United States
 Environmental Protection
 Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/3-89-009
June 1990
 Air
Hazardous Waste
Treatment, Storage, and
Disposal Facilities --
Background Information
for Promulgated Organic
Emission Standards for
Process Vents and
Equipment Leaks

-------
                            EPA-450/3-89-009
  Hazardous Waste Treatment,
Storage, and Disposal Facilities
     Background Information
   for Promulgated Standards
    for Process Vents and
      Equipment Leaks
            Emissions Standards Division
          U.S. Environmental Protection Agency
              5, Library (P|_-12J)
         U.S. ENVIRONMENTAL PROTECTION AGENCY
            Office of Air and Radiation
          Offict of Air Quality Planning and Standards
          Research Triangle Park, North Carolina 27711
               June 1990

-------
This report has been reviewed by the Emission Standards 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 Office (MD-35), U.S. Environmental Protection Agency, Research Triangle Park NC 27711, or from
National Technical Information Services, 5285 Port Royal Road, Springfield VA 22161.

-------
                        ENVIRONMENTAL PROTECTION  AGENCY

                    Background  Information  for Promulgated
                        Organic Emission  Standards  for
                       Process  Vents  and  Equipment  Leaks
                                 Prepared by:
"Uacj/R.  Farmer
 Director,  Emission Standards Division
 U.S.  Environmental Protection Agency
 Research Triangle Park,  NC  27711

 1.    The standards would limit organic air emissions as a class at hazard-
      ous waste treatment,  storage,  and disposal  facilities (TSDF)  that are
      subject to regulation under Subtitle C of the Resource Conservation
      and Recovery Act (RCRA).  The rule establishes final standards limit-
      ing organic emissions from process vents associated with distillation,
      fractionation, thin-film evaporation, solvent extraction, air or steam
   .   stripping operations  that manage hazardous  wastes with at least 10
      ppmw total organics concentration, and leaks from equipment (e.g.,
      pumps and valves) that contain or contact hazardous waste streams with
      10 percent or more total organics.  The final standards are promul-
      gated under authority of Section 3004 of the Hazardous and Solid Waste
      Amendments to RCRA.  The EPA. is required by Section 3004(n) of RCRA to
      promulgate standards  for monitoring and control of air emission-s from
      hazardous waste TSDF  as necessary to protect human health and the
      environment.

 2.    Copies of this document have been sent to the following Federal
      Departments:  Labor,  Health and Human Services, Defense, Office of
      Management and Budget, Transportation, Agriculture, Commerce,
      Interior, and Energy; the National Science Foundation; and the Council
      on Environmental Quality.  Copies have also been sent to members of
      the State and Territorial Air Pollution Program Administrators; the
      Association of Local  Air Pollution Control  Officials; EPA Regional
      Administrators; and other interested parties.

 3.    For additional information on the regulatory aspects of these
      standards, contact:

      Mr. Rick Colyer
      Standards Development Branch (MD-13)
      U.S. Environmental Protection Agency
      Research Triangle Park, NC  27711
      Telephone:   (919) 541-5262.
                                      n i

-------
     For additional information on the technical aspects of these
     standards, contact:

     Mr. Robert Lucas
     Chemicals and Petroleum Branch (MD-13)
     U.S. Environmental Protection Agency
     Research Triangle Park, NC 27711
     Telephone:  (919) 541-0884

     For additional information on the test methods associated with these
     standards, contact:

     Mr. Terry Harrison
     Emission Measurement Branch (MO-13)
     U.S. Environmental Protection Agency
     Research Triangle Park, NC  27711
     Telephone:  (919) 541-5233

4.   Copies of this document may be obtained from:

     U.S. EPA Library (MD-35)
     Research Triangle Park, NC  27711

     National Technical Information Service
     5285 Port Royal Road
     Springfield, VA  22161
                                     IV

-------
                                  CONTENTS


Chapter                                                                Page

         Figures 	      ix
         Tables 	      ix

  1.0    Summary 	     1-1
         1.1  Summary of Changes Since Proposal 	     1-2
         1.2  Summary of Impacts of Promulgated Action 	     1-6

  2.0    Summary of Public Comments 	     2-1

  3.0    Regulatory Issues 	     3-1
         3.1  Selection of Source Category 	     3-1
              3.1.1  Land Disposal Restrictions 	     3-1
              3.1.2  Ozone 	     3-5
              3.1.3  Equity of Coverage 	     3-6
              3.1.4  Pollutants 	     3-7
              3.1.5  Equipment and Sources 	     3-9
         3.2  Statutory Authority and Regulatory Approach 	    3-21
             .3.2.1  RCRA vs. the CAA 	    3-21
              3.2.2  Level of Control Under RCRA	    3-24
              3.2.3  Consideration of Costs 	    3-27
              3.2.4  RCRA Section 3004(n)  	    3-28
              3.2.5  RCRA Authority to Control Reclaimed
                     Commercial Product 	    3-29
              3.2.6  Total Organics Approach  	    3-29
              3.2.7  Transfer of the Benzene NESHAP Technology 	    3-33
              3.2.8  Accelerated Approach 	    3-34

  4.0    Applicability and Exemptions 	.	     4-1
         4.1  Applicability of Standards for Storage Tanks
              and Accumulator Vessels 	     4-1
         4.2  Selection of the 10-Percent Cutoff 	     4-2
         4.3  Exemptions for Small Solvent Recovery Operations ....     4-5
         4.4  Totally Enclosed Treatment Facility Exemption 	     4-8
         4.5  Exemption for 90-Day Storage	    4-10
         4.6  Leak Detection and Repair (LDAR) Program 	    4-11
              4.6.1  Exemption for Light/Heavy Liquids 	    4-11
              4.6.2  Exemption for Small Facilities 	    4-13
              4.6.3  Exemption for Vacuum Systems 	    4-14
              4.6.4  Flanges and Pressure Relief Devices 	    4-15

-------
                               CONTENTS (con.)


Chapter                                                                Page

               4.6.5   Directed Maintenance, Leak Definition,
                       and Petition for Reconsideration of the
                       Benzene Standards 	    4-16
         4.7   References 	    4-19

  5.0    Environmental Impacts 	     5-1
         5.1   Data, Analyses, and Methodology 	     5-1
         5.2   Model Plant Emission Estimates 	     5-4
         5.3   Flow Rate 	     5-8
         5.4   Adjustment for Volatility	    5-10
         5.5   Number of Facilities 	    5-11
         5.6   Adequacy of Data and Information
               Supporting the Impact Analyses 	    5-12

  6.0    Health Risk Impacts  	     6-1
         6.1   Representativeness of Waste Streams 	     6-1
         6.2   Unit Risk Factors 	    6-13
         6.3   Methodology....	    6-17
         6.4   Additive and Synergistic Effects 	    6-20
         6.5   Dispersion Modeling	    6-22
         6.6   Worker Exposure	    6-23
         6.7   Level for Standards 	    6-24

  7.0    Selection of Standards 	     7-1
         7.1   Emission Cutoff for Process Vents 	     7-1
         7.2   Feasibility of a 95-Percent Control 	    7-10
         7.3   Feasibility of Condensers 	    7-12
         7.4   Feasibility of Flares	    7-15
         7.5   Feasibility of Carbon Adsorbers 	    7-16
         7.6   Feasibility of Leak Detection and Repair (LDAR)
               Program 	    7-21
         7.7   Feasibility of Using Controls in Series 	    7-25
         7.8   Reference 	    7-27

  8.0    Cost Impacts  	     8-1
         8.1   Carbon Adsorber Cost Estimates 	     8-1
         8.2   Condenser Costs	     8-3
         8.3   Compliance Costs	     8-5
         8.4   Compliance Costs for LDAR Program 	     8-6
         8.5   Recovery Credit for TSDF 	     8-9
         8.6   Cost Effectiveness of Process Vent Control
               Techniques 	-.	    8-10
         8.7   Cost of a Fixed 95-Percent Emission Reduction  	    8-14

  9.0    Economic Impacts	     9-1
         9.1   Small Business Impacts	     9-1
         9.2   Closures and Other Economic Impacts	     9-4
                                     VI

-------
                               CONTENTS (con.)
Chapter
Page
  10.0   Test Methods	    10-1
         10.1  Measurement of the 10-Percent Total Organics	    10-1
         10.2  Determination of Equipment in VHAP Service	    10-7
         10.3  Fugitive Emission Monitoring by Method 21	    10-8

  11.0   Implementation and Compliance Provisions 	    11-1
         11.1  Compliance Dates 	    11-1
               11.1.1  Overall Standards 	    11-1
               11.1.2  Period for LDAR 	    11-5
         11.2  Permits 	    11-7
               11.2.1  State and Local Role 	    11-7
               11.2.2  Part B Information Requirements  	    11-8
         11.3  Compliance Provisions  	    11-9
               11.3.1  Documentation  	    11-9
               11.3.2  Special Requirements 	    11-9
               11.3.3  Alternative Means of Emission Limitation  ...    11-10
         11.4  Recordkeeping and Reporting 	    11-11
               11.4.1  Frequency of Reports 	    11-11
               11.4.2  Duplication of Reports 	    11-12
               11.4.3  Notification	    11-13
         11.5  Implementation and Enforcement 	    11-14
               11.5.1  Guidance on the 95-Percent Limit 	    11-14
               11.5.2  Omnibus Permitting 	    11-15
               11.5.3  State and Regional Role 	    11-19
               11.5.4  LDAR Enforcement Approach  	    11-19
         11.6  Miscellaneous 	    11-20

Appendix

  A      Evolution of Promulgated Standards	     A-l

  B      Estimating Health Effects	     B-l

  C      Control Costs	     C-l

  D      Source Assessment Model	     D-l
                                     VI 1

-------
                                   FIGURES
Number                                                                 Page

  7-1   Emission rate cutoff vs. maximum individual risk
        (high emissions) 	     7-6
  7-2   Emission rate cutoff vs. incidence (high
        emissions)	     7-7
  7-3   Emission rate cutoff vs. emission reduction
        (high emissions) 	     7-8

  8-1   Emission rate vs. cost effectiveness (condensers
        @ 95%) 	    8-11
  8-2   Emission rate vs. cost effectiveness (carbon
        adsorption)		    8-12
  8-3   Emission rate vs. cost effectiveness (incinerators
        (3 98%) 	,....    8-13
                                   TABLES
Number                                                                 Page

  2-1   List of Commenters on Proposed Air Emission Standards
        for Hazardous Waste Treatment, Storage, and Disposal
        Facilities 	     2-2

  4-1   Air Stripper Data Summary 	     4-6

  5-1   Model Unit:  Process Vent Emissions 	     5-6

  6-1   Summary of WSTF/TSDF Waste Stream Constituents 	     6-3
  6-2   Information for TSDF Constituents	     6-5
  6-3   Vent Emissions Constituent Data Summary for
        Steam Stripper and Distillation Units	     6-8
  6-4   Vent Emissions Constituent Range and Frequency	    6-11

  7-1   Summary of Process Vent Risk Analysis  (Incidence) 	     7-4
  7-2   Summary of Process Vent Risk Analysis  (MIR)	     7-5

  8-1   Estimates of Mid-Sized Recycling Model Unit Compliance
        Costs 	     8-7

  9-1   Price and Quantity Adjustments in the Markets for Goods
        and Services by Generating Sector,  Equipment Leak
        Fugitive Emission Control	     9-3

                                     ix

-------
                                1.0  SUMMARY

     On February 5, 1987 (52 FR 3748), the U.S. Environmental Protection
Agency (EPA) proposed standards limiting emissions of volatile organic
pollutants from process vents and equipment leaks at new and existing
hazardous waste transfer, storage, and disposal facilities (TSDF) where
(a) equipment at the facilities contain hazardous wastes or derivatives of
hazardous wastes, and (b) these wastes or their derivatives contain
10 percent or more total organics.  These regulations were proposed under
the authority of Section 3004(n) of the Resource Conservation and Recovery
Act (RCRA) as amended.  The proposed standards would have required that
emissions from all process vents that emit organics in concentrations of
10 percent or greater on all TSDF waste management units, such as vents on
condensers, steam strippers, air strippers, thin-film evaporators,
distillation operations, and vents on product accumulator vessels, be
reduced by at least 95 percent.  The impacts of the proposed standards were
based on the use of secondary condensers, although the standards also would
have allowed the use of carbon adsorbers, flares, or other controls.  In
addition, facilities would have been required to monitor operation of the
control devices after installation.
     The proposed standards also would have required implementation of a
leak detection and repair (LDAR) program for valves and pumps and equipment
leak standards for compressors, pressure relief devices, sampling connec-
tion systems, open-ended valves or lines, pipeline flanges, and closed vent
systems and control devices used to comply with the standards that handle
hazardous waste and their derivatives at TSDF.  Control systems, leak
detection methodology, leak definitions, and repair schedules were based on
existing standards developed under Sections 111 and 112 of the Clean Air
Act (CAA).
                                     1-1

-------
     The process vents that would have been covered by the proposed
standards generally are associated with equipment (including product
accumulator vessels such as distillate receivers, surge control vessels,
product separators, or hot wells) used to distill, evaporate, or steam and
air strip organic components from hazardous waste.  Such separation tech-
niques can occur in solvent reclamation operations,  wastewater treatment,
or other pretreatment processes.  Therefore, the proposed standards would
have regulated the process of reclamation (and resulting air emissions) for
the first time.  However, only those facilities already subject to RCRA
permit requirements, including facilities that perform reclamation and
require a permit for another part of their operation, would have been
subject to the proposed standards.  Therefore, the proposed rule would have
added 40 CFR 261.6(d) so that reclamation activities at permitted facili-
ties would be subject to air emission standards.
     Public comments were requested on the proposal  in the Federal
Register.  A total of 60 comments were received, predominantly from
affected companies and trade associations.  Also commenting were other
Federal agencies and departments, various State air pollution control and
solid and hazardous waste management departments, one environmental group,
and one private citizen.  The comments that were submitted, along with
responses to these comments, are summarized in this document.  The summary
of comments and responses serves as the basis for the revisions made to the
standard between proposal and promulgation.
1.1  SUMMARY OF CHANGES SINCE PROPOSAL
     Since proposal, EPA has evaluated the comments received, collected and
assessed additional information, and made changes to the final rules as a
result of comments on the proposed rules.  First, the applicability of the
process vent standards has been limited to (1) process vents on distilla-
tion, fractionation, thin-film evaporation, solvent extraction, and air or
steam stripping operations that manage hazardous wastes with total organic
concentrations of at least 10 ppmw (annual average)  and vents on condensers
serving these operations, and (2) process vents on tanks (e.g., distillate
receivers, bottoms receivers, surge control tanks, separator tanks, and hot
wells) associated with distillation, fractionation,  thin-film evaporation,
solvent extraction, and air or steam stripping processes that manage
                                     1-2

-------
hazardous wastes with total organic concentrations of at least 10 ppmw
(annual average) if emissions from these process operations are vented
through the tanks.  In the final standards, uncondensed overhead from the
process unit is subject to air emission controls.  If emissions from the
process unit or a condenser serving the unit are not vented through these
tanks  (i.e., are vented directly to the atmosphere or through a vacuum
pump), then emissions from vent(s) that may be present on the tanks are not
subject to the process vent standards.  The applicability of today's final
standards for process vents has been limited because the control technology
assessments and the emissions and risk analyses needed to support extension
of the process vent standards to other closed (covered) and vented tanks
not associated with these processes have not yet been completed.  They are
being developed in conjunction with the Phase II Section 3004(n) air
emission standards now being developed for proposal in mid-1990.
     To avoid confusion with tanks that are not associated with the pro-
cessing of waste streams, and therefore are not subject to RCRA authority,
the term "product accumulator vessel" has been deleted from the final
standard and affected equipment are more specifically defined.  In addi-
tion, the terms "volatile hazardous air pollutant (VHAP)" and "in VHAP
service" have been deleted, and the wording of the final process vent
regulation has been revised to reflect applicability based on clearly
specified hazardous waste management processes or unit operations that have
vents emitting organics to the atmosphere.
    .The final rules for process vents add a facility-based emission rate
limit of 1.4 kg/h (3 Ib/h) and 2.8 Mg/yr (3.1 ton/yr)  representing the
total emissions from all affected process vents at the facility.
Facilities with organic emissions from affected vents  below these emission
rate limits will not have to install controls or monitor process vent
emissions.  Based on the final emissions and risk analyses,  limits of 1.4
kg/h (3 Ib/h)  and 2.8 Mg/yr (3.1 ton/yr) exempt from control  requirements
only emissions that do not pose significant risk.  Facilities that exceed
the limits must either reduce organic emissions from process  vents to below
the emission rate limits or reduce,  by use of a control  device,  total
organic emissions from process vents at the facility by 95 weight percent.
If enclosed combustion devices (i.e.,  incinerators,  boilers,  or process
heaters)  are used,  the owner/operator has the option of reducing the
                                     1-3

-------
organic concentration of each affected vent stream at the facility to no
more than 20 parts per million by volume (ppmv).   The final  regulations
place the burden of proof on the facility to demonstrate through test data
or engineering calculations that the facility would never be expected to
exceed the emission limits.
     Another major change affects the applicability of the final standards
for equipment LDAR.  The LDAR requirements for pumps and valves have been
revised to include the light and heavy liquid provisions contained in EPA's
new source performance standard (NSPS) for equipment leaks of volatile
organic compounds (VOCs) in the synthetic organic chemicals  manufacturing
industry (SOCMI) (40 CFR 60, Subpart VV).  Thus,  in response to comments,
the equipment standards take volatility into consideration.   Equipment are
in light liquid service if the vapor pressure of one or more of the
components is greater than 0.3 kPa at 20 °C, if the total concentration of
the pure components having a vapor pressure greater than 0.3 kPa at 20 °C
is equal to or greater than 20 percent by weight, and if the fluid is a
liquid at operating conditions.  Because equipment processing organic
liquids with vapor pressures below 0.3 kPa tend to leak at lower emission
rates and frequencies than do equipment processing streams with vapor
pressures above 0.3 kPa, the final rules exempt equipment processing lower
vapor pressure substances (i.e., heavy liquids)  from the routine LDAR
requirements of the standards unless there is evidence of a leak by visual,
audible, olfactory, or any other detection method.  As proposed, the LDAR
program would have been implemented at those facilities where equipment
contacts waste streams containing 10 percent by weight or more total
organics without regard to component volatility.  A facility may choose any
of the applicable test methods identified in the final rules for determin-
ing the organic content.
     Because of commenters1 concerns with the administrative burden (i.e.,
time and cost) involved in obtaining a major permit modification, the final
standards do not require modifications of permits issued before the effec-
tive date.  Hazardous waste management units and associated process vents
and equipment affected by these standards must be added or incorporated
into the permit at review under Section 270.50 or reissue under Section
124.15.
                                     1-4

-------
     Several changes have been made to the monitoring, reporting, and
recordkeeping requirements of the standards since proposal in response to
concerns with the compliance provisions.  Under RCRA statutory requirements
(Section 3010(b) of RCRA), compliance must be attained by the effective"
date of the rule (i.e., promulgation date plus 6 months).  However, a
facility that documents that an emission control device required by the
rules cannot be installed by the effective date can complete an implementa-
tion schedule.  The implementation schedule for the installation of
controls may allow up to 2 years after promulgation to complete installa-
tion, but the interim dates for design, commencement of construction, and
completion of construction have been deleted.  The implementation schedule
must be in the facility's operating record no later than 6 months following
promulgation of the final rules.  Additional, more detailed information on
implementation aspects is presented in Chapter 11.  As a clarification,
specific parameters that must be considered in the control device design
have been added to the standards for interim-status and permitted facili-
ties.  Similarly, the monitoring requirements for control device operating
parameters have been made more specific for interim-status facilities and
have been added to the requirements for permitted facilities.  The
semiannual reporting requirements for facilities with final permits have
been reduced to include only exceedances of the process vent and equipment-
leak provisions; there are no reporting requirements for interim-status
facilities.  For clarity, specific definitions of exceedances for process
vents in terms of the monitored operating parameters have been added.
Records of exceedances also must be kept in the facility operating log.
     The proposed air emission standards for process vents and equipment
leaks would have added Subpart C of Part 269, Air Emission Standards for
Owners and Operators of Hazardous Waste Treatment, Storage, and Disposal
Facilities.  For consistency with standards for other TSDF sources under
RCRA, the final standards have been incorporated into Parts 264,  for
permitted facilities,  and 265,  for interim-status facilities.  Subpart AA
applies to process vents and Subpart BB to equipment leaks.  In addition,
whereas at proposal  the equipment-leak requirements of 40 CFR'Part 61,
Subpart V, were incorporated by reference,  these provisions have been
written into Subpart BB with editorial revisions appropriate for a standard
promulgated under RCRA authority rather than CAA authority.  For example,
                                     1-5

-------
the terms "in VHAP service" and "product accumulator vessel" have been
deleted, and exceedances have been specifically defined such that only
exceedances must be reported on a regular basis, which is consistent with
RCRA enforcement procedures.
1.2  SUMMARY OF IMPACTS OF PROMULGATED ACTION
     The EPA revised and expanded the impact analyses, including estimates
of emissions, risks, costs, and the economic impact on small businesses and
on the industry as a whole.  Although the results of these analyses differ
from the impact estimates at proposal, EPA's conclusion that TSDF should be
regulated remains unchanged.  Based on the revised impact analyses, nation-
wide uncontrolled organic emissions from process vents at about 450 TSDF
with solvent recovery operations range from 300 Mg/yr or 330 ton/yr (based
on lower bound emission rates) to 8,100 Mg/yr or 8,900 ton/yr (based on
upper bound emission rates).  This wide emission range occurs because of
variations in primary condenser recovery efficiencies and the presence of
secondary condensers at some sites.
     A weighted-average unit risk factor of 4.5 x 10'6 cases//
-------
in Chapters 5.0 and 7.0, respectively, of this document.  Information on
the costs and economic analyses is presented in Chapters 8.0 and 9.0,
respectively.
     With the application of the final process vent standards, approxi-
mately 45 percent of the 450 TSDF with process vents will not require
additional control.  However, nationwide organic emissions from process
vents at the affected facilities will be reduced to a range of 260 Mg/yr or
290 ton/yr (lower bound emission rates) to 900 Mg/yr or 990 ton/yr (upper
bound emission rates).  Based on lower bound emission rates, the MIR will
be reduced to 2 x 10*6 from 3 x 10'5; annual incidence will be reduced to
O.OOL from 0.015 case/yr.  Based on the upper bound emission rates, the MIR
will be reduced to 4 x 10'5 from 8 x  10'4, and the annual incidence will be
reduced to 0.027 case/yr from 0.38 case/yr.
     With the implementation of LDAR  programs for fugitive emissions,
nationwide organic emissions will be-reduced to about 7,200 Mg/yr
(7,900 ton/yr).  The MIR and incidence associated with fugitive emissions
will be reduced to about 1 x 10~3 and 0.32 case/yr, respectively.  The EPA
believes that the emissions and risk  reductions achievable by these stand-
ards are needed to protect human health and the environment.
                                     1-7

-------
                        2.0  SUMMARY OF PUBLIC COMMENTS

     A total  of 60 letters commenting on the proposed standards and the
background information document (BID) for the proposed standards were
received.  A  list of commenters,  their affiliations,  and the EPA docket number
assigned to their correspondence is given in Table 2-1.
     For the  purpose of orderly presentation, the comments have been cate-
gorized under the the following topics:
     •    Chapter 3.0     Regulatory Issues
          Chapter 4.0
          Chapter 5.0
          Chapter 6.0
          Chapter 7.0
          Chapter 8.0
          Chapter 9.0
          Chapter 10.0
          Chapter 11.0
Applicability and Exemptions
Environmental Impacts
Health Risk Impacts
Selection of Standards
Cost Impacts
Economic Impacts
Test Methods
Implementation and Compliance Provisions.
                                      2-1

-------
                                           TABLE 2-1.  LIST OF COUUENTERS ON PROPOSED AIR EMISSION STANDARDS FOR
                                                HAZARDOUS WASTE TREATMENT, STORAGE,  AND DISPOSAL FACILITIES
         Docket  item number   Commenter name and address
                                                                          Docket  item number   Commenter name and address
         F-86-AESP-00001
         F-86-AESP-00002
        F-86-AESP-00003
tsj
l
ro
        F-ea-AESP-00004
F-86-AESP-00006
        F-86-AESP-00006
        F-B6-AESP-00007
Calvin L. Sonner
Route 2, Box  191A
Strasburg, VA 22667

David J. Carlson
Environmental Control Supervisor
McLaughlin Gormley King Company
8810 Tenth Avenue North
Minneapolis,  MN  66427

Allan A. Griggs, P.E.
Project Manager
Diamond Shamrock
  Refining and Marketing Company
P.O. Box 696000
San Antonio,  TX  78269-6000

Peter Schnieder
Vice President
Romlc Inc.
2081 Bay Road
East Palo Alto, CA  94303

George L. Will)amson
Compliance Manager
M1M Chemical Company
P.O. Box 291
Gadsden, AL  36902

Verrlll M. Norwood, Jr.
Vice President
Environmental Affairs
Olio Chemicals
P.O. Box 248
Lower River Road                '
Charleston, TN  37310

P.E. Gerwert, Manager
Industrial Waste and Toxic Substances
General Motors Corporation
General Motors Technical Center
30400 Mound Road
Warren, MI  48090-9016
                                                                          F-88-AESP-00008
                                                                                   F-ea-AESP-00009
                                                                                   F-80-AESP-00010
                                                                                   F-86-AESP-00011
                                                                                  F-aa-AESP-00012
                                                                                  F-80-AESP-00013
                                                                                  F-86-AESP-00014
Harry H. Hovey, Jr., P.E.
Director, Division of Air Resources
New York State Department of
  Environmental Conservation
60 Wo If-Road
Albany, NY  12233-3260

Robert Heitzer
Manager of Technical Services
Milwaukee Solvents & Chemicals Corporation
P.O. Box 444
Butler, WI  63007

R. R. Kienle, Manager
Environmental Affairs
Shell Oil Company
One Shell Plaza
P.O. 4320
Houston, TX  77210

B. F. Ballard, Director
Environmental Control
Phillips Petroleum Company
13 A4 Phillips Building
Bartlesvllle, OK  74004

Richard A. Lemen, Director
Division of Standards Development
  and Technology Transfer
Department of Health and Human Services
National Institute for Occupational
  Safety and Health
Robert A. Taft Laboratories
4678 Columbia Parkway
Cincinnati, OH  46226-1998

Carol J. Battershell
Corporate Environmental Specialist
The Standard OiI Company
200 Public Square
Cleveland, OH  44114-2376

John F Chadbourne, Ph.D
Director of Environmental Services
General Portland Inc.
12801 North Central Expressway
North Central Expressway Plaza HI
Suite 1700
P.O. Box 324
Dallas, TX  76221
                                                                                                                                         (continued)

-------
                                                          TABLE 2-1  (continued)
Docket Item number   Commenter name and address
                                                     Docket  Item number    Commenter  name end address
F-86-AESP-00016
F-86-AESP-00016
F-8B-AESP-00017
F-66-AESP-00018
F-86-AESP-00019
F-86-AESP-00020
F-86-AESP-00021
Gary D. Vest
Director for Environment,  Safety,  end
Occupational Health
Department of the Air Force
Washington, DC  20330-1000

James R. Hulm
President
National Asociation of Solvent Recyclers
1333 New Hampshire Avenue,  NW
Washington, DC  20036

U. L. Mull ins, Director
Regulatory Management
Monsanto Company
1101 17th Street, NW
Washington, DC  20036

Phillip J. Bateman
Corporate Director
Environmental Assurance
The Lubriiol Corporation
29400 take I and Boulevard
Wickliffe, OH  44092

Alan D. Wasserman
Honigman Miller Schwarti and  Cohn
2290 First National Building
Detroit, MI  48226

Lewis D. Walker
Deputy for Environment, Safety,  and
Occupational Health (OASA-UL)
Office of the Assistant Secretary
Department of the Army
Washington, DC  20310-0103

James D. Boyd
Executive Officer
State of California Air Resources  Board
1102 Q Street
P.O. Box 2816
Sacramento, CA  9S812
                                                                          F-B6-AESP-00022
F-86-AESP-00023
                                                                          F-88-AESP-00024
                                                                          F-86-AESP-00026
                                                                          F-86-AESP-00028
                                                                          F-88-AESP-00027
                                                                          F-88-AESP-00028
                                                                          F-86-AESP-00029
Cllot Cooper
Director of Environmental Affairs
Waste-Tech Services, Inc.
18400 W. 10th Avenue
Golden, CO  80401

Carlene Basse)I, P.E.
Manager, Environmental Technology
Lederle Laboratories
Pearl River, NY  10966

Alan T. Roy, P.E.
Manager, Pollution Control
Allied Fiber*
Allied Corporation
Chesterfield Plant
P.O. BOH 831
Hopewell, VA  23860

Kevin J. Igli
Regulatory Affairs Manager
Government Affairs Department
Waste Management, Inc.
1166 Connecticut Avenue, NW
Suite 800
Washington, DC  20036

Richard A. Svanda
Director
Solid And Hazardous Waste Division
Minnesota Pollution Control Agency
620 Lafayette Road North
St. Paul, UN  66166

Kirk Thomson, Manager
Environmental Affairs
The Boeing Company
P.O. Box 3707
Seattle, WA  98124-2207

Dr. J.F. Terenzi. Director
Engineering, Toxicology, and
  Environmental Services
Chemicals Group
American Cyanmid Company
One Cyanmid Plaza
Wayne, NJ  07470

G. Mahoney
Environmental Coordinator
Heritage Environmental Services,  Inc.
7901 West Morris Street
Indianapolis, IN  46231
                                                                                                                                 (continued)

-------
                                                                     TABLE 2-1  (continued)
           Docket item number   Co/nnenter name and address
                                                                                     Docket  item number   Comnenter name and address
           F-86-AESP-00030
           F-86-AESP-00031
ro
i
           F-86-AESP-00032
           F-86-AESP-00033
           F-86-AESP-00034
           F-86-AESP-00036
           F-86-AESP-00036
William F. O'Keefe
V?ce President
American Petroleum Institute
1220 L Street, NW
Washington, DC  20006

R. Stan Jorgensen, Director
Waste Management Program
Missouri Department of Natural  Resources
Division of Environmental  Quality
P.O. Box 176
Jefferson City, MO  66102

David M. Lusk, Vice President
Michigan Disposal, Inc.
P.O. Box 6116
Dearborn, MI  48128

Leo J. Domialski
Production Engineering Manager
Nalco Chemical Company
One Nalco Center
Naperville, IL  60666-1024

William Juris, P.E.
Engineering Section
Division of Air Pollution  Control
Ohio Environmental Protection Agency
P.O. BOM 1049
361 East Broad Street
Columbus, OH  43266-1049

J. C. Edwards
Manager, Clean Environment Program
Eastman Kodak
Eastman Chemicals Division
Kingsport, TN  37662

James Parvechio, Jr.
Vice President
Arivec Chemicals, Inc.
7962 Huey Road
P.O. Box 64
Douglasville, GA  30133
                                                                                     F-B6-AESP-L0001
                                                                                     F-86-AESP-L0002
                                                                                     F-86-AESP-L0003
                                                                                     F-86-AESP-L0004
                                                                                     F-86-AESP-L0006
                                                                                     F-86-AESP-L0006
                                                                                     F-86-AESP-L0007
                                                                                     F-86-AESP-L0008
Ceraldlne V. Cox
Chemical Manufacturers Association
2601 U Street NW
Washington. DC  20037

Sandra S. Newman
Manager of Environmental Affairs
Sterling Chemicals
P.O. Box 1311
Texas City, TX  77692-1311

Stanley A. Walczynskl
Treatment Operations
Safety-Kleen
777 Big Timber Road
Elgin, IL  60123

C.M. Dueiler
Commanding Officer
Naval Air Rework Facility
North Island
San Diego, CA  92136

Walter R. Quanstrom
General Manager (Mail Code 4902)
Environmental Affairs and Safety Department
Amoco Corporation
200 East Randolph Drive
Chicago, IL  60601

Thomas A. Robinson, Ph.D.
Director, Environmental Affairs
Vulcan Chemicals
P.O. Box 7689
Birmingham, AL  36263

C. P. Gorman, Director
Environmental Affairs Division
Eli Lilly and Company
Lilly Corporate Center
Indianapolis, IN  46286

W. Caffey Norman, III
Heron, Burchette, Ruckert 1 Rothwell
Suite 700
1026 Thomas Jefferson Street, NW
Washington, DC  20007
                                                                                                                                            (continued)

-------
                                                                      TABLE 2-1 (continued)
            Docket item number   Commenter name and address
                                                                                      Docket item number   Conmenter name end address
            F-86-AESP-L0009
            F-86-AESP-L0010
            F-86-AESP-L0011
ro
 i
tn
F-86-AESP-L0012
           F-86-AESP-L0013
           F-86-AESP-L0014
           F-80-AESP-L001E
 Charlene Esparza-Baca
 Environmental Engineer
 U.S.  Department of Energy
 Albuquerque Operations
 Los Alamos Area Office
 Los Alamos, NU  67644

 Colonel  Rockwood S. Dunham
 Staff Director
 Installation Services and
   Environmental Protection
 U.S.  Department of Defense
 Defense  Logistics Agency
 Cameron  Station
 Alexandria, VA  22304-0100

 R. B. Morris, Jr.
 Director, Environmental Programs
 IBM
 208-282  Harbor Drive
 P.O.  Bon  10601
 Stamford, CT  06904-2601

 Warren W. Tyler
 Ohio  Environmental Protection Agency
 P.O.  Box  1049
 361 E. Broad Street
 Columbus, OH  43268-1049

 Peter D. Venturing Chief
 Stationary Source Division
 California Air Resources Board
 1102 Q Street
 P.O. Box 2816
 Sacramento, CA  96812

 James R. Hulm
National Association of Solvent Recyclers
 1333 New Hampshire Avenue,  NW
Washington, DC  2003d

Raymond P. Berube
Acting Director
Office of Environmental  Guidance and Compliance
U.S Department of Energy
Washington, DC  20686
                                                                                      F-86-AESP-L0016
                                                                                      F-86-AESP-L0017*
                                                                                      F-88-AESP-L0018
F-88-AESP-L0019

F-B6-AESP-L0020b
                                                                                      F-86-AESP-L0021b
                                                                                      F-86-AESP-L0022b
                                                                                      F-88-AESP-L0023b
Deborah A.  SheIman
Resource Specialist
Natural Resources Defense  Council
1360 New York Avenue, NW
Washington, DC  20006

Kevin  J. Igli
Regulatory  Affairs Manager
Government  Affairs Department
Waste  Management, Inc.
1166 Connecticut Avenue, NW
Su i te  800
Washington, DC 20036

Kevin  J. Igli
Regulatory  Affairs Manager
Government  Affairs Department
Waste  Management, Inc.
1166 Connecticut Avenue, NW
Suite  800
Washington, DC  20036

OMB

James  R. Hulm
President
National Asaociation of  Solvent Recyclers
1333 New Hampshire Avenue,  NW
Washington, DC  20036

Alan D. Wasserman
Honigman Miller Schwartz and Cohn
2290 First National Building
Detroit, MI  48228

Kevin  J. Igli
Regulatory Affairs Manager
Government Affairs Department
Waste Management, Inc.
1166 Connecticut Avenue, NW
Suite 800
Washington, DC  20038

Robert Heitzer
Manager of Technical Services
Milwaukee Solvents & Chemicals Corporation
P.O.  Box 444
Butler, WI  63007
                     at end of table.
                                                                                                                                            (continued)

-------
                                                                      TABLE  2-1  (continued)
            Docket item number   Commenter name »nd address
                                                      Docket Item number   Commenter name end address.
            F-86-AESP-L00240
Richard A. Svanda
Director
Solid and Haiardou* Waste Division
Minnesota Pollution Control Agency
620 Lafayette Road North
St. Paul, UN  66166
Rahway, NJ  07066-0900
F-86-AESP-L0026
Dorothy P. Bowers
Executive Director
Environmental Resources
Merk A Co., Inc.
P.O. Box 2000
            •Request for comment period extension.
            ^Submission of additional information.
ro
l
crt

-------
                           3.0  REGULATORY ISSUES

3.1  SELECTION OF SOURCE CATEGORY
3.1.1  Land Disposal Restrictions
     Comment:  Five commenters questioned the presumption that the land
disposal restrictions (LDR) will cause an increase in the volume of solvent
for recycling as the basis for selecting hazardous waste TSDF (including
commercial recyclers) as a source category for regulation.  In support,
Commenters AESP-00004, AESP-00016, AESP-00024, AESP-00025, and AESP-L0003
explain the volume of solvent destined for recycling actually has decreased
because: (1) generators are recycling to the extent possible and solidify-
ing the rest; the resulting solids are incinerated if they cannot be land
disposed; (2) the solvents banned by the restrictions often are nonrecycl-
able and are used instead as supplemental fuel or are incinerated; (3)
generators are allowing the solvent waste (once sent to a landfill) to
remain mixed with the used solvent they do recycle, which reduces the
solvent yield and makes using the solvents as supplemental fuel  more
economically attractive; and (4) the restrictions affect only the clearly
residual wastes of low solvent content that are being shifted from land
disposal to incineration.  Commenter AESP-00025 also submitted a survey
showing that their wastes being recycled actually have decreased as a
result of the restrictions, whereas the waste volume shipped to industrial
furnaces and boilers has increased.
     Response:  It is important to point out that LDR was not the basis for
selecting hazardous waste TSDF (which includes commercial recyclers)  equip-
ment leak and process vent emissions as a source category for regulation.
The ambient air concentrations of organics emanating from these sources at
TSDF increase the individual's (general public's) probability or odds of
getting cancer,  and they add to the prevailing cancer incidence.  The EPA
has determined the public health risk of ambient air exposure to organics
                                     3-1

-------
emitted from TSDF to be significant.   Therefore,  the Administrator has
decided to take measures to reduce the atmospheric release of organic air
pollutants from these sources,  regardless of the LDR.  The LDR was a con-
sideration in EPA's decision to accelerate the regulations for equipment
leaks and process vents independent of other TSDF emission sources.  The
EPA pursued this phased approach to regulating TSDF emissions to fulfill
the Agency's desire to substantially reduce emissions from TSDF as soon as
possible and thus to protect human health and the environment from TSDF air
emissions.
     The EPA chose to develop the process vent and equipment leak portion
of its TSDF rulemaking as the first phase of the TSDF air emission rules,
partly to prevent uncontrolled air emissions from LDR treatment technolo-
gies.  The LDR developed under Section 3004(m) of the Hazardous and Solid
Waste Amendments (HSWA) require that hazardous waste be treated to reduce
concentrations of specific chemicals or hazardous properties to certain
performance levels or by certain methods before the waste may be disposed
in or on land.  The first set of land disposal restrictions, for certain
dioxins and solvent-containing hazardous wastes, was promulgated on
November 7, 1986 (51 FR 40572); the second set of restrictions, the
"California list," was promulgated on July 8, 1987 (52 FR 25760); the
"First Third" list was promulgated on August 17, 1988 (53 FR 31138); and
the  "Second Third" was promulgated on June 23, 1989  (54 FR 26597).
     Treatment technologies evaluated under the LDR  for nonwastewater spent
solvents include distillation/separation processes subject to the require-
ments of the proposed rule that are considered to be the best demonstrated
available technologies  (BOAT).   Thus, the LDR has the potential of
increasing emissions from waste treatment technologies such as steam
stripping.  The EPA is therefore setting standards to reduce emissions  from
the  technologies expected to be used prior to land disposal to protect
human health and the environment.
     To better evaluate current emissions and the potential impacts of  the
accelerated rule, EPA developed an industry profile  of onsite and offsite
waste solvent recyclers using the most comprehensive information currently
available to reevaluate the estimated organic emissions from recycling,  the
health risk, and EPA's  decision to regulate.  The survey results are sum-
marized below.
                                     3-2

-------
     The industry profile was developed using the results of the 1986
National Screening Survey of Hazardous Waste Treatment, Storage, Disposal,
and Recycling Facilities (the Screener Survey).  The EPA Office of Solid
Waste (OSW) conducted the Screener Survey under RCRA Section 3007 author-
ity.  The purpose of the survey was to identify and characterize hazardous
waste facilities to determine the size and composition of the sample for
the detailed survey of U.S. hazardous waste treatment, storage, disposal,
and recycling facilities that is currently being conducted.  The Screener
Survey data represent all of the TSDF active in 1985 with interim status or
final RCRA permits, which totalled about 3,000 facilities.  Although EPA
recognizes that many changes have occurred since 1985, this is the latest
year from which such comprehensive data on recycling currently are avail-
able.  A copy of the screener data is in the Docket No. F-90-AESF-FFFFF
(see item S0015).
         f
     A limited amount of information has also been received in response to
a Section 3007 questionnaire sent to a small number of selected sites (Doc-
ket No. F-90-AESF-FFFFF, items S0017, S0018, S0019; S0020, S0022, S0024
through S0026, S0028 through S0036, and S0038 through S0041).  Information
available from the Section 3007 responses includes the waste volume
recycled in 1985 reported by each facility,  recycling operation category
(solvent recovery, reuse as fuel, or other recycling), and waste category
recycled.  This information was used to corroborate Screener Survey results
to the extent possible.
     The Screener Survey results show that 847 TSDF reported, onsite hazard-
ous waste recycling.  Of the 847 TSDF recyclers, 448 facilities reported
solvent recovery operations by batch distillation, fractionation, or steam
stripping,  which are processes with vents subject to the promulgated rules.
The hazardous waste recovered in batch distillation, fractionation,  or
steam stripping operations consisted predominantly of solvents, other halo-
genated organics, and other hazardous wastes.  A total of 406 of the 448
facilities handled waste solvents, including 280 facilities reporting
recovery operations handling only solvents and 126 facilities reporting
solvent recovery operations handling solvents as well as other halogenated
organics and other hazardous waste.  The total volume of hazardous waste
treated by batch distillation,  thin-film evaporation, fractionation,  and
steam stripping was about 1.7 billion L (450 million gal), based on  data
from 365 of the 448 facilities reporting on  solvent recovery operations.
                                     3-3

-------
     About one-third of the recycling facilities (273 of 847) indicated
some reuse of hazardous waste as fuel.  Of these 273 facilities, 108 of the
facilities indicating solvent reuse as fuel reported no solvent recovery
operation of any type onsite.  Therefore, transfer and handling operations
at these 108 facilities as well as at the 448 facilities reporting solvent
recovery by batch distillation, fractionation, thin-film evaporation, or
steam stripping are expected to be subject to the equipment leak provisions
of the promulgated rules.  About 503 million L (133 million gal) of waste,
including solvents, were reported by the 94 facilities not having onsite
solvent recovery operations that reported a 1985 recycled quantity as being
reused as fuel.
     Consequently, the Screener Survey shows a minimum total volume of
about 2.2 billion L  (583 million gal) of hazardous waste recycled in 1985,
including 1.7 billion L  (450 million gal) of waste being treated by solvent
reclamation and 503 million L  (133 million gal) being reused as fuel.  Not
included in this estimate is the volume of solvent being incinerated or
treated by "other recycling" (processes other than batch distillation,
fractionation, or steam  stripping).  The EPA considers the 2.2  billion L
(583 million gal) a minimum estimate because only about two-thirds of the
facilities indicating the presence of recycling operations reported waste
volumes.  If waste volumes are  projected for these facilities based on per-
facility reported volumes, the  total waste volume treated by solvent recla-
mation increases to  an estimated 2.1 billion L  (550 million  gal) and the
amount reused as fuel  increases to 617 million  L -(163 million gal),  for a
total volume of 2.7  billion  L  (713 million gal).  This estimate still ex-
cludes the waste volume  being  incinerated  or treated by  "other  recycling."
     Although EPA  acknowledges  that  the  overall waste volumes may have
decreased since  1985 due to waste minimization  techniques at certain major
facilities,  the waste  volume being recycled,  based on the data  cited above,
still results  in  a  significant  level  of  emissions and health risk that
warrants  regulation  for  protection of human  health and  the  environment.
Based on  the revised process vent  emission  factors and  model units  (see
Chapter  5.0), nationwide organic  emissions  from process  vents  at  affected
recycling  facilities are estimated to range  between  300  Mg/yr  (330  ton/yr)
and  8,100 Mg/yr  (8,900  ton/yr).   The range in  nationwide organic  emission
estimates  reflects  the  range of emission factors  developed  for process
                                      3-4

-------
vents, representing extreme lower and TSDF process vent on process vent
emission rates.  Based on the TSDF process vent unit risk factor, estimated
cancer cases occurring each year as a result of exposure to process vent
emissions may be as high as 0.38 case/yr, and the MIR of developing cancer
is about 8 x 10~4.  The health risks posed by these sources are considered
significant; therefore, the Administrator has decided to take measures to
reduce the atmospheric release of organic air pollutants from these
sources.  Additional information on the emission and risk estimates is
presented in Chapters 5.0 and 6.0, respectively, of this document.
3.1.2  Ozone
     Comment:  Whereas Commenter AESP-L0016 supports efforts to attain the
ambient ozone standard, other commenters question the role of ozone in
EPA's selection of TSDF as a source category for regulation.  Commenters
AESP-L0001, AESP-00035, AESP-L0007, and AESP-L0008, state that, on a na-
tional and global basis, the chemicals in the solvents that influence ozone
depletion yield emissions that represent only a small fraction of the
worldwide contribution of these compounds.  They also state that ozone is
under the purview of the CAA, not RCRA.  Commenter AESP-L0025 also states
that EPA has presented no evidence that the level of protection necessary
for benzene, a carcinogen, is needed for ozone precursors.
     Response:  The EPA agrees with the commenters in that ozone is a na-
tional and global problem.  In this regard, EPA is a major supporter of the
recent international agreement to limit the manufacture of chlorofluorocar-
bons to reduce ambient ozone formation and stratospheric ozone depletion.
However, EPA does not agree that the impact of the standards on reducing
ozone formation is minimal.  As discussed in the preamble to the proposed
rules at 52 FR 3752, organic emissions contribute to ambient ozone forma-
tion in the same way that VOCs contribute to this problem.  In fact,  TSDF
sources as a whole are estimated at nearly 12 percent of all VOC emissions
from stationary sources, and any reductions in these emissions also are
helpful in reducing ozone formation and associated health problems.
     Although VOC emissions have been regulated under the CAA, this does
not mean that organic emissions from TSDF may not be regulated under RCRA.
Section 3004(n) standards, like all RCRA Subtitle C standards, are to pro-
tect "human health and the environment."  VOC and ozone are threats to
human health and the environment and thus are well within the regulatory
                                     3-5

-------
scope of Section 3004(n).  In addition, organic emissions may contain a
variety of toxics, such as carbon tetrachloride,  acrylonitrile, and methy-
lene chloride.  The control of ozone precursors and the reduction of the
air toxics potential of organic emissions are important goals of these
standards.
3.1.3  Equity of Coverage
     Comment:  Ten comments were received that criticized the source cate-
gory for regulation in that generators, onsite solvent reclaimers, and
offsite reclaimers with no prior storage are excluded from control (AESP-
00002, AESP-00026, AESP-00007, AESP-00019, AESP-L0003, AESP-L0002, AESP-
00027, AESP-00029, AESP-00033, and AESP-00031).  Several commenters contend
that manufacturing processes emit the vast majority of emissions and should
be the focus of regulatory concern; they also dispute EPA's assumption that
there would be few offsite reclaimers with no prior storage or onsite rec-
lamation operations.  Commenter AESP-00019 points out that there must be
hundreds or even thousands of facilities that recycle offsite without prior
storage based on the sales by distillation unit vendors.  Commenter AESP-
00026 indicates knowledge of one offsite- recycler without storage and one
that emits volatile organometallics.  Commenter AESP-00031 states that more
than 50 facilities in his State that recycle onsite would not be covered.
Commenter AESP-L0016 states that EPA has not provided adequate justifica-
tion for limiting the coverage of'the proposed rule.  This commenter does
not find EPA's rationale compelling because:
     •    The EPA has not shown that the need to protect human health
          and the environment is any less a concern at the facilities
          that are not required to have a permit than at the facil-
          ities that are.
     •    The fact that CAA standards may apply to some of the facil-
          ities that would not be covered does not address the concern
          that RCRA standards are required to be stronger than what
          EPA proposed; if stronger standards had been developed, they
          would have applied to all facilities in order to achieve the
          same level of protection at all facilities.
     •    The practical problems of permitting these facilities are
          not a sufficient reason for such exemption; had Congress
          intended an exemption for these facilities not already need-
          ing a permit, it would have done so.
                                     3-6

-------
     •    The EPA  should obtain the data  it needs on the  implications
          of the rule on small quantity generators and small busi-
          nesses;  the commenter fails to  see how the need for obtain-
          ing data for certain reclamation processes justifies
          creating an exemption for all facilities that are not cur-
          rently subject to permit requirements.
     Response:  The facilities of concern to the commenters are those that
do not require a RCRA permit to operate.  The EPA is promulgating these
standards under the authority of Section  3004(n).  Therefore, the standards
only apply to facilities that need authorization under Section 3005 of RCRA
to operate; i.e.,  the permitted community is not being expanded to
incorporate the air emission regulations.  However, facilities that do not
need a RCRA permit have not been "exempted."  They are simply not covered
under this rulemaking.  The EPA will continue to assess the need to develop
air emission standards under the CAA for  these facilities.
     In considering the regulation of air emissions under RCRA Section
3004(n) and within the RCRA regulatory framework, EPA has concluded that
air emissions from hazardous waste management units that are subject to
RCRA Subtitle C should be regulated under the authority of RCRA Section
3004(n).  Air emissions from hazardous wastes from facilities or units that
manage solid wastes that are not regulated as hazardous wastes pursuant to
40 CFR 261 (e.g.,  cement kiln dust waste) and air emissions from hazardous
waste from units or facilities that are exempt from the permitting provi-
sions of 40 CFR 270.1(c)(2) (e.g.,  wastewater treatment units with National
Pollutant Discharge Elimination System permits) will be subject to control
technique guidelines or standards developed as needed under either the CAA
or RCRA authority.   Air emissions from wastes managed in units subject to
Subtitle D (nonhazardous solid wastes,  such as those managed in municipal
landfills) also will  be subject to guidance or standards issued under CAA
or RCRA authority,  as appropriate.
3.1.4  Pollutants
     Comment:   Commenter AESP-L0018 questions why the proposed standards do
not cover other pollutants, such as emissions from volatile metals.
     Response:   The standards do cover emissions of organometallics  if they
are contained  in the waste stream and emitted from affected vents  on dis-
tillation units,  fractionation units,  thin-film evaporators,  solvent
extraction units,   steam and air strippers, or from equipment leaks.
However,  emissions  of volatile metals can occur in other types of  processes

                                     3-7

-------
not covered by these rules.  The additional  TSDF air standards currently
under development will  apply to organic emissions from open and closed
(covered) hazardous waste management units and are scheduled for proposal
in late 1989 or early 1990.
     Air standards also have been developed for the control of organic air
emissions from permitted hazardous waste incinerators (40 CFR 264, Subpart
0).  These standards require that incinerators be operated to achieve a
destruction and removal efficiency (ORE) of at least 99.99 percent for
those primary organic hazardous constituents listed in the facility permit.
Higher efficiencies are required when the incinerator is burning certain
specified waste types.   These standards also limited air emissions of or-
ganics, hydrochloric acid,, and particulates from incinerator stacks.  Air
standards for interim status hazardous waste incinerators  (40 CFR 265,
Subpart 0) require monitoring of visible emissions and operating condi-
tions.  When burning specified wastes, these incinerators must receive a
certification from the Assistant Administrator stating that the incinerator
can meet the performance standards specified for permitted incinerators  in
40 CFR 264, Subpart 0.   Interim status standards to control air emissions
for other thermal treatment units are found in 40 CFR 265, Subpart P.
These standards apply to facilities that thermally treat hazardous waste in
devices other than enclosed devices using controlled flame combustion.   The
standards require monitoring of visible emissions and operating conditions
of the combustion devices  and prohibit open burning except for open burning
and detonati-on of waste explosives.
     The Agency has also proposed standards covering the burning of
hazardous waste in boilers and  industrial furnaces  (52 FR  16987, May  6,
1987).  These standards would require such burning to achieve a ORE of
99.99 percent for each principal organic hazardous constituent identified
in the facility permit.  In addition, a ORE of 99.9999 percent must be
achieved when burning  certain specified constituents.  The proposed
standards also have provisions  for burning low-risk wastes that allow an
owner or operator to demonstrate that the burning of hazardous waste  will
not result  in significant  adverse health effects.   To qualify for  the low-
risk waste  exemption,  an owner  or operator would  have to  use  emission
modeling to demonstrate that emissions  of carcinogenic compounds  would  not
result in offsite ground-level  concentrations  that  pose  a  risk to  the most
                                      3-8

-------
exposed  individual of greater than  1 x  10'5.   For noncarcinogenic
compounds, the emission modeling would  demonstrate that  the  resulting air
concentrations would not exceed the reference  air concentration  (RAC) of
individual hazardous compounds.  The standards proposed  also would  limit
emissions of carbon monoxide, metals, and hydrochloric acid  from boilers
and furnaces burning hazardous wastes.
     Volatile metals also can be emitted in the form of  particulates.
Standards that regulate particulate emissions  from TSDF  sources have been
promulg-ated in 40 CFR 264 (40 CFR 264.251(f),  264.301(1), and 264.273(f)).
The general standards require the implementation of design and operating
practices at RCRA-permitted wastepiles, landfills, and land treatment
operations to limit the release of particulate air emissions.  As a
separate effort, EPA has prepared a technical  guidance document to  aid in
implementation of these standards.  The document ("Hazardous Waste  TSDF-
Fugitive Particulate Matter Air Emissions Guidance Document," EPA-450/3-89-
019) includes guidance on how to estimate particulate emissions from TSDF
sources, how to assess the health risk associated with these emissions, and
what controls are available to reduce particulate emissions and their
health risks.
     Additionally, 40 CFR 264 Subpart X contains provisions that require
prevention of air releases that may have adverse effects on human health or
the environment at miscellaneous hazardous waste management units.
3.1.5  Equipment and Sources
     3.1.5.1  Definition of "Process Vent."  Comment:  Several commenters
asked for clarification of the definition of "process vent" as follows:
(1) to state whether a control device is needed on a condenser vent because
the vent is in VHAP service if the condenser vent stream is greater than
10-percent total  organic carbon (TOC)  (the commenter does not believe
control of condenser vents is intended and suggests drafting the definition
and standard in terms of requiring a 95-percent reduction of uncontrolled
emissions from specific pieces of equipment [AESP-00024]);  (2) to clarify
that discharges to the atmosphere from unenclosed (uncovered) process
vessels such as product accumulator vessels are excluded (AESP-00030);  and
(3) to clearly exclude storage tanks or fixed-roof tanks, blending tanks,
and transfer facilities (AESP-00003  and AESP-00006).
                                     3-9

-------
     Response:  First,  it should be noted that for clarity the term "in
VHAP service" has been  dropped from the promulgated regulations and the
applicability has been  more specifically defined.   These regulations apply
to process vents associated with distillation, fractionation,  thin-film
evaporation, solvent extraction, or air or steam stripping operations that
manage hazardous wastes with total organic concentrations of at least
10 ppmw.  Therefore, an exhaust gas or vent stream from a condenser serving
a distillation, fractionation, thin-film evaporation, solvent extraction,
or stripping operation  that manages hazardous wastes with organic concen-
trations of at least 10 ppmw is subject to the requirements of these
regulations.  Total facility process vent emissions must be controlled to
less than 1.4 kg/h  (3 Ib/h) and 2.8 Mg/yr (3.1 ton/yr) or the owner/oper-
ator must install a control device(s) that reduces total organic emissions
from all affected vents at the facility by 95 weight percent or, for
enclosed combustion devices, to a total organic compound concentration of
20 ppmv.  For the purpose of determining the 95 percent emission reduction
performance requirement, primary condensers  (i.e., those condensers whose
primary function is the recovery of product  for commercial recycling, sale,
or reuse) are considered as process equipment and are not considered a
control device  in the performance calculations.
     The final  standard for process vents has been clarified since  proposal
to exclude  air  emissions from vents on other closed  (covered)  and vented
tanks not associated with the specified distillation/separation processes.
Thus, the final  process vent  standards apply to (1)  process vents on dis-
tillation,  fractionation, thin-film evaporation,  solvent extraction, and
air or  steam  stripping operations and  vents  on condensers serving these
operations,  and (2) process vents on tanks or vessels  (e.g., distillate
receivers,  bottoms  receivers, surge control  tanks, separator tanks,  and  hot
wells associated with distillation, fractionation, thin-film evaporation,
solvent extraction, and  air or  steam stripping processes)  if emissions from
these operations are vented through the  tank.  For example, uncondensed
overhead  emitted from  a  distillate  receiver  (i.e., a tank)  serving  a haz-   •
ardous  waste distillation  process  unit is  subject to air  controls  under
these regulations.  On  the  other  hand,  if  emissions  from  the distillation
unit  or a condenser serving the unit are not vented  through the tank (i.e.,
they  are  vented directly  to the atmosphere or through a vacuum pump),  then
                                     3-10

-------
vent(s) that may be present on the tank are not subject to the final
standard for process vents.  Unenclosed or uncovered processes, storage
tanks, fixed-roof tanks,  and transfer facilities are not covered by these
regulations, but they will  be covered by the additional TSDF air emission
standards scheduled for proposal in mid-1990.
     3.1.5.2  Definition  of "Facility."  Comment:  Commenter AESP-00007
disagrees with two of the examples provided in the preamble (52 FR 3754)
regarding the types of facilities to be regulated.  He does not agree with
example 2b that would automatically apply the standards because an interim
status or Part B permit is  required for solvent storage prior to distilla-
tion and because this application does not recognize the difference between
small onsite reclamation  operations and large offsite operations.
     Commenter AESP-00007 also disagrees with example 2c based on EPA's
definition of "facility."  The commenter does not believe that distillation
operations should be subject to the standards merely because of the exis-
tence of a hazardous waste  management unit requiring a RCRA permit.  If the
organic emission requirements are to be applied to onsite recycling,
according to the commenter, they should apply only if the unit on the same
site requires a permit and  has as its intended purpose the treatment, stor-
age, or disposal (TSD) of the hazardous solvents that are processed in the
distillation operation.  In support, the commenter cites EPA's definition
of "facility" under 40 CFR  260.10 that excludes portions of TSDF property
not used for hazardous waste TSD from RCRA and corrective action require-
ments.
     Commenters AESP-00015  and AESP-L0010 also recommend that EPA modify
the terminology of the proposed standards so as not to inadvertently apply
the Agency's current interpretation of the term "facility"; if the termi-
nology is not modified, then industrial operations such as paint spray
booths or organic emissions from all materials, tanks, and waste handling
equipment on the entire site (in these cases, military bases)  would be
subject to the standards.
     Response:  In response to Commenter AESP-00007, in promulgating these
regulations under Section 3004(n),  it is clearly within EPA's legal author^
ity to issue air emission standards that apply to any unit that treats,  ,
stores, or disposes of hazardous wastes, regardless of whether the facility
contains other units that also treat, store,  or dispose of hazardous
                                    3-11

-------
wastes.  The EPA has determined that,  as a matter of regulatory policy,  it

would be best to address under the CAA those units that are located at

facilities that do not otherwise need a RCRA permit.  However,  this policy

decision does not preclude EPA from regulating under RCRA any TSD unit

located at a facility containing another TSD unit that requires a permit.

In this context, the term "facility" retains its traditional definition set

forth in 40 CFR 260.10.

     In the preamble to the hazardous waste management system,  final codif-

ication rule (50 FR 28702, July 15, 1985), EPA elaborated on the facility

definition.  The preamble states that:

     [The] EPA must assume that in" using the term "facility," Congress
     intended,  in the absence of contrary statutory language or legis-
     lative history, to adopt the definition of this term traditional-
     ly employed by the Agency.  The preamble to the July 26, 1982,
     regulation elaborates on the definition of this term.  Speci-
     fically, the preamble notes that "[w]hen using the term 'facil-
     ity,' EPA is referring to the broadest extent of EPA's area
     jurisdiction under Section 3004 of RCRA... [meaning] the entire
     site that is under the control of the owner or operator engaged
     in hazardous waste management," 47 FR 32288-9 (July 26, 1982).
     The legislative history of the conference bill makes it clear
     that Congress was aware of the Agency's definition.  In discuss-
     ing new Section 3004(v) (see subsection e, infra), the Congress
     noted EPA's position limiting the scope of its remedial author-
     ities to the property of the polluting facility.  130 Cong. Rec.
     H11129  (daily ed. Oct. 3, 1984).  Accordingly, for purposes of
     Section 3004(a), the term "facility" is not limited to those
     portions of the owner's property at which units for the manage-
     ment of solid or hazardous waste are located, but rather extends
     to all contiguous property under the owner or operator's control.

This facility definition was upheld by the United States Court of Appeals
for the District of Columbia Circuit  (see United Technologies Corp. v. EPA,

F. 2d  [D.C. Cir. 1987]).  Under this definition, a solvent reclaim

operation and surface impoundment unit, for example, may often be con-

sidered to be part of a single facility.  If they are part of the same

facility, then the reclaim operation  is subject to the Section 3004(n)

regulations  (provided that all other  regulatory prerequisites are met).

     With regard to the concerns of Commenters AESP-00015 and AESP-L0010,

the Subparts AA and BB regulations apply to process vents and equipment,

respectively, associated with hazardous waste management.  Subpart  BB air

emission standards apply to owners and operators of equipment if it

contains or contacts hazardous waste with organic concentrations at least


                                    3-12

-------
10 percent by weight.  A process vent is subject to Subpart AA requirements
if it is associated with distillation, fractionation, thin-film evapora-
tion, or solvent extraction, or with air or steam stripping operations that
manage hazardous wastes with organic concentrations of at least 10 ppmw.
Therefore, organic emissions from vents from industrial operations not
associated with hazardous waste management, such as paint spray booths or
product distillation columns, are not subject to these rules.  However,
organic emissions from affected process vents anywhere on the entire
facility (provided the facility requires a permit under 40 CFR 270 due to
other hazardous waste activity at the facility) are subject to the
standards.  Similarly, organic emissions from affected equipment anywhere
on the entire facility (provided a RCRA permit is required under 40 CFR 270
due to other hazardous waste activity at the facility) are subject to the
standards.
     3.1.5.3  Fugitive Sources.  Comment:'  Commenter AESP-L0015 asks if
pumps or valves for organic liquids at a hazardous waste incinerator are
subject to the equipment leak standards, and if so, is it according to the
40 CFR 60 or the 40 CFR 61 requirements.  Commenter AESP-00024 points out
that all components such as pumps, flanges, and valves associated with
hazardous waste tanks are required to be inspected on a daily basis under
40 CFR 264.195.  If a visual leak is detected, the component must be taken
out of service.  The commenter believes this is more than adequate for
equipment in heavy-liquid service and should be the only standards required
for fugitive emission control.
     Response:  In response to Commenter AESP-L0015, pumps and .valves con-
taining or contacting hazardous wastes with at least 10 percent organic
concentrations at a hazardous waste incinerator are subject to the stand-
ards.  Specifically, the applicable requirements include Section 264.1052
and Section 264.1057 for pumps in light-liquid service and valves in
gas/vapor service or in light-liquid service, and Section 264.1058 for
pumps and valves in heavy-liquid service,  pressure relief devices in light-
liquid or heavy-liquid service, and flanges and other connectors.
     The visual inspections required for pumps, valves, and flanges refer-
red to by Commenter AESP-00024 are associated with ground-water protection
rules and by themselves are not appropriate or adequate for monitoring air
emissions.  Whereas an operator can make a visual inspection to ensure that
                                    3-13

-------
no liquids are dripping onto the ground,  air emissions cannot be detected
as easily.  In addition, the commenter states that these inspection rules
are for equipment in heavy-liquid service.  However,  most air emissions
from equipment leaks are from equipment in light-liquid service.  The re-
sponse to comment 11.4.2 contains suggestions as to potential ways opera-
tors may be able to combine monitoring efforts to reduce duplication of
labor and recordkeeping/reporting provisions.
     3.1.5.4  Definition of "Product Accumulator Vessel" (PAV).   Numerous
comments were made on the definition of "product accumulator vessel" that
requested clarification of the intended scope of equipment covered under
the definition.  The promulgated standards cover process vents and equip-.
ment associated with the handling and management of hazardous waste.
     Comment:  The EPA should confirm that coverage excludes open tanks and
is limited to PAVs used for treating waste solvents that are currently en-
closed and vented by means of pipes or stacks so as not to be interpreted
that all PAVs must be equipped with a closed-vent system that captures and
transports emissions to a 95-percent control device (AESP-L0005, AESP-
00030, and AESP-00017).
     Response:  With respect to the comments concerning open tanks, the
proposal preamble states at 52 FR 3753 that "the standards apply only to
the specifically enumerated types of equipment.  Devices such as...open
tanks would not be 'equipment' for purposes of today's rule and  so would
not be covered by today's proposal."  The same is true under the final
rules.
     Comment:  Are recovered material storage tanks,  hazardous waste stor-
age tanks, or other types of storage tanks covered by the rules  (AESP-
L0006, AESP-L0007, AESP-L0011, AESP-00033, AESP-00006, and AESP-00014)?
Also, Commenter AESP-00024 asks if all hazardous waste accumulation tanks
are covered by the rules if they are associated with a VHAP hazardous waste
and the facility has an interim status or Part B (final) permit  for any
reason.  Are a tank and associated equipment that vent to the atmosphere
during loading or due to breathing losses covered since this is  not venting
by mechanical or process-related means (AESP-00022)?
     Response:  In the promulgated rule,  the applicability of the process
vent provisfons has been more precisely defined in terms of the  types of
emissions and emission sources covered.  A tank vent is covered  only if
                                    3-14

-------
 process emissions  from distillation,  fractionation,  thin-film  evaporation,
 solvent extraction, or steam or  air  stripping  operations managing  hazardous
 wastes with organic concentrations of at  least 10  ppmw  on  an annual  average
 basis, are vented  through the tank (e.g.,  distillate receiver,  condenser,
 bottoms receiver,  surge control  tank,  separator tank, or hot well) associ-
 ated with one of these operations.   Vented means that the  passage  of
 liquids, gases, or fumes is caused by mechanical means  such as  compressors
 or vacuum-producing systems or by process-related  means such as evaporation
 produced by heating and not caused by tank loading and  unloading  (working
 losses) or by natural means such as  diurnal temperature changes.   A  tank
 vent that discharges process emissions as  defined  above is subject to the
 process vent final rules.
     Regarding storage tanks, the passage  of liquids, gases, or fumes from
 the tank vent must be caused by mechanical means or  process-related  means
 and not by natural means such as diurnal temperature changes.  Therefore, a
 storage tank is not covered by this  rule unless process emissions  associ-
 ated with the specified processes are vented through it.   The definition of
 venting cited above also applies in  response to the  question of Commenter
 AESP-00022 concerning whether equipment that vent  during loading or  due to
 breathing losses are covered.  Loading and breathing losses are not  caused
 by mechanical or process-related means such as  by  heating  in distillation
 processes, and hence, are not covered  by this  rule.
     Comment:  If, as stated in the proposal preamble, the product storage
 exemption is not affected by the control of process  accumulator vessels,
 then such units as product distillation columns generating organic hazard-
 ous waste still bottoms over 10 percent total  organics would not be  subject
 to the standards while they are in the distillation  column unit.   If the
 still bottoms are not considered to be a hazardous waste while in the dis-
 tillation column,  then the column is   not in VHAP service and not subject to
 the proposed standards (AESP-L0006).
     Response:   With regard to the product storage exemption,  paragraph
 261.4(c)  of the RCRA regulations exempts from  regulation hazardous wastes
that  are generated (emphasis added)  in process-related equipment such as
product or raw material  storage tanks, or product or raw material  pipe-
 lines.   This exemption applies until  the waste is physically removed from
the unit  in which  it was  generated,  unless the unit  is a surface
                                    3-15

-------
impoundment or unless the hazardous waste remains in the unit more than 90
days after the unit ceases to be operated for manufacturing,  or for storage
or transportation of product or raw materials.  This exemption is not
affected by the final rule; therefore,  units such as product (emphasis
added) distillation columns generating  organic hazardous waste still
bottoms containing greater than 10 ppmw organics would not be subject to
this regulation while the wastes are in the distillation column unit.
However, distillation columns that receive hazardous wastes and are used in
the hazardous waste treatment are covered by this regulation.  It also
should be noted that RCRA authority does not extend to equipment handling
commercial product; dropping the term "product accumulator vessel" should
help clarify this point.
     Comment:  There are contradictions in the regulation arid the preamble
regarding PAVs.  The preamble refers to PAV as types of equipment that
generate process emissions such as process vents, distillate receivers,
surge control vessels, product separators, or hot wells (52 FR 3753) and at
other points as units used to steam strip or air strip volatile components
from hazardous waste (52 FR 3749) or as "most distillation columns"  (52 FR
3753) (AESP-L0006).
     Response:  The wording of the proposal preamble and the proposed rule
inadvertently led to confusion concerning the scope of the standards.  For
clarity, the term  "product accumulator vessel" has been dropped from the
standards.
     Comment:  The preamble at 52 FR 3762 states that "...recovery devices
such as steam strippers treating organic-rich hazardous wastes that would
otherwise qualify  for the wastewater treatment exemption are not covered by
today's rule"; by  this reasoning, air strippers  in wastewater service would
be exempt and so would distillation columns and  air and steam strippers
(AESP-L0006).
     Response:  Paragraphs 264.1(g)(6)  and 265.1(c)(9) exempt wastewater
treatment tanks and elementary neutralization units (defined in 40 CFR
260.10) from the Subtitle C regulation.  The scope of that exemption is not
affected by this regulation.  Therefore, exempted recovery devices such as
air strippers, steam strfppers, or distillation  columns that are regulated
under either Section 402 or 307(b) of the Clean  Water Act  (CWA) are  not
                                    3-16

-------
covered by this rule.  However, air emission sources that are not control-
led under RCRA may still be covered under the CAA.
     Comment:  Does the 90-day accumulation exemption apply to tanks that
have 100 percent flowthrough (AESP-L0001)?
     Response:  As noted in the proposal preamble, 40 CFR 262.34 of RCRA
states that generator (emphasis added) tanks that accumulate hazardous
waste for 90 days or less are not subject to interim status or final permit
standards, provided they comply with most of the substantive standards for
tanks storing hazardous wastes.  Therefore, the promulgated standards do
not apply to generator tanks that accumulate hazardous waste for 90 days or
less.  The exemption for 90-day storage is discussed in the response to
comment 4.5.
     Comment:  The commenter asked EPA to clarify the closed-loop tank
exemption in the definition of solid waste; the commenter also asked
whether distillation columns (which have constant flowthrough and do not
accumulate) meet the definition of tank (i.e., distillation columns are not
tanks and application of standards to them is incorrect) (AESP-L0001).
     Response:  The final rules regulate the activity of reclamation at
certain types of RCRA facilities for the first time.  The EPA is amending
40 CFR 261.6 under its RCRA authority over reclamation to allow covering
reclamation of hazardous wastes in waste management units affected by the
final rules for process vents and equipment leaks.  It should be recog-
nized, however, that these final rules apply only at facilities otherwise
needing a RCRA permit.  In addition, the closed-loop reclamation in Part
261.4(a)(8) is not changed by these rules.  Therefore, not all reclamation
units will necessarily be affected by these rules.
     In response to a court opinion (American Mining Congress v. EPA, 824
F.2d 1177, D.C. Circuit Court of Appeals, July 31, 1987) concerning the
scope of EPA's RCRA authority, EPA proposed amendments to the RCRA defini-
tion of "solid waste" that would clarify when reclamation operations can be
considered to be managing solid and hazardous wastes (53 FR 519, January 8,
1988).  The- EPA has accepted comments on its interpretation and proposed
amendments.  The EPA has not yet taken final action on this proposal.
Thus, EPA is addressing the scope of its authority over reclamation opera-
tions under RCRA in the context of that rulemaking.  The process vent rule
is based on EPA's current interpretation of its RCRA authority, as
described in the January 1988 proposal.
                                    3-17

-------
     The following summarizes EPA's proposed position.  In general, the
proposed amendments would exclude from RCRA control only those spent
solvents reclaimed as part of a continuous, ongoing manufacturing process
where the material to be reclaimed is piped (or moved by a comparably
closed means of conveyance) to a reclamation device, any storage preceding
reclamation is in a tank, and the material is returned, after being
reclaimed, to the original process where it was generated.  (Other condi-
tions on this exclusion relate to duration and purpose of the reclamation
process.  See proposed Section 261.4[a][8].)
     However, processes (or other types of recycling) involving an element
of "discard" are  (or can be) within RCRA Subtitle C authority.  When spent
materials are being reclaimed, this element of discard can arise in two
principal ways.   First, when spent materials are reclaimed by someone other
than the generator, normally in an off-site operation, the generator of the
spent material is getting rid of the material and so is discarding it.  In
addition, the spent material itself, by definition, is used up and unfit
for further direct use; the spent material must first be restored to a
usable condition.  This type of operation has been characterized by some of
the worst environmental damage incidents involving recycling  (50 FR
658-661, January  4, 1985).  Moreover, storage preceding such reclamation
has been subject  to the Part 264 and 2.65 standards since November  19, 1980.
(See generally 53 FR 522 "and underlying record materials.)  The American
Mining Congress opinion itself indicates that such materials are solid
wastes  (824 F.2d  at 1187).
     When a spent material  is reclaimed on-site in something other than a
closed-loop process, EPA also considers that the spent material is
discarded (i.e.,  spent solvents removed from the process, transferred to an
on-site distillation unit,  and regenerated have been removed from  the
production process).  The EPA's reasoning is that these materials  are no
longer available  for use in an ongoing process and have been disposed from
that operation, even if the reclamation operation  is on-site.  Finally, EPA
also considers that when hazardous secondary materials are reclaimed but
then burned as fuels, the entire operation—culminating in thermal
combustion—constitutes discarding via destructive combustion  (53  FR 523).
Consequently, under this reading, any  intermediate reclamation step  in
these types of fuel production operations remains within EPA's Subtitle C
authority.
                                    3-18

-------
     In summary, under EPA's current interpretation of the court's opinion,
air emissions from distillation, fractionation, thin-film evaporation,
solvent extraction, and stripping processes involving reclamation of spent
solvent and other spent hazardous secondary materials can be regulated
under RCRA Subtitle C whenever the reclamation system is not part of the
type of closed-loop reclamation system described in proposed Part
261.4(a)(8).  Any changes to this interpretation as part of the solid waste
definition final rule may affect the scope of this rule.
     Comment:  The EPA should remove "condenser" from the PAV definition
because a primary or secondary condenser may be installed to .meet the
standard but also be a source for control and it is difficult to differen-
tiate a process condenser vent from a control device condenser vent for
determining compliance; the commenters suggest replacing reference to
condenser with "control device" and by expressing the standard in terms of
uncontrolled emissions from specific equipment (AESP-00024 and AESP-00034).
     Response:  With respect to the commenters1 concerns about differen-
tiating between a process condenser vent and a control device condenser
vent, for determining compliance, the primary condenser is considered a
component of the process (i.e., the primary function is the commerical
recovery of product), and any additional vent condenser is considered a
control device.  This distinction is important in determining the required
95-percent facility process vent emission reduction for facilities that
cannot achieve the process vent emission rate cutoff.
     Comment:  The EPA needs to give examples of equipment covered (AESP-
00020), to clarify sources in the chemical plants and petroleum refineries
that would be newly regulated (AESP-00010), and to clarify the specific
units covered so that EPA only regulates units for which sufficient sup-
porting information is available (AESP-00013).
     Response:  The process vent standards apply to vents emitting organic
liquids, gases, or fumes that are associated with hazardous waste man-
agement units treating wastes with a 10 ppmw or greater total organics
concentration and are specific to:  (1) process vents on distillation,
fractionation, thin-film evaporation,  solvent extraction, and air or steam
stripping operations and vents on condensers serving these operations, and
(2) process vents on tanks (e.g., distillate receivers, bottoms receivers,
surge control tanks, separator tanks,  and hot wells) associated with
                                    3-19

-------
distillation, fractionation,  thin-film evaporation, solvent extraction,  and
air or steam stripping processes if emissions from these process operations
are vented through the tank.   For example, uncondensed overhead emitted
from a distillate receiver (which fits the definition of a tank) serving a
hazardous waste distillation  process unit is subject to air controls.
Emissions from vents on tanks or containers that do not derive from a
process unit specified above  are not covered by these rules.  For example,
if the condensed (recovered)  solvent is pumped to an intermediate holding
tank following the distillate receiver mentioned in the above example, and
the intermediate storage tank has a pressure relief.vent (e.g., a conserva-
tion vent) serving the tank,  this vent will not be subject to the process
vent standards.  Emissions from vents that are not covered under today's
rules will be regulated by other air standards developed under 3004(n) of
RCRA.
     3.1.5.5  Other TSDF Sources.  Comment:  Commenters AESP-L0016 and
AESP-00019 ask why the proposed standards do not cover other TSDF sources
such as open tanks, waste solvent piping, storage, and transfer operations.
Commenter AESP-00004 states that open air sources, not solvent refining,
should be the focus of regulatory concern.  Commenter AESP-00006 recommends
that the regulations should be clarified to include or exclude storage
tanks, blending tanks, and transfer facilities at TSDF.  Five commenters
(AESP-L0018, AESP-L0016, AESP-00014, AESP-00008, and AESP-00019) also
question why the proposed standards apply to closed tanks, but not to open
tanks (including storage vessels) per Congressional intention under Section
3004(n).  The commenters state that the levels of emissions from vents and
fugitive releases from these sources are far higher than from equipment
leaks, particularly when the materials stored have high vapor pressures or
are filled and emptied routinely.  They point out that excluding open tanks
is unjustified environmentally and unfair to those TSDF that already have
installed controls.  Excluding open tanks also will undercut the proposed
rules because many TSDF will  switch to open tanks to avoid control.  In
comparision, one commenter (AESP-L0003) submits that the 1981 national
emission standard for hazardous air pollutants  (NESHAP) for benzene storage
tanks was not promulgated because the cost of controls greatly exceeded the
benefit and that similiar controls on TSDF storage tanks would also fail a
cost-benefit evaluation.
                                    3-20

-------
     Response:  As previously stated, other TSDF sources mentioned by the
commenters are being studied under the additional Section 3004(n) air emis-
sion rules scheduled for proposal in late 1989.  Treatment and storage
tanks,  surface impoundments, and containers are among the sources that will
be covered under these rules.
     The EPA agrees that storage tanks are a significant source of emis-
sions at TSDF.  Preliminary estimates presented in TSDF BID Volume I
(Docket No. F-90-AESF-FFFFF, item S0042) indicate that nationwide uncon-
trolled organic emissions from TSDF storage tanks may be as high as 756,000
Mg/yr (838,000 ton/yr).  As discussed in response to comment 3.2.3, whether
or not such standards would fail a cost-benefit evaluation is not germane
under RCRA.  Finally, the benzene storage proposal was withdrawn only on
the basis that the risks did not warrant Federal regulation under the CAA.
3.2  STATUTORY AUTHORITY AND REGULATORY APPROACH
3.2.1  RCRA vs. the CAA
     Comment:  A total of 22 commenters recommend that the CAA is the pro-
per statutory authority and regulatory mechanism for the standards rather
than RCRA.  They point out that volatile organic (VO) emissions and ozone
are the province of CAA programs, including: (1) National Emission
Standards for Hazardous Air Pollutants (NESHAP), NSPS, prevention of
significant deterioration (PSD), and State implementation plans (SIP).
Specific regulations such as the NSPS for fugitive emissions "from the SOCMI
and regulations under development such as the NSPS for volatile organic
liquid (VOL) storage, the SOCMI distillation NSPS, and the hazardous
organic NESHAP (HON) address the processes and chemicals (including inter-
mediates and products) likely to require fugitive or process controls at
new or existing TSDF or WSTF.  Several commenters also point out that these
standards would not allow the construction or expansion of recycling opera-
tions without emission controls.
     Many commenters state that the proposed standards are duplicative of
these CAA programs.  For example, several commenters explain that they are
subject to State or local reasonably available control technology (RACT)
controls in ozone nonattainment areas under Section 172 of the CAA.  Com-
menter AESP-L0025 provides information showing that using 95-percent con-
trol with RACT provides a better emission reduction in most cases and
yields control of noncarcinogens based on the physical properties of the

                                    3-21

-------
compounds.  Commenter AESP-L0003 states that, by adding redundant regula-
tions inconsistent with existing CAA standards,  the Agency is fragmenting
its rules and making them more confusing to the regulated community;
Commenter AESP-00019 points out that Section 1006(b) of RCRA provides that
regulations should avoid duplication to the extent possible with the CAA.
The commenters also state that the CAA already has a regulatory structure
in place to address hazardous air emissions for specific organic chemicals
on a risk priority basis that considers the relative toxicities and differ-
ing physical characteristics of the chemicals.  According to the commen-
ters, the objectives of Section 112 of the CAA and RCRA Section 3004(n) are
identical in protection of public health and the environment, and it is
more appropriate for EPA to extend existing CAA standards to TSDF than to
devise duplicative approaches under RCRA (AESP-00024, AESP-L0007, AESP-
L0003, AESP-L0001, AESP-00029, AESP-00033,  AESP-00007, AESP-00010, AESP-
L0006, AESP-00019, AESP-00016, AESP-00025,  AESP-00027, AESP-00028, AESP-
00030, AESP-00031, AESP-00035, AESP-L0025,  AESP-00004, AESP-00002, AESP-
00023, and AESP-00017).
     Response:  Both RCRA and the CAA authorize EPA to require monitoring
and control of air emissions at TSDF.  Section 3004 is not limited to
certain media, although EPA regulations have focused on preventing the
contamination of soil and water.  Indeed, Section 3004(n) specifically
commands EPA to issue regulations controlling air emissions from TSDF.
There is no indication that Congress intended that all air regulations be
issued within the confines of the CAA.  On the contrary, when adding
Section 3004(n), Congress specifically recognized EPA's dual authority to
regulate these air pollutants (S. Rep. 98-284 page 63).  As described
below, EPA has conducted an analysis of current State and Federal controls
and concluded that further regulation under Section 3004(n) of RCRA is
necessary to protect human health and the environment.
     In response to the commenters' concerns, EPA examined State and
Federal regulations, including CAA SIP requirements, to determine the po-
tential for overlapping rules and duplicative permitting (Docket No. F-90-
AESF-FFFFF, item S0016).  In general, the SIP contain strategies for
obtaining greater VOC emission reductions,  particularly in the 37 States
with ozone nonattainment areas.  Under this CAA strategy, any new or
modified source with the potential to emit over 40 ton/yr of uncontrolled
                                    3-22

-------
VOC emissions must obtain a permit to operate.  Because control technique
guidelines (CTG) are not available for TSDF, many States have applied the
CTG for VOC leaks from synthetic organic chemical and polymer manufacturing
equipment (EPA-450/3-83-006, March 1984) and/or extended other CAA stand-
ards such as the NSPS for VOC emissions from petroleum refineries or for
SOCMI.  A total of 6 States have established air toxics programs that would
apply to TSDF, and an estimated 60 additional States and localities
currently are in the process of developing such air toxics programs.  More-
over, a total of 21 States have established generic standards for VOC
sources or processes that are separate from Federal NSPS, national emission
standards, ambient standards, and CTG programs.  These regulations primar-
ily affect non-CTG sources of VOC emissions such as process vents, VOC
water separators, equipment leaks, and vapor blowdown systems.
     Because the standards vary widely in scope and application, it is not
possible to quantify a typical baseline level of control.  However, in most
cases, it appears that, while CAA permits to operate have been issued for
process vents at onsite and offsite TSDF and perhaps have been incorporated
in the RCRA permitting process, controls generally are not required for
these significant organic emission sources.  For example, several respon-
dents to EPA Section 3007 questionnaires that were sent to a limited number
of TSDF reported requirements for air contaminant source operating permits,
but they indicated no permit requirements for controlling process vent
emissions.  The nonuniformity among the States in applying organic emission
controls at TSDF supports the Agency's view that nationwide standards are
needed.
     The potential effect on waste solvent treatment facilities of Federal
standards currently under development also was examined.  In this regard,
EPA reviewed such actions as the proposed NSPS for SOCMI distillation
operations (48 FR 57538, December 30, 1983), the HON under development, and
potential listing decisions under Section 112.  The SOCMI NSPS, if promul-
gated, would control VOC emissions from condensers, accumulators, hot
wells, steam jet ejectors, vacuum pumps, and pressure relief valves on new,
modified, or reconstructed fractional distillation units at organic
chemical manufacturing plants.  Existing units at SOCMI plants would not be
covered by the NSPS, nor would distillation units used for recycling in
many of the 150 other industries with onsite TSDF or at commercial
                                    3-23

