 3, p. IV-8.
                                    6-8

-------
                       7.   ENVIRONMENTAL IMPACT

     The environmental  impacts that would result from implementing the
regulatory alternatives being considered in this study are examined in
this chapter.   Included in  this chapter are estimations of the controlled
VOC fugitive emissions  and  the incremental  reductions in uncontrolled VOC
emissions that could be achieved under each of the alternatives.   Also,
the impacts of these regulatory alternatives on water quality, waste water
generation and treatment, solid waste generation and treatment or disposal,
and energy consumption  or savings are discussed.
7.1  IMPACT ON ATMOSPHERIC  EMISSIONS
     Implementation of Regulatory Alternatives II, III, or IV, would
reduce VOC fugitive emissions from the SOCMI.  To quantify these reduc-
tions, the controlled VOC  emission levels from emission sources in the
model units (described in  Chapter 6) were estimated for each alternative.
These emission levels are  developed below for individual emission sources, for
model units in SOCMI,  and then for SOCMI as  a whole.
7.1.1  Emission Source Characterization   •
     As indicated in Chapter 6, a SOCMI model unit typically consists of
several types of process equipment that contribute to fugitive VOC emis-
sions.  Under Regulatory Alternative I (baseline case), all these sources
are "uncontrolled" emission sources.  However, if Regulatory Alternative
 II,  III,  or IV were  implemented, the emissions  from  some  uncontrolled
sources would be reduced;  these sources would subsequently become "con-
trolled"  sources.  Both the controlled and uncontrolled sources are
important  because the  total fugitive VOC emissions from the model units
and ultimately the SOCMI are  the sum of emissions from  both types of
sources.
                                   7-1

-------
7.1.2  Development of VOC Emission Levels .
     The uncontrolled emission levels were previously presented in Chapter
3 (Table 3-1).  Controlled emission levels were developed for those
sources that would be controlled by the implementation of a regulatory
alternative.  These controlled fugitive emission levels were calculated by
multiplying the uncontrolled emissions from this equipment by a "control
efficiency" presented in Chapter 4, Tables 4-2 through 4-4.  The resulting
controlled VOC emission factors for each source are presented in Tables
7-1, 7-2, and 7-3 for Regulatory Alternatives II, II, and IV, respectively.
     The total VOC fugitive emissions from Model A, Model B, and Model C
units in the SOCMI were determined under each regulatory alternative.
Initially, emissions from each source type within a model unit were
estimated by using the model unit equipment inventories presented in
Table 6-1 and the source emission factors presented in Tables 7-1, 7-2,
and 7-3.  These emissions were then used to estimate the VOC fugitive
emissions from each of the three model units.  An example calculation is
presented in Table 7-4 to illustrate the procedure used to estimate the
total VOC fugitive emissions from a model unit under Regulatory Alterna-
tive II.  The total VOC fugitive emissions calculated for the respective
model units under each regulatory alternative are presented in Table 7-5.
Also, presented in this table are the average percent reductions in the
baseline emission levels that result from implementing Regulatory Alterna-
tive II, III, or IV.  The incremental reductions in fugitive emission
levels achieved by implementing the alternatives are also presented in  .
Table 7-5.
7-1-3  Fujbure Impact on V_OC_FucjHive Emissions
     In order to assess the future impacts of the various regulatory
alternatives on VOC fugitive emissions from the SOCMI, the levels of
these emissions were estimated for a period of five years after adoption
of  a regulatory alternative.  These emissions were estimated by using:
     1)   the emission factors presented in Tables 7-1, 7-2,
          and 7-3;
     2)   the industry population for the assumed base year
          of  1980;
                                  7-2

-------
                TABLE 7-1.  EMISSION FACTORS FOR SOURCES CONTROLLED UNDER REGULATORY ALTERNATIVE II  '
I
CO
Uncontrolled
emission source
Pumps
Light liquid service
Valves
Gas service
Light liquid service
Safety/relief valves
Gas service
Compressors
Inspection9
interval
Yearly

Quarterly
Yearly
Quarterly
Quarterly
Uncontrollf
emission
factor,
kg/hr
0.

0.
0.
0.
0.
120

021
010
160
440
;d Correction
factors
Ac
0.87

0.98
0.84
0.69
0.84
Bd
0.80

0.90
.0.80
0.90
0.90
. ce
0.98

0.98
0.98
0.98
0.98
Df
0.92

0.99
0.94
0.97
0.97
Control
efficiency
(AxBxCxD)
0.

0.
0.
0.
0.
63

86
62
59
72
Controlled9
emission
factor,
kg/hr
0.

0.
0.
0.
0.
044

003
004
067
126
       aFrom  Table 6-2.
       bFrom  Table 3-1.
       theoretical maximum control efficiency.1
        Leak  occurrence and reoccurrence correction factor - assumed to be 0.80 for yearly inspection, 0.90
        for quarterly inspection, and 0.95 for monthly inspection.2
       eNon-instantaneous repair correction factor - for a 15-day maximum allowable repair time, the 7.5-day
        average repair time yields a 0.98 yearly correction  factor  [365 - (15/2)] * 365.3
        Imperfect repair correction factor - calculated as 1 - (f v F).  Where f = average emission rate for
        sources at 1000 ppm and F = average rate for emission sources  greater than 10,000 ppm.  *
       ^Controlled emission factor = uncontrolled emission factor x [1 -  (A x B x C x D)].

-------
          TABLE  7-2.   EMISSION  FACTORS  FOR  SOURCES  CONTROLLED UNDER REGULATORY  ALTERNATIVE III
Uncontrolled
emission source
Pumps
Light liquid service
Valves
Gas service
Light liquid service
Safety/relief valves
Gas service
Compressors
Inspection9
interval

Monthly

Monthly
Monthly

Monthly
Monthly
Uncontrolled
emission
factor,
kg/hr

0.

0.
0.

0.
0.

120

021
010

160
440
b
AC

0.87

0.98
0.84

0.69
0.84
Correction
factors
Bd

0.95

0.95
0.95

0.95
0.95
ce

0.98

0.98
0.98

0.98
0.98

Df

0.92

0.99
0.94

0.97
0.97
Control
efficiency
(AxBxCxD)

0.

0.
0.

0.
0.

75

90
74

62
76
Controlled9
emission
factor,
kg/hr

0.

0.
0.

0.
0.

030

002
003

061
108
 From Table 6-2.
bFrom Table 3-1.
C                                       6
 Theoretical maximum control efficiency-
 Leak occurrence and reoccurrence correction factor - assumed to be 0.80 for yearly inspection, 0.90
 for quarterly inspection, and 0.95 for monthly inspection.
 Non-instantaneous repair correction factor - for a 15-day maximum allowable repair time, the 7.5-day
 average repair time yields a 0.98 yearly  correction  factor [365 - (15/2)]  * 365. 8
 Imperfect repair correction factor - calculated as 1 - (f -r F).  Hhere f = average emission rate for
 sources at 1000 ppm and F = average rate for emission sources greater than 10,000 ppm.  '10
^Controlled emission factor = uncontrolled emission factor x [1 - (A x B x C x D)].

-------
                TABLE 7-3.  EMISSION FACTORS FOR SOURCES CONTROLLED UNDER REGULATORY ALTERNATIVE IV
-vl
I
01
Uncontrolled
emission source
Pumps
Light liquid service
Valves
Gas service
Light liquid service
Safety/relief valves
Gas service
Compressors
Sampling connections
Inspection3
interval

None

Monthly
Monthly

None
None
None
Uncontrolled
-- -emission
factor,
kg/hr

0.120

0.021
0.010

0.160
0.440
0.015
b

Ac
u
NAh

0.98
0.84

NA
NA
NA
Correction
"factors -
j
Bd

NA

0.95
0.95

NA
NA
NA
Q
ce

NA

0.98
0.98

NA
NA
NA
--•-••- --—-Control " •"
^ ell 1 L 1 cilLy
DT (AxBxCxD)

NA

0.99 0.90
0.94 0.74

NA
NA
NA
Controlled9
-'emission
factor,
kg/hr

0.0

0.002
0.003

0.0
0.0
0.0
                                              11
 From Table 6-2.
bFrom.Table 3-1.
°Theoretical maximum control efficiency.
 Leak occurrence and recurrence correction factor - assumed to be 0.80 for yearly inspection, 0.90 for
 quarterly inspection, and 0.95 for monthly inspection.12
eNon-instantaneous repair correction factor - for a 15-day maximum allowable repair time, the 7.5-day
 average repair time yields a 0.98 yearly correction factor [365 - (15/2)] -r 365.13
 Imperfect repair correction factor - calculated as 1  - (f 4- F).  Where f = average emission rate for
 sources at 1000 ppm and F = average rate for emission sources greater than 10,000 ppm.  '
^Controlled emission factor = uncontrolled emission factor x  [1 - (A x B x C x D)].
 Since the equipment associated with this regulatory alternative essentially eliminates fugitive
 emissions, these correction factors are not applicable.

-------
   TABLE 7-4.   EXAMPLE CALCULATION OF VOC FUGITIVE EMISSIONS FROM MODEL
               UNIT A UNDER REGULATORY ALTERNATIVE II
Number of
sources in
model unit9
(N)
Emission Source:0
Pumps
Light liquidd single
mechanical seal
Light liquidd double
mechanical seal
Heavy liquid6 single
mechanical seal
Heavy liquid6 packed seal
In-line valves
Vapor service
Light liquidd service
Heavy liquid6 service
Safety/ relief valves
Vapor service
Light liquidd service
Heavy liquid6 service
Open-ended valves
Vapor service
Light liquidd service
Heavy liquid6 service
Compressors
Sampling connections
Flanges



5

3

5

2

90
84
84

11
1
1

9
47
48
1
26
600
Total
Emission
factor, b
kg/hr-source
(E)


0.044

0.044

0.020

0.020

0.003
0.004
0.0003

0.067
0.006
0.009

0.003
0.004
0.003
0.126
0.015
0.0003
emissions
Emissions
from sources
kg/hr
(N x E)


0.220

0.132

0.100

0.040

0.270
0.336
0.025

0.737
0.006
0.009

0.027
0.188
0.014
0.126
0.390
0.180
2.800
 Model  units  are characterized in  Table  6-1.

 Emission factors from Tables  3-1  and  7-1.

cSources in VOC  service.

 Light  liquid  service means  that the fugitive  emission  source  contains  a
 liquid which  has a  vapor pressure equal  to or greater  than 0.3  kPa  at
 20°C.
"Heavy  liquid  service means  that the fugitive  emission  source  contains  a
 liquid which  has a  vapor pressure less  than 0.3  kPa  at 20°C.

 Open-ended valve factor  is  equivalent to the  in-line valve factor because
 capping the  open end is  assumed to eliminate  emissions from this source.
                                  7-6

-------
            TABLE 7-5.   ESTIMATED EMISSIONS  AND  EMISSION  REDUCTIONS  ON  A MODEL  UNIT  BASIS'
Regulatory
Alternative
I
II
III
l\l
Estimated emissions^ >c
(Mg/yr)
Model unit
ABC
67 260 800
24 94 290
21 80 250
8 34 106
Average percent
reduction from emissions
estimated under
Regulatory Alternative I
—
63
69
87
Average incremental
percent reduction
in emissions
—
63
6
18
 The emissions  and  percentage  reductions  presented  in  this  table were  calculated  using  the  following:
 •   controlled  and  uncontrolled  emission  factors  (see  Tables  7-1,  7-2,  and  7-3),  and
 •   emission  sources  given  in  Table  6-1.
3A  year is  assumed  to be  equivalent  to  8,760  hours.
C1.0 Mg/yr  =  2200 pounds/yr

-------
     3)  annual replacement of the industry population based on
         a twenty-year equipment life ;  and
     4}  annual growth rate of 5.9 percent for the industry.
Using these bases and the techniques presented in Appendix E, the total
number of affected facilities (model units) in operation in 1981  were
estimated to be 148.  In 1985 the total  number of affected facilities were
                    I Q
estimated to be 831.
     Under Regulatory Alternative I, the total VOC fugitive emissions from
affected facilities were estimated to increase from 35 to 199 gigagrams
per year (Gg/yr) during the same five-year (1981-1985) period (see Table 7-6).
In the same time period, implementation  of Regulatory Alternative II could
be expected to reduce the baseline case  (Regulatory Alternative I) fugitive
emissions 65 percent.  Implementation of Regulatory Alternative III would
reduce the baseline emissions by 69 percent.   As Table 7-5 indicates,
Regulatory Alternative IV, the most stringent of all the proposed alterna-
tives, would reduce the baseline emissions by about 87 percent.
7.2  IMPACT ON WATER QUALITY
     In the absence of standards to reduce fugitive emissions of VOC from
SOCMI and under normal equipment operation, liquid leaks from various
equipment components could increase the  quantity of wastewater generated
by a "typical" SOCMI facility.  Under Regulatory Alternative I, liquid leaks
could originate from pumps and process valves in light or heavy liquid
service as well as valves on open-ended  lines in light or heavy liquid service
and enter the wastewater system as runoff.  Although the uncontrolled emission
rates for these sources are given in Chapter 3, the gas-liquid split  of
these emissions i.s not defined.  Consequently, the increase in wastewater
from SOCMI due to liquid leaks from potential fugitive emission sources
cannot be quantified.
     Implementation of Regulatory Alternative II could reduce the wastewater
from a "typical" SOCMI facility by reducing the fugitive liquid emissions
resulting under Alternative I.  The reduced emissions would be due to the
use of caps, plugs or second valves on open-ended lines in gas and light or
heavy liquid service.  For example, caps, plugs, or second valves required
                                     7-8

-------
under Alternative II  would  reduce  the  VOC  fugitive  emission rate from open-
ended lines in light  or heavy  liquid service  from 0.01  kg/hr under Alternative
I to 0.004 kg/hr.   This reduction  would  reflect  a reduction in gaseous
emissions and liquid  leaks.  Since the gas-liquid split of the emission from
a given source is site specific, the impact of Alternative II on waste-
water from SOCMI  cannot be  quantified.   However, it is  likely that this
impact would be minor.
     Implementation of Alternative III would  result in  impacts on wastewater
from SOCMI similar to those resulting  from Alternative  II.  However,  the
impacts under Alternative III  would be more pronounced  due to the more
frequent inspection intervals  required by  this alternative.  The more
frequent intervals would reduce the VOC  fugitive emission rate from values
in light or heavy liquid service from  0.004 kg/hr under Alternative II to
0.003 kg/yr under Alternative  III.  Similarly, the  fugitive emission  rate
from pumps in light liquid  service would be 0.044 kg/hr under Alternative II
and 0.03 kg/hr under  Alternative III.  Consequently, the potential for
wastewater production by possible  fugitive emission sources in SOCMI  would
be less under Alternative III  than under Alternative II.
     Of the alternatives being considered, Regulatory Alternative IV
could have the greatest impact on  the  quality of water  that is discharged
from a "typical"  SOCMI facility.   Implementation of this alternative  could
have positive (and possibly some negative) impacts  on wastewater depending on
the specific control  device requirements at each unit.   Implementation of
Regulatory Alternative IV could reduce the amount of wastewater from a
SOCMI facility by reducing  the fugitive  liquid emissions resulting under
Alternative I.  The reduction  of these emission  levels  is primarily due to
the reduction of leaks from equipment  in light liquid service, e.g.,  from
the use of double mechanical seals for pumps  and closed loop sampling.  Under
Regulatory Alternative IV,  a double mechanical seal-degassing vent arrangement
reduces the emission rate of a pump seal in  light liquid service under
Regulatory Alternative I from 0.12 to  0.0  kg/hr.  A portion of this emission
reduction would be a reduction in  liquids  leaked to the ground or ditch.
However, the amount of liquids leaked  to the  ground or ditch that could enter
a plant wastewater system is not known.

                                    7-9

-------
                  TABLE 7-6.  TOTAL VOC FUGITIVE EMISSIONS FROM AFFECTED MODEL UNITS
                                      FOR  REGULATORY ALTERNATIVES
Year
1981
1982
I! 1983
o
1984
1985
Number of affected
model units9
A B C
77 49 22
158 100 46
244 155 71
335 213 97
432 274 125
Total fugitive emissions estimated
under Regulatory Alternative^5'0
I
(Gg/yr)
35.4
73.1
113.0
155
199
II
(Gg/yr)
12.9
26.7
41.2
56.5
72.8
III
(Gg/yr)
11.0
22.8
35.2
48.3
62.1
IV
(Gg/yr)
4.6
9.5
14.8
20.2
26.0
 The bases for estimating the number of model  units,-as detailed in Appendix E,  are:
 •  an industry growth rate of 5.9 percent per year,'
 •  unit replacement based on a 20-year equipment life,  and
 •  a base year (1980) total  of 872 Model  A, 554 Model  B,  and 252 Model  C Units.

 Estimated total  VOC fugitive emissions from Model  Units A, B, and C.
Q
 Does not include emissions from units in  existence prior  to 1981.

-------
     Implementation of Regulatory Alternative IV could also result in a
negative impact on water quality due to the operation of an ultimate con-
trol  device which "captures"  the fugitive VOC's.   If a carbon adsorption
device were used to capture any VOC released at  the degassing vent, a waste-
water containing suspended solids and some dissolved organics could be pro-
duced during the carbon regeneration process if  the carbon is regenerated
at the unit.  The use of a refrigeration process  as the ultimate control
device could possibly result in a condensate containing dissolved organics.
The wastewater flow rates would be quite small and would generally be suitable
for treatement in the existing unit wastewater treatment process.  Overall
the impacts, both positive and negative, of Alternative IV on wastewaters
from SOCMI would be minor.
7.3  IMPACT ON SOLID WASTE
     In the absence of standards to reduce fugitive emissions of VOC from
SOCMI and under normal operation, solid wastes  that could result from SOCMI
include replaced seals, packing, rupture disks,  equipment components such
as pumps and valves, spent catalysts, and polymerization products.  Metal
solid wastes such as mechanical seals,  rupture disks and disposed valves can
be sold as scrap metal to companies which can recycle the metal.  This would
help to minimize the impact on solid waste.  The quantity of used valve
packings and used batteries for monitoring instruments would not signifi-
cantly contribute to solid waste.
     Implementation of Alternatives II  and III would require the use of caps,
plugs, or second valves on open-ended lines in  light or heavy liquid service,
and more frequent monitoring intervals.   Implementing either of these
alternatives would have no greater impact on solid waste than Alternative I.
This is due to the relatively long life of caps,  plugs, and second valves on
open-ended lines as well as the ability to sell  discarded components such
as valves, mechanical seals, and rupture disks  as scrap metal.
     Implementation of Regulatory Alternative IV could res.ult in the
generation of solid waste if carbon adsorption were used as a control
device and the carbon disposed of instead of being regenerated.  However,
the carbon could be sent back to the manufacturer for regeneration, and
thereby reducing the.solid waste problem at the  facility.  It is antici-
pated that.the manufacturer could incinerate or  commercially dispose of
                                    7-11

-------
any carbon that could not be regenerated (such as carbon fines) without
any serious environmental problems.  Consequently, the negative impact of
implementing Alternative IV would be minor.
7.4  ENERGY IMPACT
     Regulatory Alternatives II, III and IV call for passive controls on
euipment handling VOC streams (i.e., pump seals, process vent enclosures,
degassing vents, etc.); so implementing any of these alternatives will not
significantly increase the energy usage of a typical SOCMI plant.  If a
control device such as carbon adsorption were used, steam (or another hot
regenerating medium) would be needed to regenerate the carbon at the unit;
however, the energy requirements would be quite small.  The energy require-
ments of vapor recovery systems and of closed loop sampling would also be small
Because fugitive emissions of VOC have an energy value, implementation of
any of the alternatives being considered will result in a positive energy
impact.
     The average energy value of the fugitive VOC emissions from SOCMI is
                                     6          19
estimated to be approximately 31 x 10  joule/kg.   The energy savings
resulting from the fugitive VOC emission reductions associated with
Alternatives II, III, and IV are presented in Table 7-7.  Because Alterna-
tive IV is the most stringent, it will result in the greatest emission
reduction.  As Table 7-7 indicates, implementation of this regulatory alter-
native would reduce the uncontrolled fugitive emissions by 173 Gg in the
fifth year and by a total of 520 Gg over a five-year period after implemen-
tation.  These "recovered" VOC emissions have a total energy value of
         13                                                    6
1.55 x 10... joules based on an average heating value of 31 x 10  joule/kg.
                                    6                             20
Assuming an energy value of 5.8 x 10  Btu per barrel of crude oil,  the
energy value of the total fugitive emissions recovered over the five-year
period is approximately equal to 2.5 million barrels of crude oil under
Regulatory Alternative IV.  This corresponds to an average daily savings of
1390 bbl/day of crude oil over the five-year period.
                                    7-12

-------
             TABLE 7-7.  ENERGY IMPACT OF EMISSION REDUCTIONS FOR REGULATORY ALTERNATIVES
Year
1981
1982
1983
1984
1985
5-year
total
Reduction from baseline Energy value of emission
emissions under reductions under Crude oil equivalent
Regulatory Alternatives, Regulatory Alternatives , of emission reductions,
Gga terajouleb thousand barrels
II III IV II III IV IIC IIIC IVC
22.4 24.4 30.8 694 756 955 113 124 . 156
46.4 50.3 63.6 1,440 1,560 1,970 235 255 322
71.8 77.8 98.2 2,230 2,410 3,040 364 394 497
98.3 106 135 3,050 3,290 4,180 498 538 683
127 137 173 3,940 4,250 5,360 644 695 876
366 396 500 11,350 12,270 15,500 1,855 2,005 2,530
 Estimated total  VOC fugitive emission reduction from Model  Units A, B, and C.


 Based on 1.55 x 10   joules/kg  :   This may be slightly over estimated if safety/

 relief valves are controlled by a  closed vent and flare system.


r                 6
 Based on 5.8 x 10  Btu/bbl crude oil.

-------
7.5  OTHER ENVIRONMENTAL CONCERNS
7.5.1  Irreversible and Irretrievable Commitment of Resources
     Implementation of any of the various alternatives is not expected to
result in any irreversible or irretrievable commitment of resources.  As
previously noted, the regulatory alternatives should help to save resources
due to the energy savings associated with the reductions in emissions.
7.5.2  Environmental  Impact  of  Delayed Standards
     As  it was  indicated above,  implementation of the standards will  only
have minor impacts  on water  and  solid wastes.  Consequently, delaying the
standards would have  essentially no  impact  on these problems.   However,  a
delay  in implementing the alternatives would have a greater  impact  on air
pollution and associated energy  losses.  The air and energy  impacts of
delayed  standards are shown  in Table 7-7.   The emission  reductions  and
associated energy savings shown  would be irretrievably lost  at  the  rates
shown  for each  of the five years.
7.6  REFERENCES
1.   Wetherold,  R.  G., L. P. Provost, D. D. Rosebrook, and C. D. Smith.
     Emission Factors and Frequency of Leak Occurrence for Fittings in'
     Refinery Process Units.  EPA Contract  No.  68-02-2665, 600/2-79-44.
2.   Tichenor,  B. A., K. C.  Hustvedt, and R. C.  Weber.   Controlling
     Petroleum  Refinery Fugitive Emissions  Via Leak Detection and Repair.
     Draft.  Symposium on Atmospheric Emissions from Petroleum  Refineries
     Austin, Texas.                                                      '
3.   Ref. 2.
4.   Ref. 1.
5.   Ref. 2.
6.   Ref. 1.
7.   Ref. 2.
8.   Ref. 2.
9.   Ref. 1.
                                    7-14

-------
10.    Ref.  2.

11.    Ref.  1.

12.    Ref.  2.

13.    Ref.  2.

14.    Ref.  1.

15.    Ref.  2.

16.    Letter from .Charles A.  Muela,  Radian Corporation, to K. C. Hustvedt,
      EPA Office of Air Quality Planning and Standards.  May 11, 1979.
      Replacement Rate of Process Unit in the Organic Chemical  Industry.

17.    The American Economy,  Prospects for Growth to 1991.  New York:
      McGraw-Hill, 1979."

18.    Letter from Vincent Smith, Research Triangle Institute, to
      Russell  L. Honerkamp,  Radian Corporation.   November 30, 1979.
      Projected Number of Affected Facilities and Average Product Value in
      SOCMI.

19.    Memo from J. R.  Blacksmith, Radian Corporation, to K. C.  Hustvedt,
      EPA.   December 14,  1979.   Average Energy Value of SOCMI Chemicals.

20.    American Petroleum Institute.   Petroleum Facts and Figures.
      Washington, D.C.  1971.

21.    Ref.  19.
                                   7-15

-------
                          8.   COST ANALYSIS

8.1   COST ANALYSIS OF REGULATORY ALTERNATIVES
8.1.1   Introduction
                                   c
     The costs of implementing the regulatory alternatives for controlling
fugitive emissions of volatile organic compounds (VOC) from the synthetic
organic chemicals manufacturing industry (SOCMI) are presented in the
following sections.  A detailed description of the model units and regula-
tory alternatives being considered in this analysis is presented in
Chapter 6.  Three different model units (A, B, and C) have been chosen to
represent the range of emission source populations that exist in SOCMI
units.  Regulatory Alternative I (no controls) is used as the baseline,
and Regulatory Alternatives II, III,  and IV are increasingly restrictive
control alternatives.
8.1.2  New Facilities
     8.1.2.1  Capital Costs.   The bases for the capital costs for the
model  units are presented in Table 8-1.  The capital cost estimates for
each model unit under each regulatory alternative are given in Table 8-2.
Regulatory Alternative I requires no control of VOC emissions.  Conse-
quently, there are no capital costs associated with this alternative.
     The capital costs for the model  units are the same under Regulatory
Alternative II as under Regulatory Alternative III, since the only change
is the monitoring frequency.  These costs reflect the purchase of two VOC
monitoring instruments and caps for all open-ended lines.  It is assumed
that one monitoring instrument is used as a standby spare.
     Under Regulatory Alternative IV, like II and III, two monitoring
instruments and caps for all open-ended lines are purchased.  In addition,
there are several other capital expenditures.  All single seal pumps in
                                  8-1

-------
                          TABLE  8-1.    CAPITAL COST  DATA
                    Item
   Value used in analysis
   (last quarter 1978 $)
1.  Monitoring instrument

2.  Caps for open-ended lines

3. . Double mechanical seals
4.  Flush oil  system for double mechanical  seals
5.  Closed vents for. degassing reservoirs  of  compressors
    and double seal pumps

6.  Rupture disks for relief valves

7.  Closed loop sampling connections
2 x 4250 = 8500/model  unit0

45/lineb

575/pump (new)0    H
850/pump (retrofit)

1500/pumpe>f

6530/compressor^
3265/pump3>h

1730/relief valve (new)1.
3110/relief valve (retrofit)J
                     k
460/sample connection
aOne instrument used as a spare (Ref.  1).   Cost  from  Ref.  2.

bBased on installation of a 2.5 cm.  screwed valve.  Cost  (1967)= $12 (Ref.  3,  p.  450).
 Cost index = 278.1 -f 113 (Ref. 4 and  5).   Installation =  1  hour  at  $15/hour (Ref. 6,7, 8).

cFrom Ref. 6, p.  IV-3.  Seal  cost =  $560.   Single  seal credit  = $225.  Shop installation = $240.

dFrom Ref. 10, p.  IV-3.  Seal  cost =  $560.   Field installation  = $290.

eFrom Ref. 11, p.  IV-3.  Pressurized  reservoir  system  = $700.   Flush  system cooler = $800.

 Pumps that have double mechanical seals without regulatory  requirement may not have the cost
 of a flush system added.  The flush system is assumed to  be an  integral  part of the double .
 seal system.                                                .

9From Ref. 12, pp. IV-8,9..  Based on  installation of a 122  m. length  of 5.1 cm. diameter,
 schedule 40 carbon steel pipe at a  cost of $5200; plus three  5.1 cm. ca'st steel plug valves
 and one metal gauze flame arrestor  at a cost  of $1330.   These costs  include connection of the
 degassing reservoir to an existing  enclosed combustion device or vapor recovery header.  Cost
 of a control device added specifically to control .the degassing  vsnts is, therefore, not
 included.

 This cost is based on the assumption that two pumps  (such as  a  pump and  its spare) are
 connected to a single degassing vent.
      of rupture disk assembly from Ref. 13,  p.  IV-8.   One  7.6  cm.  rupture disk,  stainless'
 steel = $195.  One 7.6 cm. rupture disk holder,  carbon steel  =  $325.  One  0.6 cm. pressure
 gauge, dial face = $15.  One 0..6 cm.  bleed  valve,  carbon  steel,  gate  =  $25.  Installation =
 $240. .In order to allow in-service disk replacement,  a block valve must De  installed
 upstream of the rupture disk.  Cost (1967)  from  Ref. 14, p.  451,  for one 7.6  cm. gate valve -
 $240.  Cost- index = 278.1 T 113 (Ref. 15 and 16).  Installation =  10 hours at  $15/hour
 (Ref .17,18,19).  In order to prevent damage to the relief valve  by disk fragments, an
 offset mounting is required.  Cost (1967) from Ref. 20, p.  450 for one 10.2 cm.  tee and
 one  10.2 cm. elbow = $7.. 30.  Cost index = 278.1  v  113 (Ref. 21 and 22).  Installation = 8
 hours at $15/hour (Ref -23,24,25).

JCosts for the rupture disk, holder, and block valve are the same as for the  new applications.
 An additional cost is added to replace the de-rated relief valve. No credit is assumed  for
 the  used relief valve.  Cost (1967) for one 7.6  cm.  pressure  reducing valve, stainless steel
 body and trim from Ref. 3, p. 452 = $500.  Cost  index = 278.1 v  113 (Ref .26  and 27).
 Installation = 10 hours at $15/hour (Ref. 28,29,30)'.
 k
 Based on  installation of a 6 m. length of 2.5 cm.  diameter,  schedule 40,  carbon  steel  pipe
 and  three 2.5 cm. carbon steel ball valves.   Costs from Ref.31 ,  p.  IV-8.   Installation =
 18 hours  at $15/hour.
                                              8-2

-------
     TABLE  8-2.   CAPITAL COST ESTIMATES  FOR NEW MODEL UNITS
                 (thousands of last quarter 1978 dollars)

Model
1.
2.
3.


4.
5.
6.
7.


8.

Model
1.
2.
3.


4.
5.
6.
7.


8.

Model
1.
2.
3.


4.
5.
6.
7.


8.


Capital cost item3 I
Unit A
Monitoring instrument
Caps for open-ended lines
Double mechanical seals
• Seals
• Installation
Flush oil system for double mech. seals
Vents for compressor degassing reservoirs
Vents for pump degassing reservoirs
Rupture disks for relief valves
• Disks
• 'Holders, block valves, installation
Closed loop sampling connections
Total 0.0
Unit B
Monitoring instrument
Caps for open-ended lines
Double mechanical seals
• Seals
• Installation
Flush oil system for double mech. seals
Vents for compressor degassing reservoirs
Vents for pump degassing reservoirs
Rupture disks for relief valves
• Disks
• Holders, block valves, installation
Closed loop sampling connections
Total 0.0
Unit C
Monitoring instrument
Caps for open-ended lines
Double mechanical seals
• Seals
• Installation
Flush oil system for double mech. seals
Vents for compressor degassing reservoirs
Vents for pump degassing reservoirs
Rupture disks for relief valves
• Disks
• Holders, block valves, installation
Closed loop sampling connections
Total 0.0
Regulatory alternative
II III IV

8.50 8.50 8.50
4.68 4.68 4.68

1.68
1.20
7.5
6.53
26.1

2.14
16.8
12.0
13.2 13.2 87.1

8.50 8.50 8.50
18.7 18.7 18.7

6.36
4.56
28.5
13.1
94.7

8.19
64.4
47.8
27.2 27.2 295

8.50 8.50 8.50
57.5 57.5 57.5

20.1
14.4
90.0
52.2
297

25.4
199
147
66.0 66.0 911
From Tables 6-1 and 8-1.
                                 8-3

-------
light liquid service must have double mechanical  seals installed.   This
is at a cost of $575/pump.  A flush oil  system ($1500/pump)  must also be
used in conjunction with the double mechanical  seals.   Existing pumps
with double mechanical seals are assumed to have  a flush oil system
already incorporated.  Hence, there is no additional, capital expenditure
for the double seals or flush system.
     Also, under Regulatory Alternative IV, compressor seals and pump
seals must have the seal oil degassing vents that are connected to a
control device such as a vapor recovery system or an enclosed combustion
device.  The cost is estimated to be $6530 per compressor and $3265 per
pump.  This cost is based on the assumption that  one closed  vent system
is required for each compressor.  Since main pumps and spares are
generally located in close proximity to each other, one closed vent
system is required for each pair of pumps.  These costs are  based on
connecting the closed vent system to an existing  control device.
     The costs of purchasing and installing rupture disks is $1590 per
relief valve.  The rupture disks are to be installed upstream of relief
valves in gas service.  The cost includes the purchase of a  shutoff
valve to allow the disk to  be replaced after overpressure relief.
     The closed loop sampling connection costs are based on  an estimate
of $460 per sampling connection for installation  of 6m. of pipe and three
valves.
     8.1.2.2  Annual Costs.  With the implementation of Regulatory
Alternatives II, III, or  IV, visual and/or instrument monitoring of
potential sources of fugitive VOC emissions will  be required.  A summary
of the requirements for the different alternatives is presented in
Chapter 6.  Tables 8-3, 8-4, and 8-5 give the monitoring labor-hour
requirements for Regulatory Alternatives  II, III  and IV, respectively.
The  labor-hour requirements are calculated by taking the product of the
number of workers needed  to monitor a component (1 for visual, 2 for
instrument), the time required to monitor, the number of components in
the  model unit, and the number of times the component is monitored per
year.  Monitoring labor costs are then calculated  based on  $15 per hour.'32»33>3£+
Regulatory Alternative  III  hasthe highest annual  monitoring costs.
                                 8-4

-------
                      TABLE  8-3.   ANNUAL  MONITORING AND  LEAK  REPAIR LABOR  REQUIREMENTS

                                     FOR REGULATORY  ALTERNATIVE  II
Monitoring
Number of
components per
model unit
Source type ABC
Pumps (light liquid)
Single mechanical 5 19 60
seals
Double mechanical 3 10 31
seals
Valves (in-line)
Gas 90 365 1117
Light liquid 84 335 1037
CD . Safety/relief valves 11 42 130
cj, (gas service)
Valves on open-ended
1 ines^1
Gas 9 37 115
Light liquid 47 189 581
Compressor seals 1 2 8
Monitoring
Type ofa time.b
monitoring min

Instrument 5
Visual 0.5
Instrument 5
Visual 0.5

Instrument 1
Instrument 1
Instrument 8



Instrument 1
Instrument 1
Instrument 10
Times
monitored
per year

1
52
1
52

4
1
4



4
1
4
Leak repair
Estimated
Monitoring labor- number of . Repair Leak repair labor-
hours required0 leaks per year time, hours required6
A B

1.0 3.2
2.2 8.2
1.0 1.7
V-l3'"* ( 4-3

12.0 49.0
2.8 11.2
11.7 44.8



1.2 4.9
1.6 6.3
1.3 2.7
C ABC

10.0 1 1 3
26.0
5.2 112
13.4

149.0 4 15 45
34.6 3 9 25
139.0



15.3 125
19.4 2 6 14
10.7 1 1 2
hrs ABC

80 b 80 80 240

80b 80 80 160


1.13f 4.5 17.0 50.9
1.13f 3.4 10.2 28.3
O9 0 0 0

	

1.13e 1.1 2.3 5.7
1.13e 2.3 6.8 15.8
40b 40 40 80
 '2 workers for instrument monitoring, 1  for visual.   Ref. 35,  p. 4-3.
uRef.  36.
^Monitoring labor-hours = number of workers a number of components a time to monitor  (total is minimum of 1 hr).
 From  Table 4-2.
p

 Leak  repair labor-hours =  number of leaks x repair time.



f
                                                                            h°UP
                                                                                                            °f the
                                        .

    is assumed  that these  leaks are corrected  by  routine maintenance at  no additional labor requirements.  Ref. 38.

-------
                              TABLE. 8-4..  ANNUAL MONITORING  AND  LEAK REPAIR LABOR  REQUIREMENTS
                                             FOR  REGULATORY  ALTERNATIVE . 111.
oo
 I
en
Monitoring
Number of
. components per
model unit
Source type ABC
Pumps (light liquid)
Single mechanical 5 19 60
seals
Double mechanical 3 10 31
seals
Valves (in-line)
Gas 90 365 11'17
Light liquid 84 335 1037
Safety/ relief valves 11 42 130
(gas service)
Valves on open-ended
1 inesh
Gas 9 37 115
Light liquid 47 189 581
Compressor seals 1 2 8
Type ofa
monitoring

Instrument
Visual
Instrument
Visual

Instrument
Instrument
Instrument



Instrument
Instrument
Instrument
Monitoring
time, b
min

5
0.5
5
0.5

1
1
8



1
1
10
Times
monitored
per year

12
52
12
52

12
12
12



12
12
12
Leak repair
Estimated
Monitoring labor- number of Repair Leak repair labor-
hours required0 leaks per yeard time, hours required6
A

10.0
2.2
6.0
1.3

36.0
33.6
35.2



3.6
18.8
4.0
B

38.0
8.2
20.0
4.3

146.0
134.0
134.4



14.8
75.6
8.0
C ABC

120.0 1 3 9
26.0
62.0 1 2 5
13.4 .

446.8 6 2'2 68
414.8 7 25 75
416.0



46.0 1 3 7
232.4 4 14 42
32.0 1 1 2
nrs A B C

80 b 80 240 720

80 b 80 160 400


1.13f 6.8 24.9 76.8
1.13f 7.9 28.3 84.8
O9 0 0 0



1.13e 1.1 3.4 7.9
1.13e 4.5 15.8 47.5 •
40b 40 40 • 80
       a2  workers for instrument monitoring, 1  for visual.  Ref. 39.
       bRef.  40.                                                                           .
       GMonitoring labor-hours  = number of workers x number of components  x  time to monitor (total is minimum of 1  hr] .
       dFrom Table 4-2.
       eLeak repair labor-hours = number of leaks  x repair time.
        Weighted average based  on 75 percent of the leaks repaired  on-line,  requiring 0.17 hour  per repair, and on  25  percent of the leaks repaired
        off-line, requiring  4 hours per repair.   Ref. 41  .
       ^It is assumed that these leaks are corrected by  routine maintenance  at no additional  labor 'requirements.   Ref. 42 .
        The estimated number of leaks per year  for open-ended valves  is  based on the same percent of sources used for  in-line valves.   This represents
        leaks occurring  through the stem and gland of the open-ended  valve.  Leaks through the seat of. the valve are eliminated by addinq caps  for
        Regulatory alternatives II, III, IV.                                                                            • '

-------
                            TABLE 8-5.   ANNUAL  MONITORING AND LEAK  REPAIR  LABOR REQUIREMENTS FOR
                                           REGULATORY  ALTERNATIVE  IV.
CO
I
Monitoring
Number of
components per
model unit
Source type ABC
Pumps (light liquid)
Single mechanical 5 19 60
seals converted to
double seals
Double mechanical 3 10 31
seals
Valves (in-line)
Gas 90 365 1117
Light liquid 84 335 1037
Safety/relief valves 11 42 130
(gas service)
Valves on open-ended
lines'* ••
Gas 9 37 115
Light' liquid 47 189 581
Compressor seals 1 2 8
Type ofa
monitoring

Instrument
Visual

Instrument
Visual

Instrument
Instrument
Instrument



Instrument
Instrument
Instrument
Monitoring
time, b
min

5
0.5

5 .
0.5

1
1
8



1
1
10
Times
monitored
per year

Of
52

Of
52

12
12
of



12
12
Of '
Monitoring labor-
hours requiredc
A

0
2.2

0
1.3

36.0
33.6
0



3.6
18.8
0
B

0
8.2

0
4.3

146.0
134.0
0



14.8
75.6
0
C

0
26.0

0
13.4

446.8
414.8
0



46.0
232.4
0
Estimated
number of
leaks per year''
A B

of of


of of


6 22
7 25
Of Of



1 3
4 14
Of O'f
C

of


of


68
75
Of



7
42
Of
Leak repair
Repair Leak repair labor-
time, hours required6
hrs .A B • C

80b 0 0 0


80b 000


1.13g 6.8 24.9 76.8
1.139 7.9 28.3 84.8
Of)h 0 0 0



1.139 1.1 3.4 7.9
1.139 4.5 15.8 47.5
40b ' 0 0 0
       2 worker's for instrument monitoring,  1  for visual.  Ref. 43.
      bRef. 44.
      "Monitoring  labor-hours = number of workers a number of components x time to monitor (total is a minimum of 1 hr)
       From Table  4-2.
      6
       Leak repair labor-hours = number of leaks x repair time.

      ^No monitoring or leak repair required because equipment specifications eliminate leak  potential

      Jfffine^reTu^^                'Kf!^! rePai>ed °n-line'  reqU1>in9 °'17 h°Ur P6r repai>' and on 25 Patent of the leaks repaired .

      .It is assumed that these leaks  are  corrected by routine maintenance  at no additional labor requirements   Ref  46

-------
     Leak repair labor is the  cost of  repairing  those components  in which
leaks develop after initial  repair.  The  leaks are  discovered  during  the
periodic monitoring required by the regulatory alternatives.   The number
of estimated leaks and the labor hours required  for repair  are given  in
Tables 8-3, 8-4, and 8-5.  Leak repair labor  is  calculated based on  $15  per
hour1/7,1*8,49 Maintenance labor costs are  greatest under Regulatory Alter-
native III and least under Alternative IV.  The  costs are  reduced under
Alternative IV because the required installation of double  mechanical
seals with seal oil degassing  vents eliminates the  most time-consuming
repair items.         .
     Administrative and support costs  are estimated at  40  percent of  the
sum of monitoring and leak repair labor costs.   Monitoring  labor, leak
repair labor, and administrative/support  costs are  the  recurring  annual
costs for each Regulatory Alternative.
     8.1.2.3  Annualized Costs.  The bases  for the  annual ized  control
costs are presented in Table 8-6.  The annualized capital,  maintenance,
and miscellaneous costs are calculated by taking the 'appropriate  factor
from Table 8-6 and applying it to the  corresponding capital  cost  from
Table 8-2.  The capital recovery factors  were calculated using the
equation:
                         CRF -           "
                               (1  + i)n - 1
where i = interest rate, expressed as a decimal,
      n = economic life of the component, years.
The interest rate used was 10 percent (last  quarter  1978).   The  expected  1
life of the monitoring instrument is 6 years compared to 10 years for other
control equipment components.  Double seals  and rupture disks are assumed
to have a 2 year life.
     The implementation of any of the Regulatory  Alternatives (except I)
will result in the initial discovery of leaking components.  It  is
estimated that fewer leaks will be found at  subsequent inspections.   The
cost of repairing initial leaks is amortized over a  10-year period,  since
this is a one-time cost.  Repair of leaks found at subsequent inspections
is included as a recurring annual  cost in 8.1.2.2.  The estimated
                                 8-8

-------
          TABLE  8-6.   DERIVATION  OF ANNUALIZED  LABOR,  ADMINISTRATIVE,
                      MAINTENANCE AND  CAPITAL CHARGES
1.   Capital  recovery  factor  for  capital
    charges

    *  Double seals  and  rupture disks
    '  Other  control equipment
    *  Monitoring instruments

2.   Annual maintenance  charges

    *  Control  equipment
    *  Monitoring instruments

3.   Annual miscellaneous  charges
    (taxes,  insurance,  administration)
    •  Control  equipment
    •  Monitoring instruments

4.   Labor charges

5.   Administrative  and  support costs  to
    implement regulatory  alternative

6.   Annualized charge for initial  leak
    repairs
0.58 x capital .
0.163 x capital'
0.23 x capital0
0.05 x capital
$2700e
0.04 x capitaU
0.04 x capital1"

$15/hour9

0.4 x (monitoring labor +
maintenance labor)n

^(estimated number of leaking
components per model unit1 x
repair time1) x $15/hr9 x 1.4"
x 0.163J
aApplies to cost of seals ($335)  and disk ($195)  only.   Two year life,
 ten percent interest.

 Ten year life, ten percent interest.  From Ref.  50, pp. IV-3,4.
cSix year life, ten percent interest.  From Ref.  51, pp. IV-9,10.

dFrom Ref. 52, pp. IV-3,4.

elncludes materials and labor for maintenance and calibration.  Cost (last
 quarter 1977) from Ref.  53, p.  4-2.  Cost index  = 221.7 ^ 209.1 (Ref. 54
 and 55).

fFrom Ref. 56, pp. IV-3,4,9,10.

Includes wages plus  40 percent  for labor-related administrative and
 overhead costs.  Cost  (last quarter 1977) from Ref. 6, pp. 4-4,5.   Cost
 index = 190.3 ^ 180.9  (Ref. 58  and 59).

hFrom Ref. 60, pp. IV-9,10.

""Shown in Tables 8-18,  8-19, 8-20.

JInitial leak repair amortized for ten years at ten percent interest.
                                 8-9

-------
percentage of initial  leaks per component is shown in Table  4-2.   This
percentage is applied to the number of components in the model  unit
under consideration.  Fractions are rounded up to the next integer,  since
in practice it is the whole .valve, or seal , that is replaced and  not just
part of one.  The time required to repair each component type is  given
in Table 8-7.  The initial repair cost is determined by taking  the product
of the number of initial leaks, the repair time, and the labor  rate, $15
per hour.' '   '  Forty percent is added for administrative and support
costs.  Finally, the total is multiplied by 0.163, the capital  recovery
factor.  As shown in Table 8-7, the cost of initial leak repair under
Regulatory Alternative IV is substantially less for each of the model
units than under Alternatives II and III.  The main reason for  this
reduction is the required installation of double mechanical  seals and
seal oil degassing vents that reduce the leak potential of pumps  and
compressors.  Although the total number of pumps and compressors  is  not
great, the repair time for a single pump or compressor seal  is  very much
greater than the repair time for a valve.
     8.1.2.4  Recovery Credits.  The annual VOC emissions, total  emission
reductions, and annual recovered product credits for each model unit
under each Regulatory Alternative are shown in Table 8-8.  Regulatory
Alternative I represents the uncontrolled emissions from each model  unit.
The annual emission reduction is calculated by subtracting the.controlled
emission factor from the uncontrolled emission factor for each source.
To obtain an annual rate, the result is multiplied by 8760 hours per year.
The recovery credit is figured at $360 per Mg of recovered product.
     8.1.2.5  Net Annualized Costs.  The net annualized costs,  shown in
Tables 8-9, 8-10, and 8-11, are determined by subtracting the annual
recovered product credit from the total cost before credit.  For example,
Model  Unit A, under Regulatory Alternative II has a net annualized credit
of $3300, as a  result of $12,100 in costs and $15,400 for recovery
credits.
     8.1.2.6  Cost  Effectiveness.  The cost effectiveness of each regula-
tory alternative for each model unit is shown in Table 8-12.   Regulatory
Alternatives  II and III have a net annualized credit for all model units,
and cost  effectiveness numbers are negative.  Since Regulatory Alternative
                                 8-10

-------
                                       TABLE  8-7.    LABOR-HOUR  REQUIREMENTS  FOR  INITIAL LEAK  REPAIR
oo
 i
Regulatory alternative II
Number of
components
per model
unit
Source type
Pumps (light liquid)
Single mechanical seal
Double mechanical seal
Valves (in-line)
Gas
Light liquid
Safety/relief valves3
(gas service)
Valves on open-ended lines
Gas
Light liquid
Compressor seals
A

5
3

90
84
11


9
47
1
B C

19 60
10 31

365 1117
335 1037
42 130


37 115
189 581
2 8
Estimated
number of
initial
leaks b
A

2
1

9
11
0


1
6
1
B C

5 14
3 8

37 112
41 125
0 0


4 12
23 70
1 3
Repair
time,
hrs

80C
80C

1.13d
1.13d
0


1.13d
1.13d
40C
Regulatory alternative III Regulatory alternative IV
Estimated
Labor-hours "^^f
required '»^b
ABC A

160 400 1120 2
80 240 640 1

10 42 127 9
12 46 141 11
0000


1 5 14 1
7 26 79 6
40 40 120 1
B C

5 14
3 8

37 112
41 125
0 0


4 12
23 70
1 3
Repair
time,
hrs

80C
80C

1.13d
1.13d
0


1.13d
1.13d
40C
Estimated
number of
Labor-hours initial.
required leaks
ABC ABC

160 400 1120 Oe Oe Oe
80 240 640 Oe Oe O6

10 42 127 9 37 112
12 46 141 11 41 125
00 0 Oe Oe Oe


1 5 14 1 4 12
7 26 79 6 23 70
40 40 120 Oe Oe Oe
Repair
time,
hrs

80C
80C

1.13d
1.13d
0


1.13d
1.13d
40C
Labor-hours
required
A

0
0

10
12
0


1
7
0
B

0
0

42
46
0


5
26
0
C

0
0

127
141
0


14
79
0
       alt is  assumed that these leaks  are corrected by routine maintenance at no additional labor requirements.   Ref.65

        Based  on  the percent of sources leaking at > 10,000 ppm.   From Table 4-2.
       cRef.66
        Weighted  average based on 75 percent of the leaks repaired on-line, requiring  0.17  hours per repair,  and on 25 percent of the  leaks  repaired off-line,
        requiring 4 hours per repair.   Ref. 67.

       eNo maintenance required  because equipment specification eliminates leak potential.
        The estimated number  of  initial leaks for open-ended valves is based on the  same percentage of sources used for in-line valves.  This represents  leaks
        occurring through the stem and  gland of the open-ended valve.  Leaks through the valve seat are eliminated by adding caps for  Regulatory Alternatives
        II, III,  IV.

-------
                                                TABLE  8-8.   RECOVERY  CREDITS.
Regulatory
alternative
I
II
III
co
-L iv

VOC
emissions,
Mq/yr
67.2
24.5
20.8
8.46
Model unit A
Emission reduction
from uncontrolled,
Mq/yr

42.7
46.4
58.7

Recovered3
product
value,
$/yr

15,400
16,700
21,100

VOC
emissions,
Mg/yr
257
93.7
79.8
34.3
Model unit B
Emission reduction
from uncontrolled,
Mg/yr
—
163
177
223

Recovered"1
product
value,
$/yr
--
58,800
63,800
80,200

VOC
emissions,
Mg/yr
800
293
249
106
Model unit C
Emission reduction
from uncontrol led,
Mq/yr
—.
507
551
694

Recovered3
product
value,
S/yr
--
182,500
198,400
249,800 •
Last quarter 1978 dollars.   Based on an average price of $36o/Mg.  Ref. 68.

-------
TABLE  8-9.  ANNUALIZED  CONTROL  COST ESTIMATES  FOR MODEL UNIT A
             (thousands  of last  quarter  1978 dollars).

Cost item . I
Annuali zed capital charges
1. Control equipment
a. Instrument
b. Caps
c. Double seals
• Seals
• Installation
d. Flush oil system
e. Vents - pumps and compressors
f. Rupture disks
• Disks
• Holders, etc.
g. Closed loop sampling
2. Initial leak repair
Operating costs
1. Maintenance charges
a. Instrument
b. Caps
c. Double seals
d. Flush oil system
e. Vents - pumps and compressors
f. Rupture disks •
g. Closed loop sampling
2. Miscellaneous (taxes, insurance,
administration)
a. Instrument
b. Caps
c. Double seals
d. Flush oil system
e. Vents - pumps and compressors
f. Rupture disks
g. Closed loop sampling
3. Labor
a. Monitoring labor
b. Leak repair labor
c. Administrative and support
Total before credit 0.0
Recovery credits 0.0
Net annual i zed cost 0.0 .
Regulatory
II


1.96
.763









1.06


2.70
.234







.340
.187






0.54
3.17
1.51
12.1
15.4
(-3.3)
alternative
III


1.96
.763









1.06


2.70
.234







.340
.187






2.26
3.30
2.22
15.0
16.7
(-1.7)

IV


1.96
.763

0.974
.196
1.22
5.32

. 1.24
2.74
1.96
0.10


2.70
.234
0.144
.375
1.63
.950
0.60


.340
.187
.115
0.30
1.31
.758
.48

1.43
0.304
0.692
29.0
21.1
7.9
 3Based on 40 percent of monitoring plus leak repair labor.   Ref.69.
 Based on an average price of $360/Mg.  Ref.70
 c(-xx) =*> net credit
                                 8-13

-------
TABLE 8-10.   ANNUALIZED CONTROL COST  ESTIMATES FOR  MODEL UNIT B
               (thousands of  last quarter 1978 dollars)

Cost item
Annual ized capital charges
1. Control equipment
a. Instrument
b. Caps
c. Double seals
• Seals
• Installation
d. Flush oil system
e. Vents - pumps and compressors
f. Rupture disks
• Disks
• Holders, etc.
g. Closed loop sampling
2. Initial leak repair
Operating costs
1. Maintenance charges
a. Instrument
b. Caps
c. Double seals
d. Flush oil system
e. Vents - pumps and compressors
f. Rupture disks
g. Closed loop sampling
2. Miscellaneous (taxes, insurance,
administration)
a. Instrument
b. Caps
c. Double seals
d. Flush oil system
e. Vents - pumps and compressors
f. Rupture disks
g. Closed loop sampling
3. Labor
a. Monitoring labor
b. Leak repair labor
c. Administrative and support
Total before credit
Recovery credits
Net annual ized cost0
Regulatory
I II


1.96
3.05









2.73


2.7
.935







0.34
.748






2.04
3.54 .
2.23
0.0 20.3
0.0 58.8
0.0 (-38.5)
alternative
III


1.96
3.05









2.73


2.7
.935







0.34
.748






8.75
7.69
6.58
35.5
63.8
(-28.3)

IV


1.96
3.05

3.69
.743
4.65
17.6

4.75
10.5
7.79
0.41


2.7
.935
.546
1.42
5.39
3.63
2.39


0.34
.748
.437
1.14
• 4.31
2.90
1.91

5.74
1.09
2.73
93.5
80.2
13.3
  aSee footnote from preceeding Table 8-9, Ref. 71.