-------
facilities.  As noted in the EPA Semiannual Regulatory Agenda at
52 FR 40903 (October 26, 1987), new issues have arisen in this SOCMI rule-
making, and the Agency is deciding whether to promulgate these standards.
The potential  overlap of the HON also was cited by several commenters in
support of promulgating these standards under the CAA.  Although still
under development, this project would govern emissions of butadiene, carbon
tetrachloride, chloroform, ethylene dichloride, ethylene oxide, methylene
chloride, perchloroethylene, and trichloroethylene from process vents,
equipment leaks, and storage facilities.  Regulated source categories would
include facilities used to produce butadiene, chlorinated hydrocarbons,
chlorine, chlorofluorocarbons, ethylene dichloride, ethylene chloride,
neoprene, pesticides, Pharmaceuticals, polybutadiene,  and stryene butadiene
rubber, as well as chlorinated hydrocarbons used in chemical production.
It is estimated that the rule would cover about 5 percent of the SOCMI's
total production, or between 120 and 150 individual processing units.  The
air controls under consideration for these process sources (and not for
waste treatment sources) would be similar to those already required under
the benzene fugitive standards.  Thus, EPA does not believe that these
standards for air emissions from waste treatment sources are duplicative of
NSPS or NESHAP programs under development.
3.2.2  Level of Control Under RCRA
     Comment:   Commenter AESP-L0016 criticizes the level of control
selected for the proposed standards as not meeting the mandate of RCRA
Section 3004(n) for two major reasons: (1) the standards are protective in
most, but not all cases (further protection is needed in the form of
enforceable regulations and not through the imposition of more stringent
controls by permit writers); and (2) the standards are inconsistent with
Section 3004(m) that requires treatment standards based on BOAT.  On the
latter point,  the commenter argues that the proposed air rules also must be
based on BOAT (or the most effective techniques that are available), but
instead employ inadequately stringent controls despite the fact that more
effective techniques are applicable and available.
     Response:  The question of whether these standards implement the
requirements of RCRA Section 3004(m) is irrelevant.  Regulations imple-
menting Section 3004(m), which is a pretreatment-based program that defines
when hazardous wastes can be land disposed, have been (and will continue to
                                    3-24

-------
be)  separately promulgated by EPA.  For example,  see 51 FR 40636 (November
7,  1986)  and 52 FR 25787 (July 8,  1987).  Today's regulations are promul-
gated as  a first step in implementing Section 3004(n) of RCRA, which
requires  that the regulations be those that are necessary to protect human
health and the environment.  Therefore, in developing today's rule, EPA has
focused on achieving acceptable levels of health and environmental  protec-
tion, rather than on meeting a certain treatment level.
     Moreover, as stated in other responses to comments, EPA is developing
additional standards beyond those promulgated today for other emission
sources and for particular constituents that may pose unacceptable risk
after implementation of these and the additional  TSDF regulations.   To
fully address the health and environmental effects of TSDF air emissions,
the wide variety of sources, and the need for different data with increas-
ingly complex analyses, EPA has developed a multiphased standards develop-
ment approach for regulating TSDF organic air emissions.  This approach is
described below.
     The EPA has determined that organic emissions from TSDF managing haz-
ardous wastes increase cancer risks and contribute to ambient ozone
formation.  Phases I and II of the TSDF emission standards approach will
significantly reduce emissions of air toxics and carcinogens, and ozone
precursors by controlling emissions of organics as a class rather than
controlling emissions of -individual waste constituents.  The regulation of
organics as a class has the advantage of being relatively straightforward
because it can be accomplished with the minimum number of standards,
whereas the control of individual toxic constituents will require multiple
standards.  Regulating organics as a class also makes efficient use of EPA
resources, avoids many of the complexities of having multiple standards,
and reduces the number of constituents for which separate standards
ultimately will be required.
     The health and environmental effects of ambient ozone are well docu-
mented and, in terms of monetary losses, they total hundreds of millions of
dollars each year.  Health impacts of TSDF organic emissions are discussed
in this BID and in the draft BID for Phase II organic standards titled
"Hazardous Waste TSDF - Background Information for Proposed RCRA Air
Emission Standards," available in Docket F-90-CESP-FFFFF.  The substantial
reductions in organic emissions achievable through implementation of
                                    3-25

-------
Phase I and Phase II  controls will  reduce atmospheric ozone formation as a
result of reductions  in TSDF emissions of ozone precursors and will reduce
nationwide cancer incidence and maximum individual risk due to exposure to
air toxics and carcinogens emitted  from TSDF.
     Specifically,  Phase I entails  the promulgation of standards for the
control of organic  air emissions from selected hazardous waste management
processes and equipment leaks.  Publication of final rules for air emis-
sions from hazardous  waste management process vents from distillation,
fractionation, thin-film evaporation, solvent extraction, and air or steam
stripping processes and from leaks  in piping and associated equipment
handling hazardous  wastes marks the completion of this first stage.
     Like the process vent and equipment leak standards, the standards
developed under Phase II will control organic air emissions as a class.  In
the second stage, EPA will propose  to extend standards (in mid-1990) under
Section 3004(n) to  include organic  air emissions from other significant
TSDF air emission sources not covered or not adequately controlled by
existing standards.  Those sources  include surface impoundments, storage
and treatment tanks (including vents on closed, vented tanks), containers,
and miscellaneous units.
     Under Phase III  of the program, EPA may develop additional standards
applicable to sources regulated in  the process vent and equipment leak
rule, and to sources  covered in Phase II.  The Agency recognizes that the
potential remains for high risk at  some facilities because of individual
chemical constituents or other physical characteristics of the facility
even after the substantial emission reductions achieved through Phase I and
II controls.  For this reason, EPA  will examine the need for standards for
the control of emissions of specific hazardous waste constituents from all
TSDF sources that may continue to pose a high health risk after the
standards for organics are implemented.  The constituents to be evaluated
will include those reported as being present in hazardous wastes managed by
existing TSDF for which health effects have been established through the
development of unit risk factors for carcinogens and reference doses for
noncarcinogens.  General regulatory approaches that will be considered vary
widely and range from requiring additional technology controls to limiting
the quantities of specific constituents that can be managed at a TSDF.  A
more detailed discussion of Agency  plans to address TSDF residual risk is
                                    3-26

-------
presented in Section VIII of the preamble to the process vent and equipment
leak rules.  Because the development of these Phase III regulations will
require more detailed analysis and the bulk of the TSDF air emission
problems will be dealt with under Phases I and II, the Phase III rules will
be proposed and promulgated after those developed in the first two phases.
     The EPA has selected this multiphased approach, in part, because of
the complexity of the TSDF industry and the considerable uncertainty in
currently available data, particularly with regard to individual chemical
constituent emissions and risk estimates.  Industry surveys soon to be
completed by the EPA are expected to provide detailed, up-to-date data on
waste constituent concentrations, waste quantities, and processing
locations needed to provide more accurate estimates of health impacts.
These data will be fully reviewed and incorporated into emission and risk
analyses prior to promulgation of the Phase II standards and proposal of
Phase III standards.
3.2.3  Consideration of Costs
     Comment:  Five commenters question the role of costs as a RCRA deci-
sion factor.  Commenter AESP-L0016 states that the proposed rule is in
violation of RCRA for rejecting tighter controls because, according to the
commenter, neither the statute nor the RCRA legislative history allows
consideration of costs.  In comparison, other commenters ask why costs and
cost effectiveness were not more explicitly considered and why the risk/
cost/benefit value of $35.3 million/yr cited at 52 FR 3765 of the preamble
is higher than the cost-benefit level used by EPA in previous CAA control
decisions (AESP-L0018, AESP-00034, AESP-00024, and AESP-00010).
     Response:  The role of costs as a decision criterion under RCRA in
Subtitle C is not explicitly addressed in the statute.  The Agency's
position on this topic was discussed in 1980 when EPA established the basic
Subtitle C regulatory framework (45 FR 33089, May 19, 1980).  The EPA
believes that it has the authority under RCRA Subtitle C to collect cost
information for presentation to the Administrator and the public.  The EPA
also may consider costs as a basis for choosing between alternatives that
are equally protective.  However, under Subtitle C, EPA may not select
controls or standards that do not adequately protect human health and the
environment based on explicit consideration of costs.  The EPA does not
believe that the cost burden on industry is a basis for reducing the
                                    3-27

-------
stringency of standards the agency considers necessary to protect human
health and the environment.
     Emission levels considered protective of human health and the environ-
ment on a nationwide basis were selected for both process vents and equip-
ment leaks without regard to cost (see comments 7.1 and 7.6, respectively).
Because the standards for process vents are performance-based, the stand-
ards do not specify how a facility owner or operator must meet the stand-
ards, but allow him to use whatever equipment he chooses as long as the
standard is met.  Cost impacts of the standard are estimated by EPA based
on the least expensive means of control.  Cost estimates were also devel-
oped to evaluate the economic impacts of the equipment leak standard only
after a the control level was selected.
3.2.4  RCRA Section 3004(n)
     Comment:  Conimenter AESP-L0006 questions EPA's authority under RCRA
Section 3004(n) to propose controls for leaking equipment at TSDF because
Congressional intention  refers to major sources such as open tanks, surface
impoundments, and landfills, and it is arguable, according  to the commen-
ter, that this authority extends to fugitive sources traditionally regu-
lated under the CAA.  Commenter AESP-00019  states that Section 3004(n) does
not  imply that WSTF should be regulated before other major  TSDF sources.
     Response:  The EPA  disagrees that Congressional  intention under Sec-
tion 3004(n)  limits regulations to tanks, surface impoundments, and land-
fills.  The section clearly  specifies  regulations "for the  monitoring  and
control of air emissions  at  hazardous  waste TSDF, including but not limited
to  open tanks, surface  impoundments,  and  landfills, as may  be  necessary  to
protect human health  and the environment"  (emphasis added).   The  EPA
believes  that any source at  a hazardous waste  TSDF that  endangers  human
health  and the environment can,  and should,  be  regulated  under this author-
ity.   Uncontrolled  fugitive  emissions  from  WSTF  and other TSDF equipment
leaks  are estimated to  total about 26,000 Mg/yr  (29,000  ton/yr)  and result
in  an  estimated  1.1 cases  of cancer per year.   The  fact  that  equipment leak
sources  at TSDF  are  regulated  before  other  sources  is not germane to  this
 issue.   As discussed  in response to comment 3.2.8,  these standards preceded
the additional  Section  3004(n)  air emission rules  because much of the
 information  needed  to develop  these  standards  was  already available.
                                     3-28

-------
3.2.5  RCRA Authority to Control Reclaimed Commercial Product
     Comment:  Four commenters (AESP-00019, AESP-L0003, AESP-L0018, and
AESP-L0025) question EPA's authority under RCRA to regulate equipment used
to store and handle commercial product, such as distillate receivers.  The
commenters state that, while the CAA provides this authority, RCRA does
not.  Under 40 CFR 261.3(c)(2), commercial products reclaimed from haz-
ardous wastes are products, not wastes; commercial products are not covered
by Subtitle C regulations, and regenerated solvents are not wastes.
     Response:  The EPA agrees that commercial products are not covered
under RCRA.  Therefore, vents from tanks such as distillate receivers that
emit organics derived only from the reclaimed products are not subject to
the standard.  However, if the distillation column overhead stream
containing uncondensed organics from processing hazardous waste with a
total organic concentration above the applicability criterion is vented to
the atmosphere through a tank associated with the distillation operation,
then the tank vent is subject to the requirements of the standard.
     A more detailed discussion of EPA's authority to regulate reclamation
(or regeneration) operations is presented in Section 3.1.5.4 of this BID.
3.2.6  Total Organics Approach
     Comment:  Numerous commenters contend that the proposed rules for
process vents should be based on volatility and not their total organic
content because the relative amount of organic content by weight does not
determine its potential air emissions and subsequent health effects.  They
state that the definition of "VO" (1) should include only volatile organics
instead of total organics; (2) should exclude ozone precursors to be con-
sistent with the definition of "VOC" under the CAA and CAA State regula-
tions, which specifically exclude methyl chloroform, methylene chloride,
and CFC-113 because they are negligibly photochemically reactive; and (3)
should be limited to known or suspected carcinogens listed in 40 CFR 261,
Subpart D, that are volatiles.  According to these commenters, it is inap-
propriate to lump thousands of RCRA hazardous chemicals together and assume
that all should be controlled in the same manner because emissions and need
for control depend on vapor pressure, solubility, heat content, corrosiv-
ity-f and combustibility, as well as health effects.  One commenter states
specifically that EPA's proposed control of nonvolatile, nonhazardous pol-
lutants is not authorized under risk-based Section 3004(n).  (AESP-L0011,
                                    3-29

-------
AESP-00010, AESP-L0005, AESP-L0006, AESP-00011, AESP-00003, AESP-L0008,
AESP-00018, AESP-00013, AESP-00030, AESP-L0003, AESP-L0001, AESP-00016,
AESP-00019, AESP-00033, AESP-L0002, AESP-L0007, AESP-L0015, AESP-L0018,
AESP-00035, AESP-00002, AESP-00034, AESP-00035, AESP-L0009, AESP-00017,  and
AESP-00024).
     Response:  First, it should be pointed out that EPA is concerned about
ozone formation as well as cancer risk resulting from TSDF organic emis-
sions.  Ozone presents a threat to human health and the environment that
warrants control under RCRA.  The EPA agrees that total organic content may
not be a completely accurate guage of potential environmental (e.g., ozone)
or health  (e.g., cancer) impacts for a source such as process vents, but it
is a readily measurable indicator.  In addition, the final rule's substan-
tive control requirements do apply only to vents and equipment containing
volatile components.
     The final vent standard applies to certain process vents emitting
organics,  if the vent is associated with one of the process units specified
in the rule and the unit manages a hazardous waste with at least 10 ppmw
total organic concentration on an annual average basis.  A process vent is
determined to be affected by the standard if the vent is part of the
hazardous waste distillation, fractionation, thin-film evaporation, solvent
extraction, or air or steam stripping unit; this includes vents on tanks
(e.g., distillate receivers or hot wells) serving the units if emissions
from the process operations are vented through the tank.  Total organic
content of the process vent stream is not a consideration in determining
process vent applicability.  As public commenters pointed out, the 10-
percent total organics process vent concentration cutoff does not limit
total emissions or relate to emissions that escape capture by existing
control devices and therefore was not included in the final rules.
     Furthermore, the process vents covered by this rule are typically
associated with distillation/separation processes used to recycle spent
solvents and other organic chemicals.  By definition, distillation is a
process that consists of driving gas or vapor  from liquids or solids by
heating and then condensing the vapor(s) to liquid products.  Wastes
treated by distillation are expected to contain organics that are driven
off  in the process.  Thus, by their nature, process vent emissions contain
volatile organics.
                                    3-30

-------
     The EPA agrees, however, that total organic content may not accurately
describe the air emissions in terms of the potential impacts for a source
such as process vents.  Therefore, the total facility emission rate limit
of 1.4 kg/h (3 Ib/h) and 2.8 Mg/yr (3.1 ton/yr) below which additional,
quantifiably significant risk reductions are not attained has been added to
the final rule to exclude some facilities from the process vent emission
reduction requirements.  In determining the emission rate limit, emissions
and health risks have been considered.  (The cutoff determination is dis-
cussed in response to comment 7.1.)
     As the commenters suggest, thousands of RCRA chemicals could be con-
tained in various hazardous wastes treated by distillation and in hazardous
wastes treated by WSTF.  Developing and implementing regulations based on
specific constituents would be needlessly cumbersome and complex.  As de-
scribed in Chapter 6.0, a range of unit risk factors was derived that
represent the health risks of the hazardous constituents most frequently
found in hazardous waste treated by WSTF.  A midrange value of process vent
constituent risk factors was used in evaluating the health risks associated
with process vent emissions and in determining the cutoff.  Compounds
excluded by States in their definitions of VOC were not excluded because
several are suspected carcinogens that currently are being studied by EPA.
     Under the final standards, the term "organic emissions" is used in
lieu of "volatile organic emissions" to avoid confusion with "volatile
organic compounds (VOC)."  As at proposal, the final rule applies to total
organics.  Because of the hundreds of hazardous constituents that could be
contained in and contacted by the equipment covered by today's rules, the
Agency recognizes the potential for risk at some facilities to remain
higher than the residual risk for other standards promulgated under RCRA.
Regulations based only on specific constituents will therefore be
developed, as necessary, in Phase III of the Agency's regulatory approach.
The constituents to be evaluated will include those reported as being
present in hazardous wastes managed by existing TSDF for which health
effects have been established through the development of unit risk factors
for carcinogens and reference doses for noncarcinogens.
     Similarly, in analyzing the technical feasibility of achieving
95-percent control efficiencies, compounds frequently recycled by WSTF were
considered.  Compounds used in evaluating control techniques (Docket No.
                                    3-31

-------
90-AESF-FFFFF,  item S0047)  include methyl  ethyl  ketone (MEK),  methylene
chloride, toluene,  and 1,1,1 trichloroethane,  which have properties repre-
senting the range of chemical  and physical characteristics of  recycled
solvents.  Chapter 7.0 of this document includes discussions  of the techni-
cal feasibility of applying condensers and carbon adsorption  systems to
process vent streams to comply with the regulation.
     Comment:  Many comments also were received stating that  the use of
total organics is not appropriate for defining equipment "in  VHAP service"
because many of the organics at TSDF have low volatilities (they are col-
lected at ambient conditions).  Many find the definition confusing when
compared to the established definition-of "in VHAP service" under CAA
standards because the CAA definitions generally are limited to specific
chemicals and distinguish between light and heavy liquids by  vapor pres-
sure.  These commenters suggest that EPA adopt the established CAA defini-
tion based on vapor pressure because including nonvolatiles will (1) result
in unnecessary monitoring,  and (2) conflict with other statutes such as the
CAA, the CWA, Superfund, etc., where these nonvolatiles are not considered
hazardous (AESP-L0011, AESP-Q0010, AESP-L0005, AESP-L0006, AESP-00011,
AESP-00003, AESP-L0008, AESP-00018, AESP-00013, AESP-00030, AESP-L0003,
AESP-L0001, AESP-00016, AESP-00019, AESP-00033, AESP-L0002, AESP-L0007,
AESP-L0011, AESP-L0015, AESP-L0018, AESP-00035, AESP-00002, AESP-00034,
AESP-L0009, AESP-00017, and AESP-00024).
     Response:  To clarify that the equipment leak standards apply to total
organic emissions from hazardous wastes, the term  "in VHAP service" has
been dropped from the promulgated rule.  However, the EPA agrees that the
equipment leak standards should take component volatility into considera-
tion.  Previous EPA and industry studies have shown that the volatility of
stream components, as a process variable, does correlate with fugitive
emission rates (Docket No. F-90-AESF-FFFFF, items  S0001 through S0008).  An
analysis of the vapor pressures and emission rates has shown that  sub-
stances with vapor pressures  of 0.3 kPa or higher  had significant  emission
rates, while those with lower vapor pressures did  not.  For example,  EPA
analyses indicate that emissions from  valves in  light-liquid service  are
more than 30 times higher  than valves  in  heavy-liquid service  (see Section
4.6 of this BID).  This result led to  the separation of equipment  component
emissions by service: gas/vapor,  light liquid, and heavy  liquid.   These
                                    3-32

-------
classifications have been used in most fugitive emission standards to
effectively direct the major effort toward equipment most likely to leak.
Therefore, the proposed rules for equipment leaks have been revised to
account for volatility.  For example,  pumps and valves in heavy-liquid
service must be monitored only if evidence of a potential leak is found.
Light- and heavy- liquid service is based on the vapor pressure of the
components in the stream (less than 0.3 kPa at 20 °C).
3.2.7  Transfer of the Benzene NESHAP Technology
     Comment:  Eight commenters (AESP-00035, AESP-00024, AESP-00002, AESP-
00005, AESP-00028, AESP-L0015, AESP-L0016, and AESP-00018) object to the
regulatory approach of incorporating the benzene NESHAP as the basis of the
standards because: (1) this regulation is not applicable to the broad spec-
trum of wastes found at TSDF, and (2)  TSDF differ from the chemical plant
and petroleum refineries upon which the benzene standard is based.  Most
commercial recyclers comment that they do not believe that the benzene
NESHAP is appropriate as the basis for emission standards at their facil-
ities because benzene use has been almost entirely phased out except as a
raw material in manufacturing and as a gasoline additive.  Commenter AESP-
L0016 also objects because EPA has not assessed how the cost effectiveness
of benzene controls differs when applied to TSDF.  Commenter AESP-00018
agrees that the benzene study that is the basis of the original data used
to support the proposed standards is inapplicable because of benzene's high
volatility, which they say is atypical of TSDF.
     Response:  The data used by EPA in establishing the benzene fugitive
standards are based on emission and process data collected at a variety of
petroleum refinery and SOCMI operating units.  Data were obtained for
equipment and chemical component mixtures that include many of the same
organic compounds treated in TSDF hazardous waste management units, not
only benzene.
     The EPA's Industrial Environmental Research Laboratory (IERL) coordi-
nated a study to develop information on fugitive emissions in the SOCMI
(Docket No. F-90-AESF-FFFFF, item S0010).  A total of 24 chemical process
units were tested; these data covered thousands of screened sources (pumps,
valves, flanges, etc.) and included process units handling such chemicals
as acetone/phenol, MEK, ethylene dichloride, 1,1,1- trichloroethane, tri-
chloroethylene, and perchloroethylene.  Refinery studies on fugitives also
                                    3-33

-------
include tests on units handling both toluene and xylene (Docket No. F-90-
AESF-FFFFF,  items S0001 through S0005).   These same chemicals are included
in those listed by the National Association of Solvent Recyclers (NASR) as
solvents commonly recycled by member facilities.  (Docket No. F-86-AESP-
FFFFF, item L0014).
     One of the primary purposes of SOCMI plants is to manufacture organic
solvents of the variety found at TSDF.   Some refineries also produce or-
ganic chemicals and solvents.  Because hazardous waste management units
such as distillation operations have the same sources of fugitive organic
emissions (such as pumps and valves) involved in handling SOCMI chemicals,
it is reasonable to expect similar control technology performance and ef-
ficiency at hazardous waste management units.  The EPA has no reason to
believe that the equipment standards would not be applicable to TSDF.
Moreover, although EPA has not conducted actual equipment leak testing at
TSDF, observations of equipment during plant visits have confirmed that the
assumptions and analyses used in other equipment leak standards apply to
TSDF as well.
     With regard to the concerns of Commenter AESP-L0016 that the cost
effectiveness of benzene controls applied to TSDF be assessed, as noted in
the response to comment 3.2.3, RCRA excludes selection of controls or
standards based on consideration of costs.  However, costs were developed
to evaluate the economic impacts of the equipment leak standard after con-
trol levels were selected.   Estimates of the costs for fugitive emission
control at WSTF/TSDF model units are presented  in Appendix C.
     The commenters also express concern about  differences in volatilities
between SOCMI streams and TSDF waste streams.   To avoid unnecessary emis-
sion monitoring of equipment handling low volatility streams, the  final
rules have incorporated the  light/heavy  liquid  designation found in the
SOCMI NSPS for equipment leaks.
3.2.8  Accelerated Approach
     Comment:  Five commenters  (AESP-L0018, AESP-00002, AESP-00030, AESP-
00035, and AESP-L0025) do not  support EPA's regulatory approach of promul-
gating the proposed standards  on an accelerated basis, separate from the
comprehensive TSDF standards currently under development.  Commenter
AESP-L0018 specifically recommends  that  EPA proceed only with the  LDAR
provisions and re-propose other requirements when sufficient data  and
                                    3-34

-------
supporting analyses can be provided.  Commenter AESP-00002 suggests that
EPA delay promulgation until sufficient data are available.  Commenter
AESP-00025 points out that proceeding with these standards as a separate
project would require industry to apply controls that may be obsolete or
not cost-effective when considered in conjunction with the comprehensive
rules.  Commenter AESP-00016 believes EPA should withdraw the entire
proposal because it is not adequately supported by the analyses.
     Response:  The EPA accelerated this portion of the air rules in part
to prevent uncontrolled air emissions from LDR treatment technologies.
This was possible because control techniques used in meeting current CAA
standards for the same types of emission points can easily be applied to
TSDF.  After proposal, EPA contacted commenters and collected additional,
TSDF-specific information for use in vent analyses to respond to the com-
ments received.  The EPA has collected and analyzed site-specific data,
reviewed control techniques, and revised estimates of emissions, risks,
costs, and economic impacts.  The results of these analyses are discussed
in Chapters 5.0 through 9.0 of this document.  The revised impacts analysis
conducted by EPA includes a facility-specific analysis for both WSTF (448
facilities) process and fugitive (equipment leak) emissions, and for TSDF
(more than 1,400 facilities) fugitive emissions from those waste management
processes that handle hazardous waste streams with at least 10 percent
organics.  While a wealth of data is not available on all aspects of these
standards, EPA has no reason to believe that the standards should be
delayed.  In fact, the impacts of delay are compelling reasons to move
ahead, given the available control techniques and existing standards that
form the basis for these rules.
     As discussed in other responses to comments in this document, the
Phase II Section 3004(n) rules currently under development affect other air
sources at TSDF that require different types of data and technical
approaches.  These standards are scheduled for proposal in mid-1990.  The
EPA does not believe that the implementation of standards that require
add--on controls for process vents and a combination of an LDAR program,
work practices, and controls for equipment leaks will impair or be rendered
obsolete by design and engineering plans for other sources.
                                    3-35

-------
                      4.0  APPLICABILITY AND EXEMPTIONS

4.1  APPLICABILITY OF STANDARDS FOR STORAGE TANKS AND ACCUMULATOR VESSELS
     Comment:  Commenters AESP-L0001 and AESP-00024 object to the regula-
tory approach of applying a single standard to the wide varieties of
accumulator vessels (AV) irrespective of the chemical constituents that are
present and the size of the vessel because the proposed standards result in
the control of already low emission rates at disproportionately high costs.
The commenters state that the standards for tanks (whether accumulation or
storage tanks) should be conditioned by the size of the vessel, the vapor
pressure of the material being stored, and the type of units that pose a
risk to human health and the environment.  Commenter AESP-00022 recommends
that EPA establish a minimum tank size that will not be regulated and urges
that a minimum vapor pressure be used to determine which tank contents will
be classified as VO or as semi volatile (unregulated) organics because
accurate estimates of total VO cannot be made for tank contents unless a
specific cutoff is established.  Commenters AESP-L0002 and AESP-00033 also
support the development of tank standards that consider size and vapor
pressure and suggest that EPA's approach should be similar to or consistent
with the CAA NSPS for petroleum liquid storage vessels (40 CFR 60, Subpart
Ka).  These standards exempt vessels that store liquids less than 1.5 psia
or that store less than 40,000 gal.
     Response:  Commenters recommending that the air emission standards be
conditioned by the size of the tank and the vapor pressure of the material
being stored have misinterpreted the applicability of the proposed stand-
ards.  To clarify the applicability of the standards, the term "product
accumulator vessel" has been dropped from the promulgated rule, including
the equipment definition, and the process vent definition has been revised
to be specific to the applicable emission sources.   "Process vent" is
                                     4-1

-------
defined to mean "any open-ended pipe or stack that is vented to the atmos-
phere either directly, through a vacuum-producing system,  or through a tank
(e.g., distillate receiver, condenser, bottoms receiver,  surge control
tank, separator tank, or hot well) associated with distillation, frac-
tionation, thin-film evaporation, solvent extraction, or air or steam
stripping operations."  Similarly, the definition of "vented" has been
revised to specifically exclude the passage of liquids, gases, or fumes
"caused by tank loading and unloading (working losses)."  Because tank
working and breathing losses are not considered process emissions, the
comments concerning vapor pressure and tank size exemptions are not
relevant.  It should be noted, however, that EPA is considering the
regulation of hazardous waste storage tanks, along with various other TSDF
air emission sources, in the Phase II, Section 3004(n) TSDF air standards
now being developed and evaluated by the Agency.  In the development of
these standards, EPA will consider such factors as size of the tank and the
volatility of the material being stored.
      In conducting the impact analysis of the TSDF process vent standards,
EPA considered and took into account the relative size of TSDF process
units and the wide range of chemicals processed in the TSDF industry.  For
example,  three sizes  of model units  (each with a range of operating param-
eters) were defined  for analysis of  emissions, health  risks,  and  economic
impacts  in the final  rulemaking.  The final  standards  for process vents
promulgated by EPA are effectively conditioned by the  size of the unit
(i.e., process throughput) by requiring controls at  facilities whose  total
process  vent emissions are greater than 1.4  kg/h  (3  Ib/h) and 2.8 Mg/yr
(3.1  ton/yr).  Emission estimates were based  on facility throughputs;
therefore, adequate  consideration has been  given to  the TSDF  facility and
process  unit size  in  establishing the standards.  A  more detailed discus-
sion  of  the process  vent cutoff  is provided  in  response to  comment  7.1.
4.2   SELECTION OF  THE 10-PERCENT CUTOFF
      Comment:  Commenters  AESP-00014 and AESP-L0015  ask EPA  to evaluate  the-
health  and environmental  impacts associated  with  the proposed 10-percent
cutoff;  Commenter  AESP-00008  recommends  that a  reduction  in  the limit or a
different parameter be used  because, at  the current  level,  the standards
would not apply  to very  many  facilities.   As discussed in  comment 3.2.6,
                                      4-2

-------
numerous coranenters state that the volatility of the wastes should be
considered.  Commenter AESP-L0016 states that the 10-percent limit will
allow excessive emissions from leaking equipment and is based on costs, not
technical feasibility.  The commenter argues that the 10-percent limit does
not adequately protect the environment because emissions could be substan-
tial if there are numerous leaking components with relatively dilute
streams because controls, such as carbon adsorbers,  are available to
capture emissions from dilute streams.  The commenter cites Raoult's law to
show how, at higher temperatures, affected solvents  could vent more than
10,000 ppm (1 percent) even if their concentration were less than
10 percent and generate significant emissions that remain uncontrolled.
According to Commenter AESP-L0016, affected vessels, like distillation
columns, contain materials with differing concentrations throughout their
length.  If the 10-percent limit relates to the input stream to such a
column, then parts of the same piece of equipment would contact wastes at
lower concentrations.  For these reasons, Commenter  AESP-L0016 states that
EPA should abandon the cutoff approach based on concentration and focus on
the emissions that escape capture.
     Response:  As proposed,  the 10-percent total organics cutoff level for
applicability of the standards covered both equipment leak (fugitive) emis-
sions and process vent emissions.  Control technologies for fugitive emis-
sions comprise the use of control equipment, inspection of equipment, and
repair programs to limit or reduce emissions from leaking equipment.  These
control technologies have been studied and evaluated for equipment contain-
ing fluids with at least 10 percent organics (EPA-450/3-80-032b, EPA-450/3-
80-033br EPA-450/3-82-010, and EPA-450/3-86-002.)  The 10-percent criterion
was chosen in EPA's original  benzene/SOCMI studies to focus the analyses on
air emissions from equipment  containing relatively concentrated organics
and presumably having the greatest potential for air emissions.  Available
data from the original benzene/SOCMI studies do not  suggest that fugitive
emissions from leaking equipment (e.g., pumps and valves) handling streams
containing less than 10 percent organics are significant or that the
10-percent cutoff allows excessive emissions from dilute streams.  However,
to re-evaluate this would require several years to conduct field studies to
collect and analyze additional emissions and control effectiveness data for
                                     4-3

-------
equipment leaks.  Because available data support the need for and effec-
tiveness of standards for equipment handling streams containing at least
10 percent organics, the EPA does not believe that a delay in rulemaking to
assess emissions and controls for equipment handling streams containing
less than 10 percent organics is warranted.
     The effectiveness of fugitive emission control technologies has been
thoroughly evaluated for equipment containing fluids with at least
10 percent organics, and fugitive emission standards have been proposed or
established under both Sections 111 and 112 of the CAA.  (See 46 FR 1136,
January 5, 1981; 46 FR 1165, January 5, 1981; 48 FR 279, January 4, 1983;
48 FR 37598, August 18, 1983; 48 FR 48328, October 18, 1983; 49 FR 22598,
May 30, 1984; 49 FR 23498, June 6, 1984; and 49 FR 23522, June 6, 1984.)
As elaborated in these rulemakings, a 10-percent cutoff deals with the air
emissions from equipment most likely to cause significant human health and
environmental harm.
     With regard to process vent emissions, EPA agrees with the commenter.
Emission test data show that the 10-percent cutoff potentially may allow
significant emissions  from process vents on a mass-per-unit-time basis
(e.g.,  kilograms per hour or megagrams per year).  As public commenters
pointed out, the 10-percent cutoff for process vents does not  limit total
emissions, nor  does  it relate to emissions that escape capture by existing
control devices.  Therefore, the 10-percent cutoff may not  be  appropriate;
as  a  result, EPA has eliminated the  10-percent cutoff as  it  applies to
process vents.  The  EPA believes that  an emission  rate  limit more effec-
tively  relates  to emissions, emission  potential,  and  health  risks than  does
a  10-percent organic concentration cutoff.  Accordingly,  a  health-risk-
based  facility  process vent  emission  rate  limit has been  added to the  final
rules  in  lieu of the  10-percent cutoff.
      Because the emission rate  limits  (3  Ib/h  and  3.1 ton/yr)  provide
reasonable  health-based cutoffs,  EPA considered dropping  completely the
organic content criterion  (i.e.,  at  least  10  percent  total  organics).
However,  EPA decided  not  to  completely eliminate  the  organic content
criterion  because  it  is not  clear  that the same controls  can be  applied to
very-low-concentration streams  as  can be  applied  to the higher-
concentration  streams  that  generally are  associated with  emission  rates
                                      4-4

-------
greater than the limits.  For low-concentration streams, EPA questions
whether controls are needed on a national or generic basis, but is unable
to resolve this question at this time.  Thus, EPA decided to defer
controlling very-low-concentration streams until it is able to better
characterize and assess these streams and the appropriate controls.
     Once EPA decided to consider facilities that manage very-low-concen-
tration organic wastes as a separate category, there remained the problem
of determining the appropriate criterion.  The EPA examined existing data
on air strippers, the treatment device most commonly used with low-
concentration streams; and it appeared that the quantity of emissions and
the risk associated with air strippers treating streams with concentrations
below 10 ppmw may be relatively small, thus minimizing the potential harm
of deferring control until a later time.  Table 4-1 presents a summary of
the data available on air stripping operations treating either ground water
or wastewater.  Examples of facilities managing low-concentration wastes
are sites where ground water is undergoing remedial action under the
Comprehensive Environmental Response, Compensation and Liability Act
(CERCLA) or corrective action pursuant to RCRA.  Based on the limited set
of precise data available, and the comments that the 10-percent criterion
was too high, EPA determined that an appropriate criterion would be 10
parts per million (ppm) total organics in the waste on a time-weighted,
annual average basis.
     The 10 ppmw is not an exemption from regulation;  it is intended only
as a way for EPA to divide the air regulations into phases.  The EPA is
deferring action on very-low-concentration streams (i.e., ones with less
than 10 ppmw total organic content) from the final rule today but will
evaluate and announce a decision later on whether to regulate these waste
streams.
4.3  EXEMPTIONS FOR SMALL SOLVENT RECOVERY OPERATIONS
     Comment:  Commenters AESP-00020, AESP-00017,  AESP-L0004, and AESP-
L0011 ask EPA to consider exemptions for small solvent recovery operations.
Specific recommendations from commenters included exempting: (!) batch
processes of 55 gal or less,  including drums or tanks  holding solvent and
drum storage facilities (AESP-L0004); (2) systems where waste is
                                     4-5

-------
TABLE 4.1  AIR STRIPPER DATA SUMMARY1'4
Aira
stripper
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Current
controls
Y
Y
N
N
N
. Y
Y
N
N
N
N
Y
N
N
N
Y
Y
N
Y
7
Y
N
N
Y
Y
N
?
•
Y
Y
N
N
N
Y
Y
Y
7
N
Y
• ' Y
Y
N
Organic'3
emissions,
Ib/h
5.1
0.5
0.17
1.17
0.5
0.15
0.42
2.31
0.05
1.85
0.72
0.3
0.01
0.4
0.01
0.0003
0.14
0.0066
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
30
320
0.8
Maximum
influent
organic
concentration,
ppmw
2,278
359
1,317
3.37
ND
3.964
0.59
0.72
6.77
0.39
10.67
4.37
3,688
0.03
0.04
0.19
ND
3.3
61.86
29.5
0.186
0.19
0.22
9.62
35.62
0.34
0.24
1.56
1.05
10.18
4.79
323.8
138.85
4,017
0.88
0.04
0.87
1.344
13,495
999
746.5
Average
influent
organic
concentration,
ppmw
113.9
17.95
65.85
0.1685
ND
0.1982
0.0295
0.036
0.3385
0.0195
0.5335
0.2185
184.4
0.0015
0.002
0.0095
ND
0.165
3.093
1.475
0.0093
0.0095
0.011
0.481
1.781
0.017
0.012
0.078
0.0525
0.509
0.2395
16.19
6.9425
200.85
0.044
0.002
0.0435
0.0672
8,100
750
440
Influent
flow rate,
gal /mi n
300
53
50
400
259
100
400
2,847
90
3,704
, 1,400
600
80
6,250
500
40
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
31
637
2
                                                (continued)
                  4-6

-------
                      TABLE 4.1  AIR STRIPPER DATA SUMMARY*'4

Aira
stripper
42
43
44
45

Current
controls
Y
N
Y
Y

Organ ic&
emissions,
Ib/h
0.01
36
8.5
0.39
Maximum
influent
organic
concentration,
ppmw
99
30.7
19.8
9.9
Average
influent
organic
concentration,
ppmw
55
18.6
12
6

Influent
flow rate,
gal/min
18
2,353
865
79
ND = No data available.

aAir strippers numbered 1 through 38 are located at CERCLA remedial  action sites;
 those numbered 39 through 45 are located at RCRA hazardous waste TSDF sites.
     those units that currently have controls, the emission value represents the
 emissions with the controls in place.

cln some cases, average concentrations were estimated using a correlation between
 maximum and average concentrations for the available data.
                                        4-7

-------
accumulated but pumped for a few hours each month; and (3) small (e.g.,
5 gal) sumps containing 2 to 4 gal  of solvent each (AESP-00017).
     Response:  As proposed, all WSTF/TSDF would have to install control
devices on all condenser and other process vents regardless of emission
levels.  Through reanalysis of the WSTF/TSDF air quality and health impacts
using updated model unit, emission rate, and facility throughput data, EPA
has determined that a minimum allowable process vent emission rate
applicable to all facilities is appropriate.  The health impact analysis
conducted to assess the magnitude of cancer risk from exposure to air
emissions from WSTF process vent emissions shows for both MIR and nation-
wide annual cancer incidence that there is an emission level from process
vents that is protective of human health and the environment, and below
which additional emission control results in no appreciable reduction  in
health risks.
     The risk analysis results indicate that provision of a "small -
facility" cutoff of 1.4 kg/h (3 Ib/h) and 2.8 Mg/yr (3.1 ton/yr) to total
facility process vent emissions provides essentially the same level of
protection for public health as does the proposed standard with coverage
including all WSTF/TSDF nationwide.  This cutoff effectively eliminates
concerns over controlling very small sources.  Control of facilities with
process vent emissions less than the emission rate limit would not result
in further reductions of either cancer risk or annual incidence on a
nationwide basis.  Facilities with organic emissions from process vents
that do not exceed these emission rates will not have to install controls.
A detailed discussion of the determination of the small-facility cutoff  is
presented in the response to comment 7.1.
4.4  TOTALLY ENCLOSED TREATMENT FACILITY EXEMPTION
     Comment:  Commenter AESP-00026 disagrees with EPA's interpretation
that the definition of "totally enclosed treatment units"  (which are exempt
from regulation) may  in certain circumstances include onsite treatment
units that use engineered controls to prevent the release of emissions.
The-commenter states  that onsite treatment facilities associated with
process equipment have the  same potential for emissions as other sources
not exempted by  the proposed regulation.  Commenter AESP-L0015, however,
agrees with EPA's proposed  interpretation.
                                     4-8

-------
     Response:  The EPA notes that this rule does not create or modify any
exemption for totally enclosed treatment facilities; rather, the existing
definition of and exemption for totally enclosed treatment facilities (40
CFR 264.1(g)(5) and 40 CFR 265.1(c)(9)) remain in effect, and existing
regulatory interpretations remain in effect as well.  As presented in the
preamble to the proposed rule, under 40 CFR 264.1(g)(5) and 40 CFR
265.1(c)(9), totally enclosed treatment facilities are exempt from RCRA
regulation.  A "totally enclosed treatment facility" is a facility treating
hazardous waste that is "directly connected to an industrial production
process and which is constructed and operated in a manner which prevents
the release of any hazardous waste or constituent thereof into the
environment during treatment" (40 CFR 260.10).  Treatment facilities
located off the site of generation are not directly connected to an
                                       *
industrial process.  Thus, commercial waste treatment facilities with
equipment affected by the final standards, such as solvent reclamation
facilities, by definition ordinarily would not be totally enclosed.  In
addition, storage facilities, disposal facilities, and ancillary equipment
not used for treating hazardous waste do not fall within the definition of
a totally enclosed treatment facility.
     The EPA believes that many onsite treatment facilities also are not
totally enclosed.  Process emissions from distillation columns and other
treatment technologies generally are designed to release air emissions of
the hazardous waste or constituent into the air environment.  Therefore, by
definition, these on-site technologies are generally not totally enclosed.
(See 45 FR 33218, May 19, 1980 [no constituents released to air during
treatment].)
     There are two important characteristics that define a totally enclosed
treatment facility.  The key characteristic of a totally enclosed treatment
facility is that it does not release any hazardous waste or constituent of
hazardous waste into the environment during treatment.  Thus, if a facility
leaks,  spills, or discharges waste or waste constituents, or emits wastes
or waste constituents into the air during treatment, it is not a totally
enclosed treatment facility within the meaning of these regulations.
     The EPA agrees with Commenter AESP-00026 that onsite treatment facili-
ties associated with process equipment generally are designed to release
                                     4-9

-------
air emissions and thus are not "totally enclosed."  The EPA specifically
stated this in the preamble to the proposed rule.  To be considered
"totally enclosed," units must meet the test of preventing the release of
any hazardous constituent from the unit not only on a routine basis but
also during a process upset.  Thus, the risks from these units are expected
to be less than from units that are not totally enclosed.
     This rule does not create or modify any exemption for totally enclosed
treatment facilities; rather, the existing definition of an exemption for
totally enclosed treatment facilities remains in effect, and existing
regulatory interpretations remain in effect as well.  Although the preamble
to the proposed rule repeated the existing definition, it also contained a
request for comments on an interpretation of the totally enclosed facility
exemption whereby the "use of effective controls such as those required by
the proposed standard" would meet the criteria of 40 CFR 260.10.  Upon
consideration of the comments, the Agency has determined that this
interpretation would have conflicted with the regulatory definition and
previous interpretations of the exemption and therefore has decided to
withdraw it.
4.5  EXEMPTION FOR 90-DAY STORAGE
     Comment:  Comnienter AESP-L0018 objects to the exemption for AV storing
or treating hazardous wastes that are emptied every 90 days and meet the
tank standards of 40 CFR 262.34, provided a permit is not needed.  Accord-
ing to the commenter, this exclusion is justified based on cost or tech-
nological grounds, but not on risk as RCRA requires.  The commenter states
that this exclusion will increase the use of the 90-day storage exemption
and the resultant air emissions.
     Response:   In 40 CFR 270, hazardous waste generators who accumulate
waste onsite  in  containers or tanks for less than the time periods provided
in Section 262.34 are specifically excluded from RCRA permitting require-
ments.  In order to qualify  for the exclusions in Section 262.34, genera-
tors who accumulate hazardous waste onsite for up to  90 days must comply
with 40 CFR 265  Subpart  I or J  (depending on whether  the waste  is accumu-
lated in containers or tanks) and with other requirements specified in
Section 262.34.  Small quantity generators  (i.e., generators who generate
more than  100  kg but  less than  1,000 kg per calendar  month) are allowed to
                                    4-10

-------
accumulate waste on-site for up to 180 days or, if they must ship waste
off-site for a distance of 200 mi or more, and if they meet certain other
requirements set out in Section 262.34, for up to 270 days.
     The promulgated regulation does not create a new exemption for 90-day
accumulation; nor does it modify the existing regulation.  As the commenter
notes, EPA is considering what changes (if any) should be made to Section
262.34 (the "90-day rule") under a separate rulemaking (51 PR 25487, July
14, 1986).  As part of that effort, the Agency currently is evaluating
whether air emissions from these and other accumulator tanks at the
generator site should be subject to additional control requirements.
Preliminary analysis indicates that 90-day tanks and containers may have
significant organic air emissions; consequently, as part of the second
phase of TSDF air emission regulations, the Agency intends to propose to
modify the exemption to require that 90-day tanks meet the control
requirements of the Phase I and II standards.  (The multiphased standards
development approach for regulating organic air emissions is discussed in
response to comment 11.5.2.)  Until a final decision is made on regulating
the emissions from these units, they will not be subject to additional
controls.  However, EPA does not believe that more generators will use the
90-day exemption if air emission controls are not imposed.  Those genera-
tors who are eligible for inclusion under Section 262.34 are probably
already taking advantage of the provision now by storing their hazardous
wastes for less than 90 days.
4.6  LEAK DETECTION AND REPAIR (LDAR) PROGRAM
     4.6.1  Exemption for Light/Heavy Liquids
     Comment:  Several commenters criticize the incorporation of the
benzene NESHAP because of differences in scope from the SOCMI NSPS in that:
(1) the NSPS distinguishes between light and heavy liquids and the proposed
standards based on the benzene NESHAP do not; (2) the NSPS does not require
testing of all SOCMI units because process fluid vapor pressure is the
overriding consideration in predicting leak frequencies and leak rates (the
proposed standards incorporating the NESHAP do not recognize vapor pressure
and require testing of all SOCMI units);  and (3) the NSPS exempts facili-
ties from routine fugitive emission monitoring, inspection, and repair
provisions if a heavy-liquid product from a heavy-liquid raw material is
                                    4-11

-------
produced and limits monitoring of equipment in heavy-liquid service only to
where there is evidence of a potential  leak.
     Response:  The EPA agrees with the commenters that the provisions for
light and heavy liquids in the SOCMI NSPS should be incorporated in the
Section 3004(n) standards, even though  the Subpart V NESHAP does not
contain the distinction.  No distinction was made for the benzene NESHAP
because benzene is a light liquid.  By  their nature, heavy liquids exhibit
much lower volatilities than do light liquids, and because equipment leak
rates and emissions have been shown to  vary with stream volatility,
emissions for heavy liquids are less than those for lighter,  more volatile
ones.  For example, EPA analyses have determined that the emission rate for
a valve in heavy-liquid service is more than 30 times less than the
emission rate for a valve in light-liquid service.  Therefore,  a routine
LDAR monthly inspection is not necessary for equipment in heavy-liquid
service.
     Thus, the final regulations have been changed to incorporate the
light-liquid service provisions for pumps and valves.  Equipment is in
light-liquid service if the vapor pressure of one or more of the components
is greater than 0.3 kPa at 20 °C, if the total concentration of the pure
components having a vapor pressure greater than 0.3 KPa at 20 °C is equal
to or greater than 20 percent by weight, and if the fluid is a liquid at
operating conditions.  The 0.3 kPa vapor pressure criterion is based on
fugitive emission data gathered in various EPA and industry studies.
Equipment processing organic liquids with vapor pressures above 0.3 kPa
leaked at significantly higher rates and frequencies than did equipment
processing streams with vapor pressures below 0.3 kPa.  Therefore, EPA
elected to exempt valves processing lower vapor pressure substances (i.e.,
heavy liquids) from the routine LDAR requirements of the standards.  That
is,  facilities are exempted from routine fugitive emission monitoring,
inspection, and repair provisions for equipment if a heavy liquid is
processed.   In addition, instrument monitoring of equipment in heavy-
liquid service is required only where there is evidence by visual, audible,
olfactory or any other detection method of a potential leak.
                                    4-12

-------
4.6.2  Exemption for Small Facilities
     Comment:  Several commenters ask EPA to consider exemptions from
fugitive emission monitoring for small facilities based on volume, as was
done in the benzene NESHAP and the SOCMI NSPS  (AESP-00018, AESP-00034, and
AESP-00002), based on an emission threshold (AESP-L0011), based on a
product applicability threshold or equipment component count  (AESP-L0001
and AESP-L0025), or based on equipment size, such as for small tanks and
pumps (AESP-00026).  In support, the commenters point to similar exemptions
in the CAA rules, but not in the proposed standards, and state that costs
are higher for control of their very small operations compared to the
emission reduction that would be achieved.
     Response:  The commenters suggest that EPA consider other exemptions
for fugitive emission monitoring that are applied in the benzene NESHAP or
  t
SOCMI NSPS (e.g., small facilities with the design capacity to produce less
than 1,000 Mg/yr).  The EPA also recognizes that estimated emissions and
health risks from small facilities should be considered in the final rules.
With regard to the SOCMI NSPS small facility exemption, the cutoff was
based on a cost-effectiveness analysis.  Under Section 112 of the CAA, EPA
may exempt units for which costs of the standards are unreasonably high in
comparison to the emission reduction achievable.
     Under RCRA,  the statutory criterion is to protect human health and the
environment.  Therefore, any cutoff for RCRA standards must be risk-based.
Cost effectiveness is only a relevant factor under RCRA in selecting among
equally effective control technologies (45 FR 33089, May 19, 1980).
     In addition, data from fugitive emissions tests do not show any
definite relationship between emissions and process throughput above some
minimal quantities.   Fugitive emissions are more closely related to the
equipment count in a facility than to the facility size or process through-
put.  Furthermore,  each type of equipment has a different equipment leak
fugitive emission rate.  Fugitive emissions are also a function of the
chemical  characteristics of the hazardous wastes being handled.
     Typically, TSDF have a variety of hazardous waste management processes
(e.g.,  container storage, tank storage,  treatment tanks,  incinerators,
injection wells,  and terminal loading operations) located at the same
                                    4-13

-------
facility, all of which have associated pumps,  valves,  sampling connections,
etc., and therefore,  fugitive emissions from equipment leaks.  Also,
several different types of hazardous waste typically are managed at a
facility.  Because of the various factors affecting facility fugitive
emissions from equipment leaks (e.g., equipment leak emissions are a
function of equipment component counts rather than throughput of the unit),
it would be very difficult to determine a small facility exemption based on
risk and expressed as volume throughput.  For these reasons, EPA did not
include exemptions for fugitive emission monitoring such as those applied
in the benzene NESHAP or SOCMI NSPS (i.e., small process units with the
design capacity to produce [process] less than 1,000 Mg/yr).
4.6.3  Exemption for Vacuum Systems
     Comment:  Commenter AESP-L0002 notes that the benzene NESHAP allows
exemptions for vacuum systems, systems with no emissions, and systems for
which the leakage rate is demonstrated to be below 2 percent.
     Response:  The EPA has included in the final TSDF standards (Sections
264.1050(e) and 265.1050(e)) the exemption for equipment "in vacuum
service" found in the benzene NESHAP (40 CFR Subpart V, 61.242-1).  Also
included are the identification requirements contained in the regulation.
"In vacuum service" means that equipment is operating at an internal
pressure that is at least 5 kPa below ambient pressure.  The EPA has
concluded that it is unnecessary to cover equipment (e.g., pumps, valves,
compressors, and closed vent systems) "in vacuum service" because such
equipment have little if any potential for emissions and therefore, do not
pose a threat to human health and the environment.  Therefore these
equipment have been excluded from the equipment leak fugitive emission
requirements.
     The proposed standards stated that owners and operators of facilities
subject to the provisions of the proposed rule shall comply with the
requirements of 40 CFR Part 61, Subpart V, except as provided in the rule
itself.  The provisions of the proposed rule did not exclude Section
61..243, alternative standards for valves  in VHAP service, and the alterna-
tive standards have been incorporated as Sections 264.1061, 264.1062,
265.1061, and 265.1062 of the final  rule.  Therefore, an owner or operator
may elect to have all valves within  a TSDF hazardous waste  management unit
                                    4-14

-------
comply with an alternative standard that allows no greater than 2 percent
of the valves to leak (Sections 264.1061 and 265.1061) or may elect for all
valves within a hazardous waste management unit to comply with one of the
alternative work practices specified in paragraphs (b)(2) and (3) of
Sections 264.1062 and 265.1062.
4.6.4  Flanges and Pressure Relief Devices
     Comment:  Commenter AESP-L0016 recommends that any release from
pressure relief devices in gas service be directed to control equipment at
least equal in performance to those for other process sources or an alter-
native means provided to prevent an uncontrolled discharge.  According to
the commenter, rupture discs or closed-vent systems restrict small leaks
but not major releases; a closed-vent system connected to a control device
is needed to capture releases.  The commenter also states that pipeline
flanges can be controlled by tightening bolts.-  The commenter concluded
that EPA has provided no data to support exempting flanges and pressure
relief devices in liquid service from LDAR requirements.  The EPA should
not rely on operators to see, hear, or smell leaks from these-equipment.
     Response:  Pressure relief devices allow the release of vapors or
liquids until system pressure is reduced to the normal operating level.
The standards are geared toward control of routine low-level equipment
leaks that may occur independent of emergency discharges.  Pressure relief
discharges are an entirely different emission source than are equipment
leaks or process vents and were not covered in the original equipment leak
standards under the CAA.  The new Subpart BB rules require that pressure
relief devices in gas service be tested annually by Method 21 (and within 5
calendar days of any relief discharge) to ensure that the device is
maintained at no detectable emissions (i.e., less than 500 ppm above
background, as measured by EPA Reference Method 21) by means of a rupture
disc.  The EPA does not believe that additional measures are warranted.  In
addition, because a pressure discharge constitutes a process upset that in
many cases can lead to hazardous waste management unit downtime and might
also pose a risk to workers, a facility has the incentive to minimize the
occurrence of these events.
     The frequency, duration, and air emissions associated with such
emergency discharges at TSDF waste management units cannot currently be
                                    4-15

-------
estimated with any certainty on a nationwide basis.  However,  if a pressure
discharge does occur, records and reports maintained at the site (under
Sections 264.1054, 264.1064, 265.1054,  and 265.1064 of Subpart BB) will
indicate the frequency of such discharges, the estimated volume of excess
emissions, and other relevant information.  If pressure discharges appear
to be a problem at any facility, the RCRA permitting system provides State
or EPA permit writers the flexibility to require closed-vent systems for
these discharges on  a site-specific basis.
     The LDAR program transferred from the CAA standards does not exempt
pressure relief devices and flanges in heavy-liquid service, but requires
formal monitoring of these  sources if operators see, smell, or hear dis-
charges.  The EPA considers that this is the most practical way to manage
these sources.  Although scheduled routine maintenance  (e.g., tightening
bolts on pipeline flanges)  may be a way of avoiding the need for formal
monitoring,  it may not be a successful method for all sites in eliminating
leaks due to the  numerous variables affecting leak occurrence.  For
example, flanges  may become fugitive emission sources when  leakage occurs
due to  improperly chosen gaskets, poorly  assembled flanges, or thermal
stress  resulting  in  the deformation of the seal between the flange faces.
In these situations, operators will be able to detect such  leaks  by sight,
smell,  or sound.  Support for  this approach was presented  and evaluated  in
developing  background and support for several CAA  rulemakings  (EPA-450/3-
83-016b,  EPA-450/3-80-033b, and  EPA-450/3-81-015b).
4.6.5   Directed Maintenance.  Leak Definition, and  Petition  for
        Reconsideration of the  Benzene Standards
      Comment:  Commenter AESP-L0016  includes  in their  comments  their
petition  for reconsideration  of  the  benzene fugitive  standards.   Although
the  comments are  directed at  the CAA  standards, they  contain  numerous
suggestions for modifications  to the  LDAR program  that  have previously been
raised  with EPA.   Also  included in  their comments,  however,  are recommenda-
tions regarding  directed maintenance  and the  10,000-ppm leak  definition
that are specific to the  proposed standards/   On  the question  of directed
maintenance, the  commenter  believes  that the  LDAR  program should  require
preventive  maintenance,  such as the periodic  replacement  of valve packings
                                     4-16

-------
before waiting for the valve to fail.  In support, the commenter argues
that EPA's own data show that directed maintenance could reduce leaks from
valves to below 10,000 ppm.  The commenter also criticizes the 10,000-ppm
leak definition as being too high and states that EPA must consider the
level in terms of the health effects.
     Response:  The key criterion for selecting a leak definition is the
overall mass emission reduction demonstrated to be achievable.  The EPA has
not concluded that an effective lower leak definition has been demonstrat-
ed.  Most data (EPA-450/3-82-010) on leak repair effectiveness have applied
10,000 ppm as the leak definition and therefore do not indicate the effec-
tiveness of repair for leak definitions between 1,000 and 10,000 ppm.  Even
though limited data between these values were collected for support of CAA
standards, they are not sufficient to support a leak definition below
10,000 ppm.  A net increase in mass emissions might result if higher
concentration levels result from attempts to repair a valve with a screen-
ing value between 1,000 and 10,000 ppmv.  Data are insufficient to deter-
mine precisely at what screening value maintenance efforts begin to result
in increased emissions.
     As the commenter noted, although there is some evidence that directed
maintenance* is more effective, available data are insufficient to serve as
a basis for requiring directed maintenance for all sources.
     The EPA believes that there is only a small potential emission reduc-
tion for leaks having total organic concentrations between 1,000 and
10,000 ppmv.  Therefore,  using a lower leak definition would not increase
emission reductions significantly, even if EPA judged that repair was
effective for leaks of 1,000 ppmv.  In the proposal  BID for the petroleum
refinery fugitive emission NSPS (Docket Item II-A-43, p.  4-8), there is a
comparison of the percentage of total mass emissions affected by selecting
a 10,000-ppmv leak definition over a 1,000-ppmv leak definition.  These
     *In "directed maintenance" efforts, the tightening of the packing is
monitored simultaneously and is continued only to the extent that it
reduces emissions.  In contrast, "undirected" repair means repairs such as
tightening valve packings without simultaneously monitoring the result to
determine if the repair is increasing or decreasing emissions.
                                    4-17

-------
percentages represent maximum theoretical emission reductions that could be
expected if the sources were instantaneously repaired to a zero leak rate
and no new leaks occurred.  For valves in gas service, the estimated
decrease is only 1 percent.  This small potential decrease in emissions may
be more than offset by counterproductive attempts to repair sources with
low leaks.
     The EPA's rationale for selecting the 10,000-ppmv leak definition and
for not requiring directed maintenance under the CAA LDAR program also has
been discussed in the proposal and promulgation BIDs for benzene emissions
from coke by-product recovery plants (EPA-450/3-83-016 a and b),  for SOCMI
fugitive emissions (EPA-450/3-80-033 a and b),  for petroleum refinery  .
fugitive emissions (EPA-450/3-81-015 a and b),  and for benzene fugitive
emissions (EPA-450/3-80-032 a and b).  (See also the "Response to Public
Comments on EPA's Listing of Benzene Under Section 112" [EPA-450/5-82-003]
"Fugitive Emission Sources of Organic Compounds—Additional Information on
Emissions, Emission Reductions, and Costs" [EPA-450/3-82-010],  and EPA's
"Response to Petition for Reconsideration" [50 FR 34144, August 23, 1985].)
     In summary, EPA cannot conclude that additional emission reductions
could be achieved by reducing the leak definition from 10,000
to 1,000 ppmv.  In the absence of data supporting the repair potential
associated with leak definitions such as 1,000 ppmv, EPA is selecting the
clearly demonstrated leak definition of  10,000 ppmv  instead of a lower
level.  •
     The commenter also criticizes EPA for not reanalyzing the health
effects of the  10,000-ppmv level before  applying the  limit to TSDF under
RCRA.  Because  the CAA and RCRA are almost identical  in their recognition
of health risk  as the predominant decision factor, the EPA believes that
the leak definition has been  adequately  analyzed under the CAA and that
further evaluation is not needed prior to transferring it as part of the
LDAR program under RCRA.   It  must also be pointed out that transfer of the
CAA equipment  leak standards  is only the first phase of EPA's  regulatory
actions related to control of TSDF air emissions.   In this phase, EPA
transferred a  known technology to reduce emissions.   If new data show  that
a  lower leak definition is appropriate,  the  Agency will then consider
whether it  is  appropriate  to  change the  rules.
                                     4-18

-------
4.7  REFERENCES

1.   U.S. Environmental Protection Agency.  Air Stripping of Contaminated
     Water Sources Air Emissions and Controls.  Air Toxics Control Center.
     Research Triangle Park, NC.  July 20, 1987.

2.   U.S. Environmental Protection Agency.  Air Strippers and Their
     Emission Controls at Superfund Sites.  Hazardous Waste Engineering
     Research Laboratory.  Cincinnati, OH.  EPA/600/D-88/153.  August 1988.

3.   Memorandum from Vickery, James, EPA/OSWER, to Susan Wyatt,
     EPA/OAQPS/ESD.  May 17, 1989.  Ground water concentration data for 41
     Superfund sites.

4.   Memorandum from Kong,  E. J., and Coy, David., RTI, to Accelerated Rule
     Project File.  May 9,  1989.  Data and information for air strippers
     treating hazardous wastes.
                                    4-19

-------
                         5.0  ENVIRONMENTAL  IMPACTS

5.1  DATA, ANALYSES, AND METHODOLOGY
     Comment:  Several commenters criticize  the lack of data and informa-
tion supporting the impact analyses.  Numerous commenters criticized the
emission estimates because:  (1) no actual data from operating facilities
were used (AESP-L0001 and AESP-00028); (2) the estimates are not supported
by any technical data base, and EPA has used out-of-date assumptions
regarding the magnitude of the emissions  (AESP-L0001, AESP-00019, and
AESP-L0003); and (3) the role of State controls has not been considered
(AESP-L0001, AESP-L0017, and AESP-00016).  Commenters AESP-00002 and
AESP-00035 add that EPA has not shown that air emissions from WSTF present
a human health or environmental threat because the analysis is based on
extrapolated data on the number of WSTF nationwide, the number of WSTF that
are TSDF and are thus affected by the rule,  and the number of WSTF that are
commercial recycling facilities.  Extrapolation also was used to estimate
equipment leak emissions from WSTF based on  the benzene data; according to
the comments, EPA should base the analyses on factual knowledge and not on
extrapolated data.   Commenters AESP-L0015 and AESP-L0017 state that EPA has
not justified applying the same control technology for all of the sub-
stances in recycled waste streams.
     Response:  In  response to these comments,  EPA reviewed all available
WSTF site-specific  data, including vent emission measurements from primary
and secondary condensers,  condensers vented to tanks, and vacuum distilla-
tion vents.   In addition,  emission data were requested in the proposal
preamble and through Section 3007 questionnaires mailed to nine small  and
nine large TSDF after proposal.   The responses  to these information
requests have provided considerable data for inclusion in the post-proposal
analysis.   For example,  one respondent supplied data on emissions from
                                     5-1

-------
batch stills at four different facilities.  The data reviewed for WSTF
process vents represented the most prevalent configurations for venting the
process emissions subject to this regulation and led to the development of
wide ranges of flow rates and emission rates.. Based on all this informa-
tion, EPA has revised both model units and WSTF emission factors for
inclusion in the impacts analysis.  The revised model unit emission factors
developed from the emission test data are discussed in response to comment
5.2.
     State regulations also were reviewed to help better establish baseline
emissions.  In general, PSD pe-rmit requirements apply to a new or modified
source with the potential to emit more than 40 tons/yr of VOC.  For sources
where CTG are not available, such as TSDF, VOC standards applicable in the
State or a particular attainment or nonattainment area could be applied. (A
total of 37 States have ozone nonattainment areas.)  However, from the
review of these standards-, it does not appear that there would be general
control requirements for TSDF process vents.  No CTG have been developed
for WSTF/TSDF sources that would generally have potential VOC emissions of
less than 40 tons/yr.  Responses to the survey of nine large TSDF also do
not indicate control requirements for process vents.  Several of the
facilities that were requested to provide information reported requirements
for obtaining air contaminant source operating permits, but they reported
no permit requirements for controlling process vent emissions.  Therefore,
baseline emissions should be related to the efficiency of existing primary
condensers and to the existence and efficiency of vent control devices.
These factors are reflected in the revised range of emission estimates
based on site-specific emission data.
     As is discussed in response to comment 3.2.1, EPA has developed an
industry profile using results of the 1986 Screener Survey.  The Screener
Survey data represent all of the TSDF active  in 1985 with  interim status or
final RCRA permits, which totalled about 3,000 facilities.  Review of the
Screener Survey data shows that a total of 448 facilities  reported some
type of solvent recovery operation at the facility.  Operations categorized
as solvent recovery for the survey were those recovering !he organic
constituents from waste for the primary purpose of recovering organic
                                     5-2

-------
compounds for reuse.  Solvent recovery includes units that are referred to
as batch distillation, fractionation, or steam stripping units.  The EPA
used these actual facility counts together with their reported 1985 waste
solvent throughputs as the basis for the post-proposal impacts analysis.
     The constituents selected for analyses of control technologies are
considered to be representative of the industry, based on a review of rele-
vant information and literature, including a survey of member companies
submitted by the NASR, 23 site-specific plant visit reports, responses to
17 EPA questionnaires from small and large facilities (Docket No. F-90-
AESF-FFFFF, items S0017, S0018, S0019, S0020, S0022, S0024 through S0026,
S0028 through S0036, and S0038 through S0041), the Industrial Studies Data
Base (ISDB) (Docket No. F-90-AESF-FFFFF,  item S0027), and a data base
created by the Illinois EPA (Docket No. F-90-AESF-FFFFF, item S0045).  The
NASR survey provided information on the types of solvents most frequently
recycled at member facilities; the site-specific information and EPA survey
responses included waste compos.ition data.  The ISDB is a compilation of
data from ongoing, in-depth surveys conducted by EPA's OSW of designated
industries that are major hazardous waste generators.  The Illinois EPA
data base contains information from about 35,000 permit applications.  .
Generators must submit one application, for each hazardous and special
nonhazardous waste stream managed in the State of Illinois.  The ISDB and
Illinois EPA data bases are discussed in more detail in Appendix D.
     Each of these data bases contains waste stream characterization data
for numerous generic spent solvent waste streams (EPA Hazardous Wastes
F001-F005) and D001 wastes (ignitable), which information from the Screener
Survey indicates are also recycled.  The constituents that were found most
frequently in the site-specific data and their frequency of occurrence
(percent) include toluene (60 percent), ethyl benzene (44 percent), 1,1,1-
trichloroethane (41 percent),  xylene (37 percent),  isopropyl alcohol
(37 percent),  and trichloroethylene (33 percent).  Halogenated compounds
were present in more than 60 percent of the waste streams for which
composition data were available.  It should also be noted that the list of
major constituents present in the waste streams examined agrees well with
the list of the types of solvents recycled at the facilities surveyed by
the NASR.
                                     5-3

-------
     Three of the four solvents used in the analyses of control technolo-
gies are the same as were used at proposal (toluene, MEK, and 1,1,1 trich-
loroethane).  Methylene chloride was added to the list of constituents to
provide a broader, more comprehensive range of solvent volatilities for the
analyses.  Methylene chloride is at the lower end of the boiling point
range and is therefore more difficult to condense than are most other
solvents.  Therefore, the technical feasibility and costs of applying the
recommended control techniques were evaluated for constituents representing
the range of characteristics of commonly recycled solvents at TSDF.
5.2  MODEL PLANT EMISSION ESTIMATES
     Comment:  Several commenters believe that EPA has overestimated model
plant emissions and question the methodology and supporting assumptions for
the estimate of 86 Mg/yr (95 ton/yr) of VO per facility and the 8 Gg/yr
(882,000 ton/yr) typical reclamation rate.  Commenter AESP-00019 criticizes
the reclamation rate but provides no supporting data.  Commenters AESP-
L0001 and AESP-L0017 object to the assumption that one-half of the WSTF
will have process vent emissions of about 3.2 kg/h (7 Ib/h) and the other
half will have emissions of about 34 kg/h (75 Ib/h) because the effect of
controls already applied is not considered.  Commenters AESP-00009, AESP-
L0001, and AESP-00028 ask why no actual data were used as the basis of the
environmental impact estimates and what facilities were studied to form the
basis of these estimates.  In support of their contention that emissions
are overestimated, Commenter AESP-00009 cites a test at their Menomonee
Falls, Wisconsin, facility on the product receiving tank, waste feed tank,
recirculation tank, and bottoms receiving tank that shows a total emission
rate of 0.805 Mg/yr  (0.887 ton/yr) if operations were conducted 24 h/day
for 365 day/yr.  This measurement reflects one of three runs in a typically
large Luwa still sized at 4 w£ (43 ft^);  the material run was 50 percent
toluene.  Commenter AESP-00016 also cites air monitoring results at one
facility that indicate outdoor concentrations up to 3 ppm across the street
from the facility.
     Commenter AESP-L0001 also states that it is not appropriate to apply
to TSDF the fugitive emission factors that were developed to estimate leaks
from a typical hydrocarbon plant because they do not relate to the design,
                                     5-4

-------
operating conditions, maintenance practices, or controls associated with
the processing of toxic chemical wastes.
     Response:  With regard to the model plant annual production rate, the
industry profile referred to in response to comments 3.2.1 and 5.1 includes
a frequency distribution of the waste processed in gallons during 1985.  Of
448 facilities in the Screener Survey reporting solvent recovery by batch
distillation, fractionation, or steam stripping operations involving some
form of waste at the facility, 365 reported the total quantity of waste
(all types, including waste categories other than solvents) recycled in
1985.  The median facility throughput was slightly more than 189,000 L
(50,000 gal); and the mean throughput was about 4.5 million L/yr (1.2 mil-
lion gal/yr).  These figures exclude one plant that reported more than
439 million L (116 million gal) of wastes recycled through its solvent
recovery units in 1985 because the waste volume reported was almost an
order of magnitude greater than the volume reported by any other facility.
     Based on the industry profile frequency distribution, three sizes of
model units (small, medium, and large) were defined to facilitate the anal-
yses of control  costs, emission reductions, health risks, and economic
impacts.  Table 5-1 presents the model unit parameters, including general
plant operation characteristics and vent stream emission characteristics.
The bases for the model unit parameters are presented in Docket No. F-90-
AESF-FFFFF, item S0037.  AS can be seen in Table 5-1, for each model unit,
several emission rates were chosen.  The organic emission rates are based
on emission source testing conducted for EPA and summarized in Docket No.
F-90-AESF-FFFFF, item S0021.  The organic emission rates for primary
condensers varied from a few hundredths of a kg (Ib) to nearly 4.5 kg/h  (10
Ib/h) with six of the nine measurements less than 0.45 kg/h (1 Ib/h).  The
two secondary condensers tested showed emission rates of 0.9 and 2.3 kg/h
(2 and 5 Ib/h).   The analysis of emission factors is discussed further in
the response to comment 5.3.
     The test reports reviewed include the test cited by Commenter
AESP-00009; however, the emission rate of 0.805 Mg/yr (0.887 ton/yr) is
derived from ambient monitoring results; therefore, it does not represent a
process vent emission rate and was not used in developing the model unit
                                     5-5

-------
                                           TABLE B-l.  MODEL UNIT:  PROCESS  VENT EMISSIONS
WSTF modal units
Item
General plant operation
Waste solvent
Throughput8
Mg/yr
gal/yr
gal/hr
L/m
Large


8,000
2.5 x 10s
600
38
Med i urn


160
60 x 103
24
1.5
Small


32
10 x 103
4.8
0.3
    Operating hours^                     4,160                                2,080                               2,080

    Process vent stream*:    L Toluene  (110  °C)                 1. Toluene (110 °C)                  1. Toluene (110 °C)
       cases                2. MEK  (79  <>C)                      2. MEK  (79 <>C)                       2. MEK (79 or.)
                            3. 1,1,1-Trichloroethane  (74 °C)    3. 1,1,1-Trichloroethane (74 °C)    3. 1,1,1-Trichloroethane  (74 °C)
                            4. Methylene  chloride  (40 <>C)       4. Methylene  chloride (40 <>C)       4. Me thy I one chloride  (40 °C)

  Vent stream emission
    characteristics
01  	
(^  Temperature**                         24 <>C                                24 <>C                               24 «C

    Flow rate®

       (scfm)                             8.3                                  1.2                                  0.6
       (L/s)                              3.9                                  0.6                                  0.3

    Emission factors*
     (Ib/hr)
(1)
(2)
(3)
(4)
(6)
0.4
4.2
10.6
—
—
—
0.17
0.42
1.20
6.00
—
—
0.08
0.24
1.00
  MEK = Methyl ethyl ketone.
  TO  = Total organics.

  aModel plant throughputs  are  based  on  information  contained  in  the Screener Survey.

  bHours of operation for the  large model  unit are the  same  as those used at proposal:  intermittent operation at 2 shifts per  day,
   5 days per week, and 62  weeks  per  year.   For the  other  model units,  operating hours recommended" are based on a review of
   facility-specific data collected in response to Section 3007 questionnaires;  only eight (8) operating times were reported.   The
   22 TSDF plant trip reports did not contain  any  information  regarding annual  hours of operation.
                                                                                                                          (continued)

-------
                                                     TABLE  5-1   (continued)
 constituents selected for analysis are considered  to  be  representative  of  the  industry;  selection  was a  subjective judgment
  based on review of relevant information and  literature that  included  the NASR  survey  as  well  as  numerous site-specific plant
  trip reports.   Three of the solvents are the  same  as  were  used  at  proposal.  Methylene chloride  was  added to the list of  sol-
  vents for analysis in order to provide a more realistic  range of volatilities  for  the condenser  analysis.  Methylene chloride is
  at the lower end of the boiling point rang* and  is therefore more  difficult  to condense  (worst-case).  Gas streams are assumed
  to consist of  organics and a noncondensible gas  (air).   The numbers  in  parentheses are boiling points.

 "Temperature is the same as at proposal;  a review of site-specific  data  confirms the appropriateness  of the value selected.

 •Flow rates are based on a review of site-specific  data from 16  units  in the  plant  visit  reports.   The large and medium model
  unit flow rates selected also agree with the  SOCMI Distillation NSPS  BID Cases 1 and  2;  Case  5 was used  in the proposal
  analysis.   The flow for the small  model  unit  is a  subjective judgment based  on the field data from WSTF.

 * Organic  emission factors are based on plant source test  data contained  in  the  site visit reports conducted for EPA by the
  Research Triangle Institute (RTI),  Engineering Science,  Radian, and other  contractors.   There were three groupings of test
1  results  that led to the selection  of the factors at 0.1, 2.6, and  7.0 g of organics per  kilogram of  solvent treated.   The 1.0-g
1  factor was added as an intermediate to cover  the wide range between the lower  two  groupings;  the 30.0-g  factor was added  to
  provide  an upper end value for the medium and small model  units to bring overall emissions  into  line with the actual  emission
  rates on a mass per time (kg/h)  basis and to  provide  an  additional high value  for  the cutoff  determination.   These values
  compare  reasonably well  to the range of  values used in the AP-42 emission  factor,  0.26 to 4.17 g/kg.