   Based on an average
  c(-xx)=> net credit
Based on  an average price of $360/Mg.  Ref.72.
                                    8-14

-------
TABLE 8-11.   ANNUALIZED CONTROL COST ESTIMATES  FOR MODEL UNIT  C
               (thousands of  last quarter  1978 dollars)
Cost item
Annuali zed capital charges
1. Control equipment
a. Instrument
b. Caps
c. Double seals
• Seals
• Installation
d. Flush oil system
e. Vents - pumps and compressors
f. Rupture disks
• Disks
• Holders, etc.
g. Closed loop sampling
2. Initial leak repair
Operating Costs
1. Maintenance charges
a. Instrument
b. Caps
c. Double seals
d. Flush oil system
e. Vents - pumps and compressors
f. Rupture disks
g. Closed loop sampling
2. Miscellaneous (taxes, insurance,
administration)
a. Instrument
b. Caps
c. Double seals
d. Flush oil system
e. Vents - pumps and compressors
f. Rupture disks
g. Closed loop sampling
3. Labor
a. Monitoring labor
b. Leak repair labor
c. Administrative and support
Total before credit
Recovery credits
Net annual uzed cost0
Regulatory
I II


1.96
9.37









7.67


2.70
2.88







0.340
2.30






6.33
8.71
6.02
0.0 48.3
0.0 182.
0.0 (-134.)
alternative
III


1.96
9.37









7.67


2.70
2.88







0.340
2.30






27.14
21.3
19.4
95.1
198.
(-103.)

IV


1.96
9.37

11.7
2.35
14.7
56.9

14.7
32.4
24.0
1.23


2.70
2.88
1.72
4.50
17.5
11.2
7.35


0.340
2.30
1.38
3.60
14.0
8.98
5.88

17.7
3.25
8.38
283.
250.
33.0
aBased on 40 percent of monitoring plus leak repair labor. Ref.73

 Based on an average price of $360/Mg.  Ref. 74.

c(-xx) =i»-net credit.
                                  8-15

-------
                                    TABLE 8-12.   COST  EFFECTIVENESS  FOR MODEL  UNITS
                                                    (last  quarter  1978  dollars)
Model unit Aa
Regulatory alternative
Total capital cost ($1000)
Total annualized cost ($1000)
Total annual recovery credit ($1000)
oo . Net annualized cost ($1000)d
cr>
Total VOC reduction (Mg/yr)
Cost effectiveness ,
(annual $/Mg VOCr
I II
0.0 13.2
.0.0 12.1
0.0 15.4
0.0 (-3.3)
0.0 42.7
- (-77.3)
III
13.2
15.0
16.7
(-1.7)
46.4
(-36.6)
IV
87.1 •
.29.0
21.1
7.9
58.7
135.
Model unit Bb
I II
0.0 27.2
0.0 20.3
0.0 58.8
0.0 (-38.5)
0.0 163
- (-236.)
Ill
27.2
35.5
63.8
(-38.3)
177
(-160.)
IV
295
93.5
80.2
13.3
223
59.6
Model
I II
0.0 66.0
0.0 48.3
0.0 182.
0.0 (-134.)
0.0 507
- (-264.)
unit Cc
III
66.0
95.1
198.
(-103.)
551
(-187.)

IV
911
283
250.
33.0
694 .
47.6
 52 percent of the units  in the SOCMI are similar to Model Unit A.  Ref.  75.
b33 percent of the units  in the SOCMI are similar to Model Unit B.  Ref.  76.
C15 percent of the units  in the SOCMI are similar to Model Unit C.  Ref.  78.
. (-xx) = Control method net credit

-------
IV is the only one with a positive net cost,  comparisons  of cost
effectiveness in the normal  sense are meaningless.   The highest cost of
VOC control  under Regulatory Alternative IV is for  model  unit A.  Although
this cost ($128/Mg) is much  larger than the cost for model  unit C ($40/Mg),
the total annualized cost for model  unit A is only  $7500.  This amount is
insignificant compared to the annual  operating cost of the  process unit
itself.
8.1.3  Modified/Reconstructed Facilities
     8.1.3.1   Capital  Costs.  The bases for determining the capital  costs
for modified/reconstructed facilities are presented in Table 8-1.  The
capital  costs for these units are the same under Regulatory Alternatives
II and III as are those for  new units.  There are no costs  associated
with Alternative I.  The capital  costs for the monitoring instruments,
the caps for open-ended lines, the flush-oil  systems, the vents for
degassing reservoirs,  and the closed loop sampling  connections are also
the same as for new units.
     The estimated cost of retrofitting double mechanical seals for
single seal  pumps is estimated at $850 per pump.  This figure includes
$560 for a new double mechanical  seal plus $290 labor for field installa-
tion.
     Rupture disks for relief valves, required under Regulatory Alterna-
tive IV, are estimated to cost $2970 per relief valve.  The original
relief valve must be replaced with a larger relief valve.  Credit for
the removed valve is not included.
     The total capital cost estimates for modified/reconstructed facilities
are presented in Table 8-13.  As noted above, the costs associated with
Regulatory Alternatives I, II, and III are the same as for new units.
     8.1.3.2  Annualized Costs.  The annualized control costs for
modified/reconstructed units, presented in Table 8-14, are derived from
the same basis as new units  (see Table 8-2).   The only changes from new
unit costs occur under Regulatory Alternative IV because of the increased
capital  costs for double mechanical  seals and rupture disks.  The
recovered product credits for the modified/reconstructed units are the
same as for the new model units.
                                  8-17

-------
             TABLE  8-13.  CAPITAL COST ESTIMATES FOR MODIFIED/
                         RECONSTRUCTED FACILITIES
                 (thousands of  last quarter 1978 dollars)



1.
2.
3.
4.
5.
6.


Capital cost item3
Monitoring instrument
Caps for open-ended lines
Double mechanical seals
•. Seals
• Installation
Flush oil systems for double
mechanical seals
Vents for compressor degassing
reservoirs
Vents for pump degassing
Regul

A
8.5
4.68
2.8
1.45
7.50
6.53
26.1
atory alternative
Model unit
B
8.5
18.7
10.6
5.51
28.5
13.1
94.7
IVb

C
8.5
57.5
33.6
17.4
90.0
52.2
297
    reservoirs

7.   Rupture disks for relief
    valves
    . Disks                                2.14          8.19       25.4
    • Holders, block valves,              16.8          64.4       199
        installation    :                                    .
    • Replacement relief valve and         15.2          58.0       179
        installation

8.   Closed loop sampling connections       12.0     .     47.8       147

    Total                                 104          358       1107
aFrom Tables 6-1 and 8-1.

bFor Regulatory Alternatives I,  II,  III  the capital  costs for modified/
 reconstructed facilities  are the same as for new units  (Table 8-2).
                                8-18

-------
 TABLE 8-14.   ANNUALIZED  CONTROL  COST ESTIMATES  FOR  MODIFIED/
                 RECONSTRUCTED  MODFL UNITS  UNDER REGULATORY ALTERNATIVE IVa
                 (thousands  of  last  quarter 1978 dollars)

                                            Model  h         Model            Model  .
	Cost item	unit A	unit B	unit C

Annualized capital  charges

    1.   Control equipment

        a.  Instrument                         1.96            1.96             1.96
        b.  Caps                               .763          3.05             9.37
        c.  Double  seals
           • Seals                           1.62            6.15            19.5
           • Installation                      .236          0.898            2.84
        d.  Flush oil system                   1.22            4.65            14.7
        e.  Vents for pumps and compressors     5.32           17.6             56.9
        f.  Rupture disks
           • Disks                           1.24            4.75            14.7
           • Holders,  etc.                    2.74           10.5             32.4
           • Relief valves                    2.48            9.45            29.2
        g.  Closed loop sampling               1.96            7.79            24.0

    2.   In.itial leak repair                    0.10            0.41             1.23

Operating costs

    1.   Maintenance charges
a. Instrument
b. Caps
c. Double seals
d. Flush oil system
e. Vents for pumps and compressors
f. Rupture disks
g. Closed loop sampling
2. Miscellaneous (taxes, insurance,
administration)
a. Instrument
b. Caps
c. Double seals
d. Flush oil system
e. Vents for pumps and compressors
f. Rupture disks
g. Closed loop sampling
3. Labor
a. Monitoring labor
b. Leak repair labor
c. Administrative and support6
Total before credit
Recovery credits
Net annual i zed cost
Total VOC reduction (Mg/yr)
Cost effectiveness ($/Mg VOC)
2.70
.234
0.213
.375
1.63
1.71
.60


.340
.187
0.170
0.30
1.31
1.37
.480

1.43
0.304
0.692
33.7
21.1
12.6
58.7
215.
2.70
.935
0.806
1.42
5.39
6.53
2.39


.340
.748
.644
1.14
4.31
5.22
1.91

5.74
1.09
2.73
111.
80.2
30.8
223.
138.
2.70
2.88
2.55
4.50
17.5
20.2
7.35


.340
2.30
2.04
3.60
14.0
16.2
5.88

17.7
3.25
8.38
338.
250.
88.
694
127.
 aFor Regulatory Alternatives  I, II, III,  the annualized  control costs  and cost
  effectiveness for modified/reconstructed facilities  are the same as for new units
  (Tables 8-7, 8-8, 8-9).

  52 percent of existing units are similar to Model Unit  A.  Ref. 78.

 C33 percent of existing units are similar to Model Unit  B.  Ref. 79.

  15 percent of existing units are similar to Model Unit  C.  Ref. 80.

 eBased on 40 percent of monitoring plus  leak repair labor.  Ref. 81.

  Based on an average price  of $360/Mg.   Ref. 82.
                                       8-19

-------
     8.1.3.3  Cost Effectiveness.  The cost effectiveness figures for
modified/reconstructed facilities are also shown in Table 8-14.   The cost
effectiveness under Regulatory Alternatives I,  II, and III is the same as
for the new model units.  The cost effectiveness under Regulatory Alter-
native IV is a net cost of $134 per Mg for Model Units B and C,  and $208
per Mg for Model Unit A.
8.1.4  Projected Cost Impacts
     The regulatory alternatives are assumed to go into effect by 1981,
using 1980 as the base year.. The industry is estimated to grow at. a rate
of 5.9 percent.83 SOCMI facilities are estimated to be replaced at a rate
based on a 20-year equipment life (see Appendix E).  The estimated numbers
of projected new units are presented in Tables  7-5, 7-6, and 7-7.  The
estimated costs to the industry for the years 1981 through 1985 are
presented in Tables 8-15 through 8-17.  Capital costs shown are only for
units which begin operation in the indicated year.  All other costs shown
are for all units subject to NSPS in the indicated year.
                                    8-20

-------
PO
                   TABLE  8-15.   NATIONWIDE  COSTS  FOR  THE  INDUSTRY  UNDER  REGULATORY  ALTERNATIVE  II
                                            (last quarter 1978  dollars)
Cost item9
Total capital cost ($1000)b
Total annualized cost ($1000)
Total annual recovery credit
Net annualized cost ($1000)
1981
3,800
c 2,990
($1000) 8,070
(-5,080)
1982
4,040
6,160
16,700
(-10,500)
1983
4,280
9,530
25,800
'(-16,300)
1984
4,490
13,100
35,300
(-22,200)
1985
4,790
16,800
45,500
(-28,700)
      ^aFrom  Tables 8-2, 8-9, 8-10, 8-11.
        Capital  costs  for  model  units which  begin operation  in  the years  shown.
       GAnnualized costs for all model  units  subject  to NSPS  in the years shown.
        (-xx) => net credit

-------
ro
ro
                   TABLE 8-16.  NATIONWIDE COSTS FOR THE INDUSTRY UNDER REGULATORY ALTERNATIVE III
                                            (last quarter 1978 dollars)
Cost item3
Total capital cost ($1000)b
Total annual ized cost ($1000)
Total annual recovery credit
Net annualized cost ($1000)
1981
3,800
C 4,990
($1000) 8,770
(-3,780)
1982
4,040
10,300
18,100
(-7,800)
1983
4,280
15,900
28,000
(-12,100) .
. 1984
4,490
• 21,800
38,400
(-16,600)
1985
4,790
23,100
49,400
(-21 ,300)
       aFrom Tables 7-6,  8-7,  8-8,  8-9.
        Capital  costs for model  units which begin operation in the years shown.
       cAnnualized costs  for all  model  units subject to NSPS in the years shown.
        (-xx) => net credit

-------
                   TABLE 8-17.   NATIONWIDE COSTS FOR THE INDUSTRY UNDER REGULATORY  ALTERNATIVE  IV
                                           (last quarter 1978 dollars)
ro
oo
'Cost item3
Total capital cost ($1000)b
Total annual ized cost ($1000)°
Total annual recovery credit
($1000)
Net annual ized cost ($1000)d
1981
41,200
13,000
11,100
1,900
1982
44,000
27,000
22,900 .
4,100
1983
46,500
41,700
35,400
6,300
1984
48,700
57,100
48,500
8,600
1985
51,900
73,500
62,400
11,100
       Trom'Tables 7-6, 8-7, 8-8, 8-9.
        Capital  costs for model  units which begin operation in  the  years  shown.
       'Annualized costs for all  model  units subject  to  NSPS in the years shown.
        (-xx)=*> net credit

-------
8.2  OTHER COST.CONSIDERATIONS
     Environmental Safety and Health Statues which are applicable  to
SOCMI plants are listed in Table 8-18.   The provisions, requirements,  and
regulations, listed are those which may  cause an outlay of funds  by an
organic chemical  manufacturer.
     Specific costs of each of these provisions or requirements  to the
industry .defined as SOCMI were unavailable.  Total costs to SOCMI  for
complying with environmental, safety and health standards were also
unavai Table.
     The entire chemical industry is planning to spend an estimated $639
million on pollution control in 1979 according to a McGraw-Hill  Survey.8tf
Although this is a sizeable sum of money, the industry has enjoyed three  .
decades of rapid growth and high profits.  The economic health of  the
industry is better than that of many other industries.85  The substantial
pollution problems encountered in the industry and the large expenditures
necessary for this solution are expected to affect the smaller firms more
adversely than the larger firms.  However, few plant closings are  expected
due solely to costs of compliance with  standards and regulations.86
     The costs incurred by SOCMI in complying with all health, safety,
and environmental requirements are not  expected to prevent compliance
with the proposed NSPS for fugitive emissions.
                                 8-24

-------
                                     TABLE  8-18.   STATUTES  THAT MAY  BE  APPLICABLE TO SOCMI
                       Statute
   Applicable provision, regulation or
  	requirement of statute	
                                                                                                     Approximate cost  incurred due to
                                                                                                           enactment of  statute
Model unit
            Industry
             Clean  Air  Act and Amendments
State implementation plans
National emission standards for hazardous
  air pollutants
    Benzene fugitive emissions
New source performance standards
    Air oxidation
    Volatile organic liquid storage
PSD construction permits
Non-attainment construction permits
Total
$249 million3
             Clean Water Act  (Federal
               Water  Pollution Act)
oo
ro
en
Discharge permits
Effluent limitations guidelines
New source performance standards
Control of oil spills and discharges
Pretreatment requirements
Monitoring and reporting
Permitting of industrial projects  that
  impinge on wetlands or public waters
Environmental impact statements
Total
$414 million0
            Resource Conservation and
               Recovery Act
Permits for treatment,  storage,  and
  disposal of hazardous wastes
Establishes system to track  hazardous
  wastes
Establishes recordkeeping, reporting,
  labelling and monitoring system for
  hazardous wastes
Superfund
Total
$200 million0
Superfund-less than 2* of profits
  or $200 million maximum annual
  rate on petrochemical
  feedstock^
Production costs for the industry
  are expected to increase by an
  average of 0.6%'and a maximum
  of 5%.e
            Toxic Substances Control
              Act
Premanufacture notification
Labelling, recordkeeping
Reporting requirements
Toxicity testing
Total
$100-200 million per year
Preinventory notification cost:
  $1200-1500 per chemical
                                                                                                              (Continued)

-------
                                                               TABLE  8-18.    (Cont.
                               Statute
                                     Applicable  provision,  regulation or
                                           requirement  of statute 	
                                                                                                             Approximate cost incurred due to
                                                                                                                   enactment of statute
                                                 Model  unit
            Industry
                    Occupational  Safety  and  Health
                      Act
                                   Walking-working  surface standards
                                  • Means  of  egress  standards
                                   Occupational  health  and environmental
                                     control  standards
                                   Hazardous  material standards
                                   Personal  protective  equipment  standards.
                                   General environmental  control  standards
                                   Medical and  first aid  standards
                                   Fire protection  standards
                                   Compressed gas and compressed  air
                                     equipment
                                   Welding,  brazing, and  cutting  standards
                                                Total
S220/year per worker
CO
 i
rv>
cr>
                    Coastal  Zone  Management  Act
Power Plant and Industrial
  Fuel  Use Act
                                   States  may  veto  federal permits for
                                     plants  to be sited  in coastal zone
Prohibits new, major, industrial  power-
  plants which utilize fuel  oil  or
  natural gas
                    National  Environmental Policy
                      Act
                                   Requires  environmental  impact statements
                    Safe  Drinking Water Act
                                   Requires  underground  injection control
                                     permits
                   Marine  Sanctuary Act
                                   Ocean pumping permits
                                   Recordkeeping and reporting
                   Expenditure, by entire chemical industry, on air pollution control;. SOCMI's portion of expenditure not delineated.  (Ref.87  )
                    Expenditure, by entire chemical industry, on water pollution control; SOCMI's portion of expenditure not  delineated.  (Ref.83
                   cCost reflects entire organic industry; SOCMI's cost not delineated.   (Ref.89 ) .       '    •
                    Cost reflects entire organic industry; SOCMI's cost not delineated.   (Ref.90,91)
                   eCost reflects entire organic industry; SOCHI's cost not delineated.   (Ref.  92)        .  .
                    Cost incurred by entire chemical industry; SOCMI's portion, of expenditure not delineated..  (Ref.  93)
                    Cost incurred by entire chemical industry; SOCMI's portion of expenditure not delineated.   (Ref.  94)
                   nCost incurred by entire chemical industry; SOCMI's portion of expend•Ituve not delineated.   (Ref.  ?3)

-------
 8.3   REFERENCES

 1.     Erikson,  D.  G.,  and  V.  Kalcevic.   Emissions Control  Options for
       the Synthetic  Organic  Chemical  Manufacturing Industry, Fugitive
       Emissions Report.  Draft Report.   EPA Contract No.  68-02-2577.
       Knoxville, Tennessee,  Hydroscience,  Inc., March 1979.   p. IV-9.

 2.     Letter from Guy  C. Amey, Century  Systems Corporation,  to James C.
       Seme, PES,  Inc.   October 17,  1979.   Cost data for VOC monitoring
       instrument.

 3.     Peters, Max S.,  and  K.  D.  Timmerhaus.  Plant design and Economics for
       Chemical  Engineers.  Second Edition.   New York, McGraw-Hill.  1968.

 4.     Kohn, P.  M.   CE  Cost Indexes Maintain 13-Year Ascent.   Chem.  Eng.
       18(11):189-190.   May 1978.

 5.     Economic Indicators.  Chem. Eng.   Vol. 87 #1.  January 14, 1980.

 6.     Ref. 5.

 7.     Environmental  Protection Agency,  Chemical and Petroleum Branch.
       OAQPS Guideline  Series.  Control  of Volatile Organic Compound Leaks
       from Petroleum Refinery Equipment.  EPA-450/2-78-036,  OAQPS
       No. 1.2-111.  June 1978.  p. 4-5.

 8.     Economic Indicators.  Chem. Eng.   Vol 86 #2.  January 15, 1979.

 9.     Ref. 1, p. IV-3.

10.     Ref. 1, p. IV-3.

11.     Ref. 1, p. IV-3.

12.     Ref. 1, pp.  IV-8, 9.

13.     Ref. 1, p. IV-8.

14.     Ref. 3, p. 451.

15.     Ref. 4.

16.     Ref. 5.

17.     Ref. 5.

18.     Ref. 7.
                                   8-27

-------
19.     Ref.  8.

20.     Ref.  3,  p. 450.

21.     Ref.  4.

22.     Ref.  5.

23.     Ref.  5.

24.     Ref.  7.

25.     Ref.  8.

26.     Ref.  4.

27.     Ref.  5.

28.     Ref.  5.

29.     Ref.  7.

30.     Ref.  8.

31.     Ref.  1,  p. IV-8.

32.     Ref.  5.

33.     Ref.  7.

34.     Ref.  8.

35.     Ref.  7,  p. 4-3.

36.     Letter with Attachments from J. M. Johnson, Exxon Company, U.S.A.,
       to Robert T. Walsh, U. S. EPA.  July 28, 1977.

37.     Ref.  1,  p. B-12.

38.     Ref.  36.

39.     Ref.  7,  p. 4-3.

40.     Ref.  36.

41.     Ref.  1, p. B-12.

42.     Ref.  36.
                                     8-28

-------
43.    Ref. 7, p. 4-3.

44.    Ref. 36.

45.    Ref. 1, p. B-12.

46.    Ref. 36.

47'.    Ref. 5.

48.    Ref. 7.

49.    Ref. 8.

50.    Ref. 1, pp. IV-3, 4.

51.    Ref. 1, pp. IV-9, 10.

52.    Ref. 1, pp. IV-3, 4.

53.    Ref. 7, p. 4-2.

54.    Ref. 5.

55.    Ref. 8.

56.    Ref. 1, pp. IV-3, 4, 9, 10.

57.    Ref. 7, pp. 4-4.  5.

58.    Ref. 5.

59.    Ref. 8.

60.    Ref. 1, pp. IV-9, 10.

61.    Ref. 5.

62.    Ref. 7.

63.    Ref. 8.

64.    Letter from Vincent Smith, Research Triangle  Institute  to  Russell
       L. Honerkamp, Radian Corporation.  November 30,  1979.   Projected
       Number of Affected Facilities and Average Product Value in SOCMI.

65.    Ref. 36.

66.    Ref. 36.
                                   8-29

-------
67.    Ref. 1, p. B-12.

68.   .Ref. 64.

69.    Ref. 1, pp. IV-9, 10.

70.    Ref. 64.

71.    Ref. 1, pp. IV-9, 10.

72.    Ref. 64.

73.    Ref. 1, pp. IV-9, 10.

74.    Ref. 64.

75.    Ref. 1, p. IV-1.

76.    Ref. 1, p. IV-1.

77.    Ref. 1, p. IV-1.

78.    Ref. 1, p. IV-1.

79.    Ref. 1, p. IV-1. .

80.    Ref. 1, p. IV-1.

81.    Ref. 1, pp. IV-9, 10.

82.    Ref. 64.

83.    Letter  from Vincent Smith, Research Triangle Institute, to Russell
       L. Honerkamp, Radian Corporation.  August 13, 1979.  Growth Rate of
       SOCMI.

84.    News Flashes.  Chemical Engineering, Vol. 86,. No. 12.  1979.  p. 77.

85.    Environmental Quality, The Ninth Annual^Report of the Council on
       Environmental Quality.  U.S. Government Printing Office, Washington,
       D.C.  December 1978.

86.    Ref. 85.

87.    Ref. 85.

88.    Ref. 85.

89.    Solid Waste Facts, A Statistical Handbook.  U. S. Environmental
       Protection Agency, Office of Public Awareness.  U. S. Government
       Printing  Office, Washington, D. C.  August 1978.


                                    8-30

-------
90.  EPA Charges Chemical  Trade Seeks Lowest Denominator.as its Position
     on Superfund.  Chemical Marketing Reporter.   N.Y. (216) 10.  McGraw-
     Hill, Sept. 3, 1979.  p.3.

91.  Tough Version of Superfund Would Cook Industry, $1.6 Billion for
     Cleanup.  Chemical  Marketing Reporter.   N.Y. (215) (25).
     McGraw-Hill, Jan.  18, 1979.

92.  Ref. 85.

93.  Preproposal of Premanufacture Notification Notice Form and Provision
     of Rules 40 CFR Part 720.   44(201) Oct. 16,  1979.

94.  Ref. 93.

95.  Cost of Government Regulation Study.  Arthur Anderson and Co.,
     Washington, D.C.  March 1979.
                                    8-31

-------
                           9.  ECONOMIC ANALYSIS
9.1  Industry Profile
9.1.1  Introduction
       The  synthetic  organic  chemicals manufacturing  industry  (SOCMI)  has
been defined as  the  producers  of 378  synthetic  organic  chemicals, listed in
Appendix  F.  This  profile  gives  a general  qualitative description of the in-
dustry,  supported  by  quantitative   information  wherever  possible.   Because
SOCMI does not directly correspond to industrial classifications used for re-
porting information by secondary data sources, a weighting technique was used
to develop industry statistics (see Appendix El).
     Synthetic organic  chemicals (SOCs) are substances  containing  at least
carbon  and  hydrogen.   They  exhibit three  basic molecular structures:   ali-
phatic or acyclic,  cyclic,  and combinations of aliphatic and cyclic.   Acyclic
compounds are  composed groups  of atoms arranged  in a  straight chain.  Ex-
amples  are  alcohols,  ethers,  ketones, and  carbohydrates.   Cyclic compounds
have  the atoms  of their component  elements arranged  in the  form  a  closed
ring.   Examples  include  aromatic   hydrocarbons,  napthenes,  and thiazoles.
Certain amino acids and terpene hydrocarbons represent combinations of  cyclic
and aliphatic molecular structures.1
     SOCMI  chemicals  may  be used as primary feedstocks,  chemical intermedi-
ates,  or  end use chemicals.   Primary  feedstocks  are produced from crude raw
materials and  used in the  manufacture of other chemicals.  Chemical  interme-
diates  are  the  product  of  primary  feedstocks  and  are  also used to produce
other  chemicals.   End use  chemicals are  products  of chemical  intermediates
and/or  primary feedstocks  and  are used either as final goods  or  as inputs to
production  processes  outside the chemical  industry.   Many synthetic organic
chemicals are  used in more  than one of these  categories.   Figure 3.1 illu-
strates   the  general  relationships  among  the  various   organic chemicals.
                                    9-1

-------
Detailed flow charts identifying inputs and product uses for many of the SOCMI

chemicals have been presented elsewhere.2

9.1.2  Production Processes and Capacities

       Most of the  SOCMI  chemicals produced in the United States are derived

from crude  petroleum and  natural  gas.   Oil, shale, coal,  and  biomass (non-

prehistoric plant  tissue) are  also  sources of primary  feedstocks.3   A wide

variety of processes are used to  manufacture the 378 synthetic organic chem-

icals  included  in  the definition of SOCMI.   Frequently individual  chemicals

can  be manufactured  in several different  ways.   Consequently,  as relative

prices  change,  chemical producers may alter the  mix of  primary feedstocks

used to produce SOCs.