-------
emission rates.  Similarly,  the ambient monitoring results cited by Com-
menter AESP-00016 are not appropriate for defining process vent emission
factors.
     The EPA does not agree with the statement of Commenter AESP-L0001
concerning the appropriateness of applying fugitive emission factors to
TSDF that were developed to estimate leaks from a typical hydrocarbon
plant.  As is pointed out in the response to comment 3.2.7, the data used
in establishing the fugitive emission standards for TSDF are based on
emission and process data collected at a variety of petroleum refinery and
SOCMI operating units.  The EPA's IERL coordinated a study to develop
information on fugitive emissions in the SOCMI.  A total of 24 chemical
process units were tested; these data covered thousands of screened sources
(pumps, valves, flanges, etc.) and included process units handling such
chemicals as acetone/phenol, MEK, ethylene dichloride, 1,1,1 trichloro-
ethane, trichloethylene, and perchloroethylene (Docket No. F-90-AESF-FFFFF,
item S0009).  Refinery studies on fugitives also included tests on units
handling both toluene and xylene (Docket No. F-90-AESF-FFFFF, items S0001
through S0005).  These same chemicals are included in those listed by the
NASR as solvents commonly- recycled by member facilities and are found in
other sources of waste solvent constituent information, as described in the
response to comment 5..1.  The chemicals commonly recycled at TSDF are those
produced in SOCMI operating units and handled in petroleum refineries, and
the equipment involved in these  industries are typically the same (i.e.,
pumps, valves, etc.).  Therefore, it is reasonable to conclude that similar
design, operating conditions, maintenance practices, and controls are
applied and that the emissions associated with these chemicals and equip-
ment are similar.
5.3  FLOW RATE
     Comment:  Commenter AESP-00016 questions the application to WSTF of
the 12 L/s (26 scfm) primary condenser exhaust gas flow rate developed
initially for the conditions and materials in the chemical manufacturing
industry.  Recalculation of emissions using the measured, emission ratio of
1.7 g/kg in the EPA Source Assessment report  (EPA-600/2-78-004f) instead of
the SOCMI distillation data results in an overall rate of  1,300 Mg or  less
                                     5-8

-------
than 3 million Ib/yr.  Emissions from solvent recycling operations gener-
ally are low according to the commenter because most systems are closed.
     Response:  The EPA agrees that the flow rate of 12 L/s (26 scfm) is
not generally valid for application to all spent solvent recyclers.  As is
discussed in response to comment 5.2, small, medium, and large model units
were defined to facilitate the analyses of control costs, emission reduc-
tions, health risks, and economic impacts.  The flow rates specified for
the model units are 3.9, 0.6, and 0.3 L/s (8.3, 1.2, and 0.6 scfm) for the
large, medium, and small model units, respectively.  These are based on a
review of site-specific data from 16 units documented in site visit
reports.  The large and medium model WSTF/TSDF process vent unit flow rates
also agree with the SOCMI Distillation NSPS BID (see Docket No. F-90-AESF-
FFFFF, item S0013) concerning Cases 1 and 2 (for characterization of
distillation unit with low overhead gas flows); Case 5 (the average for
low, medium, and high flows) was used in the proposal analysis for the WSTF
standard impacts.
     The commenter1s primary concern seems to be with the upper limit
emission rates (based on SOCMI distillation data) used in the proposal
analysis.  In specifying the small, medium,  and large model units, revised
emission factors were developed.  The organic emission factors are based on
emission source test data from tests conducted for EPA.  There were three
groupings of emission factors:  one group from 0.02 to 0.18 g organics/kg
total organics (TO); another at 2.5 g/kg TO; and a third from 6.0 to 7.7
g/kg TO.  These groupings led to the selection of emission factors at 0.1,
2.5, and 7.0 g/kg of solvent treated.  A 1.0 g/kg factor was added as
intermediate to cover the wide range between the lower two groupings; this
allowed an additional value in the health risk and cost analyses and aided
in the cutoff determination for large- and medium-sized units.  A 30.0-g/kg
TO factor was used to provide on upper end value for the medium and small
model units and to provide an hourly emission rate (model unit hourly
emission rates can be derived using the throughput and operating hours
presented in Table 5-1)  comparable to the measured emission rates in terms
of mass per unit time (e.g., Ib/hr).  (Docket No. F-90-AESF-FFFFF, item
S0021.)  The revised range of emission factors is comparable to the range
                                     5-9

-------
of 0.26 to 4.17 g/kg used to derive the average AP-42 emission factor for
waste solvent reclamation of 1.7 g/kg.
     The minimum and maximum uncontrolled emission factors for the small,
medium, and large model units were applied to the frequency distribution of
waste throughput referred to in the response to comment 5.2 to generate
lower bound and upper bound nationwide estimates of uncontrolled process
vent emissions (see Appendix C).  The estimated nationwide uncontrolled
process vent organic emissions range from 300 to 8,100 Mg/yr (330 to 8,900
ton/yr).  The large range in uncontrolled process vent organic emissions is
due to the range of primary condensers' removal efficiencies and the use of
secondary condensers on some primary condenser vents.
5.4  ADJUSTMENT FOR -VOLATILITY
     Comment:  Commenter AESP-00034 asks why no adjustment was made to
emission factors, model units, or model facilities (and subsequent risk and
cost analyses) to account for volatility.  Several commenters (AESP-00035,
AESP-L0025, and AESP-00016) submit information on the volatility of materi-
als in the streams in support of this comment.  They point out that benzene
has a relatively high volatility of 8.3 kPa at 116 °C (1.2 psia at 60 °F)
or 22.7 kPa at 38 °C (3.3 psia at 100 °F).  However, the volatilities of
the three constituents used for the model facilities are toluene's 2.1 kPa
at 16 °C (0.3 psia at 60 °F), MEK's 8.3 kPa at 16 °C (1.2 psia at 60 °F),
and 1,1,1-trichloroethane's 11.0 kPa at 16 °C  (1.6 psia at 60 °F).  Accord-
ing to Commenters AESP-00030, AESP-L0025, AESP-00016, and AESP-L0017, risks
as well as emissions are overestimated by not  accounting for differences in
vapor pressure.  Commenters AESP-00030 and AESP-L0025 also state that EPA
should reestimate costs based on excluding nonvolatiles because only a'
marginal risk reduction is achievable at  a very high cost.
     Response:  The EPA agrees that the standards should take component
volatility into consideration.  As is stated  in the  response to comment
3.2.6, previous EPA and industry studies  have  shown  that the volatility  of
stream components, as  a process variable, does correlate with fugitive
emission rates.  Analyses of  the vapor pressures  and emission rates has
shown that substances  with vapor pressures of  0.3 kPa or higher had signif-
icantly higher emission rates  than did those with lower vapor pressures
(Docket No.  F-90-AESF-FFFFF,  items S0001  through  S0009).  This  result led
                                     5-10

-------
to the separation of equipment component emissions by service:  gas/vapor,
light liquid, and heavy liquid.  These classifications have been used in
most fugitive emission standards to effectively direct the major effort
toward equipment most likely to leak.  Therefore, the rules have been
revised to account for volatility.  For example, pumps and valves in heavy
liquid service must be monitored only if evidence of a potential leak is
found.  Light and heavy liquid service are based on the vapor pressure of
the components in the stream (less than 0.3 kPa at 20 °C defines a heavy
liquid).
     All of the constituents used in the WSTF model unit analysis,  repre-
senting the ranges of characteristics of commonly recycled solvents, are
light liquids to which the benzene and SOCMI fugitive emission factors are
applicable.  Therefore, the revised risk and cost analyses related to WSTF
equipment leak fugitive emissions are based on the fugitive emission
factors used in the proposal analyses, although a medium and a small model
unit have been added.  The analyses of risk and cost impacts on TSDF with
affected fugitive emission sources were revised to account for the differ-
ences in light and heavy liquids.  Nationwide fugitive emissions and con-
trol costs for TSDF fugitive sources were estimated using a computerized
model developed to aid in the analyses supporting the additional TSDF air
emission standards.  The computerized model, which is described in Appen-
dix D, was used to estimate TSDF fugitive emissions and control costs
because it can access the TSDF industry profile developed for the analysis
of the additional standards to identify facilities handling light liquid
hazardous wastes with greater than 10-percent organic content.  The model
also accounts for throughput in generating control cost estimates.   The
results of the revised risk and cost analyses are presented in Chapters 6.0
and 8.0, respectively.
5.5  NUMBER OF FACILITIES
     Comment:  Commenter AESP-00015 contends that the proposed rules will
affect more than the 100 WSTF estimated in the preamble because small sol-
vent recovery stills that are commonly part of industrial degreasing opera-
tions are not accounted for.
     Response:   The EPA agrees that the standards will affect more than the
100 WSTF.  At proposal,  EPA estimated the number of WSTF potentially
                                    5-11

-------
affected by the standards based on the nationwide volume of waste solvent
requiring management.  The EPA's estimate also included extension of the
proposed standards to approximately 1,300 TSDF (the midpoint of an
estimated range of 269 to 2,332 facilities) that handle hazardous waste
streams or derivatives with at least 10 percent organics.  Since proposal,
EPA has identified 448 onsite and offsite waste solvent recyclers.  The EPA
estimates that a total of about 1,400 onsite and offsite TSDF, including
TSDF that practice solvent recovery, manage light liquid hazardous waste
streams containing 10 percent or more total organics and would be subject
to LDAR requirements.
     The final standards contain a facility process vent emission rate
limit of 1.4 kg/h (3 Ib/h) and 2.8 Mg/yr (3.1 ton/yr) of organic emissions.
The EPA's estimates based on the high emission rates and 1985 waste solvent
throughput data indicate that about 45.5 percent of the WSTF identified in
the industry profile will have process vent emissions of less than 2.8
Mg/yr (3.1 ton/yr).  Therefore, it is anticipated that the small solvent
recovery stills to which the commenter refers would not be affected by the
process vent standards.
5.6  ADEQUACY OF DATA AND INFORMATION SUPPORTING THE IMPACT ANALYSES
     Comment:  Commenters AESP-00019 and AESP-00029 state that the proposed
rules should be limited to process vents for reclaimers because the BID
does not adequately address or evaluate the impacts from TSDF sources.
     Response:  At proposal, EPA concluded that the proposed standards for
WSTF could be broadened to cover all equipment handling TSDF hazardous
waste streams with at least 10 percent organics.  There is no technical
basis for limiting the proposed controls to equipment at only solvent
treatment operations.  The revised impact analyses conducted by EPA include
a facility-specific analysis for both WSTF (448 facilities) process vent
and associated fugitive (equipment leak) emissions, and for TSDF (1,428
facilities) fugitive emissions associated with those waste management
processes other than solvent recovery type processes that handle light
liquid hazardous waste streams with at least 10 percent organics.  The
revised impact estimates do not include estimated emissions or health risks
associated with process vents at TSDF not engaged in distillation,
                                    5-12

-------
fractionation,  thin-film evaporation,  solvent extraction,  or air or steam
stripping operations.   Data are insufficient at this time to provide such
estimates.  Therefore,  EPA is limiting the applicability of the process
vent standards; only process vents on  operations involved in solvent or
other organic chemical  reclamation or  waste treatment separation or removal
by distillation, fractionation, thin-film evaporation,  solvent extraction,
or air or steam stripping of hazardous wastes with total organics concen-
trations of at least 10 ppmw are subject to the promulgated rules.  Emis-
sion controls for other TSDF processes, area sources, and waste categories
are under development and are scheduled to be proposed in 1989 under a
separate rulemaking.  However, TSDF with valves and pumps containing or
contacting hazardous waste with at least 10 percent organics and otherwise
required to obtain a RCRA permit are covered by the LDAR requirements of
this standard.
                                    5-13

-------
                          6.0  HEALTH RISK IMPACTS

6.1  REPRESENTATIVENESS OF WASTE STREAMS
     Comment:  Commenters AESP-00016 and AESP-00034 object to the lack of
information on the composition of emissions and the analysis of only three
pollutants for WSTF and a fourth for TSDF.  Numerous commenters also
believe that the constituents analyzed do not fairly represent the material
in recycled streams.  Commenters AESP-00004 and AESP-00005 state that with-
in the solvent recycling industry the waste streams do not contain toxics
or volatiles, and treatment operations do not change the character of the
wastes.  In support, Commenter AESP-00016 submits an informal survey of the
materials handled by commercial recyclers that shows halogenated solvents
(including perch!oroethylene) but no listed carcinogens.  Commenter AESP-
00004 agrees, adding that these facilities do not produce emissions with
PCBs, dioxins, acids, reactives, pesticides, or other highly carcinogenic
compounds.
     Response:  At proposal, EPA believed that the types of streams in the
TSDF industry had similar characteristics to those found in the synthetic
organic chemical manufacturing industry  (SOCMI).  In an effort to provide
further information and data on the composition of waste streams treated at
WSTF/TSDF and the associated air emissions, EPA has conducted a review of
the available information regarding WSTF/TSDF waste streams constituents
and concentrations.  The EPA TSDF Waste  Characterization Data Base  (WCDB),
information  from numerous plant trip reports, and data from a limited
number of emission test reports were used to characterize waste streams
treated by WSTF and TSDF in terms of volatilities, concentrations of
components,  and unit risk factors   (Docket No. F-90-AESF-FFFFF, item
S0047).  The  selection of the  range of unit risk  factors used in the  risk
analysis is  discussed  in Section 6.2.
                                      6-1

-------
Plant-Specific Data
     Plant trip reports and emission test reports from site visits, con-
ducted in connection with the evaluation of organic removal through pre-
treatment processes, provided organic composition and concentration data
for 22 waste stream cases from 16 WSTF/TSDF.  The most obvious conclusion
to be reached from an examination of these stream data is that WSTF waste
streams are not readily characterized.  With regard to total organic con-
tent (TOC),  the examined WSTF waste streams, treated for solvent recovery,
varied from less than one (1) percent TOC to practically 100 percent TOC.
Of the 19 cases that reported a value for waste stream TOC content, five
(5) were  less than one (1) percent, six  (6) were from one  (1) to 25 per-
cent, (four) 4 were between 25 and 90 percent, and four  (4) were greater
than 90 percent TOC.  It is important to note that two of the feed waste
streams with TO concentrations of less than 1 percent generated overhead
vapor and condensate streams, when steam stripped, with  a TOC of greater
than 10 percent.  Therefore, a waste stream TOC is not a totally reliable
index or  indicator of vent streams that  exceed 10 percent total organics.
     In general, the stream contents were comprised of one or two  major
constituents with two or. more minor constituents.  The constituent that
showed up most frequently in the site-specific data was  toluene, which was
reported  as being present in about 55% of the streams examined.  Toluene
concentrations, however, were typically  less than one (1)  percent  but did
range up  to 50 percent of feed content.  Other major constituents  and their
percent occurrence  included MEK  (36%), acetone  (32%), xylene  (27%), ethyl
benzene  (36%), isopropyl  alcohol  (36%),  methylene chloride (32%) 1,1,1
tricholoroethane  (45%),  trichloroethylene  (36%) chloroform (18%),
tetrachloroethylene  (18%),  and carbon tetrachloride  (23%).   Table  6-1
summarizes  the most  commonly occurring constituents.
      Halogenated  compounds  were  present  in  more than 68  percent  of the
waste  streams  for which  composition  data were  available.  It  also  should  be
noted  that  the  list  of major constituents  present  in  the 22  waste  streams
examined  agrees well with  the  list  of the  types  of  solvent recycled  at  the
facilities  surveyed by the  National  Association  of  Solvent Recyclers.
 (Docket  F-86-AESP-FFFFF, item  L0014).
                                      6-2

-------
        TABLE 6-1.  SUMMARY OF WSTF/TSDF WASTE STREAM CONSTITUENTS3
Concentration range
Constituent Frequency of occurrence,3 % (weight %)
Toluene
Ethyl benzene
1,1, 1-Trichloroethane
Xylene
Isopropyl alcohol
Trichloroethylene
Methyl ethyl ketone
Acetone
Methylene chloride
Chloroform
Tetrachloroethy lene
Carbon tetrachloride
55
36
45
27
36
36
32
32
30
18
18
23
0.4 -
0.1 -
0.3 -
5.9 -
0.2 -
0.2 -
0.2 -
21 -
0.3 -
0.03
0.6 -
0.2 -
50
11.4
100
49.2
42.9
8.0
60.0
74
40
- 4
3.0
0.5
aFor 22 waste streams from 16 WSTF/TSDF,
                                     6-3

-------
     As regards constituent concentrations,  for nonhalogenated organics
there were no perceivable patterns;  both high and low constituent concen-
trations were observed.  For halogenated compounds,  constituent concentra-
tions tended to be quite low (i.e.,  generally less than one (1) percent of
the waste stream), the two exceptions being methylene chloride and 1,1,1-
trichloroethane which did show up in very high concentrations in a few
cases (see Table 6-1).  No discernible connection was observed between TOC
and the presence of individual constituents.
     With respect to volatility, it is useful to look at vapor pressure and
boiling point  (see Table 6-2) in comparison to benzene since EPA has based
its regulatory approach for the TSDF equipment leak standards on existing
standards issued under Section 112 of the CAA for control of benzene emis-
sions.  Of the 12 major constituents previously listed in Table 6-1, six
(6) have vapor pressures greater than and boiling points less than benzene
(95 mm Hg and 80 °C, respectively):  MEK, acetone, methylene chloride,
1,1,1 trichloroethane, chloroform, and carbon tetrachloride.   It could,
therefore, be  concluded that organic constituents treated at WSTF do tend
to be moderately to highly volatile chemicals.
     Of the  chemical constituents reported  as present  in the 22 WSTF waste
stream cases,  the only ones  having established unit risk factors were  the
halogenated  organics.  For the halogenated  organics that were mentioned
above  as being major actors, the unit risk  factors ranged from  1.5 x  10'5
(/*g/m3)-l for  carbon tetrachloride to 4.7 x 10~7  (/^g/m3)'1  for  methylene
chloride.  This  range  agrees most favorably with  the  range  of  unit risk
factors used at  proposal,  2  x  10'5 to 2  x  10'7  (/*g/m3)-l.   The  one notable
addition to  the  list was  1,2-dichloroethane [2.6  x 10'5  0*g/m3)-l], which
appeared  in  only one  case.
      Testing of  vent  emissions  has been  conducted at  a number of  sites in
connection with  the evaluation  of organics  removal through  pretreatment
processes.   However,  plant trip  reports  and emission  test  reports  at  six
 (6)  sites  provided  the organics  composition and  concentration data  in the
detail  required  for this  analysis.   Table  6-3 presents a summary of  vent
emissions  test results from these  sites.  The information  in this  table is
the  total  quantified  organics in each  vent  stream.   The site numbers  are
the  identifiers  used by EPA.  Sites  20,  3,  22,  and 0 are chemical
                                      6-4

-------
TABLE 6-2.  INFORMATION FOR TSDF CONSTITUENTS3
Compound name
1,2-Dichloroethane
1 , 1 , 1-Trichloroethane
Acetaldehyde
Acetic acid
Acetone
Acetonitrile
Acrylonitrile
Aniline
Benzene
Benzyl chloride
Bisphenol (A)
Butanol-1
Butyl cellosolve
Carbon disulfide
Carbon tetrachloride
Chlorobenzene
Chloroform
Chlorotoluene
Cresol
Cumene (isopropyl benzene)
Cyclohexanone
Dichlorobenzene (1,2)
Molecular
weight
99.0
133.40
44.05
60.05
58.08
41.03
53.06
93.10
78.10
126.60
228.31
74.12
118.20
76.14
153.80
112.56
118.40
126.60
108.13
120.20
98.15
147.00
Vapor
pressure,
mm Hg
80
123
760
15.41
266
90
114
1
95.2
1.21

6.5
1.61
366
113
11.8
208
2.8
0.3
4.6
4.8
1.5
Boiling
point, °C
83.4
81.0
20.8
118.0
56.2
81.6
77.4
184
80.1
179.4

117.7
170.0
46.3
76.8
132.0
61.2
162.0
195.0
153.0
157.0
179.0
Unit risk
value,
(/*g/m3)-l
2.6 x ID'5

2.2 x 10'6



6.8 x ID'5
7.4 x ID'6
8.3 x ID'6





1.5 x ID'5

2.3 x ID'5





                                               (continued)
                      6-5

-------
TABLE 6-2.  (continued)
Compound name
Dimethyl formamide
Ethanol
Ethyl acetate (I)
Ethyl benzene
Ethylene glycol
Formic acid
Freon
Furan
Furfural (I)
Heptane
Hexane
Isopropanol
Methanol
Methyl acetate
Methyl acrylate
Methyl ethyl ketone
Methyl isobutyl ketone
Methyl (A) styrene
Methylene chloride
Morpholine
Nitrocellulose
Phenol
Molecular
weight
73.09
46.10
88.10
106.16
62.07
46.00
120.92
68.08
96.09
100.02
86.20
60.09
32.04
74.10
86.10
72.12
100.16
236.18
85.00
87.12
10,000
94.10
Vapor
pressure,
mm Hg
3.995
50
100
10
1.11
1.22
5,000
596.207
2
48
150.3
42.8
114
235

100
15.7

438
10

0.341
Unit risk
Boiling value,
point, °C (/ig/m3)-l
152.8
78.4
77
136.2
<
198.0
100.7
-29.8
31.4
161.7
98.4
69.0
82.4
65.0
54.0

79.6
115.8

39.8 4.7 x ID'7
129.0

182.0
                                   (continued)
          6-6

-------
                          TABLE 6-2.   (continued)
Compound name
Polybutadiene
Polypropylene
Polyvinyl alcohol
Propyl acetate
Sodium hydroxide
Styrene
Tetrachloroethylene
Toluene
Trichloroethylene
Tri ethyl ami ne
Urethane
Vinyl acetate
Xylene
Molecular
weight
1,000.00
20,000.00
120,000.00
102.13
40.00
104.20
165.83
92.00
131.39
101.22
89.09
86.09
106.20
Vapor
pressure,
mm Hg



35

7.3
19
30
75
400
10
115
8.5
Boiling
point, °C



101.6

145.0
121.4
110.6
87.0

-40.3
73.0
144.4
Unit risk
value,
0*g/m3)-i






5.8 x 10-7

1.7 x 10-6




Constituents representing at least 1  percent  of  the composition  of  EPA
 hazardous waste streams F001-F005 and D001  contained in  the  ISDB and
 Illinois EPA data base for which  properties were available.
                                     6-7

-------
I*
"*:-

M S
3'

m
                         89

      *****%************* J
                             !
      ********************
      *****6************S6
       *********** 6***%ti*~*
        ZZ"»*%*************

                          *t
                         1-65
       > a ^^^t5 ^z 5 ^ i
                        — -s  u
                       -   "
                    6-8

-------
                                                           TABLE 6-3.  (continued)




Ac* ton*
2,2 DiiMthyl ox Iran*
Methanol
Chloromethane
Site 20*

vent

NA
NA
NA
NA
Sit* 3C
condeneer
vent

NA
NA
NA
NA
Sit* 21. ,
batch ld '
condenser
v*nt
NA
NA
NA
NA
Sit* 21 .
batch 2°
condenaer
v*nt
NA
NA
NA
NA
Sit* 22
r*c*iv*r
v*nt

NA
NA
NA
NA
Sit* 26f
unit 1
condenaer
v*nt
*£,
120"
a. a)
NA
Sit* 26d
unit 2
cond*n*w
v*nt
1.1
NA
NA
NA
Sit* 0*»f
prlmry
cond*n«*r
v*nt
NA
NA
NA
14.1
Total quantified organ lea  240           1.8         24           160         6.O          200            1.2              3200

Composite unit risk        2.1 x Iff*    O           2.4 x 10"8   3.1 x lO'9  2.8 x Iff*   2.9 x «T7     2.3 x Iff*       8.6 x Iff*
  factor/ (Wg/m3)"1
NA = Not analyzed for

     *av*rag* of aix maaauramants.
     bav*rag* of fiv* maaauramanta.
     ^av*rag* of thr** meaauramanta.
     "avarag* of four maaauramanta.
     *av*rag* of eight maaauramanta
     'secondary condenser vent concentrations were similar.
     fltwo maaauremanta of <2.O mg/L averaged in aa zero.
     yone measurement of <0.1 averaged in aa zero.
     •three measurementa of 
-------
manufacturers that strip organics from wastewater both for recycling and as
wastewater treatment.  Sites 21 and 25 are organic solvent recyclers.
     Table 6-4 provides a summary of the ranges of constituent concentra-
tions and frequencies of occurrence for the tests presented in Table 6-3.
Waste Characterization Data Base
     The most comprehensive source of waste characterization information is
the TSDF waste characterization data base (WCDB).  As described in Appen-
dix D, to support the development of comprehensive air emission regulations
for hazardous waste TSDF, hazardous wastes were characterized by merging
the following five existing data bases:
    •  National Survey of Hazardous Haste Generators and Treatment. Storage,
       and Disposal Facilities Data Base
    *  Industrial Studies Data Base (ISDB)
    •  Listing documentation of 40 CFR 261.32 hazardous wastes from specific
       sources (i.e., waste codes beginning with the letter K)
    •  WET Model  Hazardous Waste Data Bases
    •  A data base created by the Illinois EPA.
    Of these five data bases, the best sources  of constituent and percent
composition of constituent information were the ISDB and  Illinois EPA data
bases.
    Generic spent solvent waste  streams  include EPA hazardous wastes F001-
F005.  In addition,  information  from the Screener Survey  indicates  that  D001
wastes (ignitable) are also recycled by  commercial waste  solvents recyclers.
Therefore,  information characterizing  F001-F005 and D001  waste streams was
extracted from the ISDB  and Illinois EPA data  bases.
    The  ISDB  is  a compilation  of data  from ongoing, in-depth  EPA/OSW surveys
of designated  industries that  are major  hazardous waste generators.  The ISDB
currently addresses  only 18 SIC  codes.   Data were gathered from  detailed ques-
tionnaires  completed  by  industry, engineering  analyses, and a waste sampling/
analysis  program. The  data base contains detailed  TSDF  information on  a site-
by-site  basis.   Due  to  the  confidential  nature of much of the data,  waste
 information  was aggregated.
     The  Illinois EPA Data Base contains  information  from  permit  applications.
 Each  hazardous  and  special  nonhazardous  waste  stream  managed  in  the State of
                                     6-10

-------
         TABLE  6-4.   VENT  EMISSIONS CONSTITUENT  RANGE AND  FREQUENCY
Compound
Vinyl chloride
Chloroethane
1,1 Dichloroethene
1,1 Di chloroethane
1,2 Dichloroethene
1,2 Di chloroethane
Chloroform
Tetrachloroethylene
1,1,1 Trichloroethylene
Methyl ene chloride
Carbon tetrachlorlde
Chloromethane
1,5 Hexadlyne
Toluene
Methyl ethyl ketone
Isopropanol
Xylene
Ethyl benzene
Acetone
2,2 Dimethyl oxirane
Methanol
Range, mg/l
32
17
14
4.8
3.4
130
0.12 - 222
1.0
1.8 - 110
1.0 - 1480
0.15 - 1440
14.1
1.4
0.14 - 6.7
7.7 - 49
1.9
0.47 - 3.4
1.5
1.1 - 19
120
3.8
Frequency3
1
1
1
1
1
1
4
1
2
3
3
1
1
3
3
1
2
1
2
1
1
aTimes observed at quantifiable levels out of a possible seven occurrences,
                                  6-11

-------
Illinois must be permitted through the Division of Land/Noise Pollution
Control.  Generators must submit one application for each waste.  To date,  the
data base contains about 35,000 permits.  An analysis of each waste must be
provided with each permit application.
    An examination of the D001 and spent solvent waste stream information in
the Illinois EPA Data Base indicates the same general trends as noted for the
plant-specific data; i.e., in general the waste streams were composed of one
or two major constituents with two or more minor constituents present in
lesser amounts.  The D001 streams tend to have a wider range of constituents
than the spent solvent streams.  The major constituents included those listed
for the plant-specific data as well as perchloroethylene, freon, and mineral
spirits.
    Organic constituent concentrations in the  ISDB and Illinois EPA data base
range from less than one-half percent to 100 percent.  To identify constit-
uents representing the range of properties characteristic of recycled waste,
constituents found in F001-F005 and D001 wastes in concentrations greater than
1 percent were extracted  from the data bases.  This  produced a  listing of more
than 100 constituents.  Properties for 57 of these constituents were available
in the  literature.  Table 6-2 presents a listing of  these 57 constituents and
available information on  properties  (molecular weight, vapor pressure, boiling
point and unit risk factor).  Molecular weights range  from 32.04  (methanol)  to
120,000  (polyvinyl alcohol);  vapor pressures from 0.3  mm Hg  (cresol) to  5,000
mm Hg  (freon); boiling points  from -40.3 °C  (urethane) to 195  °C  (cresol);  and
unit risk factors  (for  10 constituents) from 4.7 x  10'7  0»g/m3)-l  (methylene
chloride) to 6.8  x  10'5  (/ig/m3)'1  (acrylonitrile).   Again, the wide  range of
constituent  properties  indicates  the  difficulty in  characterizing WSTF waste
streams,  however,  the constituents chosen  for  the model  unit analysis  have
properties  spanning  the  range of  properties  of most  of the constituents  from
the  ISDB and  Illinois EPA data base.
Conclusion
     The waste  constituents  selected  for use  in analysis  of  the process  vent
standards were toluene,  methyl  ethyl  ketone,  1,1,1-tricholoroethane,  and meth-
ylene  chloride.   Their  selection  was  a  subjective judgment  based  on  review of
the  information  and data presented above.   Three  of these  solvents  are the
 same as were used at proposal.  Methylene  chloride  was added to the list of
                                     6-12

-------
solvents for analysis in order to provide a broader,  more comprehensive range
of chemical and physical properties.  Thus, based on  the available data
regarding waste solvent compositions and concentrations, EPA concluded that
the chemical constituents used in the post-proposal  analysis for process vents
are representative of the constituents found at waste solvent treatment facil-
ities.

6.2  UNIT RISK FACTORS
     Comment:  Four commenters (AESP-L0001, AESP-00009, AESP-00016, AESP-
L0008, and AESP-L0016) object to the procedures associated with the appli-
cation of unit risk factors.  In support, Commenter AESP-L0001 points out
that the BID lists 26 chemicals likely to be covered  by the land ban regu-
lations, but only 4 have unit risk factors and are used in the analysis.
The remaining 22 chemicals are not currently considered to be carcinogens,
although some are under review as potential hazardous air pollutants.  The
analysis assumes, however, that all of the emissions  have a risk comparable
to methylene chloride at the lower bound, and all of  the emissions have a
risk comparable to carbon tetrachloride at the upper  bound.  Also, the use
of unit risk factor estimates calculated for trichloroethylene, tetra-
chloroethylene, and methylene chloride as though they posed a human cancer
risk is unjustified.  Commenter AESP-L0008 believes that the failure to
address the weight of evidence for carcinogenicity is inconsistent with
EPA's risk assessment guidelines (51 FR 33992, September 24, 1986) and with
the principles for assessing cancer risk developed by the White House
Office of Science and Technology Policy (50 FR 10372, March 14, 1985).
Commenter AESP-L0016 objects to the lack of support provided for the selec-
tion and derivation of unit risk factors and for the  choice of carbon
tetrachloride as the upper bound unit risk factor because certain chemicals
that could be treated in large quantities at TSDF have a higher unit risk
factor (e.g., acrylonitrile or ethylene oxide).  Commenters AESP-00016 and
AESP-00009 object to the risk analysis because carbon tetrachloride,
selected to represent the upper bound on the range of unit risk factors, is
not representative of wastes recycled at commercial  facilities.  Several
commenters from TSDF and WSTF also state that the risk analysis is not
representative of the wide variety of wastes handled.
                                     6-13

-------
     Response:  The selection of the range of unit risk factors used at
proposal to estimate the cancer risk resulting from TSDF emissions was
based on a review of available information regarding WSTF/TSDF waste
streams.  Of the chemical constituents reported as present in the actual
WSTF/TSDF waste stream cases examined by EPA, the ones having established
unit risk factors were the halogenated organics (including carbon tetra-
chloride) and nitrobenzene.  In response to Commenters AESP-00016 and AESP-
00009, available TSDF waste characterization data indicate the presence of
carbon tetrachloride in recycled wastes in 23 percent of the waste streams
examined.  For the six halogenated organics that were identified as being
frequently occurring constituents in WSTF waste streams (see comment 6.1),
the unit risk factors for the chemicals ranged from 1.5 x 10~5 (/tg/nv3)-!
for carbon tetrachloride to 4.7 x 10'7  (/ig/m3)'1 for methylene chloride.
This range agrees most favorably with the range of unit risk factors
[2 x 10~5 to 2 x 10'7 (/jg/m3)-1] used at proposal for estimation of both
process  vent and equipment  leak health  risks.  However, this range of unit
risk factors was not used  in the final  analysis.
     To  estimate the cancer potency of  TSDF  air emissions in the revised
analysis, an emission-weighted composite unit  risk estimate approach was
used by  EPA to address the  problem of dealing  with the  large number of
toxic chemicals that are present at TSDF.  Using the emission-weighted
composite factor rather  than  individual component unit  risk factors
simplifies the risk assessment  so that  calculations do  not need  to  be
performed for each chemical  emitted.  The composite risk  factor  is  combined
with estimates of  ambient  concentrations of  total organics and population
exposure to estimate risk  due  to TSDF emissions.   In calculating the
emission-weighted  average  unit  risk  factor,  the  emission  estimate for a
compound is  first  multiplied by the  unit  risk  factor  for  that  compound;
then  the emission-weighted average  is computed by  summing these  individual
risk  numbers  and dividing  the sum by  the  total TSDF emission  value.   Using
this  type of  average would give the  same  result  as  calculating the risk for
each  chemical  involved.   However, only  those carcinogens  for which unit
 risk  factors  are available would  be included in  the analysis  of  cancer risk
 under this  approach.
                                     6-14

-------
     Through use of the EPA's TSDF WCDB and a computerized model developed
for analysis of the regulatory options for TSDF emission sources (see
Appendix D for a description of the model), EPA estimated total nationwide
TSDF emissions by specific waste constituent.  Thirty-nine chemicals were
identified as TSDF organic air pollutant emission constituents emitted from
equipment leaks at all types of TSDF waste management processes.  Unit risk
factors were then averaged based on both carcinogenic and noncarcinogenic
nationwide TSDF equipment leak emissions to calculate an emission-weighted
composite mean TSDF cancer unit risk factor.
     With regard to the concerns of Commenter AESP-L0016, numerous constit-
uents with higher unit risk factors than carbon tetrachloride (including
acrylonitrile and ethylene oxide) were included in the calculation of the
emission-weighted unit risk factor for TSDF equipment leaks.  The TSDF
equipment leak, emission-weighted unit risk factor value was determined to
be 4.5 x 10'6 (/ig/ra3)-l, anc( was usec[ to determine the health-related
impacts associated with TSDF equipment leak (fugitive) emissions rather
than using the range of the unit risk factors that represents the limited
number of chemical compounds (organic solvents) emitted at WSTF.  A more
detailed discussion of the hazardous waste TSDF unit risk factor determina-
tion is contained in the Appendix B of this document.
     Characterization of WSTF waste streams in the final analysis indicates
that the constituents and range of unit risk factors used at proposal in
the risk analysis are appropriate and representative of the waste solvent
recycling industry.  Results of the EPA waste stream characterization are
discussed in the response to comment 6.1.
     A similar approach to developing a composite unit risk factor for
process vents was also considered for the revised impacts analysis.  How-
ever, insufficient nationwide data on WSTF (a subset of the TSDF industry)
waste stream chemical constituent quantities and concentrations were avail-
able to develop an emission-weighted, arithmetic mean cancer unit risk
factor for WSTF process vents.  As shown in Table 6-3, the composite or
emission-weighted unit risk factors for individual process vent emission
streams varies widely.
     The WSTF waste streams and their associated process vent emissions
were found to contain a variety of chemical constituents.  Those
                                    6-15

-------
constituents with established risk factors were, in all  cases for the
plant-specific data, the halogenated organics; these halogenated organic
constituent concentrations tended to be quiet low,  generally less than one
(1) percent of organics emitted.  Therefore, EPA judged, based on the
limited data available, that use of a midrange unit risk factor would be
appropriate in estimating nationwide health impacts associated with WSTF
process vents.  The unit risk factor assumed at proposal, 2 x 10'6
(/ig/m3)-1, is the geometric midrange between the highest and lowest unit
risk factors for the constituents found in the WSTF process vent streams.
The unit risk factor calculated for the equipment leak emissions agrees
favorably with the process vent number used at proposal.  Because it is not
unreasonable to assume a similar mix of constituents in process vents as
from equipment leaks, and available data do not suggest otherwise, for the
purpose of estimating impacts the same unit risk factor was used for both
process vents and equipment  leaks, 4.5 x 10'6  (/*g/n»3)-l.  Any underestima-
tion of risk resulting from  the use of midrange risk factor for process
vents would tend to be offset by applying the  unit  risk factor to the
entire waste stream emission (i.e., all organic emissions are assumed
carcinogenic) and by use of  the high range  (upper bound) emission factor  in
determining process vent emissions.
     With  regard to the comment that the failure to address the weight of
evidence  for carcinogenicity is inconsistent  with EPA's  risk assessment
guidelines  and the  principles for assessing cancer  risk, it  should be noted
that early  in the rulemaking for TSDF, EPA  looked at the contribution to
total estimated  risk  (annual incidence) by  weight of evidence.   At that
time, C carcinogens accounted for about 5 percent of the total  risk,  and  A
carcinogens about 10  percent.   For  all practical purposes,  calculating
separate  risk estimates  for chemicals  in  each weight of evidence category
adds  little to the  risk  assessment.   However,  the  EPA  "Guidelines  for
Carcinogen Risk  Assessment"  (51  FR  33992,  September 24,  1986)  and  "Guide-
 lines  for the Health  Risk  Assessment  of Chemical Mixtures"  (51  FR  34014)  do
not describe a means  to  quantitatively incorporate  weight  of evidence into
 risk  assessments.   Thus,  according  to the guidelines,  making no distinction
among  carcinogens  regarding weight  of evidence is  appropriate.
                                     6-16

-------
6.3  METHODOLOGY
     Comment:  Commenter AESP-00016 challenges the methodology applied in
the assessment to estimate the cancer risk to the MEI, which results in an
unrealistic estimate of nationwide cancer incidence and risk that is higher
than any other previous risk estimate for any EPA rulemaking.  Commenters
AESP-L0016 and AESP-00010 state that the assessment is questionable because
the methods and procedures are inadequately described, are not based on
actual data, and have never been reviewed by the Science Advisory Board,
the National Air Pollution Control Techniques Advisory Committee (NAPCTAC),
or industry.  Commenter AESP-L0001 states that the methodology and assump-
tions produce a value that bears no relationship to the actual situation
and recommends that OAQPS perform a detailed assessment of the specific
sources and sites.  Commenter AESP-L0018 does not believe that EPA has
adequately analyzed whether hazardous waste TSDF pose a risk, whether the
risk is significant, and what sources or sites are causing the risk.
     Response:  At proposal, order-of-magnitude health impacts were esti-
mated for cancer risks from exposure to organic air emissions from WSTF and
TSDF.  The Human Exposure Model (HEM) (Docket No. F-90-AESF-FFFFF,  item
S0044) was used to calculate the magnitude of risks posed by WSTF at both
typical and maximum emission rates.  Based on an estimated urban/rural
distribution, EPA selected six WSTF to represent the nationwide WSTF
industry in performing the risk assessment.  Using the results of the
analysis of these "typical" facilities,  health impacts were extrapolated to
all WSTF and TSDF in general to provide nationwide estimates.
     In the revised health impacts analysis, annual cancer incidence and
MIR were again used to quantify health impacts for the control alternatives
for process vents and equipment leaks.  Annual cancer incidence is a
measure of aggregate risk per year to the entire population exposed to TSDF
emissions nationwide.  It represents the sum of each individual's estimated
risk to exposure to TSDF emissions, for individuals who live within a
distance of 50 km of any TSDF.  The MIR,  on the other hand, represents the
potential risk to those individuals who may live closest to the expected
worst-emitting TSDF for a lifetime of 70 years.  MIR is by no means repre-
sentative of the entire industry,  and in fact is experienced by few, if
any,  individuals.
                                    6-17

-------
     However,  in this followup analysis,  the HEM was run using site-
specific data on facility waste throughputs, emission rates,  meteorology
(i.e.,  nearest National  Weather Bureau meteorology station),  and 1980
census  population density for each WSTF and TSDF nationwide identified in
the various data bases.   The facility-specific information was obtained
from three principal sources.  Waste quantity and solvent recycling data
were taken from the 1986 National Screener Survey.  Waste management pro-
cessing schemes and waste types managed in each facility were based on the
Hazardous Waste Data Management System's (HWDMS) RCRA Part A applications;
the National Survey of Hazardous Waste Generators and Treatment, Storage,
and Disposal Facilities Regulated Under RCRA in 1981 (Westat Survey); and
the 1986 National Screener Survey.  In addition, health impacts were
evaluated using an equipment leak specific  risk factor  [i.e., 4.5 x 10'6
Oig/m3)-1] applicable to TSDF equipment leaks nationwide.  The unit risk
factor used for waste solvent recyclers was based on an analysis of the
organic chemicals associated with WSTF operations;  refer to comment 6.2 and
Appendix B for more  information  regarding  selection of  the unit risk
factors.
     In revising the methodology  applied in assessing cancer  risks, EPA
conducted  facility-specific  HEM  computer runs for each  of the 448 WSTF that
reported recycling  and/or reuse  of  solvents and other organic compounds,
and for each  of  the more than  1,400 TSDF in the  industry profile of 2,300
TSDF that  were  determined to manage wastes with greater than  10-percent
organic content.  These  HEM  results were used to  estimate nationwide  cancer
incidence  for both  TSDF  equipment leaks  and process vents; the  HEM  results
were also  used  in estimating the MIR  or maximum  risk for TSDF/WSTF  process
vents.   For estimates of maximum risk associated  with TSDF equipment  leaks,
a separate methodology was  used.
     At  proposal, no attempt was made to estimate MIR for  TSDF  equipment
 leaks  because of a  lack  of  adequate site-specific information.   Since pro-
posal,  the results  of a  multiyear project  to  collect information  on the
 Nation's  generation of  hazardous waste and the capacity available to treat,
 store,  dispose,  and recycle that waste have become  available on a limited
 basis.   In the survey,  all  active treatment,  storage,  disposal  and recycl-
 ing facilities  (TSDR)  were  sent a detailed package  of  questionnaires
                                     6-18

-------
appropriate to the processes they operate.  The completed questionnaires
were reviewed for technical accuracy; after independent verification, the
information collected was entered into a complex data base.  The TSDR
survey questionnaire responses contain the most detailed, up-to-date
nationwide information regarding the hazardous waste management technolo-
gies each facility has on site.  For each facility, detailed information
are available in the data base including facility area, numbers of hazard-
ous waste management units by process type (i.e., number of surface
impoundments, incinerators, recycling units), annual throughput by process
unit, and types of waste (i.e., RCRA waste codes) managed by each unit at
the facility.  The availability of this information in computerized format
made it possible to use the TSDR survey data base to identify facilities
that represent the population of worst case facilities with regard to
equipment leak emissions and the potential for high individual risk (MIR)
values.
     The MIR estimate was made, first, by screening detailed TSDR Survey
data for more than 1,400 TSDF to identify the facility that has the highest
potential equipment leak emissions and the highest potential for these
emissions to result in high ambient air concentrations (i.e., high emis-
sions on a small facility area).  Next, it was assumed that this facility
handles hazardous wastes that have carcinogens with an emission-weighted
potency equal to that of the nationwide average and that an individual was
residing at the shortest distance from the TSDF management units to the
nearest apparent residence.  The highest annual-average ambient concentra-
tion, resulting from this high emission-rate facility, predicted to occur
at the residence nearest the facility was then determined by dispersion
modeling.  The ISCLT dispersion model was used in the equipment leak MIR
analysis to model the worst case facility as a true area source with the
actual facility area of about one acre as input.  The highest annual
average out of the results of 5 years of meteorological data modeled for
each of the eight cities used to characterize nationwide meteorology was
selected for use in the MIR calculation.  Thus, the MIR estimate is con-
sidered a worst-case estimate for the industry and should not be inter-
preted to represent the risk posed by any facility in the industry other
than a worst-case facility or the risk experienced by anyone other than the
                                    6-19

-------
theoretically most-exposed individual.   A detailed discussion of the health
impacts methodologies is presented in Appendix B of this document.
     The results of this revised risk assessment were used as the basis for
evaluation of the proposed standards and for support of changes (such as
the addition of the facility process vent cutoff) that were made to the
standards since proposal.
6.4  ADDITIVE AND SYNERGISTIC EFFECTS
     Comment:  Commenter AESP-L0016 states that the assessment is flawed
because large uncertainties are introduced when the additive or synergistic
effects of carcinogens and the individual variability in response is not
factored in.  The commenter also notes that the analysis fails to consider
the substantial environmental and public health benefits in addition to
cancer prevention that would result from stricter standards, such as tera-
togenic, neurological, immunotoxic, and respiratory effects and the bene-
fits in ozone reduction.
     Response:  As noted  in response to comments in Section 6.3, annual
cancer incidence and MIR were the only indexes used to quantify health
impacts.  The EPA does recognize that other health effects may be associ-
ated with both  short-term and long-term human exposure to the organic
chemicals emitted to the  air at WSTF/TSDF.  The  EPA believes, however,  that
a  risk assessment based on cancer serves as the  clearest basis for evalu-
ating the health effects  associated with exposure to  air emissions from
TSDF.  A quantitative assessment of the potential nationwide noncancer
health impacts  (e.g., developmental, neurological,  immunological, and
respiratory  effects) was  not conducted due to deficiencies  at this time in
the health data base for  these  types of effects.
     Although  unable to  numerically quantify noncancer  health risks,  EPA
did conduct  a  screening  analysis of the potential adverse noncancer  health
effects  associated with  short-term  and  long-term exposure to  individual
waste  constituents emitted  from TSDF.   This analysis  was  based  on  a
comparison  of  relevant  health data  to the  highest short-term  or  long-term
modeled  ambient concentrations  for  chemicals  at  each  of two selected TSDF.
 (A detailed  presentation of the screening  analysis  is contained  in
Appendix B.)
                                     6-20

-------
     Results of this analysis suggest that adverse noncancer health effects
are unlikely to be associated with acute or chronic inhalation exposure to
TSDF organic emissions.  It should be noted that the health data base for
many chemicals was limited, particularly for short-term exposures.  The
conclusions reached in this preliminary analysis should be considered in
the context of the limitations of the health data; the uncertainties
associated with the characterization of wastes at the facilities; and the
assumptions used in estimating emissions, ambient concentrations, and the
potential for human exposure.  Additional evaluation of noncancer health
effects may be undertaken as part of the third phase of the TSDF regulatory
program.  To that effect, in the forthcoming proposal preamble for the TSDF
air rules for surface impoundments, tanks, and containers, EPA is
specifically requesting comments from the public on methodologies and use
of health data for assessing the noncancer health effects of TSDF organic
emissions.  In addition, because there is a potential for cancer and
noncancer health effects from TSDF chemicals from indirect pathways, such
as ingestion of foods contaminated by air toxics that have deposited in the
soil, EPA will evaluate the need to include an indirect pathway element in
the TSDF health risk analysis in the future.
     The EPA is aware of the uncertainties inherent in predicting the
magnitude and nature of toxicant interactions between individual chemicals
in chemical mixtures.  In the absence of toxicity data on the specific
mixtures of concern,  and with insufficient quantitative information on the
potential interaction among the components (i.e., additivity, synergism, or
antagonism), the EPA has assumed additivity to estimate the carcinogenicity
of the mixtures of concern.  This is consistent with guidance provided in
the 1986 EPA "Guidelines for the Health Risk Assessment of Chemical
Mixtures" (51 FR 34014).
     The EPA also recognizes that there are uncertainties associated with
the variability of individual  human responses following exposures to
toxicants.  As stated in the 1986 EPA "Guidelines for Carcinogen Risk
Assessment" (51 FR 33992),  human populations are variable with respect to
genetic constitution,  diet,  occupational  and home environment, activity
patterns, and other cultural factors.   Because of insufficient data,
however, the EPA is unable  to determine the potential  impact of these
                                    6-21

-------
factors on the estimates of risk associated with exposure to carcinogens
emitted from TSDF.
     With regard to ozone,  EPA is confident that the organic emissions
emitted from WSTF (and from TSDF in general) do contribute to ambient ozone
formation.  The magnitude of these impacts, however, is difficult to quan-
tify because ozone is a secondary pollutant and is not emitted directly
from air pollutant sources.  Unlike many other pollutants, ozone results
from a series of chemical reactions in the atmosphere between oxidant pre-
cursors in the presence of sunlight.  Ozone modeling is quite complex, and
large uncertainties are associated with the modeling results.  Therefore,
no attempt was made to estimate the contribution of WSTF/TSDF organic emis-
sions to ambient ozone levels or to quantify the benefits attained through
reduction of ozone levels as a result of implementing these standards.
However, EPA anticipates that this rule will result in some reduction in
ambient ozone levels.
6.5  DISPERSION MODELING
     Comment:  Commenter AESP-L0016 states that the Industrial Source Com-
plex (ISC) model should be used to predict risks instead of the HEM because
the HEM underestimates incidence by a factor of 2 to 3.  According to the
commenter, it is acceptable to use the HEM instead of the more costly ISC
as long as the exposure estimates are adjusted to account for the HEM's
underestimation.
     Response:  The alleged underestimation of exposure estimates to which
the commenter refers  involved area source modeling of specific model units
used to characterize  benzene emissions at petroleum and chemical plants.
For that particular situation, the exposure estimates generated using the
HEM and the  ISC did vary by a factor of 2 to 3.  That, however, was for  a
particular set of circumstances  unique to those process units and their
configuration.  To state that HEM will in all cases underestimate exposure
(ambient concentration)  is unfounded.  In  fact, it  could  be  argued that  the
dispersion modeling techniques used  in the  HEM could overestimate ambient
concentrations  rather than underestimate them depending on  the circum-
stances.  The HEM collocates all  sources at a single point,  resulting  in a
potential overestimate of  ambient  concentrations  in the vicinity of this
point.   In the  ISC,  sources can  be  located  throughout  the facility based on
                                     6-22

-------
 actual plant  layouts, thereby allowing the model to account  for  this
 separation of  sources.
      If the  ISC were to be used as the model for incidence estimation  and
 the advanced features of this state-of-the-art dispersion model  were to  be
 used,  it would be necessary to have  individual site configurations  for each
 of the facilities modeled.  This level of detail is simply not available
 for all WSTF and TSDF.  There were more than 1,400 different  facilities
 modeled using  the HEM; facility-specific emissions based on  actual  through-
 put data and site-specific meteorology and population data were  used in  the
 HEM dispersion modeling.
      The HEM has been used successfully in many EPA risk assessments,  but
 it does add an additional element of uncertainty because of  the  inherent
 assumptions of the model.  The HEM is not a true area source  model; as
 stated above,  the model collocates all emission sources at one central
 point  and models the emissions as a point source.  The Agency does  not
 think  that these shortcomings are serious, however, for estimating  nation-
 wide  incidence for point or area sources or maximum risk for  point  sources
 such  as process vents.  In fact, this is primarily what the  HEM  is  designed
 to do.
     A more detailed MIR risk assessment was performed using  ISCLT  for TSDF
 equipment leaks because more detailed information and data were  available
 in the 1987 National TSDR Survey data base for that specific  TSDF.  The  ISC
 is unquestionably a more sophisticated dispersion model that  is  capable of
 handling complex source and topographic configurations; and where one
 facility is involved and the data,  including source configuration,  are
 available it is the appropriate model to use.  Otherwise, there  is  basi-
 cally  little difference between the two dispersion models and the output
 they would generate based on similar input data.   Appendix B  provides more
 detailed information on the dispersion modeling conducted in  the health
 impacts analysis.
6.6  WORKER EXPOSURE
     Comment:  Commenter AESP-L0016 believes the risk assessment ignores
workers in the facilities,  yet the  analysis assumes that emissions  from
control devices will  be released near the ground  and will  expose workers to
                                    6-23

-------
much higher concentrations than would affect persons living around the
facilities.
     Response:  The EPA agrees that there is the possibility that TSDF
workers may be exposed to higher concentrations than those persons in the
general public who are living in the vicinity of the facilities.  This,
however, is not always the case; the area of maximum ambient concentration
resulting from an emission source is a function of emission characteristics
(e.g., height of release, release temperature, and velocity) as well as the
local meteorology (i.e., windspeed, wind direction, and atmospheric stabil-
ity).  In many situations, the location of the maximum ground-level concen-
tration is a considerable distance downwind of the source.  Nonetheless,
the risk assessment methodology applied in evaluating the impacts of these
standards is consistent with EPA's historical approach.  Furthermore, other
government agencies are charged with the responsibility of protecting and
guarding the safety of the workplace.
6.7  LEVEL FOR STANDARDS
     Comment:  Commenters AESP-00016, AESP-L0003, and AESP-L0006 state that
even if the risk estimates were reasonably accurate, then a reduction of
0.87 case of cancer and a cancer incidence of 1 case/yr are insignificant
compared to the 800,000 new cancer cases expected to occur nationwide each
year.  Consequently, the  level of protection achieved is negligible.
     Response:  The revised risk assessment shows that the reduction in
cancer  incidence resulting from implementation of the standards  is greater
than was originally expected.   In the revised analysis, EPA estimates that
for process vents and  equipment leaks combined, the cancer incidence reduc-
tion could be  as high  as  1.1 cases/yr, and the probability of contracting
cancer  to  the  individual  living in the vicinity of the uncontrolled worst-
case TSDF  could be as  high as  5 in 1,000.  The EPA does not consider these
health  risks  to be insignificant.
     The EPA  does recognize that most human  activities and events  involve
some degree of inevitable risk.  Either  risk may be  imposed on  the  indivi-
dual  (involuntary) or  an  individual  may  elect to accept a certain  level  of
risk  as  a  result of certain action,  such as  cigarette smoking.   This  latter
risk  is  considered a  voluntary imposition  of  risk.   In the  case of  organics
                                     6-24

-------
exposure from the TSDF industry,  the risk of cancer is imposed on the
vidual without consent or acceptance.  The ambient air concentrations of
organics emanating from these facilities increase the individual's prob-
ability or odds of getting cancer,  and they add to the prevailing cancer
incidence.  The EPA has determined  the public health risk of ambient air
exposure to organics emitted from TSDF to be significant.  Therefore, the
Administrator has decided to take measures to reduce the atmospheric re-
lease of organic air pollutants from these sources.
                                    6-25

-------
                         7.0  SELECTION OF STANDARDS

7.1  EMISSION CUTOFF FOR PROCESS VENTS
     Comment:  Several commenters object to the proposed standard for proc-
ess vent equipment that requires a fixed 95-percent emission reduction.
Commenters AESP-00002, AESP-00006, AESP-L0011, and AESP-00023 believe the
requirement would be inequitable because some operations could reduce
emissions by 95 percent and still have higher emissions than do some small,
uncontrolled operations.  Commenter AESP-00016 notes that the standard of a
fixed 95-percent reduction in emissions appears to penalize facilities with
the lowest emissions and allow those with the highest emissions to meet a
less stringent standard (i.e., emission level).  Commenter AESP-00023
states that their facility's air emissions are less than 0.45 kg/h (1 Ib/h)
and are regulated by the State under the CAA.  Under the proposed regula-
tion, the commenter1s facility would have to install control devices on all
condenser and still vents because the standards do not differentiate an
acceptable emission level; according to the commenter, this would be
inconsistent with CAA regulations.  Commenter AESP-L0013 suggests that EPA
consider an alternative standard to establish a minimum allowable stack
emission rate applicable to all facilities; the commenter suggests 5,000
ppm, which is based on a 95-percent control of a 10-percent organic waste
stream.  Commenter AESP-00002 also recommends that an emission limit be
set.
     Response:  The EPA agrees with the commenters that the standard for
process vents, requiring a fixed 95-percent emission reduction,  could
result in the control  of some small facility operations not necessary to
protect the public health.  As proposed,  all  WSTF/TSDF would have to
install control  devices on all condenser and other process vents regardless
of emission levels because the standards  did not identify an acceptable
                                     7-1

-------
emission level.  The EPA reanalyzed the process vent air quality and health
impacts using updated model unit, emission rate, and facility throughput
data.  The process vent health impacts were estimated using the Human
Exposure Model (HEM) to predict nationwide health effects of exposure to
suspected carcinogens in the emissions from WSTF and TSDF.
     The HEM is a computer model that calculates exposure levels for a
population within 50 km (31 mi) of a facility using 1980 census population
distributions and local (National Weather Service) meteorological data.  A
representative unit risk factor  (i.e., 4.5 x 10'6) was used to quantify the
magnitude of risks posed by WSTF process vents  at both low and high emis-
sion rates.  The nationwide annual incidence resulting from WSTF process
vent emissions was calculated using the WSTF industry profile  (Docket No.
F-90-AESF-FFFFF, item S0043) to  identify facility locations for population
patterns and to determine overall 1985 facility waste solvent  throughput
for emission estimates.  The HEM was run for each WSTF using the site-
specific data, and  results were  summed to get the aggregate risk or total
nationwide cancer incidence.
     The risk  analysis  is  based  on overall facility operations and process
vent emissions.   It was not possible to determine WSTF unit capacities  from
the  1986 Screener Survey data.   Although total  facility  waste  solvent
throughputs were  available, the  data base did not contain  any  information
on the number  of  process units  at  each site; thus,  an analysis of  individ-
ual  process units or  process  vents was not possible.  A  review of  limited
site-specific  data  indicates  that, for small quantity recyclers  (i.e.,
<189,000  I or  50,000  gal/yr)  that  make up most  of the industry,  the  facil-
ity  would  typically consist of one waste  solvent  recycling unit  with
process emissions released through one process  vent.  For large  facilities,
there  may  be  one,  two,  three,  or more  units  at  the  facility that are  used
to process the total  WSTF  throughput.  No  information is currently
available to  determine specific process  unit counts,  and no basis  is
available to  assign process  unit counts  to  the facilities nationwide.
Therefore,  the WSTF model  unit emissions,  cost, and health risk  analyses
 assume that  there is one unit per facility  and one  process vent  per unit as
 an  emission  source.  This  method of  facility characterization is not
 expected to significantly  affect the overall analysis.   For example,
                                      7-2

-------
emission estimates are based on total facility waste solvent throughput
data; therefore, they reflect overall facility operations, not individual
unit throughputs.  In the risk analysis, the HEM colocates all sources, and
emission characteristics (i.e., stack height and diameter, and exit gas
velocity and temperature) are generally similar for all units; as a result,
risk estimates are not affected by this method of characterization.
     The nationwide maximum individual risk (MIR) was assumed to be the
highest individual risk calculated for the actual WSTF analyzed.  The MIR
is intended to reflect the most exposed individual's chance of getting
cancer if exposed continuously for 70 years to the highest annual average
ambient concentration resulting from WSTF process vent emissions.  Tables
7-1 and 7-2 present a summary of the WSTF process vent risk analysis for
the high emission rate case for incidence and MIR using a unit risk factor
of 4.5 x 10-6.
     As can be seen in Figures 7-1 and 7-2, the analysis results show that
reductions in both MIR and nationwide cancer incidence level off at a
facility emission rate of about 2.8 Mg/yr (3.1 ton/yr).  Figure 7-3 shows
that only minimal emission reductions are achieved by controlling those
facilities with total process vent emissions below about 2.8 Mg/yr (3.1
ton/yr).  A typical rate of 2,080 h/yr of operation was assumed for the
risk analysis; which corresponds to 1.4 kg/h (3 Ib/h) of organic emissions.
With an emission rate limit of 2.8 Mg/yr (3.1 ton/yr), an incidence
reduction of 0.34 case/yr is achieved; lowering the limit does not result
in further reductions in incidence.  Consequently, the risk analysis
results indicate that provision of a facility emission rate limit of 1.4
kg/h (3 Ib/h) and 2.8 Mg/yr (3.1 ton/yr) for process vent emissions
provides essentially the same level of protection for public health (in
terms of risk and incidence) as does covering process vents at all
facilities.
     Therefore, the final rule requires control of only those facilities
emitting 1.4 kg/h (3 Ib/h)  and 2.8 Mg/yr (3.1 ton/yr) or greater of organic
emissions from affected process vents.  The EPA has formatted the "small
facility cutoff" in terms of both an hourly emission rate (1.4 kg/h
[3 Ib/h]) and an annual emission rate (2.8 Mg/yr [3.1 ton/yr]).  The
                                     7-3

-------
                TABLE 7-1.  SUMMARY OF PROCESS VENT RISK ANALYSIS (INCIDENCE)
Emission
rat* cutoff i
kg/h Mg/yr
Base 1 i ne
11.8
4.6
2.3
1.4
0.9
0.5
0.3
0.2
0.1
0.0B
0

24.6
9.6
4.7
2.8
1.9
0.9
0.7
0.6
0.3
0.1
0
Percent of
facilities
below cutoff
NA
86.0
70.4
62.8
46.6
41.9
32.2
27.2
22.3
18.6
9.0
0
Uncontro 1 led
annual
incidence,
cases/yr
0.36
0.12
0.047
0.014
0.007
0 • 00D
0 .006
0.002
0.002
0.002
0.002
NA
Control led*
annual
incidence,
cases/yr
__
0.11
0.020
0.020
0.020
0.020
0.020
0.022
0.022
0.022
0.022
0.022
Total t>
incidence,
cases/yr
0.36
0.23
0.067
0.034
0.027
0.026
0.026
0.024
0.024
0.024
0.024
0.022
Inc i dence
reduction,
cas«*/yr
NA
0.13
0.29
0.33
0.33
0.34
0.34
0.34
0.34
0.34
0.34
0.34
NA = Not applicable.

•Control facility incidence is the residual incidence resulting from process vent
 emissions to the atmosphere after controls are in place.

bTotal incidence is the sum of the incidence resulting from uncontrolled facilities
 plus the incidence from controlled facilities.

-------
           TABLE 7-2.  SUMMARY OF PROCESS VENT RISK ANALYSIS  (MIR)
Emission
rate cutoff,
kg/h Mg/yr
Baseline
11.8
4.5
2.3
1.4
0:9
0.5
0.3
0.2
0.1
0.05
0

24.6
9.5
4.7
2.8
1.9
0.9
0.7
0.5
0.3
0.1
0
Percent of
facilities
below cutoff
NA
86.0
70.4
52.8
45.5
41.9
32.2
27.2
22.3
18.6
9.0
0
Maximum
concentration,
/*g/m3
176
50
25
10
7a
7
7
7
7
7
7
7
Population
exposed
11
2
30
1
11
11
11
11
11
11
11
11
MIR
8 x lO'4
3 x ID'4
1 x ID'4
6 x 10-5
4 x ID'5
4 x 10-5
4 x 10-5
4 x ID'5
4 x 10-5
4 x 10-5
4 x 10-5
4 x 10-5
 NA = Not applicable.
MIR = Maximum individual risk.

aThis concentration occurs at a facility that has been controlled to the
 95% level.
                                     7-5

-------
1
cr>
                                              8             12            16

                                                Emission Rate Cutoff (Mg/yr)
24
                      Figure 7-1.  Emission rate cutoff vs. maximum individual risk (high emissions).

-------
IT
I
I
0)
;g
o
     0.00
                                     8              12             16
                                        Emission Rate Cutoff (Mg/yr)
20
                     Figure 7-2. Emission rate cutoff vs. incidence (high emissions).

-------
I
00
                                      8
      12         16          20

Emission Rate Cutoff (Mg/yr)
                                Figure 7-3. Emission rate cutoff vs. emission reduction.

-------
overall risk analysis was based on annual  throughput,  emission, and ambient
air concentration values; therefore,  the annual limit was selected.  How-
ever, to avoid potential  short-term,  acute air pollution situations, the
corresponding hourly emission limitation was included, reflecting the
number of annual operating hours.
     A total TSDF/WSTF facility emission rate limit also provides the
owner/operator with a degree of flexibility to control sources in a more
cost-effective manner than is allowed without the limit format on a
facility basis.
     Because the promulgated standards contain the facility-based emission
rate limit that is more effective in controlling emissions from affected
sources and excluding facilities with little emission reduction potential,
the 10-percent concentration criterion for process vents has not been
included in the final rules.  Based on the post-proposal emissions and
health risk analyses, this emission rate limit represents an emission level
from process vents that is protective of human health and the environment,
and below which additional meaningful reductions in nationwide health risk
and environmental impacts attributable to process vents cannot be achieved.
Control of facilities with process vent emissions less than the emission
rate limit does not result in further reductions of either cancer risk or
incidence on a nationwide basis.  Facilities with organic emissions from
process vents that do not exceed these emission rates will not have to
install controls on affected process vents.
     Because the emission rate limits (3 Ib/h and 3.1 ton/yr) provide
reasonable health-based cutoffs, EPA considered dropping completely the
organic content criterion (i.e., at least 10 percent  total organics).
However, EPA decided not to completely eliminate the  organic content
criterion because it is not clear that the same controls can be applied to
very-low-concentration streams as can be applied to the higher-concentra-
tion streams that generally are  associated with emission rates greater than
the  limits.  For low-concentration streams, EPA questions whether controls
are  needed on a national or generic basis, but is unable to resolve this
question at this time.  Thus, EPA decided to defer controlling very-low-
concentration streams until it is able to better characterize  and assess
these  streams and the appropriate controls.
                                     7-9

-------
     Once EPA decided to consider facilities that manage very-low-concen-
tration organic wastes as a separate category, there remained the problem
of determining the appropriate criterion.  The EPA examined existing data
on air strippers (see Table 4-1, Chapter 4.0); it appeared that the quan-
tity of emissions and the risk associated with air strippers treating
streams with concentrations below 10 ppmw may be relatively small, thus
minimizing the potential harm of deferring control until a later time.
Examples of facilities managing low-concentration wastes are sites where
ground water is undergoing remedial action under CERCLA or corrective
action pursuant to RCRA.  Given the limited set of precise data available,
and the comments that the 10-percent criterion was too high, EPA determined
that an appropriate criterion would be 10 parts per million (ppm) total
organics in the waste by weight.
     The 10 ppmw is not an exemption from regulation; it is intended only
as a way for EPA to divide the air regulations into phases.  The EPA is
deferring action on very-low-concentration streams (i.e., ones with less
than 10 ppmw total organic content) from the final rule today but will
evaluate and announce a decision later on whether to regulate these waste
streams.
7.2  FEASIBILITY OF A 95-PERCENT CONTROL
     Comment:  Whereas Commenters AESP-L0003 and AESP-00007 support the
95-percent level of control for process vents, Commenters AESP-L0011,
AESP-00002, AESP-L0025, and AESP-00006 state that the proposed regulation
does not consider the limits of the best available technology.  A 95-per-
cent reduction may not be technologically achievable for a liquid waste
stream emitting vapors with a low organic content with currently available
pollution control equipment.  Commenter AESP-00025 does not believe the
level is feasible given baseline requirements under CAA State regulations.
     Response:  Emission control technologies applicable to WSTF process
vent emissions  (those emissions released to the atmosphere through process
vents such as the primary condenser exhaust)  include secondary condensers,
carbon adsorbers, flares, thermal afterburners, incinerators, and scrub-
bers.  The emission reduction potentially achievable by each control tech-
nology depends on the physical parameters associated with the process vent
stream and the design and operation of the control device.  For example,
                                    7-10

-------
 the  efficiency  of  a  condenser  is dependent  on  the  physical/chemical  proper-
 ties of  the  solvents being  condensed,  the organic  concentration  in  the  gas
 stream,  and  the operating temperature  of the condenser.
      Review  of  the available information on WSTF operations  indicates that
 condensers,  carbon adsorbers,  incinerators, and scrubbers  are  widely used;
 however, condensers  and carbon adsorbers appear to be the  technology of
 choice to  reduce process emissions of  organics at  WSTF.  Properly designed,
 operated,  and maintained, each of these control technologies can achieve  a
 95-percent emission  reduction, although there  are  situations in which each
 particular technology may not be applicable as a control method for  WSTF
 process  vents.   For  example, secondary condensers  used for emission  control
 (or  supplemental product recovery) are not  well-suited for vent streams
 containing organics  with low boiling points, high  moisture content,  large
 quantities of inerts (e.g., C02 or N2), or  low concentrations  of organics
 (i.e., too low  thermodynamically to support a  liquid phase).   Conditions
 for  which  carbon adsorption is not ideally  suited  for WSTF vents, primarily
 because  of cost  considerations, include streams with high  organic concen-
 trations,  very  high  or low molecular weight compounds, high moisture
 content, and mixtures of high- and low-boiling-point organics.  Nonethe-
 less, based on  an  analysis of these two control technologies (condensation
 and  carbon adsorption), at least one should be technically applicable to
 each  affected WSTF process vent stream and  should  be capable of achieving
 the  95-percent control efficiency for  removal  of organic emissions that is
 the  basis  of the standard.
      With  the establishment of a facility emission rate limit,  it is
 expected that condensers will  remain the technology of choice  for control
 of process vent emissions because many of the low-concentration streams
 likely will not require control to meet the 1.4 kg/h (3 Ib/h)  and 2.8 Mg/yr
 (3.1  ton/yr)  facility process  vent organic emission limit.
      Because the final rules could apply to dilute process  vent streams and
 the rule is formatted in terms  of a weight-percent reduction standard,  it
 is necessary to include a volume concentration standard in  the final rule
to account for the technological  limitations of enclosed combustion
devices,  one of the control  technologies examined  as part of the rulemak-
 ing,  used for treating dilute  streams.   Below a critical  concentration
                                    7-11

-------
level,  the maximum achievable efficiency for enclosed combustion devices
decreases as inlet concentration decreases;  thus,  for streams with low
organic vapor concentrations, the 95-percent mass  reduction may not be
technologically achievable in all cases.  Available data show that 20 ppmv
is the lowest outlet concentration of total  organic compounds achievable
with control device inlet streams below approximately 2,000 ppmv total
organics (48 FR 48932, October 21, 1983; and 48 FR 57538, December 30,
1983).  Therefore, a concentration limit of 20 ppmv has been added as an
alternative control device standard for incinerators, boilers, and process
heaters to allow for the drop in achievable destruction efficiency with
decreasing inlet organics concentration.  For consistency, the 20 ppmv  is
expressed as the sum of the actual individual compounds, not carbon equiva-
lents, on a dry basis corrected to 3 percent oxygen.  For facilities that
do not meet the emission rate limit, the final process vent  standards
require that control devices achieve a 95-percent reduction  in total
organic emissions for the facility or,  in the case of enclosed combustion
devices, a  reduction of each process vent stream to  a concentration of
20 ppmv total  organic compounds.
7.3  FEASIBILITY  OF CONDENSERS
     Comment:  Several commenters do not agree that  condensers  provide  a
feasible means of meeting the 95-percent emission  reduction  requirement.
Commenter  AESP-L0007  believes that condensers could  not  achieve  a 95-per-
cent efficiency without  cooling,  and cooling would  freeze the  water  in  the
waste  stream in the TSD  portion of the  facility.   In comparison,  Commenter
AESP-L0007 suggests that the 95-percent efficiency  be reduced  to 85  percent
so  that a  brine-cooled vent  condenser could be used.  Commenters AESP-
00035,  AESP-L0009,  AESP-00010,  AESP-00024,  AESP-L0025,  and AESP-00018 do
not believe EPA  has justified applying  the  same control  to the wide  variety
of  waste solvents,  and that, in many cases, condensers are not technically
 feasible due to  the nature  of the constituents.   Commenter AESP-L0016
objects to the estimate  of  condenser efficiency  applied in the analysis.
 According to the commenter,  about six times more  condenser area is needed
 to achieve the required  control efficiency for the higher flow rates of
 about 18 kg/h (40 Ib/h)  assumed in the proposed  standard.  Larger
                                     7-12

-------
 condensers could also be used effectively and economically to reduce
 emission levels by more than 95 percent.   This commenter also questions the
 selection of CFC-12 as a condenser refrigerant in  the calculations because
 the temperature difference between condensing vapors  and coolant is an
 important factor governing control efficiency.  The commenter suggests use
 of CFC-22 or R-50 or others with lower boiling points that will  lead to a
 higher reduction efficiency.
      Response:   The EPA examined the  feasibility of using condensers for
 organic emission control  using a state-of-the-art  chemical  engineering
 computerized process simulator known  as the  Advanced  System for  Process
 Engineering (ASPEN).  A variety of chemical  constituents and operating
 conditions was  examined to determine  the  organic removal  efficiency achiev-
 able through condensation.   The ASPEN condenser configuration consisted of
 (1)  a floating-head,  one-pass,  shell-and-tube heat  exchanger,  (2)  a refrig-
 eration unit capable of producing  a coolant  (e.g.,  chilled  brine)  at a
 temperature as  low as -28  °C (~20  °F),  and (3)  an optional  primary water-
 cooled  heat exchanger.   This final  item is considered  necessary  in some
 instances  to reduce the size of the refrigeration unit and  to remove water
 vapor in  order  to  avoid freezing problems because the  condenser  temperature
 is  low  enough to cause  ice  buildup on  heat transfer surfaces.
      The  constituents selected  for the  condenser analysis  (toluene;  MEK;
 1,1,1 trichloroethane;  and  methylene  chloride)  are considered  to be repre-
 sentative  of the solvents  recycled  by the WSTF  industry;  selection  was  a
 subjective judgment  based on  a  review of relevant information  and  litera-
 ture, which  included  an NASR  survey as well as  numerous  site-specific  plant
 trip  reports and responses  to EPA  Section 3007  information  requests.   Three
 of these four solvents  are  the  same as were used in the proposal  analysis.
 Methylene  chloride, a commonly  recycled solvent, was added  to  the  list  of
 solvents for analysis in order  to provide a comprehensive range of  volatil-
 ities for  the condenser analysis.  Methylene chloride is at the lower end
of the boiling point  range and  is therefore more difficult to condense.
Methylene chloride  (a relatively low-boiling/high-volatility compound)  has
a boiling point of 40 °C (104 °F); this compares with  boiling points of
                                    7-13

-------
110 eC, 79 eC, and 74 eC (230 °F, 174 °F, and 165 °F) for toluene, MEK, and
trichloroethane, respectively.
     Ten organic emission rates were analyzed for each of the four solvent
types; emission rates varied from 0.04 kg/h (0.08 Ib/h) up to 4.8 kg/h
(10.60 Ib/h).  Three exhaust gas flow rates (3.9, 0.6, and 0.3 L/s [8.3,
1.2, and 0.6 scfm]) were examined with the 10 emission rates to provide a
wide range of organic concentrations, from 1 to 60 percent organics.
Emission streams were assumed to consist of organics and a noncondensible
gas (e.g., air).  Thus, a total of 40 emission stream cases were included
in the condenser analysis in order to represent the wide range of stream
conditions encountered in the industry.
     The results of the ASPEN condenser simulation (Docket No. F-90-AESF-
FFFFF, item S0023) indicate that in only 15 of the 40 cases examined was a
removal efficiency of 95 percent achieved.  In 7 of the 40 cases, ASPEN
shows that, for those particular situations, appreciable condensation would
not occur.  This results from the partial pressure of the organic constitu-
ent in the vapor phase being too low to thermodynamically support a liquid
phase.  Six of the seven cases that would not condense involved methylene
chloride at concentrations of about 10 percent or less (by weight).  The
seventh case was a low concentration (1 percent) MEK.
     In general, the WSTF condenser analysis indicates that condensers
cannot universally achieve a 95-percent emission reduction when applied to
WSTF process vents.  However, they do provide, where applicable, an effec-
tive, economical method of emission reduction.  It must also be pointed out
that EPA did not base the facility process vent emission reduction require-
ments solely on the use of condensers.  Any technology capable of meeting
the standards is acceptable.  The 95-percent control efficiency option for
WSTF/TSDF process vents was analyzed on the basis of applying a condenser
to the vent as the initial choice of control technology; for those cases
where condensation is not feasible, carbon adsorption can be used, for
example.  Carbon adsorption is considered capable of attaining a 95-percent
control efficiency for organics removal on all WSTF process vent streams.
     With regard to the comment that increased condenser area or an alter-
native condenser coolant will lead to a higher emission reduction or
                                    7-14

-------
 control  efficiency,  the condenser analysis shows that there are technical
 limits on the efficiency of condensers that go beyond condenser area and
 operating parameters.   The efficiency of vent condensers is dependent on
 the physical  properties of the solvents being condensed and the solvent
 concentration in  the gas stream as well as the operating temperature and
 surface area  of the  condenser.  The analysis conducted by EPA has shown
 that low solvent  concentration and the high volatility of some solvents
 typically recycled at  WSTF are more limiting to condenser efficiency than
 is  operating  temperature.   The emission rate limit  will reduce the require-
 ment for control  of  some low-concentration streams;  however,  solvent
 volatility will remain a limiting factor in condenser efficiency.
 7.4  FEASIBILITY  OF  FLARES
      Comment:   Eight commenters object to the use of flares at recycling
 facilities because of  technical  and safety concerns  (AESP-00016,
 AESP-00019, AESP-00009,  AESP-L0007,  AESP-00005,  AESP-L0001,  AESP-L0018,  and
 AESP-L0025).   Commenters AESP-L0007,  AESP-00022,  and AESP-L0005 state that
 flares generally  require a constant emission source  for efficient operation
 and are  not suitable on  intermittent  sources or low-level  emissions  typical
 of  recycling  operations.   Commenter AESP-L0001  describes  flare pilot ther-
 mocouples  as  "not reliable" and  states  that  they  usually  last  only 3 years
 as  flare  pilot  flame detectors.   This  commenter recommends  that EPA  recog-
 nize  that  visual  inspection and  inherent  reliability  of the  fuel  supply
 should suffice  for pilot flame  detection methods.  Commenters  AESP-00016,
 AESP-00019, and AESP-00009 state  that  flares  present  the  danger of explo-
 sion, especially  if  they malfunction,  and  according  to  Commenter  AESP-
 00016, many State laws prohibit the use of flares at  recycling  facilities.
      Response:  Available  information on WSTF operations  indicates that
 condensers, carbon adsorbers, and  incinerators  are the most widely used
 control technologies; therefore,  they are expected to be  the technologies
 of choice to reduce process organic emissions at WSTF.  The final  technical
 analyses have shown that a 95-percent control efficiency can be achieved
with secondary condensers for many WSTF process vents or with carbon
 adsorbers in cases where secondary condensers are not feasible.  Flares are
                                    7-15

-------
not required controls but are an available option for facilities so
equipped, provided that they meet the criteria in the promulgated
standards.
     With regard to the safety of flares, the Agency has determined that
the use of flares to combust organic emissions from TSDF process vents
would not create safety problems if engineering precautions are taken in
the design and operation of the system.  The following are typical
engineering precautions.  First, the flare should not be located in such
close proximity to a process unit being vented that ignition of vapors is a
threat to safety.  In the analysis conducted for this standard at proposal,
it was assumed that the flare would be located as far as 122 m (400 ft)
from the process unit.  Second, controls such as a fluid seal or flame
arrestor are available that would prevent flashback.  These controls were
considered in EPA's cost analysis for the proposed rule (EPA-450/3-86-009).
Finally, the use of a purge gas, such as nitrogen, plant fuel gas, or
natural gas, and/or the careful control of total volumetric flow to the
flare would prevent flashback in the flare stack caused by low off-gas
flow.
7.5  FEASIBILITY OF CARBON ADSORBERS
     Comment:  Several commenters object to the identification of carbon
adsorption as a control technique because of technical and safety concerns.
Commenters AESP-L0007 and AESP-00022 claim that devices are needed for each
vent due to the low, intermittent emissions (which increase the costs), and
that the devices are not efficient for low concentration operations.  These
devices also are emission sources themselves because the used carbon must
be regenerated or buried, thus producing more emissions and increasing
costs.  According to Commenter AESP-00019, carbon adsorbers also are gen-
erally inappropriate because most WSTF use thin-film evaporators for recyc-
ling multicomponent solvent streams and only very few use fractional dis-
tillation columns to separate out single solvents.  According to Commenters
AESP-00019, AESP-00016, and AESP-00009, use of carbon adsorbers at facili-
ties handling multicomponent solvent waste streams also increases the
likelihood of hot spots resulting in fire or explosion because this control
is not compatible with all solvents.  Commenter AESP-L0016 objects because
                                    7-16

-------
EPA did not provide support for the removal efficiency of carbon adsorbers
and cites other authorities that support a 98-percent removal efficiency
for this type of control device.
     Response:  First, it should be noted that carbon adsorption is one of
several control technologies that could be used to attain the standards.
Other technologies include condensers, flares, incinerators, boilers, and
process heaters.  The commenters are correct in asserting that carbon
adsorption devices are potential emission sources because of the regenera-
tion or disposal of spent carbon.  However, the emissions from these
sources should be minimal compared to the emissions from uncontrolled
process vents if proper procedures are followed in managing the regenera-
tion and disposal of spent carbon.  The costs of regeneration and disposal
are not unreasonable and are included in the carbon adsorption control cost
analysis.
     Regarding carbon adsorption applications, EPA acknowledges that safety
is an important consideration,  but concludes that any safety and technical
problems can be avoided through proper design, operation, and maintenance,
including sorbent selection.  The problem of hot spots is associated with
multicomponent systems, especially large beds (large diameter) with low
flow rates, for cases involving low concentrations as well as saturated
(high) concentrations.  The problem results from a failure of the bed to
adequately dissipate the heat generated through release of heat of the
adsorption.  Physical adsorption is an exothermic operation that is most
efficient within a narrow range of temperature and pressure.  Excessive
heat buildup can result in problems.  It has been found that some carbon
types are worse than others in  regard to heat generation.  For example,
coal-based carbons have fewer problems than do wood-based carbons.   This is
in part due to the temperature  profiles generated in the carbon manufac-
turing stage;  coal-based carbons are exposed to higher temperatures at the
tail end of the process, whereas wood-based carbons are exposed to the
higher temperatures in the middle of the manufacturing process.  Another
contributing factor is bed size; i.e., larger diameters result in poor heat
transfer.  It  is difficult to predict what particular solvents (or
                                    7-17

-------
organics) would generate excess heat when adsorbed by a particular sorbent;
solvent/sorbent testing would likely be required.
     With regard to carbon-bed efficiencies, multicomponent solvent vapor
streams can experience reduced efficiencies for particular components.
This is typically a result of higher molecular weight compounds displacing
the lighter molecular weight compounds as the various compounds compete for
adsorption locations.  The lighter compounds move down the bed as they are
displaced; this causes a "wave front" for each of the compounds and can
result in early breakthrough.  In some cases, lighter components can exit
the bed at a higher concentration than the entering concentration.  This is
of concern in dealing with solvents because lower explosive limit (LEL)
values may be exceeded.
     Multicomponent systems generally present a much more complex design
problem than do single component systems.  With a wide mix of solvents in  a
multicomponent stream, selection of proper sorbents would typically require
design simulations to evaluate the possible combinations and alternatives.
A change  in the component mix would effectively alter the operating (and
design)  situation; breakthrough could occur earlier  if less adsorbable
components are present.  Another consideration in dealing with multicom-
ponent gas streams is component interaction.  This may include cases  where
components are mutually  soluble or  reactive with  carbon;  loss of  bed  capa-
city  (or efficiency)  through  fouling  is  a  common  consequence.  Component
interactions  are  difficult  to predict.   Fouling  also occurs when  monomers
present  in the  stream are polymerized  as a result of elevated temperatures
or  other factors. This  is  a  particularly  critical problem when  beds  are
designed for  regeneration.
      Moisture also reduces  the capacity  of the  carbon.   Adsorbents, espe-
cially oxygenated adsorbents, have a strong preferential  affinity for water
vapor over organic gases and would have  difficulty  attaining  high organic
 removal  efficiencies for the high  moisture gas  streams (i.e.,  relative
 humidities greater than 50  percent)  from some WSTF and TSDF  distillation
 vents.  If there is  a swing in moisture  levels  that is outside the design,
 then early breakthrough can occur.
      The performance of carbon adsorption  systems is improved by reducing
 the relative humidity of the controlled  gas stream to 50 percent or less.
                                     7-18

-------
If the gas stream is below 21 °C (70 °F) at a 100-percent relative humid-
ity, a reduction of the relative humidity can be achieved by heating the
gas stream to approximately 32 °C (90 °F).  If the gas stream is above
21 °C (70 °F) at a 100-percent relative humidity, the gas stream can first
be cooled to condense the water, then be reheated to achieve a relative
humidity of 50 percent or less.  These conditions can be obtained by using
only a duct heater, or by using both a condenser and a duct heater.  The
most effective means of avoiding the complication of multicomponent early
breakthrough is proper design followed by close monitoring of the bed ef-
fluent; this allows carbon regeneration or replacement to occur as needed.
The final rulemaking requires monitoring of carbon beds for breakthrough as
part of the standard's implementation procedures.
     In response to comments, EPA examined carbon adsorption design, opera-
tion, and performance data from a number of plants in a wide variety of
industries; and EPA has reexamined,  with the help of carbon manufacturers
and custom carbon adsorption equipment designers, the elements that affect
carbon adsorption efficiency.  This analysis1 has reinforced the Agency's
original conclusion that a well-designed, operated, and maintained adsorp-
tion system can achieve a 95-percent control efficiency for all organics
under a wide variety of stream conditions over both short-term and long-
term averaging periods.  The major factors affecting performance of an
adsorption unit are temperature, humidity, organics concentration, volumet-
ric flow rate, "channelling" (nonuniform flow through the carbon bed),
regeneration practices, and changes in the relative concentrations of the
organics admitted to the adsorption system.  The WSTF/TSDF process vent
stream characteristics are typically well within design limits in terms of
gas temperatures, pressure, and velocity for carbon adsorbers.  For
example, the adsorption rate in the bed decreases sharply when gas temper-
atures are above 38 °C (100 °F).  A review of plant field data shows that
these high-temperature streams are generally not found on WSTF/TSDF process
vents.  If high-temperature gas streams are encountered,  the gas stream can
be cooled prior to entering the carbon bed.  Also,  gas velocity entering
the carbon bed should be quite low to allow time for adsorption to take
place; the WSTF/TSDF stream flows are very low and, as a result,  bed depth
                                    7-19

-------
should not be excessive.  For cases where the organic concentration in the
gas stream exceeds the optimum range for use in carbon adsorption, the gas
stream can be diluted using nitrogen to achieve the desired organic
concentration (e.g., 25 percent of the LEL).
     In summary, EPA has concluded that, for WSTF/TSDF process vent
streams, carbon adsorption can reasonably be expected to achieve a 95-per-
cent control efficiency provided (1) the adsorber is supplied with an ade-
quate quantity of high-quality activated carbon, (2) the gas stream
receives appropriate conditioning (e.g., cooling, filtering) before enter-
ing the carbon bed, and (3) the carbon beds are regenerated or replaced
before breakthrough.  The EPA does not support a higher control efficiency
(i.e., 98 percent as opposed to 95 percent) for carbon adsorption units
applied to WSTF/TSDF process vents, particularly in light of the design
considerations related to controlling multicomponent vent streams when the
organic mix is subject to frequent change.
     With regard to commenters1 concerns that carbon adsorbers are emission
sources themselves, EPA agrees that there would be no environmental benefit
in removing organics from an exhaust gas stream using adsorption onto acti-
vated carbon if the organics were subsequently released to  the atmosphere
during desorption or during carbon disposal.  The Agency therefore expects
owners or operators of TSDF using carbon adsorption systems to control
organic emissions to take steps to ensure that proper emission control of
regenerated or  disposed carbon occurs.   For onsite  regenerable carbon
adsorption  systems, the owner or operator must account for  the emission
control of  the  desorption and/or disposal process in the control efficiency
determination.  In  the case of offsite  regeneration or disposal, the owner
or operator should  supply a certification,  to be placed in  the operating
record  of the TSDF, that all carbon removed from a  carbon  adsorption system
used  to comply  with Subparts AA and BB  is either:   (1) regenerated or
reactivated by  a  process that prevents  the  release  of organics to  the
atmosphere  (NOTE:   EPA  interprets  "prevents" as  used  in this  paragraph to
include the application of effective  control devices  such  as  those required
by these  rules),  (2)  incinerated  in a device that meets the performance
standards of  Subpart  0, or  (3) disposed in  compliance with Federal and
State regulations.
                                     7-20

-------
7.6  FEASIBILITY OF LEAK DETECTION AND REPAIR  (LDAR) PROGRAM
     Comment:  Commenter AESP-L0018 supports the LDAR program as proposed;
however, Commenter AESP-L0016 opposes the fugitive standards because they
fail to require the proper technology to control releases from pumps and
valves.  The commenter states that EPA should  set standards that require a
100-percent control, based on what available control (e.g., sealed bellows
valves, seal less pumps, or dual mechanical seals for pumps) can achieve.
The commenter believes that the degree of control represented by sealed
bellows valves should be required unless the owner or operator shows that,
for wastes treated at a particular facility, unacceptable rapid corrosion
would ensue or the controls are not suitable in the particular setting.
The commenter believes that superior emission  controls cannot be rejected
under RCRA solely because they are more costly than less effective con-
trols.
     Response:  Control technologies for fugitive emissions from equipment
leaks, as required by the proposed standards,  comprise the use of control
equipment, inspection of process equipment, and repair programs to limit or
reduce emissions from leaking equipment that handle streams with total
organic concentrations of at least 10 percent  by weight.  These control
technologies have been studied and evaluated extensively by EPA for
equipment containing fluids with 10 percent or more organics and are
similar to those required by national emission standards for chemical,
petrochemical, and refining facilities under the CAA.
     A monthly LDAR program would have been required by the proposed
standards for WSTF/TSDF pumps and valves.  Based on results of the EPA's
LDAR model, once a monthly monitoring plan is  in place, emission reductions
of 73 percent and 59 percent can be expected for valves in gas and light
liquid service,  respectively, and a 61-percent reduction in emissions can
be achieved for pumps in light liquid service.  For compressors,  the use of
mechanical seals with barrier fluid systems and control of degassing vents
(95 percent) are required,  although compressors are not expected to be
commonly used at WSTF/TSDF.  The use of control equipment (rupture disc
systems or closed vent systems to flares or incinerators) is the technical
basis for control of pressure relief devices.  Closed purge sampling is the
                                    7-21

-------
required control for sampling connection systems and is the most stringent
feasible control.  For open-ended valves or lines,  the use of caps,  plugs,
or any other equipment that will close the open end is required; these are
the most stringent controls possible.  Flanges and pressure relief devices
in liquid service are excluded from the routine LDAR requirements, but must
be monitored if leaks are indicated.  For operations such as those expected
at WSTF/TSDF, total reductions in fugitive emissions from equipment leaks
of almost 75 percent are estimated for the entire program.
     The EPA agrees with the commenter that the level of control required
by the LDAR program does not result in the highest level of control that
could be achieved for fugitive emissions from pumps and valves in certain
applications.   In some cases, there are more stringent, technologically
feasible controls.  For example, leakless equipment for valves, such as
diaphragm and sealed bellows valves, when usable, eliminates the seals that
allow fugitive  emissions; thus, control efficiencies in such cases are
virtually 100 percent as long as the valve does not fail.   In appropriate
circumstances,  pumps can be controlled by dual mechanical seals that would
capture nearly  all fugitive emissions.  An overall control  efficiency of  95
percent could be achieved based on venting of the degassing reservoir to  a
control device.
     With regard to leakless valves, the applicability of these types of
valves  is limited  for TSDF, as  noted by EPA in the proposal preamble.  The
design  problems associated with diaphragm valves are the temperature and
pressure  limitations of the elastomer used for the diaphragm.   It has been
found that  both temperature extremes and process liquids tend to  damage or
destroy the diaphragm in the valve.  Also, operating pressure constraints
will  limit  the  application of diaphragm valves to  low-pressure  operations
such  as pumping and product storage  facilities.
      There  are  two main disadvantages to sealed  bellows  valves.   First,
they  are, for the  most part, only  available commercially  in configurations
that  are  used for  on/off valves rather  than for  flow control.   As a  result,
they  cannot be  used  in all situations.  Second,  the  main  concern  associated
with  this type  of  valve  is the  uncertainty of  the  life of  the  bellows  seal.
The metal bellows  are subject to corrosion and fatigue under  severe
operating conditions.
                                     7-22

-------
     Over  150 types of industries are included in the TSDF community, and
EPA does not believe that leakless valves can be used in an environmentally
sound manner on the wide variety of operating conditions and chemical
constituents found nationwide in TSDF waste streams, many of which are
highly corrosive.  Corrosivity is influenced by temperature and such
factors as the concentration of corrosive constituents and the presence of
inhibiting or accelerating agents.  Corrosion rates can be difficult to
predict accurately; underestimating corrosion can lead to premature and
catastrophic failures.  Even small amounts (trace quantities) of corrosives
in the stream can cause corrosion problems for sealed bellows valves; these
tend to aggressively attack the metal bellows at crevices and cracks
(including welds) to promote rapid corrosion.  Sealed bellows valves
particularly are subject to corrosion because the bellows is an extremely
thin metallic membrane.
     At proposal, it was estimated that 20 percent of all plants process
halogenated compounds that tend to be highly corrosive.  The subsequently
obtained 1986 Screener Survey data show that, of the TSDF indicating
solvent recovery operations, at least 33 percent of the total handle
halogenated organics.  Furthermore,  of the 12 major chemicals determined
from site-specific data to be commonly occurring in WSTF waste streams, all
of the chemicals determined to be carcinogenic are halogenated (methylene
chloride, chloroform, and carbon tetrachloride).  Similarly,  of the 52
constituents in TSDF waste streams contributing  to the emission-weighted
unit risk factor, about 50 percent are halogenated and account for the vast
majority of the estimated nationwide emissions of carcinogens.  Thus, TSDF
are known to routinely handle and treat chemicals that may destroy sealed
bellows and diaphragm valves.
     The durability of metal bellows is highly questionable if the valve is
operated frequently;  diaphragm and bellows valves are not recommended for
general service.  EPA does not believe that the  application of sealed
bellows and diaphragm valves is technologically  feasible for all  TSDF valve
conditions or that their application would lead  to a significant reduction
in emissions and health risks.  Valve sizes,  configurations,  operating
temperatures and pressures,  and service requirements are some of the areas
                                    7-23

-------
in which diaphragm,  pinch,  and sealed bellows valves have limitations that
restrict service.   With regard to the emission reductions achieved by
sealed bellows,  diaphragm,  and pinch valve technologies,  these valves are
not totally leakless.  The technologies do eliminate the conventional seals
that allow leaks from around the valve stem; however, these valves do fail
in service from a variety of causes and when failure occurs, these valves
can have significant leakage.  This is because these valves generally are
not backed up with conventional stem seals or packing.  The EPA currently
is Devaluating the control efficiencies assigned to these technologies.
Because these leakless types of equipment are limited in their applica-
bility and in their potential for reducing health risks, EPA did not
consider their use as an applicable control  alternative for nationwide TSDF
standards.  The EPA has requested,  in a separate FEDERAL REGISTER notice
(54 FR 30228, July 19, 1989), additional  information on the applicability
and use of leakless valves  at TSDF.
      For pumps, the most effective  controls  that are technologically
feasible  (dual  seals)  in some cases  also  were not selected  as the basis  for
equipment-leak  standards.   The  impact  analysis  indicates that including
LDAR  results  in  less  emission  and  risk  reduction than does  including
equipment  requirements for pumps.   However,  the difference  in the emission
and health risk reductions attributable to  implementing  a monthly LDAR
program rather  than  the  more stringent equipment standards  for  pumps
appears to be small  in comparison  to the  results of the  overall  standards
 (about 5 percent).   The  overall standards,  including a  LDAR program for
pumps and valves, would  achieve an expected emission reduction  for  TSDF
 equipment leaks of about 19,000 Mg/yr (21,000 ton/yr).   The estimated MIR
 from equipment leak emissions would be reduced to  1 x 10'3 from 5 x 10'3
 based on the TSDF equipment leak emission-weighted unit risk factor; cancer
 incidence would be reduced to 0.32 case/yr from 1.1 cases/yr.  In compari-
 son,  including dual seals for pumps could achieve an additional fugitive
 emission reduction of about 1,200 Mg/yr  (1,320 tons/yr)  and an additional
 incidence reduction of about 0.06 case/yr.  The MIR, with leakless controls
 for  pumps, would be unchanged at 1 x 10'3.
                                     7-24

-------
     Given the small magnitude and the imprecise nature of the estimated
emission and risk reductions associated with including dual seals for pumps
in the overall standard, EPA considers the two control alternatives  (i.e.,
LDAR and dual seals) as providing the same level of protection.  The data
and models on which the risk estimates are based are not precise enough to
quantify risk meaningfully to a more exact level.  The data and models
include uncertainties from the emission estimates, the air dispersion
modeling, and the risk assessment, which involves unit risk factor,
facility location, population, and meteorologic uncertainties.
     The EPA considered these factors when deciding whether to require TSDF
to install dual seals on pumps to control air emissions rather than to rely
on monthly LDAR.  Considering the limited applicability of additional
equipment controls and the low potential for additional reductions in
health risks of applying equipment controls for valves at TSDF and the
estimated emissions and risk reductions if leakless equipment for pumps
were required, EPA is not requiring leakless equipment at this time.
     In Phase III (the multiphased approach to TSDF emission standards
development is described in response to comment 3.2.2), EPA will further
examine the feasibility and impacts of applying additional control technol-
ogy beyond the level required by today's standards.  For example, dual
mechanical seals may be an appropriate emission control method when applied
selectively to wastes with high concentrations of toxic chemicals.  In such
applications, the reduction in toxic emissions (and consequently the reduc-
tion in residual risk) may be. significant for select situations.
7.7  FEASIBILITY OF USING CONTROLS IN SERIES
     Comment:  Commenter AESP-L0016 states that EPA should evaluate carbon
adsorption in series with a condenser because condensers work best with
concentrated streams and carbon units work best with low-concentration
streams.  The two systems together could yield an overall  efficiency of 99
percent even if each unit were only 90 percent effective.
     Response:  The EPA did evaluate the feasibility of using a carbon
adsorption device in series with a condenser to control WSTF/TSDF process
vent emissions.  The objective of the analysis was to determine whether the
                                    7-25

-------
combination of control devices would yield an overall  control efficiency
greater than the 95 percent that is achievable through use of a single
device.  If, for example, a 99-percent overall control efficiency is
desired and it is assumed that the carbon adsorption system is capable of
achieving a 95-percent control efficiency in all cases (a reasonable
assumption for a properly designed, operated, and maintained system), then
a minimum efficiency of 80 percent would be required for the condenser in
series with the carbon bed.  However, the EPA condenser analysis conducted
for the WSTF model units shows that an 80-percent control was not achieved
in 16 of the 40 cases examined.  In 7 of the 40 cases, the analysis showed
that no appreciable condensation would occur, because the partial pressure
of the organic in the vapor phase  (a function of concentration) would be
too low to thermodynamically support a liquid phase.  Because the model
unit cases used in the analysis are considered representative of the wide
range of current WSTF operations,  EPA does not believe that condensers are
capable of achieving the required minimum control efficiency in all cases.
Therefore, the use of carbon adsorption and condensation in series to
achieve a 99-percent control is not considered technically feasible on an
industrywide basis and was not selected as the basis for the standard.
     In addition to the question of technical applicability  (see Section
7.4, Feasibility of Flares), EPA reviewed the health and environmental
impacts of secondary condensers/carbon adsorbers (reflecting a 95-percent
control) compared to flares and incinerators  (reflecting a 98-percent
control) as further information in deciding whether to require a 98-percent
control level.  Application of flares or  incinerators for WSTF process vent
emissions would be expected to reduce nationwide emissions from about 8,100
Mg/yr  (8,900 tons/yr) to about 890 Mg/yr  (980 tons/yr).  Condensers  and
carbon adsorbers would reduce the  emissions to  about 900 Mg/yr  (990
tons/yr).   Thus, an incremental emission  reduction of about  10 Mg/yr (11
tons/yr) results with flares  and  incinerators  (with the emission rate
cutoff in place).  The EPA's  estimates that  include flares or  incinerators
in the overall proposed  standard would reduce annual  cancer  incidence
attributable to process  vents  from 0.38 case/yr to about 0.02  case/yr  in
comparison  to  0.03 case/yr for  including  condensers in the overall  proposed
                                     7-26

-------
standard.  The estimated MIR associated with WSTF emissions would be
reduced from about 8 x 10'4 to about 2 x 10'5 with flares or incinerators
and 4 x 10'5 with condensers or carbon adsorbers.  Therefore, based on
available information and data, a 95-percent control appears to provide
essentially the same level of protection for public health as does a
98-percent control and is achievable by all WSTF processes.  For these
reasons, EPA selected the 95-percent control over a 98-percent control as
the basis of the standard for WSTF process vent emissions.
7.8  REFERENCE
1.   U.S. Environmental Protection Agency.  Carbon Adsorption for Control
     of VOC Emissions:  Theory and Full Scale System Performance.  Office
     of Air Quality Planning and Standards.  Research Triangle Park, NC.
     June 6, 1988.
                                    7-27

-------
                              8.0  COST IMPACTS

8.1  CARBON ADSORBER COST ESTIMATES
     Comment:  Commenter AESP-00016 questions the cost estimates for carbon
absorbers; at his facility, a carbon absorption unit costs about $25,000
and has operational costs that average $12,000/yr.  Other commenters ques-
tion the carbon adsorption costs and cost effectiveness because (1) a
device is needed for each vent (manifolding is not permitted due to the
potential for cross-contamination of new or recycled chemicals), and (2)
additional costs are incurred (the carbon must be regenerated or buried).
     Response:  Since proposal,  controls were evaluated for 40 model unit
cases (see Appendix C) representing ranges and combinations of solvent
physical properties, total flow rates, and organic concentrations in the
vent stream.  Secondary condensers were analyzed for each model unit as the
most likely preferred control technology.  For 25 of these cases (predomi-
nantly streams with low concentrations of low-boiling/high-volatility
compounds), it was determined that condensation will not achieve a 95-
percent organic reduction.  Both carbon canisters and fixed-bed regenerable
carbon systems were costed for each of these vent streams.  Appendix C
presents the cost estimates for each model unit case.  It was found that,
for a stream with an organic emission rate greater than 0.45 kg/h (1 Ib/h),
a carbon bed can achieve the same emission reduction at lower cost than a
carbon canister; for a stream with an organic emission rate less than 0.45
kg/h (1 Ib/h), a carbon canister is more cost effective.  The capital costs
of the fixed-bed regenerable carbon systems ranged from $97,300 up to
$202,000, and annual operating costs ranged from $40,200 to $43,500 (from
$33,100 to $43,100 when a recovery credit is included).  The capital cost
of a carbon canister was $1,050,  and annual operating costs ranged from
$7,890 to $24,800 (carbon canisters are not regenerated onsite, and a
                                     8-1

-------
recovery credit is not appropriate).  The fixed-bed regenerable carbon
system's operating costs include regeneration/disposal of spent carbon; the
carbon canister's operating costs include carbon replacement and disposal.
Therefore, the cost estimates for a fixed-bed, regenerable carbon system,
including the costs of spent carbon regeneration/disposal, are
substantially greater than the costs cited by Commenter AESP-00016.  Carbon
canister capital costs are much lower than are the costs cited, but the
range of operating costs spans the cited operational costs.  Because of the
facility emission rate limit of 1.4 kg/h (3 Ib/h) and 2.8 Mg/yra
(3.1 ton/yr), carbon canisters are less likely to be used than are fixed-
bed regenerable systems.
     With regard to the requirement of a control device for each vent, EPA
acknowledges that there are instances where vent manifolding is not allowed
because of potential product contamination (e.g., across units recycling
chemicals in Pharmaceuticals manufacturing).  However, the product has
already been recovered from the vents that are sources of organic emissions
to the atmosphere; therefore, manifolding of  the vent streams that emit to
the atmosphere should not  lead to a product contamination problem.  Conse-
quently, for the purpose of estimating the impacts of the standards,  it was
assumed that one control device would be needed per WSTF.  This was done  in
the absence of the site-specific information  needed to determine control
device requirements.  Although this may underestimate the control cost for
a facility that chooses to install carbon adsorption control devices  on
more than one vent, it  is  potentially a very  small underestimate.  For
example, the total annual  cost of a carbon canister is comprised almost
totally of annual operating costs, which are  directly proportional to the
emissions removed.  Thus,  the potential underestimate in  total annual  cost
resulting from  assuming one carbon  adsorber per facility  is not signifi-
cant.  Similarly, for the  model  unit fixed-bed, regenerable carbon systems,
total annual operating  costs  ranged from 57 to 72 percent of total annual
costs.  Furthermore, the  addition of the emission  limit to the rules  based
on  the  facility  organic emission rate  lessens the  likelihood that  a  facil-
ity will  need  to  control  multiple process vents to  attain the  allowable
organic emission  rate of  1.4  kg/h  (3 Ib/h) and 2.8  Mg/yr  (3.1  ton/yr).
                                      8-2

-------
8.2  CONDENSER COSTS
     Comment:  Commenter AESP-00016 questions the overall credit of
$620,000 in nationwide costs for secondary condensers.  According to the
coraroenter, the assumption that secondary condensers will more than pay for
themselves is based on an estimated increased yield of 20 to 30 percent.
The commenter alleges that a primary condenser may provide such an
increase, but the yield of a secondary condenser would not exceed 1 per-
cent.  Considering the labor costs for maintenance, he challenges the esti-
mated credit and states that secondary condensers actually would result in
substantial costs.  Other commenters also question the costs because a
device is needed for each vent to prevent cross-contamination due to mani-
folding of streams.  Commenter AESP-L0018 states that the costs do not
account for the more sophisticated systems needed in high-humidity areas to
allow for equipment deicing or prewater removal.
     Response:  Because of commenters1 concerns regarding the estimated
condenser yields and the requirement for more sophisticated systems in
high-humidity areas, a state-of-the-art computer process simulator known as
ASPEN (Docket No. F-90-AESF-FFFFF, item S0012) was used in the final
analyses of condenser design and cost for the model unit cases representing
ranges and combinations of vent stream characteristics.  The ASPEN
condenser configuration consists of (1) a floating head, one-pass, shell -
and-tube heat exchanger, (2) a refrigeration unit capable of producing
chilled brine at a temperature of -28 °C (-20 °F), and (3) an optional
primary water-cooled heat exchanger.  This final item might be necessary in
some instances of organics condensation to reduce the size of the
refrigeration unit and to remove water vapor in order to avoid freezing
problems because the condenser temperature is low enough to cause ice
buildup on heat transfer surfaces.  Therefore, cost estimates generated by
ASPEN account for water removal.
     Model unit condenser cost estimates are presented in Appendix C.  For
the large model unit cases (3.9 L/s or 8.3 scfm total  flow rate),  total
annual cost with recovery credit ranged from a credit of $4,980 up to a
cost of $2,800.  For the medium model  unit cases (0.6 L/s or 1.2 scfm total
flow rate), total annual cost with recovery credit ranged from $630 up to
$2,120.   For the small  model unit cases (0.3 L/s or 0.6 scfm total flow
                                     8-3

-------
rate), total annual cost with recovery credit ranged from $1,770 up to
$2,030.
     The model unit control  cost estimates and the WSTF industry profile
were used to generate nationwide control cost estimates of implementing the
process vent regulations (see Appendix C).  The cost estimates are for 73
large facilities and 167 medium facilities.  The 208 small facilities (less
than 189,000 L [50,000 gal]  throughput/yr as defined in the final analysis)
would not be required to install emission controls under the facility
process vent organic emission rate limit.  Because there was insufficient
site-specific information available to determine which facilities could
apply condensation rather than carbon adsorption, upper and lower bound
cost estimates were generated for the 95-percent emission reduction control
option.  The upper bound cost estimate is based on the assumption that
fixed-bed, regenerable carbon adsorption systems would be required to
control process vents at all facilities with organic emissions above the
emission rate limit.  Similarly, the lower bound cost estimate is based on
the assumption that condensers could be used to control process vents at
all facilities with emissions above the emission rate limit.  The range in
estimates of nationwide total annual cost is from a credit of $68,000 up to
a cost of $12.9 million, assuming the installation of one control device
per facility with  emissions above the limit.
     As is discussed in the response to comment 8.1, multiple control
devices should not be required to prevent product contamination.  In addi-
tion,  available site-specific information indicates that the process emis-
sions  are released through one vent per  recycling process and that many
facilities only have one recycling process unit.  Because emission esti-
mates  are based on total facility waste  solvent throughput data, the
assumption of one  vent does not significantly  affect the overall analysis.
Furthermore, the addition of the emission limit to the rules based on the
facility organic emission rate  lessens  the likelihood that a facility will
need  to control multiple process vents  to attain the allowable organic
emission  rate of 1.4 kg/h (3 Ib/h) and  2.8 Mg/yr  (3.1 ton/yr).   Therefore,
it was assumed that one control device  would  be needed per WSTF  with emis-
sions  above the limit and that  facility  process emissions would  be mani-
                                     8-4

-------
folded and vented to the control device if control of more than one process
vent were required.
8.3  COMPLIANCE COSTS
     Comment:  Commenter AESP-00019 disagrees with the estimated compliance
costs provided at 52 FR 3765 (February 5, 1987) of the preamble.  The com-
menter estimates that it will cost more than $160,000 for distillate
receiver controls alone, exclusive of the three distillation units for
which secondary condensers already have been added.  Commenter AESP-L0003
estimates a per-unit capital cost of between $500,000 and $750,000 per
facility compared to EPA estimates of $30,200 to $174,500; per-unit operat-
ing costs per facility are estimated at $60,000 to $100,000 compared to EPA
estimates of $13,600 to $115,000.  Although the company operating costs
estimates are within EPA's estimated ranges, the commenter believes that
the EPA ranges are too broad (a factor of 10) to be realistic.
     Response:  In response to a request for additional information rela-
tive to the cost estimate of $160,000 for distillate receiver controls,
Commenter AESP-00019 submitted supporting documentation.  The cost estimate
is based on a system that would provide a vent condenser on each storage
unit in the commenter1s system.  There are 39 such storage units for which
condensers are estimated to range in cost from $2,000 to $5,000 per unit
(similar to the range in estimated condenser capital costs of $1,200 to
$5,700 for small and medium model unit cases).  The cost estimate of
$160,000 represents an average cost of $3,500 per unit plus the cost of a
utility system needed to support the operation.  Although the capital cost
estimates for 39 vent condensers appear to be reasonable, this regulation
only covers distillate receivers releasing emissions vented from the dis-
tillation column.  Therefore, storage tank vents that are not venting emis-
sions from the distillation process are not covered by these rules.
In addition,  control of process vent emissions is only required if facility
process vent  organic emissions exceed 1.4 kg/h (3 Ib/h) and 2.8 Mg/yr
(3.1 ton/yr).
     With regard to the costs referenced by Commenter AESP-L0003,  the com-
pliance costs  for a facility are the costs of controlling process vent
emissions and  the costs of controlling fugitive emissions from equipment
                                     8-5

-------
leaks.  As is discussed in the response to comment 8.1,  in the post-propo-
sal analyses of feasible control  technologies,  vent controls were evaluated
for 40 model unit cases representing ranges and combinations of solvent
physical properties, total flow rates,  and organic concentrations in the
vent stream.  The sources of data used  to define the model unit cases are
described in response to comment 5.1.  For each of the 40 model unit cases,
a chemical engineering process simulator known as ASPEN was used to esti-
mate the control costs for condensation.  For cases where condensation
would not achieve a 95-percgnt control  in organic reduction, carbon
adsorption also was costed using the EAB Control Cost Manual (Docket No. F-
90-AESF-FFFFF, item S0014).  Details of the cost estimates are presented in
Appendix C.
     The costs cited by Commenter AESP-L0003 represent compliance costs for
large and mid-sized recycling model units as specified in the post-proposal
analysis  (Appendix C).  Table 8-1 presents lower bound and upper bound
estimates of mid-size  recycling model unit compliance costs.  The ranges
are not as broad as the ranges estimated at proposal because the lower
bound and upper bound  process vent compliance cost estimates represent  the
application of condensation and fixed-bed, regenerable carbon adsorption,
respectively.  Either  condensation or carbon adsorption  is judged to be
feasible  in all cases; therefore, flares and incinerators are not included,
although  their use  is  not precluded  by  the standard.  The ranges of cost
estimates are still broad and would  be  unacceptably broad if meant to
represent the range of compliance costs  for a  specific facility.  However,
the  range in estimated compliance costs  for process vents reflects the
diversity of constituents and operating  characteristics  across  WSTF
nationwide,  and  therefore is  realistic.
8.4   COMPLIANCE  COSTS  FOR LDAR  PROGRAM
      Comment:   Commenter AESP-00017  questions  the estimated compliance
costs for implementation of the  proposed LDAR  program.   The commenter  cites
costs at  one plant  with  180 valves,  2  pumps,  and 1 rupture  disc of  $12,000
 for setup (including  $2,000 for instrumentation and the  rest for manpower
 costs), plus $2,500/yr for manpower for monthly tests.   Five monthly tests
 have found no leaks and one pump seal  failure.  At another site,  the Table
                                      8-6

-------
 TABLE 8-1.  ESTIMATES OF MID-SIZED RECYCLING MODEL UNIT COMPLIANCE COSTS

                              	Per plant cost	
                              Process       Fugitive
                              control3      control"       Total
Lower bound
Capital cost, 1986 $
Annual cost, $/yr
Upper bound
Capital cost, 1986 $
Annual cost, $/yr
8,740
(2,820)

101,000
51,300
26,970
11,910

26,970
11,910
35,710
9,090

127,970
63,210
aThe lower bound cost estimates for process vent control represent the
 average costs of applying a condenser to model unit cases 6, 7, 10, 11,
 and 12.  The upper bound cost estimates represent the average costs of
 applying a fixed-bed, regenerable carbon adsorption system to model unit
 cases 5, 8, and 9.  In Appendix C, the model unit cases are specified in
 Table 1 of the January 26, 1988, memorandum from Robert Zerbonia, RTI, to
 Rick Colyer, EPA/SOB, and Robert Lucas, EPA/CPB, regarding model unit
 condenser in carbon adsorber cost estimates.

bln Appendix C, the fugitive control costs are displayed in Table 2 of the
 September 22, 1988, memorandum from Robert Zerbonia, RTI, to Rick Colyer,
 EPA/SDB, and Bob Lucas, EPA/CPB, regarding costs for fugitive emissions
 control at WSTF/TSDF model units.
                                    8-7

-------
initial setup cost was $550,000 (including $400,000 for scrubbers, instru-
mentation, and piping changes and $150,000 for labor).  The annual cost is
estimated at $16,000 for manpower to run the tests.  These costs are for
two different operating departments; together they have 1,350 valves, 40
pumps, and 10 relief valves.  During 2 years of compliance, 2 leaking
valves and 19 pump seal leaks have been detected.
     Response:  The facility compliance costs for the first site cited by
the commenter are in line with the estimated compliance costs for implemen-
tation of the LDAR program.  The estimated compliance costs are based on
unit capital and annualized cost factors for each type of fugitive emission
source.  The unit costs were applied to model unit equipment counts to
generate facility cost estimates.  The mid-sized recycling model unit con-
tains a total of 173 emission sources, including 121 valves, 5 pumps, and 3
pressure relief valves, which is a similar arrangement to the sources at
the first plant mentioned by the commenter.  The estimated capital costs
for this facility are $27,000, and the estimated total annualized costs are
$11,900 (including amortization of capital costs); both of these costs are
greater than the costs cited by the commenter by factors of more than 2.
     The equipment counts provided by the commenter for the second site are
almost an order of magnitude greater than the mid-sized recycling model
unit equipment counts and nearly four times as large as the large-sized
recycling model unit used in the analysis.  In this case, the capital costs
are more than two times greater than the capital costs estimated  for the
large  recycling model unit,  increased in proportion to the same equipment
counts.  However, the capital costs include $400,000 for scrubbers,  instru-
mentation,  and piping changes.  Because vent controls such as scrubbers are
not required  for  implementation of  an LDAR program, most of the capital
costs  cited by the commenter would  not be  incurred as a result of this
rulemaking.   Subtracting the costs  for scrubbers and piping changes would
bring  the setup costs  in line with  the large model unit's  capital costs,
taking the  differences  in equipment counts  into  consideration.  The  annual
costs  cited by the commenter are  less than one-half the estimated propor-
tional model  unit's  annual  costs, with capital equipment amortization  fac-
tored  out.
                                      8-8

-------
      Therefore, for  the  two  plants  for which the commenter  has  provided
  information on equipment counts  and compliance costs, the model unit's
  estimated capital costs  are  more than twice the cited capital costs  in one
  case and comparable  to the cited capital costs in the other case, after
  adjusting for differences in equipment counts and controls  not  required for
  an  LDAR program.  Estimated  annual  costs are more than twice the cited
  annual costs in both cases.  Because of the diversity of characteristics of
  TSDF/WSTF nationwide, the model  unit's cost estimates are not expected to
* agree closely with the actual control costs incurred by specific facili-
  ties.  However, the model units  are based on aggregated data from numerous
  facilities and provide a sound basis for estimating nationwide  control
  costs.
  8.5  RECOVERY CREDIT FOR TSDF
                                                 *
      Comment:  Commenters AESP-00018, AESP-L0003, and AESP-00024 do  not
  believe that a recovery  credit for  the value of captured emissions is
  applicable to TSDF.  Without a recovery credit, the costs will  be higher
  for TSDF than for WSTF,  and  this  merits adoption of a less  burdensome
  standard for TSDF according  to Commenter AESP-00018.  Commenter AESP-00024
  states that no recovery  credit should be applied because there  is no bene-
  fit of capturing the nonvolatiles or low organics included  in the defini-
  tion of "VHAP."  If the  definition  is not changed to reflect the concerns
  associated with recycling, then  no  recovery credit should be applied.
      Response:  The EPA  agrees with Commenters AESP-00018,  AESP-L0003, and
  AESP-00024 in that a recovery credit is not applicable to TSDF  in general
  because most of the hazardous wastes handled at a TSDF are  destined for
  disposal.  In contrast,  at a WSTF,  the air emissions resulting from
  equipment leaks are potentially  recyclable solvents.  Thus,  no recovery
  credit was applied for TSDF other than WSTF in the analyses for the final
  equipment leak standards.  Although it is also true that in the absence of
  a recovery credit the costs will  be somewhat higher for TSDF than WSTF,
  because these standards  are not based on costs,  there is no basis for
  setting a less stringent standard for TSDF.  Regarding the concerns of
  Commenter AESP-00024 with the definition of VHAP,  as is explained in
  response to comment 3.2.6,  the equipment-leak standards have been revised
  to take component volatility into consideration.   Also it should be noted
                                      8-9

-------
that to clarify the applicability of the standards, VHAP has been dropped
from the promulgated rule.
8.6  COST EFFECTIVENESS OF PROCESS VENT CONTROL TECHNIQUES
     Comment:  Commenter AESP-00009 asks if EPA has performed any study of
the cost effectiveness of incineration and carbon absorption for reducing
VOC on a ton-per-year basis.  Commenter AESP-L0018 asserts that, assuming
the controls are appropriate from a risk standpoint, EPA should ensure the
controls are cost effective for WSTF.  Commenter AESP-00019 states that the
BID does not include cost estimates for the control of product accumulator
vessels (including distillate receivers, condensers, bottoms receivers,
product separators, and hot wells).
     Response:  In the post-proposal analysis, the feasibility and costs of
applying condensation, carbon adsorption, and incineration were evaluated
for WSTF process vent streams.  Ranges of vent stream and organic flow
rates were used in the analyses, based on a review of available WSTF site-
specific data (see responses to comments 5.2 and 5.3).  Vent emission meas-
urements from primary and secondary condensers, condensers vented to tanks,
and vacuum distillation vents were reviewed.  Because the regulations only
apply to tank vents releasing uncondensed organics from the distillation
column overhead, the emission data reviewed are representative of the wide
ranges of flow rates and organic emission rates from these emission
sources.
     Figures 8-1, 8-2, and 8-3 present graphs of the organic emission rate
(kg/h) versus the cost effectiveness ($/Mg) of condensers, carbon adsorb-
ers, and incinerators, respectively.  In each case, the cost effectiveness
is fairly constant for organic emission rates above 0.45 to 1.4 kg/h (1 to
3 Ib/h), but rapidly worsens as the organic emission rate drops below 0.45
to 1.4 kg/h  (1 to 3 Ib/h).  Because the three types of add-on controls are
equally protective, the choice of a control is expected to be made on the
basis of cost effectiveness.  From this perspective, condensation should be
the preferred type of control, with carbon adsorption used on vent streams
for which condensation will not achieve an adequate degree of control.
Incineration might be used by a facility with a hazardous waste incinerator
                                    8-10

-------
00
I
               o>
               I
A  w"

II
UJ

8
u
      10



       9



       8



       7



       6




       5
                      2




                      1




                      0




                     -1
                  I
                                                                23


                                                                  Emission Rate (kg/h)


                                             Figure 8-1. Emission rate vs. cost effectiveness (condensers @ 95%).

-------
OD
 I
t-»
ro
                 O)
                   •s
                       90
                       80
                       70
                       60
                       50
•p 3

  1  40
                       30
                       20
                       10
                                                                    1
                                                                         1
                                                                     2                    3

                                                                       Emission Rate (kg/h)
                                                 Figure 8-2. Emission rate vs. cost effectiveness (carbon adsorption).

-------
CO


»-•

€*>
                        6
II
H-     o
l|l     O
                   o
                   o
                                                                                         J_
                                                                    2                     3


                                                                       Emission Rate (kg/h)
                                                 Figure 8-3.  Emission rate vs. cost effectiveness (incinerators @ 98%).

-------
already onsite or by facilities that wish to use incineration to dispose of
other hazardous wastes.
8.7  COST OF A FIXED 95-PERCENT EMISSION REDUCTION
     Comment:  Commenters AESP-00002 and AESP-00006 state that the require-
ment of a fixed 95-percent emission reduction would be more costly for
small operations compared to equipment (facilities) emitting higher levels
of waste organics.
     Response:  The EPA agrees that the proposed requirement of a fixed
95-percent emission reduction could be more costly for small operations
than for operations emitting higher levels of waste organics, both in terms
of the small facility's ability to pay the control cost and in terms of the
cost effectiveness of controls.  The facility process vent organic emission
rate limit of 1.4 kg/h  (3 Ib/h) and 2.8 Mg/yr (3.1 ton/yr), based on the
health risk analysis for process vents (see response to comment 7.1), that
has been added to this  rulemaking does not require small facilities to
control process vent emissions.  The emission rate limit thus has the
effect of "exempting" facilities for which installing emission controls
would not be cost effective from the process vent requirements.  As the
graphs presented  in the response to comment 8.6  illustrate, for condensers,
carbon adsorbers, and incinerators, the cost effectiveness  ($/Mg) is fairly
constant for organic emission  rates above 0.45 to 1.4 kg/h  (1 to 3 Ib/h),
but  rapidly worsens as  the organic emission rate drops below 0.45 to 1.4
kg/h  (1 to 3 Ib/h).
                                     8-14

-------
                               9.0  ECONOMIC IMPACTS

*  9.1  SMALL BUSINESS IMPACTS
        Comment:   Three commenters (AESP-00016,  AESP-00036,  and AESP-00009)
   object to the  preamble certification that the proposed standards will  not
   have a significant impact  on a substantial  number of small  business enti-
   ties.   In support,  Commenter AESP-00016 states that all  but three or four
   member companies  in the NASR are small  businesses with an average of fewer
   than 30 employees.   Commenter AESP-00036 says that 90 percent of all com-
   mercial recycling facilities in the United  States are family-owned busi-
   nesses.  Commenter AESP-00009 states that the typical recycler has one or
   two stills and employs fewer than 100 people.  Commenter AESP-00016
   requests to review and comment on the documentation used to support the
   certification  and states that EPA should perform a Regulatory Flexibility
   Analysis.   Commenters AESP-00002,  AESP-00005, and AESP-L0018 generally feel
   that the standards  would have an adverse economic impact on small  busi-
   nesses and would  have a significant impact  on large businesses.
        Response:  After completing the revised  estimates of the compliance
   costs  resulting from the regulation,  EPA examined whether the regulation
   would  have "a  significant  economic impact on  a substantial  number of small
   entities."   (See  "Economic Impact of Air Emissions Regulations:   Waste
   Solvent Recycling,"  Docket No.  F-90-AESF-FFFFF,  item S0046.)   A  "substan-
   tial  number" is generally  thought to imply  greater than  20  percent of  the
   small  entities, although this is not a  fixed  rule.   A "significant economic
   impact"  is  said to  occur whenever any of the  following criteria  is
   satisfied:
        •     Annual  compliance costs  (including  annualized  capital,  oper-
             ating,  and  reporting  costs) increase total  costs  of produc-
             tion  for  small entities  for the relevant  process  or product
             by more than  5 percent.
                                       9-1

-------
     •     Compliance costs  as  a  percent  of  sales  for small  entities
          are at least  10 percent  higher than  compliance costs  as  a
          percent of sales  for large entities.
     •     Capital costs of  compliance represent a significant portion
          of capital available to  small  entities, considering internal
          cash flow plus external  financing capabilities.
     •     The requirements  of  the  regulation are likely to result  in
          closures of small entities.
     An estimated 450 recycling facilities  will be directly affected by
this rule.  From financial  data bases, information was obtained regarding
the number of employees and the company's sales for the affected facil-
ities.   Using the Regulatory Flexibility Act guidelines, these data were
compared to the Small Business Administration criteria that apply to each
Standard Industrial Classification (SIC) code to determine the extent of
small firms.  It should be noted that the criteria are focused on the firm,
not the facility.  For some of the facilities, it was impossible to find
any information about their sales or number of employees.  Therefore, to
bound the estimates of the number of small  businesses, it was assumed that
the firms for which no information is available  are either all small busi-
nesses or none  (small business) at all.  Using this sensitivity analysis
approach, a  range of 12 to 146 captive  facilities are owned by small
businesses;  a range of 0 to 70 commercial  facilities are owned by small
businesses.
     Faced with  higher recycling costs,  firms that own captive facilities
will presumably  choose the least-cost option  between  incurring those costs,
using offsite management,  or  changing their production process to reduce
generation of spent  solvents.  While some  firms  may decide to shut  down
their recycling  units,  this is not  expected to result  in much of  an impact
on  the overall  activities  of  the  firm.   The projected  product price
increases in Table  9-1  reflect the  insignificant component of product  costs
represented  by  recycling.  Given  price  changes of much less  than  0.1 per-
cent,  it  is  reasonable to  conclude  that captive facilities owned  by small
businesses  will  not be disadvantaged as compared with larger firms.
     The  price  increases are  larger in  the commercial  sector in which  the
service  provided is recycling rather than, for example,  the  production of
                                      9-2

-------
               TABLE 9-1.   PRICE AND QUANTITY  ADJUSTMENTS  IN THE MARKETS  FOR  GOODS AND SERVICES BY GENERATING SECTOR,
                                              EQUIPMENT LEAK FUGITIVE  EMISSION CONTROL
U>
Low-cost process vent
•missions controls
Sector
Mining
Grain and textile mill products
Furniture, paper products, printing
Industrial chemicals, inorganic and organic
Plastics, fibers
Biological, pharmaceutical, medical chemicals
Assorted chemical products
Paint and allied products, petroleum and coal
Rubber, plastics
Cement companies
Primary metals
Metal fabrication
Nonelectrical machinery
Electrical machinery and supplies
Transportation equipment
Instruments
Miscellaneous manufacturing
Electric and gas utilities
Nondurable goods: wholesale sales
Research laboratories, hospitals, universities, consultants
Change In
quantity
produced, X
-0.00069
-0.00423
-0.00083
-0.00460
— 0 * 00068
-0.00894
-0.00866
-0.00038
-0.00010
-0.00044
-0.00066
-0.00044
-0.00062
-0.00184
-0.00066
-0.02286
-0.00691
-0.00116
-0.00051
-0.00179
Change in
price, X
0.00086
0.00606
0.00118
0.00687
0.00423
0.01004
0.00994
0.00164
0.00074
0 • 00083
0.00080
0.00063
0.00276
0.00176
0.00066
0.03266
0.00616
0.00061
0.00073
0.00061
High-cost process vent
emissions controls
Chsnge in
quantity
produced, X
-0.00074
-0.00624
-0.00102
-0.00670
-0.00084
-0.01107
-0.01072
-0.00047
-0.00012
-0.00066
-0.00069
-0.00064
-0.00066
-0.00228
-0.00068
-0.02832
-0.00732
-0.00144
-0.00063
-0.00221
Change in
price, X
0.00106
0.00749
0.00146
0.00861
0.00624
0.01244
0.01232
0.00203
0.00091
0.00078
0. 00099
0.00078
0.00340
0.00217
0.00082
0.04046
0.00762
0.00076
0.00090
0.00076

-------
industrial chemicals.  Such increases range from 1.76 to 2.18 percent.
However, recycling is only one of many hazardous waste management services
offered by many of these facilities.  In that case, it is unlikely that
decreased recycling will result in shutting down the facility.  In some
cases, however, recycling is the only form of hazardous waste management
service offered by the facility.  In such instances, some very small-volume
facilities may be uneconomical.  Whether these facilities will actually
close is unclear.  There are several mitigating reasons for this:  (1) The
volume of recycling is expected to increase over time.  Thus, at the same
time that the regulation is decreasing the supply of commercial recycling
services, the demand for these services is increasing.  The net effect on
volume recycled is impossible to predict.  (2) In fact, reductions in
recycling are likely to be distributed across all the facilities in the
industry, rather than being concentrated in only the facilities operating
very small units.  Thus, all the firms would decrease their recycling
activities slightly, and no facilities would be shut down.  (3) The
permitting process is long and costly; a permit represents an asset to the
firm, which would be made more valuable as the price of recycling services
increases.  The firm may decide not to sacrifice the permit by shutting the
facility down.  (4) The costs of closure can be very substantial.  Thus,
whereas a firm may be adversely affected by the increased cost of
recycling, it will probably be reluctant to shut down the facility even if
it is temporarily uneconomical to operate.
9.2  CLOSURES AND OTHER ECONOMIC IMPACTS
     Comment:  Numerous commenters stated that the proposed rules will
cause closures and ask  if EPA has evaluated this economic impact and the
additional economic impacts not included in the cost estimates, such as the
cost of more insurance  and the additional costs that will be  incurred for
becoming repermitted at the State and Federal level to obtain approval for
the previous exemption  for recycling.  Commenter AESP-00009 cites a cost of
over $1 million to reopen with new permits.  Commenters AESP-00009 and
AESP-00016 request EPA  to evaluate the economic impact of lost recycling
capacity, including  (1) decreased production, increased material remaining
as natural resources for future use, decreased imports, and decreased
                                     9-4

-------
material for waste  disposal.  Commenters AESP-00002, AESP-00005, and AESP-
00009 claim that the economic impact will discourage recycling, will put
commercial recyclers at a competitive disadvantage with onsite and unper-
mitted recycling operations, and will increase the volume of solvent into
the alternative fuels program at the expense of recycling.  Commenter
AESP-L0003 states that, if higher costs of emission control are imposed on
a TSDF reclaimer, then materials will shift to non-TSDF reclaimers without
emissions controls.  This is the opposite of the Agency's intent and will
cause TSDF to shift material to less efficient units.
     Response:  With regard to comments concerning the costs of
re-permitting, the standards have been revised to require the inclusion of
units affected by these rules only at permit reissue under Section 124.15
or review under Section 270.50, provided the permit has been issued prior
to the effective date (promulgation date plus 6 months) of these rules.
Therefore, the additional costs associated with re-permitting are expected
to be minimal.
     In response to comments regarding the economic impacts on recycling,
EPA performed an economic analysis (Docket No. F-90-AESF-FFFFF, item S0046)
using the control costs described in Chapter 8.0.  The unit costs of
control based on average process vent compliance costs exhibit significant
economies of scale.  Small recycling facilities (facilities recycling up to
189,000 L [50,000 gal/yr]) are estimated to cost $27/Mg of recycling
throughput to control; large facilities are estimated to cost $'ll/Mg of
recycling throughput.  The compliance costs are associated with an increase
of 1.76 to 2.18 percent in the price of commercial waste solvent recycling
services and a 1.59- to 2.27-percent decrease in the quantity of these
services, assuming the implied waste minimization is economically feasible.
The increases in the prices of goods and services where solvents are used
as a productive impact are virtually nonexistent.  Spent solvent management
costs represent less than 0.1 percent of production costs for all
generators of spent solvents.
     Because of the economies of scale in compliance and because onsite
recycling facilities tend to be smaller (119 Mg median volume recycled)
than commercial facilities (1,103 Mg median volume recycled), some shift in
waste solvent recycling from onsite to offsite commercial facilities is
                                     9-5

-------
projected.  This shift, if it does occur, would potentially be accompanied
by the closure of a large number (around 195 units) of very small recycling
units, with an associated decrease of between 10 and 12 jobs  (full-time
equivalent employees per year).  It is important to note, however, that
these are not plant closures.  The closure would represent taking off-line
small volume distillation units within a plant that produces, for example,
industrial chemicals.
     Furthermore, there would be a number of legal and economic reasons why
captively owned recycling facilities would not close even if  the analysis
performed indicated that they would.  For example, there may  be insuffi-
cient commercial capacity located near the plant, so that transportation
costs would make shipping offsite uneconomical.  Or the firm  may decide it
can reduce its legal risk by recycling the spent solvent onsite.  Simi-
*
larly, a firm may feel that, for a small volume of throughput, the advan-
tages of having control over scheduling the recycling operation outweigh
the cost savings of using a commercial recycler.  Thus, unit  closures at
captive plants are probably overestimated by some unknown amount.  Recyc-
ling unit costs for captive recyclers represent only a very small fraction
(averaging less than 0.03 percent for captive recyclers) of their total
production costs.  The increase in recycling costs resulting  from the regu-
lation may in some cases cause a facility to shut down its recycling unit,
but it will have virtually no impact on the production activities of cap-
tive facilities.  (Sector output of goods and services decreases by at most
0.03 percent in response to the regulation.)  Because of the  existence of
economies of scale in compliance, the commercial recycling sector is pro-
jected to experience a smaller decrease in output than the onsite recycling
sector.
                                     9-6

-------
                             10.0  TEST METHODS

10.1  MEASUREMENT OF THE 10-PERCENT TOTAL ORGANICS
     Comment:  Five commenters (AESP-00013, AESP-00019, AESP-00024, AESP-
00032, and AESP-L0001) provide their views on the test methods proposed for
determining applicability of the equipment leak standards.  Commenter
AESP-00013 states that volatile compounds nonreactive in the formation of
ozone should not be included in calculation of the stream's organic
content; the measurement method should be for volatile chemicals and not
total organics.  Commenter AESP-00019 recommends using a total hydrocarbons
measurement because equipment that handle chlorinated hydrocarbons also
should have to comply with the standards.  Commenter AESP-00024 recommends
Method 8240 of SW-846, which detects both volatile and relatively low-
volatility compounds, for determining applicability over Method 9060, which
detects nonvolatile or nonhazardous materials and would be costly and
duplicative.  Commenter AESP-00032 agrees in that no conclusions can be
drawn regarding the concentration of VOC in a waste using a TOC analysis.
The commenter points to the inconsistency of requiring additional monitor-
ing if a TSDF managed a hazardous waste with constituents in the low parts
per million ranges while not regulating up to 10 percent of the pure
constituent in an aqueous waste based on TOC, an unrelated parameter.
Commenter AESP-L0001 agrees with the other commenters that the test method
should consider the volatility of the wastes.
     However, Commenter AESP-L0001 notes than no analytical method cur-
rently exists that will measure 10-percent TOC in both a solid and liquid
waste.  He states that Method 9060 determines the TOC of water and waste
via the conversion of all organic carbon to carbon dioxide, making no dis-
tinction as to the types of organic compounds present (e.g., volatile vs.
nonvolatile organics).  Many solid or water-insoluble viscous liquid wastes
                                    10-1

-------
cannot be tested by this method.  Method 8240 is designed to measure only
selected VOC in water and wastes; use of this method would require develop-
ment of compound-specific criteria.  Extensive interlaboratory studies also
would be needed to determine the precision and accuracy at the high concen-
trations (i.e., 10 percent) at issue in the standards because the method is
designed to measure concentrations in the parts per million range.
Finally, TOC is not equivalent to total hydrocarbon because of the presence
of hydrogen, oxygen, or other heteroatoms in the organic molecules.  Alter-
natively, Method ASTM D 2267-68 does not provide a measure of total hydro-
gen, but only the aromatic components of petroleum-derived fuels.  Methods
ASTM E  169-63, E 168-67, and E 260-73  provide general guidance to users of
ultraviolet  (UV), infrared (IR), and gas chromatography  (GC) techniques,
but these practices do not contain any detailed methods  for determining TOC
or specific  organic compounds.  The commenter explains that determining
total VO would be a more realistic measure of emission potential than total
organics, but existing methods  for VO are compound-specific.  The commenter
encourages  EPA to continue developmental efforts for  a total VO method  and
apply  it.
      Response:   The determination  of applicability  was intended to be based
on the  total organic content of the waste stream.   The standards cover
organic emissions  and  are  intended to  reduce impacts  on  human health  and
the  environment  associated with total  organics.  The  basic  reason  for
measuring total  organics  is  that the original  collection of data  used to
support the standards  measured  total organics.   In  practice, most  organic
wastes  and  their derivatives  affected  by these standards are considered
volatile.   In  addition,  the  promulgated equipment  leak standards  have been
drafted to  distinguish between  sources containing  or coming into  contact
with wastes which are  more or less volatile, i.e.,  the  light versus  heavy
 liquid criterion.  If  the total organic content of the waste stream never
 equals or exceeds 10 weight  percent,  no controls are required.   Commenters
 AESP-00019, AESP-00024,  AESP-00032,  and AESP-L001  noted  that the various
 proposed test methods  all  had limitations and none were  universally
 applicable.  The EPA recognized this and therefore proposed the use of
 several test methods.   The determination of applicability should not
                                     10-2

-------
require precise measurement of the 10 weight percent total organic in most
cases.  The EPA anticipates that most waste streams will have an organics
content much lower or much higher than 10-percent.  Furthermore, because
the regulation requires control if the organic content of the waste stream
ever equals or exceeds the 10-percent value (Sections 264.1063[d] and
265.1063[d]), EPA believes that owners or operators will not claim that a
waste stream is not subject to the requirements of the standard based on a
sample analysis with results near 10 percent.  The precision of the
measurement results will be considered in determining if the organic
content of the waste stream is ever expected to exceed the 10 percent value
and is thus not critical.
     If the facility does decide to test the waste, the choice of the
appropriate method must be based on a knowledge of the process and waste.
The EPA has prepared a guidance document that includes information to aid
TSDF owners/operators and enforcement and permitting personnel in
implementing the regulations.  Additional detail is provided in the
guidance document to aid in choosing the most appropriate test method
(refer to "Hazardous Waste TSDF--Technical Guidance Document for RCRA Air
Emission Standards for Process Vents and Equipment Leaks").
     After proposal, EPA considered the commenters1 concerns that volatil-
ity of the waste stream should be considered in the equipment leak rules.
As a result, the LDAR provisions of the regulation were changed to estab-
lish two potential levels of required monitoring.  Those processes with the
greater emission potential are designated to be in gas/vapor service or
light liquid service and are required to implement a more restrictive LDAR
program.  Those processes with a lesser emission potential are designated
to be in heavy liquid service and are required to implement a less restric-
tive LDAR program.  The determination of being in light liquid service as
opposed to heavy liquid service is based on the concentration of organic
components in a waste whose pure vapor pressure exceeds 0.3 kPa.  This
addresses the commenters' concerns that volatility of the waste stream
should be considered.
     To determine whether a particular waste managed in a hazardous waste
management unit of the type specified in the rule (e.g., a steam stripping
                                    10-3

-------
or air stripping unit) is subject to the provisions of Subpart AA of
Parts 264 and 265, the owner/operator is required to conduct a determina-
tion of the waste's total organic concentration initially (by the effective
date of the standards or when the waste is first managed by a waste
management unit) and thereafter on a periodic basis (for continuously
generated wastes).
     A waste determination would only be necessary for Subpart AA applica-
bility when an owner/operator manages the waste in a distillation, frac-
tionation, thin-film evaporation, solvent extraction, or air or steam
stripping operation that is not controlled for organic process vent
emissions.  Waste determinations would not be necessary for wastes managed
in  (affected) units that are controlled for organic emissions to meet the
substantive requirements of Subpart AA.
     Determination that  the time-weighted, annual  average total organic
concentration of  the waste managed  in the unit is  less than 10 ppmw must be
performed by direct measurement or  by knowledge of the waste as described
later  in this section.   Direct measurement of the  waste's total organic
concentration must be performed by  collecting individual grab samples of
the waste under  process  conditions  expected to result in the waste having
maximum  organic  concentration and analyzing the samples using one of the
approved reference methods identified  in  the rule.
     The EPA  is  also  requiring that analytical results  for  a minimum of
four representative  samples  be used to  determine the  total  organic concen-
tration  for each waste  stream managed  in  the unit.  In  setting the minimum
number of  samples at  four, EPA will obtain  sufficient data  to characterize
the total  organic concentration  of  a waste  without imposing an unnecessary
burden on  the owner/operator to  collect and analyze the samples.
      As  an  alternative  to using  direct  measurement,  an  owner/operator  is
 allowed  to use knowledge of  the  waste  as  a  means  of determining  that  the
 total  organic concentration  of  the  waste  is less  than 10  ppmw.   Examples of
 information that shall  be considered by EPA to constitute sufficient  knowl-
 edge include:  (a)  documentation that  organics are not  involved  in the
 process  generating the waste;  (b)  documentation  that the waste is generated
 by a process that is identical  to a process at the same or another facility
                                     10-4

-------
which has previously been demonstrated by direct measurement to have a
total organic content less than 10 ppmw; or (c) previous speciation anal-
ysis results from which the total concentration of organics in the waste
can be computed.  The final standards include the provisions that EPA can
require that the waste be analyzed using Method 8240 if EPA finds the
documentation is insufficient to determine an exception by knowledge of the
waste.
     In order to address the temporal variability that can occur both
within a particular waste stream and within the various waste streams
managed in a hazardous waste management unit, the final rules require a
time-weighted, annual average concentration to characterize the waste
managed in the unit.  The final rules require that an owner/operator repeat
the waste determination whenever tliere is a change in the waste being
managed or a change in the process that generates or treats the waste or,
if the waste and process remain constant, at least annually.  For example,
continuous processes are more likely to generate a more homogenous waste
than batch operations; batch operations involve processes that may
frequently involve change in materials or process conditions.  Batch opera-
tions, therefore, usually generate wastes with varying characteristics,
including such characteristics as organics content.  Ground water concen-
trations would also be expected to show significant variation if more than
one well provides influent to a waste management unit such as an air
stripper and the wells that feed the unit are varied over time or if the
proportions from the wells that make up the influent are changed.  This is
because there is typically considerable spatial variability in contaminated
ground water concentrations.  The situation where feed wells are changed
and the change is not accounted for in the initial waste determination
would be considered a process change or change in the waste being managed
that would require a new determination.
     With the time-weighted, annual average applicability criterion, a
hazardous waste management unit would not be subject to this rule if it
occasionally treats wastes that exceed 10 ppmw if at other times the wastes
being treated in the unit were such that the weighted annual average total
organic concentration of all wastes treated is less than 10 ppmw.  The
                                    10-5

-------
time-weighted,  annual  average is calculated using the annual  quantity of
each waste stream managed in the unit and the mean organic concentration of
the waste stream.,
     The location where the waste volatile organic content is determined is
also of importance since sampling location can greatly affect the results
of the determination.   This occurs because the concentration  level  can
decrease significantly after generation as the waste is transferred to
various waste management units.
     If the waste is directly or indirectly exposed to ambient air at any
point, a portion of the organics in the waste will be emitted to the
atmosphere, and the concentration of organics remaining in the waste will
decrease.  For highly volatile organic compounds such as butadiene, all of
the compound would evaporate within a few seconds of exposure to air.
Similarly, emissions of organics from open waste transfer systems (e.g.,
sewers, channels, flumes) are expected to be very significant.  To ensure
that the determination of total organic concentration is an accurate
representation of the emission potential of a waste upon generation, it is
essential that the waste determination be performed at a point as near as
possible to where the waste is generated, before any exposure to the
atmosphere can occur.
     For the reasons stated above, the waste determination must be based on
the waste composition before the waste is exposed, either directly or
indirectly, to the ambient  air.  Direct exposure of the waste to the
ambient air means the waste surface  interfaces with the ambient air.
Indirect exposure of the waste to the ambient air means the waste surface
interfaces with  a gas stream that subsequently is emitted to the ambient
air.   If the waste determination  is  performed using direct measurement,  the
standards would  require  that waste samples be collected from an enclosed
pipe  or other closed system which is used  to transfer the waste after
regeneration to  the first  hazardous  waste  management unit.   If the waste
determination  is performed using  knowledge of the waste,  the standards
would  require that the  owner or  operator  have documentation  attesting  to
the volatile organic concentration of  the  waste  before  any exposure  to the
ambient  air.
                                     10-6

-------
     The location where the waste determination would be made for any one
facility will depend on several factors.  One factor is whether the waste
is generated and managed at the same site or the waste is generated at one
site and transferred to a commercial TSDF for management.  Another
important factor is the mechanism used to transfer the waste from the
location where the waste is generated to the location of the first waste
management unit (e.g., pipeline, sewer, tank truck).  For example, if a
waste is first accumulated in a tank using a direct, enclosed pipeline to
transfer the waste from its generation process, then the waste
determination could be made based on waste samples collected at the inlet
to the tank.  In contrast, if the waste is first accumulated in a tank
using an open sewer system to transfer the waste from its generation
process, then the waste determination would need to be made based on waste
samples collected at the point where the waste enters the sewer before the
waste is exposed to the ambient air.  For situations where the waste is
generated off-site, the owner or operator may make the determination at the
inlet to the first waste management unit at the TSDF that receives the
waste provided the waste has been transferred to the TSDF in a closed
system such as a tank truck and the waste is not diluted or mixed with
other waste.
     If a waste determination indicates that the total  organic concen-
tration is equal to or greater than the applicability criteria,  then the
owner or operator would be required to comply with the standards.
10.2  DETERMINATION OF EQUIPMENT IN VHAP SERVICE
     Comment:  Commenter AESP-00003 objects to the use of weight percent
when defining "in VHAP service" in gaseous samples.  The commenter contends
that it would be unnecessarily difficult to determine percent by weight of
a class of components (i.e.,  VO) in gaseous samples where volume fractions
are more commonly measured.  Commenter AESP-00014 states that determining
whether a vent on a covered tank is in VHAP service will be difficult to
measure because the organic vapors in the headspace will not mix homogen-
eously in the tank's headspace.  At times, the displaced gases will  contain
less than 10-percent VHAP when more than 10-percent VHAP may be present.
                                    10-7

-------
     Response:   First,  it should be noted that for clarity the term "in
VHAP service" has been  dropped from the promulgated regulations and the
applicability has been  more specifically defined.   Weight percentage is the
unit of choice when the determination of applicability is made on a solid,
liquid, or sludge waste and is commonly associated with these types of
wastes.  It is true that volume fractions are more commonly reported for
gaseous streams as asserted by Commenter AESP-00003.  However, it is not
easier to calculate the volume fraction rather than weight fraction.  Addi-
tional information on the calibration standard used, the carrier gas in the
standard, and both the organic and other inorganic gases in the sample are
required in both cases.  For simplicity, the units of the standard are
uniformly weight percent.  Additional detail will be provided  in the
guidance document concerning the conversion of the results of  the methods
to weight percent as required by the regulation.
     With regard to the difficulties of  obtaining a representative head-
space  sample in  covered storage tanks, this comment is valid  in certain
cases, but  is only relevant  if this  is the  location chosen to  collect  a
sample to determine if  an  associated process  vent  is subject  to the
standard.   Additional detail will  be provided  in  the guidance  document on
proper siting of a sampling  location.  This  is the  type  of issue that  will
be  worked out  in the permit.   It must  also  be pointed  out that if  it  is
determined  that  equipment  ever  contains  or  contacts wastes with  greater
than 10  percent  organics,  it is  covered  by  the standards.  Applicability
does not vary  from waste to  waste  (or  test  to test) for  the  same equipment.
 10.3  FUGITIVE  EMISSION MONITORING BY  METHOD 21
      Comment:   Two commenters (AESP-00019 and AESP-L0011)  do not support
 the use of portable  organic  vapor  analyzers for  leak  detection.   Commenter
 AESP-00019 points  to the difficulty of obtaining accurate readings at WSTF
 because VO levels  tend to be elevated due to the large number of pieces of
 equipment,  amount  of processing,  time and humidity conditions, and the need
 to continually use connect/disconnect systems to operate a facility.   These
 problems are particularly true in  enclosed settings such as  solvent storage
 areas; the commenter recommends that some allowance be made for measuring
 VO under these conditions.  Commenter AESP-L0011 objects because the
                                     10-8

-------
monitoring device is labor intensive, inefficient, costly, and not nearly
as sensitive as the automated, computer-controlled instrument installed at
his facility for the continuous monitoring of fugitive emissions.  This
system, according to the commenter, is capable of monitoring fugitive
emissions at multiple interior locations at a fraction of the concentra-
tions required by the proposed standards.
     Commenter AESP-00012 is concerned that workers conducting monitoring
of equipment with more than 10-percent VO using a nonspecific, direct-read-
ing organic vapor analyzer will not be adequately protected.  He submits
that the National Institute of Occupational Safety and Health (NIOSH)
recommends that workers be provided with appropriate protective equipment
and clothing (including eye protection) while performing these duties.  For
further information on occupational safety and health considerations at
TSDF, he recommends NIOSH publication 85-115 (Occupational Safety and
Health Guidance Manual for Hazardous Waste Site Activities).
     Response:  Commenter AESP-00019 felt that it would be difficult to
obtain accurate readings during equipment leak detection inspections
because of varying background conditions and suggests that some allowance
be made.  The EPA proposed two levels of instrument readings.  The one that
defines when a leak is detected, i.e., a 10,000-ppm reading using EPA
reference Method 21, should not be sensitive to ambient levels.  The no
detectable emission reading is defined in the Method 21 as the difference
between the instrument reading at the point of interest and the local
background reading.  Therefore, the regulation does take into account the
possibility of elevated background levels.
     Commenter AESP-L0011 proposes to use an automated, computer-controlled
instrument installed to monitor fugitive emissions at multiple locations
within the plant in lieu of a leak detection inspection with a portable
vapor analyzer.  The periodic leak detection inspection is intended to
pinpoint leaks at specific process equipment and to target them for repair.
It is unclear how the suggested automated ambient fugitive emission moni-
toring system would pinpoint which pump, valve, flange, etc., was leaking
unless each pump, valve, etc., was monitored.  Provided this could be ade-
quately designed to pinpoint leaks at specific process equipment and the
                                    10-9

-------
detector used to measure the concentration met the performance criteria
specified in Method 21, EPA has no objections to the use of an automated
system.  In this regard, the promulgated standards for equipment leaks
contain provisions for alternative means of emission limitations.
     Commenter AESP-00012 was concerned that the workers conducting the
leak detection inspection would not be adequately protected.  The EPA
agrees that adequate safety procedures should be followed.  Because of the
wide range of situations expected to be encountered, it is impractical, as
well as outside of the regulation's scope, to specify safety procedures in
this regulation.  However, there is a copy of NIOSH publication 85-115
(Occupational Safety and Health and Guidance Manual for Hazardous Waste
Site Activities) in the docket for this rulemaking (see Docket No. F-86-
AESP-FFFFF, item 00012.A).  There are also provisions in Sections
264.1064(g) and 265.1064(g) that allow the owner or operator to designate
and exempt from monthly monitoring specific valves as unsafe to monitor.
The owner or operator of the valve must demonstrate that the valve...is
unsafe to monitor because monitoring personnel would be exposed to an
immediate danger as a consequence of monitoring; and a written plan must be
followed that requires monitoring of the valve as frequently as practicable
during safe-to-monitor times.
                                    10-10

-------
                11.0   IMPLEMENTATION  AND  COMPLIANCE  PROVISIONS

 11.1   COMPLIANCE  DATES
 11.1.1  Overall Standards
      Comment:   Seven  commenters  (AESP-L0001,  AESP-L0006,  AESP-00020,
 AESP-00024,  AESP-00033,  AESP-00003,  and  AESP-00035)  believe  that  more time
 is  needed  to comply to the  standards than  is  allowed by  Section 269.33 of
 the proposed standards.  Specifically, Commenter  AESP-L0001  states  that the
 proposed rule  does not address compliance  dates for units or equipment that
-may be subject to the rule  in the  future.   The proposed  rule ties compli-
 ance  to 6  to 24 months after promulgation  of  the  regulations and  so does
 not address  units or  equipment that  will be first subject to the  air
 emissions  regulations months or  years after the regulations  are promul-
 gated.  According to  Commenter AESP-L0001,  because  the effective  date of a
 RCRA  regulation is usually  6 months  after  the date  of promulgation, the
 proposed rule  would require the  design to  be  completed 12 months  after
 promulgation,  yet construction would commence 9 months after promulgation.
 This  schedule  would require construction to commence before  the design is
 completed.   The commenter suggests allowing compliance within  24  months
 after the  date of promulgation or  within 24 months  after the equipment
 becomes subject to the regulation, whichever  occurs  later.   This  commenter
 also  contends  that the standards should  allow for procedural delays result-
 ing from major permit modifications  that are  required to add monitoring and
 control  devices to permitted facilities; a major  modification  takes a
 minimum of 12  to  18 months  to account for  public  participation requirements
 and control  devices cannot  be installed  until after  the  modification  is
 made.   The commenter  recommends  that EPA require  compliance  within  a
 reasonable time after the permit modification has been issued  or  to amend
 40  CFR 270.42  to  specify that all  changes  necessary  to comply  with  the air
                                     11-1

-------
standards are minor permit modifications.  Commenters AESP-L0006 and AESP-
00033 note that the compliance period for interim status facilities is
specified, but no similar period is provided for TSDF that have a final
permit.  According to Commenters AESP-00020 and AESP-00035, a lead time of
greater than 24 months is needed to program needed funds, design cost-
effective systems, and start construction; they recommend between 36 and 48
months.  Commenter AESP-00033 agrees and asks that provisions be made for
extensions on a case-by-case basis.  In comparison, Commenter AESP-00003
states that the 24-month deadline appears to be achievable, but that the
standards should not dictate a step-by-step compliance schedule for
interim-status facilities.
     Response:  Under RCRA statutory requirements, compliance must be
attained by the effective date of the rule  (promulgation date plus 6
months).  However, the final rules allow facilities that cannot install the
control devices and closed-vent systems required to comply by the effective
date to complete an implementation schedule that ensures compliance no
later  than 18 months after the effective date of the statutory or regula-
tory amendments under RCRA (e.g., a new listing or identification of a
hazardous waste) that render the facility subject to Subpart AA or BB.  The
implementation schedule must be in the operating record  on the effective
date of the statutory or  regulatory amendment.  The facility also must
document  a rationale of why installation of the emission controls cannot  be
completed at  an earlier date.  The EPA considers the implementation sched-
ule  approach  as reasonable for facilities that cannot  install emission  con-
trols  in  6 months  and the 24-month schedule a realistic  estimate of the
maximum timeframe  needed  to install a control device and closed-vent  sys-
tems,  based on Section  112, CAA regulations which  also contain  provisions
that  allow up to  2 years  for  installation of  a control  device.   The 24-
month  schedule allows time for testing,  design, evaluation,  fabrication,
and  installation.  Therefore, no provisions have been  made in the  standards
for  extensions beyond 2 years on a case-by-case basis.
      The  final standards  have been  revised  to require  a TSDF with  a final
permit issued before  the  effective date  to  incorporate the requirements of
the  standards only at permit  review  under Section  270.50 or reissue under
Section  124.15.   Therefore,  the  standards do  not  require a permit  modifica-
tion.   A 24-month implementation  schedule is  not  allowed for waste

                                     11-2

-------
management units or equipment starting operation after the effective date
of Subparts AA and BB because owners or operators will be aware of the
requirements of the rule and are expected to plan in advance for control
devices and closed-vent systems.  However, waste management units or
equipment that become newly regulated units because of a newly listed
hazardous waste will have 24 months from the effective date of the rule
establishing the new listing or identification to complete installation and
begin operation.  The EPA agrees that the Agency should not dictate step-
by-step implementation schedules because each affected facility needs some
flexibility to budget funds, perform engineering evaluations, and complete
construction; therefore, the interim dates in the schedule have been
dropped.
     Under the approach discussed above, the standards promulgated today
for process vents and equipment leaks would be implemented on the following
schedule for existing TSDF:
     •    180 days following promulgation, the standards become effective;
          all facilities become subject to the new standards.
     •    On the effective date of the standards, compliance with the
          standards is required.  Each facility that does not have the
          control devices required by the standards in place must have one
          of the following in the facility's operating record:  (1) an
          implementation schedule indicating when the controls will be
          installed and in operation, or (2) a process vent emission rate
          determination that documents that the emission rate limit is not
          exceeded (therefore, process vent emission controls are not
          required).
     •    No later than 18 months following the effective date (2 years
          following promulgation), any control devices required by the
          standards for process vents and equipment leaks must be installed
          and in operation at all facilities.
     •    All permits issued after the effective date must incorporate the
          standards.
     An existing solid waste management unit may become a hazardous waste
management unit requiring a RCRA permit when a waste becomes newly listed
or identified as hazardous.  Owners and operators of facilities not
previously requiring a RCRA permit who have existing units handling newly
listed or identified hazardous waste can submit a Part A application and
                                    11-3

-------
gain interim status.  The air emission standards promulgated today would be
implemented at these facilities on the following schedule:
     •    180 days following the date the managed waste is listed or
          identified as hazardous, the standards become effective;
          facilities become subject to the standards.
     •    On the effective date of the standards, each facility that does
          not have the control devices required by the standards in place
          must have one of the following in the facility's operating
          record:  (1) an implementation schedule indicating when the
          controls will be installed and in operation, or (2) a process
          vent emission rate determination that documents that the emission
          rate limit is not exceeded (therefore, process vent emission
          controls are not required).
     •    No later than 18 months following the effective date, the
          controls required by the standards must be installed and in
          operation at all facilities.
     Newly constructed TSDF are required to submit Part A and Part B permit
applications, and to receive a final permit prior to construction as
required by Section 270.10.  Following the effective date of the standards
promulgated today, a Part B application for a new facility must demonstrate
compliance with the standards as contained in Part 264, if applicable.
Therefore, all controls required by the standards would have to be in place
and operating upon startup.
     Similarly, new waste management units added to existing facilities
would have to be equipped with the required controls prior to startup.  For
a new unit added to an existing permitted facility, a permit modification
would be necessary.  Where a new unit is added to a facility in interim
status, the owner or operator must submit a revised Part A application
(Section 270.72[c]), including an explanation of the need for the new unit,
and then receive approval from the permitting authority.
     For facilities with hazardous waste management units that previously
were not subject to control requirements because the wastes  in the units
did not contain organics in concentrations equal to or greater than the
applicability criteria of 10 ppmw as a time-weighted annual  average or
10 percent, the owner or operator would be required to comply with all
Subpart AA or BB requirements on the date that the facility  or waste
                                    11-4

-------
management unit becomes affected by the rules (i.e., the date the facility
begins to manage wastes in the units with time-weighted annual average
organic concentrations greater than 10 ppmw for Subpart AA or greater than
10 percent for Subpart BB irrespective of any change in permit status that
is required by the change in waste concentration.  In this situation,
should the facility owner or operator elect to use a control device to
comply with the process vent or equipment leak provisions, the control
device must be installed and operating on the effective date; the 24-month
extension beyond the promulgation date is not applicable in this case.  For
the process vent emission rate limit, the situation is somewhat different.
TSDF process vents associated with the distillation/separation operations
specified in the rule that manage wastes with time-weighted annual average
organics concentrations of 10 ppmw or greater are affected by the regula-
tion regardless of whether the facility emissions are above or below the
emission rate limit.  Therefore, any change in the facility operations that
results in a TSDF going above or below the emission rate limit does not
cause a change in the applicability of the facility to Subpart AA.  The
rules require that affected TSDF reduce total process vent organic emis-
sions from all affected vents by 95 percent or reduce the facility's total
process vent emissions to or below 1.4 kg/h and 2.8 Mg/yr.  One of these
conditions must be met at all times; the facility's emission rate deter-
mination, which documents the facility's status regarding compliance with
the process vent standards, must also at all times reflect current design
and operation and wastes managed in the affected units.
11.1.2  Period for LDAR
     Comment:  Commenters AESP-00008 and AESP-L0016 believe that the 5-day
period allowed for returning a release through a pressure relief device to
a "no detectable emissions" condition is too long and that the first repair
of a leak should occur within 24 hours of detection.  Complete repair
should occur within 5 days.  The standards should not retain the current
requirements and still allow repair delays of up to 1 year if a process
shutdown is needed.  Commenter AESP-L0013 believes that any delay in repair
should be conditioned by a determination that no adverse health impact
would result from the delay.
                                    11-5

-------
     Response:  The selected repair intervals of the 5-day "first attempt
at repair" requirement and the 15-day repair period for pumps, valves, and
pressure relief devices provide maximum effectiveness of the LDAR program
by requiring expeditious emission reduction, while allowing the owner or
operator the time to maintain a reasonable overall maintenance schedule for
the plant.
     During development of the CAA standards for petroleum refineries and
chemical plants (EPA-450/3-80-032a, EPA-450/3-80-033a, and EPA-450/3-82-
010), EPA personnel made a concerted effort to investigate and gain
knowledge of maintenance practices.  In EPA's technical judgment, an
initial attempt at repair within 5 days is ample for all simple field
repairs.  A 24-hour period following leak detection is often not long
enough to allow maintenance personnel to identify the cause of the leak and
then to attempt repair.  Although plants could schedule repair personnel to
accompany the monitoring team, emergency situations or critical equipment
problems could easily upset these arrangements.  Although some or most
repairs can be made within 24 hours, it is not practical to require an
attempt to repair all equipment within 24 hours.  The EPA has not been able
to distinguish between equipment that could and could not always be
repaired within 24 hours.  In addition, with the commenter's approach,
repair crews would spend much of their time on an inspection with few
needed repairs.  Furthermore, the owner or operator of a recycling opera-
tion has an incentive to repair leaks as quickly as possible to prevent
additional product losses.
     A 15-day repair interval provides time for isolating leaking equipment
for other than simple field repairs.  A shorter interval could cause
scheduling problems in repairing valves that are not conducive to simple
field repair and that may require removal from the hazardous waste manage-
ment process for repair.  A 15-day interval provides the owner or operator
enough time to determine precisely which spare parts are needed and to
schedule repairs.  In addition, a 15-day repair interval allows more
efficient handling of more complex repair tasks while maintaining an
effective reduction in equipment leaks.  Again, the owner or operator of a
recycling operation has an incentive to repair leaks promptly.
                                    11-6

-------
     Section* 264.1059(c) and 265.1059(c) require that for delay of repair
for valves to be allowed, the owner or operator must demonstrate that
emissions of purged material resulting from immediate repair are greater
than fugitive emissions likely to result from delay of repair and that,
when repair procedures are effected, the purged material is collected and
destroyed or recovered in a control device complying with Section 264.1060
or 265.1060.  Therefore, the delay of repair should not result in increased
emissions or health risks.  Furthermore, the delay-of-repair provisions
allow the facility operator sufficient time to obtain necessary repair
parts and maintain some degree of flexibility in overall plant maintenance
scheduling.
11.2  PERMITS
11.2.1  State and Local Role
     Comment:  Coimnenter AESP-00009 asks if EPA has analyzed the
feasibility of obtaining State and local operating permits for the
incinerators and flares proposed as control devices.  According to the
commenter, State and local governments may define organic incinerators as
hazardous waste incinerators and regulate them to a 99.99-percent
destruction or prohibit them completely through State and local siting
procedures.
     Response:  Although flares and incinerators were evaluated as control
devices and are approved alternatives, the standards do not require them.
The EPA acknowledges that an incinerator, cement kiln, or industrial boiler
accepting a RCRA-listed waste would be regulated to a 99.99-percent
destruction and removal efficiency under permitting standards.  Where
facilities do have permitted incinerators, ducting emissions to these units
is acceptable under the standards and may be more cost-effective depending
on site-specific conditions.  No analysis has been conducted on the feasi-
bility, costs, or impacts of obtaining State and local operating permits
for these control devices (flares and incinerators) because they are not
required by the standards.
     It is also important to note that noncontainerized gases emitted from
hazardous wastes are not themselves hazardous wastes because the statute
                                    11-7

-------
implicitly excludes them.   EPA does,  however,  have the authority to regu-
late them as emissions from hazardous wastes.   Therefore, thermal destruc-
tion of process vent emissions would not necessarily be classified as
hazardous waste incineration.
11.2.2  Part B Information Requirements
     Comment:  Commenter AESP-00035 recommends that Section 270.22 for
specific Part B information requirements should be limited to the units
already included in the Part B permit application.  Units that must comply
with this regulation because the facility is subject to  RCRA permit
requirements for other reasons should not be required to be added to the
Part B permit.  The call for documentation of compliance with this regula-
tion should not delay issuance of a RCRA permit.
     Response:  First, it  should be noted that in the final standards,  Part
B  information  requirements for process  vents are contained in Section
270.24 and  those for  equipment are in Section 270.25.  The EPA  is  aware
that extending specific Part  B information requirements  to those hazardous
waste management units that  are not  subject to RCRA  permitting  but are
located  at  facilities that are otherwise  subject  to  RCRA permit require-
ments  could result in the  need for  those  facilities  to modify RCRA permits
or their Part  B  applications.  However, EPA believes that extending  the
Part  B information requirements to  hazardous  waste management units  not
subject  to RCRA  permitting is necessary to  ensure compliance with the
Subpart  AA and Subpart  BB standards.
      The EPA also  agrees  that requiring a modification  of RCRA  permits (and
 Part B applications)  as part of  this rule could  result   in delays in  proc-
 essing and issuing final  RCRA permits.   Therefore, in the final rules,
 facilities are not required as a  result of the TSDF process vent and
 equipment-leak standards to modify permits  issued before the effective
 date.  Consistent with Section 270.4,  a facility with a final  permit  issued
 prior to the effective date is required to add or incorporate the
 requirements of these standards into their permit at permit review under
 Section 270.50 or reissue under Section 124.15.
      Facilities that have obtained RCRA interim status,  as specified  in
 Section 270.70 (i.e., compliance with  the requirements  of Section 3010(a)
                                      11-8

-------
 of RCRA pertaining to notification of hazardous-waste activity and the
 requirements of Section 270.10 governing submission of Part A applica-
 tions), will be subject to the Part 265 standards on the effective date.
 Interiurn-status facilities that have submitted their Part B application
 prior to the effective date of the regulation will  be required to modify
 their Part B application to incorporate the requirements of these air
 rules.
      A facility operating under interim status that has not submitted a
 Part B application prior to the effective date of the regulation  must
 incorporate the requirements of these standards in  the Part B application.
 11.3  COMPLIANCE PROVISIONS
 11.3.1   Documentation
      Comment:   Commenter AESP-00034 suggests  that requirements for documen-
 tation  of compliance  with Section  61.242-11 should  not reference  "Control
 of Gaseous Air  Pollutants";  this should  be replaced with engineering
 evaluation,  a material  balance  calculation, or an air emission source test
 for the control  efficiency  requirement so as  not  to limit documentation  of
 compliance to one  publication.  Commenter AESP-00034 states  that  Section
 61.242-11  for documentation  of  compliance should  require a compliance test
 instead of design  criteria  because  the technical  requirements  in  this
 section are for  flares,  and  the documentation  requirements are actually  for
 the control  efficiency  requirements  for control devices  that are  not
 flares.
      Response:   First,  it should be  noted  that  Section 61.242-11  has  been
 incorporated into  the standards as Sections 264.1060  and  265.1060.  The
 document  in  question establishes review protocol  for  engineering  design  and
 evaluation.  The promulgated regulation has been  revised  to allow other
 engineering  texts  acceptable to the Regional Administrator that present
 basic design information.  In addition, specific operating parameters that
must be addressed  in the control device design have been  added for each
type of applicable control device,  including flares.  Also, the owner or
operator must certify in writing that the control  device meets the require-
ments of the standards.  The EPA also notes that documentation demonstrat-
ing compliance with the process vent standards can be based on engineering
                                    11-9

-------
calculations or source tests that must be made with the process operations
(e.g., distillation, fractionation,  or steam stripping operations) running
at full operating conditions and at  maximum flow rates.
11.3.2  Special Requirements
     Comment:  Commenter AESP-00035  recommends that Section 270.22 for
"special requirements" be revised to allow equipment (e.g., piping system
to incinerator) determined to be in  VHAP service also to be used for non-
VHAP waste transport.  The commenter states that engineering judgment
should be considered in revisions of the VHAP determination and the proce-
dures specified in the section at issue should be allowed in a valid
determination revision.  Commenter AESP-00003 agrees in theory in that
Section 269.34(a)(3) should not be included in the standards because this
provision ignores the possibility of revisions that would reduce VHAP
concentrations sufficiently that engineering judgment would be appropriate
and because EPA retains the authority to require the procedures specified
in Section 269.34(a)(l) in the event of a disagreement.
     Response:  First it should be noted that for clarification, "in VHAP
service" has been dropped from the standards and applicability is defined
strictly by the characteristics of the hazardous waste contained or
contacted by equipment.  Under the standards, once equipment is determined
to be subject to the requirements, the equipment retains this status.  Only
when the operator determines that the equipment never will contain or
contact a hazardous waste with 10 percent or more organics can that
determination change.  The purpose of this provision is to discourage
switching back and forth between being subject to and not being subject to
the standard requirements; with this approach, compliance would be much
more straightforward and would avoid possible confusion that might lead to
contravention of the standards.  Equipment designated as containing or
contacting hazardous waste with at least 10 percent organics can transport
hazardous waste streams with less than 10 percent organics but the controls
required by the standards must be maintained.
11.3.3  Alternative Means of Emission Limitation
     Comment:  Commenter AESP-00035 recommends that alternative means of
emission limitation be allowed for permitted and interim-status facilities
                                    11-10

-------
under Sections 269.32 and 269.33 to encourage the development of new con-
trol methods.
     Response:  The provisions of 40 CFR 61.244 Subpart V, which provide a
formal mechanism for applying for use of an alternative means of emission
limitation, were specifically not included in the proposed TSDF process
vent and equipment leak rules and have not been included in the final
standards.  The equipment leak standards comprise the use of control
equipment, inspection of process equipment, and repair programs.  The
alternative means of emission limitation provisions for the equipment leak
standards are not considered self-implementing; i.e., these provisions
cannot be satisfied without the need for detailed explanation or
negotiation between the facility owner/operator and EPA.  Therefore, the
standards do not allow for the use of alternative means of emission
limitation for equipment leaks.
     The closed-vent system and control device provisions are performance
standards, requiring 95 percent emission reduction or control of facility
process vent emissions below the emission rate limit.  Compliance with
these provisions can be demonstrated and documented through the use of a
performance test or engineering calculations.  The standards allow an
owner/operator to use a control device other than a thermal vapor incinera-
tor, catalytic vapor incinerator, flare, boiler,  process heater, condenser,
or carbon adsorption system provided that documentation is developed
describing control device operation and demonstrating performance.  The
negotiation process associated with issuance of a final permit can be used
by the owner/operator and EPA to review and establish the appropriate
recordkeeping requirements.
11.4  RECORDKEEPING AND REPORTING
11.4.1  Frequency of Reports
     Comment:  Commenter AESP-L0001 contends that a reporting frequency of
6 months is unnecessary.  According to the commenter, simple certification
of compliance with the standards, with records available for inspection at
the respective sites, should be adequate.  Commenter AESP-L0001 proposes a
three-tiered approach to reporting requirements for LDAR.  Specifically,
                                    11-11

-------
semiannual  reports should be required for processes with high leak rates
and with records of quarterly and monthly maintenance at the plants.  The
reporting philosophy is similar to that used for spill prevention plans
under the CWA, where reporting is not required for exceptional sites as
long as records are maintained, or where reporting increases with the
frequency of spills.  This approach should be applied to processes with
(1) leak rates of less than 2 percent, and (2) the best design features or
exceptionally low emission rates.  Plants using the skip period approach
under 40 CFR 61.243-2 should report only annually.
     Response:  In the promulgated rules, reporting on  a routine  basis  is
only required for exceedance situations for facilities  with  final RCRA
permits; there  are no reporting  requirements  for  interim-status facilities.
For equipment leaks, these would be cases where leaks have  not been
repaired in the required time.   However,  if a facility  does  not have  any
exceedances during the reporting period,  no report  is required.   Interim-
status  facilities  also are  not required  to  submit reports.   The purposes of
the reporting requirements  for permitted  facilities  are to  alert  EPA
offices of situations  in which leaks  have not been repaired as  required and
to assist  the enforcement program to prioritize inspection  and  enforcement
 actions.   Most of the  detailed information  on LDAR compliance is  maintained
 at the plant  site for  inspection.  This level of  recordkeeping  and report-
 ing is the minimum level needed for enforcement and compliance  purposes,
 and previous  experience with similar regulations  has shown that the report-
 ing requirements do not create an unreasonable burden.
 11.4.2  Duplication of Reports
      Comment:  Commenter AESP-00003 commends EPA on its efforts  to elimi-
 nate duplicative recordkeeping  requirements  and  suggests that such a provi-
 sion also be included in Section 269.32 of the proposed rule.
      Response:  The EPA  agrees  that any duplicative recordkeeping and
 reporting should generally be eliminated to  the  extent possible.  Because
 of the difficulties in  foreseeing all situations in which  duplication  could
 occur, a  provision to this effect has not been added to the standards.
 However,  when  records and  reports required by  State  programs are substan-
 tially similar to those required by  these  standards, a copy to EPA of  the
 information  submitted to the  State will  generally  suffice.   When similar
                                      11-12

-------
records and-reports are required by other EPA programs (e.g., the visual
observations required for pumps and valves associated with storage tanks
and incinerators), EPA suggests that one set of records be maintained, with
emphasis on the more detailed monitoring records required by these stan-
dards.  The EPA considers that the monitoring required for equipment leaks
under these standards differs significantly from the monitoring required
for ground-water protection purposes under other RCRA rules.  However, the
monitoring programs and records for the air and ground-water programs can
be combined to reduce labor and costs in situations where duplication
exists.
11.4.3  Notification
     Comment:  Commenter AESP-L0015 asks whether TSDF are subject to any
notification requirements if they determine that their waste stream is less
than 10-percent VHAP by weight.
     Response:  The proposed standards, through incorporation of the
provisions of 40 CFR Part 61 Subpart V, required the owner or operator of
any piece of equipment to which Subpart V applied to submit a written
statement to the Administrator that the requirements of the standards were
being implemented.  Such a notification is a normal requirement of CAA
regulations.  However, upon reviewing the standards within the RCRA
regulatory framework, EPA determined that the notification is not needed.
Enforcement of standards under RCRA is primarily through inspections during
site visits, particularly the review of records maintained in the facility
operating record.  The semiannual exceedance reports required of permitted
facilities by the standards can be used to direct enforcement efforts
toward facilities with potential compliance problems.  Therefore, there are
no notification requirements in these rules for equipment or process vents,
whether or not they manage wastes with organic concentrations above the
applicability criteria.  States may require such notification, however.
Facilities found to have process vents and equipment that are subject to
the standards and are not in compliance with the requirements of the
standards will be subject to enforcement orders.
                                    11-13

-------
11.5  IMPLEMENTATION AND ENFORCEMENT
11.5.1  Guidance on the 95-Percent Limit
     Comment:  Commenter AESP-00022 states that, although the proposed rule
identifies a 95-percent control for tanks and provides a method to be used
to demonstrate compliance, no guidance is provided regarding whether the
95-percent is an average or instantaneous value and what time period should
be used for averaging.  If the figure is an average, the minimum could be
achieved by scheduled maintenance and idle time shutdowns during the aver-
aging period to limit the time the tank vents are controlled by wet scrub-
bers.  A properly set averaging time would provide operator flexibility to
control emissions without investing in equipment for short-duration
shutdowns; guidance should be provided to clarify the 95-percent goal.  The
commenter is referring to hazardous waste storage tanks vented to an incin-
erator that employ wet scrubbers when the incinerator is not operating.
     Response:  It should be noted that, unless process emissions are
vented through the hazardous waste storage tanks, the storage tank emis-
sions are not subject to the requirements of this rulemaking.  Nonetheless,
in response to the commenter1s concerns, the 95-percent emission control
requirement  is considered a continuous requirement and as such must be
achieved at all times during operation of the process being vented to the
atmosphere.  The control technologies that are  typically used for process
vent emissions  (condensers and carbon adsorbers) are capable of achieving
the 95-percent control over extended periods if properly designed,
operated, and maintained.  The EPA has not identified any situations
particular to this  industry that would warrant  any short-term variance  to
the 95-percent control requirement.  Although there are  likely to be normal
variations in process stream flow  rates,  it  is  not expected that these  will
result  in extreme  emission fluctuations  that could  result in short-term
failure to meet the 95-percent emission  reduction requirement.
      To ensure  that the  control devices  perform according to design and are
properly  operated  and maintained  on  a day-to-day basis,  the final rules
require the  continuous monitoring  of specific parameters on all control
devices needed  to  meet the standards.   In  general,  the monitoring require-
ments for control  device operating parameters  include (1) coolant fluid
                                     11-14

-------
temperature  and exit gas temperature or the  concentration  level  of  the
organic compounds  in the exhaust  vent  stream for  condensers,  (2)  exhaust
vent stream  concentration  level of organics  or  a  parameter that
demonstrates  regularly scheduled  carbon bed  regeneration for  carbon
adsorbers, (3) combustion  temperature  for  incinerators  and other enclosed
combustion devices, and  (4) pilot flame detection  for flares.  The
monitoring readings must be checked at least  once  each  operating  day  and,
if necessary, corrective measures must be  implemented immediately to  ensure
that the control device operates  efficiently.   The monitoring  informa-
tion/data must also be recorded in the facility operating  record.   Periods
when monitoring indicates  control device operating parameters  are outside
or exceed established tolerances  on design specifications  that are  not
corrected within 24 hours  must be recorded.   For facilities with  final RCRA
permits, exceedances must  be reported to the  Regional Administrator on a
semiannual basis; there are no reporting requirements for  interim-status
facilities.  The records and reports must  include  exceedance dates, dura-
tion, cause,  and corrective measures taken.
     The facility process  vent emission rate  limit that has been  added to
the standards provides operators  of facilities  for which process  vent
emissions can be controlled to less than 1.4  kg/h  (3 Ib/h) and 2.8  Mg/yr
(3.1 ton/yr)  the flexibility of determining the compliance method.  For
example, if a WSTF has multiple process vents and  total facility  process
vent emissions of 15 Mg/yr (16.5  ton/yr), the facility emission rate  limit
conceivably could be met by reducing the emissions from each vent by 82
percent (2.8 Mg/yr [3.1 ton/yr] is approximately 18 percent of 15 Mg/yr
[16.5 ton/yr]) or by reducing the emissions from one vent by 95 percent if
that vent accounts for 85 percent or more of the total facility process
vent emissions.   Detailed  implementation information is provided  in the
guidance document that is being published concurrently with these
standards.
11.5.2  Omnibus  Permitting
     Comment:  Commenters AESP-L0016 and AESP-L0018 object to statements in
the preamble  regarding the role of omnibus permitting authority under RCRA
Section 3005(c)(3).  Commenters AESP-L0018 and AESP-L0015 question the
absence of criteria for when  such  authority would be applied to require
                                    11-15

-------
more stringent controls for individual constituents that continue to pose a
risk after control.  Commenter AESP-L0018 also does not support the use of
this authority to pursue data gathering through ambient monitoring.  In
support, Commenter AESP-L0018 argues that data gathering does not meet the
protection of human health and environment mandate of Section 3005(c).
Commenter AESP-L0016 objects because the standards are not protective in
all cases.  Commenter AESP-L0016 also contends that Congress intended the
authority of this section to be used as a temporary tool, pending promulga-
tion of regulations providing needed protection.  The commenter argues that
authorizing permit writers to impose more stringent controls based on an
unenforceable guidance document is not a substitute for the regulations.
The commenter submits that it was precisely because hazardous air emissions
at TSDF have largely been unregulated through the RCRA permit system, and
because EPA had failed to promulgate permit standards under Section 3005
governing air emissions at TSDF, that Congress enacted Section 3004(n).
     Response:  Air emissions from hazardous wastes include photochemically
reactive and nonphotochemically reactive organics, some of which are  toxic
or carcinogenic,  and also may include toxic or carcinogenic inorganic
compounds.  Depending  on the source,  particulates  (including metals,
aerosols of organics,  dust, as well as toxics and carcinogens) also may  be
released or generated.  These emissions, which are released to the
atmosphere from a wide variety of sources within TSDF,  present diverse
health and environmental  risks.  Therefore,  EPA has developed  a  multiphased
approach  for  regulating TSDF organic  air emissions.    This  approach,
described  generally  below,  reflects the  Agency's  understanding of  the
problem and  knowledge  of  applicable,  effective controls at  this  time.
      Organic  emissions from TSDF managing  hazardous wastes  contribute to
ambient ozone formation and increase  cancer and other health  risks.
 Phases I  and  II  will  address  these  two problem areas  (i.e.,  air  toxics  and
carcinogens,  and  ozone precursors)  by controlling emissions of organics as
 a class rather than  controlling  emissions  of individual waste constituents.
 The regulation of organics as a  class has  the advantage of being relatively
 straightforward because it can  be accomplished with  the minimum number of
 standards,  whereas the control  of individual toxic constituents will
                                     11-16

-------
require multiple standards.  Regulating organics as a class also makes
efficient use of EPA resources, avoids many of the complexities of having
multiple standards, and reduces the number of constituents for which
separate standards ultimately will be required.  The environmental effects
of ambient ozone are well documented and, in terms of monetary losses, they
total hundreds of millions of dollars each year.  The substantial reduc-
tions in organic emissions achievable through implementation of Phase I and
Phase II controls will reduce atmospheric ozone formation as a result of
reductions in TSDF emissions of ozone precursors and will reduce nationwide
cancer incidence and maximum individual risk due to exposure to air toxics
and carcinogens emitted from TSDF.
     Specifically, Phase I entails the promulgation of standards for the
control of organic air emissions from selected hazardous waste management
processes and equipment leaks.  The EPA chose to develop this portion of
its TSDF rulemaking first to prevent uncontrolled air emissions from land
disposal restriction (LDR) treatment technologies.  The technologies used
in lieu of land disposal include the distillation/separation processes
subject to the Phase I rules.  Publication of the final rules for air
emissions from hazardous waste management process vents from distillation,
fractionation, thin-film evaporation, solvent extraction, and air or steam
stripping processes and from leaks in piping and associated equipment
handling hazardous wastes marks the completion of this first phase.
     In the second phase, EPA will propose to extend standards (in
mid-1990) under Section 3004(n) to include organic air emissions from other
significant TSDF air emission sources not covered or not adequately
controlled by existing standards.  These sources include surface impound-
ments, storage or treatment tanks (including process vents on closed,
vented tanks), containers, and miscellaneous units.
     The analyses of impacts indicate that at some facilities, residual
cancer risk to the most exposed individuals after implementing the first
two phases of regulation will remain outside the historical range achieved
by other RCRA regulations (i.e., 1 x 10'4 to 1 x 10'7).  The EPA is
therefore planning a third phase of the effort to control TSDF emissions in
which various means for further reducing risk will be examined.  In the
                                    11-17

-------
interim, the omnibus permitting authority of RCRA is an available option
for requiring additional  emission and risk reductions beyond that achieved
by today's final  rules if it is decided,  on a case-by-case basis, that
additional control  is needed to protect human health and the environment.
     The EPA is currently involved in an effort to improve the data used in
the current risk analyses and,  in the third phase, will make use of any new
data obtained.  If additional constituent control is found necessary, the
number of constituents for which additional control is needed is expected
to be significantly less than if a constituent approach were used as the
only means of regulating TSDF air emissions.  Therefore, the EPA is
convinced that the control of organics as a class followed by controls for
individual toxic constituents,  if necessary, will ultimately result in
comprehensive standards that are protective while providing effective
interim control.
     As noted above, permit writers can use their omnibus permitting
authority to require more stringent controls at facilities where a high
residual risk remains after implementation of the standards for volatile
organics.  The permitting authority cited by Section 3005 of RCRA and
codified in Section 270.32(b)(2) states that permits issued under this
section "...shall contain such terms and conditions as the Administrator or
State Director determines necessary to protect human health and the
environment."  This section, in effect, allows permit writers to require,
on a case-by-case basis, emission controls that are more stringent than
those specified by a standard.  The EPA has a mandate to use this authority
for situations in which regulations have not been developed or in which
special requirements are needed to protect human  health and the environ-
ment.   For example, this authority can be used in situations where,  in the
permit  writer's judgment, there  is an unacceptably high residual risk after
application of controls required by the process vent and equipment organic
emission  standards.
     The  EPA  is currently preparing guidance to be used by permit writers
to help identify facilities  that would potentially have high residual risk.
The guidance will  include procedures to be used to identify potentially
high-risk  facilities  and will  include guidance for making a formal,  site-
                                     11-18

-------
specific risk assessment.  Methods for providing additional emissions
control at facilities identified as having high residual risk after
implementation of the standards for volatile organics will be included in
additional guidance developed by the Agency.  Examples of both risk
assessments and the provision of additional emissions control will be
included in the respective guidance documents.
     As has been described above, the approach that EPA is using to control
TSDF air emissions is to proceed with promulgation of regulations to
control organic emissions (Phases I and II) and to follow this with a phase
of regulations that would require more stringent controls for individual
hazardous constituents where necessary.  With regard to data-gathering by
ambient air quality monitoring, the feasibility of requiring site-specific
risk analyses that involve the determination of ambient air quality impacts
through dispersion modeling or ambient monitoring for TSDF is currently
being studied under the Phase II and III standards and will not be required
for the purposes of these standards.
11.5.3  State and Regional Role
     Comment:  Commenters AESP-00031, AESP-00032, and AESP-L0015 ask that
EPA clearly delineate whether implementation and enforcement responsibili-
ties are under the CAA or RCRA.
     Response:  Implementation and enforcement responsibilities for these
TSDF air rules will be under RCRA.  Although EPA acknowledges that CAA
offices have been active in permitting decisions and have experience and
expertise in air concerns, RCRA personnel must implement and enforce RCRA
rules.  Internally, EPA will delineate how these offices can work together
to eliminate duplication of permitting and enforcement labor.  To this end,
EPA has developed an Implementation Work Group to discuss these and similar
concerns.  An implementation guidance document is being issued in
conjunction with promulgation of these TSDF air rules.
11.5.4  LDAR Enforcement Approach
     Comment:  Commenter AESP-00021 recommends that EPA adopt the LDAR
enforcement approach used in California where, if a leak has not been
recorded or repaired within a specified timeframe, then the leak is a vio-
lation.  In the alternative, EPA could define as a violation any leak found
                                    11-19

-------
by an inspector that has not been recorded or repaired properly after a
company inspection.  These approaches give the control authority some
recourse if the facility is not conducting the program properly or not
conducting it at all.
     Response:  Seals and packings inherently leak and not even the use of
leakless equipment can prevent occasional leakage.  Because an occasional
leak cannot be prevented, EPA cannot accept the commenter's suggestion that
a leak (a reading over 10,000 ppm) should be considered a violation when
documented during a compliance inspection.  Instead, the compliance burden
has been placed on the owner or operator to attempt to repair the leak
within 5 days of its detection and to completely repair the leak within 15
days.
     The commenter implies that enforcement is unlikely because it must be
proven that recordkeeping and reporting requirements were not met or that
the leaking component was not repaired.  The EPA disagrees.  The regulation
states that compliance will be determined by review of records, reports,
performance test results, and inspections.  By comparing records and
reports of plant performance to the actual sources during an onsite inspec-
tion, enforcement  personnel will  be able to detect unrepaired sources,
unsubstantiated records  regarding delayed repair, falsified records, and
lack of records or reports.  Under these standards, the records and reports
(or lack thereof)  provide usable  evidence of a violation, and enforcement
action  is  likely.  Although the  recordkeeping and reporting requirements,
coupled with  onsite  inspections,  are the only measures to determine
compliance, EPA believes these provisions are adequate to ensure diligent
monitoring and repair of leaks by plant  personnel and effective enforcement
by EPA.
11.6  MISCELLANEOUS
     Comment: Commenter AESP-00001, a private citizen from Strasburg,
Virginia,  does not address  the specific  proposal, but describes instead  the
health  problems for  area residents  exposed to organics emissions caused  by
a printing company located  near  his  home.  He also  describes  the difficul-
ties  in working with company  and government  officials.
                                     11-20

-------
     Response:  The EPA is concerned about the health problems associated
with air toxics contained in the organic emissions subject to control by
these standards.  These emissions can cause short-term and acute health
problems.  Although the company in question may not be subject to these
RCRA standards unless it has certain processes onsite for hazardous waste
treatment, existing standards under the CAA would likely apply to this
site.
                                    11-21

-------
            APPENDIX A



EVOLUTION OF PROMULGATED STANDARDS

-------
     Date
                   TABLE A-l.  EVOLUTION OF THE FINAL RULES
               Event
February 5, 1987


March 23, 1987


September, 1987


August 4, 1988



April 3, 1989



June 13, 1990
Publication of proposed rules in the
Federal Register, 52 FR 3748.