     After chemical  feedstocks  are manufactured from petroleum, natural gas,

and  other  raw materials,  they are processed  into  chemical  intermediates and

end  use  chemicals.   Some  of the chemicals  included  in  SOCMI are the product

of a simple  distillation  process,  while others are produced from a series of

cracking processes.

     The 1976  Organic  Chemical  Producers Data Base4 reports 1,270 units pro-

ducing SOCMI chemicals  in the United States.*  Table 9-1 presents a distribu-

tion of  those units and estimated capacity by state.  New Jersey, Texas, and

California have the  largest number of units producing SOCMI chemicals.  Texas

and  Louisiana have the  largest total  production capacities.  These states are

major producers because of their petroleum deposits and good sea port facili-

ties.   Table  9-2 presents  a geographical distribution  of  units by reported

capacity.  Approximately  12  percent  of these units  produce fewer than 5,000

Mg.  Another  12  percent of the units have production capacities in excess of
*The 1976 version of the Organic Chemical Producers Data Base is used because
it was the most recent version available.
                                    9-2

-------
    TABLE 9-1.   ESTIMATED ANNUAL  PRODUCTION CAPACITY
                         BY  STATE,  19764
State
Total
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawai i
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New York
North Carolina
Ohio
Oklahoma
Oregon
Pennsylvania
Puerto Rico
Rhode Island
South Carolina
Tennessee
Texas
Utah
Vermont
Virgin Islands
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Number
of units
1,270
25
2
1
12
120
5
18
14
14
20
1
2
85
31
11
1
27
54
1
17
27
28
6
15
16
1
4
2
5
131
52
50
88
9
17
75
13
8
27
24
126
4
1
2
22;
13
24
18
1
Percentage of
units reporting
capacity
40
52
50
00
33
' 23
20
26
50
43
30
- 00
50
31
26
64
00
56
74
00
35
48
25
17
47
37
100
100
50
40
24
19
38
26
78
53
27
85
00
48
46
76
25
100
100
68
46
63
22
100
Estimated total .
capacity,
(103 Mg)
319,835
5,174
399
91
1,982
19,650
644
2,765
2,031
3,257
3,459
91
• 97
16,517
3,551
1,698
390
6,062
31,810
390
2,160
4,835
9,735
574
1,999
4,072
222
103
122
483
28,070
10,586
7,283
14,576
702
3,838
14,634
7,259
815
3,875
6,809
77,189
628
2
643
3,581
2,502
9,242
3,514
24
Capacities were estimated by calculating  the mean of reported  unit capacity
for each chemical.   This was  substituted  for  any  missing  values  of unit
capacity  for  each chemical.   If no units  reported capacity  for  a chemical,
then the  mean of all  chemicals was substituted for the missing  value.  Esti-
mated capacity  represents the sum of reported-capacities,  means  of reported
capacity  for  some  specific  chemicals,  and  industry  mean  reported capacity
for other chemicals.
                                9-3

-------
                     TABLE 9-2.  DISTRIBUTION OF UNITS BY UNIT CAPACITY AND REGION, 19764
Number of
units

Unit capacity ranges (103 Mg)

Region
North east
New England
Mid-Atlantic
North central
East
West
South
East south
central
West south
central
South
Atlantic
West
Mountain
Pacific

0-5
17
7
10
16
13
3
23

2

4

17
3
0
3

5-10
5
0
5
5
3
2
7

3

3

1
2
1
1

10-25
12
4
8
13
10
3
34

12

10

12
17
3
14

25-50
16
5
11
13
9
4
38

8

18

12
12
2
10

50-100
17
4
13
13
9
4
44

5

20

19
5
0
5

100-250
10
0
10
17
16
1
50

9

24

17
8
0
8

250-500
5
1
4
6
6
0
34

3

24

7
3
0
3

500+
1
0
1
3
2
1
50

4

44

2
0
0
0

Units
reporting
capacity
83
21
62
86
68
18
280

46

147

87
50
6
44

Units not
reporting
capacity
235
39
196
202
182
20
200

45

54

101
119
10
109


Total
units
318
60
258
288
250
38
480

91

201

188
169
16
153
Total
59
19
76
79
79
85
48
54
499
756
1,255

-------
500,000 Mg.   Seventy-five  percent of  these large facilities are  located in

the southwest central region  of  the  United States, which  includes Texas and

Louisiana.   Table 9-3  presents  the  total  reported capacity  for  each region

by unit size.

9.1.3  Production and Sales

     Production and sales data for the SOCMI are presented in Table 9-4.   The

production of SOCMI  chemicals  increased from 58,050 Gg  in 1968 to 84,530 Gg

in 1978, at an average annual  growth rate of approximately 3.5 percent.  How-

ever,  output  levels have  fluctuated  widely since 1974.   The effects of the

oil embargo, the increase in energy and feedstock prices, and the sharply re-

duced demand resulting from a major economic recession caused the industry to

cut  back  production  by  13.2  percent  in  1975.   In  1976 output  rose  only

slightly,   but,  in 1977, as  real  prices for energy and  feedstocks fell, the

economy recovered,  and the need to increase inventories  became urgent,  pro-

duction increased by 50.4 percent.  In 1978 energy and feedstock prices began

to increase again  and  the need to replenish inventories disappeared.  Output

declined  that year  by 28.6 percent.    Nevertheless,  production  in  1978 was

greater than  in 1974, suggesting that  the  industry  may  have  substantially

adjusted to the shocks experienced in 1974 and 1975.

     Sales  and production  trends were  virtually identical  over  the period

1967-1978.*   The two variables are likely to remain highly correlated in the

future, because the  industry's feedstock requirements are  closely tied to its

production levels.  The absolute  level of sales was much lower  than the  level

of production (45.6 percent of production) over the period 1967 to 1978.   The
*The estimated correlation coefficient for the two variables over this period
is 0.97.
                                    9-5

-------
TABLE 9-3.   DISTRIBUTION OF INDUSTRY CAPACITY BY UNIT CAPACITY AND REGION, 19764
Industry capacity
Unit capacity
Region
North east
New England
Mid-Atlantic
North central
East
West
South
DEast south
^ central
West south
central
South Atlantic
West
Mountain
Pacific
Caribbean
Total
0-5
37.7
19.1
18.6
42.2
38.1
4.1
62.9
1.3
12.2
49.4
5.0
5.0
--
147.8
5-10
43.1
43.1
38.8
22.5
16.3
58.8
23.6
28.8
6.4
13.2
6.4
6.8
--
153.9
10-25
199.2
70.8
128.4
205.9
142.4
63.5
605.6
204.6
194.6
206.4
298.4
48.5
249.9
—
1,309.1
25-50
548.4
176.9
371.5
478.1
350.2
127.9
1,442.0
299.4
710.8
431.8
477.2
72.6
404.6
109.8
3,055.5
50-100
1,177.6
263.1
914.5
953.0
664.1
288.9
3,236.1
316.2
1,532.3
1,387.6
342.5
342.5
72.6
5,781.8
(103 Mg)
100-250
1,392.6
1,392.6
2,544.7
2,406.8
137.9
8,369.
1,494.6
4,411.7
2,463.5
1,055.1
1,055.1
--
13,362.2

150-500
1,811.2
299.4
1,511.8
2,024.9
2,024.9
11,910.9
984.5
8,596.6
2,329.8
1,024.2
1,024.2
1,324.1
18,095.3

^500
742.5
742.5
2,430.8
1,906.9
523.9
67,600.5
3,132.6
63,197.8
1,270.1
--
5,659.1
76,432.9

Total
5,952
829
5,123
8,718
7,556
1,163
93,287
6,457
78,685
8,145
3,216
127
3,088
7,166
118,339

-------
   TABLE 9-4.   ANNUAL  PRODUCTION  AND  SALES  OF  SYNTHETIC ORGANIC CHEMICALS'
Year
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
Production3
(Gg)
51,380
58,050
65,210
68,140
69,020
76,740
81,220
83,720
72,660
76,030
114,320
84,530
Sales volume3
(Gg)
23,440
26,960
30,360
32,090
33,020
36,930
39,420
38,450
32,920
32,520
49,470
35,310
Sales value3
($106)
3,085.55
3,411.91
3,590.07
3,702.20
3,724.03
4,173.97
4,991.53
9,357.99
8,411.34
10,187.76
15,317.72
12,951.16
Average
unit value3
($Ag)
0.13
0.13
0.11
0.12
0.11
0.11
0.13
0.24
0.26
0.29
0.31
0.36
3See Appendix El for a discussion of the methodology used to compute these
 data.
                                    9-7

-------
difference between  output  and  sales represents captive consumption,  indicat-



ing that  the  industry has  a relatively high degree  of vertical  integration.



9.1.4  Resource Use



       Estimates of  employment,  assets,  cost of materials and energy used in



SOCMI from 1972 to 1976 are presented in Table 9-5.   In general,  resource use



increased with  production.   Total  industry employment, including administra-



tive, clerical, marketing  and  service employees as well  as  production work-



ers, increased 5 percent from 1972 to 1976.  Employment of production workers



increased 4.1  percent during this period, although the  number  of production



workers declined during  the adjustment period following  the  1973-74 oil  em-



bargo.   Cost  of materials  increased substantially during the 1972-1976 peri-



od;  however,  much  of this increase can be  attributed to rapidly increasing



prices  of crude  petroleum.  Consumption  of energy for  heat and power  has



fluctuated, resulting in  an overall decrease of about 1 percent from 1972 to



1976.  Value  of assets increased each year from 1972 to 1976.  The total in-



crease during  that period  was approximately 53 percent, much of which can be



accounted for  by  changes in the value of buildings and equipment.  The stock



of physical assets  increased at a much slower rate.



9.1.5  Consumption



       The chemicals in  SOCMI  have a wide variety of  end uses as fuels, sol-



vents,  pesticides,  and  pigments,  and as  feedstocks  for the  production of



plastics,  synthetic  fibers  and  textiles,  soaps  and  detergents,  rubber pro-



ducts, medicines and fertilizers.   It  is not possible  to estimate consistent-



ly  apparent  consumption,  because  import and  export data presented  in Table



9-6  for SOCMI  are  not compatible with  the production and  sales data presented



in Table  9-4.   However,  it  is probable that  historical consumption trends have
                                    9-8

-------
                       TABLE  9-5.   SOCMI  RESOURCE  USE6

Year
1972
1973
1974
1975
1976
Total
employment
. do3)
130.6
132.3
130.1
132.7
137.1
Production
workers
(103)
83.2
85.1
84.0
82.7
86.6
Total
assets
($106)
12,287.8
13,048.3
13,919.5
16,198.2
18,788.3
Cost of
materials
($106)
5,338.5
6,311.8
10,388
11,569
14,503.1
Energy
purchased
for heat
and power
(109 joules)
l,220.1a
l,286.6a
l,322.7a
l,154.4a
1,202.4
 These data were estimated  by  multiplying the 1976 estimate of energy use by
the  ratio  of production  in each of  the previous  years  to  1976  production
levels.  Thus,  for  example,  energy  use  in 1972 was estimated by multiplying
energy use in  1976  by  the ratio of  production in 1972 to  production in 1976.
                                    9-9

-------
                  TABLE 9-6.   INDUSTRIAL ORGANIC CHEMICALS:
                      U.S.  IMPORTS AND EXPORTS,  1966-779
Year
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
Imports
($106)
48
48
67
84
91
129
150
169
259
205
294
326
Exports3
($106)
211
231
292
290
336
304
320
484
930
779
1,008
995
 Includes  exports  of   some  finished  products.    Figures  include  estimates
and are not strictly comparable with imports or production.
                                    9-10

-------
been similar to historical  production  and sales trends.   Certainly, over the
period 1967-1978,  consumption  increased,  although  since 1974,  if  the sales
data presented in  Table  9-4 can be regarded  as an  indicator of consumption,
consumption exhibited wide  year-to-year  variations  for the reasons discussed
in Section 9.1.3.
9.1.6  Prices
       The general  level  of prices for SOCMI chemicals more than tripled be-
tween 1967 and 1978.  Most of the increase occurred after 1973.  From 1967 to
1973, the  average  unit  price of SOCMI chemicals  remained close to $0.12/kg.
Following  the  1973-1976  adjustments  in  oil  prices,  average  prices in SOCMI
doubled, rising to $0.24/kg.   After that  time  average  unit prices increased
at  a rate of  approximately 11  percent  annually, to a  price  of $0.36/kg in
1978.  It  is  important  to  realize  that  these are average prices per unit of
all  SOCMI  chemicals.   In  1976,  prices  for  individual  chemicals ranged from
$0.11/kg for  formaldehyde to $4.30/kg for benzophenone.  Changes in the unit
price  for  individual  chemicals  may vary  substantially  from  the  changes in
average  prices.   Table   9-4 presents  annual  statistics  of production, sales
volume,  sales  value and  average unit value  for  the  industry.   The data are
weighted using the procedures described  in Appendix El   to reflect  the behav-
ior  of the industry as accurately as possible.
9.1.7  International Trade
       Chemical imports  were first made subject  to  tariffs  at the beginning
of the 20th century.  The tariffs were initiated  to protect the  infant chemi-
cal  industry  from  foreign  competition.    Since 1936,  tariffs  have been pro-
gressively lowered on chemical  products.7  The U.S.  International  Trade Com-
mission  reports  824 benzenoid intermediates  on which tariffs are  collected.
Of  these,  179 are assessed duties competitively using  import prices as the
                                    9-11

-------
basis for tariffs.   Another 430 of these products are classified noncompeti-
tive, with tariffs  based on U.S.  domestic prices.  The competitive status of
15 products is not available.8  The remainder are not tariffed.
     Accurate data  concerning  imports  and exports of SOCMI chemicals are not
available.  The  most reasonable approximation of trade  statistics  for SOCMI
are  provided  by the  U.S.  International Trade Commission.5   Annual  value of
imports and exports  for the period 1966-1977 is  presented in Table 9-6.  In
each of these years, U.S. exports exceeded U.S.  imports of industrial organic
chemicals.  Table 9-7  presents imports, exports  and  trade  balance  of indus-
trial organic  chemicals in  1976  and  1977 between the United States and its
principal trading partners.  These countries include West Germany, Italy, the
United Kingdom, Switzerland, France, Belgium, the Netherlands, Canada, Japan,
Mexico, Brazil  and  Argentina.   In 1977 the U.S.  experienced a deficit in its
balance  of  trade in  chemicals with West  Germany,  Japan,  Italy,  the United
Kingdom,  Switzerland  and France.   It experienced a surplus in its balance of
trade  in  chemicals  with Belgium, Canada,  the Netherlands,  Mexico,  Argentina
and  Brazil.   Table  9-8  presents  the  value  of  imports  for consumption from
principle  sources  from  1972  to 1977.   These imports amounted to  a total of
about $326 million in 1977.
9.1.8  Industry Growth
       A  number  of  forecasts of economic  growth  in  the  organic chemical in-
dustry  are  available.   The  annual  growth  rate  used here,  5.9  percent, was
estimated  by  McGraw  Hill10  for the basic organic chemicals  industry.   The
McGraw Hill estimate was selected for the  following reasons. First, the growth
rate was  calculated  for  a group of chemicals which closely corresponds to the
378  SOCMI  chemicals.   Second, the method  used by McGraw Hill to develop the
growth rate is internally consistent and takes account of forecasted develop-
                                    9-12

-------
                 TABLE 9-7.   INDUSTRIAL ORGANIC  CHEMICALS:
          U.S. TRADE, BY PRINCIPAL TRADING PARTNERS,  1976  AND  19779
                                 ($103)
           Source
Imports1
  Exports
Trade balance
     1976:
       West  Germany
       Japan
       Italy
       United  Kingdom
       Switzerland
       France
       Belgium
       Canada
       Netherlands
       Mexico
       Argentina
       Brazil
       All other

          Total

     1977:
       West  Germany
       Japan
       Italy
       United  Kingdom
       Switzerland
       France
       Belgium
       Canada
       Netherlands
       Mexico
       Argentina
       Brazil
       All Other

          Total
 94,768
 61,228
 30,678
 24,709
 17,280
 12,371
  2,154
  8,081
  8,987
  3,452
  1,927
     98
 28,103

293,836
105,172
 65,770
 32,711
 31,132
 21,956
 15,763
  9,839
  7,270
  4,858
  4,673
  3,353
    538
 22,865

325,900
   10,487
   27,380
     N.A.
   15,497
    2,681
   11,401
   46,779
   93,471
  178,111
   63,964
     N.A.C
   59,444
  498,985

1,008,200
    5,038
   30,736
     N.A.C
   27,458
    6,541
     N.A.C
   61,126
   82,676
  156,581
   62,965
    6,283
   78,512
  477,469

  995,385
   • 84,281
   • 33,848
   • 30,000
   •  9,212
   • 14,599
       970
    44,625
    85,390
   169,124
    60,512
   -  1,500
    59,346
   470,882

   714,364
   •100,134
   - 35,034
   • 32,500
   -  3,674
   • 15,415
   • 15,500
    51,287
    75,406
   151,723
    58,292
     2,930
    77,974
   454,604

   669,485
 Data represent customs import value,  the value appraised by the U.S.  Customs
Service in  accordance  with the  legal  requirements  of sec. 402  and  402a of
the Tariff Act of 1930, as  amended.
 Includes exports of  some  finished products.
are not strictly comparable with imports.

CN.A.  = Not available.
                 Figures  include estimates and
                                    9-13

-------
                  TABLE 9-8.   INDUSTRIAL ORGANIC CHEMICALS:
        U.S.  IMPORTS FOR CONSUMPTION,  BY PRINCIPAL SOURCES,  1972-779
                                  ($103)3
    Source           1972        1973       1974       1975       1976     1977
West Germany
Japan
Italy
United Kingdom
Switzerland
France
Belgium
Canada
Netherlands
Mexico
Argentina
All other
Total
66,085
36,181
11,305
7,605
11,593
1,611
1,220
4,301
5,067
35
3
5,031
150,037
72,715
29,793
10,705
10,433
16,063
4,233
7,919
5,515
4,724
486
--
6,892
169,478
84,059
65,027
17,323
21,119
15,846
8,585
10,494
4,826
10,291
1,812
--
19,190
258,572
62,145
49,243
19,073
18,820
14,773
9,797
1,871
4,352
6,738
388
657
17,625
205,482
94,768
61,228
30,678
24,709
17,280
12,371
2,154
8,081
8,987
3,452
1,927
28,201
293,836
105,172
65,770
32,711
31,132
21,956
15,763
9,839
7,270
4,858
4,673
3,353
23,403
325,900
aCustoms  import  value, the  value  appraised by  the U.S.  Customs  Service in
accordance with the legal  requirements of sec.  402 and 402a of the Tariff Act
of 1934, as amended.
                                    9-14

-------
ments in the  U.S.  economy.   Third,  the projections are developed for the the
period 1979-1991, entirely  covering  the forecast period of  interest in this
study (1981-1985).
     In order to estimate  the number of new model units covered by the regu-
latory alternatives,  it is  assumed  that the  number  of operating facilities
will grow  at  the  same rate as the  industry's  output.   It is further assumed
that  any   regulatory  alternative will  take effect on  January  1,  1981,  and
therefore  that  the  fifth  year of the  impact  analysis  is 1985.  In 1976 (the
most recent year for which data are available), 1,334 facilities manufactured
SOCMI chemicals in the U.S.11  If the industry grows at an annual rate of 5.9
percent, by the beginning of 1981 this number will have risen to 1,678 facil-
ities  and  by the  end of  1985  to 2,235 facilities.   Thus,  an estimated 557
units built  to  provide additional capacity for  the  industry will be covered
by  the regulatory alternatives.
     The regulatory alternatives will also cover  units constructed to replace
existing capacity which "wears out" during the period.  The  number of replace-
ment  units is  estimated  on the basis  of the  following assumptions.  First,
units have a working  life of 20 years. Second, the historical growth rate for
SOCMI prior  to  1977 was 6  percent per year.   Using these assumptions, it  is
estimated  that  274 new  units will  be required  to  replace the  part  of the
existing  capacity  that will  "wear out" over  the period  1981 to 1985.  The
methodology  used to compute this estimate  is described in detail in Appendix
E2.
     The  estimates  for  entirely new units,  combined  with  estimates  for
replacement  facilities,  indicate a total of 831  units that will be affected
by  the regulation.   To estimate  the number of A,  B and C model  units (identi-
fied  in  Section 6.1)  that  will  be constructed between 1981 and 1985,  it  is
                                    9-15

-------
assumed that  the  mix of model units will  not  change over time, and that the



percentages of A, B and C model units are as follows:



                   Model unit               Percent of existing units



                       A                               52



                       B                               33



                       C                               15                   !



     If these  percentages  are applied to the estimate of the total number of



units  presented  above, they  imply that 432 A units, 274 B  units  and  125 C



units will be affected by the regulatory alternatives.



9.2  Economic Impact Analysis



9.2.1  Market Structure and Financial Profile



     SOCMI producers manufacture  378  chemicals,  each of  which has  its  own



national  and  regional markets.   Consequently,  SOCMI firms  encounter a wide



range  of  market  situations  for  the different  chemicals  they produce.   Many



SOCMI chemicals,  for example, formaldehyde, urea and benzene, are manufactured



by a  relatively  large number cf firms using an array of different processes.



The products  have a wide range of end uses in which substitute materials can



often be  used.   Thus industry-wide elasticities of  demand  for the chemicals



are relatively high.  In this type of market situation,  producers have little



or no  ability to pass on  cost increases to consumers in the form of higher



market  prices.   Other SOCMI  chemicals,  for  example, succinonitrile, isoamy-



lene, and methyl  butynol, are manufactured by a small number of producers and



in some cases only one producer,  and  have  no  close substitutes in their end



uses.  In these oligopolistic and monopolistic markets,  producers may be able



to exercise  considerable  influence  on market  prices and to  pass  on a large



part  or all  of any  production cost increases   in the form  of higher prices.
                                    9-16

-------
     The ability of firms  to  pass on cost increases in the form of price in-

creases is influenced by  the  extent to which  the  industry is vertically and

horizontally integrated.   There  is  extensive vertical integration within the

SOCMI.   Captive consumption in  the  industry averaged  53.7  percent* of total

output during the  period  1967-1978,  and this ratio varied only slightly from

year to year.   The precise degree of horizontal  integration  within SOCMI is
                                        I
difficult to evaluate because  it varies iconsiderably among products.  However,

a general assessment of the industry-wide situation may be made using the ca-

pacity share data presented in Table 9-9i   These data suggest that no one com-

pany or  group of  companies has  a dominant position  within the industry.  In

1976, the top four companies owned only 18.3 percent and the top twenty firms
                                          1
45.4 percent  of total SOCMI  capacity.    There is  no  reason  to believe that
                                          1
the  extent  of  industry-wide  market  concentration  has altered significantly

since that time.                          I

     Data on  the  returns  on  equity, returns  on  debt,  returns on preferred
                                           i
stock,  debt-asset  ratios, equity-asset ratiios  and preferred stock-asset ra-

tios were collected  for  a sample of 100 chemical manufacturing firms for the

most recent available years.t  These data are  presented  in Table  E3-2.  The

data  have  been  used to estimate  the cost of capital  to  firms  in the SOCMI,

using  the assumption  that the sample of  firms in Table E3-2 is unbiased and

normally distributed.  A detailed discussion of the methodology used to esti-

mate the cost of capital  is presented in Appendix  E3.

     The estimated cost  of capital,  presented  in Table 9-10, is used in Sec-

tion 9.2.3 to estimate the economic  impacts of  SOCMI  fugitive emissions regu-
This figure is estimated from data presented in Table 9-4.

tData on  the  ratio variables and rates of return were available for 1977 and
1978, respectively.
                                    9-17

-------
                  TABLE 9-9.  INDUSTRY CONCENTRATION,  19764
 Number        Percent             Estimated                 Percent of
of firms      of firms           capacity (gg)            industry capacity


 Top 4          0.72                58.75                       18.3

 Top 8          1.43                91.82                       28.6

 Top 20         3.58               145.75                       45.4

 Top 40         7.17               186.68                       58.1
                                    9-18

-------
          TABLE 9-10.   ESTIMATED  COST OF  CAPITAL FOR FIRMS IN SOCMI3

Mean
Standard
deviation
Minimum
Maximum
Aftertax cost
of capital
10.807%
0.930
8.015%
12.798%
Pretax cost
of capital
20.783%
1.789
15.414%
24.612%
aSee Appendix E3 for details  of the data and methodology used to estimate
 the cost of capital for firms in SOCMI.
                                    9-19

-------
latory  alternatives.   Note  that the  average aftertax  cost of capital  for

chemical firms  is  10.8 percent.   On a pretax basis, this figure increases to

20.8 percent.   If, as  was assumed, capital  costs  are  normally distributed,

then 95 percent of the firms in the  industry face aftertax costs  of capital

in  the  range of 9.0 percent  to  12.7  percent and pretax costs  of  capital in

the range of 17.2 percent to 24.4 percent.

9.2.2  Regulatory Alternatives

     The four  regulatory alternatives being  considered  are  described  in de-

tail  in Section  6.2.    The  baseline  regulatory  alternative (alternative I)

does not require  producers to implement additional control techniques.  Con-

sequently,   model  units  complying with  this  alternative would  not incur any

incremental  costs*  and no economic  impacts would result from its implementa-

tion.   Regulatory  alternatives  II,  III,  and  IV  require  successively more

stringent  equipment inspections  and equipment specifications.  Firms comply-

ing with regulatory alternatives II, III, and IV would therefore incur incre-

mental costs, and consequently economic impacts would result from their imple-

mentation.

9.2.3  Economic Methodolgy

9.2.3.1  Regulatory Scenarios

         Economic  impacts are estimated for  regulatory  alternatives II, III

and  IV,  but not for regulatory alternative I, since firms will not  incur in-

cremental  costs  in complying with that  alternative.    The  economic  impacts

associated  with alternatives II, III  and IV  are estimated under two alterna-
*Incremental  costs  of a regulatory alternative are  those  additional  costs a
firm  incurs  in meeting the  regulatory alternative that it would not incur in
meeting the baseline  alternative.
                                    9-20

-------
tive assumptions about firm  pricing  behavior:   (1) full  cost  absorption and
                i
(2) full cost pricing.   Combining  the three regulatory alternatives with the

two alternative  pricing models yields six regulatory scenarios:

                     Regulatory Alternative                Pricing Policy



Scenario 1                Alternative II                Full Cost Absorption



Scenario 2                Alternative II                Full Cost Pricing



Scenario 3                Alternative III               Full Cost Absorption



Scenario 4                Alternative III               Full Cost Pricing



Scenario 5                Alternative IV                Full Cost Absorption



Scenario 6                Alternative IV                Full Cost Pricing



Under  full  cost  absorption,  the  affected  firm  bears the  full incremental

costs of environmental controls, accepting a lower rate of return on  its cap-

ital investment.  Under full cost pricing, the firm adjusts product prices so

as  to  maintain its  current aftertax  rate of return  on capital investment.

     The  alternative assumptions about  firm  pricing  behavior are associated

with different  market conditions  in the  affected  industry.   In both cases,

firms are assumed to have no monopsony power in resource  markets.  Thus, they

cannot  pass back  cost increases to  resource  suppliers.   In the cost absorp-

tion case,  the  domestic  industry as  a  whole  is  assumed  to be a price taker,

unable  to affect the market price of  its product either because  of the exist-

                                    9-21

-------
ence of close  product substitutes,  or because of strong international compe-



tition in domestic and foreign markets.   However, full cost pricing will  take



place if  the  industry produces a commodity for which no domestic or imported



substitutes exist, or if the industry has constant costs. A constant-cost in-



dustry  is one  in  which unit  costs  remain constant  as industry  output  in-



creases.   Firms in  such  industries  experience  constant  returns  to scale.



     In fact, firms in SOCMI face a wide variety of product market situations



(see Section 9.2.1).   Some firms will be able to fully pass through cost in-



creases to consumers in the form of higher prices.  Some will be able to pass



on only a part of the cost increases.  Others will be forced to fully absorb



all regulatory  control  costs,  leaving product prices unchanged.  Consequent-



ly, the full  cost pass through  and  full  cost absorption scenarios evaluated



below provide  estimates of the  maximum range of possible price  and rate of



return impacts  for the different products and firms in SOCMI.



9.2.3.2   Estimation of Regulatory Price Impacts Under Full  Cost Pricing



     Under  full cost  pricing,  the  firm  is  assumed  to  respond to  cost  in-



creases by adjusting product price to maintain a target rate of return on in-



vestment.   The  required price change (dP) may be calculated using the follow-



ing equation:*



      HD  _ dTOC + r dK/(l-t)                          (1)
      dK  -        Q






where

      dP  = required change in product price



    dTOC  = total annual operating costs of compliance



      dK  = total initial costs of compliance



       Q  = total annual unit output
*The  derivations of  Equations  (1)  and  (2)  are  presented  in  Appendix E4.


                                    9-22

-------
       r = target rate of return
       t = tax rate
9.2.3.3  Estimation of Rate of Return Impacts Under Full Cost Absorption
     Under full cost absorption, an increase in facility production costs re-
sults  in  a lower rate of  return  on investment for the  firm,  because market
conditions prohibit  it from  passing on cost increases  to  the consumer.  The
impact on the facility's  rate of return on investment is given by the follow-
ing equation:

    -dr  =  r * dK  + P-V dTOC                       (2)
                      K
where
     dr  =  change in rate of return, and
      K  =  preregulation level of capital investment.
Note that  pretax  rate of return impacts may be calculated by setting the tax
rate variable,  t, equal   to  zero  in  Equations (1) and  (2).   Also note that
price and  rate  of return impacts are  estimated  on the assumption that capa-
city utilization  rates  remain  constant (that is, Q  remains unchanged).   To
the  extent that the regulatory alternatives result  in decreases (increases)
in capacity utilization  rates, price and rate of return impacts will be larg-
er (smaller) than those  estimated using Equations (1) and (2) because of eco-
nomies of  scale in the use of control techniques.
9.2.3.4  Other Economic  Impacts
     The price  and  rate  of return  impacts  estimated  by the above techniques
are used to make a quantitative assessment of the probable impacts of the reg
ulatory alternatives  II, III,  and IV  on  industry  growth,  new facility open-
ings,  the  replacement of existing  facilities, and investment levels.  These
data are  then  used  to assess the  extent of interindustry and macroeconomic
impacts associated with  the various regulatory alternatives.
                                    9-23

-------
9.2.3.5  Estimation Data
     Estimation of price and rate of return impacts for different model units
requires data  on  the following variables:  (1) total acquisition and instal-
lation costs of  the control equipment (dK), (2) total annual operating costs
of the control  equipment and monitoring procedures (dTOC), (3) the preregula-
tion capital stock  (K),  (4) the target  rate  of return (r), (5) the tax rate
(t), and  (6) model  unit production levels (Q).  Data on dK and dTOC for each
of  the three model  units identified as  representative  of the industry* were
obtained from  Section 8.1.   The tax rate  is  assumed to be 48 percent.  Data
on  model  unit  production levels were obtained from the 1976 Organic Chemical
Producers  Data Base for each of the units covered by the regulatory alterna-
tives.  The 831  model units are assumed  to  be distributed by capacity in an
identical  manner  to the  1,105 units for which both value of product and quan-
tity  data  are available in the 1976  Organic  Chemical  Producers Data Base.
Thus,  the  number of new units  with  a  given capacity, say 100 Gg, is assumed
to  be  equal to  the  number  of  units in  the  Organic Chemical  Producers Data
Base  with  the same capacity,   multiplied  by  the ratio of  the  number of new
units  to the number of units in the data base (831/1,105).  Actual unit output
levels are obtained by applying a capacity utilization rate to the estimated
unit capacities.
     To evaluate  industry-wide  impacts,  the cost data  from Section 8.1 were
adjusted  to  allow for higher or lower product  recovery credits for chemicals
with  a value  greater or less than $0.36/kg.  For such chemicals, product re-
covery credits were  estimated by multiplying estimated product savings by the
price  of the chemical  in question.
 *See  Chaper  6  for  a  detailed  discussion of the model units.
                                     9-24

-------
     Data on the value  of  the preregulation capital  stock for plants of dif-
ferent capacities were calculated as follows.   A capital-capacity coefficient
for firms in SOCMI  was  obtained by dividing the estimated total value of in-
dustry assets in 1976   by  the volume of  output produced in that year.*  The
estimate  of  the capital-output coefficient  was  converted into  a capital-
capacity  coefficient by  multiplying the capital-output  coefficient  by  an
assumed industry-wide capacity utilization rate.  The assumed capacity utili-
zation rate for  1976 was  50 percent.  This capacity utilization estimate was
based on  the assumption that the typical  capacity utilization for the indus-
try is 75 percent.   In  1976, output was  9.2  percent below the industry-wide
high  level of output achieved in 1974.   Between 1974 and 1976 it is probable
that some additions to industry capacity were made.  Hence, the assumption of
a 50  percent capacity utilization rate for 1976,  though somewhat arbitrary,
is  not  unreasonable.  The  1976 capital-capacity  coefficient  was updated to
last  quarter 1978  dollars  using the machinery and equipment price  index com-
puted by  the  United  States Department of  Commerce.12   The capital-capacity
coefficient estimated  by  the  above procedure was $125/Mg  of product.   This
coefficient was  multiplied  by model unit capacity to obtain an estimate of K
for each model  unit considered in the analysis.
     Estimates of  pretax  and  post-tax  rates of return  used  in the analysis
are presented in Table  9-10.  These data were obtained from an analysis of a
sample  of 100  firms in  the  SOCMI industry.   Details  of the  analysis are
contained in Appendix E3.
 ''See  Tables  9-4 and  9-5  for data  on production  and  total  industry assets.
                                    9-25

-------
9.2.4  Economic Impacts
9.2.4.1  Rate of Return Impacts
     Data on unit  capacity,  product value, capital investment  and  tax rates
are  available  for 1,105  units in  the  1976 Organic  Chemical  Producers  Data
Base.4  Price data were updated using the Chemical Marketing Reporter.13   Cap-
ital  stock  estimates were  also  expressed  in  1978 prices.  These  data  were
used in conjunction with the cost information presented in Section 8.1 to cal-
culate full cost absorption rate of return impacts of regulatory alternatives
II,  III, and IV for the 831 model units projected to be built.   It is assumed
in estimating the rate of return impacts presented here that the 831 new model
units will have the same capacity and product value distributions as the units
in  the  Organic Chemical  Producers  Data Base,  and  that capacity utilization
for  each  unit  is  50 percent.  This  relatively  low capacity utilization rate
is  used  to  estimate  unit  output  levels   because  it  represents  a feasible
worst-case  economic  scenerio for the industry  (that  is,  economic conditions
similar to those experienced in 1976).  As a-result, actual impacts are like-
ly to be  less adverse than those presented below.
     Rate of return  impacts are estimated  on  the basis of these assumptions
for  each  of the 831 new  model  units  covered by  the regulatory alternatives.
It  is  probable  that the  assumption of  a  constant product price distribution
also  results in an overestimate of adverse rate  of return and price impacts,
since  the prices  of  the  products manufactured by  SOCMI  are  expected to in-
crease  between  1979 and  1985  as energy and feedstock  costs  rise.   Any real
increase  in  product  prices will raise the  value  of product recovery credits,
lower the net costs of compliance associated with any given regulatory alter-
native, and thereby reduce adverse  rate of  return and price impacts.
     Rate of return impacts for A,  B  and C  average model  units  under each reg-
                                    9-26

-------
ulatory alternative are presented in Table 9-11.   Each of these average model
units is assumed to manufacture products valued at approximately $0.36/kg, to
have an annual  capacity  of  84,678 Mg and to have an existing cost of capital
of 10.81 percent.   These average model  units differ only in terms of the com-
plexity of the  processes  they  use to manufacture the chemicals.  The product
value and  rate  of return data  represent the means for  each  variable in the
samples used in  the analysis.   Under regulatory alternatives II and III, each
average model unit  experiences  a very small increase, not a decrease, in its
aftertax rate of return  on  investment,  regardless  of  the  process it uses.
This result is obtained because at a price of $0.36/kg for recovered product,
product recovery credits exceed total annualized costs of control.  Under reg-
ulatory alternative  IV,  average model  unit of types A and B experience small
decreases in aftertax rates of return on investment.  Model C units experience
rate of return  decreases  amounting to  1.12  percentage points,  still  quite
small adverse impacts.
     The data presented in Table 9-11 suggest that some  firms may benefit from
the  implementation  of  any regulatory alternative.  These results are subject
to  the  following  qualifications.   In the above  analysis,  it is assumed that
firms will  not independently implement the emissions controls proposed in the
regulatory alternatives.   In  fact,  if there are  significant net cost reduc-
tions to be achieved from additional emissions controls, firms  will voluntar-
ily  adopt  them.   Under such  circumstances,  the  cost  reductions  associated
with any regulatory alternative will be considerably smaller.  Note  that in-
centives for  voluntary  emissions controls increase as the value of the manu-
factured product increases. As some SOCMI producers manufacture highly valued
products with prices  in excess of $0.50/kg, they  are likely to be willing to
use  extensive emissions control techniques in the  absence of any NSPS.
                                     9-27

-------
                 TABLE 9-11.   AVERAGE RATE OF RETURN IMPACTS3
Change in rate of return (percentage points)
Model
Units
Unit A
Unit B
Unit C
Alternative II
+0.000
+0.003
+0.006
Alternative III
+0.000
+0.001
+0.005
Alternative IV
-0.16
-0.37
-1.12
almpacts are estimated on the assumption that the initial aftertax rate of reti;vn
on investment is 10.807 percent, the mean cost of capital presented in Table 9-10;
the initial price of the product is $0.36/kg; plant capacity is 84,678 Mg; and
the capacity utilization rate is 50 percent.
                                    9-28

-------
     Although in general firms will  not be  affected by the implementation of
regulatory alternatives  II,  III,  and  IV, because of wide variation in produc-
tive capacity  and  value of  product  among model units,  some  will  experience
adverse rate of return impacts.  Estimates of the numbers of model units exper-
iencing rate of  return  decreases  in  excess of  one  and two percentage points
as a result of the  implementation of  each regulatory alternative are present-
ed in Table 9-12.  These estimates were obtained by calculating rate of return
impacts for each of the  831 new model units  under the assumption that 52 per-
cent,  33  percent,  and  15  percent of all units  of  all sizes are A,  B and C
model  units,  respectively.  Under regulatory  alternatives  II and  III,  the
estimated  number of adversely affected units  is  very small;  only  6 and 12
units, out of a total of 831 model units, experience rate of return decreases
of more than  one percentage point.  Under  regulatory  alternative  IV, a much
larger number of units,  93 in all, are estimated to experience rate of return
decreases  in  excess  of  1 percent  under  a full  cost absorption scenario.  It
should be noted that most of these adversely affected units are B and C model
units rather than A model units.   In  fact, all B and C model units with capa-
cities in  excess of 26,464 and 89,121  Mg,  respectively,  producing chemicals
with prices exceeding $0.15/kg, will  experience rate of return impacts small-
er  than  one percentage  point  even under alternative  IV.   The EPA estimates
that virtually  all  B and C model  units do  in fact have capacities in excess
of this figure,14  and furthermore, industry sources  indicate  that most pro-
duce  chemicals  that  have  prices  in  excess of  $0.15/kg.13   If the estimated
impacts on  B  and C model units are  ignored,  only 25  units  are likely to be
adversely affected by regulatory alternative IV.
                                    9-29

-------
  TABLE 9-12.   MODEL UNITS EXPERIENCING SIGNIFICANT RATE OF RETURN IMPACTS
                         UNDER FULL COST ABSORPTION3
Model
units
Unit A
Unit B
Unit C
Total
Alternative
dr < -1% dr
6
0
0
6
II
< -2%
4
0
0
4
Alternative
dr < -1% dr
7
4
1
12
III
< -2%
5
2
0
7
Alternative
dr < -1% dr
25
34
34
93
IV
< -2%
12
16
20
48




adr denotes the percentage point change in firms' rates of return on investment.
                                    9-30

-------
9.2.4.2  Price Impacts
     The potential price impacts  of  regulatory alternatives II,  III,  and IV
are also  estimated under the  assumption  that capacity and  value of product
distributions will remain constant over  the forecast period, 1979-1985.   The
price  impact  estimates  are  therefore subject to the  same  limitations as the
rate of return impact estimates discussed above.   Potential price impacts for
A, B, and C model units with average  capacities of 84,678 Mg and product val-
ues of  $0.36/kg  are  presented  in Table 9-13.  Under  regulatory alternatives
II and III,  price impacts  are  negative  for each type of  model  unit because
annual product recovery credits exceed the total  annualized cost of the moni-
toring  procedures  and capital   equipment  required  under  these alternatives.
Under  regulatory  alternative   IV,  extremely  small  positive  price  impacts
occur.  In general most units will not increase product prices as a result of
the implementation of regulatory alternatives II,  III, or IV.   However, be-
cause  of  the variations  in capacity and product value within the industry,
some  firms  may  have  to raise   product  prices in order  to maintain existing
rates  of  return  on investment.   In  some  cases,  the price increases required
by individual facilities are in excess of 5 percent and even 6 percent of the
current product  price.   Data on the estimated numbers of  such units are pre-
sented  in Table  9-14.  These estimates are also obtained by calculating price
impacts for  each of  the 831 new model units  under the assumption that units
are distributed  among A,  B, and C model  units in the manner described above
and operate  at  50 percent of unit capacity.   Under alternatives II and  III,
only  five  A  and eight B model  units would have to  increase product prices by
more  than  5  percent.   Under alternative  IV, 30 units must increase prices by
more  than  5  percent  to maintain preregulation rates of return on investment.
However,  it  should be noted that these estimates may overstate the extent of
                                     9-31

-------
  TABLE 9-13.   AVERAGE PERCENTAGE PRICE IMPACTS OF  REGULATORY  ALTERNATIVES3
Model
units
Unit A
Unit B
Unit C

Alternative II
-0.000
-0.002
-0.009
Price changes (percent)
Alternative III
-0.000
-0.002
-0.007

Alternative IV
+0.000
+0.000
+0.733
 Impacts are estimated on the assumption that the target rate of return is
10.807 percent, the average cost of capital  presented in Table 9-10;  the
initial price of the product is $0.36/kg; plant capacity is 84,678 Mg;  and
the capacity utilization rate is 50 percent.
                                    9-32

-------
TABLE 9-14.   MODEL UNITS REQUIRING SIGNIFICANT PRICE INCREASES
              TO MAINTAIN TARGET RATES OF RETURN
Model
Units
Unit A
Unit B
Unit C
Total
Alternative II
Price increase
£ 5% ^6%
4
0
0
4
4
0
0
4
Alternative III
Price increase
^5% £ 6%
4
2
0
6
4
2
0
6
Alternative IV
Price increase
^5% ^6%
6
10
14
30
5
9
12
26
                             9-33

-------
significant price impacts under regulatory alternative IV.   Most of the units
estimated  to  require  price  increases  in  excess  of 5  percent are  C model
units.   In  fact,  C  model units that manufacture chemicals with prices in ex-
cess of  $0.15/kg and  have  capacities  greater than 71,550 Mg do  not  have to
increase product prices by more than 5 percent to maintain their target rates
of return  on  investment.   The EPA estimates  that  virtually  all plants using
processes with the same degree of complexity as that assumed for C model units
have larger capacities and produce products with higher values than these.13
In addition, model units are assumed to operate at the relatively low capacity
utilization rate of 50 percent.
     A  final  caveat concerning  price and  rate  of return  impacts  should be
noted.   The impact estimates presented in Tables 9-12 and 9-14 were developed
on the  assumption that feedstock prices are unaffected by the  implementation
of any regulatory alternative.  However, as the industry extensively uses its
own products  as  feedstocks, this assumption is not strictly valid and intro-
duces a  systematic  upward bias in the  estimated  size of adverse rate of re-
turn and potential price impacts.  The upward bias occurs because, in general,
firms adopting  alternative II, III, and IV control technologies will achieve
net cost reductions and, at  least in competitive markets, will  tend to reduce
rather  than increase  the prices of products used as feedstocks by the indus-
try.
9.2.4.3  Investment Impacts
     It  is  difficult to  assess the impact of any of the standards on the num-
ber of  units  to  be constructed between January 1, 1981, and December 31, 1985,
because of  the variations  in  these impacts across units.  Some  smaller facili-
ties may not  be  erected  as  a  result of the standard because of  adverse impacts
on  rates  of return and  price  competiveness.   Other  larger facilities may be
                                    9-34

-------
built because production costs  fall  as a result of  emissions reductions and

product recovery credits.   Therefore, in this analysis it is assumed that im-

plementation of regulatory alternatives  II,  III, IV will  have no measurable

impact on the number of new facilities constructed between 1981 and 1985, the

5-year period  following proposal of any regulatory  alternative.  Industry-

wide  investment  impacts are  therefore simply the  incremental capital  costs

associated with the  acquisition of  the capital  and  monitoring equipment re-

quired under each  regulatory  alternative by the 831 new units expected to be

constructed between 1981 and 1985.

     Data on these investment impacts are presented in Table 9-15.  The esti-

mates  are  obtained by assuming  that 432 A  model plants,  274 B model units,

and 125 C model units will  be constructed and that, as a result of each regu-

latory alternative, these units incur incremental capital costs equal to those

presented  in  Section  8.2.    Under  regulatory  alternatives  II  and  III,

industry-wide  investment  impacts  are  quite  small,  less than  $22 million.

Under  regulatory  alternative  IV, they increase  substantially to almost $233

million.   Nevertheless, even under regulatory alternative IV,  the total 5-year

investment impacts  of  any  of the regulatory  alternatives  would be less than

1.14  percent of  total  industry assets in 1976  and less than  1.64 percent of

the value of industry sales in 1978.*  It appears, therefore,  that the indus-

try as a whole will not have much difficulty in obtaining the  investment funds

to  acquire  required control  equipment under any  of the regulatory alterna-

tives.
*Data on  total  industry value of assets  and  industry sales are presented in
Table 9-4 and 9-5.
                                    9-35

-------
                       TABLE 9-15.  INVESTMENT IMPACTS
Model
units
              Number of
             model  units
  Incremental
   model unit
costs of control
    ($106)
Incremental  indus-.:
  costs of control
       ($106)
Regulatory
alternative
    Unit A
    Unit B
    Unit C
II
          Total
                 432
                 274
                 125

                 831
          Total
                 831
    0.0132
    0.0272
    0.0660
          Total
                 831
       5.7
       7.5
       8.3

      21.5
Regulatory
alternative III
Unit A
Unit B
Unit C

432
274
125

0.0132
0.0272
0.0660

5.7
7.5
8.3
                         21.5
Regulatory
alternative IV
Unit A
Unit B
Unit C

432
274
125

0.0871
0.2950
0.9110

37.6
80.8
113.9
                        232.3
                                     9-36

-------
9.2.4.4  Employment Impacts



     Regulatory alternatives II,  III,  and IV will  each have small but measur-



able impacts on  employment  in  SOCMI because they  require  firms to intensify



monitoring and maintenance schedules to control fugitive emissions.  Estimates



of the  number  of additional  workers required  as  a result of each regulatory



alternative are presented in Table 9-16.   The estimates were obtained by mul-



tiplying the projected numbers of each type of affected facility by the unit-



by-unit, person-year monitoring  and maintance requirements for each standard



presented  in  section  8.1.*   The  largest employment impacts (400 workers)



are associated with regulatory alternative III, which requires more stringent



monitoring programs than alternative II.   Under alternative IV, some alterna-



tive III monitoring requirements are replaced by equipment controls, reducing



incremental employment requirements to approximately 225 workers.  The employ-



ment impacts of  each  of the standards are small relative to total employment



in the  industry, representing no more than 0.6 percent of the 1976 SOCMI work



force in each case.



9.2.4.5  Total Annualized Costs of Control



     Total incremental annualized costs of control for the fifth year follow-



ing promulgation of alternatives II,  III, or  IV are presented  in Table 9-17.



Product  recovery  credits  are  calculated  using  the  fourth   quarter  1978



industry-wide  average  product price of  $0.36/kg.   Under regulatory alterna-



tives  II  and III, the industry  as  a  whole is estimated to reduce annual ized



production  costs by  $28.73 million and  $21.35 million,  respectively.   Under



regulatory  alternative  IV,  annualized   production  costs  are  estimated  to
*A person-year is assumed to consist of 2,000 person-hours.
                                    9-37

-------
                      TABLE 9-16.  EMPLOYMENT IMPACTS
                               (Person-years)
Model
unit
                   Alternative II
                       Alternative III
 Unit   Industry
 Unit   Industry
                         Alternative IV
 Unit   Industry
Unit A

Unit B

Unit C
0.1237    53.43

0.1863    51.05

0.5017    62.71
0.1855   80.14

0.5079  139.16

1.4532  181.65
0.0579    25.01

0.2277    62.39

1.0982   137.27
All units
         167.19
        400.95
         224.67
                                    9-38

-------
       TABLE 9-17.   MODEL UNIT AND INDUSTRY ANNUALIZED CONTROL COSTS
Regulatory
alternative
       No.  of
       model
       units
  Incremental
     unit
annualized costs
 without product
 recovery credit
     ($103)
  Incremental
     unit
annualized costs
  with product  .
 recovery credit0
     ($103)
 Incremental
  industry
annualized costs
 with product
recovery credit
    ($103)
Alternative II
    Unit A
    Unit B
    Unit C

          Total

Alternative III
    Unit A
    Unit B
    Unit C
          Total
Alternative
    Unit A
    Unit B
    Unit C
IV
        432
        274
        125
        432
        274
        125
        432
        274
        125
      12.1
      20.3
      48.3
      15.0
      35.5
      95.1
      29.0
      93.5
     283.0
     •  3.3
     • 38.5
     -134.0
     -  1.7
     • 28.3
     -103.0
      7.9
     13.3
     33.0
          Total
- 1,430
-10,550
-16,750

-28,730
    730
- 7,750
-12,870

-21,350
  3,410
  3,640
  4,120

 11,170
 Product recovery credits estimated on the basis of an assumed product value of
$0.36/kg.
                                    9-39

-------
increase by $11.17 million.  If the above estimates are accurate in the mini-



mal  sense  that they  indicate  the direction  in which  production  costs will



move and their  approximate order of magnitude, then it may be concluded that



none of the  regulatory alternatives  will result in  any measurable industry-



wide increase in prices.



9.2.4.6  Interindustry Impacts



     Interindustry  impacts will  be negligible, because net annualized costs



of control  are extremely small relative to the value of total industry output,



representing  less  than 0.03 percent of the value  of 1978 output in even the



most adverse case  (regulatory alternative IV).



9.3  Socio-Economic and Inflationary Impacts



     The socio-economic  and inflationary impacts  of alternatives II, III and



IV will be very small.



(1)  Annualized Costs:  In  the fifth year following promulgation, the regula-



     tory  alternatives,  if implemented,  are  estimated to  result  in either



     annualized  cost  reductions or  very small  annualized  cost  increases.



     Consequently, none of  the alternatives violates the regulatory criterion



     of $100 million.



(2)  Price Impacts:   Because industry-wide annualized costs  of compliance for



     alternatives  II, III  and  IV are estimated to  be  negative or extremely



     small relative to the  value of industry output, none of the standards is



     likely to cause  any  industry-wide price increases.
                                     9-40

-------
                                 REFERENCES
 1.   Condensed  Chemical  Dictionary.    Rev.  by  Gessner  Hawley,  8th  ed.  New
     York:   Van Nostrand Reinhold Co,  1971.

 2.   Comer,   James   F.    Synthetic Organic Chemicals Manufacturing Industry:
     Inputs  and Product  Uses.   Prepared for the U.S.  Environmental Protection
     Agency, 1979.

 3.   Proceedings  of  the  Conference   on  Chemical  Feedstock  Alternatives.
     American Institute  of Chemical Engineers,  Houston, TX, 1977.

 4.   Radian  Corp.   "Organic  Chemical   Producers  Data Base,  1976."  Prepared
     for the  U.S.   Environmental   Protection  Agency  under EPA  Contract No.
     68-03-2623.  1978.

 5.   United  States  International  Trade Commission.  Synthetic Organic Chemi-
     cals, U.S.  Production and Trade.   Washington, D.C.,  1967-1978.

 6.   United   States  Department  of Commerce.   Annual Survey of Manufactures,
     Industry Profiles,  1976.   Washington, D.C.

 7.   Russell,  T.W. F. ,  M.W.  Swartzlander,  and  J.  Wei.   The Structure of the
     Chemical Processing Industries.    New  York:   McGraw  Hill,  1979.  pp.
     321-334.

 8.   United   States   International  Trade  Commission.   Imports of Benzenoid
     Chemicals  and Products,  1978.  Washington, D.C., July 1979.

 9.   United   States  International  Trade  Commission.   "Import  Penetration of
     U.S.  Markets   for  Cyclic  Intermediates."   Synthetic Organic Chemicals:
     U.S.  Production and Sales -1977.   Washington, D.C., 1978.

10.   The American Economy, Prospects for Growth to 1991.   New  York:   McGraw
     Hill,  1979.

11.   Radian  Corp.   "Organic  Chemical  Producers  Data  Base,  1978."  Prepared
     for the  U.S.  Environmental  Protection  Agency  under EPA  Contract No.
     68-03-2623.  1979.