Public hearing regarding the proposed
rules.

Information requests, as authorized under
Section 3007 of RCRA sent to select TSDF.

Work Group Closure meeting on final TSDF
process vent and equipment leaks stand-
ards (#1).  Closure not reached.

Work Group Closure meeting on final TSDF
process vent and equipment leak stand-
ards (#2).  Closure reached.

Final rules signed by EPA Administrator.
                                     A-3

-------
       APPENDIX B



ESTIMATING HEALTH IMPACTS

-------
                                 APPENDIX B
                          ESTIMATING HEALTH  IMPACTS

      Many  adverse  health.effects can result from exposure to air emissions
 from  hazardous waste treatment,  storage, and disposal  facilities (TSDF).
 The major  pathways for human exposure to environmental contaminants  are
 through  inhalation, ingestion, or dermal contact.  Airborne contaminants
 may be toxic to the sites of immediate exposure, such  as the skin, eyes,
 and linings of the respiratory tract.  Toxicants may also cause a  spectrum
 of systemic effects following absorption and distribution to various target
 sites such as the  liver,  kidneys, and central nervous  system.
      Exposure to contaminants in air can be acute, subchronic, or  chronic.
 Acute exposure refers to  a very short-term  (i.e., <24  h), usually  single-
 dose, exposure to a contaminant.  Health effects often associated  with
 acute exposure include:   central nervous system effects such as headaches,
 drowsiness, anesthesia, tremors, and convulsions; skin, eye, and respira-
 tory tract irritation;  nausea; and olfactory effects such as awareness of
 unpleasant or disagreeable odors.  Many of these effects are reversible and
 disappear with cessation of exposure.  Acute exposure  to very high concen-
 trations or to low levels of highly toxic substances can, however,  cause
 serious and irreversible tissue damage,  and even death.  A delayed toxic
 response may also occur following acute exposure to certain agents.
     Chronic exposures  are those that occur for long periods of time (from
many months to several  years).  Subchronic exposure falls between  acute and
chronic exposure,  and usually involves  exposure for a period of weeks or
months.   Generally, the health effects  of greatest concern following inter-
mittent or continuous long-term exposures are those that cause either irre-
versible damage and serious  impairment  to the normal  functioning of the
individual, such  as cancer and organ dysfunctions,  or death.
                                    B-3

-------
     The risk associated with exposure to a toxic agent is a function of
many factors, including the physical  and chemical characteristics of the
substance,  the nature of the toxic response and the dose required to
produce the effect,  the susceptibility of the exposed individual, and the
exposure situation.   In many cases,  individuals may be concurrently or
sequentially exposed to a mixture of compounds, which may influence the
risk by changing the nature and magnitude of the toxic response.
B.I  ESTIMATION OF CANCER POTENCY
     The unit risk estimate (LIRE, unit risk factor) is used by the Environ-
mental Protection Agency (EPA) in its analysis of carcinogens.  It is
defined as the lifetime cancer risk occurring in a hypothetical population
in which all individuals are exposed throughout their lifetime  (assumed to
be 70 years) to an average concentration of 1 /jg/m^ of the pollutant in the
air they breathe.  Unit risk estimates can be used for two purposes:  (1)
to compare the carcinogenic potency of several agents with one another, and
(2) to give a rough indication of the public health risk that might be
associated with estimated air exposure to these agents.1
     In the development of unit risk estimates, EPA assumes that if experi-
mental data show that a substance is carcinogenic in animals, it may also
be carcinogenic in humans.  The EPA also assumes that any exposure to a
carcinogenic substance poses some risk.2  This nonthreshold presumption is
based on the view that as little as one molecule of a carcinogenic sub-
stance may be sufficient to transform a normal cell into a cancer cell.
Exposed individuals are represented by a referent male having a standard
weight, breathing rate, etc.  (no reference is made to factors such as race
or state of health).
     The data used for the quantitative estimate can be of two  types:   (1)
lifetime animal studies, and  (2) human studies where excess cancer risk has
been associated with exposure to the agent.   It  is assumed, unless evidence
exists to the contrary, that  if a carcinogenic response occurs  at the dose
levels used  in a study, then  responses will occur at all  lower  doses with
an incidence determined by the extrapolation model.
     There is no solid scientific basis for any mathematical extrapolation
model that relates carcinogen exposure to cancer risks at the extremely  low
concentrations that must be dealt with in evaluating environmental hazards.
                                    B-4

-------
 For practical  reasons,  such low levels  of risk cannot be measured  directly
 either by animal  experiments or by epidemiologic studies.  We must,  there-
 fore,  depend on our current understanding of the mechanisms of carcinogen-
 esis for guidance as to which risk model  to use.  At  present,  the  dominant
 view of the carcinogenic process is that  most agents  that cause cancer also
 cause irreversible damage to DNA.   This position is  reflected by the fact
 that a very large proportion of agents  that cause cancer are also  muta-
 genic.  There  is  reason to expect  that  the quantal type  of biological
 response,  which is characteristic  of mutagenesis,  is  associated with a
 linear nonthreshold dose-response  relationship.   Indeed,  there is  substan-
 tial  evidence  from mutagenesis  studies with both ionizing radiation  and a
 wide variety of chemicals that  this type  of dose-response model  is the
 appropriate one to use.   This  is particularly true at  the lower end  of the
 dose-response  curve.  At  higher doses, there can be an upward  curvature
 probably reflecting the effects of multistage processes  on  the mutagenic
 response.   The linear nonthreshold dose-response relationship  is also  con-
 sistent  with the  relatively  few epidemiologic studies  of  cancer responses
 to  specific  agents  that  contain  enough information to make  the evaluation
 possible (e.g., radiation-induced  leukemia,  breast and thyroid cancer,  skin
 cancer induced  by  arsenic  in drinking water,  liver cancer induced by afla-
 toxins  in  the  diet).  There  is  also  some evidence from animal  experiments
 that is  consistent  with the  linear nonthreshold  model  (e.g.,  liver tumors
 induced  in mice by  2-acetylaminofluorene in  the  large-scale  EDgi study  at
 the National Center for Toxicological Research and the initiation stage of
 the two-stage carcinogenesis model  in rat  liver  and mouse skin).
     Because of this evidence,  the  linear nonthreshold model is considered
 to be  a  viable model for any carcinogen, and  unless there is direct evi-
 dence  to the contrary,  it  is used  as the primary basis for risk extrapola-
 tion to  low  levels of exposure.3
     The mathematical formulation chosen to describe the  linear non-
threshold dose-response relationship at  low doses is  the  linearized multi-
stage model.  The linearized multistage  model is applied to the original
unadjusted animal  data.   Risk estimates  produced by this model from the
animal  data are then scaled to a human equivalent estimate of risk.  This
is done by multiplying the estimates by  several factors to adjust for
                                    B-5

-------
experiment duration, species differences, and, if necessary,  route conver-
sion.  The conversion factor for species differences is currently based on
models for equitoxic dose.4  The unit risk values estimated by this method
provide a plausible, upperbound limit on public risk at lower exposure
levels if the exposure is accurately quantified; i.e.,  the true risk is
unlikely to be higher than the calculated level and could be substantially
lower.
     The method that has been used in most of the EPA's quantitative risk
assessments assumes dose equivalence in units of mg/body weights/3 for
equal tumor response in rats and humans.  This method is based on adjust-
ment for metabolic differences.  It assumes that metabolic rate is roughly
proportional to body surface areas and that surface area is proportional to
2/3 power of body weight (as would be the case for a perfect sphere).  The
estimate is also adjusted for lifetime exposure to the carcinogen consider-
ing duration of experiment and animal lifetime.5.6
     For unit risk estimates for air, animal studies using exposure by
inhalation are preferred.  When extrapolating results from the inhalation
studies to humans, consideration is given to the following factors:
     •    The deposition of the inhaled compound throughout the
          respiratory tract
     •    Retention half-time of the inhaled particles
     •    Metabolism of the inhaled compound
     •    Differences in sites of tumor  induction.
     Unit risk estimation from animal studies is only an approximate indi-
cation of the actual risk in populations exposed to known concentrations of
a carcinogen.  Differences between species  (lifespan, body size, metabo-
lism, immunological responses, target site  susceptibility), as well as
differences within species  (genetic variation, disease state, diet), can
cause actual risk to be much different.  In human populations, variations
occur in genetic constitution, diet, living environment, and activity
patterns.  Some populations may demonstrate a higher susceptibility due to
certain metabolic or inherent differences in their  response to the effects
of carcinogens.  Also, unit risk estimates  are based on exposure to a
referent adult male.  There may be an increased  risk with exposure to
                                    B-6

-------
 fetuses,  children,  or young  adults.   Finally, humans  are exposed to  a  vari-
 ety  of  compounds,  and the  health effects, either synergistic,  additive,  or
 antagonistic,  of exposure  to complex  mixtures of chemicals  are not
 known.7.8
 B.I.I   EPA  Unit Risk Estimates
     The  EPA  has developed unit risk  estimates for about 71 compounds  that
 are  either  known or suspect  carcinogens and that could be present at a TSDF
 (Table  B-l).   Constituents were drawn from the Agency's final  rule on  the
 identification and  listing of hazardous waste (Appendix VIII)9  and from  the
 Industry  Studies Data Base,  a hazardous waste data base developed by EPA's
 Office  of Solid Waste.10  The unit risk estimates in  Table  B-l  have  been
 derived by  the Agency's Carcinogen Assessment Group,!1 and most have been
 verified  by the Agency's Carcinogen Risk Assessment Verification Enterprise
 (CRAVE) or  are under CRAVE review.  As shown in Table B-l, these estimates
 range in  value from 4.7 x  10-7 (/jg/m3)'1 for methylene chloride to
 3.3  x 10'5  (pg/m3)'1 for dioxin.
 B.I.2   Composite Unit Risk Estimate
     To estimate the cancer  potency of TSDF air emissions, a composite unit
 risk estimate approach was adopted to address the problem of dealing with
 the  large number of toxic chemicals that are present at TSDF.  Using a
 composite estimate  rather than individual  unit risk estimates simplifies
 the  risk  assessment so that  calculations do not need to be performed for
 each chemical emitted.   The  composite risk estimate is combined with esti-
 mates of  ambient concentrations of total organics and population exposure
 to estimate the additional cancer incidence in the general population  and
 the maximum individual  risk  due to TSDF emissions.
     Because detailed equipment leak emission estimates are available and
 because cancer incidence and maximum individual  risk are proportional to
 both the unit risk  estimates and emissions,  an emission-weighted averaging
 technique was used.  In  calculating the emission-weighted average,  the
emission estimate  for a  compound is multiplied by the unit risk estimate
for that compound.   The  emission-weighted  arithmetic average is computed as
follows:
                                    B-7

-------
TABLE B-l.  TSDF CARCINOGEN LIST

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Constituent
acetaldehyde
(75-07-0)
aery 1 amide
(79-06-1)
acrylonitrile
(107-13-1)
aldrin
(309-00-2)
aniline
(62-53-3)
arsenic
(7440-38-2)
benz( a) anthracene
(56-55-3)
benzene
(71-43-2)
benzidine
(92-87-5)
benzo(a)pyrene
(50-32-8)
beryllium
(7440-41-7)
bis(chloroethyl)
ether (111-44-4)
bis(chloromethyl)
ether (542-88-1)
1,3-butadiene
(106-99-0)
cadmium
(7440-43-9)
Unit risk
estimate.
to/"3)'1
2.2xlO-6
1.1x10-3
6.8xlO-5
4.9xlO-3
7.4xlO-6
4.3x10-3
8.9xlO'4
8.3X10'6
6.7xlO-2
1.7xlO-3
2.4x10-3
3.3xlO-4
2.7x10-3
2.8xlO-4
1.8x10-3
Basis3
CRAVE verified
(class B2)
CAG URE
(class B2)
CRAVE verified
URE (class Bl)
CRAVE verified
URE (class B2)
CAG URE
(class C)
CRAVE verified
(class A)
CAG URE
(class B2)
CRAVE verified
(class A)
CRAVE verified
URE (class A)
CAG URE
(class B2)
CAG URE
(class B2)
CRAVE verified
URE (class B2)
CAG URE
(class A)
CRAVE verified
URE (class B2)
CRAVE verified
URE (class Bl)
                                   (continued)
                B-8

-------
TABLE B-l (continued)

16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
Constituent
carbon tetra-
chloride (56-23-5)
chlordane
(12789-03-6)
chloroform
(67-66-3)
chloromethane
(74-87-3)
chloromethyl methyl
ether (107-30-2)
chromium VI
(7440-47-3)
DDT
(50-29-3)
dibenz(a.h)
anthracene
(53-70-3)
l,2-dibromo-3-
chloropropane
(96-12-8)
1,2-dichloroethane
(107-06-2)
1,1-dichloro-
ethylene (75-35-4)
dieldrin
(60-57-1)
2,4-dinitrotoluene
(121-14-2)
Unit risk
estimate.
Oig/m3)-i
1.5x10-5
3.7xlO'4
2.3xlO-5
3.6x10-6
2.7x10-3
1.2x10-2
9.7x10-5
1.4xlO-2
6.3x10-3
2.6x10-5
5.0x10-5
4.6x10-3
8.8x10-5
Basis3
CRAVE verified
URE (class B2)
CRAVE verified
URE (class B2)
CRAVE verified
(class B2)
ECAO URE
(class C)
CAG URE
(class A)
CRAVE verified
URE (class A)
CRAVE verified
URE (class B2)
CAG URE
(class B2)
CAG URE
(class B2)
CRAVE verified
URE (class B2)
CRAVE verified
URE (class C)
CRAVE verified
URE (class B2)
CAG URE
(class B2)
                            (continued)
         B-9

-------
TABLE B-l (continued)

29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
Constituent
1,4-dioxane
(123-91-1)
1 , 2-di pheny 1 hydrazi ne
(122-66-7)
epichlorohydrin
(106-89-8)
ethyl ene di bromide
(106-93-4)
ethylene oxide
(75-21-8)
formaldehyde
(50-00-0)
gasoline
(8006-61-9)
heptachlor
(76-44-8)
heptachlor epoxide
(1024-57-3)
hexachlorobenzene
(118-74-1)
hexachlorobutadiene
(87-68-3)
hexachlorocyclohexane
(no CAS #)
alpha-hexachloro-
cyclohexane
(319-84-6)
beta-hexachloro-
cyclohexane
(319-85-7)
Unit risk
estimate.
0*g/m3)-i
1.4xlO-6
2.2xlO'4
1.2x10-6
2.2xlO'4
l.OxlO'4
1.3xlO-5
6.6xlO-7
1.3xlO-3
2.6xlO-3
4.9xlO-4
2.2xlO-5
5.4xlO'4
1.8x10-3
5.3xlO'4
Basis3
CAG URE
(class B2)
CRAVE verified
(class B2)
CRAVE verified
URE (class B2)
CRAVE verified
URE (class B2)
CAG URE
(class B1-B2)
CAG URE
(class Bl)
CAG URE
(class B2)
CRAVE verified
URE (class B2)
CRAVE verified
URE (class B2)
CAG URE
(class B2)
CRAVE verified
URE (class C)
CRAVE verified
URE (class B2)
CRAVE verified
URE (class B2)
CRAVE verified
URE (class B2)
                             (continued)
         B-10

-------
                       TABLE B-l  (continued)

43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54,
Constituent
gamma-hexachloro-
cyclohexane
(lindane) (58-89-9)
hexach 1 orod i benzo-
p-dioxin,l:2 mixture
(57653-85-7 or
19408-74-3)
hexachloroethane
(67-72-1)
hydrazine
(302-01-2)
3-methyl chol anthrene
(56-49-5)
4,4'-methylene-bis
(2-chloroaniline)
(101-14-4)
methylene chloride
(75-09-2)
methyl hydrazine
(60-34-4)
nickel refinery
dust (7440-02-0)
nickel subsulfide
(12035-72-2)
2-nitropropane
(79-46-9)
n-nitrosodi-n-
Unit risk
estimate.
0*g/m3)-i
3.8xlO-4
1.3x10°
4.0x10-6
2.9x10-3
2.7x10-3
4.7x10-5
4.7x10-7
3.1x10-4
2.4x10-4
4.8x10-4
2.7x10-3
1.6x10-3
Basis3
CRAVE verified
URE (class C)
CRAVE verified
URE (class B2)
CRAVE verified
URE (class C)
CAG URE
(class B2)
CAG URE
(class B2)
CAG URE
(class B2)
CAG URE
UCR (class B2)
ECAO URE
(class B2)
CRAVE verified
URE (class A)
CRAVE verified
URE (class B2)
CAG URE
(class B2)
CRAVE verified
     butyl amine
     (924-16-3)

55.  n-nitroso-
     diethylamine
     (55-18-5)
4.3x10-2
                   URE (class B2)
CRAVE verified
URE (class B2)
                                                   (continued)
                              B-ll

-------
                      TABLE B-l (continued)
                               Unit risk
                               estimate.
     Constituent
    Basis3
56.  n-nitroso-                1.4x10-2
     dimethyl amine
     (62-75-9)

57.  n-nitroso-n-              8.6x10-2
     methyl urea
     (684-93-5)

58.  n-nitroso-                6.1x10"^
     pyrrolidine
     (930-55-2)

59.  pentachloronitro-         7.3x10-5
     benzene
     (82-68-8)

60.  polychlorinated           1.2x10-3
     biphenyls
     (1336-36-3)

61.  pronamide                 4.6x10*6
     (23950-58-5)

62.  reserpine                 3.0x10-3
     (50-55-5)

63.  2,3,7,8-tetrachloro-      3.3xlQ-5
     dibenzo-p-dioxin          (pg/m3)"1
     (1746-01-6)

64.  1,1,2,2-tetra-            5.8x10-5
     chloroethane
     (79-34-5)
CRAVE verified
URE (class B2)
CAG URE
(class B2)
CRAVE verified
URE (class B2)
CAG URE
(class C)
CAG URE
(class 82)
CAG URE
(class C)

CAG URE
(class B2)

CAG URE
(class B2)
CRAVE verified
URE  (class C)
65.
66.
67.
tetrach 1 oroethy 1 ene
(127-18-4)
thiourea
(62-56-6)
toxaphene
(8001-35-2)
5.8xlO-7
5.5xlO-4
3.2x10-3
CAG URE
(class B2)
CAG URE
(class B2)
CRAVE verified
URE (class B2)
                                                   (continued)
                              B-12

-------
                      TABLE B-l (continued)
     Constituent
Unit risk
estimate.
0*g/m3)-i
    Basis3
68.  1,1,2-trichloro-
     ethane
     (79-00-5)

69.  trichloroethylene
     (79-01-6)

70.  2,4,6-trichloro-
     phenol
     (88-06-2)

71.  vinyl chloride
     (75-01-4)
1.6xlO-5



1.7xlO-6


5.7x10-6



4.1xlO-6
CRAVE verified
URE (class C)
CAG URE
(class 82)

CRAVE verified
URE (class B2.)
CAG URE
(class A)
( )  = Chemical Abstracts Service (CAS) Number.

aCancer unit risk estimates (UREs)  were either (1) verified by
 the Carcinogen Risk Assessment Verification Enterprise (CRAVE)
 work group or (2) established by the Carcinogen Assessment
 Group (CAG), but not yet verified by CRAVE.  The unit risk
 estimates for chloromethane and methyl hydrazine were derived
 by  the Environmental Criteria and Assessment Office (ECAO).

Note:  The constituents on this list and the corresponding unit
       risk estimates are subject to change.
                              B-13

-------
                               N
                               I (REi  • ER.)
                                      	  ,                       (B-l)
where
      RE = weighted average unit risk estimate
     RE. = unit risk estimate for compound i
     ER. = TSDF emissions for compound i
     ER. = total emissions for TSDF.

Using this type of average would give the same result as calculating the
risk for each chemical involved.
     The calculation of the composite unit risk estimate for the TSDF
equipment leak baseline case is illustrated in Table B-2.  The table lists
the compounds included in the development of the composite risk estimate,
total nationwide baseline TSDF equipment leak emissions by compound, the
unit risk estimate by compound, and the weighted-average unit risk
estimate.  The composite unit risk estimate calculated in this analysis for
the baseline TSDF equipment leaks is 4.5 x 10'6 (/*g/m3)-l.  In the
calculation of the TSDF equipment leak unit risk factor, -the emissions
shown in Table B-2 (i.e., 18,413 Mg/yr) represent leaks from equipment
associated with hazardous waste management units other than solvent
recycling units.  The data base used for the analysis did not contain
information on TSDF solvent recycling units or waste solvent treatment
facilities.  Emissions from these units were calculated separately and were
estimated to be an additional 7,830 Mg/yr, bringing the total TSDF
equipment leaks emission estimate to about 26,200 Mg/yr.  Available data in
the waste characterization data base (see Appendix D, Section D.2) were not
adequate to allow recycling unit equipment leak emissions to be estimated
on a constituent basis and included in the calculation.  Because it is not
unreasonable to assume a similar mix of constituents in recycling units as
in other waste management units, and because available data do not suggest
otherwise, for the purpose of estimating impacts the same unit risk factor
                                   B-14

-------
             TABLE B-2.  CALCULATION OF THE EQUIPMENT LEAK EMISSIONS-WEIGHTED COMPOSITE UNIT RISK ESTIMATE  (URE)
                                                      FOR THE BASELINE"
DO
I

Compound
Acetaldehyde
Aery 1 amide
Aery loni tr i le
Aldrin
Ani 1 ine
Benzene
Benzo (a) pyrene
Benz (a) anthracene
b i s (ch 1 oromethy 1 ) ether
b i s (2-ch 1 oroethy 1 ) ether
Butadiene (1,3)
Carbon tetrach loride
Ch lordane
Ch lorof orm
Ch loromethane
Ch 1 oromethy 1 (methy 1 ) ether
DDT (4,4*)
D i benz (a , h) anthracene
Di bromo- (1 , 2) -ch 1 oropropane(3)
D i bromoethane
Dich loroethane(l ,2)
Dichl oroethy lene (1,1) (54)
D i n i troto 1 uene (2,4)
Dioxane(l,4)
Dipheny lhydrazine(l ,2)
Epi ch 1 orohydr i n
Ethylene di bromide
Ethy leneoxide
Forma Idehyde
Gasol ine (58)
Heptach lor
Hexach 1 orobenzene
Hexach 1 orobutad iene
Hexach 1 oroethane
Hydrazine
Lindane
Methyl hydrazine
Methy 1 cho 1 anthrene (3)
Methy lene chloride
Ni tropropane(2)

emissions, Mg/yr
477.153
0
30.06
0.002
10.514
37.03
0.003
0
0.008
0.001
2.362
118.1
0.003
128.504
0.124
0
0.005
0.0E+00
0.003
0
392.863
1.965
0.187
12.065
0
2.884
8.703
0
31.61
39.24
0
6.613
76.91
93.93
1.087
0.0E+00
9.212
0.001
369.974
8.878

UREb
2 . 2E-06
1 . 1E-03
6.8E-05
4.9E-03
7.4E-06
8.3E-06
1 . 7E-03
8 . 9E-04
2 . 7E-03
3 . 3E-04
2 . 8E-04
1 . 5E-05
3 . 7E-04
2.3E-05
3.6E-06
2 . 7E-03
3.0E-04
1.4E-02
6.3E-03
2 . 2E-04
2.6E-05
5.0E-05
8.8E-05
1 . 4E-06
2 . 2E-04
1 . 2E-06
2 . 2E-04
1.0E-04
1 . 3E-05
6.6E-07
1 . 3E-03
4 . 9E-04
2 . 2E-05
4.0E-06
2.9E-03
3.8E-04
3 . 1E-04
2 . 7E-03
4 . 7E-07
2.7E-03
URE x emissions for chemical^
Total TSDF emissions
5 . 70E-08
0.00E+00
1.11E-07
5.32E-10
4 . 23E-09
1.67E-08
2.77E-10
0.00E+00
1 . 17E-09
1.79E-11
3.59E-08
9.62E-08
6.03E-11
1.61E-07
2.42E-11
0.00E+00
8.15E-11
.0.00E+00
1.03E-09
0.00E+00
5 . 55E-07
5.34E-09
7.98E-10
9.17E-10
0.00E+00
1.88E-10
1.04E-07
0.00E+00
2.23E-08
1 . 14E-09
0.00E+00
1.76E-07
9.19E-08
2.04E-08
1.71E-07
0.00E+00
1.55E-07
1.47E-10
9.44E-09
1 . 30E-06
                                                                                                           (continued)

-------
                                                    TABLE  B-2  (continued)
DO
I
CTi
Basel ine
Compound emissions, Mg/yr
N-Ni trosopyrrol idine
N-n i troso-n-methy 1 urea
Polychlorinated bipheny Is (11)
Tetrach 1 orod i benzo (2,3,7,8) -p-d i ox i n
Tetrach 1 oroethane (1 ,1,2,2)
Tetrach 1 oroethy 1 ene
Thiourea
Toxaphene
Trichl oroethane (1,1, 2)
Tr i ch 1 oroethy 1 ene
Trichlorophenol (2,4,6)
Vinyl chloride
Total nationwide
baseline emissions
0
0
3.1E-02
0.0E+00
182.478
.334.269
8.731
0
483.978
837.562
0.007
7.087

18,413=
UREb
6.1E-04
8.6E-02
1.2E-03
3 . 3E+01
5.8E-05
6.8E-07
5.6E-04
3.2E-03
1.6E-06
1.7E-06
5 . 7E-06
4.1E-06


URE x emissions for chemical**
Tota 1 . TSDF emi ss i ons
0.00E+00
0.00E+00
2.02E-09
0.00E+00
5.7SE-07
1.06E-08
2.61E-07
0.00E+00
4.21E-07
7.73E-08
2.17E-12
1.68E-09

4.5 x 10~s
TSDF = Treatment, storage, and disposal facility.
•This table illustrates the calculation of the composite unit risk estimate for the baseline.  This procedure
 was repeated for each of the control options.

bUnits are (/Jg/m3)'1'
<=The emissions shown, i.e., 18,413 Mg/yr, represent leaks from equipment associated with hazardous waste
 management units other than solvent recycling units.   Emissions from these units were calculated
 separately and were estimated to be an additional 7,830 Mg/yr, bringing the total TSDF equipment  leaks
 emission estimate to about 26,200 Mg/yr.  Available data in the waste characterization data base
 (See Appendix 0, Section D.2) were not adequate to allow recycling unit equipment leak emissions to be
 estimated on a constituent basis and  included in the calculation.

-------
 was used for all  equipment leak and recycling unit health impacts.  (See
 Chapter 7 of the BID for further discussion of the representativeness of
 the waste streams.)
      The unit risk factor calculation was repeated for each of the control
 options based on the compound-specific emission estimates associated with
 the option.  The resulting composite unit risk estimates for control
 options were not significantly different from the baseline value.   The
 emission rates of specific compounds under the various control options
 changed proportionally under each option.
      In addition  to  the uncertainties in estimates of emissions, other
 difficulties arise in averaging the UREs for specific constituents to
 develop a composite  URE.   Unit risk estimates have not been developed for
 all  of the pollutants of  concern,  due,  in part,  to insufficient data.
 Various options  for  dealing  with  this problem were considered.  The EPA
 selected an approach in which  only those carcinogens  for which unit risk
 estimates were available  would be included  in the analysis  of  cancer risk.
 Consideration was  also given  to adding  the  weighted risk estimates for only
 those  compounds having similar EPA classifications,13  i.e.,  to present the
 composite unit risk  estimate  and  associated cancer risks separately for
 Group  A compounds  (human  carcinogens),  Group  B  compounds (probable human
 carcinogens),  and  Group C compounds  (possible human carcinogens).   However,
 because only  about 1  percent of the  weighted  composite risk  estimate is
 attributed  to Group  A compounds and  about 5 percent for Group  C, EPA
 elected to  present the risk associated  with all  three  groups combined.
 B.2  DETERMINING NONCANCER HEALTH  EFFECTS
     Although  cancer  is of great concern  as  an  adverse  health  effect asso-
 ciated  with exposure  to a chemical or a mixture  of  chemicals,  many  other
 health  effects may be  associated with such  exposures.    These effects may
 range from subtle  biochemical,  physiological, or  pathological  effects  to
 gross effects  such as  death.   The  effects of  greatest  concern  are  the  ones
 that are  irreversible  and impair the normal functioning  of the  individual.
 Some of these effects  include  respiratory toxicity, developmental  and
 reproductive toxicity, central nervous system effects,  and other systemic
effects such as liver and kidney toxicity, cardiovascular toxicity,  and
 immunotoxicity.
                                   B-17

-------
B.2.1  Health Benchmark Levels
     For chemicals that give rise to toxic endpoints other than cancer and
gene mutations, there appears to be a level of exposure below which adverse
health effects usually do not occur.  This threshold-of-effect concept
maintains that an organism can tolerate a range of exposures from zero to
some finite value without risk of experiencing a toxic effect.  Above this
threshold, toxicity is observed as the organism's homeostatic, compensat-
ing, and adaptive mechanisms are overcome.  To provide protection against
adverse health effects in even the most sensitive individuals in a
population, regulatory efforts are generally made to prevent exposures from
exceeding a health "benchmark" level that is below the lowest of the thres-
holds of the individuals within a population.
     Benchmark levels, termed reference doses (RfDs),  are operationally
derived from an experimentally obtained no-observed-effect level or a
lowest-observed-effect level by consistent application of generally order-
of-magnitude uncertainty factors that reflect various types of data used to
estimate the RfD.  The RfD is an estimate  (with uncertainty spanning
perhaps an order of magnitude or greater) of daily exposure to the human
population (including sensitive subpopulations) that is likely to be
without an appreciable risk of deleterious effect.
     The Agency has developed verified oral RfD for a large number of chem-
icals, but has only recently established an internal work group to begin
the process for establishing inhalation RfDs.  Agency-verified inhalation
reference doses for acute and chronic exposures will be used  in this analy-
sis when they become available.  Unverified inhalation reference doses that
have been developed by the Agency may be used on an interim basis after
careful review of the supporting data base.
B.2.2  Noncarcinoqenic Toxic Chemicals of  Concern
     A preliminary list of 179 TSDF chemicals of concern for  the noncancer
health assessment is shown in Table B-3.   Constituents were drawn from the
Agency's final rule on the identification  and listing of hazardous waste
(Appendix VIII)!4 and from the Industry Studies Data Base, a  hazardous
waste data base developed by EPA's Office  of Solid Waste.15   To be selected
from these sources, the chemical must have had either an Agency-verified
                                   B-18

-------
     TABLE B-3.  TSDF CHEMICALS - NONCANCER HEALTH EFFECTS ASSESSMENT
         Chemical
      Chemical
acetone (67-64-1)
acetaldehyde3 (75-07-0)
acetonitrile (75-05-8)
acetophenone (98-86-2)
acetyl chloride (75-36-5)
l-acetyl-2-thiourea (591-08-2)
acrolein3 (107-02-8)
acrylic acid (79-10-7)
acrylonitrile3 (107-13-1)
aldicarb (116-06-3)
aldrin3 (309-00-2)
ally! alcohol (107-18-6)
ally! chloride3 (107-05-1)
aluminum phosphide  (20859-73-8)
5-aminomethyl-3-isoxazolol
  (2763-96-4)
4-aminopyridine (504-24-5)
ammonia (7664-41-7)
ammonium vanadate  (7803-55-6)
antimony (7440-36-0)
arsenic3 (7440-38-2)
barium (7440-39-3)
barium cyanide (542-62-1)
benzidine3 (92-87-5)
benzoic acid (65-85-0)
beryllium3 (7440-41-7)
1,1-biphenyl (92-52-4)
bis(2-ethylhexyl)phthalatea
  (117-81-7)
bromodichloromethane (75-27-4)
bromoform (75-25-2)
butanol (71-36-3)
cadmium3 (7440-43-9)
calcium chromate3  (13765-19-0)
calcium cyanide  (592-01-8)
carbon disulfide (75-15-0)
carbon oxyfluoride  (353-50-4)
carbon tetrachloride3 (56-23-5)
chlordane3 (12789-03-6)
chlorine (7782-50-5)
chloroacetaldehyde  (107-20-0)
2-chloro-l,3-butadiene
(126-99-8)
chloroform3 (67-66-3)
chloromethane3 (74-87-3)
3-chloropropionitrile (542-76-7)
chromium III (7440-47-3)
chromium VI (7440-47-3)
copper cyanide (544-92-3)
cresols3 (1319-77-3)
crotonaldehyde (4170-30-3)
cumene (98-82-8)
cyanide (57-12-5)
cyanogen (460-19-5)
cyanogen bromide3  (506-68-3)
cyanogen chloride  (506-77-4)
cyclohexanone (108-94-1)
2,4 D (dichlorophenoxyacetic
  acid) (94-75-7)
DDT3 (50-29-3)
                      (continued)
                                   B-19

-------
                           TABLE B-3 (continued)
         Chemical
       Chemical
decabromodiphenyl oxide (1163-19-5)
di-n-butyl phthalate (84-74-2)
1,2-dichlorobenzene (95-50-1)
l,4-dichlorobenzenea (106-46-7)
dichlorodifluoromethane (75-71-8)
l,l-dichloroethanea (75-34-3)
l,l-dichloroethylenea (75-35-4)
2,4-dichlorophenol (120-83-2)
l,3-dichloropropenea (542-75-6)
dieldrina (60-57-1)
diethyl phthalate  (84-66-2)
dimethoate (60-51-5)
dimethyl amine (124-40-3)
dimethyl aniline  (121-69-7)
(alpha, alpha) dimethyl
  phenethylamine  (122-09-8)
dimethylterephthalate (120-61-6)
2,4-dim'trophenol  (51-28-5)
dinoseb (88-85-7)
diphenyl amine (122-39-4)
disulfoton (298-04-4)
endosulfan (115-29-7)
endothall (129-67-9)
endrin  (72-20-8)
epichlorohydrina  (chloro-2,3-
  epoxy-propane)  (106-89-8)
ethyl acetate  (141-78-6)
ethyl benzene  (100-41-4)
ethylene glycol  (107-21-1)
ethylene oxide3  (75-21-8)
ethylene thiourea3 (96-45-7)
fluoracetic acid, sodium salt
  (62-74-8)
fluoride (16984-48-8)
fluorine (7782-41-4)
formaldehyde9 (50-00-0)
formic acid (64-18-6)
freon 113  (76-13-1)
furan (110-00-9)
gamma-hexach1orocyclohexane
  (lindane) (58-89-9)
heptachlor3 (76-44-8)
heptachlor epoxide3  (1024-57-3)
hexachlorobutadiene3  (87-68-3)
hexachlorocyclopentadiene  (77-47-4)
hexachloroethane3  (67-72-1)
hydrogen chloride  (7647-01-0)
hydrogen cyanide  (74-90-8)
hydrogen sulfide  (7783-06-4)
isobutyl alcohol  (78-83-1)
lead  (7439-92-1)
maleic hydrazide3  (123-33-1)
malononitrile (109-77-3)
mercury  (7439-97-6)
methacrylonitrile  (126-98-7)
methomyl  (16752-77-5)
methoxyclor  (72-43-5)
methyl bromide  (bromomethane)
   (74-83-9)
                                                                (continued)
                                    B-20

-------
                            TABLE  B-3  (continued)
          Chemical
       Chemical
methyl  chloroform  (1,1,1-
  trichloroethane)  (71-55-6)
methylene  chloride3  (75-09-2)
methyl  ethyl  ketone  (78-93-3)
methyl  iodide3  (74-88-4)
methyl  isobutyl  ketone  (108-10-1)
methyl  isocyanate  (624-83-9)
2-methyl lactonitrile  (75-86-5)
methyl  parathion  (298-00-0)
nickel  carbonyl3  (13463-39-3)
nickel  cyanide  (557*19-7)
nickel  refinery dust3  (7440-02-2)
nitric  oxide  (10102-43-9)
nitrobenzene3 (98-95-3)
4-nitroquinoline-1-oxide (56-57-5)
osmium  tetroxide  (20816-12-0)
pentachlorobenzene3  (608-93-5)
pentachloroethane3 (76-01-7)
pentachloronitrobenzene (82-68-8)
pentachlorophenol3 (87-86-5)
phenol  (108-95-2)
m-pheny1enedi amine3  (25265-76-3)
phenylmercuric acetate  (62-38-4)
phosgene (75-44-5)
phosphine  (7803-51-2)
potassium cyanide (151-50-8)
potassium silver cyanide (506-61-6)
pronamide3 (23950-58-5)
propanenitrile (107-12-0)
n-propylamine (107-10-8)
2-propyn-l-ol (107-19-7)
pyridine (110-86-1)
selenious  acid  (selenium  dioxide)
   (7783-00-8)
selenourea (630-10-4)
silver  (7440-22-4)
silver  cyanide  (506-64-9)
silvex  (93-72-1)
sodium  azide  (26628-22-8)
sodium  cyanide  (143-33-9)
styrene3 (100-42-5)
strychnine (57-24-9)
1,2,4,5-tetrach1orobenzene
   (95-94-3)
1,1,1,2-tetrachloroethane3
   (630-20-6)
tetrachloroethylene3  (127-18-4)
2,3,4,6-tetrach1orophenol
   (58-90-2)
tetraethyl dithiopyrophosphate
   (3689-24-5)
tetraethyl lead (78-00-2)
thallic oxide (1314-32-5)
thallium (7440-28-0)
thallium (1) acetate  (563-68-8)
thallium (1) carbonate  (6533-73-9)
thallium (1) chloride (7791-12-0)
thallium (1) nitrate  (10102-45-1)
thallium (1) selenite (12039-52-0)
thallium (1) sulfate  (10031-59-1)
thiomethanol (methyl mercaptan)
   (74-93-1)
thiosemicarbazide (79-19-6)
                                                                  (continued)
                                   B-21

-------
                           TABLE B-3 (continued)
         Chemical
       Chemical '
thiram (137-26-8)
toluene (108-88-3)
1,2,4-trichlorobenzene (120-82-1)
1,1,2-trichloroethanea (79-00-5)
tri chloromonof1uoromethane
  (75-69-4)
2,4,5-trichlorophenol3 (95-95-4)
1,2,3-trichloropropane (96-18-4)
vanadium pentoxide (1314-62-1)
warfarin (81-81-2)
xylene(s) (1330-20-7)
zinc cyanide (557-21-1)
zinc phosphide (12037-79-5)
zineb* (12122-67-7)
( ) = Chemical Abstracts Service (CAS) Number.
aCarcinogen.
                                    B-22

-------
oral  reference dose  (as of September 30,  1987),16 or a  Reference Air Con-
centration  (RAC) found in the Agency's proposed rule on the burning of
hazardous waste  in boilers and industrial furnaces.1?   Additional chemicals
were  added  to Table  B-3 based on knowledge of a high toxicity associated
with  that substance.
B.3   EXPOSURE ASSESSMENT
      Three  models were used to assess exposure, and ultimately risks, for
air emissions from TSDF.  The Human Exposure Model  (HEM) was used to calcu-
late  the number  of people exposed to predicted ambient  concentrations of
volatile organics from equipment leaks at each of about 2,300 TSDF in the
United States which  included the more than 1,400 TSDF that are likely to be
affected by the  process vent and equipment leak standards.  The results of
these analyses were  used to quantify annual cancer  incidence.  To determine
the maximum lifetime cancer risk to the most exposed individual, the
Industrial  Source Complex Long-Term (ISCLT) model was used to estimate the
highest ambient  concentrations of organics in the vicinity of a worst-case
TSDF.  In addition,  the ISCLT model was used in the evaluation of chronic
noncancer health effects.  Finally, the Industrial Source Complex Short-
Term  (ISCST) model was used to estimate ambient concentrations of
individual chemicals of concern for the acute noncancer health effects
assessment and as a  preliminary screen for the chronic noncancer health
effects assessment.  Each of these is briefly described below.
B.3.1  Human Exposure Model
      In addition to  the composite unit risk estimate, a numerical expres-
sion of public exposure to the pollutant is needed to produce quantitative
expressions of cancer incidence.  The numerical expression of public
exposure is based on two estimates:  (1) an estimate of the magnitude and
location of long-term average air "concentrations of the pollutant in the
vicinity of emitting sources based on air dispersion modeling; and (2)  an
estimate of the number of people living in the vicinity of emitting
sources.
     The EPA uses the Human Exposure Model (HEM) to make these quantitative
estimates of public exposure and risk associated with a pollutant.  The HEM
uses an atmospheric dispersion model  that includes local meteorological
                                   B-23

-------
data (i.e.,  nearest National Weather Bureau meteorology station data) and a
population distribution estimate based on 1980 Bureau of Census data to
calculate public exposure.18
     The dispersion model in HEM used data for a model plant that was
placed at each TSDF location (initially about 2,300 sites).  The location
of each TSDF was obtained from the EPA's TSDF Industry Profile.  Inputs to
the initial  run included a unit cancer potency estimate (1.0) and a unit
emission rate (10,000 kg VOC/yr).   In addition, an exit velocity and an
effluent outgas temperature of 0.1 m/s and 293 °C were assumed.  These
inputs were used to estimate the concentration and distribution of the
pollutant at distances of 200 m to 50 km from the source.  The population
distribution estimates for people residing near the source are based on
Bureau of Census data contained in the 1980 Master Area Reference File
(MARF) data base.19  The data base is broken down into enumeration
district/block group (ED/BG) values.  The MARF contains the population
centroid coordinates (latitude and longitude) and the 1980 population of
each ED/BG (approximately 300,000) in the United States.  By knowing the
geographic location of the plant (latitude and longitude), the model can
identify the ED/BG that fall within the 50-km radius used by HEM.
     The HEM multiplies the concentration of the pollutant at ground level
at each of the 160 receptors around the plant by the number of people
exposed to that concentration to produce the exposure estimates.  The total
exposure, as calculated by HEM, is illustrated by the following equation:
                                       N
                     Total exposure =  E (P.)(C-)  ,                   (B-2)
                                      i=l  1   n
     E  = summation over all grid points where exposure is calculated
     Pi = population associated with grid point i
     Ci = long-term average pollutant concentration at grid point i
     N  = number of grid points.
The HEM assumes that:   (1) people stay at the same location  (residence)  and
are exposed to the same concentrations of the pollutant for 70 years;  (2)
the terrain around the plant is flat; and  (3) concentrations of the  pollut-
ant are the same inside and outside the residence.
                                   B-24

-------
B.3.2   ISCLT Model
     As  noted above, the  ISCLT model was used to estimate ambient concen-
trations of organics for  estimating most exposed individual  (MEI) risk or
maximum  lifetime risk for the cancer health effects assessment and the
chronic  noncancer effects study.  The  ISCLT model  is a steady-state,
Gaussian plume, atmospheric dispersion model that  is applicable to multiple
point, area, and volume emission sources.  It is designed specifically to
estimate long-term ambient concentrations of pollutants in the vicinity of
industrial source complexes.  The model was applied to the worst-case
equipment  leak TSDF to estimate the highest concentrations of organics and
individual chemicals at various distances from the TSDF.  As described
later  in Section B.4, the highest ambient organic concentrations are used
with the equipment leak composite unit risk estimate to calculate MEI risk.
B.3.3  ISCST Model
     The ISCST model was  used to estimate ambient concentrations of indi-
vidual hazardous waste constituents for purposes of evaluating acute,
noncancer  health risks.   It was also used as a screening tool to identify
which  of the chemicals of concern in Table B-3 should be further evaluated
with the ISCLT.  The ISCST is similar  in nature to the ISCLT, except that
it is  suitable for estimating short-term ambient concentrations (e.g.,
concentrations averaged over 1 h, 3 h, 8 h, 24 h, etc.) as well as long-
term averages.  ISCST was applied to two worst-case TSDF to estimate the
highest constituent concentrations resulting from all types of sources at
the TSDF,  not just equipment leaks, for variable averaging times at the
fenceline  or beyond.
B.4  CANCER RISK ASSESSMENT
     Three pieces of information are needed to assess the cancer risks of
exposure to TSDF air emissions:  (1) an estimate of the carcinogenic
potency, or unit risk estimate, of the pollutants in TSDF air emissions;
(2) an estimate of the ambient concentration of the pollutants from a TSDF
that an individual  or group of people breathe;  and (3)  an estimate of the
number of people who are exposed to those concentrations.
     Multiplying the composite unit risk estimate by (1)  the numerical
expressions of public exposure obtained from HEM and (2)  the maximum con-
                                   B-25

-------
centration predicted by ISCLT gives two types of cancer risk measures:
(1) annual incidence,  a measure of population or aggregate risk,  and
(2) individual risk or maximum lifetime risk.  The definition and calcula-
tion of annual incidence are discussed in the next section.  Maximum life-
time risks are discussed in Section B.4.2.
B.4.1  Annual Cancer Incidence
     One expression of risk is annual cancer incidence, a measure of
aggregate risk.  Aggregate risk is the summation of all the risks to people
estimated to be living within the vicinity (usually within 50 km) of a
source.  It is calculated by multiplying the estimated concentrations of
the pollutants by the unit risk estimate by the number of people exposed to
different concentrations.  This estimate reflects the number of excess
cancers among the total population after 70 years of exposure.  For
statistical convenience, the aggregate risk is divided by 70 and expressed
as cancer incidence per year. 20
     A unit cancer potency estimate of 1.0 and a unit emission rate of
10,000 g/yr were used as input data for HEM.  Annual incidence attributed
to each TSDF, as calculated by using HEM, is proportional to the cancer
potency estimate and emissions.  Thus, another model was used to scale the
annual incidence for each TSDF by the estimated composite unit risk esti-
mate and by the estimated organic emission that were attributed to equip-
ment leaks at each TSDF:
                                           Composite    Organic
                                           unit risk    emissions
                                           estimate    for TSDF XX
Annual incidence = HEM annual incidence x  - — - x
                                              j— g -   — IQ 000 kq   '   ~

The annual incidences were then summed over all TSDF.  This scaling  and
final aggregation was performed with the Source Assessment Model  (SAM).
Appendix D contains a detailed description of the SAM and the emission and
incidence estimation methodology used in the model.
      In the estimation of annual cancer incidence for process vents, the
HEM was run using site-specific emissions, local National Weather Bureau
meteorology data, and 1980 population distributions for each of the  WSTF
identified in the 1986 National Screener Survey  (of treatment, storage,
                                   B-26

-------
disposal, and recycling facilities).  Emissions for each facility were
based on the quantity of solvent recycled at the facility; this quantity
was also taken from the 1986 National Screener Survey.  The upper and lower
bound emission factors were applied to the facility throughput to determine
the process vent emission range for the facility.  Development of emission
factors is discussed in BID Chapter 5, Sections 5.2 and 5.3.  These upper
and lower bound emission rates were used as input to the HEM to model each
of the facilities in the data base that reported recycling solvent (or
halogenated organics).  The HEM results for individual facilities were
aggregated to estimate nationwide cancer incidence resulting from WSTF
process vent emissions.  These HEM results were also used in estimating the
MIR for WSTF process vents.  The highest risk value predicted by HEM for
all of the facilities modeled was used as the MIR risk for process vents
for both the upper and lower bound emission cases.
B.4.2  Maximum Individual Risk
     Maximum individual risk (MIR) refers to the persons who may live in
the area of highest ambient air concentrations of the pollutant(s) as
determined by the detailed facility modeling.  The MIR reflects the
probability of developing cancer as a result of continuous exposure to the
estimated maximum ambient air concentration for 70 years.  It is based only
on the maximum exposure estimated by the procedure used,21 and it does not
incorporate uncertainties in the exposure estimate or the unit risk
estimate.
     MIR is calculated by multiplying the highest ambient air concentration
by the composite unit risk estimate.  The product is the probability of
developing cancer for those individuals assumed to be exposed to the
highest concentration for their lifetimes.  Thus,
                                               Highest   ]
                                         x   ambient air    .         (B-4)
                                            .concentration]
MTD
 1
               f  Composite unit risk
                              f
               lestimate at 1 l/jg/m
B.4.3   Methodology for Specifying a "Worst-Case Facility" for TSDF
        Equipment Leak Emissions and Risk
     B.4.3.1   Introduction.  This section of Appendix B describes the
methodology for developing a realistic, "worst-case" hazardous waste TSDF
                                    B-27

-------
to estimate the maximum individual  risk to a hypothetical  person living
nearby.  The information available and the assumptions and their bases for
characterizing such a facility are presented.  This facility is intended to
represent a TSDF that could realistically and reasonably exist and that
could result in "worst-case" exposure to emissions from hazardous waste
management process equipment leaks.  It is not intended to be
representative of all facilities in the hazardous waste management industry
nor to represent a worst case that could not reasonably be expected to
exist.  The approach for approximating the hazardous waste TSDF industry
and identifying the facility exhibiting the highest maximum risk, relying
to the extent possible on the existing and available data, is presented
below.
     B.4.3.2  Methodology.  The EPA is conducting a multiyear project to
collect information on the Nation's generation of hazardous waste and the
capacity available to treat, store, dispose, and recycle that waste.  The
initial phase of the project was the 1986 National Screening Survey that
identified and collected summary information from all hazardous waste
treatment, storage, disposal, and recycling facilities  (TSDR) in the United
States.  These data served as the basis for the solvent recycling industry
profile and the waste quantity used in the SAM.  Phase  II involved
verification and updating the data provided by the active TSDR.  In Phase
III, all active TSDR facilities identified as having treatment, disposal,
or recycling technologies were sent a detailed package  of questionnaires,
appropriate to the processes they operate, for the 1987 National TSDR
Survey  (hereafter referred to as the TSDR Survey).  The completed
questionnaires are reviewed for technical accuracy, and after verification
the information collected is entered into a complex data  base.  The TSDR
Survey  is described  in detail under Phase III of Attachment  1,  National
Surveys of Facilities that Generate, Treat, Store, Dispose,  or  Recycle
Hazardous Waste.  This information is the most detailed and  up-to-date
available.
     A  portion of the computerized TSDR Survey data base  was  available  for
limited use and included  information and data from 1,402  facilities out of
a total of 2,600. The facilities in the data base  include all TSDR  facili-
ties that are commercial  operations and all  TSDR  facilities  that have  some
                                    B-28

-------
 form  of waste management  process  that  involves  "land  disposal."   Land  dis-
 posal  processes  include surface impoundments, underground  injection,  land
 treatment, wastepiles, and  landfills.
      The  TSDR Survey questionnaire  responses contain  up-to-date  nationwide
 information  regarding the hazardous waste management  technologies  each
 facility  has onsite.  For each facility, detailed  information  is  available
 in  the data  base  including  facility area, numbers  of  hazardous waste
 management units  by process type  (i.e., number  of  surface  impoundments,
 incinerators, and  recycling units), annual throughput  by process  unit,  and
 types  of  waste (i.e., RCRA waste  codes) managed by each unit at  the
 facility.  The availability of this information in computerized  format made
 it  possible  to screen the data base to  identify the facilities that are
 expected  to  represent the population of worst-case facilities with regard
 to  equipment leak  emissions.
      The  1,402 facilities entered in the system represented over  50 percent
 of  the total TSDF  population.  Because  all commercial  TSDF were contained
 in  the data base,  a representative sample of large TSDF (i.e., those treat-
 ing large waste volumes in a variety of waste management units and as  a
 result those facilities expected  to have the largest equipment counts,  the
 major  factor in determining equipment leak emissions), were included in the
 data base at the time.  With regard to  solvent  recycling operations, the
 solvent recycling  industry profile developed from  results of the  1986
 Screening Survey indicates that about 30 percent of the facilities report-
 ing solvent recovery operations are commercial facilities, and commercial
 facilities report much higher throughputs than noncommercial facilities.
 For example,  facilities reporting recycling of only halogenated organics
 (NOTE: Of all recycled solvents with EPA unit risk factors, nearly all  are
 halogenated organics) in  the solvent recovery units, noncommercial facili-
 ties averaged about 20,000 gallons throughput in 1985; commercial opera-
 tions  averaged more than 610,000 gallons throughput in 1985.  Therefore, it
 is expected that  the TSDR Survey data base used in the screening was repre-
 sentative of the  worst-case solvent recycling population.
     The following paragraphs describe the step-wise procedure to specify a
worst-case facility for TSDF equipment leaks based on use of the informa-
tion and data contained in the TSDR Survey.   A worst-case facility was  used
                                   B-29

-------
as input to dispersion modeling to estimate maximum annual  average con-
centration.  The maximum annual average concentration was used to estimate
maximum lifetime risk to nearby residents associated with equipment leak
fugitive emissions.
     B.4.3.2.1  Data base screened to identify facilities expected to have
the highest maximum risk.  Equipment counts are the most important deter-
mining factor for emission estimates from facility equipment leaks, not the
amount of waste handled.  Equipment counts do not double or triple as
facility throughput is increased accordingly.  The size of the individual
equipment is also not a consideration in estimating equipment leak emis-
sions.  In addition, equipment counts (i.e., the number of pumps, valves,
sampling connections, and pressure relief devices) tend to be similar for
similar types of process units.  These facts allow a screening approach to
be used with a reasonable degree of accuracy in estimating TSDF equipment
counts and equipment leak emissions.
     For screening, a weighting factor that reflects relative equipment
counts  (i.e., the number of pumps and valves associated with each process
type) was assigned to each type of hazardous waste management process.  For
example, a waste solvent recycling unit or a hazardous waste incinerator
will have more associated pumps and valves than a storage tank or surface
impoundment.  Equipment counts and weighting factors based on expected
equipment counts have previously been developed for a wide variety of haz-
ardous waste management process types for use  in SAM, and these  factors
were used to assign weights to the hazardous waste management units
identified  in the TSDR  Survey  data base.  Weighting factors, rather than
equipment counts, were  used in the screening analysis for accounting
convenience only.   The  equipment count weighting factors are presented  in
Tables  B-4  and B-5.
     A  computerized  screening  of the TSDR data base was  performed  to
identify the number of  units used  at each facility  for  each type of
hazardous  waste management process.  The  number of  units were then
multiplied  by the  appropriate  equipment  count  weighting  factor for that
management  process.   Summing these products  for all  hazardous waste
management  process  types  at a  facility yielded a  total  facility  equipment
                                    B-30

-------
                     TABLE B-4.  MODEL UNIT PARAMETERS
No. of components
per model unit
Equipment type
(weighting factor)
Pumps
Valves
Sampling connections
Open-ended lines
Pressure relief valves
A
(1.0)
5
165
9
44
3
B
(0.6)
3
99
5
26
2
(0.2)
1
33
2
9
1
Emission
factors,
Mq/yr/component
Liquid liquid
0.43
0.062
0.13
0.015
0.91
Heavy liquid
0.19
0.002
0.13
0.015
0.91
      TABLE B-5.  MODEL UNITS ASSIGNED TO WASTE MANAGEMENT PROCESSES
Management process
(weighting factor)
Uncontrolled emissions, Mg/yr
Model Unit A (1.0)

    Incinerators
    Solvent recovery units

Model Unit B (0.6)

    Tank systems
    Underground injection units

Model Unit C (0.2)

    Surface impoundments
    Fuel blending tanks
    Solidification/stabilization
    Fuel reuse systems
            16.90
            10.16
            3.39
                                   B-31

-------
count score that reflected relative equipment counts expected at each
facility.  The facilities were then ranked by total  facility equipment
count score to identify the facility with the highest potential  for total
equipment leak emissions.
     Emissions are not the only source-based parameter that affects ambient
air concentrations and risk; the area of the facility is also of signifi-
cance.  For a constant emission rate, i.e.,  constant mass per unit time,
increased area size yields lower maximum concentrations.  For example, a
10-acre source with emissions of 1 Mg/yr will yield greater ambient concen-
trations than a 20-acre source with the same emissions.
     The TSDR Survey data base contains information on the area of the
TSDF; therefore, these data were also used in the screening process.  The
total facility equipment count score was divided by the facility area to
get a screening ratio that reflected equipment counts per unit area.  This
screening ratio served as an index to identify the facility, with the
highest potential for equipment leak emissions in the smallest facility
area.  Sections B.4.3.2.5 and B.4.3.2.8 describe how this screening ratio
and total facility equipment leak emissions were used to identify a worst-
case facility when emission dispersion is considered.
     B.4.3.2.2  Facilities selected for detailed analysis.  About 15  to  20
facilities to be analyzed in detail were chosen based on the combination of
total equipment counts (representing total emissions) and the ratio of
total equipment count score to facility area  (representing emission concen-
tration).  Facilities with both high equipment counts and high equipment
count to area ratios served as the bases for characterizing the worst-case
facility.
     Two prioritized listings of facilities  (one ranked on the basis  of
high equipment count and the second  ranked on the basis of high equipment
count to area ratio) were generated  from the screening of the TSDR data
base.  Tables B-6 and B-7 present the results of the screening for the  top
50 facilities in each of the two rankings.   From each  listing, the  10
highest-ranked facilities were selected for  detailed analysis.  Those
facilities appearing near the top of both lists  (thus  representing both
high emissions and small emitting area) were  singled out.  The facilities
                                   B-32

-------
                    TABLE B-6.  1987 TSOR SURVEY SCREENING RESULTS RANKED
                   BY TOTAL SCORE (i.e., EQUIPMENT LEAK EMISSION POTENTIAL)
Observation
1
2
3
4
5
8
7
8
9
10
11
12
13
14
IS
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Faci 1 ity area*
(acres)
0
384
750
2,000
832
5,122
30
803
1,600
315
1,000
688
120
6
1
470
56
3,000
12
17
3,992
35
1,100
2,154
3
6
18
728
26
3
43
1,273
14
1,262
6
163
86
1,649
780
550
160
6
1
13
4
22
10
15
1,150
2
Tot* 1 score^
109.2
65.8
59.6
59.0
53.0
53.0
52.8
51.8
44.4
43.2
42.2
40.8
40.4
39.6
36.8
36.6
36.0
34.6
32.2
32.0
31.2
30.0
30.0
29.8
29.8
26.8
26.8
26.4
26.0
26.0
25.8
25.8
25.8
25.0
24.4
24.4
24.4
24.2
23.6
23.4
23.2
23.0
22.4
21.6
21.2
21.2
21.0
20.8
20.8
20.6
Ratio of total score
to faci 1 ity area
0.0000
0.1714
0.0795
0.0295
0.0637
0.0103
1 . 7600
0.0645
0.0278
0.1371
0.0422
0.0593
0.3367
6.6000
36 . 8000
0.0779
0.6429
0.0115
2.6833
1.8824
0.0078
0.8571
0.0273
0.0138
9.9333
4.4667
1.4889
0.0363
1.0000
3.6667
9 . 6000
0.0203
1.8429
0.0198
4.0667
0.1497
0.2837
0.0147
0.0303
0.0425
0.1450
3.8333
22 . 4000
1.6615
5 . 3000
0 . 9636
2.1000
1.3867
0.0181
10 . 3000
»A
iro value indicates that no value was reported in questionnaire response.
  score represents relative equipment leak emission potential.
                                          B-33

-------
                  TABLE B-7.  1987 TSOR SURVEY SCREENING RESULTS RANKED BY
             THE RATIO OF TOTAL SCORE TO FACILITY (i.e., EMISSION CONCENTRATION)
Observation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Faci 1 ity area
(acres)
1
1
1
1
1
2
3
1
2
2
3
1
1
1
1
1
6
1
3
1
1
2
3
1
1
1
4
2
1
2
4
3
2
2
3
4
6
1
2
1
1
2
1
6
1
6
1
1
1
1
Tota 1 score*
36.8
22.4
18.4
11.0
10.8
20.6
29.8
9.6
18.8
18.0
26.0
8.4
8.4
8.2
7.4
7.2
39.6
6.6
19.2
6.4
6.2
12.2
18.2
5.8
5.6
5.6
21.2
10.4
5.2
10.2
20.2
15.0
9.4
9.4
13.6
18.0
26.8
4.4
8.6
4.2
4.2
8.4
4.2
24.4
4.0
23.0
3.8
3.6
3.6
3.6
Ratio of total score
to faci 1 ity area
36 . 8000
22.4000
18 . 4000
11.0000
10.8000
10.3000
9.9333
9.6000
9.4000
9.0000
8.6667
8.4000
8.4000
8.2000
7.4000
7.2000
6.6000
6.6000
6.4000
6.4000
6.2000
6.1000
6.0667
5.8000
5.6000
5.6000
5.3000
5.2000
5.2000
5.1000
5.0500
5.0000
4.7000
4.7000
4.5333
4.5000
4.4667
4.4000
4.3000
4.2000
4.2000
4.2000
4.2000
4.0667
4.0000
3.8333
3 . 8000
3.6000
3.6000
3.6000
•Total score represents relative equipment leak emission potential
                                            B-34

-------
 chosen  for  the  sample  represent the facilities with the greatest potential
 for posing  the  maximum lifetime risk to a hypothetical person  living
 nearby.   The  detailed  analysis of the TSDR  Survey data for these facilities
 is described  in Sections B.4.3.2.3, B.4.3.2.4, B.4.3.2.6, and  B.4.3.2.7.
     B.4.3.2.3  Facility Rankings Verified  and/or Adjusted.  For each
 facility  in the sample selected (see Section B.4.3.2.2), detailed
 information including  facility area, numbers of hazardous waste management
 units by  process type,  type of waste (RCRA  codes) managed by unit, and
 number  of units planned for operation by January 1992 was available in hard
 copy from the TSDR Survey responses.  As a  quality assurance check, the
 information on  facility area and numbers of hazardous waste management
 units was used  to recalculate total facility equipment count scores and
 ratios  of equipment counts to facility area.  Because of the potential
 double- counting of some tanks (e.g., the same tank could have been
 reported  in the fuel blending and tank systems questionnaires), when the
 manual  check  of equipment count scores and  ratios was performed,
 corrections were made  for tanks and other processes that had been double-
 counted or were otherwise in error.
     After the  quality assurance checks were completed, the prioritized
 lists were adjusted.   The information on number of units planned for opera-
 tion by January 1992 was also used in a later step (see Section B.4.3.2.6)
 in specifying hazardous waste management units for the worst-case facility.
 The type of waste managed by unit was used  (see Section B.4.3.2.4)  to
 screen equipment handling wastes with less than 10-percent organics from
 the analysis.
     B.4.3.2.4  Screening of equipment not subject to control requirements
 of accelerated rule.  The equipment leak provisions of the rules apply to
 equipment that either contains or contacts hazardous waste with organic
 concentrations greater than 10-percent by weight.   Therefore, equipment
 associated with hazardous waste management units that could handle hazard-
ous wastes with greater than 10-percent  organics were identified.   The TSDR
Survey  data base contains information on the types of hazardous wastes
 (RCRA  codes) managed by each hazardous waste management unit at each TSDR
 in the  sample.  This information  along with  characterizations of each RCRA
                                   B-35

-------
waste stream was used to screen equipment not handling wastes with organic
concentrations of 10 percent or more from the analysis.
     The TSDF Waste Characterization Data Base (WCDB)  was used to charac-
terize each RCRA waste stream in terms of average organic content. The WCDB
was developed to support the development of the comprehensive air emission
regulations for hazardous waste TSDF.  Wastes listed in this data base were
characterized, primarily using the following five existing data bases:
     •    The Westat Survey
     •    The Industry Studies Data Base
     •    A data base of 40 CFR 261.32 hazardous wastes from specific
          sources  (i.e., waste codes beginning with the letter K)
     •    The WET Model Hazardous Waste Data Base
     •    A data base created by the Illinois EPA.
Each hazardous waste stream was characterized in terms of constituents and
their percent composition.
     The characterizations of hazardous waste streams in the WCDB represent
national average compositions and cannot be used to accurately characterize
specific waste streams or their organic content variability at a facility.
That is, although a stream may average 5-percent organic content, it may at
times exceed 10-percent.  Therefore, it was assumed that wastes that
average less than 1-percent total organics would be wastewaters and would
never contain greater than 10-percent total organics.  This is consistent
with the Agency's definition of solvent-containing wastewaters classified
as F001-F005.
     Using the waste characterizations from the WCDB, hazardous waste
management units that do not manage wastes with greater than 1-percent
organics were eliminated from the analysis.  The analysis included all
streams with greater than 1-percent organic content, even those that would
never exceed 10-percent, and thus overestimated emissions.  Facility equip-
ment count scores and ratios of equipment count scores to facility area
were recalculated, and a new prioritized list of facilities was generated.
The revised prioritized list of facilities was used (see Section B.4.3.2.6)
                                   B-36

-------
to specify types and number of hazardous waste management units and
facility area  for the worst-case facility.
     B.4.3.2.5  Facility selected to serve as basis for the worst-case
facility.  From the sample of facilities developed  (see Section B.4.3.2.2),
a facility was chosen to serve as the basis for the worst-case facility.
The choice was mad.e on the bases of the ratio of total equipment count
score to facility area and of the total equipment count score.  Facilities
with the highest screening ratio of total equipment count score to facility
area were chosen as candidate "worst-case facilities."  The total equipment
count score for these facilities was compared to the total equipment count
scores of the  facilities with the highest scores to ensure that a facility
with high potential equipment leak emissions was chosen.  A facility that
had both a high screening ratio and was also high on the list of facilities
prioritized by total equipment count score was chosen as the basis for the
worst-case facility.
     The facility selected for the worst-case facility analysis ranked 15
out of the more than 1,400 facilities in the data base in terms of total
equipment scores (i.e., a direct index of potential equipment leak
emissions).  However, after reviewing the detailed questionnaire responses
of the top scoring facilities, it was found that most did not handle wastes
with organic concentrations of 10 percent or greater.  For example, five
out the top six facilities were found to be processing wastewaters that are
not covered by the standards.  Adjusting the facility equipment scores
based on waste organic content puts the selected facility among the 10
highest scores.  In terms of the ratio of equipment score to facility area,
the selected facility ranks at the top of the list of the 1,400 facilities
screened.  The ratio of equipment score to facility area was used as an
index of the potential  for high ambient air concentrations.
     B.4.3.2.6  Hazardous waste management units and facility area
specified.   The types and numbers of the following waste management units
located at  the chosen facility (see Section B.4.3.2.5) were assigned to the
worst-case  facility:
     •     Incinerators
     •     Units burning hazardous waste as fuel
     •     Fuel  blending units
                                   B-37

-------
     •    Solidification/stabilization units
     •    Solvent recovery units
     •    Surface impoundments
     •    Underground injection wells
     •    Storage and treatment tanks.
In addition, transport and handling operations, such as drum and/or termi-
nal loading were added because of the type of waste management processes at
the facility.  For example, a facility with solvent recycling processes
would receive some wastes in drums and ship the recovered solvent from a
terminal.  Drum and terminal loading have associated pumps and valves but
are not included in the TSDR Survey because they are not considered hazard-
ous waste management units.  However, it is likely that they handle hazard-
ous waste with greater than 10-percent organics.  Also, according to the
TSDR Survey responses, such a facility plans to construct or begin opera-
tion of other waste management units before January 1992; these units were
added to the worst-case facility characterization.
     The actual area of the facility selected was assigned to the worst-
case facility.  Therefore, the types and numbers of hazardous waste manage-
ment units and the facility area were based on an actual facility indicated
to have a high ratio of equipment handling organic wastes to the facility
area.
     B.4.3.2.7  Equipment counts assigned.  Assigning equipment counts
(i.e., numbers of pumps, valves) followed the same procedure used in
assigning the equipment weighting factors; however, in this case, the
equipment counts associated with the hazardous waste management units were
used rather than the weighting factors.
     After the worst-case facility was specified in terms of area and
numbers and types of hazardous waste management units, an "equipment" model
unit was assigned to each hazardous waste management unit at the facility.
Various model units were developed to represent the range of waste
management processes.  The model units are simplified representations of
the equipment component mix expected to be associated with a particular
hazardous waste management process.  The model unit provides an estimate of
the number of pumps, valves, open-ended lines, pressure relief valves, and
                                   B-38

-------
  sampling connections  that  are used  in the  hazardous waste  management  proc-
  ess.  Although these  model units are not exact  for each  type  of  process,
  they provide a reasonable  approximation of what can be expected.   Precise
  equipment counts for  each  unit at each facility are not  available.
      Table B-4 presents a  list of equipment counts and emission  factors for
  each equipment source within the model units.  The equipment  counts are
  calculated to be about 60  percent and 20 percent of Model  Unit A for  Model
  Unit B and Model Unit C, respectively.  Model units were assigned  to  each
 waste management unit based on a review of available information on associ-
 ated equipment counts.  Table B-5 presents the model units assigned to the
 hazardous waste management processes that are contained in the TSDR Survey
 data base.
      B.4.3.2.8  Dispersion modeling performed.  The synthetic organic
 chemical  manufacturing industry (SOCMI)  equipment leak emission factors
 (EPA-450/3-82-010)  were used  to estimate fugitive organic emissions for the
 worst-case  facility.   The SOCMI  emission factors were  used because they are
 the best  available  for mixtures  of  chemicals  and because TSDF have been
 found  to  handle  the same chemicals  as  those found  in the SOCMI.   Uncon-
 trolled and controlled emissions  and the appropriate area source  size  and
 receptor  distance served as input to a  long-term modeling approach  to
 estimate  the maximum  annual average  concentration  for  the worst-case
 facility.
     The model selected  for this analysis was  the  Industrial  Source
 Complex-Long-Term model  (ISCLT).  The ISCLT is a steady-state, Gaussian
 plume, atmospheric  dispersion model  that is applicable to multipoint,  area,
 and  volume emission sources.  It is  designed specifically to estimate  long-
 term ambient concentrations resulting from  air emissions  from  this  source
 type.  The ISCLT is recognized by the EPA Guideline on Air  Quality  Models
 as a preferred model for dealing with complicated sources  (e.g.,  facilities
with area sources)   in either rural or urban areas located in flat or
rolling terrain.   The current UNAMAP version of ISCLT was used in all
modeling applications.  Several  years of meteorological data from various
sites around the Nation were also used  in the dispersion modeling exercise
to identify  the highest ambient  air concentrations for the worst-case
facility.
                                   B-39

-------
     The approach to characterize the facility layout was to spread the
equipment leak emissions over the facility area in a uniform manner.  This
approach was expected to best reflect the actual  facility layout.  Because
the worst-case facility has a large number of hazardous waste management
units (i.e., equipment count) concentrated in a small area, equipment will
by necessity be distributed throughout the facility area and not concen-
trated on one side of the facility. The dispersion modeling produced
estimates of the highest ambient air concentrations for the worst-case
facility for the eight cities used to characterize meteorological
variations.
     B.4.3.2.9  Distance specified to nearest resident.  Because
information on the distance from the facility boundary to  the nearest
residence was available for the facility serving  as the basis for  the
worst-case  facility, this distance was used  in the worst-case facility
analysis.   The distance from the facility boundary to the  nearest  residence
was  estimated to be  at  least 250 ft  for this  particular facility.
     In  order to determine how this  distance  compares to distances for  the
industry as a whole, EPA utilized  the preliminary results  of the 1987
Generator Survey.  The  distance  from the  facility boundary to the  nearest
resident is not  reported in  the  TSDR Survey.  A  random sample of 50
facilities  was drawn from  the respondents to the Generator Survey.  The
minimum distance to  the nearest  resident  reported for  facilities in this
sample  was  100 ft;  the  maximum  reported  distance was  10,000 ft.  The mean
 (arithmetic average) for the sample  was  about 1,500  ft;  the median distance
was   1,000  ft  (i.e., half  of the facilities  reported a distance to the
nearest resident of  more than 1,000  ft).
      B.4.3.3   Conclusions  on Selection  of the Worst-Case Facility.  The
 approach to characterizing a realistic  worst-case facility described above
 is considered to best  reflect actual worst-case conditions that could be
 expected at TSDF.   The National  TSDR Survey  data are the most up-to-date,
 detailed,  and comprehensive information available on hazardous  waste
 management process units.   Characterizing the worst-case facility  based on
 actual  highest unit/equipment counts and facility areas ensures that the
 representations  are both reasonable and that the worst-case is  based on
                                    B-40

-------
 existing data.  In addition, it is taken a step beyond "actual" by  locating
 the TSDF near the city with the worst meteorology examined and counting
 equipment that may not actually be covered (handling wastes containing
 between 1- and 10-percent organics).  Finally, it should be noted that the
 vast number of individual URF that form the composite URF represent the 95-
 percent upper confidence bound, thus adding further conservatism to the
 risk estimates.
 B-4-4  Estimates of TSDF Equipment Leak Emissions and Maximum Risk
      The worst-case facility developed as the basis of the TSDF equipment
 leak maximum risk analysis has the following waste management units:  drum
 handling operations,  storage tanks,  solvent recycling units,  waste fixa-
 tion/solidification,  fuel blending,  terminal  loading/unloading,  incinera-
 tion,  and boilers/process heaters.
      B.4.4.1   Drum Handling.   The  worst-case facility is  expected  to oper-
 ate an enclosed drum  handling  unit to unload and  load organic hazardous
 wastes.  Drum unloading  is basically  an  assembly  line operation  of pumping
 the wastes  to be processed from the  drums  or pumping the  recovered solvent
 into the drums.   Equipment associated with  this unit have been character-
 ized by a medium equipment-count model unit.   Equipment  leak  emissions  from
 this unit are estimated  at 10.2 Mg/yr, uncontrolled.
      B-4-4-2   Storage  and Process  Tanks.  The  worst-case  facility  has 28
 storage and process tanks;  tank sizes range  from 400  gallons  to  10,000
 gallons.  A small  equipment count  model  unit has been  assigned to  each
 storage/process  tank.  Total uncontrolled organic emissions from tank-
 associated equipment leaks  are  estimated at 94.9 Mg/yr, uncontrolled.
     B-4-4-3   Recycling Units.   The worst-case facility recycles organic
 solvents that  are processed in  three  fractionation units, four batch
 distillation units, and two thin-film evaporation units.  Based on reported
 unit capacities, the recycling  units' equipment counts were characterized
 by two  large recycling model units, two medium recycling model units, and
 five small recycling model units.  Total  uncontrolled equipment leak
emissions from the recycling units  are estimated at 186 Mg/yr,
uncontrolled.
                                   B-41

-------
     B.4.4.4  Fixation/Solidification.   The worst-case facility treats a
small percent of incoming waste material in a fixation process.  Treatment
consists of solidification of organic waste with inert material, after
which the drum waste is sent to a commercial hazardous waste landfill.  A
medium equipment-count model unit has been assigned to this operation, and
total equipment leak fugitive emissions are estimated at 10.2 Mg/yr, uncon-
trolled.
     B.4.4.5  Fuel Blending Tanks.  The worst-case facility has five fuel
blending units operating at the facility.  Organic wastes not suitable for
recycling are blended to meet specifications established by the end user,
usually a cement kiln, and shipped in bulk.  A small equipment-count model
unit has been assigned to each fuel loading tank operation, and total
equipment leak emissions are estimated  at  17.0 Mg/yr, uncontrolled.
     B.4.4.6  Transfer, Handling, and Loading.  Product shipped from  the
worst-case  facility  is sent either in bulk utilizing tank trucks  or in
drums.  Approximately 80 to 90 percent  of  the  recovered organics  sent to
customers are expected to be shipped in bulk form, with the  remainder sent
in  drums.   Waste  organics blended for reuse as  fuel  are also  shipped  in
bulk.   A medium equipment-count  model unit was  selected to  represent  this
operation;  total  uncontrolled, equipment leak  emissions from the  unit are
estimated  at  10.2 Mg/yr.
      B.4.4.7   Incineration.   The designation of two  incineration  units  at
the facility  is  based  on  reported plans to install  a liquid-injection
 incinerator and  a fluidized-bed  incinerator at the site.   A large
equipment-count  model  unit  is  used  to  represent equipment associated  with
each incineration operation;  total  equipment leak emissions are estimated
 at 33.8 Mg/yr,  uncontrolled.
      B.4.4.8  Boilers/Process Heaters.   The worst-case facility also
 includes the use of a boiler process heater to dispose of organic wastes.
 A medium equipment-count model unit has been used to characterize equipment
 associated with this operation.  Total  equipment leak emissions are esti-
 mated at 10.2 Mg/yr, uncontrolled.
      B.4.4.9  Total-Worst Case Facility Emissions.  Based on the estimated
 equipment  associated with the worst-case  facility operations discussed
                                    B-42

-------
above, the total  uncontrolled equipment leak emissions for this TSDF/WSTF
are estimated at  373 Mg/yr.  Table B-8 summarizes these emissions by source
category.
     B.4.4.10  Maximum Risk.  Table B-9 presents the maximum risk resulting
from uncontrolled and controlled equipment-leak emissions from the worst-
case facility for the urban situation at varying distances from the edge of
the area source.   The emission rate of 373 Mg/yr was used to calculate
maximum risk based on the ISCLT dispersion estimates for emissions from an
area source conducted by EPA/ESD/SRAB.  Eight cities around the Nation were
modeled using a minimum of 5 years of meteorological data to identify
highest ambient air concentrations for the worst-case situations.  Table
B-10 presents the dispersion modeling results.  It is of interest to note
that the highest  ambient concentrations were found to occur in the city
that is the location of the facility selected to represent the worst case.
     B.4.4.11  Equipment Leak MIR Results.  Based on results of the analy-
ses described above, the revised MIR estimates for TSDF equipment leak
emissions are as  follows:

                      Control Case            ME I
                      Uncontrolled         5 x 10'3
                      Option 1             1 x 10'3
                      Option 2             1 x 10'3
The MIR risks were calculated using the highest annual average concentra-
tion at the receptor distance of 250 ft and using the TSDF equipment leak
source-specific,  emission-weighted unit risk factor of 4.5 x 10'6
B.5  NONCANCER HEALTH EFFECTS
B.5.1  Chronic Exposures
     The assessment of noncancer health effects associated with chronic
exposures to TSDF chemicals, of concern is based on a comparison of the
chemical-specific health benchmark levels (as discussed in Section B.2.1)
to estimated ambient concentrations at various receptor locations around a
facility.  Inhalation exposure limits are compared to the highest annual
                                    B-43

-------
             TABLE B-8.  WORST-CASE FACILITY EQUIPMENT LEAK
                      EMISSIONS BY SOURCE CATEGORY*
Source category
Drum handling
Storage and process tanks
Solvent recycling units
Fixation/solidification
Fuel blending
Transfer, handling, and
loading
Incineration
Boilers/process heaters
Total
Uncontrolled
emissions,
Mg/yr
10.2
94.9
186.0
10.2
17.0
10.2
33.8
10.2
373
Option 1
emissions,1*
Mg/yr
2.78
26.0
51.0
2.78
4.64
2.78
9.26
2.78
102
Option 2
emissions,0
Mg/yr
2.36
22.0
43.2
2.36
3.93
2.36
7.84
2.36
86.4
alncludes equipment leaks from pump seals,  valves,  open-ended lines,  sam-
 ling connections,  and pressure relief valves.
^Option 1 includes  leak detection and repair requirements for pumps and
 valves, caps for open-ended lines, closed-purge sampling,  and rupture
 disks for pressure-relief devices.
CQption 2 includes  leak detection and repair for valves, dual mechanical
 seals and associated equipment for pumps,  caps for open-ended lines,
 closed-purge sampling, and rupture disks for pressure-relief devices.
                                    B-44

-------
            TABLE B-9.   MAXIMUM RISK FROM EQUIPMENT LEAKS FOR A
                      WORST-CASE FACILITY (TSDF/WSTF)3
Distance,
ft
50
100
250b
500
1,000
2,000
4,000

Uncontrolled
2 x ID'2
1 x lO-2
5 x 10-3
2 x 10-3
7 x 1C'4
2 x 1C'4
7 x 10-6
Maximum risk
Option 1
5 x 10-3
3 x 10-3
1 x ID'3
5 x lO'4
2 x ID'4
6 x lO-5
2 x 10-5

Option 2
4 x 10-3
3 x 10-3
1 x ID'3
4 x lO'4
2 x lO'4
5 x 10-5
2 x 10-5
aBased on the highest annual  average ambient concentration at the receptor
 distance for the eight cities modeled by the ISCLT, a worst-case emis-
 sion rate of 373 Mg/yr (11.8 g/s) uncontrolled, and an URF of 4.5 x lO'6
     minimum distance to the nearest residence for the facility serving as
 the basis for the worst-case facility analysis.
                                    B-45

-------
      TABLE B-10.   SUMMARY OF AMBIENT  CONCENTRATION  ESTIMATES  FROM A
         PROTOTYPE  TSDF  IN AN URBAN  LOCATION  AT VARIOUS  DISTANCES3
Distance, ft .
City
Houston
Baton Rouge
Cleveland
Charleston
Chester
Newark
Wyandotte
Santa Monica
50
235.2
291.2
267.1
218.7
261.4
200.2
191.1
329.7
100
138.1
168.1
162.3
132.7
155.7
119.4
114.6
217.8b
250
53.9
61.9
62.6
51.5
58.5
44.3
42.9
91.7
500
21.8
24.4
23.7
20.7
22.7
16.8
16.6
35.9
1,000
7.64
8.61
8.15
7.24
7.78
5.67
5.68
12.2
2,000
2.48
2.81
2.60
2.35
2.48
1.78
1.80
3.92
4,000
0.79
0.90
0.82
0.75
0.78
0.55
0.56
1.22
aAll concentrations are in micrograms per cubic meter.

bThe highest annual average concentration at the minimum receptor distance
 used in the calculation of MIR risk.
                                   B-46

-------
average ambient concentration for each chemical at the selected facilities.
These annual concentrations represent an estimation of the highest average
daily ambient concentration experienced over a year.  Ambient concentra-
tions that are less than the RfD are not likely to be associated with
health risks.  The probability that adverse effects may be observed in a
human population increases as the frequency of exposures exceeding the RfD
increases and as the size of the excess increases.
     Because Agency-verified RfD are not available, an interim screening
approach was used.  The likelihood of adverse noncancer health effects was
determined by comparing modeled ambient concentrations of individual
constituents to the available health data.  These health data are obtained
from various sources, including EPA reports and documents, data used to
support occupational exposure recommendations and standards (e.g., American
Conference of Governmental Industrial Hygienists, Documentation of the
Threshold Limit Values), and other published information.  An assessment of
the potential for adverse noncancer health effects was made case-by-case,
considering:  (1) the magnitude of the differences between the exposure
concentration and the lowest-observed-adverse-effect level or the no-
observed-adverse-effect level, and (2) the quality of the health effects
data base.  In general, the likelihood of noncancer health effects was
considered to be low if modeled concentrations are several orders of
magnitude below the health effect levels of concern.  The probability that
such effects will occur increases with increasing exposure concentrations.
This screening effort was used only to give a preliminary indication of the
potential for noncancer health effects.
B.5.2  Acute Exposures
     An assessment of the potential for noncancer health effects associated
with short-term (acute) exposure to TSDF chemicals of concern at selected
facilities was conducted as a screening effort to provide additional
qualitative support to the overall  noncancer health effects analysis.   In
addition to the lack of short-term inhalation health benchmark levels  at
this time, acute inhalation data are limited for many of the TSDF chemicals
of concern.  The assessment was conducted by comparing maximum modeled
ambient concentrations for averaging times of 15 min,  1 h, 8 h,  and 24 h to
                                   B-47

-------
available short-term health data matched to the appropriate averaging time.
A determination of the risk of adverse health effects associated with esti-
mated short-term exposures was based on a consideration of the quality of
the available health data and the proximity of the exposure concentration
to the health effect level.
B.5.3  Results of Noncancer Health Effects Analysis
     A screening analysis of the potential adverse noncancer health effects
associated with acute and chronic exposure to individual waste constituents
emitted from TSDF was conducted by EPA.  This analysis was based on a com-
parison of relevant health data to the highest short-term or long-term
modeled ambient concentrations for chemicals at each of the two selected
TSDF.  Results of this analysis suggest that adverse noncancer health
effects are unlikely to be associated with acute or chronic inhalation
exposure to TSDF organic emissions on a nationwide basis.  It should be
noted that the health .data base for many chemicals was limited, particu-
larly for short-term exposures.  The conclusions reached in this prelimi-
nary analysis should be considered in the context of the limitations of the
health data; the uncertainties associated with the characterization of
wastes at the facilities;  and the assumptions used in estimating emissions,
ambient concentrations, and the potential for human exposure.  Additional
evaluation of noncancer health effects will  be considered as part  of the
third phase of the  TSDF regulatory program.
B.6  ANALYTICAL  UNCERTAINTIES APPLICABLE  TO  CALCULATIONS OF PUBLIC HEALTH
     RISKS  IN THIS  APPENDIX
B.6.1  Unit Risk  Estimate
     The  procedure  generally  used to  develop unit  risk  estimates is  fully
described in  Reference  1,  using  nickel  as  an example.   The  low-dose
extrapolation model used  and  its  application to  epidemiological  and  animal
data have been the  subjects  of  substantial  comment by  health  scientists.
The  uncertainties  are  too complex to  be summarized  in  this  appendix.
 Readers  who wish to go  beyond the  information presented in  the reference
 should  see the  following  Federal  Register notices:   (1)  EPA's  "Guidelines
                                     B-48

-------
for Carcinogenic Risk Assessment," 51 FR 33972 (September 24, 1986), and
(2) EPA's "Chemical Carcinogens; A Review of the Science and Its Associated
Principles," 50 FR 10372 (March 14, 1985), February 1985.
     Significant uncertainties associated with the cancer unit risk esti-
mates include:  (1) selection of dose-response model,  (2) selection of
study used to estimate the unit risk estimate, and (3) presence or absence
of a threshold.  Uncertainties related to the composite risk estimate
include the assumption of additivity of carcinogenic risk.  According to
the EPA "Guidelines for the Health Risk Assessment of Mixtures," a number
of factors such as data on similar mixtures and the interactions among
chemicals must be considered before additivity can be assumed.22  Because
of the sheer number of chemicals emitted from TSDF and the lack of specific
information on particular compounds, EPA assumed additivity.
B.6.2  Public Exposure
     B.6.2.1  General.  The basic assumptions implicit in the methodology
are that all exposure occurs at people's residences,  that people stay at
the same location for 70 years, that the ambient air concentrations and the
emissions that cause these concentrations persist for 70 years, and that
the concentrations are the same inside and outside the residences.  From
this it can be seen that public exposure is based on a hypothetical rather
than a realistic premise.  It is not known whether this results in an over-
estimation or an underestimation of public exposure.
     B.6.2.2  The Public.  The following are relevant to the public as
dealt with in this analysis:
     •    Studies show that all people are not equally susceptible to
          cancer.  There is no numerical recognition of the "most
          susceptible" subset of the population exposed.
     •    Studies indicate that whether or not exposure to a particu-
          lar carcinogen results in cancer may be affected by the
          person's exposure to other substances.   The public's expo-
          sure to other substances is not numerically considered.
     •    Some members of the public included in this  analysis are
          likely to be exposed to compounds in the air in the work-
          place, and workplace air concentrations of a pollutant are
          customarily much higher than the concentrations found in the
          ambient or public air.  Workplace exposures  are not numeri-
          cally approximated.
                                   B-49

-------
     •    Studies show that there is normally a long latency period
          between exposure and the onset of cancer.   This has not been
          numerically recognized.

     •    The people dealt with in the analysis are not located by
          actual residences.  As explained previously,  they are
          "located" in the Bureau of Census data for 1980 by popula-
          tion centroids of census districts.

     •    Many people dealt with in this analysis are subject to
          exposure to ambient air concentrations of potentially toxic
          pollutants where they travel and shop (as in downtown areas
          and suburban shopping centers), where they congregate (as in
          public parks, sports stadiums, and school  yards), and where
          they work outside (as mailmen, milkmen, and construction
          workers).  These types of exposures are not dealt with
          numerically.

     B.6.2.3  Ambient Air Concentrations.  The following are relevant to
the estimated ambient air concentrations used in this analysis:

     •    Flat terrain was assumed in the dispersion model.  Concen-
          trations much higher than those estimated would result if
          emissions impact on elevated terrain or tall  building near a
          plant.

     •    The estimated concentrations do not account for the additive
          impact of emissions from plants located close to one another.

     •    Meteorological data specific to plant sites are not used in
          the dispersion model.  As explained, meteorological data from
          a National Weather Service station nearest the plant site are
          used.  Site-specific meteorological data could result in
          significantly different estimates, e.g., the estimates of
          where the higher concentrations occur.

     •    With few exceptions, the emission rates are based on assump-
          tions and on limited emission tests.  See the Background
          Information Document for details on each source.

B.7  REFERENCES

 1.  U.S. Environmental Protection Agency.  Health Assessment Document for
     Nickel and Nickel Compounds.  Publication No. EPA-600/8-83-012FF.
     Office of Health and Environmental Assessment, Washington, DC.  1986.
     p. 8-156.

 2.  Reference 1, p. 8-156.

 3.  U.S. Environmental Protection Agency.  Carcinogen Assessment of Coke
     Oven Emissions.  Publication No. EPA-600/6-82-003F.  Office of Health
     and Environmental Assessment.  Washington, DC.  1984.  p. 147.
                                   B-50

-------
 4.  Reference 1, p. 8-161.

 5.  Reference 1, p. 8-179.

 6.  Reference 1, p. 8-162.

 7.  Reference 1, p. 8-179.

 8.  U.S. Environmental Protection Agency.  Health Assessment Document for
     Carbon Tetrachloride.  Publication No. EPA-600/8-82-001F.  Environ-
     mental Criteria and Assessment Office, Cincinnati, OH.  1984.
     p. 12-10.

 9.  U.S. Environmental Protection Agency.  Hazardous Waste Management
     System; Identification and Listing of Hazardous Waste; Final Rule.
     51 FR 28296.  1986.

10.  Memorandum from Lisa Ratcliff, EPA, to Bob Scarberry and Debra
     Dobkowski, EPA.  June 29, 1987.  Inhalation exposure limits.

11.  Reference 10.

12.  Memorandum from Branscome, M., RTI, to Docket.  September 28, 1988.
     Calculation of composite unit risk estimates.

13.  U.S. Environmental Protection Agency.  Guidelines for Carcinogen Risk
     Assessment.  51 FR 33992.  September 24, 1986.

14.  Reference 9.

15.  Memorandum from Coy, Dave, RTI, to McDonald, Randy, EPA/OAQPS.  May 2,
     1986.  Listing of waste constituents prioritized by quantity in Indus-
     try Studies Data Base, combined with WESTAT Survey and WET Model.

16.  U.S. Environmental Protection Agency.  Status Report of the RfD Work
     Group.  Environmental Criteria and Assessment office, Cincinnati, OH.
     1987.

17.  U.S. Environmental Protection Agency.  Burning of Hazardous Waste in
     Boilers and Industrial Furnaces; Preamble Correction.  52 FR 25612.
     1987.

18.  U.S. Environmental Protection Agency.  User's Manual for the Human
     Exposure Model  (HEM).  Office of -Air Quality Planning and Standards,
     Research Triangle park,  NC.   Publication No. EPA/450/5-86-001.  1986.

19.  Department of Commerce.   Local Climatological Data.  Annual Summaries
     with Comparative Data.

20.  U.S. Environmental Protection Agency.  Inorganic Arsenic NESHAPs:
     Response to Public Comments  on Health, Risk Assessment, and Risk
                                   B-51

-------
     Management.  Publication No. EPA/450-5-85-001.  Office of Air Quality,
     Planning, and Standards, Research Triangle Park, NC.  p. 4-13.

21.  Reference 20, p. 4-18.

22.  U.S. Environmental Protection Agency.  Guidelines for the Health Risk
     Assessment of Chemical Mixtures.  51 FR 34014.  September 24, 1986.
                                    B-52

-------
 APPENDIX C



CONTROL COSTS

-------
RESEARCH  TRIANGLE   INSTITUTE

Centerfor Environmental Systems     __  - r^               January  fif  198g
      MEMORANDUM

      TO:      R1ck Colyer, EPA/SDB and Bob Lucas, EPA/CPB

      FROM:    Steve York, RTI

      SUBJECT: Estimation of Nationwide Control Costs and Cost  Effectiveness  of
               Implementing Accelerated Rule


          The estimated nationwide upper bound and lower bound  control  costs  and
      cost effectiveness of implementing the accelerated rule are presented 1n the
      attached tables.  Table 1 presents the nationwide capital and annual1zed costs
      of implementing solvent recycling fac1Hty/TSDF fugitive  emission controls.
      The solvent recycling facility fugitive emission control  costs are based on
      the model unit costs presented In the January 26, 1988 memorandum from  Robert
      Zerbonia to Rick Colyer, SOB and Bob Lucas, CPB regarding the costs  for
      fugitive emission control at WSTF/TSDF model units.  The  solvent  recycling
      facility costs are for 99 large facilities, 207 medium facilities, and  250
      small facilities representing 448 WSTF and 108 recycling  facilities  without
      solvent recovery on site reusing solvent as fuel.  The TSDF fugitive emission
      control costs were calculated using the Source Assessment Model (SAM) and
      represent the estimated costs of controlling emissions from fugitive sources
      handling hazardous wastes consisting of 10 percent or greater total  organic
      content.  The SAM was used to estimate TSDF fugitive emission control costs
      because it can access the comprehensive TSDF Industry profile to  identify
      affected facilities managing waste streams greater than 10 percent organics
      and 1t accounts for throughput in generating control cost estimates.  There
      were more than 1400 TSDF that were projected to handle waste affected by the
      standards.  Recovery credits are not Included for fugitive emission  controls
      because most of the control costs are associated with TSDF and commentors on
      the proposed rule pointed out that a recovery credit is not reasonable  to
      assume for a TSDF which 1s not recycling.

          Table 1 also presents the nationwide capital and annualized costs and cost
      effectiveness of implementing WSTF process vent controls.  The control  costs
      for the 95 percent emission reduction are based on the model unit  costs
      presented in the January 26, 1988, memorandum from Robert Zerbonia,  RTI to
      R1ck Colyer, EPA/SDB and Robert Lucas, EPA/CPB regarding  the model unit
      condenser and carbon adsorption cost estimates.  The WSTF costs are  for 73
      large facilities and 167 medium facilities.  The 208 small facilities as
      defined for the post-proposal analysis (see the January 27, 1988, memorandum
Post Office Box 12194        Research Triangle Park, North Carolina 27709        Telephone: 919 541-6000

-------
from Steve York and Robert Zerbonla,  RTI to R1ck Colyer,  EPA/SDB and Robert
Lucas, EPA/CPB regarding the estimation of nationwide uncontrolled VO
emissions from solvent recycling facilities for the Accelerated Rule post-
proposal analysis) are not required to add control devices under the 3 Ib/hr
facility emission rate cutoff.  The estimated number of WSTF might Include
some TSDF with process vents since only 406 of the 448 WSTF (Including the 208
small facilities) reported handling waste solvents 1n the 1986 Screener
Survey.  The remaining 42 facilities reported recycling other hazardous wastes
1n solvent recycling operations; 29 of these 42 reported recycling halogenated
organics.

    The analysis of the technical feasibility of attaining the 95 percent
emission reduction 1s based on the use of condensers to control process vent
streams for which condensation will achieve 95 percent control and on the use
of carbon adsorption systems to control the remaining process vent streams
subject to the regulations.  Since Insufficient site specific Information is
available to determine which facilities could apply condensation rather than
carbon adsorption, upper and lower bound cost estimates were generated for the
95 percent emission reduction.  The upper bound cost estimates are based on
the assumption that fixed-bed, regenerable carbon adsorption systems would be
required to control process vents at all facilities with emissions above the
emission rate cutoff.  Similarly, the lower bound cost estimate is based on
the assumption that vent condensers could be used to control process vents at
all facilities with emissions above the emission rate cutoff.

    Table 2 presents the nationwide costs of implementing the accelerated
rule.  The two control options represent combinations of fugitive emission and
process vent control yielding upper and lower bound nationwide cost estimates.


SBY/ddf

-------
                                      TABLE 1. NATIONWIDE* COSTS AND COST EFFECTIVENESS OF
                                         SOLVENT RECYCLING FACILXTY/TSDF F^mVE EMXsllON
  ControI
  Option
              Capital
               Cost,
           1980S x ID'*
 Equipment*       13.4
  L*aka

 Process Vents*   24.6
  Upper Bound      1.5
  Lower Bound
                               Annual
                               Cost,6 m
                            8/yr  K 1O"8
                                S.6
                                12.9
                               (0.07)
   Capital
    Cost,
1960S x  10-*
                                             US.2
   AnnusI
   Cost,«
S/yr » 1O-*
Capital          Annual
 Cost,           Cost,
      -«
                                                                          1966S x 10-
                                                                                             S/yr  x
 Emission       Cost
Reduction,   Effectiveness,
                S/Mg
                                                               29.S
                                                                             126.6
                                                                             24.6
                                                                              l.S
                                                                                             S2.9
                                                   12.9
                                                   (0.07)
                                               19,OOO


                                               7,200
                                                                                                                      480(1730)
                                           1,800
                                          30(2000)
 •mission control .t WSTF/TSOF model units
                                               slk
            using th* Sourc* Ass*ssmsnt M*d*l (SAM)  M
 •missions  from source, handling hazardous wa.tes with
 costs .re  b*s*d on th* mod.l unit costs pr.'.nted
 EPA/SOB snd Robert Lucss,  EPA/CPB r*asrdl
 for 73 l.rg* f.ciliti*. ind ^107^ msd?£ fl
 27, 1988,  m^orsndum from Steve York^nd  o
 •st mstlon of n.tionwld. uncontrollod VO S
.
   Include.  r«:ov.ry credit.. (  )  |ndlcst.s s net cost credit.
e  Do*, not  include recovery credits.
                                                                                     *              .
                                                                     11             , CPB regarding the costs for fugitive

                                                                   :r- ?-r£'J»i2r^s7.ftailsisjrjss

                                                                               !^* I*** !"» t V* •*'••»•'' control  costs w*r*
                                                                               costs of controlling •quip.mt leak fugitive
                                                                    8C22e*r tot*'  5r»'2le «••»**»*•  *STF process vent control
                                                                   ' 192*» TIW''*IM|U" fr*" "ob-rt '••••••»«•,  *TI to Rick  Coly.r.
                                                                   r            -dsorptlon cost .stlmst*..'  Th. WSTF  cosU.r?
                                                                    kcffl       pl/^n** IJ»«*-Prop... I  sn.ly.l. (.^ th.  Jsnusry
                                                                         '         iW-»£fcr* V'fl.W1 r***r"»9  ««•
•mission reduction
tot.,
                                                                          ** C>1. 000.000 gallons throughput).
                                                                                             (th        '
                                                                                                                Th* totsl
                                                                     -nd v"

                                                                                         '"--»". -g«,.r.b,. csrbon
Th. n.tionwld. coats ar. baa.don an sv.rsg. of thr*«tl of  «rk»! .!!!  it'*' "'S" *"'••'?"• ••«« *"• mission r.t* cutoff.
applied to th* 73  l.rg* facilities snd on th* cost J U.! «  k!   2 adsorption system, for Isrg* mod* I  unit cases 6,  8 snd 9
107 s*dlum fscilitl..:  Mod.l  un?t ?^J^£^ it^SSJlffl1^*** »^-^>« •»<«•• «»»* «— 26 .ppM J  t, th.
•stimaUs ar*  bas*d on th. assumption thst condensers c«uld 2TT.T*  '  ?  { «*»«randum cited abov*.   The lower bound coat
above th. .mission r.t. cutoff.   TUn.^^"^ colt* "^ J..td !n^.^.v.«rO' JTT*" T**.'* '" f*el11*1- •'«• mission.
cas*. 6,  7,  10, 11 and 12 appli*d to th. 73  arflrf.ci*Hi.s  Ind on in .!IT ° ',*£ "'^ °f «<>n<'-"Mr* for ''r»» "O
c.,e, 2e,  27,  and 28 applied to  the 107 medium fTclTlt *s             «v.rs8. of th* costs of cond.n.er, for medium
                                                                                                                         unlt

-------
                       TABLE 2.   Nationwide  Capital and Annualized Coata of Implementing Accelerated Rule
                                                         Emission             Coat
Nationwide Capital            Nationwide Annual          Reduction,        Effectiveneaa
Cost. 1986 8 x 1O"^          Coat, 8/yr x  1O-6           Mo/vr               8/Mq	
   Coat Range

Upper Bound                   161.2                         46.8                26,ZOO               176O

Lower Bound                   128.1                         32.8                19,000               173O

-------
RESEARCH   TRIANGLE   INSTITUTE

Center for Environmental Systems
                                                                 January 26,  1988
      MEMORANDUM

      TO:      R1ck Colyer, EPA/SDB and Robert Lucas, EPA/CPB

      FROM:    Robert Zerbonla, RTI

      SUBJECT: Model Unit Condenser and Carbon Adsorber Cost Estimates

          The WSTF control costs for Options I and II cases that Involve the use of
      condensers for VO emission control were estimated using a chemical engineering
      process simulator known as ASPEN (Advanced System for Process Engineering).
      Table 1 provides the chemical constituents and operating conditions that were
      used 1n the ASPEN runs as well as the condenser efficiencies that were
      generated.  The ASPEN condenser configuration consists of:  (1) a floating-
      head, 1-pass, shell and tube heat exchanger, (2) a refrigeration unit capable
      of producing chilled brine at a temperature of -20°F, and (3)  an optional
      primary water-cooled heat exchanger.  This final Item might be necessary 1n
      some Instances of volatile organic condensation to reduce the size of the
      refrigeration unit or to remove water vapor and avoid freezing problems.  In
      this design effort, RTI has assumed complete removal  of water vapor prior to
      VO condensation,  recognizing that the condenser temperature 1s low enough to
      cause Ice build up on heat transfer surfaces.

          ASPEN's cost  correlation for heat exchangers,  developed originally for
      plant-scale processes,  does not extend to the low flows examined for WSTF.
      Therefore,  vendor quotes (e.g.,  $2050 from Brink's Inc. for a  heat transfer
      area of 25 square feet)  were added to ASPEN to allow  cost scaling by condenser
      area for the low  flows.   A 25% over design factor was added to the calculated
      (theoretical) condenser areas to determine capital  costs.  Tables 2,  3,  and 4
      provide the costs and cost effectiveness numbers for  the large,  medium,  and
      small model unit  cases,  respectively.

          In the ASPEN  condenser model,  a small  refrigeration unit was Included to
      achieve a target  condensation efficiency of 95%.  The unit produces coolant
      (e.g.,  chilled brine)  for the heat exchanger at a  temperature  of -20°F;
      coolant flow 1s on the  shell  side of the condenser.   It 1s of  Interest to note
      that 1n only 15 of the  40 cases  examined was a removal  efficiency of  95
      percent achievable.   The efficiency of vent condensers  1s dependent of the
      physical  properties of  the solvents being condensed,  the solvent concentration
      1n the gas  stream,  and  the operating temperature of the condenser.

          In seven (7)  of the  40 cases,  ASPEN Indicated  that  for those particular
      situations  appreciable  condensation would not occur.  This results  from  the
      partial  pressure  of the  VO 1n the  vapor phase being too low to
      thermodynamically support a  liquid  phase.   Six  of  the seven  cases involved
ost Office Box 12194        Research Triangle Park, North Carolina 27709        Telephone: 919 541-6000

-------
methylene chloride as the VO constituted; methylene chloride has a boiling
point of 40°C (a relatively low boiling/high volatility compound).  This
compares with boiling points of 110°C, 79°C, and 74°C for toluene, MEK, and
1,1,1 TCE, respectively.  The seventh case that would not condense was a low
concentration (1%) MEK.

    For cases where condensation will not achieve a 95 percent control VO
reduction, carbon adsorption will be used as the control technology for the
emission reduction and cost analysis.  Those cases where carbon absorption 1s
the control technology used to achieve the 95 percent reduction 1n VO
emissions as called for under Option I are Case No. 5, 8, 9, 17, 18, 19, 20,
21, 25, 33, 36,  and 37.  Table 5 presents the costs and cost effectiveness
numbers for Option 1 controls using carbon adsorption.  All  carbon adsorption
costs were estimated using the EPA/EAB CONTROL COST MANUAL.  EPA 450/5-87-001A,
February 1987.

    There are twelve model unit cases that are below the Accelerated Rule
regulatory applicability cut-off of 10% VO content; these are Cases 1-4, 13-
16, and 29-32.  Although condensation 1s achievable and costs were generated
for some of these cases, these costs will not be included 1n the nationwide
costs of the regulation since the unit would not be affected by the regulation
as proposed.


RAZ/ddf

Enclosure
NOTE:  This memo was written prior to  the  decision  to  drop  the proposed
       10 percent concentration cutoff.

-------
 TABLE 1.  OPERATING  CONDITIONS USED IN ASPEN CONDENSER ESTIMATES
AND CONDENSER CONTROL EFFICIENCIES PREDICTED FOR MODEL  UNIT CASES,
CASE NO.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
ifi
17
18
19
20
21
22
23
24
25
26
27
2B
29
30
31
32
33
34
35
36
37
38
39
40
TOTAL FLM
(SCFN)
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
.2
.2
.2
.2
.2
.2
.2
.2
n
.i
1.2
1.2
1.2
1.2
1.2
1.2
1.2
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
VO RATE
(LB/HR)
0.40
0.40
0.40
0.40
4.20
4.20
4.20
4.20
10.60
10.60
10.60
10.60
0.17
0.17
0.17
0.17
0.42
0.42
0.42
0.42
1.20
1.20
1.20
1.20
5.00
S.OO
5.00
5.00
0.08
0.08
0.08
0.08
0.24
0.24
0.24
0.24
1.00
1.00
1.00
1.00
CONSTITUENT

HECHL
TOL
1,1,1 TCE
NEK
HE CHL
TOl
1,1,1 TCE
NEK
NE CHL
TOL
1,1,1 TCE
HEK
HE CHL
TOL
1,1,1 TCE
HEK
NE CHL
TOL
1,1,1 TCE
NEK
NE CHL
TOL
1,1,1 TCE
NEK
HE CHL
TOL
1,1,1 TCE
HEK
NE CHL
TOL
1,1,1 TCE
NEK
HE CHL
TOL
1,1,1 TCE
HEK
NE CHL
TOL
1,1,1 TCE
NEK
OP HRS
(Hfl)
BXX XX3X X3
4160
4160
4160
4160
4160
4160
4160
4160
4160
4160
4160
4160
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
2080
VO PERCENT

XXXX3SXXXSXX'
i:
12
i!
1Z
in
in
in
in
251
25Z
24!
251
31
3!
31
31
10!
10!
10!
10!
20*.
20!
20!
20!
59J
58!
551
61!
3!
3!
3!
3!
10!
10!
10!
10!
31!
30!
30!
31!
CONOEN EFF
* (!)
1 X3XXXXS XX XX
NC
45
16
NC
NC
95
95
87
44
95
95
95
NC
82
72
50
NC
82
69
80
26
95
95
95
87
95
95
95
NC
80
70
47
NC
95
90
83
58
95
95
95
    NC = NO CONDENSATION.

-------
TABLE  2.    ASPEN  CONDENSER  DESIGN  AND  COST  ESTIMATES  FOR  THE  LARGE  MODEL  UNIT  CASES  (1-12),  OPTION  1
PARAMETER CASE NO.
Coolant TeMperature I F)
Design Heat Load (Watt)
Required Condenser Area dq. »)
Refrigeration Capacity (ton*)
Heat Enchanjer tee Cott (1986 «)
Refrigeration Base Cost (1986 *)
Electricity, Refrigeration (1986 «)
WSE EQUIPMENT COST (1986 «)
TOTAL CAPITAL INVESTMENT (I9K «)
TOTfll ANNUAL COST (1386 «)
RECOVERY CREDIT ( */Yr )
COST EFFECTIVENESS (VNg VO renewed)
1 J
-«0
292
0.319
0.08
940
1600
a
255?
5050
8870
ISO
8160
3
-80
283
0.307
0.08
930
1560
a
2490
4940
2850
SO
25200
4 5 £
-30
578
0.624
0.16
1870
8540
a
MIO
7550
3420
3400
0
7
-80
468
0.51
0.13
1160
2210
a
3370
6690
3230
3400
-80
M»nsxs:
8
-20
608
0.663
0.17
1300
2650
a
J950
7840
3490
3090
60
9
-eo
557
0.607
0.16
1850'
2490
a
1740
7430
3400
3930
-60
snxmncm
10 •
-20
917
0.591
0.26
1840
3510
a
4750
9420
3830
8570
-250
II
-20
728
0.561
0.21
1210
8980
a
4190
8310
3590
8570
-860
12
-20
1184
1.28
0.34
1740
4180
a
5920
11740
4340
8570
-220
           NOTES)

           a)  Costs art IMS than 1100.

           I.  IrxtruMentationi  10* of Bate Equipment Cott  IBEC).
           2. Sales tax arid freight)  BS of KC * Instr.
           3. Purchased Equipment Cost (PEC)i  KC * Instr. + Sales Tax I Freight.
           4. Total Installation Cost (Direct  * Indirect))  67f of PEC
           5.  Total Capital Investment (TCl)i  PEC * Total  Installation Cost
           6. Supervision and Adiun. Labor i  I5X of Direct Labor.
           7. Maintenance Labor and Materials) 3X of TCI.
           6. Overhead)  60* of Op. Labor + Supv./Rdw. » feint.
           9. Property Ta»es,  Insurance, arid Admin. Charges)  4* of TCI.
           10. Capital Recovery! Kit over a IF-year service life.

-------
o
I—I

a.
o
CO
CM
 I
CO
CO
LU
CO


                                                              2 c
                                                                   ••* 'S 3
                                                                2 1 JS 13 ii**
                                                           ^  «i«Ti**5'S




                                                        S   "3  — oj M » «
*  3  JS $

-------
TABLE  4.    ASPEN  CONDENSER  DESIGN AND COST  ESTIMATES  FOR  THE  SMALL  MODEL  UNIT  CASES  (29-40), OPTION  1
PARAMETER CASE NO.
Coolant Tenperature ( F)
Design Heat Load (Uatt)
Required Condenser Area (sq. •)
Refrigeration Capacity (tons)
Heat Exchanger few Cost (1966 *)
Refrigeration Base Cost (I9S6 «>
Electricity, Refrigeration (1966 *)
BASE EOUIPMENT COST (1986 «)
TOTAL CAPITAL INVESTMENT (1966 «)
TOTAL ANNUAL COST (1986 II
RECOVERY CREDIT ( */Vr )
COST EFFECTIVENESS (*/Mg VO moved)
29 30
-20
K
0.027
0.01
320
300
a
610
1220
2030
30
30000
31
-20
23
0.025
0.01
310
280
a
590
1160
2020
20
45000
32 33
-20
24
0.026
0.01
110
290
a
600
1190
2030
20
45230
24
-20
37
0.04
0.01
380
390
a
760
1520
2100
100
9000
35
-20
31
0.33
0.01
350
340
a
690
1370
2070
90
9900
36
-20
33
0.042
0.01
390
400
a
780
1550
2110
80
11420
37
-20
55
0.06
0.02
450
510
a
960
1900
2180
250
3470
KKUUXUBE
38
»s:wax**l
-CO
79
0.046
0.02
400
660
a
1060
2100
2230
400
2060
SC3CS&M*
39
-20
61
0.041
0.02
380
140
a
930
1840
2170
400
1990
«»*n*Xl
40
««&sxsn
-20
103
0.089
0.03
540
730
i.
1320
2620
2340
400
2180
  NOTESi

  a>  Coit* are left than «IOO.

  I. Instrumentation:   10X of Bast Equipment Cost (CEO.
  2. Sales tax and freight*  8* of KC « Imtr.
  3. Purchased EquipMent Cost (PEC)i  SEC «• Instr. » Sales Tax I Freight.
  4. Total Installation Cost  (Direct * Indirect))  67* of PEC
  5. Total Capital Inve&tMent (TCI)t  PEC * Total Installation Cost
  6. Supervision arid Main. Labor: 151 of Direct Labor.
  7. Maintenance Liber  and Materials;  3* of TCI.
  8. Overhead) 60* of  Op. Libor * SUDV./MMU.  * Naint.
  9. Property Ta«es, Insurance, arid Aduin. Charges:  4t of TCI.
  10. Capital Recovery:   10*  c.ver a 15-year service life.

-------
tfl f~ w r*j •-  »  K

I????  P  s
"ISCS  S
? n 5- 3- s  i
T a< o o •»•  S
z 1t * s  *
I" Sg *  *
*Ut:  *
O. •»—•-" 5  —
s s i s *  |
* ? JS 4E i  g

53IIS  1
«| 11 *  ?
*> ~f -«» ^» V

»JS;
!sSi|
   —«»
S 2 •!"  S
g: y 5 "« 5

£i-J>i

iSSil
*|]«2i
?|ui

p
    «s
  a. - * S.

  -i'S
  if*?
   S "**"
 I S,IS
 liil
§  III
a  3 " -
|  ?|::

I   p^
5    n
!
i
,:  J
                 i

                 §
                 s
                 Jto  ro
                 3

                 "i
          S.  E.
          i'  ?
        ~  s  s
>  .«- W  .5
§  I £  i
                       u
                       3
  S    S

  S  I  i
                         s  s
                 n,  p, f.  =.


                 3  § S  i
     3i

     I  6
     iv  n>

     6  f

     S  J8
     UJ  ftj
     S  3



     i  i
                         g  I
                         B  -
                         .s  s
                 i  § i  I  i
                      -°  ^>
                         §u
                         3


                      I  "i
                            i
                            3

                            i
                 f
                                   —•  T)  «*"

                                   !  I  I
                                          I  •  I
                                          :  I
                 8



                 S



                 i


                 i
                                     S"
                                     vD
                      i  s «
                i

                i
*•  •*  S
n>  n»  H




iC?  n>  m




   «"  ft
   ro  a£
                                          ui

                                          s ^
                                 S?
                                 II BJ
                                 II §
                                 H (B
                                 N **
                           ^  i  a
                           U  X  >
                                   S 5
                                    S
                                                   S3 !
                                            I 01
                                                          m 30
                                                          < en
                                                          HH O
                                                           VO O
                                                           CJI C/>
                                                           ** o
                                                            73
                                                           O -O
                                                           O —I
                                                           Z i-i
                                                           —I O
                                                           JO Z
                                                           O
                                                           r— o
                                                            o
                                                           —I 3D
                                                           ST. O

                                                           r>
                                                           oo
                                                           z o
                                                           o co
                                                           m —i
                                                           z co
                                                           CO
                                           co -
                                           oo
                                           -oo
                                             CO
                                             m
                                             co

                                          8  ~

-------
RESEARCH   TRIANGLE   INSTITUTE

                                                                 September 22,  1988
Center for Environmental Systems


      MEMORANDUM

      TO:       R1ck Colyer,  EPA/SDB and  Bob Lucas, EPA/CPB

      FROM:     Robert Zerbonla, RTI \Jw\

      SUBJECT:  Costs for  Fugitive  Emission Control at WSTF/TSDF Model  Units


           The model unit  costs for controlling fugitive emissions  at waste solvent
      treatment facilities (WSTF) and  hazardous waste treatment, storage,  and
      disposal facilities  (TSDF) are presented In the attached  tables.   These costs
      reflect three levels of control.  Option 1 Involves  Implementation of a
      monthly leak detection  and repair program for pumps  and valves and
      Installation of leakless controls on pressure relief devices, sampling
      connections, and open-ended lines.  Option 2 Involves the Implementation  of  a
      monthly leak detection  and repair program for valves (NOTE: All valves at
      solvent recycling units are considered to be 1n light liquid  service) and
      Installation of leakless equipment  for pumps (dual mechanical seals), pressure
      relief devices, sampling connections, and open-ended lines.   Option  3 requires
      installation of leakless equipment  (sealed-bellows valves and dual mechanical
      seals) on valves and pumps as well  as leakless controls on pressure  relief
      devices, sampling connections, and  open-ended lines.  All costs are  presented
      in January 1986 dollars.

           Costs for the monthly leak  detection and repair program  are  based on
      results of January 1986 runs  of  the EPA/CPB LOTUS 123 costing program for a
      leak detection and repair program for pumps, valves, and  other potentially
      leaking sources (see RCRA TSDF A1r  Emissions Background Technical  Memoranda
      for Proposed Standards, EPA-450/3-86-009, October 1986).  Costs for  Option 2
      control equipment for pump and valves are based on the data contained in
      Appendix F (Supplemental Information for the Cost Analysis) of the May 1984
      Coke Oven By-Products'  BID (EPA-450/3-83-016a) and results of a valve and pump
      manufacturer/vendor  survey conducted after proposal.

      RAZ/ddf

      Attachments
^st Office Box 12194         Research Triangle Park, North Carolina 27709        Telephone: 919 541-6000

-------
       TABLE  1.  Cost,  Emission  Reduction,  and  Cost  Effectiveness  of
                   Implementing Option  1  Equipment Leak Controls on  the
                   WSTF/TSDF  Large Model  Unit.
aedsS unit II - option i


EMISSION SOURCE
0'JHP SEALS
LIGHT LIQUID
HEAV'f LIQUID
COMPRESSORS
FLANSES
VALVES
6A9
LIQUID
RELIEF VALVES
SAS
LIQUID
SAPLING CONNCTN
OPEN-ENDE2 LINES
SONIT3R. INSTRUMENT
TOTAL
NUHBER
OF
SOURCES

15
0
0
0

100
264

9
0
26
105
ia.b.c} HA
519
ANNUAL
EMISS.
iHq/Yri

6. a
0
0
0

C
U
16.4

B.I
0
3.5
1.5
NA
41.1
CONTROL
EFFIC.
r.)

61
0
100
0

73
59

100

100
100
NA

EMISSION
REDUCTION

-------
          TABLE 2.   Cost, Emission  Reduction,  and  Cost  Effectiveness of
                      Implementing Option  1 Equipment Leak Controls  on the
                      WSTF/TSDF Medium Model Unit.
*osel unit *2  - notion 1


EHI3SION SOURCE
PUflP SEALS
LIGHT LIQUID
HEAVY LIQUID
COMPRESSORS
FLANGES
VALVES
GAS
LIQUID
RELIEF VALVES
GAS
LIQUID
SAHPLIH6 CQNNCTN
OPEN-ENDED LINES
MONITOR. INSTRUMENT (a.t

NUHBER
OF
SOURCES

5
0
0
0

34
37

3
0
9
35
i.cj NH
173
ANNUAL
EHISS.
-.Hq/Yr)

2.2
0
0
0

1.7
5.4

2.7
0
1.2
0.5
NA
13.7
CONTROL
EFFIC.
(I)

si
0
100
A

73
59

100

100
100
NA

EMISSION
REDUCTION
•Hq.'lfr)

1.3

0.0
0.0

1.2
3.2

*"• 7

1.2
0.5
NA
10.2
CAPITAL
COST
(1986 1)

290
0
0
NA

340
(e!

11,540

6.150
2,340
6.310
2S.970
ANNUAL!!.
COST


1,790

0
NA

2,380
is)

3,280

1,530
5iO
2.370
11.910
, RECOVER i
CREDIT id)
l*,Yr>

59,;,

0
NA

5:0
i.420

1,220

530
230
NA
4,520
COST
EFFECT.
<*•'"<'•

500

NA
NA

100
,•51

7sO

530
*0
W
--,-,
                        !a)  AssUie one instruwnt per olant.
                        '.b>  Caoital recovery factor  is based on 10X interest ana i-/r ssuis
                        ici  !1aintsnce, taxss, and insurance = 25' of caoita! co=t.
                        id)  Assies anpromaateiv $450/Cg  credit for recovsrv.
                        !i)  Total for ooth gas and liquid.

-------
              TABLE  3.   Cost,  Emission  Reduction,  and  Cost Effectiveness  of
                           Implementing Option  1 Equipment Leak Controls on  the
                           WSTF/TSDF  Small  Model Unit.
•sods! unit t3 - option i


EMISSION SOURCE
 1:0
is) J50 •'?'

1,970 T30 c'O

320 320 as:
340 140 ;~0
2,370 NA NA
8,100 2.710 5'0
                        ia'i  Ass^as one instrusent per  slant.
                        bi  Camtal recovery factor is based on 10". interest and 6-/r  ecuiooert ii
                        ic.i  Haintence, taxes, and insurance = 25J of capital cost.
                        'd)  Assnases accroxiiatei / 1450/Mg credit for recovery.
                        ie1  Total  for both gas and liquid.

-------
                TABLE  4.   Cost,  Emission  Reduction,  and  Cost  Effectiveness of
                            Implementing Option  2 Equipment Leak Controls  on the
                            WSTF/TSDF  Large Model  Unit.
Ddei unit II - option 2


EMISSION SOURCE
PMP SEALS 'd
LIGHT LIQUID
HEAVY LIQUID
COMPRESSORS
FLANBES
VALVES idi
GAS
LI9UID
RELIEF VALVES
3A5
LISUID
SAflPLIHG CONNCTN
OPEN-ENDED LINES
MONITOR. INSTRUMENT
TOTAL
NUHBER ANNUAL
OF EHISS.
SOURCES !Hg/Yr!

15 6.
0
0
0

100 5,
264 16,

9 8
0
26 3
105 1
NA
519 41
CONTROL EMISSION
EFFIC.
JH

6
0
0
0

,1
.2

.1
0
.6
.5
NA
.1
REDUCTION
(Hg/Yri

95 6.3
0
0 0
0 0

73 3.7
59 9.6

100 8.1

100 3.6
100 1.5
NA NA
32.3
=========2============================"--— --——-— • — - — - — -— .
CAPITAL ANNUALS!. RECOVERY COST
COST COST CREDIT (a) EFFECT.
(1986 $1

144,750
0
0
NA

1,020
>bi

34,620

18,440
7,010
6,310
($/Yri <$/¥r) !*/Mg;

57,300 2.840 s.s-".1

0 0 M
NA NA NH

7,140 1.670 •>•••
(b) 4,3vO 'b-

9,840 3,660 lav

4,590 1,530 8*0
S,s30 s3; s70
2.370 NA HA
212,150 82, '20 14.730 2.030

 totes:
   ^A = 'tat
   •'a';  Asiuses aoorcsiBateiv I450.'«9 crsdit for recovery.
   (b>  "otai for both gas and liquid.
    ci  »c=ed 3n installation ot dual sechanical seals with barrier  fluid svstes and deoassina ,ents.

   ••:•  Bisid 3D LDAR for valves.

-------
          TABLE 5.   Cost, Emission Reduction, and Cost Effectiveness of
                      Implementing  Option 2  Equipment  Leak  Controls  on the
                      WSTF/TSDF Medium  Model  Unit.
.•»od?l unit 12 - ootion 2


EMISSION SOURCE
PiJHP SEALS ici
LiBHT LIQUID
HEAVY LI3UID
COMPRESSORS
FLANGES
VALVES id)
GAS
LIQUID
RELIEF VALVES
SAS
LIQUID
SAMPLING COHNCTN
OPEN-ENDED LINES
S1QN1TGR. INSTRUMENT
TOTAL
NUMBER
OF
SOURCES

5
0
0
0

34
87

3
0
9
35
NA
173
ANNUAL CON'
EMISS. EFF:
iMg/Yri (7.)

2.2
0
0
A
U

1.7
5.4

2.7
0
1.2
0.5
NA
* V T
i J. /
:======= =z==s:
FROL
1C.


95
0
0
0

73
59

100

100
100
NA

EMISSION
REDUCTION
(Hg/Yn

2.1

0
0

1.2
*.*• L

2.7

1.2
0.5
NA
10.9
CAPITAL ANNUAL!!. RECOVER1; COST
COST COST CREDITi a) EFFECT.
<19S6 t» i*/Yr) (*.Yr) it/Ho-

48,250 19,100 950 8.640
0
0 0 0 NA
NA NA NA NA

340 2,380 560 "0
(b) ,t!

11,540' 3,280 1.220 -oO

6,150 1.530 530 830
2,340 560 230 scO
6,310 2,370 NA NA
74,930 29,220 4, '20 2,230
   HA = Not Aaplicabie
   •a;  Assuses aosroxiaatelv 1450/Sq credit for recovery.
   •,b;  Total ror uottj gas and liquid.
   'c;  Based s~ installation of dual aechanical seals with  iiarrier fluid system ana ae2a5=:r:2 -ents.
    o   E'issd an LDAR for valves.

-------
           TABLE 6.   Cost,  Emission  Reduction,  and  Cost Effectiveness  of
                       Implementing Option  2 Equipment Leak Controls on  the
                       WSTF/TSDF  Small  Model Unit.
j.odsi unit 13 - option 2


EHISSION SOURCE
p'jHP SEALS (c;
LIGHT LIQUID
HEAVY LIQUID
COMPRESSORS
FLANGES
VALVES 
-------
      TABLE  7.  Cost,  Emission  Reduction,  and  Cost Effectiveness  of
                  Implementing Option  3 Equipment Leak Controls on  the
                  WSTF/TSDF  Large Model Unit.
isodel unit il - ootion 3


EHISSION SOURCE
PUNP SEALS >c)
LIGHT LIQUID
HEAVY LIQUID
COMPRESSORS
FLANGES
VALVES  (Z)

s. 6
0
0
0

5.1
16.4

8.1
0
3.5
1.5
NA
41.1
(


95
0
100
0

100
100

100

100
100
NA

EMISSION
REDUCTION
a)EFFECT.
!i/Yri ($/Yr) It/No..1

57,300 2.340 9,a40

0 0 NA
NA NA N;

504.870 2,300 23.030
•,b) 7,350 >L

9,540 3,550 75.;.

4,590 1,590 9iO
l.iSO &30 :"•:•
NA NA «-
578.280 13.440 13.5*;
:== =zsss=ss=ss=rsz=r=zs=r===sz====z=====
Notes:

f*A = Not nBClicable
•i'  Assuies aopro»iiatelv $450/Bg credit  for recovery.
 c1  Tot si or both qas and liauid.
 c.1  Basei on installat;3n of duai sechanical seals with barrier fluid svstes and deqassinq ^ents.
']<  Ba=9d on instailation of sealed-bellows Calves.

-------
        TABLE 8.   Cost, Emission Reduction, and Cost Effectiveness of
                    Implementing  Option 3  Equipment  Leak  Controls on the
                    WSTF/TSDF Medium Model  Unit.
icdsi unit 12 - cation 3


EMISSION SOURCE
P'JUP SEALS (c>
LIGHT LIQUID
HEAVY LIQUID
COMPRESSORS
FLANGES
VALVES idi
GAS
LIQUID
RELIEF VALVES
6A5
LIQUID
3A.1PLIH6 CQNNCTN
OPEN-ENDED LINES
MONITOR. INSTRUMENT
TOTAL
NUilBER
OF
SOURCES

5
0
0
0

34
87

3
0
9
35
NA
173
ANNUAL
EMISS.
(Mg/Yr)

2.2
0
0
0

1.7
5.4

2.7
0
1.2
0.5
NA
13.7
CONTROL
EFFIC.
(7.)

95
0
100
0

100
100

100

100
too
NA

EMISSION
REDUCTION
iMq/Yr!

2.1

•.
0

1.7
5.4

2.7

1.2
0.5
NA
13.6
CAPITAL
COST
il986 $)

48,250
0
0
NA

665,740
!b)

11,540

6,150
2.340
NA
734,020
==r:==r===
ANNUALIZ. RECOVERY COST
COST CREDIT (a; EFFECT.
tt/Yn ($/Vr) (t/flgj

19.100 950 e.e*0

0 0 NA
NA NA NA

167,830 7?0 23,190
(b) 2,430 \V

3,280 1,220 :60

1,530 530 330
560 230 6cO
NA NA NA
192.300 0,130 i3,i:*0
 (JA = ?lot Scpiicable
 •a)  Assuaes  aaDfusdsateiv J450. ilg credit for r=c2ver-'.
 •&•  Total for both gas and hqjid.
 .c;  Sasec on installation of dual aechanical seals «itn darner fluid svstea and  decassinc /=rt=.
 ,j'  £a=ed -:n installation of sealed-bellews valves.

-------
             TABLE 9.   Cost,  Emission Reduction,  Cost  Effectiveness  of
                         Implementing  Option 3  Equipment Leak Controls  on  the
                         WSTF/TSDF Small Model  Unit.
ssisl unit 33 - option 3


EHI5SION SOURCE
POMP SEALS (c-
LIGHT LIQUID
HEAVY LI3UID
CGHPRESSOR3
FLANGES
VALVES io)
SA3
LIQUID
RELIEF VALVES
GAS
LIQUID
SAILING CGH8CTN
3PEN-ENDED LINES
MONITOR. INSTRUMENT
TOTAL

NUMBER
OF
SOURCES

3
0
0
0

20
52

n
i.
0
5
21
NA
103

ANNUAL
EMISS.
!Hg/Yri

1.3
0
0
0

1
3.2

1.8
0
0.7
0.3
NA
8.3

CONTROL
EFFIC.
(I)

95
0
100
0

100
100

100

100
100
NA

_________
EMISSION
REDUCTION
(Mg/frl

1.2

0
0

1.0
3.2

1.8

0.7
0.3
NA
8.2

CAPITAL ANNUAL I Z. RECOVER '? CGST
COST COST CREDIT !a,< EFFECT.
(1986 *) ($/Yr> •:*/"'>' '$•«?'

28,950 11,4.0 540 5,iOv
0
0 0 0 "in
NA NA NH ^,

39s, 140 99,8.0 450 23. 3: :
!b) (b) 1.44'i it

6,920 i.970 "30 .^0

3,690 '20 320 !?•:•
i,400 340 140 c"-}
NA NA NA NA
437,100 114,550 ',.;•'• i~,^

Notes:

NA - Not AjslicaSle
\si  ^ssusss asproKisatelv  I45v/fig credit for  recovery.
'Q'  Total »or  bath qas and liQuid.
 •:;•  3sssd an installatiofi  of dual mechanical  seals «tn barrier  fluid systea  and
':   Basso 2P installation  of sealsfl-beilows valves.


-------
RESEARCH   TRIANGLE   INSTITUTE


Center for Environmental Systems
                                                                 September 28, 1988
      MEMORANDUM
                                  ^ .
      TO:      R1ck Colyer, EPA/SDB and Bob Lucas, EPA/CPB

      FROM:    Robert Zerbonla, RTI

      SUBJECT: Evaluation of Incineration as a Process Vent Control Technique


          The proposal analysis Included an evaluation of the use of Incineration
      (halogenated compounds)/flares (non-halogenated compounds) as a regulatory
      alternative achieving 98 percent control.  This level of control was not
      proposed because It resulted 1n minimal risk reduction at a significantly
      greater control cost than condensation.

          The proposal analysis Incineration control costs are based on the
      application of the lowest size vapor Incinerator (1 m3) at 98 percent control
      efficiency to process vent streams with 7 Ib/hr and 75 Ib/hr of organic
      emissions.  For the post proposal model units, the capital costs are the same
      (updated from June 1985 to January 1986) and the fuel-use component of the
      operating costs were factored based on the organic content of the vent stream.
      Table 1 presents estimates of the capital and annual 1 zed costs of vapor
      Incinerators for the (post proposal) model unit cases.  In estimating
      nationwide costs,  the number of facilities reporting on-s1te liquid hazardous
      waste Incinerators 1n the Industry profile could be accounted for.  The 1987
      Screener data shows 190 facilities with liquids Incineration on site; review
      of the first half of the 1987 National  TSDR Survey database Indicates that
      only 137 of the TSDF surveyed have Incinerators onslte.  As Input to the
      analysis of risk reduction,  a dBase file could be created for PAB with revised
      emission rates and vent stream characteristics (temperature, flow rates, stack
      heights, etc.).  It would also be possible to approximate the risk reduction
      achieved through use of an incinerator rather than a condenser/carbon adsorber
      by proportioning the risk based on the ratio of control efficiencies for the
      options.  Stack heights and  flow rates  would be similar for both control
      technologies.   Exhaust gas temperature  could be somewhat higher for the
      incinerator;  however,  this may not necessarily be the case 1f the incinerator
      exhaust gases  undergo  any sort of processing such as heat recovery or NOX or
      particulate controls.

          Regarding  the  control  efficiency, we feel  that  for this  regulation to be
      promulgated under  RCRA,  the  vent  stream would  not necessarily be classified  as
      a  hazardous waste,  requiring an  Incinerator that  achieves 99.99  percent
      destruction efficiency.   Recent guidance from  EPA's  Office of General  Council
      stated  that noncontalnerized gases  emitted  from hazardous wastes  are not
3St Office Box 12194        Research Triangle Park, North Carolina 27709         Telephone: 919 541-6000

-------
themselves hazardous wastes, because the statute Implicitly excludes them.
EPA does however have the authority to regulate them as emissions from
hazardous wastes.  Therefore, the revised analysis will examine vapor
Incineration at a control efficiency of 98% as the bases for process vent
controls.  Please call me to discuss the Inclusion of Incineration 1n the
analysis of process vent control  techniques.


RAZ/ddf

-------
  TABLE 1.   CAPITAL AND ANNUALIZED COSTS OF INCINERATORS FOR MODEL UNIT CASES
Model Unit Size

large
large
large
medium
medium
medium
medium
small
small
small
Model Unit
Case No.

1 thru 4
5 thru 8
9 thru 12
13 thru 16
17 thru 20
21 thru 24
25 thru 28
29 thru 32
33 thru 36
37 thru 40
Capital Cost,3
1986 S

211,000
211,000
211,000
211,000
211,000
211,000
211,000
211,000
211,000
211,000
Annual cost,b
$/vr

305,000
171,000
163,000
275,000
171,000
125,000
106,000
471,000
224,000
130,000
a Incinerator system costs Include 1 m3 combustion chamber,  150 ft duct work
  fan, stack and quench/scrub system.  It 1s unknown 1f the  system Includes a
  neat exchanger.

b The large model units operate 4,160 hr/yr, the same number of hours used In
  estimating operating costs for the proposal  cost analysis.  The medium and
  small  model units operate 2,080 hrs/yr, therefore, operating costs were
  halved before the fuel-use component was factored based on the organic
  content of the vent stream.

-------
RESEARCH   TRIANGLE   INSTITUTE
Center tor Environmental Systems
                                                                  January 6, 1989
       MEMORANDUM

       TO:      R1ck Colyer,  SDB,  and Robert Lucas,  CPB

       FROM:    Robert Zerbonla,  RTI

       SUBJECT: Results of the Source Assessment Model  (SAM)  Runs for TSDF
                Equipment Leaks


           The SAM was used to estimate nationwide uncontrolled and controlled
       organic emissions from TSDF equipment leaks.   Additional SAM runs were made to
       estimate total nationwide  cancer Incidence from equipment leak emissions and
       the nationwide costs of controlling these emissions through Implementation of
       a leak detection and repair program (LDAR) as well  as  certain equipment
       requirements.  The SAM runs were structured to estimate equipment leak
       emissions from TSDF waste  streams of 10 percent or greater total  organics and
       to reflect the estimated Increase 1n waste managed due to Implementation of
       EPA/OSW's land disposal restrictions.  The two control cases run  to estimate
       emissions and costs are:  Case A, LDAR for both pump and valves in light
       liquid service and equipment requirements for other sources; and  Case B, LDAR
       for valves and a dual  mechanical seal requirement for  pumps in light liquid
       service, with the same equipment requirements for other sources.

           Because the SAM (and the accompanying data bases)  was not designed to
       incorporate waste solvent  recycling units among the hazardous waste management
       units analyzed, equipment  leaks from these units were  estimated separately.
       Nationwide organic emissions from equipment leaks are  estimated at 18,410
       Mg/yr for the uncontrolled case for all  TSDF  waste management units excluding
       recyclers.  With recycling equipment leaks included, the total nationwide
       emissions from equipment leaks is 26,240 Mg/yr,  uncontrolled.  For control
       Case A, nationwide emissions from equipment leaks are  estimated at 5,289 Mg/yr
       with LDAR for all  TSDF equipment excluding recycling units.  With recycling
       unit equipment leaks included, the total  nationwide emissions from equipment
       leaks are 7,189 Mg/yr for  Case A.  Under the  more stringent controls which
       include dual  mechanical seals for pumps,  nationwide emissions from TSDF
       equipment leaks are estimated at 4,569 Mg/yr  excluding recycling  units which
       are not a part of the  SAM  estimates.   With recycling equipment leaks included,
       the total  nationwide emissions from equipment leaks are 6,077 Mg/yr for Case
       B.   Tables 1,  2,  and 3 present the results of the SAM  equipment leak emission
       estimates  for the  uncontrolled case,Case  A, and  Case B respectively.
       Attachment 1  is a  summary of the procedure used  by  the SAM to estimate
       emissions  from equipment leaks.   The  results  of  the SAM runs to estimate
       control  costs  are  presented  in Tables 4 and 5 for control  Case A.   Table 3
=bst Office Box 12194
Research Triangle Park, North Carolina 27709
Telephone: 919 541-6000

-------
presents the costs for Case B.  For Case A,  total  nationwide capital  costs are
estimated at $113.2 million, excluding control  of recyclers.  With recyclers,
the total capital costs are estimated at $126.6.   Total  annual costs  for
equipment leak controls at all TSDF waste management units other than
recycling units were estimated at $29.3 million per year.   With recycling
units, the total nationwide annual costs are $32.9 million/yr.  For Case B,
the capital costs are $173.8 million without recyclers and $211.7 million with
recycling units.  The total annual costs are $41.8 m1H1on/yr without
recyclers and $54.1 m1ll1on/yr with recycling units Included In the analysis
of Case B.  Attachment 2 provides a summary of the procedure developed for the
SAM to estimate control costs for equipment leaks at TSDF.


RAZ/ddf

-------
Ttble 1.  Uncontrolled Equipment Leak
  Emissions  (Excluding Recyclers) *t
         Estlnited by the SAN
HIT OIMICTC
M Facility
2
1
4
5
i
7
1
3
I* •
II
12
IJ
14
|j
Ii
17
II
13
»
21 ,
22
23 I
24
25
2i
21
23
30
31
32
33
34
35
X
37
31
33 '
40
41
42
43
44
45
+6
47
41
43
SO
Ictjl Top 50
Total All FJC
« Distributico
Ml
4.I2ME-OI
O.OMOEKtt
I.2M
•.25M
4.3W
4.4004£tOO
4. OOOOEKIO
4. 244*
4.1272
4.2345
•.2*44
4.1*41
4.JKCE-OI
•.1711
0.1332
ft. 74471-41
4.72ISE-4I
4. OOOOEtOO
0.(t73E-4l
4.t3*3E-4l
0.3I7U-OI
4. OOOOEtOO
0.3I4SE-OI
0. OOOOEKIO
0.«74SE-4I
0.75I3E-OI
•.42ME4I
4.4U2E-4I
0.4M4E-OI
0. OOOOEtOO
0. OOOOEtOO
0.33IK-01
O.OOOOEtOO
0.5KSE-OI
O.J234E-OI
4.342IE-4I
O.SW5E-OI
4.43ICE-4I
O.OOOOEKIO
0.535IE-01
O.OOOOEK«
4.4KIE-OI
0.273K-OI
4.427X-4I
O.J9KC-04
0.6094E-42
0. OOOOEKIO
O.OOOOEtOO
4.377
6.0W
Ml Ml Mt IN
1.117 4.0MMK4 fcMMEiM 4.3l27E-Ot
0.0040Et«4 tMMKM AMM4M «.MMEtM
•.4213 4.0000EK4 •.OOOOEtM «.4400Et*
•.3111 •.MNftM «.MMC
4.000MI4*
4.0MOEKM
4.4000EK4
•.OOOOEtOO
•.MMEMI
4.4I3IE-OE
O.OOMEt04
•.OOOOEK*
4. OOOOEKIO
O.MOS-M
4. OOOOEtOO
0. OOOOEKIO
0. OOOOEtOO
0.4040E««4
O.OMOEK*
0. OOOOEKIO
O.OOOOEKIO
0. OOOOEtOO
O.OOOOEKI4
0. OOOOEtOO
•.com too
•.OOOOEtOO
4. OOOOEKIO
O.OOOOEtOO
O.OOOOEtOO
0. OOOOEKIO
O.OOOOEKIO
O.OCIOOEKIO
4. OOOOEKIO
O.OOOOEtOO
0. OOOOEtOO
0. OOOOEKIO
0. OOOOEtOO
O.OOOOEKIO
O.OOOOEKIO
0. OOOOEtOO
0.4U4E-M
0. 16ME-05
0.00*
I.IU
MM
O.OMOE4W
•.OWOEKI4
•.OOOOftO*
•.MMKW
t.MME*M
4.MUE-M
O.OOOOEKI*
4.554K-41
4.IIM
4.0000EK*
O.OOHOEK*
O.OOOOEKI*
4.00MEKW
«. OOOOEKIO
O.OOOOEI04
0. OOOOEKIO
•.OOOOEtOO
0.0400EK4
4.K8E-OI
0. OOOOEKIO
O.OOOOEtOO
0.2306E-01
4.1243E-OI
O.IZ2K-M
•.•MOCK*
0.530M-4I
O.OOOOEKW
O.OOOOEt04
O.OOOOEKIO
0.737K-42
•.OOOOEtOO
O.OOOOEKIO
0. OOOOEtOO
0.4748E-OI
O.OOOOEt«0
O.OOOCCtOO
4. OOOOEtOO
O.OOOOEtOO
0. OOOOEtOO
0.21 ICE-OS
O.OOOOEKKI
0. OOOOEKIO
0.3WOE-OI
4.343
4.611
35.01*
4.0MOEKI*
*.OOME<0»
*.4000Et04
•.MMt«ft»
4XO»MEt44
4.0000EK*
•.OOOOEtOO
•.OOOOEtOO
4.0000EtO*
4,0040EK»
•.OOMEtO*
*.4000EK«
«.4000Et«4
•.MOtEK*
•.40NEKW
«.OOOOEt04
O-OOOOEtOO
O.OOOOEK*
4. OOOOEKIO
4.0000EtOO
0. OOOOEtOO
4.0400EK)*
4.0000E104
4.2343E-43
O.CMOEK*
0.4*54E-03
0.450JE-03
4.ISME-OI
•.OOOOEtOO
4. OOOOEtOO
4. OOOOEtOO
4. OOOOEKIO
0.1307E-0*
•.OOOOEtW
•.OOOOEKIO
0. OOOOEKIO
O.OOOOEtOO
4.2396E-04
4. OOOOEtOO
0. OOOOEKIO
0. OOOOEKIO
0.2507E-C*
0. OOOOEtOO
•.OOOOEtOO
O.ICWE-04
0.1424
0.1585
0.36*
O.OOOOIM4
•••OOOftOO
•.OOOOEK4
•.OOOOEKI4
•.OOOOEtOO
«.0«OOEt04
«.M«tt«
•.OOOOEtOO
•.OOOOEtOt
•.2390E-02
•.ONOEtW
*.MNC«M
•.OOOOEtOO
4.4000EIO*
•.OOOOEtOO
•.OOOOEtOO
4.0KXCKIO
•.OOOOEtOO
0. OOOOEtOO
4. OOOOEKIO
•.OOOOEtOO
4. OOOOEKIO
•.OOOOEtOO
«. OOOOEtOO)
0.0400EKW
•.OOOOEtOO
•.OOOOEtOO
4.2I43E-OI
•.OOOOEtOO
•.OOOOEtOO
0. OOOOEKIO
•.OOOOEtOO
0. OOOOEtOt
•.OOOOEtOO
•.OOOOEtOO
4. OOOOEKIO
•.604CE-02
4.0000EKIO
0.3070E-OI
0. OOOOEtOO
0. OOOOEKIO
0. OOOOEtOO
0.3H1E-02
4. OOOOEKIO
0. OOOOEKIO
•.OOOOEKIO
O.M27E^I
0.1141
0.62*
4.000KK4
4.0000H44
•.4000EK4
•.OOOOEtOO
0.«400EtOO
4.0400EKM
ft. OOOOEtOO
4.4000EKIO
•.OOOOEtOO
•.MMEK*
ft. OOOOEtOO
4. OOOOEKIO
•.OOOOEtOO
ftOWOEtOO
4. OOOOEtOO
•.OOOOEKIO
4. OOOOEtOO
•.OOOOEtOO
4. OOOOEtOO
4. OOOOEtOO
0. OOOOEtOO
0. OOOOEtOO
0. OOOOEKW
•.OOOOEtOO
4. OOOOEKIO
0. OOOOEKIO
0. OOOOEtOO
0. OOOOEtOO
0. OOOOEtOO
•.OOOOEKIO
•.OOOOEtOO
•.OOOOEKIO
0. OOOOEKIO
0. OOOOEtOO
O.OOOOEtOO
•.OOOOEtOO
0. OOOOEtOO
O.OOOOEtOO
O.OOOOEKIO
0. OOOOEKIO
0. OOOOEtOO
0. OOOOEKIO
0. OOOOEKIO
0. OOOOEKIO
0. OOOOEtOO
0. OOOOEKK)
O.OOOOEKKI
0. OOOOEKIO
0.00*
Ml
•.400CEKW
•.MNCt«0
•.MMEKW
».»»OOEt«0
•.WMEHM
4.4000EKI*
4.4000EKW
•.MOCEtet
•.MOOEtOO
•.MOOEKM
•.OOOOEtOO
*.OOHEte»
•.OOOOEKIO
•.OOOOEKIO
4. OOOOEKIO
•••OMEtW
4.4000Et«o
•.OOWCtOO
4. OOOOEKIO
O.OOOCCKXI
4. OOOOEtOO
ft.OOOOE.KKI
4. OOOOEtOO
4.«OOOEtOO
4. OOOOEtOO
0. OOOOEtOO
0. OOOOEKIO
4,0000£tOO
•.OOOOEtOO
4. OOOOEtOO
•.OOOOEtOO
4. OOOOEtOO
4.000CCtOO
•.OOOOEKW
•.OOOOEKIO
•.OOOOEKIO
•.OOOOEKIO
0. OOOOEtOO
•.OOOOEKIO
0. OOOOEKIO
4. OOOOEtOO
O.OOOOEtOO
4. OOOOEKIO
0. OOOOEtOO
•.OOOCCtOO
•.OOOOEtOO
0.0900EKIO
O.OOOCCtOO
0.00*
Ml
4.0000EI4*
4.0000EK4
4.0000EtO»
0,OOOOEK«
4.0000EK»
•.OOOOEKI4
0.4000EK*
4.0000EK*
«.«NOEt«
4. OOOOEKIO
4.0000E10*
O-OOOOEtOO
•.OOMEtOt.
4.4000EI44
•.OOOOEtO*
4. OOOOEtOt
•.OOOOEtOt
0. OOOOEtOO
0. OOOOEtOt
aooooEioo
0, OOOOEtOt
0. OOOOEtOO
0. OOOOEtOt
0. OOOOEtOt
4. OOOOEtOt
O.OCIOOEKIO
0. OOOCCtOO
0. OOOOEtW
O.OOOOEtOO
o. ooocctoo
0. OOOOEtOt
O.OOOCCKIO
0. OOOCCtOO
0. OOOOEtOO
O.OOOOEKIO
0. OOOOEtOO
0. OOOOEtOO
0. OOOOEKIO
0. OOOOEtOt
0. OOOOEtOt
0. OOOOEtOI)
0. OOOOEKIO
O.OOOOEtOO
O.OOOOEtOO
0. OOOOEtOt
0. OOOOEtOO
0. OOOOEtOO
0. OOOOEKIO
O.OOOCCKIO
o.ooocetoo
O.M64E-03
o.oo*
FniitiMt nay
Mil tem» Mfcnftt
M. Md CM*
•.KM
4.IM4
i.m
•.7312E-04
•.MMEtM
tvOOOOEtOO
0. OOOOEKIO
•.MMEtM
•.OOMEtOO
•.OOMtOt
•.MMEtM
•.IUIE-41
4.0000Et04
dlSME-M
4.I4JSE-4I
•.MMEtM
•.MMEtM
•.W7E-04
•.MMEtM
4. OOOOEtOO
•.MMEtM
•.SHCE-42
0.47WE-02
ft. OOOOEtOO
O.I5HE-OI
0.4M2E-OJ
0. OOOOEtOO
0.14ME-03
4.7373E-02
0. OOOOEKIO
4.4002E-02
•.OOOOEtOO
•.OOOOEtOt
».4*4X-«4
o.ooocetoo
0.»2HE-«2
•.OOOOEtOt
O.I7&2E-03
O.OOOOEtOO
0. OOOOEtOt
4.2047E-02
3.225
17.511
I.3H IS.3U
MM (7.i»
I.Mt I7.3H
I.Mt 47.71*
I.4K SS.4U
4.7733 33.W*
•.4553 U.I3*
•.JR4 (4.24*
4.234J «.•»
4.2444 M.34*
ft. 1333 74.0S*
4.1MB 7I.OM
•.1771 72.05*
•.1711 72. M*
4.1338 73.73*
HUTS 74.37*
•.MM 74.31*
•.1477 75.54*
4.105* 7S.12K
•.Mil 7l.*7*
4.M7IE-M 77.20*
4.37S9E-OI 77.71*
4.I335E-M 7I.7M
4.»7HE-»1 73.13*
4.I34X-OI 7S.M*
0.4035E-4I M.53*
4.77I3E-4I M.34*
4.7440E-OI «.3t*
•.7USE-M 11.77*
O.U77E-4I K.1I*
4.M47E-OI 12.44*
O.SU5E-OI 12.74*
0.35S2E-OI §3.05*
0.5456E-OI 13.34*
0.5402E-OI I3.M*
O.S3UE-OI 13.33*
•.5352E-OI M.22*
0.535IE-OI 14.51*
•.4727E-4I *4.77*
0.466IE-OI B.42*
0.4432E-OI K.1H
0.4273E-OI (5.43*
0.4067E-OI 15.71*
4.3M7E-4I 15.32*
0.3W3E-4I M.I3*
4.3I2CE-4I K.34*
15.90
It. 41

-------
                  ***************
 *************************************************
 l!!iiiIi!!i!ilIil!i!!!ill!ili!IU!liili!iH!llii!

liiififHiifffiififiiiiiiiiHiiiiiiiiiiiiiiiiii
                                         *****
* s>
ii!
   **************************************************
  i:^::s:s:ssss;:5:::ss;::::::s;s:s::::j:;j;s;:;:;;j
 ii



 ii



 ii

f r r
§£§


P*S I
•il
 ii

"ii
 ii



sii
   **************************************************

   !ff!!!f!!!!!!!f!HfH!!!f!!!!!i!ff!HH!H!!!HU!
   **************************************************
   **************************************************
   i'.
   **************************************************
   f>*f,*** ********************************************
   mmmmmmmmmmmmmmmnm
                                 **************
   *************************
    f********************************************
                             **************?.,.—^^,
                                              •  •

                                              !  f
                                                  Ifl
                                                  !B
                                                  tl!
                                                  •9
                                             i!
                                              1

-------
Air
Table 2.  Equipment Leak Emission for
 Control Case  A (Excluding  Recyclers)

N Facility
2
I
4
5
7
9
:o
II
12
14
15
16
17
It
13
30
21
22
33
24
35 	 -
27
2S
29
30
31
33
33
34
35
36
37
33
40
41
43
43
44
45
ti
47
•3
43
50
Tetil 'c.p »
T:ti! 01! Fir
l-.M U.K