12.   United   States  Department  of   Commerce.    Survey of Current Business.
     Washington, D.C.,  1976-1979.

13.   Chemical Marketing Reporter.  September 7, 1979.

14.   Memo from Hustvedt,  K.C.   December  20,  1979.   Model  Unit Capacities.
                                    9-41

-------
APPENDIX C.  EMISSION SOURCE TEST DATA

-------
                               APPENDIX C
                        EMISSION  SOURCE TEST DATA
     The purpose of Appendix C is to describe testing results used in the
development of the Background Information Document (BID) for fugitive
emissions from the Synthetic Organic Chemicals Manufacturing Industry
(SOCMI).  The information in this appendix consists of a description of
the tested facilities,  and the sampling procedures and test results of
fugitive emissions studies in SOCMI  and the petroleum refining industry.
     Considerable data  exist concerning both the incidence and magnitude
of fugitive emissions from petroleum refineries.  The purpose of the SOCMI
study was, in part, to  support the use of emission factors generated
during studies of emissions from  petroleum refineries for similar sources
in the Synthetic Organic Chemicals Manufacturing Industry.  The results
of the SOCMI study and  data from  a study of fugitive emissions from
petroleum refineries are discussed in Section C.I.
     Section C.2 consists of the  results of three studies on the effects
of maintenance on reducing fugitive  VOC emissions from valves in petroleum
refineries and one study on maintenance of valves in a SOCMI process unit.
These results are included as an  indication of the reduction in emissions
which could be expected as a function of the designated action level, and
by applying routine on-line maintenance procedures.
C.I  FUGITIVE EMISSIONS TEST PROGRAMS
     The SOCMI test program conducted by EPA personnel and EPA contractors
consisted of emissions  testing in six chemical process units.  Data were
collected pertaining to the percentage of fugitive emission sources found
to be leaking, as indicated by the VOC concentration measured at the source.
     The results of a study on fugitive emissions from petroleum refineries
are also discussed in this section.   Data on fugitive emissions were obtained
from thirteen refineries located  in  major refining areas throughout the

                                  C-l

-------
country.  Data on the effects of maintenance were obtained at the
last four of these refineries.  These results are discussed later in
Section C.2 of this Appendix.
       The test procedures and the results obtained for each of these
  studies are described in detail  in the following sections.
  C.I.I  Description and Results of SOCMI Study
       The objective of this test program was to gather data on the percen-
  tage of sources which leak (as defined by a VOC concentration at the leak
  interface of >10,000 ppmv).  To achieve this objective,  an attempt was
  made to screen all potential leak sources (generally excluding flanges)
  on an individual component basis with a portable organic vapor analyzer.
  The test crews relied on plant personnel  to identify equipment handling
  organics.  Normally all pumps and compressor seals were  examined, and the
  percentage of valves carrying VOC which were screened ranged from 33 to
  85 percent.  All tests were performed with a Century Systems Corporation
  Organic Vapor Analyzer, Model 108, with the probe placed as close to the
  source as possible.  The results of this study are shown in Table C-l.
        Six chemical process units were screened.  Unit A is a chlorinated
  methanes production facility in the Gulf Coast area which uses methanol
  as feedstock material.  The individual component testing was conducted
  during September 1978.  Unit  B is a relatively small ethylene production
  facility on the West Coast which uses an ethane/propane  feedstock.
  Testing was conducted during October 1978.   Unit  C is a chlorinated
  methanes production facility in the Midwest.  This plant also uses
  methanol as the basic organic feedstock.   Over the last  few years, several
  pieces of equipment have been replaced with equipment the company feels  is
  more reliable.  In particular, the company has installed certain types of
  valves which they have found do not leak "as much" as other valves..
                                    C-2

-------
The individual  component testing  was  conducted during January 1979.   Unit
D is an ethylene production facility  on the Gulf Coast,  using an ethane/
propane feed.   The facility is  associated with a major refinery, and
testing was conducted during March  1979.   Units E and F are part of an
intermediate size integrated petroleum refinery located in the North
Central United  States.   Testing was conducted during November 1978.   Unit E
is an aromatics extraction unit that  produces benzene, toluene, and xylene
by extraction  from refined petroleum  feedstocks.  Unit E is a new unit and
special attention was paid during the design and startup to minimize
equipment leaks.  All valves were repacked before startup (adding 2 to 3
times the original packing) and all pumps in benzene service had double
mechanical seals with a barrier fluid.  Unit F produces benzene by
hydrodealkylation of toluene.  Unit F was originally designed to produce a
different chemical and was redesigned to produce benzene.
     In general, chloromethane  plants had fewer leaks than the ethylene
production facilities.
C.I.2  Description and Results  of Refinery Fugitive Emissions Study
     Data concerning the leak frequencies and emission factors for various
fugitive sources were obtained  primarily at nine refineries.  More
complete information for compressors  and relief valves emissions was
obtained by sampling at four additional refineries.  Refineries were
selected to provide a range of  sizes  and ages and all of the major
petroleum refinery processing units were studied.  The type of process
units and the number of each studied  in the first nine refineries are
listed in Table C-2.
                                C-3

-------
TABLE C-l   FREQUENCY OF LEAKS FROM FUGITIVE EMISSION SOURCES IN
            SYNTHETIC ORGANIC CHEMICAL UNITS
. _ Unit Ac. .... _ . . Unit Bc 	 	 	 Unit_C° 	 	 Unit De 	 Unit Ef
Chloromethanes Ethyl ene Chloromethanes Ethyl ene BTX Recovery
Number Percent with Number Percent with Number Percent with Number Percent with Number Percent with
of screening of screening of screening of screening of screening
sources values sources values sources values sources values sources values
Equipment type tested >10,000 ppmv tested >10,000 ppmv tested >10,000 ppmv tested ^O.OOO ppmv tested >10,000 ppmv
Valves 600 1 2301 19 658 0.1 862 14 715 1.1
Open-ended lines 52. 2 386 11 -a 90 13 33 0.0
Pump seals : 47 15 51 21 39 3 63 33 33b 3.0
o
i
"^ Compressor seals -a 42 59 3 33 17 6 -3
Control valves 52 6 128 20 25 0 25 44 53 4.0
Pressure relief valves 7 0 -a -3 -a -a
Flanges 30 3 -a -a -a -a
Drains -a -a -a 39 10 -a
Unit Ff
Toluene HDA
Number Percent witn
of screening
sources values
tested >_! 0,000 ppmv
427 7.0
28 11.0
30 10.0
_a
44 11.0
_a
_a
_a

a No data
Pump seals in benzene service have double mechanical seals
c Source: Reference- 1
Source: Reference 2
e Source: Reference 3

-------
         TABLE C-2.   SAMPLED PROCESS UNITS FROM NINE REFINERIES
                          DURING REFINERY STUDY
                                               Number of
  Refinery process unit	sampled units


Atmospheric distillation                             7

Vacuum distillation                                  4

Thermal operations (coking)                          2

Catalytic cracking                                   5

Catalytic reforming                                  6

Catalytic hydrocracking                              2

Catalytic hydrorefining                              2

Catalytic hydrotreating                              7

Alkylation                                           6

Aromatics/isomerization                              3

Lube oil manufacture                                 2

Asphalt manufacture                                  1

Fuel gas/light-ends processing              -        11

LPG                                                  2

Sulfur recovery                                      1

Other                                                3



Source:  Ref. 5
                                  C-5

-------
     In each refinery, sources in six to nine process units were selected
for study.  The approximate number of sources selected for study and
testing in each refinery is listed below:
               Valves                      250-300
               Flanges                     100-750
               Pump seals                  100-125
               Compressor seals             10-20
               Drains                       20-40
               Relief Valves                20-40
There were normally 500-600 sources selected in each refinery.
     The distribution of sources among the process units was determined
before the selection and testing of individual  sources was begun.
Individual sources were selected from piping and instrumentation diagrams
or process flow diagrams before a refinery processing area was  entered.
Only those preselected sources were screened.  In this way, bias based on
observation of individual sources was theoretically eliminated.
     The screening of sources was accomplished with portable organic
vapor detectors.  The principal device used in this study was the
J. W. Bacharach Instrument Co. "TLV Sniffer".  The components were tested
on an individual basis, and only those components with VOC concentrations
in excess of 200 ppmv were considered for further study.
     A substantial portion of these leaking sources were enclosed and
sampled to determine both the methane and nonmethane emission rates.  An
important result of this program was the development of a correlation between
the maximum observed screening value (VOC concentration) and the measured
nonmethane leak rate.
     Emission factors and leak frequency information generated  during
this study are given in Table C-3.
C.I.3  Comparison of Fugitive Emissions Test Data
     The results of the SOCMI study and those of the refinery emissions
study are compared in Table C-4.  Fugitive emission leak frequencies for
similar source types appear to correlate, particularly for valves and
pump seals.
                                   C-6

-------
  TABLE C-3.   LEAK FREQUENCIES  AND  EMISSION  FACTORS FROM FUGITIVE
                        SOURCES IN  PETROLEUM REFINERIES
                                    Percent of      Estimated emission
                                  sources having        factor for
    Equipment                    screening values    refinery sources,
       type	  2.10,000 ppmv	kg/hr-source
Valves                                 NA.                   NA
  Gas service                          10                  0.021
  Light liquid service                 12                  0.010
  Heavy liquid service                  0                  0.0003

Pump seals                             NA                   NA
  Light liquid service                 23                  0.12
  Heavy liquid service                  2                  0.02

Compressor seals (hydrocarbon          33                  0.44
  service)
Pressure relief valves                  8                  0.086
  Gas service                                              0.16
  Light liquid service                                     0.006
  Heavy liquid service                                     0.009

Flanges                                 0                  0.0003

Open-ended lines                       NA                   NA
  Gas service                                              0.025
  Light liquid service                                     0.014
  Heavy liquid service                                     0.003

Source:  Ref. 5
                                  C-7

-------
         TABLE C-4.  COMPARISON OF LEAK FREQUENCIES FOR FUGITIVE EMISSION SOURCES IN SOCMI  UNITS AND PETROLEUM REFINERIES


                                                  Percent of SOCMI       Percent of petroleum refinery sources
                                              sources having screening   having screening values ^10,000 ppmv
              Equipment type	values ^.10,000 ppmva	
            Valves (all)                                 HC                                 NA
              Gas service                                                                   ]^
              Light liquid service                                                          ^
              Heavy liquid service                                                           "

            Open-ended lines (all)                       10                                 NA
              Gas service
              Light liquid service
              Heavy liquid service

«           Pump seals (all)                             17                                 NA
00             Light liquid service                                                          23
              Heavy liquid service                                                           2

            Compressor seals (hydrocarbon                43                                 33
              service)

            Pressure relief valves (all)                  0                                  8
              Gas service
              Light liquid service
              Heavy liquid service

            Flanges (all)                                 3                                  0


            jjsource:   Table C-l.
            DSource:   Table C-2.
            clncludes  block and control  valves.

-------
C.2  MAINTENANCE TEST PROGRAMS
     The results of four studies  on the effects of maintenance on fugitive
emissions from valves are discussed in this section.   The first two studies
were conducted by refinery personnel  at the Union Oil  Co. refinery in
Rodeo, California, and the Shell  Oil  Co. refinery in Martinez, California.
These programs consisted of maintenance on leaking valves containing
fluids with vapor pressures greater than 1.5 Reid Vapor Pressure.  The
third study was conducted by Radian Corporation, under contract to EPA.
Valves were selected and maintained at four refineries.  The fourth study
was conducted by EPA and EPA contractors at Unit D (ethylene unit).  The
study results and a description of each test program are given in the
following sections.
C.2.1  Description and Results of the Union Maintenance Study
     The Union valve maintenance study consisted of performing undirected
maintenance on valves selected from 12 different process units.  Undirected
maintenance consists of performing valve repairs without simultaneous
measurement of the effect of repair on the VOC concentration detected.
This is  in contrast to directed maintenance where emissions are monitored
during the repair procedure.  With directed maintenance, repair procedures
are continued until the VOC concentration detected drops to a specified level
or further reduction in the emission level is not possible.  Also, maintenance
may be curtailed  if increasing VOC concentrations result.
     The Union data was obtained with a Century Systems  Corporation
Organic  Vapor Analyzer, OVA-108.  All measurements were  taken at a
distance of 1 cm  from the seal.
     Correlations developed by EPA have been used to convert  this  data
from readings taken at one centimeter to equivalent readings at the  leak
interface.   This facilitates comparison of data from  different studies
and allows the estimation of emission rates based on screening value-
leak rate correlations.
     The results  of the Union study are given in Table C-5.  Two sets of
results  are provided; the first includes all repaired  valves with  initial
screening values  greater than or equal to 5300 ppmv, and the second
includes valves with initial screening values below 5300 ppmv.  A  screening
value of 5300 ppmv, obtained, with the OVA at 1 cm from the leak interface,
is equivalent to  a screening value of 10,000 ppmv measured by a Bacharach
                                  C-9

-------
                        TABLE C-5.  SUMMARY OF MAINTENANCE STUDY RESULTS FROM THE UNION OIL CO.

                                             REFINERY IN RODEO, CALIFORNIA3
o
I
All valves
with initial
screening values
>5300 ppmvb
Number of repairs attempted
Estimated emissions before maintenance, kg/hr
Estimated emissions after maintenance, kg/hr
Number of successful repairs (<5300 ppmv after maintenance)
Number of valves with decreased emissions
Number of valves with increased emissions
Percent reduction in emissions
Percent successful repairs
Percent of valves with decreased emissions
Percent of valves with increased emissions
133
9.72
4.69
67
124
9
51.8
50.4
93.2
6.8
All valves
with initial
screening values
<5300 ppmv
21
0.323
0.422
--
13
8
-30.5
--
61.9
38.1
          a
           Source:  Ref. 6.


          DThe value 5300 ppmv, taken with the OVA-108 at 1  cm., generally corresponds  to a value of 10,000 ppmv taken

           with a "TLV Sniffer" at 0 cm.

-------
Instrument Co.   "TLV  Sniffer" directly at  the  leak  interface.  The  OVA-
1  cm readings have  been  converted  to  equivalent TLV-0 cm  readings because:
     1)   EPA correlations  which estimate  leak rates  from screening
          values were developed from  TLV-0 cm  data.
     2)   Additional  maintenance study data exists  in the TLV-0  cm  format.
     3)   Method 21 specifies 0 cm screening procedures.
     The results of this study  indicate  that maintenance  on valves  with
initial screening values above  5300  ppm  is much more  effective than
maintenance on valves leaking at lower rates.   In fact, this study  indi-
cates that emissions  from valves are  reduced 51.8 percent for valves
initially over 5300 ppmv while  valves with lower  initial  screening  values
experienced an increase  in  emissions  of  30.5 percent.
                                                              w
C.2.2  Description and Results  of  the Shell Maintenance Study
     The Shell maintenance program consisted of two parts.  First, valve
repairs.were performed on 172 leaking valves.   In the second part of the
program, 163 of these valves were  rechecked and additional maintenance
was performed.  The second part of the program was conducted approximately
one month after the initial maintenance period.  It was not determined
whether the maintenance procedures were directed or undirected, based on
the information reported by Shell.
     VOC emissions were measured  using  the OVA-108 and readings were
obtained one centimeter from the  source.  This data has been transformed
to TLV-0 cm values as was the Union data.   And, the same methods of data
analysis described in Section C.2.1  have been  applied to the Shell  data.
The results of  the Shell maintenance study are given in Table C-6.
                                                                       g
     C.2.3  Description  and Results  of the EPA/Kadian Maintenance Study
          Repair data were  collected  on  valves located  in four refineries.
     The effects of both directed  and undirected  maintenance were
                                   C-ll

-------
                           TABLE C-6.   SUMMARY OF  MAINTENANCE  STUDY RESULTS  FROM THE SHELL OIL COMPANY
                                        REFINERY  IN MARTINEZ, CALIFORNIA
ro
March maintenance

Number of repairs attempted
Estimated emissions before maintenance, kg/hrc
Estimated emissions after maintenance, kg/hrc
Number of successful repairs (<5300 ppmv after
maintenance)
Number of valves with decreased emissions
Number of valves with increased emissions
Percent reduction in emissions
Percent successful repairs
Percent of valves with decreased emissions
Percent of valves with increased emissions
All repaired valves
with initial screening
values >5300' ppmvb
161
11.08
2.66
105
161
0
76.0
65.2
100.0
0.0
aSource: Ref. 8. 	 	
The value 5300 ppmv, taken with the OVA-108 at 1 cm., generally corresponds
GShell reported the screening value of all valves which measured <3000 ppmv
All repaired valves
with initial screening
values <5300 ppmv
11
0.159
0.0
—
11
0
100.0
--
100.0
0.0
to a value of 10,000 ppmv
(<1500 ppmv-TLV at 0 cm.)
April maintenance
All repaired valves with All repaired valves with
initial (March) screening initial (March) screening
values >5300 ppmv 
-------
evaluated and other data,  including valve size and type and the processes'
fluid characteristics,  was obtained.  Screening data were obtained with
the Pacharach Instrument Co.   "TLV Sniffer" and readings were taken as
close to the source as  possible.
     Unlike the Shell  and  Union studies,  emission rates were not based on
the screening value correlations.   Rather each valve was sampled to determine
emission rates and after maintenance using techniques developed by EPA during
the refinery emission  factor  study.  These values were used to evaluate
emissions reduction.
     The results of this study are given  in Table C-7.  Of interest here
is a comparison of the  emissions  reduction for directed and undirected
maintenance.  The results  indicate that directed maintenance is more
effective in reducing  emissions than is undirected maintenance, particu-
larly for valves with  lower initial leak  rates.  The results shown an
increase in total emissions of 32.6% for valves with initial screening
values less than 10,000 ppmv  which were subjected to undirected maintenance.
However, this increase  is  due to a large  increase in the emission rate
of only one valve.
C.2.4  Description and  Results of Unit D (Ethylene Unit) Maintenance Study
     Maintenance was performed by Unit D personnel.  VOC concentration
measurements were made  using  the OVA-108, and readings were obtained at
the closest distance possible to the source.   The results of this study
are shown in Table C-8.  Directed and undirected maintenance procedures
were used.  The results show that directed maintenance results in more
repairs being successfully completed than when undirected maintenance is used.
C.2.5  Comparison of Maintenance Study Results
     Generally speaking, the  results of these maintenance programs would
tend to support the following conclusions:
          A reduction  in emissions may be obtained by
          performing maintenance on valves with screening values
          above 10,000 ppmv (measured at  the source).
          The reduction in emissions due  to maintenance of valves
          with screening values below 10,000 ppmv is not as dramatic
          and may result in increased emissions.
                                 C-13

-------
                TABLE C-7.  SUMMARY OF EPA REFINERY MAINTENANCE STUDY RESULTS
Repaired values with initial Repaired values with initial
screening values >1 0,000 ppmv screening values <1 0,000 ppmv

Number of valves repaired
Measured emissions before maintenance
kg/hr
Measured emissions after maintenance
kg/hr
Number of successful repairs
(<1 0,000 ppmv after maintenance)
Number of valves with decreased
emissions
Number of valves with increased
emissions
Percent reduction in emissions
Percent successful repairs
Percent of valves with decreased
emissions
Percent of valves with increased
emissions
Directed
Maintenance
. 9

0.107
0. 01 39
8

9

0
87.0
88.9

100.0

0.0
Undirected Directed Undirected
Maintenance Maintenance Maintenance
23 10 16

1.809 0.0332 0.120
0.318 0.0049 0.159
13

21 6 15

2 4 1
82.4 85.2 -32.6
56.5

91.3 60.0 93.8

8.7 40.0 6.3
Source:  Ref.9

-------
                      TABLE C-8.   UNIT D ETHYLENE UNIT BLOCK VALVE  REPAIRS
    Action Level:  >10,000 ppm      Instrument:  "OVA-108" VOC detector
    Distance from Source:  Maximum concentration at seal interface
o

en
Tag Initial
Number Reading
32



28
16
10
7
4
367
366
364
362
>1 0,000
>10,000
>1 0,000
>1 0,000
>10,000
>10,000
>1 0,000
>10,000
>1 0,000
>10,000
>10,000
>1 0,000
>1 0,000
Date Maintenance
Screened Attempted
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
03/05/79
03/05/79
03/05/79
03/05/79
No
'No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Undirected Directed Maintenance
rid i ii icMaiiLc
Reading 1 2
>10,000 1,100
>10,000 >10,000 210,000
>1 0,000 >1 0,000 >10,000
2,000 100
2,000
>10,000 >10,000 >10,000
100
>10,000 >10,000
200
500

NCb
>10,000 >10,000
Readings
3 Comments
Only checked one valve
with tag — meter lines
Only checked one valve
with tag
Only checked one valve
with tag


700 Repaired when valve was
backseated

Bolts all the way down


Bolts need replacing

Leak at gland, not stem —
corrosion preventing good
seating of gland

-------
                  TABLE C-8.  UNIT D ETHYLENE UNIT BLOCK VALVE REPAIRS (Continued)
o
01
Tag
Number
360
359
None
358
361
None
356
354
352
65
64
Initial
Reading
>1 0,000
>1 0,000
>10,000
>1 0,000
>1 0,000
>10,000
>1 0,000
>1 0,000
>10,000
>1 0,000
>1 0,000
Date
Screened
03/05/79
03/05/79
03/05/79
03/05/79
03/05/79
03/05/79
03/05/79
03/05/79
03/05/79
03/06/79
03/06/79
Maintenance
Attempted
Yes
Yes
No
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
(gland)
Undirected Directed Maintenance Readings
Reading 123
2,000
4,000
»1 0,000 »1 0,000
NCb
>10,000 >10,000
>10,000 >10,000
NCb
900
NCb
3,000
1,000
>10,000 >10,000 7,000
Comments


Mistagged originally so no
initial repair attempted —
tightened bolts — needs
new packing

Leak reduced but needs new
packing
Near No. 361 — needs new
packing
Was not leaking before
maintenance (mistagged)




Leak detected by soap
Qnln-Hnn — miscpH hv-
                                                                                              instrument operator

-------
                   TABLE C-8.  UNIT D ETHYLENE UNIT BLOCK VALVE REPAIRS  (Concluded)
Tag
Number
315
311
316
313
312
314
0
Initial
Reading
>10,000
NCb
>10,000
>1 0,000
>1 0,000
>1 0,000
Date
Screened
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
Maintenance
Attempted
Yes
Yes
Yes
Yes
Yes
No
Undirected Directed Maintenance Readings
Reading 1 2 3
3,000
NCb
>10,000 2,000
>10,000 >10,000
>10,000 >10,000 5,000
>1 0,000

Comments
Drain still >10,000

All the
packing
All the
packing
Bad bol
replaci

way down on
way down on
ts — need
ng
JA11  readings are in parts per million by volume calibrated to hexane using OVA-108 detector.
3NC  =  No change detected in reading above ambient level.

-------
          Directed maintenance is preferable to undirected maintenance

          for valve repair.
     The information presented in Tables C-5, C-6, C-7, and C-8 has been

compiled with the objective of placing the data on as consistent a basis

as possible.  However, some differences were unavoidable and others may

have gone unrecognized, due to the limited amount of information concerning

the details of methods used in each study.  Therefore, care should be

exercised before attempting to draw specific quantitative conclusions based

on direct comparison of the results of these studies.

C.3.  REFERENCES

1.  Muller, Christopher, memo to files, U.S. Environmental Protection
Agency, Emission Standards and Engineering Division, Research Triangle
Park, N.C., January 18, 1979.  (Plants A & B).

2.  Muller, Christopher, memo to files, U.S. Environmental Protection
Agency, Emission Standards and Engineering Division, Research Triangle
Park, N.C., March  19,  1979.   (Plant C).

3.  Muller, Christopher, memo to files, U.S. Environmental Protection
Agency, Emission Standards and Engineering Division, Research Triangle
Park, N.C.  (Plant  D).

4.  Hustvedt, K.C., trip report  to James F.  Durham, Chief, Petroleum
Section, U.S. Environmental Protection Agency, January 5, 1979  (Plants
E & F).

5.  Wetherold, R.G., and L.P. Provost, Emission Factors and Frequency of
Leak Occurrence for Fittings  in  Refinery Process  Units.  EPA-600/2-79-044.
Radian  Corporation, Austin, Texas, February  1979.

6.  Valve Repair Summary and  Memo from F.R.  Bottomley, Union Oil Company,
Rodeo,  Calfironia, to  Milton  Feldstein, Bay  Area  Quality Management District,
April 10, 1979.

7.  Honerkamp, R.L., L.P.  Provost, J.W. Sawyer, and R.G. Wetherold,
Valve Screening Study  at Six  San Francisco Bay Area Petroleum Refineries,
Final Report.  Radian  Corporation, Austin, Texas,  February 6, 1979.

8.  Valve Repair Summary and  Memo from R.M.  Thompson,  Shell Oil Company,
Martinez Manufacturing Complex,  Martinez, California.  To Milton Feldstein,
Bay Area Quality Management District, April  26, 1979.

9.  Radian  Corporation, The Assessment of Environmental Emissions  From Oil
Refining, Draft report, Appendix B, detailed results,  EPA Contract No.
68-02-2147, Exhibit B., Austin,  Texas, August  1979.