901
0.3276E-02
0.3544
0.7454E-01
o.oootetoo
0.27M
0. OOMEtOO
O.OOOOEKIO
9.77JIE-OI
0.3777E-OI
0.49UE-OI
0.59I4E-OI
O.I320E-05
O.II35E-O.I
0.38621-01
0.50»4E-OI
0.2804E-OI
0.4135E-OI
0.2312E-OI
0.3129E-OI
0.13SX-OI
0.2934E-OI
0.1*4 7E-OI
O-OOOOEtOO
O.OOOOEKIO
0.2714E-01
0.2604E-01
0. OOMEtOO
0.220a£-OI
O.I376E-01
0.1302E-OI
O.OOOOEKIO
O.OOOOEtOO
O.IMSE-OI
O.II64E-OI
O.IS71E-01
0.9770E-OJ
0.10J6E-OJ
0. I60SE-OI
0. I282E-OI
0. IS90E-OI
O.OOOOEKIO
O.OOOOEtOO
0. OOMEtOO
0.4227E-02
0. I270E-OI
O.OOOOEtOO
O.OOOOEKIO
O.OOOOEtOO
0.684SE-02
0.6645E-02
1.464
1.763

as Estimated by the SAM
90S 901 904 TOI
0.3333 hOOMEtOft 0.0000£tOO 0.27I2E-02
0.1262 O.OOOOEKIO 0. OOMEtOO O.OOOC€tOO
0. OOMEtOO O.OOOOEtOO O.OOOOEKO 0.7IIIE-04
aOMOEtOO O.OOOOEKW O.OOOOEtOO O.OOOCfKX>
0.9285E-OI O.OOOOEKIO O.OOOOEtOO O.OOOOEtCO
O.OOOOEtOO O.OOOOEtOO «.«MOEtftO O.IS37E-OI
O.OOOOEKO 0.0000£tOO O.OMOEK0 O.OOOOEtOO
0.2331E-01 O.OOOOEKO 0.00001 too ft.MUE-01
0.13S6E-OI O.OOOOEtOO O.OOOOEK* O.I047E-OS
O.OOOOEKK1 0.0000£tOO O.OMOEtOO O.OOOOEKK)
O.OOOOEKK) O-OOOCftOO «.OOOOEtM 0.0000£tOO
0.3322E-OI 0. OOMEtOO O.OOOOEtOO O.OOOOEKW
0.2MIE-OI O.OOOOEtOO O.OOOOEKIO 0.1SME-03
0.1037E-OI O.OOOOEKIO 0. OOMEtOO O.OOOOEtOO
^* COOOE*00 0* OOOOf Htt Ou MOOC^Oft A. AMtf itiA
O.OOOOEKK) O.OOOOEKO O.OMC€tOO O.OOOOEtOO
O.OOMEK4 ft.OOMEt«ft O.OOOOEtOO O.OOOOEK*
0.7336E-03 O.OOOOEKIO 0-OOOOEtM O.OOOOEKIO
0.6543E-02 O.OOOOEtOO O.OOOOEKIO (.OOMEtOO
0.7I40E-OS O.OOOOEtOO O.OOOOEKO O.OOOOEKIO
O.OOOCttOO O.OOOOEKO O.OOOOEtOO O.OOOOEKO
0.6649E-02 O.OOOCttOO O.OOOOEKO O.OOOOEKK)
fcOMOEtOO O-OOOCttOO O.OOOOEKO O.OOOOEKO
O.OOOOEK-0 O.OOOCCt« O.OOOOE<00 O.OOOOEtOO
0.2923E-04 O.OOOOEKIO 0. OOMEtOO O.OOOOEKK)
ft-OMOEtOO O.OMOEt«0 O.OOOOEtOO O.OOOOEtOO
O.OOOOEKO O.OOOOEKIO 0. OOMEtOO O.I903E-03
0.8044E-07 O.OOOOEtOO O.OOOCttOO O.OOOOE*00
0.4946E-08 O.OOOOEKIO 0. OOMEtOO O.OOOOEKIO
0.4693E-02 O.OOOOEKIO O.OOOOEKK) O.OOOOEKK)
0.1I6IE-OI O.OOOOEKO O.OOOOEKK) O.OOOOEKIO
O.OOOOEtOO O.OOOOEKIO O.OOOCttOO 0.4097E-03
0.4Z76E-0* O.OOOOEKW O.OOOOEtOO O.OOOOEKiO
0.4162E-M O.OOOOEtOO O.OOOOEtOO O.OOOOEtOO
O.OMOEt« 0.0000£tOO O.OOOOE«00 O.OOOOEtOO
0.350IE-03 O.OOOCttOO O.OOOOEtOO O.M78E-04
0.368IE-OJ O-OOOCEKO O.OOOOEtOO 0. OOMEtOO
O.OMOEKK) O.OOOOEKX) O.OOOCttOO O.OOOOEKO
O.I936E-02 O.OMOEtOO O.OOOOEtOO O-OOMEK*
O.OOOOEtOO O.OOMEKX) O.OOOOEKK) O.OOOCttOO
O.OOOOEKKI O.OOOCttOO O.OOOOEtOO O.OOOCCKIO
O.OOOOEK10 O.OOOOEKK) O.OOOOEtOO O.OOOOEKIO
O.I770E-03 O.OOOOEKO O.OOOOEKIO 0.4496E-03
0.29ME-02 O.OOMEtOO 0. OOMEtOO O.OOOOEKIO
O.OOOCttOO O.OOOOEtOO 0. OOMEtOO O.OOOOEKiO
O.OOOOEtOO O.OOOOEtOO 0. OOMEtOO O.OOOOEKIO
O.OOOCftOO O.OOOOEKIO O.OOOCttOO 0.60I4E-04
O.M75E-03 O.OOOOEtOO O.OOOOEtOO O.OOOOEKKI
0.2449E-02 O.OOOOEKIO O.OOOOEKIO O.OOOOEtOO
0.238SE-03 O.OCOOEKIO O.OOOOEKIO O.OOOOEKX)
0.7334 O.OOOOEtOO O.OOOCttOO 0.24IIE-OI
0.8648 O.OOC<€t«l 0.3455E-04 0.3590E-OI
I6.3f* O.'Xtt 0.00* 0.49*
IIV M HCILITIEB WTIIMngE
CMlSSltna M 1000 »!/yr
T02
0. OOMEtOO
0. OOMEtOO
0. OOMEtOO
O.OOOOEKIO
0. OOMEtOO
O.OOOOEKX)
O.OOOOEKIO
0. OOMEtOO
O.OOOtttOO
O.OftOOEtOO
0. OOMEtOO
0. OOMEtOO
O.OOOOEKO
0. OOMEtOO
0. OOMEtOO
O.IOS9E-06
O.OOOOEKIO
0. OOMEtOO
0. OOMEtOO
O.OOOOEKO
0. OOMEtOO
O.OOOOEKO
O.OOOOEKO
0.3SI8E-IO
O.OOOOEKO
O.OOOOEKO
O.OOOOEKO
0. OOMEtOO
O.OOMEtOO
0. OOMEtOO
0. OOMEtOO
O.OOOOEKO
0. OOMEtOO
0. OOMEtOO
O.OOOCCtOO
O.OOOOEKIO
0. OOMEtOO
0. OOMEtOO
O.OOOOEKIO
0. OOMEtOO
O.OOOOEtOO
O.OOOCttOO
O.OOOOEKIO
O.OOOOEtOO
O.OOOOEKIO
O.OOOOEKO
O.OOOOEKiO
O.OOOOEKIO
O.OOOOEtOO
O.OOOOEKIO
O.I060E-06
0.4463E-06
0.00*
TO)
0.4332
0.0000£tOO
0. OOMEtOO
0.533S
O.OOOCttOO
O.OOOOEKX)
O.OOOOEKK)
0. OOMEtOO
0.2045E-01
O.OOOOEKW
O.OOOOEKO
0.342JE-OI
O.OOOOEtM
O.IS74E-OI
O.OOOOEKX)
O.OOOC£tOO
O.OOOOEKX)
O.OOOOEtOO
0, OOMEtOO
O.OOOOEKO
ft. OOMEtOO
O.OOOOEtOO
O.OOOOEKO
ft. OOMEtOO
O.OOOOEKIO
O.OOOOEKK)
0.235K-OI
aOMOEtOO
0.3704E-02
0.365K-02
0. OOMEtOO
0.1753E-OI
0.6420E-08
0. OOMEtOO
O.OOOOEKIO
0.2376E-02
O.OOOOEtOO
0. OOMEtOO
O.OMOE«00
O.OOOOEKO
0.1J85E-0!
0. OOMEtOO
O.OOOCttOO
O.OOOOEKIO
O.OOOOEKO
O.OOOOEKO
O.I069E-OI
O.OOOOEKIO
O.OOOOEtOO
O.OOOOEKIO
1.173
1.353
33.ee*
T04
0. OOMEtOO
O.OOOOEKK)
0.37E2E-OI
O.OOOOEKO
O.OOOOEtOO
0.0000£tOO
0. OOMEtOO
0-OOOOEtOO
0. OOMEtOO
0. OOMEtOO
0. OOMEtOO
O.OOOOEtOO
0. OOMEtOO
0. OOOOEKX)
0.0000£tM
O.OOOOEtOO
0.0000£tOO
O.OOOOEKK)
O.OOOOEKIO
O.OMOEKK)
O.OOOOEKK)
O.OOOOEtOO
ft. OOMEtOO
O-OOOOEtOO
0. OOMEtOO
O.OOOOEKX)
0. OOMEtOO
0.1779E-04
O.I446E-03
O.I34IE-03
0.4709E-02
ft. OOMEtOO
0. OOMEtOO
0. OOMEtOO
ft. OOMEtOO
0.566IE-07
0. OOMEtOO
O.OODOEKO
O.OOOOEKIO
O.OOOOEtOO
O.OOOOEtOO
O.OOOOEKO
O.I905E-05
0. OOMEtOO
O.OOOOEKO
O.OOOOEKO
0.314IE-05
O.OOOOEKK)
O.OOCOEKW
O.OOOOEtOO
0.4230E-OI
0.4707E-01
o.ew
079
0. OOMEtOO
0. OOOOEKX)
O.OOOCCKIO
0. OOMEKK)
O.OOOOEtM
O.OOOOEKX)
0. OOMEtOO
0. OOMEtOO
O.OOOOEtOO
O.MOOEKIO
0. OOMEtOO
O.OOOOEtOO
0. OOMEtOO
ft.UME-«3
O.OOOOEtOO
O.OOOOEKXI
O.OOOOEtOO
0. OOMEtOO
0. OOMEtOO
ft. OOMEtOO
aOOMEtftO
ft. OOMEtOO
O.OOOOEKIO
O.OOOOEKO
O.OOOOEtOO
O.OOOOEKK)
ft.MOC€tM
O.OOOOEtOO
0. OOMEtOO
O.OOOOEKX)
0.6386E-02
O.OOOOEtOO
O.OOOOEKO
O.OCOOEKO
O.OOCOEKIO
0. OOOOEKX)
O.OOOOEKIO
O.OOOOEtM
O.OCOOEKO
O.OCOOEKIO
O.J326E-02
O.OOOOEtOO
0.9120E-03
O.OOOOEtOO
O.OOOOEtOO
O.OOOOEKX)
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEtOO
0. 1767E-OI
0.3I80E-OI
0.60t
MO
O.OOOOEKKI
O.OOOOEKK)
O.OOOOEKO
«.MMEtOft
O.OOOOEKX)
O.OOOOEtOO
O.OOOOEKO
O.OOOOEtOO
0. OOOOEKX)
O.OOOOEKK)
O.OOOOEKO
O.OOOCttOO
O.OOOOEKIO
O.OOOCtKK)
O.OOOOE400
O.OOOOEKK)
O.OOMEtOft
O.OOOOEKXI
O.OOOCttOO
O.OOOCtKK>
O.OOOOEKO
O.OOOOEKO
O.OOOOEKO
O.OOOOEKO
0. OOMEtOO
O.OOOCttOO
0. OOMEtOO
O.OOMEtM
O.OOOCtKX)
O.OOOCttOO
O.OOOOEKIO
O.OOMEtM
O.OOOOftOO
ft. OOMEtOO
O-OOOCtKO
O.OOOOEtOO
O.OOOOEtOO
O.OOCKtKO
0. OOOOEKX)
O.OOOOEKIO
O.OOOCttOO
0.(«OOEKX)
0. OOOOEKX)
O.OOOOEKIO
O.OOOOEtOO
O.OOOOEtOO
O.OOOOEtOO
O.OOOOEKO
O.OOOC€tOO
O.OOOOEtOO
O.OCK
Ml
0. OOOOEKX)
O.OOOOEKK)
O.OOOOEKIO
0. OOOOEKX)
O.OOOOEtOO
O.OOOOEKW
O.OOOOEKKI
O.OOOOEKKI
0.0000£tOO
O.OOOCttOO
O.OOOOEKIO
0. OOOOEKX)
O.OOOOEKKI
O.OOOOEKIO
O.OOOOEKO
O-OOOOEiOO
0.0000£tOO
O.OOOOEKX)
0. OOMEtOO
0. OOMEtOO
0. OOMEtOO
O.OOOOEtOO
O.OOMEtM
O.COOOEtOO
0. OOMEtOO
O.OOOCttOO
0. OOMEtOO
O.OOOCttOO
O.OMOE400
O.COOCtKO
0. OOMEtOO
O.OOOOEKIO
0. OOOOEKX)
O.OOOOEtOO
O.OOOOEtOO
O.OOOOEtOO
0. OOMEtOO
O.OOOOEKIO
O.OOOOEtOO
O.OOOOEKX)
O.OMOEKW
O.COOOEtOO
O.OOOOEtM
O.OOOOEKIO
O.OOOOEKIO
O.COOCfKKI
O.OOOOEKIO
O.CrtOCttOO
O.OOOOEKX)
O.OMOT*00
O.C'.OOEKiO
0.00*
M)
O.OOOOEKIO
0. OOMEtOO
O.OOOCttOO
0. OOMEtOO
0. OOMEtOO
0. OOOOEKX)
0. OOOOEKX)
0. OOOOEKX)
0. OOOOEKX)
O.OOOOEtOO
O.OOOOEK*
CUMMEKft
ft. OOMEtOO
ft.MMEtOO
•.MMEtftft
0. OOMEtOO
0. OOMEtOO
O.OOOOEKO
O.OOOOEtOO
O.OOMEtOO
aooocttoo
O.OOOOEKO
O.OOOOEtOO
ft. OOMEtOO
0. OOMEtOO
O.OOOOEKO'
O.OOOOEtOO
a. ooooEtoo
O.OOOOEtOO
ft. OOMEtOO
0. OOMEtOO
O.fOOCttOO
0. OOMEtOO
0. OOMEtOO
O.OMOEtOO
O.OOOCttOO
O.OOOOEtOO
O.OOOOEtOO
O.OOOCttOO
O.OOMEtOO
0. OOOOEKX)
ft. OOOOEKX)
0. OOMEtOO
O.OOOOEtM
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEtOO
O.OOOOEKIO
O.OWOEKO
O.OOOOEKO
0. 1504E-09

Fniitlvn 
-------
 I
 I*
 Sa S


  1
           iSSSISSi
           iastfaasi
	5Sgggi
ddddddddddd^
                                              «
                                              oodooooooo
o^S «>
u jjiii
?!!!!!!!!!!!!!!!!iiisiiiissifiisiisiiiiiliiisiill!is
 111111111111111111111111
      dddddddeid.

dddddddSdddddddddddddddddddddddddddddddddddddddddd


?fl??l?l??§??????????ll?ll§?f?ll???????????????f||


ddddddd<*dddddddddd«Sddd<{
-------
        Table  3.  Emissions,  Total Capital  Costs, and
           Total Annual Costs for Equipment Leaks
                   Under  Control Case B.
   Fugitive Emissions,
   Control: Altern. B
                                                    06-Jan-89
        TCI and  TAG -far  VO >10*/.
tt
1
*"?
.*-
•— •
4
5
6
7
8
9
10
11
12
13
14
15
16
17
13
19
20
21
22
f\^f
.£.<*>
PROCESS
SOI
S02
SOS
S04
TO!
T02
T03
T04
D79
D80
D81
D83
THL
LTA
INC
PT1
PT2
PT3
PT4
PT5
XS02
FXP
LDR
EMISSIONS
MG/YR
1 . 5O3E+03
7.411E+02
O.OOOE+OO
2. 163E-02
2.214E+01
4.015E-04
1. 114E+03
4.O22E+01
2.751E+01
O.OOOE+OO
0 . OOOE+00
1 . 504E-07
8. 182E+02
0 . OOOE+00
O.OOOE+OO
O.OOOE+OO
O.OOOE+OO
0 . OOOE+OO
0 . OOOE+00
O.OOOE+OO
0 . OOOE+OO
O.OOOE+OO
2.974E+02
TCI
$
4.655E+07
3.548E+07
0 . OOOE+OO
1. 117E+04
1 . 402E+06
1 . 409E+02
5.355E+07
1 . 738E+06
2.309E+06
0 . OOOE+00
O.OOOE+OO
1 . 865E+0 1
2, 133E+07
O.OOOE+OO
0 . OOOE+00
O.OOOE+OO
O.OOOE+OO
0 . OOOE+00
0 . OOOE+00
O.OOOE+OO
O.OOOE+OO
O.OOOE+OO
1 . 097E+07
TAG
*/YR
1 . 183E+07
8. 168E+06
O.OOOE+OO
2.303E+03
3.457E+05
2.941E+01
1. 1S2E+07
4.894E+O5
5. 115E+05
O.OOOE+OO
O.OOOE+OO
3.734E+00
5.919E+06
O.OOOE+OO
O.OOOE+OO
O.OOOE+OO
O.OOOE+OO
0 . OOOE+OO
0 . OOOE+00
O . OOOE+OO
O.OOOE+OO
O.OOOE+OO
2. 733EK)6
     TOTAL
4.569E+03
1.738E+08
4.1S2E+07
NOTE:  Does not include recycling units

-------
1 OH 1C * • IUfc» 1 »»wi* i »• • ~~-- • — -T -•
Tot.i Cipitit in,«t»nt Leak Control Case A (Excluding Recyclers) TOP so FKUITIES WTIOMIDE
COST IN *
N Facility

.
j
4
s
f
7
I
5
to

12
13 '

.-
It
|7
II
I)
20
21
22
23
24
25
26

21
21
30
31
32
33
34
3}

37

u
40

42
43
44
45

47

49
SO
Trial Tcp SO
Tt*»l All fx
( DistritHitlOr.
SOI
O.OOOOEtOO
O.OOOOEtOO
0.7798EKI7
0.7606EKB
0.2875EKI7
0.4032£t06
O.I487EKI7
O.I490EKI7
O.OOOOEtOO
0.644BEKI6
O.OOOOEtOO
O.OOOOEtOO
CUOOOOEtOO
0.66UEK*
0.4707EK*
0.6II3E«06
O.OOOOEtOO
O.SS64EI06
458.4
0.43I6EK*
1231.
0.8048EKI5
0.318IEK*
140.1
0. 1607EKI6
0.3I36EK*
0. 17C6EK*
O.OOOOEtOO
0. 15B4EK*
0.233IEKI6
0.232UKI6
0. IS22EK*
O.«67£t05
O.OOOOEKIO
0. 1758EK«
O.I097£Kt
O.OOOCtK*
O.I426C«06
0.9249EK0
O.OOOOEKIO
0.8883EKI5
0.844KKB
0. 13IS£0
O.OCIOOEtOO
cuoooctioo
O.OOOQEKIO
0.5IIOEKI5
O.OOOOEIOO
O.OOOOEKIO
O.OOOCEtOO
O-OOOOEtOO
O.OOCIOEtOO
0.1224EKQ
O.OOOOEIOO
O.OOCIOEtOO
O.OOOOEtOO
O.OOOOEIOO
O.OOOOEtOO
0.2033EKIS
O.OOOCttOO
O.OOOOEKIO
0.89I2EK6
0.76I3EKI5
O.OOOCttOO
0.4234EKI8
0.4323EKI8
38.251
T04
O.OOOOEtOO
O.OOOOEKIO
O.OMKCKW
O.OOOOEKW
O.OOOOKOO
0.3244EK*
O.OOOOEK>0
0.3524EKIS
O.OOOOEtOO
O.OOOOEtOO
0.000(ttOO
(UOOOOEtOO
(UOOOOEtOO
O.OOOOEtOO
(UOOOOEKiO
(UOOOOEtOO
O.OOOCttOO
O.OOOOEtOO
0.0000£tOO
O-OOOOEtOO
O.OOOCttOO
O.OOOOEKIO
O.OOOCtKIO
O.I073E-OI
O.OOOOEtOO
O.OOOOEtOO
O.OOOCtK>fl
O.OOOCttOO
(UOOOOEtOO
O.OOOOEtOO
O.OOOOEKIO
O.OOOOEtOO
O.OOOOEKIO
O.OOOCttOO
O.OOOCtKIO
O.OOOCtKIO
O.OOOCttOO
436.1
0.6000
O.OOOCtKIO
O.OOOOEKIO
1107.
O.CKttttOO
O.OOOOEKIO
725.7
0.26IIEKQ
O.OOOOEKIO
O.OOOOEKKI
O.OOOCttOCi
O.OOOOEKIO
O.J8B7EKI6
O.W34EKI6
0.801
073
O.OOOOEtOO
O.I21K*07
O.OOOCtKIO
O.OOOOEtOO
O.OOOOEKIO
O.OOOOEKX)
(UOOOOEtOO
0.4775EKI5
O.OOOOEKIO
0.202X«05
O.OOOOEtOO
O.OOOOEtOO
0.0000£tOO
CuOOOOEtOO
O.OOOCttOO
O.OOOOEKIO
O.OOOOEIOO
(UOOOOEKIO
O.OOOCtKIO
O.OOOOEtOO
O.OOOCttOO
O.OOOCttOO
O.OOOOEtOO
0.2434EI05
O.OOOCttOO
CUOOOOEtOO
O.OOOOEtOO
O.OOOCttOO
O.OOOCttOO
O.OOOOEKIO
O.OOOCtKIO
O.OOOCttOO
O.OOOCttOO
O.OOOOEKIO
O.OOOOEtOO
O.OOOCttOO
O.OOOCtiOO
O.OOOOEKIO
O.OOOCtKIO
O.OOCIOEtOO
O.OOOOEKIO
O.OOOOEKIO
(I.OOOCttOO
O.OOOOEKIO
O.OOOCtKIO
0.3524EKQ
O.OOOOEKIO
O.OCIOOEKIO
0.4694EKI5
O.OOOOEtOO
0. 14&3EKI7
0. 1938EKI7
1.711
mo
O.OOOOEtOO
O.OOOOEtOO
O.OOOOEK>0
O.OOOOEK>0
(UOOOOEtOO
O.OCIOOEKIO
O.OOOOEtOO
O.OCIOOEKIO
O.OOOOEtOO
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEIOO
O.OOOOEKIO
O.OOOOEKIO
(UOOOOEKIO
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEtOO
O.OOOOEIOO
O.OOOOEKIO
a.ooooEtoo
O.OOOOEKX)
O.OOOOEtOO
O.OOOOEtOO
O.OOOOEKXI
O.OOOOEKIO
O.OOOCttOO
(I.OOOCttOO
O.OOOCttOO
O.OOOOEtOO
O.OOOOEtOO
O.OOOOEtOO
O.OOOOEKIO
O.OOOCttOO
O.OOOOEtOO
O.OOOCttOO
O.OOOCttOO
Cl.OOOCttOO
O.OOOOEKIO
O.OOOOEKXI
O.OOOOEKIO
O.CKXXttOO
O.OOCIOEtOO
O.OOOOEtOO
O.OOOOEIOO
O.OOOOEKIO
O.OdOOEKIO
0.00*
Ml
(UOOOOEtOO
O.OOOOEIOO
*.OOOOEiO»
O.OOOOEt*0
O.OOOOEt«0
O.OOOOEIOO
(UOOOOEtOO
(UOOOOEtOO
O-OOOftEtOO
O.OOOOEKIO
O.OOOOEMW
(UOOOOEKIO
O.OOOCEKO
aooooEtoo
O.OOOOEKIO
O.OOOOEKIO
aooooEtoo
O.OOOCtKIO
(UOOOOEKIO
O.OOOCttOO
C-OOOOEKIO
O.OOOOEIOO
(UOOOOEtOO
O-OOOOEiOO
O.OOOOEK4
8.0000EKIO
O.OOOOEKIO
(UOOOOEtOO
O.OOOOEIOO
«.0000£tOO
(UOOOOEKIO
O.OOOOEKIO
(UOOOOEKIO
O.OOOOEI40
O.OOOOEtOO
O.OMOCKO
O.OOOCttOO
O.OCIOOEtOO
O.OOOOEKIO
O.OOOOEIOO
O.OOOCttOO
O.OOOCttOO
O.OOOOEKIO
O.OOOCtKIO
O.OOOCttOO
O.OOOCttOO
O.OOOOEKlO
O.OOCIOEtOO
O.OOOOEKXI
O.OOOCftOO
O.OOOCtKIO
O.OOOCttOO
0.00«
M3
O.OOOOEtOO
O.OOOOEKIO
O.OOOOEIOO
O.OOOOEKIO
O.OOOOEKIO
O.OOOtttOO
(UOOOOEKIO
O.OOOOEtOO
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEtOO
O.OXOEtOO
O.OOOOEtOO
O.OOOOEtOO
O.OOOOEtOO
(UOOOOEKIO
O.OOOOEKIO
(UOOOOEtOO
O.OOOOEtOO
O.OOOOEtOO
O.OOOOEKXI
O.OCIOOEKIO
(UOOOOEKIO
O.OOOOEKIO
CUOOOOEtOO
O.OOOOEKIO
O.OOOOEKXI
O.OOOOEtOO
O.OOOOEKXI
O.OOOOEKXI
O.OOOOEtOO
(UOOOOEKW
O.OOOOEtOO
O.OOOOEtOO
o.ooocttoo
O.OOOOEtOO
O.OOOCttOO
O.OOOOEtOO
O.OCIOOEKIO
O.OOOOEtOO
O.OOOCttOO
0.000ft tOO
O.OOOCCtOO
O.OOOCtKIO
O.OOOOEtOO
O.OOOOEtOO
O.OOOCftOO
O.OOOCtKIO
O.OOOCtKIO
o.coooetoo
O.OOOOEKXI
18.54
0.00*
Fgftti
W»l i
TW.
O.OOOOEtOO
O.OOOOEKIO
O.OOOOEK0
O.I957EK*
0. 64IXK«
0.3371EKI7
0.3I5KKI6
O.OOOOEKXI
O.OOOOEKIO
O.OOOOEKXI
O.OMXCIOO
O.I1MEKI7
0.206IEKI5
O.I3ISEKI5
A.IOKEK4
0.253X105
O.OOOOEKXI
O.OOOOEKXI
O.I86SEKI6
(UOOOOEtOO
O.OOOOEtOO
OJ37XKC
O.OOOCftOO
265.2
120.1
O.OOOOEKXI
0.3742EI05
O.OMCEKM
O.J502EKI5
O.OOOOEtOO
O.OOOOEIOO
O.OOOOEtOO
CUOOOOEtOO
O.OOOCttOO
O-OOOCttOO
0.t3KEK6
O.OOOCttOO
O.OCIOOEtOO
74M.
(UOOOOEtOO
O.I94IEKI5
O.I933EKI5
(UOOOOEKIO
(UOOOOEtOO
7726.
O.OOOCttOO
O.OOOCttOO
0.2SS6EKI5
O.OOC«tOO
O.OOOOEKXI
0.6377EKI7
0.7400EKI7
6.S4*
vtt wily
crMi tutvniti
Tot«l CUM*
O.BllEtOi 11.64*
O.I9E9EI08 41.61*
0.779K*07 55.56*
0.7691EKI7 62.37*
0.4S33EKI7 66.37*
0.4330EKI7 70.13*
0.2300EKI7 72.23*
0.2I24EKI7 74.10*
O.IM3EI07 75.77*
O.I2»3Et07 71.31*
O.I270EI07 78.03*
(UII32Et07 79.08*
O.IOMEt«7 10.03*
0.8867EKI6 I0.ll*
0.7449EK* 11.46*
0.6744EK* 12.06*
0.5264EKH 12.53*
0.4673EKC 12.34*
0.4S34EH* 13.34*
0.43I6EK4 13.73*
0.4230EK* 14.10*
0.4IS6EK« M.47*
O.J467Et06 84.77*
0.3I8IEKI6 85.05*
0.3143EKK 85.33*
0.3I36EK« 15.61*
O.K72EK* 15.85*
0.2640E««6 86.08*
0.2417EKH 86.30*
0.233IEKI6 86.50*
0.232*£t06 86.71*
O.IS22EKI6 86.871
(UI733EKI6 17.03*
O.I767EKI6 17.19*
O.I758EK* 17.34*
O.I7I1EKH 17.4.3*
O.I63CEK* 17.63*
O.ISS7E«06 17.77*
O.I430E<06 17.30*
O.I423EKI6 88.031
O.IJSSEK* 88.151
O.I34IEK* 88.27*
O.I312EK« 88.33*
O.I303EK* 88.501
O.I298Et06 M.62*
O.I254EKI6 88.731
O.I2SIEK* 88. Ml
0. 133IEK* 88.951
O.I231EK« S3. 06*
O.II9IEK* 09.16*
O.IOMEKW
O.II32EKI9


-------
 Total Capital InvntMnt
N    Facility
Table  4.   Total Capital Cost  for Equipment
Leak Control  Case A (Excluding Recyclers)
                                8     '     '   TOP SO FACILITIES WTIMIIDE
                                                      COST IN    «
                                                                                                                                              Tctal tcnw v+mHt
                  LT«
                            IK
                                      Ptl
                                                PT2
                                                          PT3
                                                                    PT4
                                                                              m
                                                                                        «SM
                                                                                                                       IM
                                                                                                                                 III
                                                                                                                                           m
                                                                                                                                                     Tot»l
1
2
1
S
i
7
a
3
10
II
12
13
14
IS
16
17 .
ia
13
20
21
22
23
24
25
36
27
2a
23
30
31
32
33
34
35
36
37
M
33
40
41
42
43
44
45
46
47
48
43
50
Total Top SO
Total All Far
1 Distribution
0. OOOOEKW
O.OOOOEKW
0. OOOOEKW
O.OOOOEtOO
O.OOOOEKW
O.OOOOEKW
0. OOOOEKW
0. OOOOEKW
O.OCWOEKW
O.OOOOEKW
O.OOOCCKIO
O.OOOOEKIO
0. OOOOEKW
O.OOOOEtOO
O.OOOOEKW
0. OOOOEKW
0.00X6*00
O.OOOOEKW
O.OOOOEKW
O.OOOOEKW
0. OOOOEKW
0.00X6*00
O.OOOCCKW
O.OOOOEKW
0.00X6*00
O.OOOOEtOO
0. OOOOEKW
O.OOOCCKW
0. OOOOEKW
O.OOOOEKW
O.OOOOEKIO
O.OOOCCKW
0.00006X10
O.OOX6KW
O.OOOOEKW
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEKW
O.OOOOEKW
O.OCWOEKW
0. OOOOEKW
O.OOOdEKW
O.OOOOEKW
O.OOOOEtOO
O.OOOOEKW
O.OOOOEKIO
O.OOOOEKW
O.OOOOEKW
O.OOOOEKW
O.OOOOEKW
O.OOOOE.OO
O.OOfc'CKW
0.001
O.OOOOEKW
0. OOOOEKW
O.OOOOEKW
0. OOOOEKW
O.OCWOEKW
0. OOOOEKW
0. OOOOEKW
O.OOOOEKW
0. 00X6100
0. 00X6*00
O.OOOOEKW
0.00X6*00
O.OOOOEKW
O.OCIOOEKW
O.OOOOEKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
0.00X6 tOO
0. OOOOEKW
•.00X6*00
O.OOOOEKW
O.OOOOEKW
0. OOOOEKW
•.00006*00
O.OOOOEKW
O.OCWOEKW
0. OOOOEKW
0.00006*00
O.OOOOEKIO
0. OOOOEKW
0.00006*00
0.00X6*00
O.OXOEKW
O.OOOOEtOO
O.OOOOEKW
0. OOOOEKW
0.00X6*00
0.00X6*00
O.OOOOEKW
O.OOOOEtOO
O.OOOOEKIO
O.OOOCCKIO
O.OOMEKW
O.OOOCCKIO
O.OOOOEKW
0. OOOOEKW
O.OCWOEKW
0. OOOOEKW
0. OOOOEKW
0.001
O.XOOEKW 0.00X6*00
O.OOXEKW 0. OOOOEKW
O.OOOOEKW O.OOOOEKW
0.00006*00 O.OOOOE*00
0.00006*00 0,00006*00
0,00006*00 O.OOOOEKW
0.00X6*00 O.OOOOEKW
tUOOOOEtOO O.OOOCEKW
O.OOOOEtO* O.OOOOEKW
O.OOOOEKW O.OOOOEKIO
0. OOOOEKW 0. OOOOEKW
O.OOXEKW O.OOOOEKW
O.OOOOEtOO 0.00X6*00
0. OOOOEKW O.OOOOEKW
0. OOOOEKW O.OOOOEKW
O.OXOEKW O.OOOOEKW
O.OOOOEKW O.OCWOEKW
•.OOOOEKW O.OOOOEKW
O.OOXEKW O.OOOOEKW
O.XOOEKW O.OOOOEKW
0.0000£*00 O.OOOOEKW
•.00X6*00 0. OOOOEKW
•.•0X6*00 O.OOOOEKW
t.OOOOEKW O.OOOOEKW
O.OOOOEKW 0. OOOOEKW
0. OOOOEKW O.OOOOEKW
0. OOOOEKW O.OOOOEKW
O.OOOOE*00 O.OOOOEKW
O.OOOOEKW O.OOOOEKW
0. OOOOEKW 0. OOOOEKW
O.OOCWEKW •.OOOOEKW
O.XOOEKW O.OOOCCKW
0.00X6*00 O.OOOCCKIO
0.00X6*00 O.OOOOEKW
0.00X6*00 O.OOOOEKW
0.00006*00 O.OOOCCKW
0. OOOOEKW O.OOOOEKIO
0.00006*00 0. OOOOEKW
0. OOOOEKW 0. OOOOEKW
•.00X6*00 0. OOOOEKW
O.OOOOEKW 0. OOOOEKW
O.OOOOEKW O.OOOOEKW
0.00X6*00 0.00006*00
0. OOOOEKW O.OOOOEtOO
O.OOOOEtOO O.OOOOEKW
O.OOOOEKIO 0. OOOOEKW
0. OOOOEKW O.OOOOEKW
O.OOOOEKW O.OOOOEtOO
O.OOOCCtOO O.OOOOEKW
O.OOOOEKIO O.OOOOEtOO
O.OOOOEtOO O.OOOOEKW
O.OOOOEKW O.OOOOEKW
O.OCK O.OC*
O.OOOOEKW
0. OOOOEKW
•.OOOOEKW
0.00006*00
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
•.OOOOEKW
0.00006*00
0.00X6*00
0.00X6*00
0. OOOOEKW
O.OOOOEtOO
O.OOOOEKW
O.OOOOEKW
0. OOOOEKW
0. OOOOEKW
O.OCIOOEKW
0. OOOOEKW
O.OOOOEKW
0. OOOOEKW
O.OOOOEtOO
O.OOOCCKW
0.0X06*00
O.OOOOEtOO
O.OOOOEKW
«. OOOOEKW
O.OOOOEKW
O.OOOOEKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
O.OOXEOO
0. OOOOEKW
0. OOOOEKW
O.OOOOE*00
O.OOOOEKW
O.OOOOEKIO .
O.OOOOEKW
0. OOOOEKW
O.OOOOEKXI
0. OOOOEKW
O.OOCKCKW
O.OOOOEKW
O.OOCWEKW
O.OOOOEKIO
O.OOOOEKO
0.00«
O.OOCWEKW 0.00X6*00
O.OOOOEKW 0.00006*00
O.OOOOEKW O.OOOOEKW
O.OOOOEKW 0.00006KW
O.OOOOEKW O.OOOCCKW
0. OOOOEKW O.OOOOEKW
0. OOOOEKW O.OOOOEKIO
0. OOOOEKW 0. OOOOEKW
O.OOOOEKW O.OOOOEKW
O.OOOOEKW O.OOOOEKW
O.OOOOEKW O.OCWOEKW
0.00X6*00 O.OOOOEKW
•.00006*00 O.OCWOEKW
O.OOOOEKW 0. OOOOEKW
O.OOOCEKW O.OOOOEKW
O.OOOOEKW O.CWOOEKW
O.OOOOEKW 0.00006*00
O.OOOOEKW 0.00X6*00
0. OOOOEKW 0. OOOOEKW
0.00006*00 0.00X6*01)
O.OOOOEKW 0. OOOOEKW
O.OOOOEtOO O.OOOOEKW
O.OOOOEKW O.OOOOEKW
O.OOOOEKKI O.OOOOEKW
O.OOOOEKW 0.00X6*00
•.•0X6*00 O.OCWOE*00
0.00006*00 0. OOOOEKW
O.OOOCCtOO O.OOOOEKW
O.OOOOEKW O.OCIOOEKW
O.OOOOEKIO 0.00X6*00
O.OOOOEKIO 0.00X6*00
O.OODOEKM 0.00X6*00
0.00006*00 O.OOCKCKW
O.OCIOOEKW 0.00X6*00
0.00X6*00 0.00006*00
O.OOOOEKW 0.40X6*00
0.00X6*00 O.OOOOEKW
0.00X6*00 O.OOOOEtOO
O.OOOOEKO 0.00006*00
O.OOOCCKIO 0. OOOOEKW
O.OOOOEKIO O.OOOOEKW
0.00006*00 O.OOOOEKW
O.OOOfCtOO O.OOOCCKW
0.00006*00 0. OOOOEKW
0.00(«KIO 0. OOOOEKW
O.OOOOEKIO 0.00006*00
O.OOOCCtOO O.OXOEKW
O.OOOOE*00 O.OOOOEKIO
O.OOOOEtOO O.OOOOEKIO
O.OOOOEKIO O.OOOOEKW
O.OOOCCKlO O.OOOOEKW
O.OOOOEtCKi O.OOCKCKW
a.oot o.ooi
O.OOOOEKW
O.OOOOEKW
0. OOOOEKW
O.OOOOEKIO
0. OOOOEtOO
0. OOOOEKW
•.WMEfM
O.OOXEKW
O.OOOOEKW
O.OOOOEKM
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
a OOOOEKW
O.OCWOEKIO
O.OOXEKW
O.OOOOEKW
O.OOOCCKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
O.OCWOEKW
O.OOCWEKW
0. OOOOEKW
0.00006400
O.COOOEKJO
O.OOOOEKW
O.CWOOEKW
0. OOOOEKW
O.OCWOEKW
O.OOOOEKIO
0. OOOOEKW
0.00X6*00
0.00006*00
O.OOOOEKIO
0.00X6*00
0.00X6*00
O.OOOOEtOO
0. OOOOEKW
0.00X6*00
O.OOOOEKW
O.OOOOEKW
0. OOOOEKW
O.OOOOEKW
O.OOOOEKW
O.OOOCCKW
0. OOOOEKW
0.00006*00
O.OCWOEKW
O.OOOOEKW
O.OOOCCtOO
O.OOOCCKW
0.00*
O.XOOEKW O.OOOOE*00
O.OOOOE*00 O.OOOOE*0*
O.OOOOEKW O.XOOEKW
0. OOOOEKW O.OOXEK*
0.00006*00 O.MM64*
O.OOOOE«00 «.23IOE*06
0.00006*00 O.OOOOEKW
O.OOOOEKW O.OOOOE«0»
0.00006KW O.IMX4W
O.OOXEKW O.OMOEtM
O.OOOOEKW O.I270EKff
O.OOOOEKW O.OOOOEtO*
O.XOOEtOO O.OOOOEKIO
O.OOOOEKW O.OOOOEKW
0. OOOOEKW O.OOOOEKI*
O.OOOOE*00 O.OOOOE*00
0. OOOOEKW O.S264EI06
O.OOOOEKW O.IMXXK
0. OOOOEKW O.OOOOEKW
0.00006*00 0.00006*0*
O.OOOCCKW 0. OOOOEKW
O.OOOOEKW O.OOOOEKW
O.OOOOEKW 0.12KEK*
O.OOOOEKIO o.ooocetoo
O.OOOOEKW O.OCWOEKW
0. OOOOEKW O.CWOO£tOO
O.OOOOEKW O.OOOOEKW
0.00X6*00 0.3MOEK*
O.OOOCCKW 0. OOOOEKW
0.0X06*00 O.OOOOEKW
O.OOOOEKW O.OOOOEKW
O.OOOOEKW O.OOOOEKW
0. 00006*00 O.OCIOOEKW
0. OOOOEKW O.I767EK*
O.OOOOEKW 0. OOOOEKW
O.CWOOEKW O.OOOOEKW
0. OOOOEKW 0.16MEK*
O.OOOOEKW 0.12536*03
O.OOOOEKW 4«3»,
O.OCIOOEKW O.I423E«K
O.OOOOEKW O.OOOOEtOO
O.OOOOEKW O.OOOCCKW
O.OOOOEKW 0.00006*00
O.OOOOEKW O.I30X*06
0.00«€KW O.OOOOEKW
0.00006*00 0, 00006*00
0.00X6*00 O.OOOOEKIO
O.OOOOEKM O.OOOOEKIO
O.OOOOEKIO O.OCWOEKW
0. OOOOEKW O.OCWOEKW
O.OCWOEKIO 0.5I20EKI7
O.OOOCCKW 0.6I04EKI7
0.001 5.351
0. OOOOEKW
•.ooooEtoa
•.OOOOEKW
0. OOOOEKW
O-OXOEtO*
•.MOOCH*
•.MOOCMO
•.00006*00
•.00X6 KW
0. OOOOEKW
0, OOOOEKW
•.OOOOEKW
O.OOOOE*00
•.00006*00
•.00006*00
0. OOOOEKW
•.OOOOEtOO
«. OOOOEKW
«. OOOOEKW
0. OOOOEKW
0. OOOOEtOO
•.OOOOEKW
O.OOOOEtOO
0. OOOOEKW
aOOOOEtOO
O.OOOOEKW
O.OOOCCKW
0. OOOOEtOO
O.OOOOEKW
O.OOOCCKW
O.OOOOEKW
0. OOOOEKW
•.OOOOEtOO
O.OOOOEKW
O.OOOOEtOO
O.OOOCCKIO
0. OOOOEKW
O.OOOCCKIO
•.00006*00
O.OOOCCKW
0. OOOOEKW
0. OOOOEtOO
0.00006*00
O.CWOCCKW
0. OOOOEKW
O.OOOCCKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
O.OOOOEKW
O.OOOOEKIO
0.001
0. OOOOEtOO
0.00006*0*
o. ooooEtoo
0. OOOOEKM
0.00006*00
0. OOOOEKW
•.OOOOE*Ot
0. MODE tOO
0-OOOOEtO*
0. OOOOEtOO
0. OOOOEKW
O.OOOOEKW
0. OOOOEtOO
O.OOOOEKX)
0. OOOOEtOO
0.00006*40
o.ooooE*oo
•.OOOOEKW
•.OOOOEtOO
0. OOOOEKW
O.XOOEKW
O.OCWOEKW
0. OOOOEtOO
0. OOOOEtOO
0. OOOOEKW
0. OOOOEtOO
O.XOOEKW
O.OOOOEKW
O.OOOOEtOO
•.OOOOEtOO
O.OOOOEKW
O.OOOOEKW
0.00X6*00
O.OOOOEKW
O.OOOCCKW
O.OOOOEKW
0.00X6*00
O.OOOOEKW
O.MWOE*00
O.OOOCCKW
0.00006*00
0.00X6*00
0.00006*00
O.OOOOEKW
O.OOOOEKW
O.OOOOEtOO
O.OCIOOEKW
0. OOOOEtOO
0. OOOOEtOO
O.OOOOEKIO
O.OOOOEtOO
O.OOOOEKW
0.001
•••0006 *••
0. OOOOEKW
t. OOOOEKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
•.00006*00
O.OOOCCKW
O.OOOOEKW
0. OOOOEKW
O.OXOEKW
0. OOOOEtOO
0. OOOOEtOO
0. OOOOEtOO
0. OOOOEKW
0. OOOOEKW
0. OOOOEKW
0. OOOOEtOO
0. OOOOEKW
0. OOOOEKW
0. OOOOEtOO
0. OOOOEKW
0. OOOOEKW
o.ocwoEtoa
0. OOOOEKW
O.OOOOEKIO
0. OOOOEKW
0. OOOOEKW
O.OOOCCKW
•.OOOOEtOO
O.OOOOEKW
0. OOOOEtOO
O.OOOOEKW
0. OOOOEKW
O.OOOCCKW
0.00X6*00
O.CWOOEKW
0.00X6*00
0.00006*00
O.OOOOEKW
0. OOOOEKW
0. OOOOEtOO
O.OOOOEKW
O.OOOCCKW
O.OOOCCKW
•.OOOOEtOO
•.OOOOEKW
0. OOOOEKW
0. OOOOEKW
0.00006*00
O.OOOCCKW
O.OOOCCKIO
0.001

O.I323EK* «J.U»
0.773(6*07 S.56*
0.763KKI7 ft. JH
0.45336*07 16.37*
•.43306*07 70.131
0.21006*07 72.23*
0.2I24EK>7 74.10*
O.I»WE«07 75.77*
0. 12a3f *07 76.31*
•.I270E*«7 Tt.03*
•.II32E»»7 73.0a*
0. |OUC*47 •0.02*

•.7443EK* M.46*
0.67446*06 K.OU
O.SJM6K* K.0t
0.4673E*06 12.34*
0.4S34E*06 U.34*
•.4230EKH M.IO*
0.4t56E*Ot M.47*
0.34i7E*06 14.77*
o.3i*iE*o6 as-os*
*.3I43E*06 IS. 33*
0.3I36EKK fS.il*
0.*72E*06 H.au
«.2W7E««i K.JH
0.233IEK* M.SO*
0.23J*EK» 16.711
O.I«22EK« tt.171
O.I73XK* 87.03*
O.I761EK« (7.131
•.ITSKHK 17.34*
•.I7IIEK6 «7.43*
O.I620EK* 17.631
0.tS57E«06 a7.77*
•. 1430£*06 17. 30*

O.I3K6tO( M.IS*
O.I34IEK* M.27*
0. 13I2EKK aa.33*
O.I303EKX M.JO*
O.I2S4E*06 M.73*
•.I25IEKK M.84*
•.I23IEKK M.951
O.I23IE*06 *3.06(
O.II3IEKK «3.I6«
O.IOOXKI3
0. 113KKI3


-------
  . s
 •;!•
mmiiiiismisiiiiiiiiiiiiiiHiiiiiiiiHiffili
loSlslillslsslsslssslsslSlssiSoolilSoooooooolodoooiod
                                                : p X.


     liililil  111 !!!!!!!!!!!!!!!!!!!!
CTO
U «
t-'o
 t-
 I
 1
 1

     liliiiiiglllllllillllll§l§§§§§§r
     rfdrf«i<5«*^^<><><*«i<*
-------
Table 5.  Total Annual Cost for  Equipment
Leak Control Case A (Excluding Recyclers)
I dill WmtMl Udftl

M Facility
1
2
3
4
5
t
7
a
3
10
11
12
13
14
IS
IS
17
ia
13
20
21
22
23
24
25
It
27
2a
23
30
31
32
33
34
35
*
37
38
33
44
41
42
43
44
45
46
47
49
43
50
Total Top 50
Tc.tal All Fac
1 Distribution
i

LTft
O.OOOOEKO
O.OOOOftOO
O.OOOOEKO
O.OOOOEKIO
O.OOOOEtOO
O.OOOCCtOO
O.OOOOEKO
O.OOOOEKO
O.OOOOEKIO
O.OOOOEKO
O.COOOEKO
O.OOOOEKO
0.0000£tOO
O.OOOOEKO
O.CCOOEKO
O.OOOCCKIO
O.OOOOEKO
O.OOOOEKO
O.COOOEKO
O.OOOOEtOO
O.OOOOEKO
O.COOOEKO
O.OOOOEKO
O.OOOOftOO
O.OOOOEKO
O.OOOOEtOO
O.OOOOEKIO
O.OOOOEKO
O.OOOOEKO
O.OOOOEKO
O.OCOCCKO
O.OOOOEKO
O.OCOCttOO
O.OOOCCtOO
(I.OflOCCKO
O.OOOOEtOO
O.OCOCCKO
O.OOOOEtOO
O.OOOOEKO
O.OOOOEtOO
O.OOOOEKO
O.OOOOEKO
O.OOOOEtOO
O.OOOOEtOO
O.OOOOEKO
O.OOOOftOO
O.OOOOftOO
O.OOOCCtOO
O.OOOOEKIO
O.OOOOEKO
O.OOOOEKO
O.OOOOEtOO
0.00*


INC
O.OOOOEKO
O.OOOOftOO
O.OOOOEKIO
O.OOOOEK10
O.OOOOEKO
O.OOOOEK-0
O.OOOOEKIO
O.OOOOEKO
O.OOOOEKO
O.OOOOEtOO
O.OOOOEKO
O.OOOOEKO
O.OOOOEtOO
O.OOOOEKO
O.OOOOEtOO
O.OCOOEKW
O-OOOCCtOO
O.OOOOEKO
O.OOOOEtOO
O.COOOEKO
O.OOOOEKIO
O.OOOOEKO
O.OOOOEKO
O.COOOEtOO
O.OOOOEKO
O.OOOOEKIO
O.OOOOEKO
O.OOOOEKO
O.OOtOEKO
o.ooooeKo
O.OOOOEtOO
O.OOOOEKO
O.OOOOEK*
O.OOOttKO
O.OOOOEtotl
O.OOOOEtOO
CUOOOEKO
O.OOOCtKO
O.OOOOEKO
O.OOOOEtOO
o.otoettoo
O.OOOOEKO
O.OCOOEKO
O.OOOOEKO
O.OOOOEtOO
O.COOOEKO
O.COOOtKIO
0.flOOOEKKI
O.OCOOEKO
O.OOOCCtOO
O.OOOCCKO
O.OOOOEtOO
0.00*


m
O.OOOOEtOO
O.OOOOftOO
O.OOOCCKO
O.OOOOftOO
O.OOOOEKO
O.OOOOEtOO
O.OOOOEtOO
O.OOOCCKO
O.OOOOEKO
O.OOOCCtOO
O.OOOOEtOO
O.OOOCCKO
0.0000£tOO
O.OOOOEKO
O.OOOOEKIO
O.OOOCCKIO
O-OOOCCtOO
O.OOOOEKO
O.OOOCCKO
O.OOOOEKIO
O.OOOCCKO
O.OOOOEKO
O.OOOOEKO
O.OOOOEKO
O.OOOOEKO
O.OOOCCtOO
O.OOOOEKO
O.OOOOEKO
O.OOOCCKO
O.OOOOEtOO
O.OOOOEKO
O.OOOOEKO
O.OOOOEKO
O.OOOCCtOO
O.OOOCCtOO
O.OOOCCKO
O.OOOOEtOO
O.OOOOEKO
O.OOOOEKO
O.OOOOEKIO
O.OOOCCKO
O.OOOOEKO
O.OOOCCtOO
O.OOOCCtOO
O.OOOOEtotl
O.OOOCCKO
O.OOOOEtOO
O.OOOOftOO
O.OCOCCKIO
O.COOOEKIO
O.OCIClOEtOO
O.COOOEKO
O.OCl*


PT2
O.OOCOEKO
O.OOOOEtOO
O.OOOOEtOO
O.OOOOftOO
O.OCOOEKO
O.OOOOEKO
O.OCOtttOO
O.OOOCCKO
O.OOOCtKO
O.OOOOEKO
O.OOOC€KO
O.OOOCtKO
O.OOOOEKO
O.OOOOEKO
O.OOOCtKlO
O.OOOOEtOO
O.OOOCCKO
O.OOOOEKO
O.OOOCCKO
O.OOOOEKO
O.OOOCttoO
O.OOOttKO
O.OCKKCKO
O.OOOCftOO
O.OOOttKO
O.OOOOEtOO
0.0005EKO
O.OCOOEtOO
O.OOOCtKO
O.OOOCtKO
O.OCOCftOii
O.OOOCtKO
O.OOOOEKO
O.OOOCttOO
O.OOOOfKO
O.OCOOEKO
aooocetco
O.OOOCttOO
6.000CCKO
O.OOOCCKO
O.OOOOEtOO
O.OOOCtKlCi
O.OCOCCtCO
O.OOOOEKIO
O.OCOOEtOO
O.OOOCCtOO
O.OOOOEKO
O.OOWEtOO
O.OOOOEtOO
O.OOOCCKIO
O.OOOOEKO
tl.OOOOEKO
0.00*


PT3
O.OOOOEKO
O.OOOOEKO
O.OCOOEKO
O.OOOCCtOO
O.OOOOEKIO
O.OOOCCtOO
O.OOOCCKO
O.OOOOEtOO
O.OOOCCKIO
O.OOOOEKO
O.OOOCCKO
O.OOOCCKO
O.OOOCCKO
O.OOOOEtOO
O.OdOOEtOO
O.OOOOEtOO
O.OOOOEKO
O.OOOOEtOO
O.CIOOOEKO
O.OOOCCKO
O.CIOOOEKO
0.0500EKO
O.OOOOEKIO
O.OOOOEKO
O.OOOCCKO
O.OOOOEtOO
O.OOOOEKO
O.OCOOEtOO
O.OOOOEKO
O.COOOEtOO
O.OOOCCKO
O.OOOCCtOO
O.OOOCCKIO
O.OOOOEtOO
O.OOOOEKO
o.coooetoo
O.OOOOEKO
O.OOOOEtOO
O.OCOCCK»
O.OOOCCKO
O.OOOCCKIO
O.OOOOEtOO
O.OdOCttOO
Q.COOOEKO
O.OOCOEKO
O.OOOCCKO
O.OOOOEtOO
O-OOOOEtOCI
O.OOOOEKO
O.OOOOEKIO
O.OOOOEtOO
O.OOOOttOO
a.«it
TOP W

PT4
O.OOOCCKO
O.OOOOEKO
O.OOOCCKO
O.OOOCCKIO
O.OOOOEKO
O.OOOCCKO
O.OOCOEKO
O.OOOOEKO
O.OOOCCKIO
O.OOOCCKIO
O.OOOCCKO
O.OOOCCKO
O.OOOOEKO
o.ooocetoo
O.OOOCCKO
O.OOOCCKO
O.OOOCCKO
O.OOOCCKO
O.OOOCCtOO
O.OOOOEtOO
O.OCOOEKO
o.ooooeKo
O.OOOOEKIO
O.OOOCCKO
O.OOOCCKO
O.OOOOEtOO
O.OCOOEKO
O.OOOCCKO
O.OOOOEKO
O.OOOCCKO
O.OCOOEKO
O.OOOCCKIO
O.OOOOEtOO
O.COOOEKO
O.OOOCCKO
O.OOOOEtOO
O.OMCCtM
O.OOOCCKO
O.OOOOEKIO
O.OOOCtKO
O.OOOOftOO
O.OOOCCtOO
O.OOOOEKO
O.OOOOEKO
O.OOOCCKO
O.OOOOEtOO
O.OOOCCKO
O.OOOOEKIO
O.OOOOEtOO
O.OOOCCKIO
O.OOOCCKO
O.OOOCCK'O
Cl.COt
FflCIUTIES *
COST IN
CIS
O.OOOCCKIO
O.OOOOEtOO
•XMOOEttt
O.OOOOEK-0
O.OOOCCKIO
0. (OWE tOO
O.OOOOEtOO
O.OOOOEKIO
O.OOOCCKIO
O.OOOCCtOO
O.OOOCCtOO
O.OOOCCtOO
O.OOOCCtOO
O.OOOOftOO
O.OOOOEKIO
O.OOOC€*W
O.OC-OOfK*
O.OOOOftOO
O.OCrOOEKIO
O.OOOCCtOO
O.OOOCCKO
O.OOOOEKO
O.OOOOEKO
O.OOOOEKO
O.OOOOEtOO
O.OOOCCKO
O.OOOCCKO
O.OOOCCKO
O.OOOCCKO
O.OOOCCtOO
O.OOOCCKIO
O.OOOCCtOO
O.OOOCCKO
O.OOOCCKO
aOOCKCKW
O.OOOCCKO
O.OCIOCCKO
O.OOOCCKO
O.OOOOEKO
O.OCOOEKO
O.OOOCCKIO
O.OOOOEKO
O.OCOOEKO
O.OOOOEKIO
O.OOOCCKIO
O.CKOCCKO
O.OCOOEtOO
O.OOOOEKIO
O.OCiOCCKO
O.OCWEKXI
O.OOOCCKIO
O.OOOCCKlCi
o.oo*
"TlOWItC Fy|itivtt only
Vyr
«S02
O.OOOCCKO
O.OOOOEKO
O.OOOCCKO
O.OOOOEKXI
O.OOOOEtOO
O.OOOCCtOO
O.OOOCCKIO
O.OOOCCKO
O.OOOCCKO
O.OOOCCKO
O.OOOCCKIO
O.OOOOEKO
O.OOOOftOO
O.OOCOftOO
O.OOOCCKO
O.OOOCCtOO
O.OOOCCKIO
O.OCOCCKO
0.0000£t»
O.OCOCCKO
O.OOOCCKO
O.OOOOEKX)
O.OCOOEKIO
O.OCOCCKO
O.OOOOEKO
O.OOOOEKIO
O.OOOOEKIO
O.OOOOEtOO
O.OOOCCKO
O.OOOOfKO
O.OOOCCKO
O.COOOEKO
O.OCOCCKO
O.OOOCCKO
O.OOOCCKO
O.COOOEKO
O.OOOOEKO
O.OOOCCKO
O.OOOCCKIO
O.OOOOftOO
O.OCPOOEKO
O.OOOCCtOO
O.OCOCCKO
O.OCOOEKO
O.OOOCttOO
O.OCOOEKO
O.OOOOEKO
O.OOOCCKO
O.OOOOEKO
O.OOOCCKlCi
O.OOOOEKIO
O.OOOOEtOO
o.«*

FJP
O.COOOEKO
O.OOOOEKO
O.OCOCCKO
O.OOOOEtOO
O-OOOOftOO
O.OOOCCtOO
cuooooftoo
O.OOOOEKIO
O.OOOCCKO
O.OOCOEKO
O.OOOCCKO
O.OOOOEKO
O.OOOCCKO
O.OCOOftOO
0.0000£tOO
o.ooocetoo
O.OOOOEtOO
O.OOOOftOO
O.OOOOEKO
0/OOCOEKO
O.OOOCCKO
O.OCOCCKO
O.OOOOEtOO
O.OOOOEKO
O.OOOCCKO
O.OCOCCKO
O.OOCOEKO
O.OOOCCtOO
O.OOOOftOO
O.OOOCCKO
O.OOOCCKO
O.OOOOftOO
O.OOOOEKO
O.OOOCCKO
O.OOOCCKO
O.OOOOEKO
O.OOOCCKO
O.OCOCCKO
O.OOOCCKIO
O.OOOCCKO
O.OCOCCKO
O.OOOOEtOO
O.OOOCCKO
o.ococetoo
O.OOdCCKO
O.OCOOEKIO
O.OOOOEtOO
O.OOOCCtOO
O.OOOOEKO
O.OOOOEKIO
O.OOOCCKO
O.OOOCCKO
0.00*

IM
O.OOOOEKO
O-OOOOftOO
O.OOOCCKO
0.343JEK-5
(uooooftoo
d-OOOOEtOO
O.OOOOEKO
O.OOdCCtOO
O.OOOCCtOO
0.394*Kt
0.3734EtOf,
O.OOOOEKIO
O.OOOCCKO
O.OMCCtOO
0.7404EK8
O.OOOOEKIO
».OOOCCK>0
O.OOOOEKO
O.OOOOEKO
O.OOOOEKO
O.OOOCCKO
O.OCOOEKO
O-IIOTEK*
0.1055EK*
O.OOOOEKO
O.OOOCCK»
O.OCOOEKO
O.OOOOEKIO
0.4UOEt05
o.ooooetoo
0.7J05EK*
O.OOOOEKO
O.OOOCCKO
O.OOOCCKO
4a7(.
o.sa*nT£KS
O.OOOCCKO
O.OOOOEtOO
O.OOOOEtOO
O.OOCOEKIO
0.5227EKI5
O.OOOOEKO
O.OOCOEKO
O.OOOOEKO
20U.
O.OOOCCtOO
O.OOOCCtOO
O.OOOCCtOO
O.COOOEKO
O.OOOOEKIO
0. 1337EK>7
O.I723EKI7
5. 3d*

HI
O.OOOCCtOO
O.OOOOEK>0
O.OOC«tOO
O.OOOCtftOO
O.OCOOEtOO
O.OOOOEK10
O.OOOCCtOO
O.OOOOftOO
CkOOOCCtOO
t.OOOOCtOO
O.OOOCCKIO
O-OOOOftOO
O.OOOCCKIO
O.OOOOEtOO
O.OOOOEKIO
O.OOOOEKIO
O.OOOCCtOO
O.OOOOEtOO
O.OOOCCKIO
0.0000£tOO
IXOOOOEtOO
O.OCOCCKO
O.OCOCCKO
O.OOOCCKO
O.COOOEKO
O.OOOOEKX)
O.OCOOEtOO
O.OOOOEtOO
O.OOOOftOO
O.OOOOEKO
O.OOOOEKO
O.OOOOEtOO
O.OCOCCKIO
O.OOOOEKO
O.OCOCCKO
O.OCOOEKO
O.OOOCCKO
O.OOOOEKO
O.OOOOfKO
O.OOOCCtOO
O.OCOCCKO
O.OOOCCtOO
O.OOOCCKIO
O.OOCOEtOO
O.OCOOEKIO
O.OOOCCtOO
O.OOOCCKO
O.OOOCCtOO
O.OOOCCKIO
O.OCOOEKIO
O.OOOOEKIO
O.OOOCCKO
o.w<

m
O.OOOCCKO
O.OOOOEKIO
O.OOOOftOO
o.ooocetoo
O.OOOOftOO
aooocctoo
O.OOOCCKIO
O.OOOCCtO»
aooooctoo
O.OCOCCKIO
O.OOOOftOO
CuOOOOftOO
O.OOOCCKO
O.OCIOCCKO
O.OOOCCKO
O.OOOOEtOO
aooocetoo
O.OOOOEtOCi
O.OOOCCtOO
O.OOOCCKIO
O.OCOCCKIO
o.ooocetoo
O.OWOEKW
O.OOOCCKO
O.OOOCCK)0
O.OOOOEtOO'
O.OOOCCtOO
O.OOOCCtOO
O.OOOCCKIO
O.OOOCCtOO
O.OOOCCKIO
O.OOOOEtOO
O.OCOOEtOO
O.OOOOEKO
O.OOOOEKIO
O.OOOCCtOO
O.OOOOEKO
O.OOOOEtOO
O.OOOCCKIO
O.OOOCCtOO
o.ococetoo
O.OOOCCKO
O.OOOCCKO
o.ooocetoo
O.OOOCCKlO
O.OOOCCKO
O.OCOCttOO
O.OOOOEtOO
O.OOOCCKIO
O.OOaOEKM
O.OOOOEKIO
O.OOOCCKW
0.00*
Mil ,

0.744(£t07 t5.4M
0.3K7E«07 34.SW
0.tiaX«07 44.72*
ft.ia33EKl7 S2.SW
O.HCSE«07 5*. 31*
0.-44SEKI7 O.IU

OlSOTXtOt U.33t
0.4258EK« H.7H
0.19ME40& 71. IK
0.37S4EtOl 72. 4W
0.3300EKH 7X5SJ
0.]OJ7EK« 74.41*
0.**39«EK« 75. 42*
0.1323EK» 74.0**
0.|8llEt04 74.4S*
O.I793EKK 77.31*
O.I7a7E*OS 77.SH
O.I7MEK* TIL 52*
O.I750EK* 73. Kt
0.1337Etor> 79.57*
0.1330EKI& a0.03<
0.1I07EK« ao.40*
0.10S*£K« aO.76*
0.9«SOEtOS ai.10*
0.9837EK*5 ai.43*
o.33aaEt05 ai.75*
0.9220EKI5 K.m
0.30aCCt05 B2.3a*
0.ai03Et«s 82. M*
0.7305Et05 82.30*
0.72I1EKI5 83,15*
0.7033EK6 83.33*
0.6383EK*5 83.6|t
O.M80EKI5 13.82*
0.5837EKI5 84.02*
0.58I7EK8 84. SJ*
0.5S20fK>5 84.41*
0.542t£KI5 84.S3*
0.5257EK6 84.77<
O.J227EKI5 84.35*
0.9073EtOS 8!. 12*
0.5002ft05 K.29*
0.4944Et05 as. 44*
0.4MitEKt5 85.63*
0.47MEKS B5. 73*
0.43J6EKI5 85.341
0.423XK6 14. C8(
0.4040EtOS 94.22*
0.37ltEKIS 84.35*
0 253IEtOJ
0.2332EKi8


-------
ATTACHMENT 1

-------
                                 ATTACHMENT 1

            PROCEDURE USED BY THE SOURCE ASSESSMENT MODEL  (SAM) TO
                    ESTIMATE EMISSIONS FROM EQUIPMENT LEAKS

     The procedure used 1n the SAM to estimate emissions from equipment leaks at
TSDF sources 1s summarized 1n this attachment 1n two sections.  The first
section describes the derivation of model units and emission factors  (called
"partition fractions" as used in SAM).  The second section provides a narrative
summary of how the SAM applies these factors to estimate nationwide impacts at
all TSDF.

1.0  Model Units and Emission Factors

     A large model unit (Model Unit A) was defined based on the model  unit
described 1n Reference 1 with 5 pumps, 165 valves, 9 sampling connections, 44
open-ended lines, and 3 pressure-relief devices.  The emission factors for each
source and the number of components were used to estimate the uncontrolled
emissions given in Table 1.  Two smaller model units were defined to represent
equipment items used in waste management processes that have fewer components
than represented by Model  unit A.  For example,  a typical storage tank may have
only one pump and is unlikely to have 5 pumps.  Model unit B was defined as 60
percent of Model unit A (for example, 3 pumps and 99 valves).  Model unit A was
defined as 20 percent of Model unit A (for example, 1 pump and 33 valves).
Uncontrolled emissions for all three model units are given in Table 2.
     There are several different waste management processes at TSDF, and the
number of equipment components varies based on the type of process and waste
throughput.  In Table 3, each waste management process is assigned to  one of the
three model units based on data from the 1986 Screener Survey3 and the quantity
of waste processed.  For each process, an emission factor (partition fraction)
is calculated from the model  unit emissions divided by the weighted average
throughput of waste for the process.   The results are summarized in Table 3 for
light liquids.

-------
                   TABLE 1.  NUMBER OF COMPONENTS AND EMISSIONS FOR
                                    MODEL UNIT A1-2
Emission factor
Emission Number of (Mg/yr/source)b
bource Components41 Light liquid Heavy liquid
Pump seals 5 0.43 0.19
Valves 165C 0.062 0.002
Sampling
connections 9 0.13 0.13
Open-ended
lines 44d 0.015 0.015
Pressure-relief
devices 3 0.91 0.91
Total
Uncontrolled
emissions (Mq/yr)
Light liquid
2.15
10.23
1.17
0.66
2.73
16.94
Heavy liquid
0.95
0.33
1.17
0.66
2.73
5.84
aDerived from Reference 1:  "RCRA TSDF Air Emissions-Background Technical
 Memoranda for Proposed Standards."  EPA-450/3-86-009.  Attachment 7.
 October 1986.

bFrom Reference 2:  "Fugitive Emission Sources of Organic Compounds-Additional
 Information on Emissions, Emission Reductions, and Costs."  EPA-450/3-82-010.
 April 1982.  p. 2-70.

cFrom Reference 1, includes 87 (liquid) + 34 (gas) + 9 (sampling connections)
 + 35 (open-ended lines).

dFrom Reference 1, includes 35 (open-ended lines) + 9 (sampling connections).

-------
             TABLE 2.  EMISSION ESTIMATES FOR EACH
                          MODEL UNIT
Model Unit
A*
Bb
Cc
Uncontrolled
Light liquid
16.94
10.16
3.39
emissions (Mg/yr)
Heavy liquid
5.84
3.50
1.17
aFrom the total in Table 1.

bModel unit B is defined as 60 percent of Model Unit A.

cModel unit C is defined as 20 percent of Model Unit A.

-------
         TABLE 3.  PARTITION FRACTIONS FOR EACH WASTE MANAGEMENT
                        PROCESS FOR LIGHT LIQUIDS
Management
Process
Uncontrolled
emissions
(Mg/yr)*
Weighted Average
Throughput
(Mg/yr)b
Partition
fraction0
Model Unit A:
Incinerator
Injection Well

VO removal devices;
 Steam stripper
 Distillation
 Incinerator
Model Unit B;
Treatment tank:
3xxd
4xx
5xx
Ixx
          16.94
          16.94
16.94
16.94
16.94
16.94
                   10,035
                  189,243
121,951
 17,299
 26,918
 74,026
                  0.00169
                  0.000090
0.000139
0.000979
0.000629
0.000229
quiescent
aerated
Vo removal devices:
Air stripper 3xx
Thin film
evaporator 2xx
Model Unit C:
Storage tank
Surface impoundment
Terminal loading
Drum loading
10.16
10.16

10.16
10.16
3.39
3.39
3.39
3.39
154,466
785,632
1 1
326,395
17,640
3,350
378,622
4,800
1,200
0.000066
0.000013

0.000031
0.000576
0.001012
0.000009
0.000706
0.002825
 aFrom Table 2.

 bQuantity of waste processed calculated as a weighted average from the
  data for all waste processes 1n SAM.

 Uncontrolled emissions divided by throughput.

 dThese are codes for waste forms, where Ixx = inorganic solid,
  2xx = aqueous sludge, 3xx = aqueous liquid, 4xx = organic liquid,
  5xx = organic sludge, and 6xx = miscellaneous.

-------
       Only one set of partition fractions  can be read  Into the SAM for each model
  unit to estimate emissions.   The partition  fractions  1n  Table 3  are for light
  liquids;  consequently an  adjusting  factor was derived to use  for heavy liquids.
  From Table 1,  the ratio of emissions  of heavy and  light  liquids  is  0.34
  (5.84/16.94).   Therefore,  when SAM  determines that a  waste  stream is  a heavy
  liquid,  the partition fractions  for light liquids are multiplied by 0.34  to
  estimate  heavy liquid emissions.  For a given waste stream, SAM  first determines
  the  total  organic content.  Waste streams with  less than  10 percent organics  are
  not  affected by  the  proposed  standard for equipment leaks.  However,  waste
  streams with over 10 percent  total  organics will be subject to the  equipment
  leak  controls.   The  model divides those waste  streams with over  10  percent
  organics  into  light  and heavy  liquids.  Waste  streams that contain  20  percent or
  more  by volume of  high volatility compounds are classified as light liquids.  In
  SAM,  the  high volatility compounds  are defined as those with a vapor pressure
  greater than 1.33  kPa or a Henry's  law constant greater than o'.l kPa m3/g mol.
  Wastes with over  10  percent total organics but less than 20 percent high
  volatility organics  are classified as heavy liquids.   For wastes that are
  identified as heavy  liquids,  the model adjusts the emissions by the factor of
 0.34 as previously described.
      Control techniques and their efficiencies are given  in Table 4.  The
 control efficiencies and  uncontrolled  emission estimates  are used in Table 5 to
 estimate controlled emissions  for two  control  options. Option 1  includes
 monthly leak detection and repair for  pumps  and valves and equipment controls
 for the other sources. Option 2 substitutes dual  mechanical  seals for leak
 detection  and repair for pumps.  The overall  emission  reductions  are given in
 Table 5 for Model  unit A and were used in  the  SAM  to estimate  the overall
 controlled emissions  for the two  options.

 2.0   SOURCE ASSESSMENT MODEL (SAM)

     Details on the structure of SAM are given- in Reference 6 and are summarized
 in this section.  The SAM is a  computer model used to estimate total nationwide
 impacts from emissions at TSDF.  The model  contains information on over 2,300
TSDF nationwide -and the various waste management processes used at each
facility.   In addition, the model contains  waste concentration data for the
various types of hazardous  wastes processed at these facilities.   The waste

-------
               TABLE 4.  CONTROLS AND THEIR EFFICIENCES FOR
                            EQUIPMENT LEAKS4'5
Control efficiency (percent)
Source
Pump seals

Valves
Sampling
connections
Open-ended
lines
Pressure-relief
devices
Control
Leak detection and repair3
Dual mechanical seals&
Leak detection and repair
Closed-purge sampling
Caps
Rupture disks
Light liquid
61
100
59
100
100
100
Heavy liquid
0
100
0
100
100
100
alncludes leak detection and repair with monthly inspections.

bDual mechanical seals with barrier fluid and degassing reservoir vent.

-------
            TABLE 5.  ESTIMATES OF CONTROLLED EMISSIONS FOR
                              MODEL UNIT Aa
Source
Pump seals
Valves
Sampling
connections
Open-ended
lines
Light
Hguid emissions (Mg/yr)
Option 1& Option 2C
0.839 0
4.19 4.19
0 0
0 0
Heavy
liguid emissions (Mg/yr)
Option 1 Option 2
0.95 0
0.33 0.33
0 0
0 0
 Pressure-relief
 devices             	0

          Total      5.03

Overall percent
 reduction          70.3
 	0

 4.19


75.3
 	0

 1.28


78.1
 	0

 0.33


94.3
 aEstimated from the uncontrolled emissions in Table 1 and the
  control  efficiencies in Table 4.

 bOption 1 includes monthly leak detection and repair for pumps and
  valves,  closed-purge sampling, caps for open-ended lines,  and
  rupture  disks for pressure-relief devices.

 C0ption 2 includes dual  mechanical  seals for pumps and controls
  for other sources as listed for Option 1.

-------
concentration data are expressed in terms of surrogate compounds classified in
terms of volatility (high, medium,  and low).  For each facility, the SAM models
emissions from various waste management processes as the waste stream travels
through different units at the facility.  The model  also adjusts (reduces)  the
organic concentration of the waste stream that enters a given process to account
for losses of orgam'cs by emissions from a previous  processing step.
     For equipment leaks, the model first determines the concentration of total
orgam'cs in a given waste stream at a particular facility.    If the
concentration is greater than 10 percent total orgam'cs, the waste stream is
identified as a candidate for control of equipment leaks.  The waste streams
with over 10 percent total orgam'cs are then identified as  either light liquids
(20 percent or more high volatility compounds) or heavy liquids (less than  20
percent high volatility compounds).  The model then  determines which waste
management processes are used for this waste stream  at the  given facility from
the model's industry profile data base.  The industry profile was developed from
TSDF survey data and contains a record of how each waste stream at each facility
is processed.  For each waste stream and facility, data are provided for storage
(in containers, tanks, waste piles, and surface impoundments) treatment (in
tanks, impoundments, incinerators,  and other) and disposal  (injection well,
landfill, land application, and impoundment).  If the waste stream is processed
in different waste management units, the model assigns a logical processing
sequence, such as storage followed by treatment and  disposal.  Terminal loading
is assigned to those facilities that are most likely to ship waste offsite.  The
facilities with terminal loading are identified as those that do not have a
disposal process listed for the waste stream and the last process used for the
waste stream is not incineration or placement in a waste pile.
     The partition fraction for the waste management process is chosen from
Table 3, and emissions are estimated from the partition fraction times the
quantity of waste processed (from the 1986 National  Screener Survey of TSDR,
which is 1n the SAM database for the given facility) times  the adjusting factor
for heavy liquids, if appropriate.  This calculation is performed for each
process at each facility, and emissions from each facility are summed to
estimate total nationwide emissions.  Controlled emissions  are estimated by
applying the overall efficiencies given in Table 5.
                                       8

-------
3.0  REFERENCES


1.  U.S. EPA.  RCRA TSDF Air Emissions-Background Technical Memoranda for
    Proposed Standards.  EPA-450/3-86-009.  Attachment 7.  October 1986.

2.  U.S. EPA.  Fugitive Emission Sources of Organic Compounds-Additional
    Information on Emissions, Emission Reductions, and Costs.  EPA-450/3-82-010.
    April 1982.  p. 2-70.

3.  Office of Solid Waste.  National Screening Survey of Hazardous Waste
    Treatment, Storage, Disposal, and Recycling Facilities.  U.S. EPA.
    Washington, DC.  June 1987.

4.  U.S. EPA.  RCRA TSDF Air Emissions-Background Technical Memoranda for
    Proposed Standards.  EPA-450/3-86-009.  Attachment 7.  October 1986.

5.  U.S. EPA.  Fugitive Emission Sources of Organic Compounds-Additional
    Information on Emissions, Emission Reductions, and Costs.  EPA-450/3-82-010.
    April 1982, p. 4-43, 4-55, 4-56.

6.  U.S. EPA.  Hazardous Waste TSDF - Background Information for proposed RCRA
    Air Emission Standards (Draft).  November 1988.  Appendix D.

-------
ATTACHMENT 2

-------
                                 ATTACHMENT 2
                    SUMMARY OF PROCEDURE DEVELOPED FOR THE
                   SOURCE ASSESSMENT MODEL (SAM) TO ESTIMATE
                       CONTROL COSTS FOR EQUIPMENT LEAKS

     The steps used to develop cost functions for control  of equipment leaks
are described below.  The descriptions are keyed to the tables,  which
provide the cost details and the results.   Two control  options were
investigated.  The first one is described  in  Steps 1  through 6 and includes
leak detection and repair for pumps and valves,  closed-purge sampling for
sampling connections,  caps for open-ended  lines,  and  rupture disks for
pressure-relief valves.   The second control option (described in Steps 7
through 10)  is similar except that  dual  mechanical  seals with a barrier
fluid are applied to pumps.
                       C?uts were  taken from Reference 1 and are summarized
           mn           ?* fntr01? 1nclude leak detfiction and repair for
           pump  seals  and  valves,  closed-purge sampling, rupture disks for
           pressure  relief devices, and caps for open-ended lines.  These
           costs were  updated to 1986 dollars and Spply to equipment in
           from  t Iq±,,S?-V1$e'   Jhe.annual operating cost wascaLlated
           PwQfi? t?    llzed C9J* »1nus the capital recovery factor
           ,rfH 2 I tlm?s Jhe caPltal cost-  (Capital recovery costs will  be
           added back  in by SAM to estimate total annual ized costs.)
                             Was def1ned for TSDR based on the model  unit
              l    t         nCV-  The nui"ber of components for TSDF
          Tab?] 1 hl^H n 9Jhe" 1"Jabl! 2'  The costs were scaled from
          i able l based on the number of components.
     3.   For equipment in heavy liquid service,  the  leak detection and
          coStrn^r"" ^ mt aPPl1"ble-  However,  the equipment
          Snnfr  M(   r!d Purge samPl1n9r  rupture disk,  caps)  are
          tn Tab?e 3.'      C0ntr   C°StS f°r heavy 11qu1ds are
          Three model  units  were  defined  for  TSDF sources.  Model Unit A is
          MnHp innftT-  Wlt|j.the ?umber  of components defined in Table 2.
          cos?   of inSpi18.."??1!!"  SiZ? and 1S  6° percent  (1n components and
          cost)  of Model  Unit A.   Model Unit  C is a small unit and is
          Jhr^'ln5!20 PrC6nt °f Model  Un1t A«  The contro1 ^sts for the
          three  model  units  are summarized in Table 4

-------
                TABLE 1.  COST DATA FOR CONTROL OF EQUIPMENT LEAKS -
                               LIGHT LIQUID SERVICE3^
Source Number^
Pump seals 5
Valves 121
Sampling
connections 9
Open-ended
lines 35
Pressure-
relief valves 3
Monitoring
device 1
Annual
TCP TCIC costb AOCd AOCC
(1985$) (1986$) (1985$/yr) (1985$/yr) (1986 $/yr)
292 292 1,787 1,720 1,722
336 336 2,378 2,301 2,304
6,146 6,154 1,530 119 119
2,338 2,341 560 23.2 23.2
11,541 11,557 3,280 630 631
6,309 6,318 2,366 917 918
TCI = Total capital investment.
AOC = Annual operating cost.
aControls include leak detection and repair for pumps and valves; equipment
 controls for other sources.
bTaken from Reference 1:  "RCRA TSDF Air Emissions-Background Technical
 Memoranda for Proposed Standards."  EPA-450/3-86-009.  October 1986.
 Attachment 7.
cBased on a cost index of 1.0013 for 4th quarter 1985 to January 1986.
dAOC = Annual Cost - (.2296) (TCI) where .2296 is the capital recovery factor
 based on a 6-yr life and 10 percent interest.

-------
                      TABLE 2.  COST DATA FOR TSDF
                      MODEL UNIT A - EQUIPMENT LEAK
                   CONTROLS FOR LIGHT LIQUID SERVICE3
Source
Pump seals0
Valves
Sampling
connections
Open-ended
lines
Number
5
165d
9
44f
Pressure-
relief
valves 3
Total
Monitoring
device
1
TCI
(1986$)
292
458e
6,154
2,9439
11,557
21,404
6,318
AOC
(1985$/yr)
1,722
3,1426
119
299
631
5,643
918
TACb
(1986$/yr)
1,789
3,247
1,532
705
3,284
10,557
2,369
TCI = Total capital investment.
AOC = Annual operating cost.
TAC = Total annualized cost.
aBased on leak detection and repair for pump seals and valves,
 equipment controls for the other sources.  For equipment in
 light liquid service.
ABased on an equipment life of 6 years and 10 percent interest
 with a capital recovery factor of 0.2296.
cFrom Table 1 (1986 dollars).
dFrom Reference 1, includes 87 (liquid) + 34 (gas) + 9
 (sampling connections) + 35 (open-ended lines).
eScaled up from Table 1 by 165/121.
fFrom Reference 1, includes 35 (open-ended lines)  + 9
 (sampling connections).
9Scaled up from Table 1 by 44/35.

-------
                       TABLE 3.  COST DATA FOR TSDF
                       MODEL UNIT A - EQUIPMENT LEAK
                    CONTROLS FOR HEAVY LIQUID SERVICE3
Source Number^
Sampling
connections 9
Open-ended
lines 44
Pressure-
relief
valves 3
Total
TCI
(1986$)b
6,154
2,943
11,557
20,654
AOC
(1986$/yr)b
119
29
631
779
TAC
(1986$/yr)b
1,532
705
3,284
5,521
TCI = Total Capital investment.

AOC = Annual operating cost.

TAC = Total annualized cost.

aControls include equipment items for the sources.  Leak detection
 and repair program is not applied to heavy liquids.

bFrom Table 2.

-------
                  TABLE 4.  SUMMARY OF EQUIPMENT LEAK CONTROL COSTS
                                FOR TSDF MODEL UNITS3
                     TCI (1986$)
AOC (1986$/yr)
TAG (1986$/yr)
Model unit
Light liquid:
AD
Bc
C
Heavy liquid:
Ae
Bc
Cd
Monitor

6,318
6,318
6,318

—
—
— -B
Equipment

21,404
12,842
4,281

20,654
12,392
4,131
Monitor

918
918
918

—
—
~~
Equipment

5,643
3,386
1,129

779
467
156
Monitor

2,369
2,369
2,369

—
__
— —
Equipment

10,557
6,334
2,111

5,521
3,313
1,104
TCI = Total Capital investment.
AOC = Annual  operating cost.
TAG = Total annualized cost.
alncludes leak detection and repair for pump seals and valves in light-liquid
 service.  Equipment controls for other sources.
bFrom Table 2.
cModel Unit B is 60 percent of Model  Unit A.
dModel Unit C is 20 percent of Model  Unit A.
eFrom Table 3.

-------
 5.
           The model units are assigned to TSDF processes based on the size
           and complexity of the process and auxiliary equipment.  For
           example, numerous pieces of equipment are associated with an
           incinerator; therefore, Model Unit A is used for this process.
           Model Unit C is assigned to a single storage tank because the
           associated equipment count would be small.  The model unit
           assignments by process are given in Table 5.  Also given in Table
           5 is the average annual throughput for the process.  The model
           unit cost divided by the throughput yields the slope (B.B1) of
           the cost function.
     Each facility must purchase a monitoring device, and the other costs
associated with controlling equipment leaks are assumed to be a function of

the process throughput.  The cost functions from Table 5 are used in SAM to
estimate costs, where:


         Total capital investment = 6,318 + B (Quantity, Mg/yr)
         for a facility

         Annual operating cost = 918 + B1 (Quantity, Mg/yr).
         for a facility


The slope (B.B1) times the quantity of waste processed yields an estimate of
costs for each source, and the cost of a monitoring device is added only

once per facility, independent of throughput.  Total nationwide costs are
estimated by summing the costs across each facility.
8.
           The cost functions in Table 5 are for equipment in light liquid
           service.  Because only one set of fugitive cost functions can be
           read into SAM for each model  unit,  an adjusting factor was
           derived for heavy liquids.  From Table 4,  the ratio of capital
           control costs for heavy/light liquids for Model Unit A is
           20,654/21,584 = 0.957.  The cost function for light liquids was
           multiplied by this factor in  SAM to estimate the control costs
           for heavy liquids.  Similarly for annual  operating costs, a
           factor of 779/6871 = 0.113 (from Table 4)  was used for heavy
           liquids.

           The second control option includes  dual mechanical  seals for
           pumps instead of leak detection and repair.   The control for
           other sources remain the same as those in  the first option.  The
           costs for Model  Unit A are summarized in  Table 6 for control  of
           light liquids.

           Table 7 summarizes the costs  for control  of  heavy liquids
           emissions from  pumps,  sampling connections,  open-ended lines, and
           pressure-relief valves for Model  Unit A.

-------
                 TABLE 5.  COST FUNCTIONS FOR EQUIPMENT LEAK CONTROL
                               COSTS - LIGHT LIQUIDS3
Process
Incinerator
Injection well
Quiescent treatment tank
Aerated treatment tank
Storge tank
Storage impoundment
Treatment impoundment
Disposal impoundment
Drum loading
Terminal loading
Model unitb
A
A
B
B
C
C
C
C
C
C
Throughput0
(Mg/yr)
10,000
189,000
154,000
786,000
3,350
5,830
379,000
31,500
1,200
4,800
B for TCld
2.14
.113
.0834
.0163
1.28
.734
.0113
.136
3.57
.892
B1 for AOCe
.564
.0299
.022
.0043
.337
.194
.00298
.0358
.941
.235
TCI = Total Capital investment.
AOC = Annual operating cost.

aThe table shows how the cost function is calculated for use in SAM.   The
 equipment cost for the given model  unit from Table 4 is divided by the
 weighted average annual throughput  to calculate B and B1.

bModel unit assigned to given TSDF process.   Model unit A is large
 (equipment count in Table 2), B is  medium (60 percent of A),  and C
 is small (20 percent of A).

cWeighted average annual throughput  for the  TSDF process from SAM.

dCost function for TCI = 6,318 + B»Q (Mg/yr)  where 6,318 is  the
 cost of a monitoring device, B is given in  the table, and Q is
 the throughput of a specific facility's process in Mg/yr.

eCost function for AOC = 918 + B^O  (Mg/yr)  where 918 is the annual
 operating cost for the monitor, B1  is given in the table, and Q is
 the throughput of a specific facility's process in Mg/yr.

-------
       TABLE 6.  COST DATA FOR TSDF MODEL UNIT A - EQUIPMENT LEAK
         CONTROLS FOR LIGHT LIQUID SERVICE WITH DUAL MECHANICAL
                            SEALS FOR PUMPS3
Source
Pump seals
Valvesc
Sampling
connections0
Open-ended
lines0
Pressure- rel
valves
Monitoring
device0
Number
5
165
9
44
ief
3
Total
1
TCI
(1986$)
48,250b
458
6,154
2,943
11,557
69,362
6,318
AOC
(1986$/yr)
3,820b
3,142
119
29
631
7,741
918
TAC
(1986$/yr)
14,900
3,247
1,532
705
3,284
23,668
2,369
TCI = Total Capital investment.

AOC = Annual operating cost.

TAC = Total annualized cost.

aControls include dual mechanism seals with a barrier fluid and
 degassing vents for pumps, leak detection, and repair for valves,
 closed-purge sampling, rupture disks, and caps for open-ended lines,

bFrom Appendix F in Reference 2:  "Benzene Emissions from Coke
 By-product Recovery Plants - BID."  EPA-450/3-83-016a.  May 1984.
 Scaled from 1979 dollars by f = 1.487.  Capital  recovery factor =
 0.2296 (6 yr. life at 10 percent interest).

cThese costs were taken from Table 2.
                                 8

-------
         TABLE 7.  COST DATA FOR TSDF MODEL UNIT A -  WITH DUAL
           MECHANICAL SEALS FOR PUMPS - HEAVY LIQUID SERVICE3
Source Number
Pumpsb 5
Sampling
connections0 9
Open-ended
lines0 44
Pressure- relief
valves0 3
Total
TCI
(1986$)
48,250
6,154
2,943
11,557
68,904
AOC
(1986$/yr)
3,820
119
29
631
4,599
TAC
(1986$/yr)
14,900
1,532
705
3,284
20,421
TCI = Total Capital investment.
AOC = Annual operating cost.
TAC = Total annualized cost.
Controls include equipment items for the sources,
bFrom Table 6.
°From Table 3.

-------
      9.    The  costs  for the  large model  unit  in Table  6  and  7 were  used  to
           estimate the  cost  for  the  two  smaller units  (at  60 and  20 percent
           of Model Unit A).   The costs for all three model units  are given
           in Table 8 for the application of dual  mechanical  seals to pumps.

      10.   The  cost functions for controls that include dual  mechanical
           seals for  pumps are given  in Table  9.   These cost  functions are
           used in SAM as described  in Step 5  to estimate the control costs
           for  this option.
REFERENCES
      1.   USEPA.  RCRA TSDF Air Standards - Background Technical Memoranda
           For Proposed Standards.  EPA-450/3-86-009.  October 1986.