10. Equipment Summary  from Phillips Petroleum  Company, Sweeney, Texas,
March 14, 1979.
                                   C-18

-------
APPENDIX D.   EMISSION MEASUREMENT AND CONTINUOUS MONITORING

-------
       APPENDIX D - EMISSION MEASUREMENT AND CONTINUOUS MONITORING

D.I  EMISSION MEASUREMENT METHODS
     To develop data in support of standards for the control  of fugitive
emissions, EPA conducted leak surveys at six petroleum refineries and
three synthetic organic chemical  manufacturing plants.  The resulting
leak determination procedures contained in Reference Method 21   were
developed during the course of this test program.
     Prior to the first test, available methods for measurement of
fugitive leaks were reviewed, with emphasis on methods that would provide
data on emission rates from each source.  To measure emission rates,
each individual piece of equipment must be enclosed in a temporary cover
for emission containment.  After containment, the leak rate can be
determined using concentration change and flow measurements.   This
                                           (1  2 T
procedure has been used in several studies,  ' ' and has been demonstrated
to be a feasible method for research purposes.  It was not selected for
this study because direct measurement of emission rates from leaks is a
time-consuming and expensive procedure, and is not feasible or practical
for routine testing.
     Procedures that yield qualitative or semi-quantitative indications
of leak rates were then reviewed.  There are essentially two  alternatives:
leak detection by spraying each component leak source with a  soap solution
and observing whether or not bubbles were formed; and, the use of a
portable analyzer to survey for the presence of increased organic compound
concentration in the vicinity of a leak source.  Visual, audible, or
olefactory inspections are too subjective to be used as indicators of
leakage in these applications.  The use of a portable analyzer was selected
as a basis for the method because it would have been difficult to establish
a leak definition based on bubble formation rates.   Also, the temperature
of the component, physical configuration, and relative movement of parts
often interfere with bubble formation.
                                D-l

-------
     Once the basic detection principle was selected, it was then necessary
to define the procedures for use of the portable analyzer.  Prior to
performance of the first field test, a procedure was reported that
                                                            (3]
conducted surveys at a distance of 5 cm from the components.  '  This
information was used to formulate the test plan for initial testing.  '
In addition, measurements were made at distances of 25 cm and 40 cm on
three perpendicular lines around individual sources.  Of the three
distances, the most repeatable indicator of the presence of a leak was a
measurement at 5 cm, with a leak definition concentration of 100 or 1000
ppmv.  The localized meteorological conditions affected dispersion
significantly at greater distances.  Also, it was more difficult to
define a leak at greater distances because of the small changes from
ambient concentrations observed.  Surveys were conducted at 5 cm from
the source during the next three facility tests.
     The procedure was distributed for comment in a draft control techniques
guideline documents.  Many commentors felt that a measurement distance
of 5 cm could not be accurately repeated during screening tests.  Since
the concentration profile is rapfdly changing Between 0 and about 10 cm
from the source, a small variance from 5 cm could significantly effect
the concentration measurement.  In response to these comments, the
procedures were changed so that measurements were made at the surface of
the interface, or essentially 0 cm.  Additional testing at two refineries
and three chemical plants was performed by measuring volatile organic
concentrations at the interface surface.
     A complication that this change introduces is that a very small
mass emission rate leak C'pin-hole leak") can Be totally captured by the
instrument and a high concentration result will be obtained.  This has
occurred occasionally in EPA tests and a solution to this problem has
not been found.
     The calibration basis for the analyzer was evaluated.  It was
recognized that there are a number of potential vapor stream compositions
that can be expected.  Since all analyzer types do not respond equally
to different compounds, i~t was necessary to establish a reference
calibration material.  Based on the expected compounds and the limited
                                  D-2

-------
information available on instrument response factors,  hexane was  chosen
as the reference calibration gas  for EPA test programs.   At the 5 cm
measurement distance, calibrations  were conducted at approximately 100
or 1000 ppmv levels.   After the measurement distance was changed,
calibrations at 10,000 ppmv levels  were required.  Commentors pointed
out that hexane standards at this concentration were not readily  available
commercially.  Consequently, modifications were incorporated in the
method to allow alternate standard  preparation procedures or alternate
calibration gases.

     The alternative of specifying  a different calibration material for
each type stream and normalization  factors for each instrument type was
not intensively investigated.  There are at least four instrument types
available that might Be used in this procedure, and there are a large
number of potential stream compositions possible.  The amount of prior
knowledge necessary to develop and  subsequently use such factors  would
make the method prohibitively complicated.  Based on EPA test results,
the number of concentration measurements in the range where a variability
of 2 or 3 would change the decision as to whether or not a leak exists
is small in comparison to the total number of potential  leak sources.
     An alternative approach to leak detection was evaluated by EPA
during field testing.  The approach used was an area survey, or walk-
through, using a portable analyzer.  The unit area was surveyed by
walking through the unft positioning the instrument probe within 1 meter
of all valves and pumps.  The concentration readings were recorded on a
portable strip chart recorder.  After completion of the walkthrough, the
local wind conditions were used'with the chart data to locate the
approximate source of any increased ambient concentrations.  This
procedure was found to yield mixed results.  In some cases, the majority
of leaks located by individual component testing could be located by
walkthrough surveys.  In other tests, prevailing dispersion conditions
and local elevated ambient concentrations complicated or prevented the
interpretation of the results.  Additionally, it was not pqs.s.tb.le to
                               D-3

-------
develop a general criteria specifying how much of an ambient increase at
a distance of 1 meter is indicative of a 10000 ppm concentration at the
leak source.  Because of the potential variability in results from site
to site, routine walkthrough surveys were not selected as a reference or
alternate test procedure.

D.2  CONTINUOUS MONITORING SYSTEMS AND DEVICES
     Since the leak determination procedure is not a typical emission
measurement technique, there are no continuous monitoring approaches
that are directly applicable.  Continual surveillance is achieved by
repeated monitoring or screening of all affected potential leak sources.
A continuous monitoring system or device could serve as an indicator
that a leak has developed between inspection intervals.  EPA performed a
limited evaluation of fixed-point monitoring systems for their effective-
ness in leak detection.  The systems consisted of both remote sensing
devices with a central readout and a central analyzer system (.gas chromatograch)
with remotely collected samples.  The results of these tests indicated
that fixed point systems were not capable of sensing all leaks that were
found by individual component testing.  This is to be expected since
these systems are significantly affected by local dispersion conditions
and would require either many individual point locations, or very low
detection sensitivities in order to achieve similar results to those
obtained using an individual component survey.
     It is recommended that fixed-point monitoring systems not be required
since general specifications cannot be formulated to assure equivalent
results, and each installation would have to be evaluated individually.

D.3  PERFORMANCE TEST METHOD
     The recommended fugitive VOC emission detection procedure is Method 21
This method incorporates the use of a portable analyzer to detect the
presence of volatile organic vapors at the surface of the interface
where direct leakage to the atmosphere could occur.  The general approach
of this technique assumes that if an organic leak exists, there will be
an increased vapor concentration in the vicinity of the leak, and that
the measured concentration is generally proportional to the mass emission
rate of the organic compound.

                                 D-4

-------
     Method 21  is designed for use in many different source categories



and does not include the specification of a specific compound in instrument



calibration or a leak definition in terms of VOC concentration.   These



criteria are given in the applicable standard.



     There are at least four types of detection principles currently



available in commercial portable instruments.   These are flame ioniza-



tion, catalytic oxidation, infrared absorption  (NDIR) and photoionization.



Two types (flame ionization and catalytic oxidation) are known to be



available in factory mutual certified versions  for use in hazardous



atmospheres.



     The recommended test procedure includes a  set of design and operating



specifications and evaluation procedures  by which an analyzer's  performance



can be evaluated.  These parameters were  selected based on the allowable



tolerances for data collection, and not on the  performance of individual



instruments.  Based on manufacturers'  literature specifications, many



commercially available analyzers can meet these requirements.



     The estimated purchase cost for an analyzer ranges from about $1000



to $5000 depending on the type and optional equipment.  The cost of an



annual monitoring program per unit, including semiannual instrument



tests and reporting is estimated to be from $3,000 to $4,500.  This



estimate is based on EPA contractor costs experienced during previous



test programs.  Performance of monitoring by plant personnel may result



in lower costs.  The above estimates do not include any costs associated



with leak repair after detection.





D.4  REFERENCES



     1.  "JOINT DISTRICT, FEDERAL, and STATE PROJECT for the Evaluation



of Refinery Emissions", Los Angeles County Air  Pollution Control District,



Nine Reports, 1957-1958.




                             D-5

-------
     2.  "Emission Factors and Frequency of Leak Occurrence Fittings in



Refinery Process Units" Radian Corporation Contract No.  68-02-2147 and



No. 68-02-2665, EPA Report No. 600/2-79-044,  February 1979.



     3.  Telephone Communication:   Paul  Harrison, Meteorology Research,



Inc. to K.  C. Hustvedt, EPA, December 22, 1977.



     4.  EMB Report No. 77-CAT-6,  "Miscellaneous Refinery Equipment VOC



Sources at Arco, Watson Refinery and Newhall  Refining Co." ESED,  EPA,



December, 1979.



     5.  "Control of Volatile Organic Compound Leaks from Petroleum



Refinery Equipment," OAQPS Guideline Series,  EPA-45Q/2-78-036, June, 1978.
                                 D-6

-------
APPENDIX £.  METHODOLOGY FOR ECONOMIC ANALYSIS

-------
        APPENDIX El:   WEIGHTING  PROCEDURES FOR THE  ESTIMATION OF SOCMI
                              TIME-SERIES DATA

     The chemicals produced by  the  synthetic organic chemicals manufacturing
industry (SOCMI) do not directly correspond to the  industrial classifications
of organic chemicals  used  by  sources reporting production  and sales statis-
tics.   Consequently,  the weighting procedure described below was used to gen-
erate data that reflect the SOCMI as accurately as  possible.
     Production and  sales  data  for synthetic  organic  chemicals are reported
annually  in  the United  States  International  Trade  Commission (ITC) report,
Synthetic Organic Chemicals:   U.S. Production and Sales.   The report presents
production, quantity  of sales  and  value  of sales  data  for 14 categories of
chemicals.  Four  of  these categories,  Tar and Tar  Crudes, Primary Products
from Petroleum and Natural  Gas for Chemical Conversion, Cyclic  Intermediates,
and Miscellaneous  Cyclic and  Acyclic Chemicals, contain  SOCMI chemicals, as
well as  other  chemicals.  To derive appropriate  estimates  of data for the
SOCMI  chemicals,  production,  quantity  of sales and value  of sales for 1977
were estimated  for all  SOCMI  chemicals included in each of the four categor-
ies.   The  estimates  for SOCMI chemicals  for  each  variable were then divided
by  the  aggregate  estimates for  all chemicals within the category.  Resulting
ratios were used as weights to calculate estimates  of production, quantity of
sales  and  value of sales of SOCMI chemicals each in category  over the period
1974-1978.
     Prior  to  1975,  the chemicals  included  in the  category, Miscellaneous
Cyclic and Acyclic Chemicals,  were reported as Miscellaneous  Synthetic Organ-
ic  Chemicals.  A weighting  scheme based  on 1974 data  for this category was
developed  using the  procedure described above and was  used to estimate pro-
                                    E-l

-------
duction and  sales of  SOCMI  chemicals  in this category for  the  period 1968-



1974.   Data on production and sales of SOCMI for the remaining three categor-



ies for the period 1968-1974 were estimated using the 1977 weights.



     Table  El-1  presents  the SOCMI  chemicals  included  in  each of  the  ITC



categories. Table  El-2 presents  the estimated ratios used to weight the ITC



data in order  to calculate production, quantity of  sales and value of sales



of SOCMI chemicals.
                                    E-2

-------
         TABLE £1-1.   SOCMI CHEMICALS INCLUDED IN ITC CATEGORY RATIOS
Tar and Tar Crudes*

Benzene
Toluene
Xylene
Solvent Naptha
Primary Products from Petroleum and Natural Gas for.Chemical Conversion
Benzene
Cumene
Cyclohexane
Ethyl benzene
Napthalene
Styrene
Toluene
Xylenes
o-Xylene
p-Xylene
All other aromatics and napthenes
Acetylene

Cyclic Intermediates

Aniline
Benzoic acid
Biphenyl
Cresols
Cresylic acid, refined
Cyclohexanone
Cylohexylamine

Miscellaneous Cyclic and Acyclic Chemicals

Benzyl alcohol
Caprolactam
Dioxane
p-Hydroxybenzoic acid
Maleic anhydride
Ethanolamines
Acenitrile
Acrylonitrile
Acetic acid
Acetic anhydride
Acrylic acid
Adipic acid
Fumaric acid
Propionic acid
Formaldehyde
Isobutyraldehyde
Acetone
Ethylene
Propylene
Butadiene and butylene fractions
1, 3-Butadiene, grade for rubber
1-Butene
Isobutylene
Isoprene
Dodecene
Petenes
Nonene
Polybutene
o-Dichlorobenzene
p-Dichlorobenzene
Hydroquinone
a-Methylstyrene
Nitrobenzene
Nonylphenol
Phenol
Butyl amines
Ethyl amines
Isopropylamirie, mono-
Methyl amines
All other amines
Pentaerythritol
Propylene glycol
All other polyhydric alcohols
2-(2-Butoxyethoxy) ethanol  (Diethy-
  lene glycol monobutyl ether)
2-[2-(2-Butoxyethoxy) ethoxy]
  ethanol (Triethylene glycol,
  monobutyl ether)
Diethylene glycol
Dipropylene glycol
2-(2-Ethoxyethoxy) ethanol  (Diethy-
  lene glycol monoethyl ether)
                                    E-3
                         (continued)

-------
                           TABLE El-1.   (continued)
Miscellaneous Cylic and Acyclic Chemicals (continued)
2-Butanone (methyl ethyl ketone)
4-Hydroxy-4-methyl-2-pentanone
  (Diacetone alcohol)
4-Methyl-2-pentanone (methyl
  isobutyl ketone)
4-methyl-3-penten-2-one (Mesityl
  oxide)
All other ketones
n-Butyl alcohol (n-propylcarbinol)
Methanol
n-Butyl acetate
Ethyl acetate
Ethyl aerylate
Isobutyl acetate
Methyl acetate
Methyl methacrylate
Vinyl 'acetate
Ethylene glycol
Glycerol, synthetic
1,1,1-Trichloroethane (methyl
  chloroform)
Trichloroethylene
Vinyl chloride, monomer (chloro-
  ethylene)
All other chlorinated hydrocar-
  bons
Chiorodi f1uoromethane
Dichlorodi.fi uoromethane
2-Methoxyethanol (ethylene glycol
  monomethyl ether)
2-(2-Methoxyethoxy) ethanol (Diethy-
  lene glycol monomethyl ether)
Polyethylene glycol
Polypropylene glycol
Propylene glycol
Triethylene glycol
Carbon tetrachloride
Chloroethane (ethyl chloride)
Chloroform
Chloromethane (methyl chloride)
1-2-Dichloroethane (ethylene
  dichloride)
Dichloromethane (methylene chloride)
1-2-Dichloropropane (propylene di-
  chloride)
Tetrachloroethylene (perchloro-
  ethylene)
Trichlorofluoromethane
Carbon disulfide
Ethylene oxide
Ethyl ether
Propylene oxide
All other epoxides, ethers, and
  acetals
Phosgene (carbonyl chloride)
^Derived from coal.  Does  not duplicate data contained in other categories.
                                     E-4

-------
                  TABLE El-2.   RATIOS USED TO WEIGHT ITC DATA3
Production
(1000 Ibs)
Quantity
(1000 Ibs)
Value
(1000 $)
Tar and Crudes

Total SOCMI Chemicals               3,265,976          1,543,188          976,660
ITC Grand Total                      4,145,815          2,009,737        1,104,285
Ratio                       .           78.78%             76.79%           88.44%

Primary Products from
Petroleum and Natural Gas

Total SOCMI Chemicals              93,517,108         38,873,142        4,120,327
ITC Grand Total                    126,133,316         61,008,376        5,820,390
Ratio             '  .   .                74.14%             63.72%           70.79%

Cyclic Intermediates

Total SOCMI Chemicals              15,699,616          6,139,015        1,878,235
ITC Grand Total                     18,725,626          7,985,790        2,596,627
Ratio                                  83.84%             76.84%           72.33%

Miscellaneous Cyclic
and Acyclic Chemicals

Total SOCMI Chemicals              65,876,154         27,695,411        4,734,676
ITC Grand Total                     86,968,069         38,753,311        7,919,082
Ratio                                  75.75%             71.47%           59.79%

Miscellaneous Chemicals
1974 Figures

Total SOCMI Chemicals              73,670,360         34,817,621        3,607,825
ITC Grand Total                    100,604,375         47,430,967        7,815,487
Ratio                                  73.23%             73.41^           46.16%

 1977 figures except where indicated.
                                    E-5

-------
              APPENDIX E2:   REPLACEMENT INVESTMENT PROJECTIONS

     The  methodology  used  to project  SOCMI  replacement  investment  is  de-
scribed in this  appendix.   The projections are based  on  two key theoretical
assumptions:    (I)  the historical  growth  rate of capacity,  p,  has  been con-
stant  over time;  and  (II) model  units  have  a fixed life  of L  years.   These
assumptions are summarized in the following equations:
     KT = (l+p)L KT_L                         (1)
     IT = PKT + RT                            (2)
     RT = IT_L                                (3)
where
     K =  industry capacity,
     I =  gross investment,
     R =  replacement  investment,
     T =  time subscript,
and  K, I and R are measured  in terms of model units.
     Equation  (1)  is  an algebraic restatement of assumption I.   Equation (2)
is  simply a  mathematical  definition  of  gross   investment,  that  is,  gross
investment, I-,-,  is  equal  to  additions to new capacity, pKT, plus replacement
investment, R-.-.   Equation  (3) is an algebraic restatement of assumption II.
Appropriately  lagging Equation (2) and back  substituting  from  (2)  into (3),
it can be shown that

     RT = p ^ KT_iL                           (4)

Further,  by substituting for  the  various KT  .. in Equation (4) using Equation
(1), and  rearranging  terms, the following result  is obtained:
                                    E-6

-------
     RT = KT-I   Z     P                          (5)
                        i I
               1=0 (1+p)1


                oo

The expression  I     p      is  a constant and,  if p  is assumed to be 0.06 and


               i=0
L to be  20,  approximately  equal  to 0.087.  Equation  (5)  can be used to pro-


ject replacement  investment  in any  year, T,  if  an estimate  of  the capital


stock  in  the (T-L)th year is  available.   For SOCMI, capital  stock  data are


available for  1976.   This information,  together with  an  assumed historical


growth rate  of 6 percent, was used  to  estimate the  capital  stock for the


years  1961  to  1965  by  means of  Equation (1).  The  resulting capital  stock


estimates are then used  in Equation (5)  to project replacement investment in


SOCMI  for each of  the  five years  following proposal of any regulatory alter-


natives (1981-1985),  on the basis  of the  empirical  assumption that each model


unit has  a  life  of  20  years.  The  annual  projections of replacement invest-


ment  are then summed  to obtain  a  projection of  the number  of  replacement


facilities  subject  to  the provisions  of any  regulatory alternative  in the


fifth year following its proposal.  The projections of replacement investment


obtained by applying this methodology are presented in Table E2-1.
                                    E-7

-------
       TABLE E2-1  PROJECTIONS OF REPLACEMENT INVESTMENT

                        Number of replacement capacity units
Year                    Annual                    Cumulative

1981                      49                          49
1982                    .  51                         100
1983                      55                         155
1984                      58                         213
1985                      61                         274
                             E-8

-------
          APPENDIX  E3:   METHODOLOGY  FOR  COMPUTING  COST OF  CAPITAL


                 TO SYNTHETIC  ORGANIC  CHEMICAL MANUFACTURERS






     This appendix  describes  the  process used to estimate  the cost of capital


for the  chemical  industry.   The  cost of  capital  for any new  project  is  the


cost  of  equity, debt,  and preferred  stock, weighted  by the  percentage  of


funds generated by  each type  of financing, that is,




     k   =  k   -  + k  -  +  k   -                  (I)
     kc      el    Ki I    kp  I                  U;
where
         k   =.  cost of capital


         k   =  cost of equity capital


         k.  =•  cost of debt capital


         k   s  cost of preferred stock capital


         E   =  the amount of equity used to finance a given investment


         D   =  the amount of debt used to finance a given investment


         P   =  the amount of preferred stock used to finance a given

                investment


         I  '=  the total funds needed for the investment
     The first  step  in  estimating Equation (1)  is  to determine the relevant


weights for  the three  types of financing.   It is assumed that the proportion


of  debt,  equity,  and preferred stock, to be  used on  any  new  project will be


the same as  currently  exists in the  firm's  capital  structure.   This implies


that the firm  is  currently using the optimal  mix of financing.  Figures for


the three  types of funds came from the COMPUSTAT tapes, supplied by Standard


&  Poor's  Corporation,  for  each  firm's fiscal  year  ending  in  1977.  Common


equity  included the par  value  of common  stock, retained  earnings, capital




                                    E-9

-------
surplus,  self-insurance  reserves, and  capital  premium, while  debt included

all obligations  due  more than a year  from  the  company's balance sheet date.

Preferred stock represented the net number of preferred shares outstanding at

year-end multiplied by the involuntary liquidating value per share.

     The  next  step  in  calculating Equation  (1)  is to  estimate  the cost of

equity financing.  Two approaches are commonly used: the results derived from

the  capital-asset pricing  model   (CAPM)  and  the  results  derived  from  the

dividend capitalization model (DCM).  The CAPM examines the necessary returns

on  a  firm's  stock  in  relation  to a  portfolio  comprised of  all existing

stocks, while the DCM evaluates the stream of dividends and the discount rate

needed to  arrive at the firm's existing share price.  The  required  return on

equity using the CAPM is:

       ke  =  i + p  (km-i)                           (2)

where
        i  =  the expected risk free interest rate

       k -i =  the expected excess return on the market, and

        P  =  the firm's beta coefficient.

The required return  on equity using the DCM is:

             Di
       ke  =  r,   *«                                (3)

where
       D,  =  the dividend expected  in period  1

       P   =  the share price at the beginning of  period 1

       g  =  the expected rate of dividend growth,  assumed  to be constant.

     Figures  for Equation  (2)  were developed  in the  following manner.  The

expected  risk-free  rate was assumed equal to the yield on  a 3-month Treasury

Bill,  as  reported in the October  1,  1979,  Wall Street Journal.  The current

yield  was  10.46 percent.  This corresponds to  the yield from  a bond with no


                                    E-10

-------
possibility  of  default  and  offering  no chance  of  a  capital  loss  and  is
therefore riskless.   The firm's beta coefficients came from the September 24,
1979,  Value Line.Investment  Survey.    The   expected  excess  return  equalled
2.9646  percent,  the  5-year  average  (July   1974-June  1979)  of the  monthly
excess returns on the Standard & Poor's .500  Stock Index multiplied by twelve.
     Figures  for  Equation (3)  came  from two sources.  Both  share  price and
expected yearly dividends  came from figures reported in the October 1, 1979,
Wall Street Journal.   The  growth  rate  was  calculated from  data contained on
the  COMPUSTAT  tapes.   Three  different growth rates  were tried:  the  5-year
average  growth  of  total  assets,  the  5-year average  growth  of  per share
earnings, and the 5-year average growth of  dividends.
     A number of  theoretical  reasons exist  for preferring  the CAPM approach
to  the DCM  for  estimating the required return  on  equity, but  the  figures
calculated  revealed  a more practical  justification.   Using growth estimated
from  per  share earnings  or  dividends' resulted  in a number  of firms  having
negative required returns  with the DCM method.   Although using the growth in
assets  resulted  in  only  one  firm with  a  negative  required  return,  several
firms  had  extremely  low returns (less  than  10 percent).   It  is unreasonable
to expect that stockholders would  demand a  return on their stock that  is less
than  the existing yield on Treasury Bills,  yet all three variants of  the DCM
method  led   to   this  conclusion   for  a   number   of   firms.    From  these
considerations it was decided  to  use  the  CAPM calculations  as the required
return on equity.
     The third  step   in  estimating Equation  (1)  is  calculating  the cost of
debt financing.   This would be a relatively  easy  estimation if interest rates
did  not change over  time.   Past yields on  old issues of bonds would suffice.
Since  interest rates  have  been increasing,  it was felt  that  a more forward-
                                    E-ll

-------
looking rate was required.  The method selected was to take the average yield

as given  in the October  1 - September 3,  1979,  Moody's  Bond  Survey  for the

firm's  bond ratings  class  as  the  necessary yield  the  firm must offer  on

long-term  debt.    The  firm's  ratings  class  came from  the  September  1979

Moody's Bond Record or the 1979 Moody's Industrial Manual.  A small number of

firms were  not  rated by Moody's.  One firm was ranked in Standard and Poor's

Bond Guide  and  this  was  used  to  approximate  a  Moody's  bond   class.

Information on other  firms was contained in the 1979 Moody's Industrial Manual

or   the   Standard & Poor's Corporation  Records,    concerning   bank   notes,

revolving  credit,   or term-loan  agreements that  tied the  interest  rate  on

these  types  of debt  to  the  current  prime rate.   This  was  used   as  the

necessary yield on  long-term debt.  Table E3-1 presents the yields by ratings

class and  the  prime  rate  (as of  October 1, 1979) used for  the  cost  of debt

funds.

      TABLE E3-1.   YIELDS  BY RATING CLASS FOR COST OF  DEBT FUNDS,  1979
                            (prime rate = 13.50 %)


          Ratings  Class                            Yield  (percent)

               AAA                                     9.25

                AA                                     9.59

                 A                                     9.72

               BAA                                     10.38

                BA                                     11.97

                 B                                     12.395
     The yield  on  long-term  debt  does  not  represent  the  aftertax cost of debt

 financing  since interest  charges  are tax deductable.  To arrive at the after-

                                    E-12

-------
tax cost,  the  yield must be multiplied  by  one minus the marginal  tax  rate.
        k1  =  k(l - t)
where
         k  =  the yield on bonds
         t  =  the marginal tax rate
It is assumed that the firms in the sample are profitable, so that taxes must
be paid, and that their marginal tax rate is 48 percent.
     The  last  step in  estimating  Equation  (1)  is to arrive at  the  cost of
preferred  stock financing.   Unlike debt,  preferred stock  does  not have a
maturity date,  so  that the current yield should approximate the yield on new
issues.  The yield is:

     V;
where
          D  =  stated annual dividend
          P  =  the price of a share of preferred stock
The figures for dividends and share price came from the October 1, 1979, Wall
Street Journal  or,  if not included in this source, from the January 1,  1979,
listing  in the  Daily Stock Price Record.   A number of  firms did  not have
their  preferred stock  listed  in either  source,  yet had  preferred stock in
their  capital  structures.  All  used  less  than  15 percent  preferred stock,
with  the  majority  less   than  5  percent.    For  these   firms  the yield on
preferred stock was set equal to the yield on  long-term debt.
     Table  E3-2 lists the cost  of  capital  for all 100  firms  in  the. sample,
along  with  some of  the components of  Equation  (1).   These firms represent
the  best available sample  of the  approximately 600 firms  in the industry.
However,  it is likely that on  the average  they  are larger  than  the  firms
                                    E-13

-------
TABLE E3-2.   FINANCIAL DATA FOR 100 FIRMS IN SOCMI
                                                  1-11
Name
Abbott Labs
Akzona
Alco Standard Corp.
Allied Chem Corp.
American Cyanamid
Armco Steel Corp.
Atlantic Richfield
Beatrice Foods
Bendix Corp.
Bethlehem Steel Corp.
Borden Inc.
Borg-Warner Chem.
Brown Co.
CPC International
Inc.
Celanese Corp.
Charter International
Oil
Cities Service Co.
Combustion
Engineering
Continental Oil
Crompton & Knowles
Dart Indust.
Dayco Corp.
De Soto, Inc.
Diamond Shamrock
Corp.
Dow Chemical
Du Pont De Nemours
Eastern Gas & Fuel
Associates
Essex Chem. Corp.
Cost of
Capital
12.014
10.276
12.151
10.091
11.083
10.588
9.749
11.232
11.118
10.913
10.484
11.863
9.813

11.638
10.181

9.175
10.395

11.494
10.881
11.298
10.689
8.270
11.499

9.790
10.060
11.328

11.605
12.502
Return
On
Equity
14.018
13.276
13.425
13.721
13.425
13.276
13.128
12.832
13.425
14.018
12.683
13.128
12.387

13.128
13.128

14.166
12.980

14.314
13.721
13.425
14.166
12.980
13.128

13.721
14.018
13.573

14.018
14.166
Return
On
Debt
9.590
10.380
15.120
9.720
9.590
9.720
9.590
9.250
9.720
9.720
9.590
9.720
12.395

9.590
11.970

12.395
9.720

9.720
9.590
14.450
9.720
11.970
13.750

9.720
9.590
9.250

14. 180
12.395
Return
On'
Preferred
Stock
__a
--
--
-- •
--
6.461
--
7.429
3.333
--
--
--
. •

--
10.084

--
--

--
2.564
--
4.211
6.071
--

--
--
8.654

--
--
Proportion
Of
Equity
.77262
. 61914
. 64134
.58118
.72252
. 66880
.51602
.79803
.72911
.65360
.71317
.82756
.56680

.81691
.53511

.27557
.67388

.68700
.67568
.53329
.63113
.30351
. 72746

.54639
.56176
.72512

.63681
. 78453
Proportion
Of
Debt
.216575
. 380859
.259343
. 418825
.277480
. 306858
.362174
.194329
. 248140
. 346402
.285155
.145263
.433202

. 183087
.396896 '

.623167
.326120

.296229
.321308
.375634
.231645
.666445
.272535

.453615
.438236
.232172

.363188
.215465
Proportion
Of
Preferred
Stock
. 010804
.000
.099317
.000 •
.000
.024337
.121802
.007644
.022754
.000
.001677
.027181
.000

.000
.067997

.101265
.000

.016774
.003009
.091078
.137221
. 030044
.000

.000
.000
.042712

.000
.000

-------
TABLE E3-2  (Continued)
Name
Exxon Corp.
FMC Corp.
Ferro Corp.
Firestone Tire &
Rubber
Ford Motor Co.
GAF Corp.
General Electric Co.
General Motors Corp.
General Tire & Rubber
Georgia-Pacific Corp.
Goodrich (B.F.) Co.
Goodyear Tire &
Rubber Co.
Gulf Oil Corp.
Hercules Inc.
Inland Steel
Insilco Corp.
Interlake, Inc.
International
Harvester
Kaiser Steel Corp.
Kraft Inc.
Marathon Oil Co.
Martin Marietta Chem.
Mead Corp.
Merck & Co.
Minnesota Mining &
Manuf .
Mobil Oil Corp.
Monsanto Co.
Morton-Norwich
Products
Cost of
Capital
11.875
10.183
12.369

10.610
12.069
9.398
12.130
12.798
11.440
10.793
10.430

10.101
11.745
11.177
10.092
9.339
11.331

10. 534
11.688
10.774
9.582
11.238
10.000
12.309

12.572
10.868
10.970

10.726
Return
On
Equity
13.276
13.573
13.276

12.980
13.276
13.573
13.721
13.425-
13.276
13.573
13.276

12.980
12.980
13.869
12.980
13.276
13.128

13.573
14.018
12.683
13.128
13.276
13.869
13.573

13.869
13.128
13.573

13.721
Return
On
Debt
9.250
9.720
9.720

9.720
9.250
10.380
9.250
9.250
11.970
9.590
10.380

9.720
9.250
9.720
9.590
11.970
9.720

9.720
14.000
9.250
9.720
9.720
9.720
9.250

9.250
9.250
9.590

9.720
Return
On
Preferred
Stock
_ _
6.250
--

--
--
7.559
--
8.715
•
--
8.864

--

--
--
7.752
--

--
--
--
--
--
4.308
--

--
--•
5.000

--
Proportion
Of
Equity
.83450
.59257
.88968 .

.70096
.85743
. 44490
. 82148
.91962
.73287
.67625
.62957

.63679
. 84880
.69461
.62702
. 41885
.77736

.63297
.63274
.75752
.56074
.75212
.56423
.85481

.85677
.72833
.69690

.65441
Proportion.
Of
Debt
.165504
.339730
.110317

.299038
.142565
.387035
.178521
.063516
.258968
.323751
.349707

.363210
.151203
.305394
.352735
.475634
.222640

.348230
.345717
. 242479
.439257
. 247882
. 398718
.143358

.143235
.271665
.300335

.345589
Proportion
Of
Preferred
Stock
.000
.067701
.000

.000 .
.000
.168061 •
,000
.016862
.008163
.000
.020723

.000
.000
.000
.020249
.105511
.000

.018796
.021539
.000
.000
.000
.037048
.001827

.000
.000
..002767

.000

-------
TABLE E3-2  (Continued)
Name
National Distillers
& Chem.
National Steel Corp.
Northwest Indust.
Owens-Corning
Fiberglass
PPG Industries
Penwalt Corp.
Pfizer
Phillips Petroleum Co.
Procter & Gamble Co.
Quaker Oats Co.
Reeves Bros. Inc.
Reichold Chems.
Republic Steel Corp.
Riegel Textile Corp.
Rockwell International
Rohn and Haas Co.
SCM Corp.
Scott Paper Co.
Shakespeare Co.
Sherwin-Williams Co.
Squibb Corp.
A. E. Staley Mfg. Co.
Stauffer Chemical Co.
Sterling Drug
Sun Chem. Corp.
Sybron Corp.
Tenneco Inc.
Texaco
Texfi Indust.
Textron Inc.
Union Camp Corp.
Union Carbide Corp.
Cost of
Capital

11.037
9.909
8.015

11.653
10.596
9.013
11.244
11.670
11.824
10.946
10.629
10.647
11.305
11.201
9.589
10.739
10.835
10.784
11.229
9.617
11.266
10.428
10.188
12.595
10.427
10.786
9.155
11-230
10.090
10.085
11.359
10.775
Return
On
Equity

13.128
12.683
13.869

13.425
13.276
13.276
14.018
13.721
13.276
13.573
12.535
13.425
13.425
12.980
12.535
13.721
14.018
13.721
13.276
12.980
14.018
13.573
13.425
13.276
13.573
13.869
12.980
12.980
13.275
13.425
13.276
13.573
Return
On
Debt

9.720
9.590
10.380

9.720
9.590
9.720
9.590
9.250
9.250
9.720
10.380
10.380
9.720
11.970
9.720
9.720
10.380
9.590
14.000
10.380
9.590
9.720
9.720
9.590
12.395
9.720
10.380
9.250
16.000
9.720
9.590
9.590
Return
On
Preferred
Stock

9.193
--
2.9412

--
.
7.529
--
--
--
9.008
--
--
--
--
5.398
'
--
--
--
10.00
--
--
--
--
--
--
3.887
--
--
6.222
--
--
Proportion
Of
Equity

.73310
.63946
.32561

. 78828
.67661
.41712
.69289
. 76982
.82842
.651578
. 732870
.571986
. 746819
.736598
.602132
.655939
.630766
.660791
.658505
. 523981
.695345
. 629947
.613351
. 917816
. 558689
.616191
. 505890
.785863
.356904
.577353
. 768639
.674170
Proportion
Of
Debt

.251565
.360538
.617085

.211721
.323394
.369200
.307113
.230179
. 171428
.262094
.267130
.295871
.253181
.263402
.309032
. 344061
.369234
.333680
. 341495
.422439
.304655
. 368508
. 386649
. 082184
.441311
.319517
.442129
.214137
.643096
.252757
.231361
.325830
Proportion
Of
Preferred
Stock

.015334
.000
.057301

.000
.000
.213675
.000
.000
.000153
.086328
.000
. 132143
.000
.000
. 088836
.000
.000
.005529
.000
.053579
.000
. 001544
.000
.000
.000
.064292
.051981
.000
.000
.169890
.000
.000

-------
                                              TABLE  E3-2   (Continued)
m



Name
Union Oil , Calif.
Uni royal
U.S. Gypsum
U.S. Steel Corp.
Upjohn Co.
Vulcan Materials Co.
Walter (Jim) Corp.
Westinghouse Electric
Corp.
.Weyerhaeuser Co.
Wheeling-Pittsburgh
Steel
Whittaker Corp.
Wit Chem. Corp.


Cost of
Capital
10.577
10.514
10.726
10.919
11.052
10.675
9.019

12.596
10.402

11.238
10.070
10.736

Return
On
Equity
13.128
13.425
13.276
13.573
13.573
12.980
13.721

14.018
14.166

13.869
14.314
13.573

Return
On
Debt
9.590
11.970
9.590
9.590
9.590
9.720
11.970

9.720
. 9.590

14.000
11.970
9.720
Return
On
Preferred
Stock
	
16.000
5.539
--
--
--
4.444

8.837
.5.957.

12.739
--
3.313

Proportion
Of
Equity
.663994
.521603
.686341
. .690912
.706383
.709218
. 398726

.838775
.583685

.512893
.457808
.673790

Proportion
Of
Debt
.295934
.423786
.223477
.309088
.293617
.290782
.491966

.155115
.357341

.381136
.517470
.292825
Proportion
Of
Preferred
Stock
. 040072.
.054611
.090182
.000
.000
.000
.109308

.006110
..058973

.105972
.024722
.033385
     Dashes  indicated  that  data  are  unavailable.

-------
excluded,  as  many  small  firms  do  not  have  to publish  detailed financial



records.   This potential  bias  in the sample of  firms  used may have resulted



in a slight underestimation of the industry's cost of capital.12
                                     E-18

-------
                 APPENDIX E4:   METHODOLOGICAL CONSIDERATIONS








Price and Rate of Return Impacts



     Let  P  denote  product  price,  Q denote  unit  output,  TOC  denote  total



operating costs,  K  denote  the  amount of  capital  invested  in the  unit,  r



denote  the  rate of  return  on capital and  t  denote the tax  rate  in  a  given



year.  The aftertax  rate  of return on capital  invested  in  the unit may then



be defined as:



            = (1-t) (PQ - TOC)
                    K



where  [PQ  - TOC]  is  the unit's  pretax net revenues  from  its operations in



that year.   Now,  assume  that  the unit  is required  to  change its operating



costs  and  level of  capital  investment in  order  to comply with the implemen-



tation of  some  regulatory alternative.   Under the  full  cost absorption sce-



narios the unit will be unable to adjust the the price of its product or unit



output.  Consequently, the rate of return on investment, r, will change.  The



formula used to estimate this  impact is obtained  by totally differentiating



Equation (1) with respect to TOC and K; that is,



           .     _  (1-t) dTOC     (1-t) (PO-TOC) dK
          or -                 +
Substituting in (2) from (1) and rearranging terms, it follows that:



          -dr = (1-t) dTOC + rdK                       (3)

                       K



Equation  (3),  identical  to Equation (2) in  section  9.2,  is the formula used



to  calculate  the  full  cost absorption  rate of  return  impacts presented in



Chapter 9.



     Price  impacts are estimated  on the basis of the  assumption that firms



will be  able  to maintain the preregulation  rate  of  return (r) by  increasing





                                     E-19

-------
product prices.   Thus,  r  is  now a  constant  and P a  variable.   Rearranging



terms in Equation (1), it may be shown that:



     p = TOC + r K/(l-t)                 (4)





In  full  cost  pass  through scenarios,  changes  in TOC  and K  leave  r and  Q



unaffected  but result  in  a change  in  P.   The  formula for estimating  this



change in P may be obtained by total  differentiating Equation  (4) with



respect to TOC and K; that is,



     dp = dTOC + rdK /(1-t)               (5)





Equation (5),  identical  to Equation  (1) in Section 9.2,  is the formula used



to  estimate  the  full  cost pass through price  impacts  presented in Chapter 9.
                                    E-20

-------
                            APPENDIX E REFERENCES
 1.   COMPUSTAT.   New York:.  Standard & Poor1 s Corporation, 1978.

 2.   Daily Stock Price Record.    New  York:   Standard  &  Poor's Corporation,
     1979.                                            .

 3.   Moody's Bond Record.   New  York:   Mo.ody's.  Investors  Service,  Inc., Sep-
     tember 1979.

 4.   Moody's Bond Survey.   New York:   Moody's  Investors Service,  Inc., Octo-
     ber 1 - September 3,  1979.

 b.   Moody's Industrial  Manual.   New  York:   Moody's Investors  Service, Inc.,
     1979.

 6.   Scherer, F.  M. ,  et  al.   The  Economics of Multi-Plant Operation.   Cam-
     bridge, Mass.:   Harvard University Press, 1975.

 7.   Standard & Poor's Bond Guide.  . New York:  Standard & Poor's Corporation,
     September 1979.

 8.   Standard & Poor's Corporation Records.   New  York:    Standard  &  Poor's
     Corporation, September 1979.

 9.   Standard & Poor's Statistical  Service.   New  York:   .Standard  &  Poor's
     Corporation, October 1979.

10.   Value Line Investment Survey.    New  York:    Arnold  Bernhard & Co., Inc.,
     September 24, 1979.                 .   ,

11.   The Wall- Street Journal.   New  York:   Dow  Jones  &  Company,  October  1,
     1979.     •
                                    E-21

-------
APPENDIX F - SYNTHETIC ORGANIC CHEMICALS MANUFACTURING INDUSTRY
                               F-l

-------

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
OCPDB No.*
20
30
40
50
65
70
80
90
100
110
120
125
130
140
150
160
170
180
185
190
200
210
220
Chemical
Acetal
Acetaldehyde
Acetaldol
Acctamide
Acetanilide
Acetic acid
Acetic anhydride
Acetone
Acetone cyanohydrin
Acetonitrile
Acetophenone
Acetyl chloride
Acetylene
Acrolein
Acryl amide
Acrylic acid and esters
Acryl oni trile
Adipic acid
Adi poni trile
Alkyl naphthalenes
Ally! alcohol
Allyl chloride
Aminobenzoic acid
*The OCPDB Numbers are reference indices assigned to the various chemicals
 in the Organic Chemical  Producers Data Base developed by EPA.
                                  F-2

-------
            OCPDB No._         Chronical



24            230              Aiiiiiioethylethanolamine



25            235              p-nminophcnol



26            240              Amyl  acetates



27            250              Amyl  alcohols



28            260             .Amyl  amine



29            270              Amyl  chloride



30     •       280              Amyl  mercaptans



31            290              Amyl  phenol



32         -   300              Aniline



33            310              Aniline hydrochloride



34            320              Anisidine



35            330              Anisole



36            340              Anthranilic  acid



37            350              Anthraqin'none



38            360              Benzaldehyde



39            370              Benzamide



40          .  380              Benzene



41            390              Benzenedisulfonic  acid



42            400              Benzenesulfonic acid



43            410              Benzil



44            420              Benin lie acid



45            430              Benzoic acid



46            440              Benzoin



47            450              Benzonitrile



48            460              Benzophenone



49            4,30              Benzotrichloride
                               F-3

-------
                              Clif ;;n'cal



!>0           490              Bc-nzoyl  chloride



!>1           1300           '   Benzyl  alcohol



:?2           510              Benzyl  ann'ne



53           520              Benzyl  bcnzoate



54.           530              Benzyl  chloride



55           540              Benzyl  dichloride



56           550              Biphenyl



57           560              .Bisphenol  A



58           570              Bromobenzene



59           580              Bromonaphthalene"



60           590              Butadiene



61           592              1-butene



62           600              n-butyl  acetate



63           630              n-butyl  acrylate



64           640              n-butyl  alcohol



65           650              s-butyl  alcohol



66           660              t-butyl  alcohol



67           670              n-butylamine



68           680              s-butylamine



69           690              t-butylamine



70           700              p-tert-butyl benzoic acid



71           710              1,3-butylene glycol



72           750              n-butyraldehyde



73           760              Butyric acid



74           770              Butyric anhydride



75           780              Butyronitrile
                                .F-4

-------
          •'   OCI'DIi No.       .  Cin-iiiical



 76           785             Cnprolactam



 II           790             Carbon disulfide



 78           800             Carbon tetrabromide



 79           810             Carbon tetrachloride



 80           820             Cellulose acetate



 81           840             Chloroacetic acid



 82           850             m-chloroaniline



 83           860             o-chloroaniline



 84           870             p-chloroaniline



 85           880             Chlorobenzaldehyde



 86           890             Chlorobenzene



 87           900             Chlorobenzoic acid



 88           905             Chlorobenzotrichloride



 89           910             Chlorobenzoyl chloride



 90           920             Chlorodifluoroethane



 91           921             Chlorodifluoroniethane



 92           930             Chloroform



 93           940             Chloronapthalene



 94           950            .o-chloronitrobenzene



 95           951             p-chloronitrobenzene



 96           960             Chlorophenols



 97           964             Chloroprene



 98           965             Chlorosulfonic acid



 99           970             in-chlorotoluene



100           980             o-chlorotoluene



101           990             p-chlorotoluene
                                  F-5

-------
            OU'UB No.          Clir.-ini_cal_s



102           992             Chlorotrifluoroiiiethane




103          1000             m-crcsol



104          1010             o-cresol



105          1020             p-cresol



106          1021      .       Mixed cresols



107          1030             Crosylic  acid



108          1040         .    Crotonaldehyde



109          1050             Crotonic  acid



110        .1060             Cuinene



111          1070             Cuinene hydroperoxide



112          1080             Cyanoacetic acid



113          1090             Cyanogen chloride



114          1100             Cyanuric acid



115          1110             Cyanuric chloride



116          1120             Cyclohexane



117          1130             Cyclohexanol



118          1140             Cyclohexanone



119          1150             Cyclohexene



120          1160             Cyclohexylamine



121          1170             Cyclooctadiene



122          1180             .Decanol



123          1190             Diacetone  alcohol



124          1200             Diaminobenzoic acid



125          1210             Dichloroanillne



126          1215             m-dichlorobenzene



127          1216             o-dichlorobenzene





                                     F-6

-------

128
129
130
131
132.
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
(H.I'hB N'o.
1220
1221
1244
1240
1250
1270
1280
1290
1300
1 304
1305
1310
1320
1330
1340
1360
1420
1430
1440
1442
1444
1450
1460
1470
1480
1490
Chemical
p-dichlorobonzene
Dichlorodi fl uoromothano
1 ,2-dichloroothane (HOC)
Dichloroethyl ether
Dichlorohydrin
Dichloropropene
Dicyclohexylamine
Di ethyl ami ne
Diethylene glycol
Diethylene glycol diethyl ether
Diethylene glycol dimethyl ether
Diethylene glycol monobutyl ether
DiethyTene glycol monobutyl ether acetate
Diethylene glycol monoethyl ether .
Diethylene glycol monoethyl ether acetate
Diethylene glycol rnonomethyl ether
Diethyl sulfate
Difluoroethane
Diisobutylene
Diisodecyl phthalate
Diisooctyl phthalate
Diketene
Dimethyl ami ne
N,N -dimethyl aniline
N,N-dimethyl ether.
N ,N-dirnethyl formamide
F-7

-------

154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
1/3
174
175
176
177
178
OCPUI3 N_o_._
1495
1500
1510
1520
1530
1540
1545
.1550
1560
1570
1580
1590
1600
1610
1620
1630
1640
1650
1660
1661
1670
1680
1690
1700
1710
Chc-inical
Dimethyl hydrazine
Dimethyl sulfate
Dimethyl sulfide
Dimethyl sulfoxide
Dimethyl terephthalate
3, 5-di nitrobenzene acid
Dinitrophenol
Dinitrotoluene
Dioxane
Dioxolane
Diphenylamine
Diphenyl oxidq
Diphenyl thiourea
Dipropylene glycol
Dodecene
Dodecyl aniline
Dodecyl phenol
Epichlorohydrin
Ethanol
Ethanol amines
Ethyl acetate
Ethyl acetoacetate
Ethyl acrylate
Ethylamine
Ethylbenzene
F-8

-------
            UU'i)B No.         Chf.-iiiicajs	



 179           1720             Ethyl bromide



 180           1730             Ethylcellulose



 181           1740             Ethyl chloride



 182           1750             Ethyl chloroacetate



 183           1760             Ethyl cyanoaceta'te



 184           1770             Ethylene



 185           1780             Ethylene carbonate



 186           1790'         "   Ethylcne chlorohydrin



 187           1800             Ethylenediamine



 188           1810             Ethylene dibromide



 189           1830             Ethylene glycol



 190           1840             Ethylene glycol diacetate



 191           1870             Ethylene glycol dimethyl ether



 192           1890             Ethylene glycol monobutyl ether



 193           1900             Ethylene glycol monobutyl ether acetate



 194          1910             Ethylene glycol monoethyl ether



 195    .      1920             Ethylene glycol monoethyl ether acetate



 196          1930             Ethylene glycol monoinethyl ether



 197          1940             Ethylene glycol inonomethyl ether acetate



 198          1960             Ethylcne glycol monophenyl e.ther



 199          1970             Ethylene glycol monopropyl ether



200          1980             Ethylene oxide



201          1990             Ethyl ether



202          2000             2-ethylhexanol



203          2010             Ethyl orthoformate



204          2020             Ethyl oxalate
                                 F-9

-------

205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
Oi.i'ltf I!o.
20:10
2040
20'JO
2060
2070
2073
2090
2091
2100
2110
2120
2145
2150
2160
2165
2170
2180
2190
2200
2210
2240
2250
2260
2261
2270
2280
Ch.-.-inical
Ethyl sodium oxnlacetate
Formaldehyde
Form a mi da
Formic acid
Fumaric acid
Furfural
Glycerol (Synthetic)
Glycerol dichlorohydrin
Glycerol tri ether
Glycine
Glyoxal
Hexachlorobenzene
Hexachloroethane
Hexadecyl alcohol
Hexamethylenediami ne
Hcxame thy lone glycol
Hexamethylenetetramine
Hydrogen cyanide
Hydroquinane
p-hydroxybenzoic acid
Isoamylene
Isobutanol
Isobutyl acetate
Isobutylene
Isobutyraldehyde
. Isobutyric acid
F-10

-------

231
232
7.33
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
GU'ljB No.
2300
2320
2321
2330
2340
2350
2360
2370
2380
2390
2400
2410
2414
2417
2420
2430
2440
2450
2455
2460
2490
2500
2510
2520
2530.
2540
Chemical
.Isodecanol
Isooctyl alcohol
Isopentane
Isophorone
Isophthalic acid
Isoprcne
Isopropanol
Isopropyl acetate
Isopropylamine
Isopropyl chloride
Isopropyl phenol
Ketene
Linear alkyl sulfonate
Linear alkylbenzene
Maleic acid
Maleic anhydride
Malic acid
Mesityl oxide
Metanilic acid
I'ethacrylic acid
Methallyl chloride
Methanol
. Methyl acetate
Methyl acetoacetate
Methyl ami ne
n-methylani 1 ine
F-ll

-------
 257       2545              Methyl  bromide



 258       2550              Methyl  butynol



 2159       2560              Methyl  chloride



 260       2570              Methyl  cyclohexane



 261        2590              Methyl  cyclohoxanone



 262       2620              Methylene  chloride



 263       2530              Methylene  dianiline



 264       2635              Methylene  diphonyl diisocyanate



 265        2640              Methyl  ethyl  ketone



 266        2645              Methyl  formate



 267        2650              Methyl  isobutyl  carbinol



 268        2660              Methyl  isobutyl  ketone



 269        2665              Methyl  methacrylate



 270        2670              Methyl  pentynol



 271        2690              a-methylstyrene



 272        2700              Morpholine



 273        2710              a-naphthalene  sulfonic acid



 274        2720              g-naohthalene  sulfonic acid



 275        2730              a-naphthol



 276        2740              g-naphthol



277        2750              Neopentanoic acid



278   .     2756               o-nitroani1ine



279        2757               p-nitroani1ine



280        2760               o-nitroanisole



281        2762               p-nUroanisole



282        2770               Nitrobenzene





                                F-12

-------
       .  fKi'DB Np_._          Chemical.  __      	



283       2780              Nitrobenzpic acid  (o, r,  and  p).



284       2790     '         Nitroethane



285       2791              Nitromethane



286       2792              Nitrophenol



287       2795              Nitropropane



288       2800              Nitrotoluene



289       2810              Nonene



290       2820              Nonyl phenol



291       2830              Octyl phenol



292       2840              Paraldehyde



293       2850              Pentaerythritol



294       2851              n-penta.ne



295   .    2855              1-pentene



296       2860 '.            Perchloroethylene



297  .     2882              Perchloromethyl mercaptan



298       2890              o-phenetidine



299       2900   .           p-phenetidine



300       2910              Phenol



301       2920              Phenolsulfonic acids



302       2930      .        Phenyl anthranilic acid



303       2940      .        Phenylenediamine



304       2950              Phosgene



305       2960              Phthalic anhydride



306       2970              Phthalimide



307       2973              B-picoline



308       2976              Piperazine
                                F-13

-------
         nr.i'Mij ::o.           ci.-.ncai	
309        3000               Polybutenes
310        3010               Polyethylene glycol
311        3025               Polypropylene glycol
312        3063               Propionaldehyde
313        3066               Propionic acid
314        3070               n-propyl  alcohol
315        3075               Propylamine
316        3080  .             Propyl  chloride
317        3090               Propylene
318        3100               Propylene chlorohydrin
319        3110               Propylene dichloride
320        3111               Propylene glycol
321        3120               Propylene oxide
322.        3130               Pyridine
323        3140               Quinone
324        3150               Resorcinol
325        3160               Resorcylic acid
326        3170              'Salicylic acid
327        3180               Sodium acetate
328        3181               Sodium benzoate
329        3190               Sodium carboxymethyl  cellulose
330        3191               Sodium chloroacetate
331        32.00               Sodium formate
332        3210               Sodium phenate
333        3220              Sorbic acid
334        3230               Styrene
335        3240               Succinic  acid
                                F-14

-------

336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
OU'DB Mo.
3250
3251
3260
3270
3280
3290 & 3291
3300
3310
3320
3330
3335
3340
3341
3349
. 3350
3354
3355
3360
3370
3380
3381
3390,3391
& 3393
3395
3400
Chemical
Succinitrile
Sulfanilic acid
Sulfolane
Tannic .acid
Terephthalic acid
Tetrachloroe thanes
Tetrachlorophthalic anhydride
Tetraethyllead
Tetrahydronapthalene
Tetrahydrophthalic anhydride
Tetramethyllead
Tetramethyl enedi ami ne
Tetramethy 1 e thy! enedi ami ne
Toluene
Toluene-2,4-diamine
To! uene-2 ,4-di i socyanate
Toluene diisocyanates (mixture)
Toluene sulfonamide
Toluene sulfonic acids
Toluene sulfonyl chloride
To! ui dines
Trichlorobenzenes
1 ,1 ,1-trichloroethane
1 ,1 ,2-trichloroethane
F-15

-------

360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
3/7
3/8
CO PUB No.
3410
3411
3420
3430
3450
3460
3470
3480
3490
3500
3510
3520
3530
3540
3541
3560
3570
3580
3590
Chemical
Trichloroethylcne
Trichlorof luoroincthiine
1 ,2,3-trichloropropane
1 ,1 ,2-trichloro-l ,2,2-trifluoroethane
Tri ethyl ami ne
Triethylene glycol
Tri ethyl ene glycol dimethyl ether
Triisobutylene
Tr i methyl ami ne
Urea
Vinyl acetate
Vinyl chloride
Vinylidene chloride
Vinyl toluene
Xylenes (mixed)
o-xylene
p-xylene
Xylenol
Xylidine
F-16

-------