      2.   USEPA.  Benzene Emissions from Coke By-product Recovery Plants-
           Background Information Document.  EPA-450/3-83-016a.  May 1984.
           p. F-5.
                                       10

-------
                  TABLE 8.  SUMMARY OF EQUIPMENT LEAK CONTROL COSTS
                        WITH DUAL MECHANICAL SEALS FOR PUMPS3
                    TCI (1986$)
AOC (1986$/yr)
TAG (1986$/yr)
Model unit
Light liquid:
A"
Bc
Cd
Heavy liquid:
Ae
Bc
°
Monitor

6,318
6,318
6,318

—
—

Equipment

69,362
41,617
13,872

68,904
41,342
13,781
Monitor

918
918
918

__

""
Equipment

7,741
4,645
1,548

4,599
2,759
920
Monitor

2,369
2,369
2,369

—
—
™ ™
Equipment

23,668
14,200
4,734

20,421
12,253
4,084
TCI = Total Capital investment.
AOC = Annual operating cost.
TAC = Total annualized cost.
alncluded dual mechanism seals for pumps, leak detection and repair for
 valves, and equipment controls for the other sources.
bFrom Table 6.
cModel Unit B is 60 percent of Model Unit A.
dModel Unit C is 20 percent of Model Unit A.
eFrom Table 7.
                                         11

-------
               TABLE 9.  COST FUNCTIONS FOR EQUIPMENT LEAK CONTROL
            COSTS (DUAL MECHANICAL SEALS FOR PUMPS) - LIGHT LIQUIDS3
Process Model
Incinerator
Injection well
Quiescent treatment tank
Aerated treatment tank
Storge tank
Storage impoundment
Treatment impoundment
Di sposal i mpoundment
Drum loading
Terminal loading
unitb
A
A
B
B
C
C
C
C
C
C
Throughput0
(Mg/yr)
10,000
189,000
154,000
786,000
3,350
5,830
379,000
31,500
1,200
4,800
B for TCId
6.94
.367
.269
.053
4.14
2.38
.0366
.440
11.6
2.89
B1 for AOCe
.774
.041
.030
.0059
.462
.266
.0041
.049
1.29
.323
TCI = Total Capital Investment.
AOC = Annual operating cost.

aThe table shows how the cost function is calculated for use in SAM.  The
 equipment cost for the given model unit from Table 8 is divided by the
 weighted average annual throughput to calculate B and B1.

bModel unit assigned to given TSDF process.  Model unit A is large
 (equipment count in Table 2), B is medium (60 percent of A),  and C
 is small (20 percent of A).

cWeighted average annual throughput for the TSDF process from SAM.

dCost function for TCI = 6,318 + B»Q (Mg/yr)  where 6,318 is the
 cost of a monitoring device, B is given in the table, and Q is
 the throughput of a specific facility's process in Mg/yr.

eCost function for AOC = 918 + B^Q (Mg/yr) where 918 is the annual
 operating cost for the monitor, B1 is given in the table, and Q is
 the throughput of a specific facility's process in Mg/yr.
                                     12

-------
RESEARCH  TRIANGLE   INSTITUTE

^            •* PSSTriJ
Center for Environmental Systems       «p ft j£ |               September 30, 1988


       MEMORANDUM

       TO:      R1ck Colyer, EPA/SDB, and Robert Lucas, EPA/CPB

       FROM:    Steve York and Robert Zerbonla,  RTI

       SUBJECT: Estimation of Nationwide Uncontrolled Emissions from Solvent
                Recycling Facilities for the Accelerated Rule Post Proposal Analysis


           The purpose  of this memo is to present estimates of nationwide
       uncontrolled organic emissions from solvent recycling facilities (WSTF and
       facilities that  reuse solvent as fuel  but do not have on-site solvent recovery
       operations).  The estimates are based on  facility and waste throughput
       information summarized in the industry profile1 and model  unit parameters
       presented in the RTI memorandum recommending WSTF/TSDF model  unit parameters.2

           Table 1 presents lower bound and upper bound estimates of nationwide
       annual  uncontrolled process vent emissions from WSTF.  The nationwide total  of
       448 facilities represents the number of facilities reporting  solvent recovery
       operations, such as batch distillation or fractional on,  involving some form
       of waste at the  facility 1n the 1986 National Screening Survey of Hazardous
       Waste Treatment, Storage, Disposal  and Recycling Facilities.   Though this may
       be an overestimate of the number of WSTF  (435 of these facilities did report
       recycling solvents or halogenated organics in their solvent recovery
       operations, 13 did not) it is the best available estimate  of  the number of
       facilities with  process vents subject  to  the proposed rule.  Of the 448
       facilities reporting solvent recovery  operations,  365 reported the total
       quantity of waste (all  types, including waste categories other than solvents)
       recycled in 1985.

           This information was used to produce  the frequency distribution of waste
       throughput presented in Table 1.0 of Reference 1.   The midpoint or average of
       each throughput  range is used as the 1985 throughput in Table 1.  The
       estimated distribution of number of facilities by 1985 throughput in Table 1
       was generated using the frequency distribution (proportion in internal, %)
       from Table 1.0.

           Lower bound  and upper bound uncontrolled process vent  emissions were
       estimated by applying the minimum and  maximum VOC emission factors (gVO/kgTO)
       for the small, medium and large draft  model  units presented in Table 1 of
       Reference 2, assuming 7 Ib/gal of throughput.  Small model unit emission
       factors were applied to facilities  with throughputs less than 50,000 gallons
       per year; medium model  unit emission factors in the throughput range from
       50,000  to less than 1,000,000  gallons per year;  and large model unit emission
       factors in the throughput range greater than or equal to 1,000,000 gallons per
      year.
Post Office Box 12194        Research Triangle Park, North Carolina 27709        Telephone: 919 541-6000

-------
    Table 2 presents estimates of nationwide annual  uncontrolled equipment
leak emissions from solvent recycling facilities.  The model  units are
presented in Table 2 of Reference 2 and are similar to those  used for the
equipment leak analysis in the Source Assessment Model and those used for the
benzene fugitive emission model units.  SOCMI VOC emissions factors were used
to estimate organic emissions for each source; operating hours were assumed to
be 8,760 hr/yr.  The total numbers of small, medium and large facilities were
derived from the industry profile using the throughput ranges described in the
previous paragraph.  Table 3 presents the estimates of nationwide annual
uncontrolled equipment leak emissions from TSDF reusing waste solvent as fuel
with no solvent recovery operations on site.  These fuel reuse operations are
characterized as fuel blending units.  It should be noted that the Source
Assessment Model (SAM) will be used to estimate nationwide annual uncontrolled
and controlled equipment leak emissions associated with other TSDF management
processes handling hazardous wastes with an organic content of greater than
10% total organics.  These TSDF management processes would include drum
handling, storage tasks, treatment tanks, incinerators, impoundments,
stabilization processes, and so on.

    Table 4 summarizes the lower bound and upper bound estimates of nationwide
annual uncontrolled organic emissions from WSTF.  The estimated range of total
nationwide organic emissions compares quite closely with the  range estimated
for the proposed rule (a range of 8,130 to 15,930 Mg/yr vs. a range of 2,550
to 14,740 Mg/yr estimated for the proposed rule).  The post proposal estimated
number of WSTF has increased from 95 to 448.  In addition, 108 solvent
recycling facilities have been identified that reuse solvent  as fuel but do
not have on-site solvent recovery operations.  However, the impact on the
uncontrolled emission estimate of increasing the number of facilities has been
offset by the estimated distribution of small, medium and large model units
and the revised emission factors.

    The most noticeable difference between the post proposal  uncontrolled
emission estimate and the proposal estimate is in the increased relative
contribution of equipment leaks emissions to total nationwide emissions from
waste solvent recycling units.  At proposal, equipment leaks  accounted for
between 9 and 51 percent of total uncontrolled emissions from WSTF.  In the
post proposal analysis, equipment leaks account for between 46 and 95 percent
of total uncontrolled emissions from WSTF.  Including the facilities reusing
solvent as fuel, equipment leaks account for between 49 and 96 percent of
uncontrolled emissions from solvent recycling units.


References

1.  Zerbonia, Robert A. (RTI).  Industry Profile-Proposed Air Emissions
    Standards for Volatile Organics Control from Hazardous Waste Treatment,
    Storage, and Disposal Facilities.  Prepared for U.S. Environmental
    Protection Agency.  Research Triangle Park, NC.  August 13, 1987.  33p.

2.  Memorandum and attachments from Zerbonia, Robert and York, Steve, RTI, to
    Colyer, Rick, EPA/SDB. September 28, 1988. 24p.  Model unit parameters
    (Accelerated Rule) post proposal analysis.


SY/RZ/ddf

-------
TABLE 1.  LOWER BOUND AND UPPER BOUND ESTIMATES OF NATIONWIDE ANNUAL
        UNCONTROLLED PROCESS VENT ORGANIC EMISSIONS FROM WSTF
1985 Throughput,
1,000 gallons
.25
.75
3.0
7.5
30
75
300
750
1,500
3,500
7,500
10,000
Nationwide
Total
Number of
Facilities
16
22
50
42
78
58
69
40
23
28
12
10

448
Annual Uncontrolled
Lower Bound
0.03
0.13
1.2
2.5
19
14
66
95
11
31
29
32

300
Process Vent Emissions, Mg Organ ics
Upper Bound
0.38
1.6
14
30
220
410
2,000
2,900
270
780
720
800

8,100

-------
                            TABLE 2.  ESTIMATES OF NATIONWIDE ANNUAL UNCONTROLLED
                         EQUIPMENT LEAK EMISSIONS FROM SOLVENT RECYCLING FACILITIES3
Emission Source
Number of Sources
Large
Model Unit
# 1
Pump Seals
Light Liquid
Compressors
Flanges
Valves
Gas
Liquid
Pressure Relief
Devices
Gas
Liquid
Sampling Connections
Open Ended Lines
Model Unit Totals
Number of WSTF

15
0
0

106
264


9
0
26
105
519
73
Medium
Model Unit
# 2

5
0
0

34
87


3
0
9
35
173
167
Small
Model Unit
1 3

3
0
0

20
52


2
0
5
21
103
208
Annual Emissions, Mq Orqanics/Model Unit
Large Medium Small
11 12 #3

6.6 2.2 1.3
_
_

5 1.7 1.0
16.4 5.4 3.2


8.1 2.7 1.8

3.5 1.2 0.7
1.5 0.5 0.3
41.1 13.7 8.3

Nationwide Total,
   Emissions
   Mg organics/yr

Grand Total, Mg/yr
3000.3      2287.9    1726.4
            7014.6 (-7010 Mg/yr)

-------
            TABLE 3.  ESTIMATES OF NATIONWIDE ANNUAL UNCONTROLLED
           EQUIPMENT LEAK EMISSION FROM TSDF FUEL BLENDING UNITS.
Emission Source

Pumps
Valves
Sampling connections
Open-ended lines
Pressure relief valves
Total number of facilities
Model Unit Annual Emissions
Model
Model Unit
14
5
121
9
35
3
26
13.7
Unit Character!'
(Reuse as Fuel)
Model Unit
#5
3
72
5
21
2
40
8.3
sties
Model Unit
#6
1
24
2
7
1
42
3.1
     (Mg/yr)

Nationwide Total Emissions
     (Mg/yr)
356.2
332
130.2
Grand Total (Mg/yr)
              818.4 (-820 Mg/yr)

-------
    TABLE 4.  LOWER BOUND AND UPPER BOUND ESTIMATES OF NATIONWIDE ANNUAL
      UNCONTROLLED ORGANIC EMISSIONS FROM SOLVENT RECYCLING FACILITIES
                         Nationwide Uncontrolled Organic Emissions, Mq VO/yr
                                Process       Equipment
                                 Vent            Leak          Total
Lower Bound Estimate                 300         7,830         8,130

Upper Bound Estimate               8,100         7,830        15,930

-------
      APPENDIX D



SOURCE ASSESSMENT MODEL

-------
                                APPENDIX D
                          SOURCE ASSESSMENT MODEL

D.I  DESCRIPTION OF MODEL
D.I.I  Overview
     The standard-setting process for hazardous waste transfer, storage,
and disposal facilities  (TSDF) involves identifying the sources of air
pollutants within the industry and evaluating the options available for
controlling them.  The control options are based on different combinations
of technologies and degrees of control .efficiency, and they are typically
investigated in terms of their nationwide environmental, health, economic,
and energy impacts.  Therefore, information and data concerning TSDF
processes, emissions, emission controls, and health risks associated with
TSDF pollutant exposure are being made available for input to the review
and decisionmaking process.
     The Source Assessment Model  (SAM) is a tool that was developed to
generate the data sets necessary for comparison of the various TSDF con-
trol options.  The SAM is a complex computer program that uses a wide
variety of information and data concerning the TSDF industry to calculate
nationwide impacts (environmental,  cost, health, etc.)  through summation
of approximate individual facility results.  It should be pointed out that
the primary objective and intended use of the SAM is to provide reasonable
estimates of TSDF impacts on a national level.  Because of the complexity
of the hazardous waste management industry and the current lack of
detailed information for individual  TSDF,  the SAM was developed to utilize
national  average data where site-specific data are not  available.   As a
result,  the SAM impact estimates  are not considered accurate for an
individual  facility.   However,  on a  nationwide basis,  the SAM impact
estimates are a reasonable approximation and provide the best available
basis  for analysis of options  for controlling TSDF air  emissions.
                                    D-3

-------
D.I.2  Facility Processor
     Information processed by the SAM includes results from recent TSDF
industry surveys, characterizations of the TSDF processes and wastes, as
well as engineering simulations of the relationships among:  (1) waste
management unit type, waste, and emission potential (emission models); (2)
pollution control technology, equipment efficiencies,  and associated
capital and operating costs; and (3) exposure and health impacts for TSDF
pollutants (carcinogen potency factors).
     Inputs to the SAM calculations have been assembled into specific data
files.  Figure D-l outlines the functions and processing sequence of the
SAM and shows the data files used as input to the model and the output
files generated by the SAM.
     The facility processor is a segment of the program that accesses the
SAM input files and retrieves the information/data required for a particu-
lar determination or calculation.  The facility processor contains, in a
series of subroutines, all the program logic and decision criteria that
are involved in identifying TSDF facilities, their waste management proc-
esses, waste compositions, and volumes; assigning chemical  properties to
waste constituents and control devices to process units; and calculating
uncontrolled emissions, emissions reductions, control  costs, and health
impacts.  The facility processor also performs all the required calcula-
tions associated with estimating emissions, control costs,  and incidence.
Other functions of the SAM facility processor include performing a waste
stream mass balance calculation for each process unit to account for
organics lost to the atmosphere, removed by a control  device, or biode-
graded; testing each waste stream for volatile organic (VO) content and
vapor pressure based on models of the laboratory tests; determining total
organics by volatility class for each waste stream; and checking for waste
form, waste code, and management process incompatability.
D.I.3  Industry Profile
     Waste management processes, waste types, and waste volumes for each
facility are included in the SAM Industry Profile.  This file contains
each TSDF name, location, primary standard industrial  classification  (SIC)
code, and the waste volume and management process reported for that par-
ticular facility for each waste type (Resource Conservation and Recovery
                                    D-4

-------
             Input RIW
         Figure D-1. Source Assessment Mode! flow diagram.


•The parentheses refer to the appropriate sections of Appendix D
 that describe in detail the SAM input files.
                                  D-5

-------
Act [RCRA]  waste code).   Where the level  of detail  contained in  the SAM
Industry Profile is not  adequate for facility-specific determinations,  the
SAM uses estimates based on national average data.   The Industry Profile
contains information on  the management processes that are in operation and
the waste quantities that are processed at a particular facility.  What is
not known are the details on process subcategories  within the general
management process category.  For example, a given  quantity of waste is
reported as processed by treatment tanks; because no further information
is available, the SAM uses data on national averages for the distribution
and use of treatment tanks to identify and assign process subcategories
(i.e., covered quiescent tanks, uncovered quiescent tanks, and uncovered
aerated tanks) and to distribute waste quantities treated within these
subcategories for each particular facility.  This nationwide averaging
results in impacts that may not be accurate for an individual facility but
when summed yields reasonable nationwide estimates.
     The SAM facility-specific  information was obtained from three
principal sources.  Waste quantity data were taken from the 1986 National
Screening Survey of Hazardous Waste, Treatment, Storage, Disposal,  and
Recycling Facilities  (1986  Screener) J.2  Waste management  scenarios  (or
processing schemes) in the  SAM  were  based on the Hazardous  Waste Data
Management System's (HWDMS) RCRA  Part  A  applications,3 the  National  Survey
of Hazardous Waste Generators  and Treatment, Storage,  and  Disposal
Facilities Regulated  Under  RCRA in  1981  (Westat Survey),4  and the  1986
Screener.  Waste  types managed  in each facility were  obtained from all
three  sources.  For a more  detailed  discussion  of  the TSDF Industry
Profile, refer  to  Section  D.2.1 of  this  appendix.
D.I.4   Waste Characterization  File
      The Waste  Characterization Data Base (WCDB) is  a SAM file  that con-
tains  waste  data  representative of  typical  wastes  for each industrial
classification  (SIC code).   The SAM links waste data to  specific facili-
ties by the  primary SIC  code  and the RCRA waste codes (waste type) identi-
fied for  that  facility  in  the Industry Profile.   For those SIC  codes for
which  no waste  data were available,  waste compositions were estimated
using  the  available  data bases.  Waste data reported for facilities with
similar processes were  reviewed,  and waste stream  characteristics typical
                                     D-6

-------
 of the particular process were identified.  Thus, each SIC code is
 assigned applicable RCRA waste codes.
      A RCRA waste may be generated in one of several physical/chemical
 forms (e.g., an organic liquid or an aqueous sludge); therefore, the RCRA
 waste codes were categorized in the waste characterization file according
 to general physical and chemical  form.  Each physical/chemical  form of a
 waste code is assumed to contain  the composition of chemical  constituents
 at the respective concentrations  for the RCRA waste code.  The  SAM uses
 this aspect of the WCDB to distribute waste forms within a RCRA waste code
 and to provide a representative chemical composition for each form of
 waste.  The quantitative distribution of physical/chemical forms within a
 waste code was developed from the quantities reported in the  Westat Survey
 data base by the physical  and chemical  form of the waste code.
      Waste composition is  used to estimate emissions on  the basis  of
 concentration and volatility of the chemicals  present in the  waste.  Once
 waste form distributions are established,  the  SAM facility processor
 searches  for chemical  compositions  to assess the volatility and emission
 potential  of each waste  code/form combination  for use in  emission
 calculations.   Waste  characteristics  and compositions used in the  SAM are
 derived from five existing  data bases,  recent  field  data,  and RCRA waste
 listing background  documents.   It should be  noted that the model waste
 compositions  defined  in  Appendix  C, Section  C.2.2, of Reference 5  are not
 used  in any  way  in  the waste characterization  file or to  estimate  uncon-
 trolled emissions  from the  industry facilities.   Section  D.2.2  of  this
 appendix  contains  information  on  the  development  and  use  of the WCDB.
 D.I.5  Chemical  Properties  File
      Emission estimation on  a  chemical constituent basis  for  each  of  the
 more  than 4,000  TSDF waste constituents  identified in the  data  bases  was
 not possible because of a lack  of constituent-specific physical   and chemi-
 cal property data and  because of  the  sheer number  of  chemicals  involved.
 Therefore, to provide  the emission models with the relevant constituent
 physical,  chemical, and biological properties that influence emissions and
 still maintain a workable and efficient method of estimating emissions,
waste constituent categorization was required.  As a result, TSDF waste
constituents were grouped into classes by volatility  (based either on
vapor pressure or Henry's law constant,  depending on the waste management

                                     D-7

-------
unit process and emission characteristics)  and by biodegradability.
Surrogate categories were then defined to represent  the actual  organic
compounds that occur in hazardous waste streams based on the various
combinations of vapor pressure (four classes), Henry's law constant  (three
classes), and biodegradability (three classes).  The surrogates substitute
for the particular waste constituents (in terms of physical, chemical, and
biological properties) in the emission calculations  carried out by the
SAM.
D.I.6  Emission Factors File
     For each waste management process (e.g., an aerated surface impound-
ment), a range of model unit sizes was developed in order to estimate
emissions.  However, because specific characteristics of these model units
were unknown, a "national average model unit" was developed to represent
each waste management process.  Each national unit is a weighted average
of the nationwide distribution of process design parameters (e.g.,  unit
capacity), using the nationwide frequency distribution of each model  unit
size as  the basis for weighting.  For each model unit, its  emission factor
(emissions per megagram of waste throughput)  is multiplied  by  the appro-
priate weighting factor.  The sum of these products  results in a weighted
emission  factor for each national average model  unit.  The  weighted emis-
sion factors  were then compiled  into an  emission factor  file for use  in
the SAM  emission estimates.   The SAM multiplies  the  annual  quantity of
organic  compound processed  (or passed) through the  unit  by  the appropriate
weighted emission factor for  the surrogate  (constituent)  and management
process,  identified  in the  Industry  Profile,  to  calculate the  amount  of
organic  compound that  is emitted to  the  air  or that  is  biodegraded.
Because  wastes  may  flow  through  a series of  process  units,  a mass balance
is  performed  for each  waste management  process unit  to account for
organics lost to  volatilization  and  biodegradation  in the unit;  the
revised  organic content  is  then  used to  estimate the emissions for  the
next  downstream unit.
D.I.7   Control  Strategies  and Test  Method Conversion Factors
      As  a tool  for  evaluating control  strategies or regulatory options,
the SAM was designed to  calculate environmental  impacts of any number of
combinations of control  technologies and control efficiencies  which are
 part  of an externally generated  control  strategy.   For example,  controls
                              »
                                     D-8

-------
  can  be  applied  based  on  the emission potential of  the  incoming waste
  stream;  in this case, emission potential  is defined as the VO content of
  the  waste stream.  The SAM can test the stream for VO content and apply,
  from an  established file, VO test method  conversion factors to the stream
  to estimate the VO concentration a particular test method would detect.
  The  waste stream VO content can then be compared to a preselected VO
  action  level  (concentration limit) to determine if controls are to be
  applied  to the waste  stream.  If the waste stream exceeds the VO action
  level,  it is controlled as part of the TSDF control strategy.   The SAM
  then estimates emissions from each controlled management process with the
  appropriate technology in place.   The SAM can calculate emissions in a
  variety of formats.  Emission estimates can be presented by waste
 management process, waste code,  waste form, and volatility class,  on a
 nationwide level.
 D.I.8  Cost  and  Other Environmental  Impact Files
      Data files  have  also been  assembled  for calculating  controlled
 emissions,  control  costs,  and other  environmental  impacts.   Files  were
 developed for  the  SAM  that  provide control efficiencies,  capital  invest-
 ment, and annual operating  costs  for  each  control  option  that  is  appli-
 cable to a particular  waste management  process.  Cross-media and  secondary
 impacts  for the  control options are calculated  external to.the SAM.   These
 are the  environmental  impacts that result  from  implementation of the  air
 pollution control option  (e.g., solid wastes generated  through use of
 control  techniques  such as carbon adsorption).  For cost, cross-media, and
 secondary impacts,  control option impacts  are calculated  as a function of
 the waste quantities identified in the  Industry Profile.  Impact estimates
 were  developed for  a national average model unit that reflects the general
 frequency of national   unit size characteristics for each waste management
 process.  The impact estimates are divided by the model unit throughput  to
 obtain a  factor from which nationwide impacts are computed.  Multiplying
 facility throughput for the management unit by the appropriate impact
 factor results  in an estimate of the  impact for the particular unit.
These impacts  are summed  to yield  national  estimates.
                                    D-9

-------
D.I.9  Incidence and Risk File
     The SAM incidence and risk file contains exposure level coefficients
to estimate annual  cancer incidence and maximum lifetime risk (MLR) for
the population within 50 km of each TSDF.  The coefficients were developed
using the Human Exposure Model (HEM) with 1980 census population distribu-
tions, local meteorological/climatological STAR data summaries,  and an
assumed emission rate (10 Mg/yr)  and unit risk factor (1 case//tg/m3/per-
son).  The SAM facility-specific incidence and risk coefficients can be
scaled by annual facility emissions and the appropriate unit risk factor
to give health impact estimates that reflect the level of emissions
resulting from a particular emission scenario or control option.  For a
more detailed examination of incidence and risk determinations,  see
Appendix E of Reference 5.
D.2  INPUT FILES
D.2.1  Industry Profile Data Base
     D.2.1.1  Introduction.  As an initial input to the estimation of air
emissions, an Industry Profile was developed to characterize TSDF waste
management practices.  The Industry Profile is based on data from the
Westat Survey and from EPA's HWDMS.  Data from the Office of Solid Waste's
(OSW) 1986 Screener, which reflect 1985 TSDF activities, are also used
heavily.
     The following sections describe the  Industry Profile contents and
outline the data base sources.  Discussion centers on the current  Industry
Profile of 2,336 TSDF.  Section D.2.1.2 describes the data  base structure
and contents, Section D.2.1.3 documents selection of the SAM TSDF uni-
verse, and Section D.2.1.4 reviews data sources.
     D.2.1.2  Data Base Contents.  Table  D-l lists the variables in the
current Industry Profile.  Each record in the  Industry Profile constitutes
a single waste stream.  A facility may have several different waste
streams.  The variables following the waste code indicate quantities and
management methods for TSDF operations.   All quantities are expressed in
megagrams per year  (Mg/yr).
     Table D-2 gives an example record of an Ohio TSDF with EPA identifi-
cation number OHDOOOOOOOOO  (variable FCID).  Its primary SIC code  is
designated as 2879  (SIC1, Pesticides and  Agricultural Chemicals).  Table
                                    D-10

-------
 Variable
             TABLE D-l.   INDUSTRY PROFILE DATA BASE CONTENTS3
                                        "•••™ \^~—^^^^—^ i,. •»
                                         Description
                EPA 12-digit facility identification number
                Primary 4-digit standard industrial classification (SIC)
                  code
                EPA hazardous waste number (RCRA waste code)
                Amount of waste for WSTCDE (Mg/yr)
                Amount of waste stored (Mg/yr)
                Storage process(es)  - one of 20 potential process combina-
                  tions13
                Amount of waste treated  (Mg/yr)
                Treatment process(es) -  one of 19 potential  process
                  combinations"
                Amount of waste disposed (Mg/yr)
                Disposal  process(es)  - one of 11  potential  process combi-
                  nations"
                Source of data  for waste quantities,  RCRA codes,  and
                  management  methods
                Facility  status
                Latitude  (expressed  in degrees, minutes,  seconds,  and
                  tenths  of seconds)
                Longitude (expressed  in  degrees,  minutes,  seconds,  and
                  tenths  of seconds)

RCRA  = Resource  Conservation and Recovery Act.
  Mg  = Megagrams.

aThis table identifies and describes  those variables of  the  Industry
 Profile data base used  to characterize treatment, storage,  and disposal
 facilities in nationwide impacts modeling.
"Hazardous waste management process combinations  are presented in
 Table D-3.
 FCID
 SIC1

 WSTCDE
 WAMT
 QTYSTR
 TYPSTR

 QTYTX
 TYPTX

 QTYDIS
 TYPDIS

 SOURCE

 ELIGSTAT
 LATT

LONG
                                 D-ll

-------
        TABLE 0-2.   INDUSTRY PROFILE DATA BASE - EXAMPLE RECORD9

              Variable                Contents

              FCID                    OHDOOOOOOOOO
              SICC1                   3879
              WSTCDE                 D001&
              WAMT                    1056954
              QTYSTR                 1056954
              TYPSTR                 1
              QTYTX                   1056954
              TYPTX                   10
              QTYDIS                 0
              TYPDIS                 0
              SOURCE                 2
              ELIGSTAT                7
              LATT                    3115000
              LONG                    08758000
    example record of how one facility waste stream would  appear  in  the
 Industry Profile data base.
bD001 = ignitible waste.   Source:   40 CFR 261.21,  Characteristic  of
 ignitibility.
                                     D-12

-------
  Ignitible  wastes  identified  as  D001  (WSTCDE)  are  managed  at  this  facility.
  This  TSDF  manages  (WAMT)  and  stores  (QTYSTR)  1,056,954  Mg  of waste  D001  in
  a  tank  (TYPSTR  =  1—see Table D-3),  but  it  also treats  the same amount
  (QTYTX  = 1,056,954 Mg) in a tank  (TYPTX  = 10-see Table D-3).  No quantity
  of this waste is disposed of  (QTYDIS and TYPDIS,  respectively).   The data
  source  for the  RCRA waste code, its  fraction  of the total  TSDF waste quan-
  tity, and  its management processes may have come  from EPA's  HWDMS (SOURCE
  =  2,  3, or 4).  Another source of such data may include the  Westat  Survey
  (SOURCE =  1).  OSW's 1986 Screener (SOURCE =  5 or 6) provided the total
 waste quantity managed in 1985—from which the waste code  quantity  was
 derived—along with verification of waste management processes active in
 1985.  The facility operating status code (ELIGSTAT) indicates the  TSDF is
 an active TSDF,  ELIGSTAT = 7  (former TSDF,  ELIGSTAT = 1; or closing TSDF,
 ELIGSTAT = 3).  Latitude (LATT)  of the site is 31  degrees, 15 minutes,  and
 no seconds, and  the longitude (LONG)  is 8 degrees, 75 minutes, and no
 seconds.
      The Industry  Profile  contains the following waste management  proc-
 esses found under  variables  TYPSTR (storage),  TYPTX (treatment),  and
 TYPDIS (disposal):
           Storage  in a  container (SOI),  tank (S02),  wastepile (S03),
          or  surface impoundment  (S04)
          Treatment in  a tank  (T01),  surface impoundment (T02)   in-
          cinerator (T03), or  other process  (T04)
          Disposal by injection well  (D79),  landfill  (D80), land ap-
          plication (D81), or  surface  impoundment  (083).
 A variety of management process combinations may occur at  facilities, some
 of  which one would expect to find  in  parallel  or in  series.   Where a series
 representation in the Industry Profile is not  appropriate,  the SAM is
 programmed to divide streams evenly between or among the listed processes.
 All potential  process combinations found  in the Industry Profile are listed
 in  Table D-3 with the assigned divisions.  The processes in column 2 become
 the parallel or series-parallel processes in column 3.   Note  that  T04
 ("other treatment") is  listed separately, but  its emissions are calculated
on the basis of T01 (treatment tanks)  operation.   T03 (incineration)  and
D79 (injection well)  are listed,  but the SAM only calculates their transfer
                                   D-13

-------
           TABLE D-3.  INDUSTRY PROFILE REFERENCE KEY FOR WASTE
                     MANAGEMENT PROCESS COMBINATIONS3
Combination
number
Storage Processes
0
1
2
3
4
5
6
7
8
9
10
lib
12b
13b
14b
15
16
,»
Process code
description0
(variable TYPSTR in Table D-l)
No storage
S02 only
SOI only
S04 only
SOS only
Other storage
SOI, S02
SOI, S04
SOI, 502, SOS
SOI, SOS
SOI, S02, S04
SOI, S04
SOI, SOS, S04
S04, sump
S02, other
SOS, S04
S02, SOS
S02, SOS, S04
Waste flow used
in modeling simulation

No Storage
•» S02
+ SOI
-> S04
- SOS
- SOI
-> SOI -> S02
+ SOI -> S04
r* SOI -> S02
*U SOS
- SOI •> SOS
+ SOI •» S02 - S04
r> SOI
*U S04
r* SOI -> S04
*u sos
-> S04
* S02 -> SOI
r» SOS
*U $04
r* S02
"U SOS
r> S02
*P sos
U S04
See notes at end of table.
(continued)
                                    D-14

-------
TABLE D-3 (continued)
Combination
number
Storage Processes (con.)
18
19b
20
Treatment Processes (vari
0
1
2
3
4
5
6b
7b
8
9b
10
lib
12
13
See notes at end of table
Process code
description0

S02, S04
SOI, S02, S03, S04
SOI, S02
able TYPTX in Table D-l)
No treatment
T01 only
T02 only
T03 only
T04 only
T01, T02
T01, other
T01, other
T01, T03
T03, other
T01, T02, T03
T01, T03, other
T02, T03
T02, T04
*
Waste flow used
in modeling simulation

r* S02
*U S04
r+ SOI
*F S03
U S04
r» SOI
*U S02



-» S02


No treatment
+ T01
+ T02
* T03
+ T04
•> T01 *
+ T01 +
+ T01 •>
.-> T01
U T03
+r T03
U T04
r-> T01
"U T03
*r'T01
U T03
r* T02
"U T03
+ T02 •»





T02
T04
T04


•> T02
+ T04

T04
(continued)
        D-15

-------
                           TABLE D-3  (continued)
Combination
number
Treatment Processes
14b
15
16
17
18
19
Disposal Processes
0
1
2
3
4
5
6
7
8b
gb
10
Process code
description0
(con.)
T01, T02, T03, T04
T01, T04
T03, T04
T01, T02, T04
T01, T03, T04
T02, T03, T04
(variable TYPDIS in Table D-l)
No disposal
079 only
D80 only
D83 only
D81 only
Other
D81, 083
080, 083
079, 083
079, 081
080, 081
Waste flow used
in modeling simulation

r* T01 -> T02 «• T04
*U T03
«• T01 + T04
*r T03
U T04
+ T01 * T02 -» T04
r-> T01 + T03
*U T04
r» T02 * T04
*U T03

No disposal
* 079
•> 080
* 083
«• 081
* 080
^r D81
U 083
p* 080
"U D83
+r D79
U D83
~ 079
U D81
r» 080
"U D81
See notes at end of table.
(continued)
                                    0-16

-------
                            TABLE D-3 (continued)
Combination               Process code                     Waste flow used
  number                  description0                 in modeling simulation

Disposal Processes (con.)


_1|	D79- D80    	
-------
and handling emissions.   This is because a separate Agency program is under
way to regulate air emissions from hazardous waste incineration and because
there are no process air emissions from injection wells.
     The Industry Profile also contains RCRA waste codes  as defined in
Title 40, Part 261, of the Code of Federal Regulations (CFR).7  The data
base contains over 450 waste codes and includes "D," "F," "K," "P," and "U"
RCRA codes.  Hazardous waste codes are described in more  detail in Chapter
3.0.
     D.2.1.3  Establishing the SAM Universe of TSDF.  The 1986 Screener
surveyed over 5,000 potential TSDF.  The Screener identifies 2,221 "active"
TSDF to be characterized in the SAM.  An active facility treated, stored,
disposed of, or recycled waste during 1985 that was considered hazardous
under Federal RCRA regulations.  Active facilities include TSDF filing for
closure if the facility managed some waste in 1985.  The Screener desig-
nates as "inactive" those facilities that fall into any of three other
categories:
     •    Former TSDF that have ceased  all hazardous waste management
          operations
     •    TSDF that are closing and did not manage waste  in  1985
          Facilities that do not  treat, store, dispose of, or  recycle
          hazardous waste.
     Active  Screener TSDF that  are  not  currently  addressed  in  the  SAM were
excluded.   Excluded TSDF  represent:
     •    TSDF that manage polychlorinated  biphenyls  (PCB)--a  waste
          that is  currently  not RCRA  hazardous
     •    TSDF whose waste  is  hazardous under State  RCRA regulations
           but  not  under  Federal RCRA  rules
     •     TSDF that treat waste in  units  exempt  from RCRA or store it
           under  the 90-day  rule (40 CFR 262.34(a))8 and,  therefore,  do
           not  require  RCRA  permits
TSDF whose  total  waste  amount  managed (including storage,  treatment, and
disposal)  is less  than  0.01  Mg/yr (about  340 TSDF)  were  considered small
potential  emitters and  were  also  excluded from the SAM to improve data base
manageability.   A total  of  about  340  TSDF were excluded  due to either
                                    D-18

-------
0.01-Mg/yr cutoff or because they only managed State-designated hazardous
waste.  Another nine active TSDF were excluded from the Industry Profile
because all available data are classified as Confidential Business
Information (CBI).  The impact on nationwide waste volume from these nine
TSDF is considered small due to their low volumes (less than 0.5 percent of
the waste volume managed nationwide).
     In addition to currently active TSDF, former or closing TSDF that had
land disposal operations were also profiled.  This is because of the poten-
tial source for air emissions from TSDF closed with waste left in place.
The Westat Survey, HWDMS, and 1986 Screener identified 115 TSDF with former
or closing land disposal operations.  Therefore, the total universe for the
SAM was set at 2,336 TSDF (2,221 active TSDF plus 115 closing or former
TSDF).
     D.2.1.4  Data Sources.  The Industry Profile represents a composite of
waste-stream-specific information collected from the 1986 Screener, the
Westat Survey, and HWDMS.  This section describes each of these sources.
Waste stream data for each facility were derived from these sources as
shown in Table D-4.
   TABLE D-4.   INDUSTRY PROFILE DATA BASE:  DISTRIBUTION OF FACILITIES
                           AMONG DATA SOURCES3



Data source
Westat Survey
HWDMS
1986 Screener
Total


Number of
active TSDF +
438
1,361
422
2,221
Number of
closed or
former TSDF
with land
disposal units
27
85
3
115



= Total TSDF
465
1,446
425
2,336
TSDF = Treatment,  storage,  and disposal  facility.
HWDMS = Hazardous  Waste Data Management  System.
aThis table shows  the number of facilities for which each Industry Profile
 data source provides waste stream information.
                                   D-19

-------
     The 1986 Screener was used to identify the universe of regulated TSDF
and their waste quantities managed annually.  The Screener data base con-
tains the most current data on TSDF operations—data from the year 1985.
However, specific waste codes and the processes by which they are managed
at each facility are not contained in the data base.  Therefore, two other
sources of waste code data were used.  The Westat Survey was the preferred
data source for assigning RCRA waste codes and management processes and
distributing waste quantities by process.  But due to the Westat Survey's
limited sample of 831 TSDF, it was necessary to access the HWDMS RCRA Part
A permit application data.  The 1986 Screener was also used to verify man-
agement processes in operation and describe a TSDF's waste streams and
management processes if the Westat Survey or the HWDMS data did not contain
the information needed.
     The Westat Survey and the HWDMS were used as initial inputs to assign-
ing an SIC code to each facility.  Section D.2.1.4.4 outlines additional
sources used to determine a facility's principal business activity.
     D.2.1.4.1  1986 Screener data.  The goals of using the 1986 Screener
data were threefold:  (1) to identify which TSDF should be included in the
SAM, (2) to profile 422 active TSDF identified by the Screener but not
included in the HWDMS or the Westat Survey, and (3) to update the total
waste quantity by TSDF to reflect 1985 data.
     As a first goal, the Screener data on TSDF operating status were com-
pared to the Industry Profile list of active and closed facilities.  Any
inconsistencies in the profile were revised, using the 1986 Screener infor-
mation as the most current source of data.
     The second goal--to profile the additional Screener TSDF--entailed
adapting the Screener data to make them compatible with the HWDMS and the
Westat Survey.  The 1986 Screener does not refer to individual RCRA waste
codes but rather to general waste types:  acidic corrosives,  metals,  cya-
nides,  solvents,  dioxins, other halogenated organics,  find other hazardous
waste.   Also,  management processes listed in the Screener differ slightly
from the processes cited in the HWDMS and the Westat Survey.   For instance,
the 1986 Screener does not list storage in tanks or containers,  specifi-
cally.   Rather,  these are combined in a category listed as "other storage."
                                   D-20

-------
 To  adapt  these Screener data, default waste categories were developed to
 replace RCRA waste codes, and management process descriptions were con-
 verted to RCRA process codes.   For example, the 1986 Screener waste type
 "acidic corrosives" was assigned to a default RCRA waste code of D002 (cor-
 rosive waste).  Cyanides were assigned to D003 (reactive waste).   (Section
 D.2.2.10  describes the development of default waste compositions.)  For
 waste management processes, most process code assignments were straight-
 forward;  however, some process  descriptions were not.  For example, the
 Screener1s wastewater treatment category was assigned the process code T01
 (treatment in a tank) when not  specified as exempt from RCRA regulation.
 Other processes included solidification, which was assigned T04 (other
 treatment), and "other storage," which was assigned a combination of SOI
 and S02 (storage in a container or tank).
     After assigning management processes and RCRA waste codes to each
 facility, the next step used to develop Screener waste streams was to as-
 sign specific waste quantities to RCRA waste codes and management proces-
 ses.  Question 3 of the Screener indicated the total amount of waste that
 was treated, stored, or disposed of onsite in units regulated under RCRA at
 each facility.  Quantity distributions were made based on information
 obtained  from the 1986 Screener, telephone inquiries conducted by the
 Screener  staff, and best engineering judgment.
     The  third goal in using 1986 Screener data was to update waste quan-
 tities (derived from the HWDMS or the Westat Survey) for the active TSDF.
 Screener  Question 2 was used to identify the total quantity of hazardous
 waste that was treated, stored,  or disposed of onsite in 1985 under Federal
 RCRA regulations.  The 1985 total  quantity of waste per facility was dis-
 tributed  among waste streams on  a weight basis.   1985 distributions were
 made proportionate to the TSDF's distribution of waste code quantities used
 previously from either the HWDMS or the Westat Survey.  For example, if a
 facility  had a waste code quantity of 1,000 Mg and a total  waste quantity
 for the facility  of 2,000 Mg,  the distribution of waste code to total  waste
quantity  is 1,000/2,000 or 0.5.   If Screener data indicate that the facil-
 ity has a 1985 total  waste quantity of 3,000 Mg,  the waste code quantity is
 increased  from 1,000 to 1,500 Mg to reflect its  ratio to the facility's
total  waste quantity (0.5 multiplied  by 3,000).
                                   D-21

-------
     D.2.1.4.2  Westat Survey.  Data were accessed from Westat's general
questionnaire to identify facility waste streams.  Question 12 asked for
the total quantity of hazardous waste that the facility treated, stored, or
disposed of onsite during 1981.  Question 17 asked the facility to complete
a table for the 10 hazardous wastes handled in largest volume in 1981.  The
table requested that the waste be listed by EPA waste code and include a
breakdown of waste by specific management processes (e.g., tank, incinera-
tor, wastepile) and by specific waste quantities for storage, treatment,
and disposal.  The Westat Survey is preferred to HWDMS as a data source
because data reflect actual annual throughputs and waste management proc-
esses for TSDF.  However, the data base covered only 831 TSDF.  Of these,
only 438 active and 27 closed TSDF were of interest.  Also, data represent
activities in the year 1981 and may no longer be accurate.  Westat Survey
data have been reviewed to exclude hazardous wastes that are exempt or
excluded from RCRA regulation.  The Westat Survey specifically excludes
waste streams sent to publicly owned treatment works (POTW), waste from
small quantity generators, wastes that are stored in containers or tanks
for less than 90 days, wastewater treatment in tanks whose discharges are
covered under National Pollutant Discharge Elimination System (NPDES) per-
mits, and wastes that have been delisted by EPA even if the delisting
occurred after 1981.9
     D.2.1.4.3  HWDMS.  HWDMS data,  retrieved in October of 1985,  consist
largely of RCRA Part A permit application information.  Existing TSDF were
required to complete Part A of the permit application by November 19, 1980,
in order to receive interim status to operate.  The Part A permit asks the
facility to list quantity of waste (by RCRA waste code) that will  be
handled on an annual basis and waste management processes that will be
used.
     HWDMS data have several disadvantages compared to Westat Survey data.
Unlike the Westat Survey data, Part A reflects estimated, not actual, waste
throughput and processes.  Part A is a record of "intent to manage" waste.
The HWDMS also does not break down the total amount of waste managed into
quantities that were treated, stored,  or disposed of,  and the year for
which data are provided is unknown.   A facility may have submitted an
                                   D-22

-------
amended Part A to reflect changes in waste types or quantities since 1980,
but the date of submission cannot be ascertained.  Finally,  some waste
streams may reflect processes that are exempt or excluded under RCRA, such
as less than 90-day storage.  These streams cannot be identified.
     D.2.1.4.4  SIC codes development.  Each of the TSDF in  the Industry
Profile was examined individually to determine a primary 4-digit SIC.  In
assigning SIC, the HWDMS and Westat Survey were used as initial points of
reference,  but because of the number of nonexistent codes and the abundance
of only 2-  or 3-digit SIC codes, each SIC was verified using all available
reference sources.
     Several steps were taken to assign an SIC code.  The Standard Indus-
trial Classification Manual 10 was used to identify SIC codes for TSDF when
no code was provided in the data sources, and the facility's name, address,
waste codes, and waste amounts were examined for identifying information.
In many instances, this information was enough to assign an  SIC.  For exam-
ple, a facility, Wood Preserving Company B, was assigned an  SIC of 2491
(wood preserving industries).  A facility with waste codes of K048-K052
would be assigned an SIC relating to the petroleum refining  industries.
Additional  sources of informational|12,13 provided corporate or plant
descriptions.  Also, the various census reportsl4-18 were used to identify
the number of facilities in each State with a given SIC code.  For example,
in trying to establish an SIC for Oil Service Company C in Arizona, waste
codes were referenced first.  No "K" waste codes were identified that
related the facility to petroleum refining.  Therefore, the  Census of Manu-
factures^  Was consulted.  It indicated zero petroleum refineries in
Arizona.  Oil Service Company C was assigned the SIC of 5172 (petroleum
products not elsewhere classified).
D.2.2  TSDF Haste Characterization Data Base (WCDB)
     D.2.2.1  Background.  To support the development of air emission regu-
lations for hazardous waste TSDF, a data base of waste characteristics was
developed.   Wastes listed in this data base were characterized, primarily
using five  existing data bases:  (1) the Westat Survey,20 (2) the Industry
Studies Data Base (ISDB),21 (3) a data base of 40 CFR 261.32 hazardous
wastes from specific sources22  (i.e., waste codes beginning  with the
letter K),  (4) the WET Model Hazardous Waste Data Base,23,24 and (5) a data
                                    D-23

-------
base created by the Illinois EPA.25  An additional  source of data,  EPA
field reports on hazardous waste facilities,  also was used.   The WCDB makes
no use of the model wastes defined in Appendix C, Section C.2.2, of Refer-
ence 5.
     The Westat Survey data base contains the most extensive information on
the physical/chemical form, quantity, and management of waste;  therefore,
it was selected to serve as the framework for the TSDF WCDB.  This data
base has been organized to present hazardous waste stream* information in
the following series of categories:
     •    Primary  SIC code
     •    RCRA waste code
     •    General  physical/chemical waste form.
For each SIC code, Westat  contains a list of waste codes.   It then divides
each waste  code  into physical/chemical forms such as  inorganic  sludges,
organic  liquids, etc.  Westat also designates a  waste quantity  for each
physical/chemical  form of  a waste  code.
     The remaining four data bases and EPA field reports were used to  pro-
vide chemical  composition  data  in  the  form of two additional data cate-
gories  in the  WCDB:   "waste constitutents" and  "percent  composition  of con-
stitutents."   Where  information was  not  available for these two categories,
a list  of constitutents and their  percent compositions was  created  (i.e.,
default composition)  based on  information found  in  the four data bases,
field  reports,  RCRA  waste listing  background  documents,  and engineering
judgment.
      Table  D-5 is  an example of a  hazardous  waste  stream in the WCDB.   This
example states that,  in the commercial  hazardous waste management industry
 (SIC code  4953), RCRA waste code U108 is managed as an  organic  liquid (form
4XX).   Its  composition is 90  percent 1,4-dioxane and 10  percent water.
      D.2.2.2  Application to the Source Assessment Model (SAM).  The SAM
 uses the WCDB to identify representative compositions for wastes managed at
 each TSDF.   SAM uses these compositions to  estimate organic emissions based
 on waste constituent concentrations  and their volatility.  The procedure is
 described  in the following paragraphs.
      *For discussion, a hazardous waste stream is a unique combination of
 SIC code, RCRA waste code, and physical/chemical form.
                                     D-24

-------
               TABLE  D-5.  WASTE CHARACTERIZATION DATA BASE:
                       EXAMPLE WASTE STREAM RECORD3
SIC code                 ,                        4953
Form codeb                                       4XX

RCRA characteristic codec.d                      T

RCRA waste coded                                 U108

Waste constituent/% composition                  l,4-Dioxane/90%

                                                 Water/10%

SIC = Standard industrial classification.
RCRA = Resource Conservation and Recovery Act.

aThis table presents an example of the information found in the Waste
 Characterization Data Base for one waste stream managed in a given
 industry.

^Physical/chemical waste forms are coded as follows:
 1XX = Inorganic solid            4XX = Organic liquid
 2XX = Aqueous sludge             5XX = Organic sludge
 3XX = Aqueous liquid             6XX = Miscellaneous.
CRCRA characteristic code reflects the hazard of the waste:
 T = Toxic
 C = Corrosive
 I = Ignitible
 R = Reactive.

dRCRA characteristic and waste codes listed in 40 CFR 261.33(f).26
                                  D-25

-------
     The SAM initially reads the Industry Profile  (described  in  Section
D.2.1) for each TSDF's primary SIC code,  RCRA waste codes,  and  the annual
quantity of each code.  It then searches  the WCDB  for this  SIC  and then  for
the TSDF's RCRA waste codes.  Because the physical  and chemical  form of  a
waste code may vary,  the chemical  composition and  emission  potential will
also vary.  Therefore, for each waste code,  the WCDB provides quantities
from the Westat Survey data base by physical/chemical form  of the waste
code.  The quantitative distribution of physical/chemical forms within a
waste code is then applied to the Industry Profile waste code's quantity
for that TSDF.  For example, if the TSDF's profile has 150  Mg of D003 and
the WCDB shows that D003 has 1,200 Mg of organic liquid and 600 Mg of
organic sludge forms present across that SIC (i.e., a two-to-one ratio by
form), the TSDF profile's 150 Mg is distributed two-to-one as 100 Mg of
organic liquid and 50 Mg of organic sludge.  This approach  allows the most
current waste quantity information to be used in a more detailed fashion,
using distribution data from a more rigorous data source (Westat Survey).
     Once form distributions are established, the SAM begins to search for
chemical compositions to assess volatility and, in turn, emission potential
of each waste code/form combination.  The search proceeds as depicted in
Figure D-2.  Six discrete sets of waste  composition  data are identified in
the  figure:
           ISDB
      •     Field data
      •     Illinois EPA data base
      •     K  Stream data base
      •     WET  Model  data  base
      •     Data set consisting  of  default values.
The  logic  shown  in Figure D-2  ranks  these data  sets  in  the  order listed
above to  reflect  the relative  certainty  in  data representativeness.  Thus,
 if a waste stream had more  than one  set  of  compositions to  choose from,  the
SAM would use  the highest ranking data base composition;  The  logic diagram
does not  include  the Westat Survey constituents because no percent composi-
tions were available.
                                    D-26

-------
Is there a unique ISOB stream
with numerical percentages?
            I
  Is there a corresponding
     field data stream?
  Is there a corresponding
    Illinois EPA stream?
  Is there a corresponding
  "K" data base stream?
 Is there a corresponding
 "WET" data base stream?
  Is there a default list of
      constituents?
   Print "Not available"
     in the final list.
 Print as the
  final list.

  Go to next
waste stream.
             Figure D-2.  Logic flow chart for selection of final list
                            of waste constituents.
                               D-27

-------
     Sections 0,2.2.4 through D.2.2.10 discuss each of the five existing

data bases, EPA's field data base, and the default values established.

     D.2.2.3  Limitations of the WCDB.  The limitations of this WCDB

coincide with those found in all contributing data bases.  Therefore, some

of the same weaknesses were shared:

     •    Compositional data were not available from the existing data
          bases on each SIC code/waste code/waste form combination
          (also referred to as a "waste stream").  Therefore, it was
          necessary to assign compositions (i.e., default composi-
          tions) to 30 percent of the organic waste streams.  This
          reduces the certainty of actual waste compositions the SAM
          uses for SIC codes.

     •    The data base consisted of 1981 waste codes (the year the
          Westat Survey was conducted).  It did not reflect additions
          to 40 CFR 26127 since 1981 such as listing of dioxins.
          However, wastes delisted since 1981 have been eliminated
          from the WCDB.  Thus,  the SAM emission estimates reflect
          delisting of wastes but not the role of wastes listed since
          1981.

     •    Certain organic constituents are generic chemical classes,
          e.g., "amino alkane,"  and thus do not have specific physical
          and chemical properties.  Therefore, volatility and biodeg-
          radation classes were designated for these generics by
          referencing a common chemical considered representative of
          that generic chemical.  Therefore, the presence of generic
          classes in the WCDB decreases the SAM's certainty of
          predicting appropriate emissions from that class.

     D.2.2.4  Westat Survey Data Base.  This survey data base compiles data
from a 1981 EPA survey of all hazardous waste generators and TSDF.  Use of

the data base for this project focused on TSDF only.

     The Westat Survey data base contains information on TSDF from approxi-

mately 230 SIC codes, covering active and closed TSDF.  A subset of the

data base was used to develop the TSDF WCDB.  This subset represents only

the active facilities in the Westat data base (covering 182 SIC codes).

The active facilities constitute about 70 percent of the complete Westat

data base, and closed facilities make up the remaining 30 percent.

     D.2.2.4.1  Use of the Westat data base.  As stated in Section D.2.2.1,

the Westat data base provides the SAM (1) quantitative distributions of

physical/chemical forms of waste codes, and (2) the framework for the SAM

to track a waste code to an appropriate chemical composition in the WCDB.
                                   D-28

-------
  (Compositions are selected from the data bases described in Sections
  D.2.2.5 through D.2.2.10.)

      The WCDB uses Westat waste stream information such as facility SIC

  code, RCRA waste codes managed, and physical/chemical forms of waste codes

  (i.e., waste streams).  This information is organized by SIC so that data

  can be applied to any TSDF in the Industry Profile with that SIC code.

      The WCDB and the SAM use the following Westat data base categories:

      •    SIC code—Primary SIC code of the survey respondent.   If the
           respondent's primary SIC code was 2-digit,  e.g.,  2800,  the
           more detailed,  secondary SIC code listed by the respondent
           was used when available, e.g.,  2812.  (For all  remaining
           2-digit codes,  more descriptive 4-digit codes were assigned
           to the WCDB based on knowledge of the TSDF's industrial
           operations.)

      •     RCRA waste  code—Survey  respondents  were asked  to list  the
           10 largest  waste streams (by RCRA waste code)  managed at
           each TSDF.   Thus,  for each  SIC  code,  TSDF respondents with a
           matching SIC will  have their top  10  waste codes  listed.

      •     Physical/chemical waste  form—Survey respondents  were also
           asked  to describe the physical/chemical  character of  each  of
           the  10 waste streams.  Based on these descriptions, the
           physical/chemical forms  were classified  as  follows:

                1XX  Inorganic  solid          4XX Organic  liquid
               2XX  Aqueous sludge          5XX Organic  sludge/solid
               3XX  Aqueous liquid          6XX Miscellaneous

          Therefore, within a SIC's waste code, one will find as many
          as six forms of  that waste code.

          Physical/chemical waste  form quantity—The quantity of each
          physical/chemical form of a waste code managed within each
          SIC code.  (Note:  These form quantities are mutually exclu-
          sive of each other and may be added.)  If more than one  TSDF
          reported the same form of waste code, their quantities were
          added to provide an indication of the volume of that  stream
          managed by the TSDF population having a common SIC code.

     D-2.2.4.2  Westat Survey Data Base limitations.  Certain limitations

of the Westat Survey data base that may affect the SAM results are dis-
cussed below:
                                   D-29

-------
Several survey respondents identified wastes by using more
than one waste code.  The EPA entered these streams into the
Westat data base as X---codes.  For the WCDB,  the X codes
were translated into their respective D, F, K,  P, and U
waste codes, and the first code listed from the multiple
codes was used in the WCDB.  For example, if X002 is a com-
bination of F003 and F005, then F003 was used in the WCDB.
Not knowing which code best represented a waste increased
the uncertainty of waste compositions used in the SAM.

Individual waste streams were not always keyed to their most
descriptive SIC code.  The WCDB identifies waste streams by
the primary SIC code listed by a TSDF.  Consequently, it is
possible that a waste stream will be identified by the
facility's primary SIC code when another SIC code is more
descriptive.  To correct this limitation, the most descrip-
tive SIC codes were chosen following an Industry Profile
review of facility SIC codes.

Invalid or missing codes were found in the Westat data base.
For example, the Westat data base may have no SIC codes
listed for some TSDF, invalid RCRA waste codes listed such
as "DOOO, 9995, 9998, 9999, Y—," and no physical/chemical
form of waste listed.

To examine those Westat Survey waste streams with invalid
waste forms and waste codes (9999, etc.), a list of such
codes was generated. Then, it was decided to remove some of
these streams from the WCDB and reassign real  waste codes to
the remaining streams based on an examination of waste con-
stituents and waste form.  The following summarizes steps
taken to resolve invalid waste codes and forms:

     For invalid waste codes:

          —Streams <18.9 Mg (5,000 gal) were not included
            in the WCDB.

          --Streams <18.9 Mg but containing PCB were reas-
            signed.

          —Streams >18.9 Mg but containing no constituent
            information were not included.

          --Streams >18.9 Mg and having useful  constituent
            information were reassigned.

     For waste streams with no physical/chemical form
     listed:

          --Streams <18.9 Mg were not included in the WCDB.
                         D-30

-------
                     --Streams having no constituents were not
                       included.
                     —Management method(s) were  reviewed for a clue as
                       to the liquid, sludge, or  solid state.  Then,
                       physical/chemical forms were assigned to such
                       streams.
     D.2.2.5   Industry Studies Data Base.  The ISDB is a compilation of
data from  EPA/OSW surveys of designated industries that are major hazardous
waste generators.  The ISDB version used addresses eight SIC codes:
     •     Industrial  inorganic chemicals - alkalies and chlorine (SIC
           2812)
     •     Industrial  inorganic chemicals - not elsewhere classified
           (SIC 2819)
     •     Plastics materials, synthetic resins,  and nonvulcanizable
           elastomers  (SIC 2821)
           Synthetic  rubber (SIC 2822)
     •     Synthetic  organic fibers, except cellulosic (SIC 2824)
     •     Cyclic crudes, and cyclic intermediates, dyes, and organic
           pigments (SIC 2865)
     •     Industrial organic chemicals, not elsewhere classified (SIC
           2869)
     •     Pesticides and agricultural chemicals,  not elsewhere classi-
           fied (SIC 2879).
Data on other SIC codes are being developed by the EPA/OSW and could be
added in the future.  Information in the ISDB was gathered from detailed
questionnaires completed by industry, engineering analyses,  and a waste
sampling/analysis program.  The data base contains detailed information on
specific TSDF sites.  Because of the confidential nature of much of the
data,  waste information was provided in a nonconfidential  form to allow its
use; e.g.,  generic chemical  constituent names such as "amino alkane" were
used where specific constituents were declared confidential.
     D.2.2.5.1  Use of the ISDB.   The WCDB contains ISDB waste composition
data.   The WCDB uses the ISDB SIC code, waste code, and its  physical/chemi-
cal  waste  form to track and identify waste stream compositions.   It then
                                   D-31

-------
uses the waste form's quantity in  the  ISDB  to  normalize constituent  concen-

trations across multiple occurrences of  the same waste stream.  The SAM  uses

the ISDB composition data via the  WCDB for  TSDF with  those  SIC  codes listed

in the previous subsection.   The SAM uses the  following ISOB waste composi-

tion data:

     •    Constituents—The  ISDB provides chemical  constituents con-
          tained in an SIC code's  waste  code/waste  form combination,
          i.e., a waste stream.  The stream data have been  compiled  in
          a way that makes all information  nonconfidential.

     •    Normalized constituent concentrations—Weighted average
          constituent concentrations were calculated  for  each  of the
          constituents to yield a normalized waste  stream composition.
          Normalizing sets all total constituent concentrations to  100
          percent.

     D.2.2.5.2  ISDB limitations.   The ISDB used  in the WCDB  provided

useful waste composition data not only for  direct  use in  the  SAM but also

to fill data gaps in the WCDB, e.g.,  to create default compositions  for SIC

codes where waste compositions were not available.   However,  it is  neces-

sary to identify some limitations of the ISDB:

     •    The  petroleum refining industry—one of  the top five indus-
          try  generators—was not available for the ISDB  version used.
          The  EPA/OSW surveyed this industry  (SIC  code 2911),  but
          questionnaire responses were not  accessible from the data
          base at the time.   However,  some raw field data were pro-
          vided for  the industry under the ISDB program.   This is
          discussed  in Section D.2.2.6.  For waste streams with no
          field data, K stream data and default compositions  were
          used.

     •    The  ISDB  used a larger number of more specific waste forms
          than the  WCDB. To  make the  data  more consistent with the
          WCDB, it  was necessary to condense  the ISDB list of waste
          forms to  the six WCDB forms listed  in Section  D.2.2.4.1.
          This task  was  straightforward with  most categories.

     •    The  ISDB  contains  confidential business information.  To  use
          the  ISDB  waste characterization,   its confidential data had
          to  be made nonconfidential  beforehand.  As a result, the
          printout  frequently  did not identify RCRA  D, K,  P, and U
          waste codes.   For  example,  instead  of printing "K054," ISDB
          used "KXXX."   It was  possible to determine that  DXXX repre-
          sented  D004 to  D017  because ISDB did list  D001,  D002, and
          D003.   However, the  large number of K, P,  and  U  waste codes
                                    D-32

-------
          would not permit use of protected ISDB KXXX, PXXX, and UXXX
          compositional data as used for DXXX.  Thus, this led to an
          increased use of default compositions by the SAM.

     •    The percent composition of waste stream constituents was
          sometimes listed as "unknown."  In these cases, their con-
          centrations were designated as zero because the other con-
          stituents with known concentrations typically added up to
          nearly 100 percent.  This was considered to have a minimal
          impact on the SAM results.

     •    The number of participants in the ISDB program was small.
          However, the ISDB was considered the most thorough and accu-
          rate of the five data base sources and therefore was used in
          many respects such as in the development of D code default
          compositions.

     •    The waste constituents were often nonspecific, i.e., the
          ISDB listed constituents as generic chemicals such as "amino
          alkane."  In these cases, a common chemical considered
          representative of the generic chemical was chosen so that
          the SAM could assign volatility and biodegradation classes
          to the constituent.  Therefore, the presence of the generic
          chemical classes in the WCDB decreases the SAM's certainty
          of predicting appropriate emissions from that class.

D.2.2.6  New Field Test Data.

     D.2.2.6.1  Data base description.  This data base is a collection of

waste composition data developed from the review of a hazardous waste TSDF

process sampling report^S and petroleum refining test data from the OSW

listing program.  It contains waste data from three industries:

     •    Petroleum refining (SIC 2911)

     •    Electroplating,  plating, polishing,  anodizing, and coloring
          (SIC 3471)

     •    Aircraft parts and auxiliary equipment,  not elsewhere
          classified (SIC 3728).

This data base contains detailed information from specific TSDF
sites.29,30,31  The petroleum refining data were collected as part of the

Industry Studies survey;  however,  they were not accessible through the
ISDB.

     D.2.2.6.2  Use of the data base.   The WCDB contains this data file's

waste compositions.   It uses the file's SIC code,  waste code, and waste
                                   D-33

-------
 form  to  track  and  identify compositions.  The data file contains the nine
 waste streams  listed  in Table D-6.
      D.2.2.6.3  Data  base limitations.  The two sampling reports and the
 petroleum  refining test data used to create the field data base did not
 always label waste stream information with RCRA waste codes.  Therefore, it
 was necessary  to assign waste codes and waste forms to stream compositions
 based on the reports' descriptions of sampling points and waste composi-
 tions.   This may limit the certainty that the SAM uses the most representa-
 tive  waste compositions for waste codes.
      The specific organic constituents for these nine streams were so
 numerous and so small in concentration that it was decided to reduce the
 chemicals to the following categories:
      •    Total paraffins
      •    Total aromatic hydrocarbons
      •    Total halogenated hydrocarbons
      •    Total oxygenated hydrocarbons
      •    Total unidentified hydrocarbons (includes oil)
      •    Total nonmethane hydrocarbons.
 Some  of these categories were already present in the TSDF chemical uni-
 verse.  Unidentified hydrocarbons proved to be the largest concentration
 category among waste streams because of their oil  content.
      D.2.2.7  Illinois EPA Data Base.
      D.2.2.7.1  Data base description.  Before an  Illinois TSDF can accept
 RCRA wastes,  they must obtain a permit from the Illinois  EPA's Division of
 Land/Noise Pollution Control.  For each waste, the applicant must detail
 its generation activities and provide analysis of  each waste.  The Illinois
 EPA has compiled this permit information in a data base.   It contains waste
compositions  for RCRA hazardous and special nonhazardous  waste streams from
 large quantity generators (>1,000 kg generated per month)  in the State of
 Illinois  and  other States that ship wastes to Illinois TSDF for management.
The data  base used contained 35,000 permits.
                                   D-34

-------
       TABLE D-6.  WASTE STREAMS BY INDUSTRY IN THE FIELD TEST DATA3
SIC code
3471
3728
2911
2911
2911
2911
2911
2911
2911
Industry
Electroplating
Aircraft Parts
Petroleum Refining
Petroleum Refining
Petroleum Refining
Petroleum Refining
Petroleum Refining
Petroleum Refining
Petroleum Refining
Waste codeb
D002
D002
D002
D006
D007
K048
K049
K051
K052
Waste formc
3XXd
3XXd
3XXd
2XX
2XX
5XX
5XX
5XX
2XX
 SIC = Standard industrial classification.
WCDB = Waste Characterization Data Base.

aThis table summarizes those waste streams compiled in a data base of field
 test results. 32, 33  it reflects the industry tested and the waste code/form
 combinations tested and notes decisions made on how to use the data as part
 of the WCDB.
       codes listed in 40 CFR 261,  Identification and Listing of Hazardous
 Waste, Subpart C, Characteristics  of Hazardous Waste, and Subpart D, Lists
 of Hazardous Wastes.34

cPhysical/chemical waste forms are  coded as follows:

 1XX = Inorganic solid                          4XX = Organic liquid
 2XX = Aqueous sludge                           5XX = Organic sludge
 3XX = Aqueous liquid                           6XX = Miscellaneous.

dThe field data contained only a very small percentage of organic
 constituents; therefore, these organics were inserted into the existing
 WCDB compositions, normalizing the original35 organics to maintain the
 original total organic percent composition.
                                   D-35

-------
     D.2.2.7.2  Use of the data base.   The Illinois EPA data used for this

program contained the following information pertinent to the WCDB:

     •    Generator SIC code (most of the codes on file were assigned
          by the State)

     •    RCRA waste code(s)

     •    Physical phase of waste

     •    Waste composition (states whether the waste was organic or
          inorganic)

     •    Key waste stream constituents by name and percent composi-
          tion.

     A total of about 4,000 SIC code/waste code combinations were evaluated

for incorporation into the WCDB.  These 4,000 records reflect over 250 SIC
codes.

     D.2.2.7.3  Data base limitations.   The Illinois EPA data expanded the

volume and quality of information used  in the WCDB.  However, certain limi-

tations were noted when the data were collected and organized:

     •    Only those permits listing RCRA waste codes were used in the
          WCDB.  (This excluded the special nonhazardous wastes and
          hazardous waste permits with  incomplete or no RCRA waste
          codes.)   This ensures that only the most accurate waste data
          are used.

     •    Only Illinois waste permits listing just one RCRA code were
          incorporated into the WCDB.  A large number of Illinois EPA
          permits  contained more than one RCRA waste code.  This deci-
          sion decreased the usage of the Illinois EPA data, but those
          data used were considered higher in quality.

     •    Only those permits for which  SIC codes could be identified
          were incorporated into the WCDB, for without SIC codes a
          waste composition cannot be properly assigned to its  most
          appropriate generating industry.  Most of the SIC codes
          found in the Illinois EPA data base were assigned by  the
          State,  not the waste permit applicant.  All remaining
          records  that were missing SIC codes were identified.   A list
          of these records was printed  by generator name.  Dun  and
          Bradstreet's 1986 Million Dollar Directory^ was researched
          to identify as many generators by company name and SIC code
          as possible.  However,  it was not possible to identify all
          of the  companies'  codes.  Only those permits for which SIC
          codes could be identified were incorporated into the  WCDB.
                                   D-36

-------
     D.2.2.8  RCRA K Waste Code Data Base.
     D.2.2.8.1  Use of the data base.  The original  K waste code data base
developed by Environ^? describes these codes in terms of waste stream
constituents, constituent concentrations,  and other waste characteristics
such as specific gravity and reactivity or ignitibility.  The data base was
derived from a combination of RCRA listing background documents, industry
studies, and open literature.  Thus, it generally provides a range of con-
centrations for any given constituent in a waste stream.
     A representative concentration for each constituent in a waste stream
was needed to develop waste stream characteristics and calculate emissions.
Because the Environ data base reported varying compositions from various
sources, Radian^ selected representative constituent concentrations from
the ranges provided in that data base.  The WCDB uses this file of repre-
sentative constituent concentrations for the SAM.  For example, a mean
would be used for a range of concentrations originating from one data
source.  However, if the waste data came from two or more sources, a more
elaborate procedure was necessary to determine representative constituent
information.  For waste data from two sources, Radian chose the highest
concentration of each constituent found in the two sources and then normal-
ized the waste composition to 1,000,000 parts.  This may have resulted in
above-average concentrations of constituents; however, the approach was
selected to ensure that at least a representative average concentration was
identified.  For waste with three or more data sources, a check was made
for outlying values,  and the remaining data were averaged to obtain repre-
sentative constituent concentrations if no mean were provided.
     D.2.2.8.2  K Stream data base limitations.  Although this data base
contained compositional information on each RCRA K stream, it had two limi-
tations:
     •     Some stream compositions totaled less than 100 percent and
          were therefore incomplete.  In such cases, the WCDB con-
          sidered the unidentified components inorganic.
     •     Some waste  constituents appeared as generic chemical
          constituents, e.g.,  "other chlorinated organics."  Volatil-
          ity and biodegradation classes were designated for those
          generic constituents by referencing a common chemical con-
          sidered representative of that generic constituent.
                                   D-37

-------
     D.?.?.9  WET Model Data Base.
     D.?.2.9.1  Data base description.  This data base contains 267 waste
streams.  Data collection for this data base concentrated on industry sec-
tors where the impact of the RCRA land disposal regulations may be most
significant.  Based on the preliminary regulatory impact analysis  (RIA) for
the land disposal regulations,39 those industry sectors potentially
impacted to the greatest degree and included in this data base are:
     •    Wood preserving (SIC 2491)
          Alkalies and chlorine (SIC 2812)
     •    Inorganic pigments (SIC 2816)
     •    Synthetic organic fibers (SIC 2823, 2824)
     •    Gum and wood chemicals (SIC 2861)
          Organic chemicals (SIC 2865, 2869)
     •    Agricultural chemicals (SIC 2879)
          Explosives (SIC 2892)
          Petroleum (SIC 2911)
          Iron and steel (SIC 331, 332)
     •    Secondary nonferrous metals (SIC 3341)
     •    Copper drawing and rolling (SIC 3351)
          Plating and polishing (SIC 3471, 3479).
     The WET Model study investigated the appropriate level of control for
various hazardous wastes by characterizing a manageable number of waste
streams, a process requiring a considerable amount of approximation and
simplification.  This process achieved two major objectives.
     The approach to waste characterization was to develop a series of
comprehensive profiles for each hazardous waste stream using available
data.   In many cases,  these profiles were developed  from partial informa-
tion using processes of approximation and extrapolation.
     D.2.2.9.2  Use of the data base.  The WCDB uses the following WET
data:
                                   D-38

-------
       •     SIC  code

       •     RCRA waste code

       •     Phase description,  i.e., composition in terms of oil, non-
            aqueous liquids, water, and solids content

       •     Constituent concentrations.

       D.2.2.9.3  WET data base limitations.  The quality of the available
 data  varied greatly and, in general,  was not as adequate for the WCDB as

 other data bases for several reasons.  Among the reasons are the following:

       •     Nontoxic hazardous wastes are excluded from the data base
           because the model is capable of assessing only the toxicity
           hazard.   Therefore,  waste compositions exclude nontoxic,
           volatile organics.

      •    Waste compositions may total less than 100 percent because
           the data might have been incomplete for particular waste
           streams  due to lack of available source material,  either  in
           absolute terms or in the time  frame of this project.   Thus,
           missing  waste  constituents  were considered inorganic.

      •    Data  availability also might have been  limited for particu-
           lar industries where there  were few generators,  e.g.,  in  the
           pesticide  industry.

      •    The data might have  been  imprecise in  the  recording of
           specific  information,  e.g.,  the reporting  of total  chromium
           with  no  quantitative information  on  the  concentration of
           hexavalent chromium, which  is  by  far the more  toxic agent.40

      Because  of the variability  in the data  quality  for  constituent  con-

 centration, this data base  was considered of lesser  quality than others
 and,  therefore, used less.

      D.2.2.10   WCDB Waste Composition Defaults.  As  previously stated, the
 ISDB,  WET,  K stream, Illinois  EPA, and field data bases were used primarily

 to provide  waste stream  constituents and their percent of the stream's
 composition.  Although these data bases were extensive, they did not

 address each and every SIC code/waste code/form combination found in the

Westat Survey data base.   Therefore,  default waste compositions were

developed to fill  these data gaps.  This  section explains how these  default
compositions were  developed.
                                   D-39

-------
     The existing ISDB D code compositions were used to develop default
compositions for each combination of DQOl/waste form,  D002/waste form,
D003/waste form, and DXXX (i.e.,  D004-D017)/waste form.  For example,  if
the ISDB had compositions of D001/4XX from four SIC codes,  the four sets of
compositions were composited to create one D001/4XX default composition.
Each time the SAM finds a TSDF managing D001/4XX whose SIC  code does not
contain the waste stream in the existing data sources, the  stream is
assigned the default composition.
     It was also necessary to develop default compositions  for F code/waste
form combinations not in the existing data bases.   The distribution of
constituents for each of the following F streams was derived from a back-
ground document41 to the 40 CFR 261 regulations that provides consumption
data on those chemicals found in RCRA waste codes F001 to F005.
     For F001,'halogenated degreasing solvents, the background document
states that trichloroethylene is the solvent used most prevalently.42
Unlike F002 to F005, there is no summary of F001 consumption by specific
chemical solvent.  Therefore, trichloroethylene serves as the solvent each •
time an F001 code appears in the TSDF data base.
     The consumption data in the background document provided a percentage
solvent distribution for waste codes F002 to F005, as shown in Table D-7.
     Although a single waste code stream would not contain all of the
chemicals listed, the distribution shown in Table D-7 allows one to address
all chemicals in a manageable way.
     Once the distribution of solvents among waste codes was completed,  it
was necessary to assign compositions by waste  form, e.g.:
     Waste  form _XX  Waste code  F	%  Solvents 	  _% Solvent  1
                                                    	  _% Solvent  2
                                                    	  _% Solvent  3
                                                    	  Js Solvent  4
For waste forms  1XX  (inorganic solid)  and  2XX  (aqueous  sludge),  general
wastewater  engineering  principles4^ were  applied:
                                    D-40

-------
       TABLE D-7.  PERCENTAGE DISTRIBUTION FOR WASTE CODES F002 TO F005a
Quantity of chemical
consumed as solvent annually
Solvent waste codesb and (ca. 1980), Percent
respective chemicals 1Q3 Mg/yr consumption
F002/Tetrachl oroethy 1 ene
Methyl ene chloride
Trichloroethene
Trichloroethane
Chlorobenzene
Trichlorotrifluoroethane
Dichlorobenzene
Trichlorofluoromethane
F003/Xylene
Methanol
Acetone
Methyl isobutyl ketone
Ethyl acetate
Ethanol
Ethyl ether
Butanol
Cyclohexanone
F004/Cresols
Nitrobenzene
F005/Toluene
Methyl ethyl ketone
Carbon disulfide
Isobutanol
Pyridine
255.8
213.2
188.2
181.4
77.1
24.04
11.8
9.072
489.9
317.5
86.2
78.0
69.9
54.43
54.43
45.36
9.072
11.8
9.072
317.5
202.3
77.1
18.6
0.907
26.6
22 2
C. £_ • £.
19 6
A J • \J
18 9
JUJ * j
8 0
w • \J
2 5
b • *J
1 2
A • ^
0.9
40 7
t \J • /
26 3
*_W • tj
7 2
1 • f-
6 5
' U • >J
5 ft
J • O
4 5
~ • *J
4 5
~ • *J
3 a
«J • vj
0.8
56 5
*J\J • O
43.5
51 5
•ij j. • *j
32 8
*J Cm • O
19 C
1 L, • Ji
3 0
•J • \J
0.2
        .                 	 ---,- of solvents in 1980.43  The percent
 usage of each solvent with a waste code is estimated based on the 1980 data.

bWaste codes listed in 40 CFR 261.31, Hazardous wastes from non-specific
 sources.^
                                   D-41

-------
     •    Raw domestic wastewater is 0.07 percent solids.
     •    Digested domestic sludge is 10 percent solids.
     •    Vacuum-filtered sludge is 20 to 30 percent solids.
These principles were used, along with data from a RCRA land  disposal
restrictions background document,^ which show that as much as 20 percent
of the F codes in aqueous liquid (3XX) form are solvents.   The same docu-
ment was used to determine waste compositions for waste forms 4XX (organic
liquid) and 5XX (organic sludge/solid).  This'document contains generic WET
Model streams and their compositions for each of the three waste forms.
     Table D-8 provides the default compositions developed for waste
streams F001 to F005.  In Table D-8, the waste stream constituent "water"
may potentially contain oil.
     Default compositions for all P and U code waste streams  are designated
90-percent pure with 10 percent water when present in the  natural physical/
chemical form of the P and U chemical.  A 90-percent purity is assumed
given the nature of the regulatory listing, i.e., any commercial chemical
product, manufacturing chemical intermediate, off-specification product, or
intermediate (40 CFR 261.33).  This manner of listing implies how close to
purity the waste chemical is.49
     D.2.2.11  Organic Concentration Limits.  During the development of the
WCDB, it was found that respondents to the Westat Survey often listed RCRA
waste codes as aqueous liquids and sludges when the codes  themselves were
described in 40 CFR 261 as organic by nature, e.g., F001--spent halogenated
solvents and organic K, P, and U waste codes.  These occurrences of aqueous
listings indicated that the concentrated organic compositions commonly
found in the WCDB were not representative of the waste code in a dilute
aqueous form and could cause an overestimation of emissions.   Also, in
reviewing ISDB data for D waste codes, it was noted that the organic con-
tent of aqueous liquids and sludges was related to the type of management
process (e.g., total organic concentrations for wastewaters managed in
uncovered tanks and impoundments were typically lower than those managed in
enclosed units such as underground injection wells).  These issues led to
the derivation of organic concentration limits for those wastes described
above.  These limits are presented in Table D-9.
                                    D-42

-------
  TABLE D-8.  DEFAULT STREAM COMPOSITIONS FOR WASTE CODES F001 TO F005a
Waste
Waste formc
  Composition, % constituent
   F001
    1XX



    2XX



    3XX


    4XX


    5XX


    6XX
15.00% Trichloroethylene
60.00% Water
25.00% Solids

18.00% Trichloroethylene
72.00% Water
10.00% Solids

20.00% Trichloroethylene
80.00% Water

60.00% Trichloroethylene
40.00% Water

20.00% Trichloroethylene
80.00% Solids

NA
   F002
    1XX
                         2XX
60.00% Water
25.00% Solids
 3.99% Tetrachloroethylene
 3.33% Methylene chloride
 2.94% Trichloroethylene
 2.84% Trichloroethane
 1.20% Chlorobenzene
 0.38% Trichlorotrifluoroethane
 0.18% Dichlorobenzene
 0.14% Trichlorofluoromethane

72.00% Water
10.00% Solids
 4.79% Tetrachloroethylene
 4.00% Methylene chloride
 3.53% Trichloroethylene
 3.40% Trichloroethane
 1.44% Chlorobenzene
 0.45% Trichlorotrifluoroethane
 0.22% Dichlorobenzene
 0.16% Trichlorofluoromethane
See notes at end of table.
                                        (continued)
                                  D-43

-------
                          TABLE D-8 (continued)
Waste code13
Waste formc
Composition,  % constituent
   F002  (con.)
                         3XX
                         4XX
                         5XX
                         6XX
                     80.00%
                                           5.
                                           4.
                                           3,
                                           3.
                                           1,
                       .32%
                       .44%
                       .92%
                       .78%
                       .60%
                      0.50%
                      0.24%
                      0.18%

                     40.00%
                     16.00%
                     13.30%
                     11.80%
                     11.30%
                       .80%
                       .50%
                      0.72%
                      0.54%
                                           4.
                                           1,
                     80.00%
                      5.32%
                      4.44%
                                             92%
                                             78%
                      1.60%
                      0.50%
                      0.24%
                      0.18%

                     NA
     Water
     Tetrachloroethylene
     Methylene chloride
     Trichloroethylene
     Trichloroethane
     Chlorobenzene
     Trichlorotrifluoromethane
     Dichlorobenzene
     Trichlorofluoromethane

     Water
     Tetrachloroethylene
     Methylene chloride
     Trichloroethylene
     Trichloroethane
     Chlorobenzene
     Trichlorotrifluoromethane
     Dichlorobenzene
     Trichlorofluoromethane

     Solids
     Tetrachloroethylene
     Methylene chloride
     Trichloroethylene
     Trichloroethane
     Chlorobenzene
     Trichlorotrifluoromethane
     Dichlorobenzene
     Trichlorofluoromethane
F003 1XX 60
25
6
3
1
0
0
0
0
0
0
.00%
.00%
.10%
.94%
.08%
.98%
.87%
.68%
.68%
.57%
.12%
Water
Solids
Xylene
Methanol
Acetone
Methyl isobutyl
Ethyl acetate
Ethyl benzene
Ethyl ether
Butanol
Cyclohexanone





ketone





See notes at end of table.
                                        (continued)
                                  D-44

-------
                           TABLE  D-8  (continued)
Waste code^ Waste formc Composition, % constituent
F003 (con.)
2XX 72.
10.
7.
4.
1.
1.
1.
0.
0.
0.
0.
3XX 80.
8.
5.
1.
1.
1.
0.
0.
0.
0.
4XX 20.
32.
21.
5.
5.
4.
3.
3.
3.
0.
5XX 80.
8.
5.
1.
1.
1.
0.
0.
0.
0.
6XX NA

00%
00%
33%
73%
30%
17%
04%
81%
81%
68%
14%
00%
14%
26%
44%
30%
16%
90%
90%
76%
16%
00%
60%
04%
76%
20%
64%
60%
60%
04%
64%
00%
14%
26%
44%
30%
16%
90%
90%
76%
16%


Water
Solids
Xylene
Methanol
Acetone
Methyl isobutyl
Ethyl acetate
Ethyl benzene
Ethyl ether
Butanol
Cyclohexanone
Water
Xylene
Methanol
Acetone
Methyl isobutyl
Ethyl acetate
Ethyl benzene
Ethyl ether
Butanol
Cyclohexanone
Water
Xylene
Methanol
Acetone
Methyl isobutyl
Ethyl acetate
Ethyl benzene
Ethyl ether
Butanol
Cyclohexanone
Solids
Xylene
Methanol
Acetone
Methyl isobutyl
Ethyl acetate
Ethyl benzene
Ethyl ether
Butanol
Cyclohexanone







ketone









ketone









ketone









ketone






See notes at end of table.
(continued)
                                  D-45

-------
                          TABLE D-8 (continued)
Waste code*3
Waste formc
  Composition,  % constituent
   F004
    1XX
                         2XX
                         3XX
                         4XX
                         5XX
                         6XX
60.00% Water
25.00% Solids
 8.48% Cresols
 6.52% Nitrobenzene
                     72.00% Water
                     10.00% Solids
                     10.17% Cresols
                      7.83% Nitrobenzene

                     80.00% Water
                     11.30% Cresols
                      8.70% Nitrobenzene

                     20.00% Water
                     45.20% Cresols
                     34.80% Nitrobenzene

                     80.00% Solids
                     11.30% Cresols
                      8.70% Nitrobenzene

                     NA
   F005
     1XX
                         2XX
                         3XX
60.00% Water
25.00% Solids
 7.72% Toluene
 4.88% Methyl ethyl ketone
 1.88% Carbon disulfide
 0.45% Isobutanol
 0.03% Pyridine

72.00% Water
10.00% Solids
 9.27% Toluene
 5.90% Methyl ethyl ketone
 2.25% Carbon disulfide
 0.54% Isobutanol
 0.04% Pyridine
                      80.00% Water
                      10.30% Toluene
                            Methyl ethyl  ketone
                            Carbon disulfide
                            Isobutanol
                            Pyridine
                                           6.56
                                           2.50
                                           0.60
                                           0.16
 See  notes  at  end  of  table.
                                         (continued)
                                   D-46

-------
                          TABLE D-8 (continued)
Waste codeb          Waste formc            Composition, % constituent
   F005 (con.)
                         4XX              20.00% Water
                                          41.20% Toluene
                                          26.20% Methyl ethyl ketone
                                          10.00% Carbon disulfide
                                           2.40% Isobutanol
                                           0.16% Pyridine

                         5XX              80.00% Solids
                                          10.30% Toluene
                                           6.56% Methyl ethyl ketone
                                           2.50% Carbon disulfide
                                           0.60% Isobutanol
                                           0.16% Pyridine

                         6XX              NA
NA = Not applicable.

aThis table presents default waste stream compositions derived from WET
 model waste stream data4? for wastewaters containing solvents and for
 organic liquids containing solvents.   These defaults are used by the
 Source Assessment Model  when Standard Industrial  Classification code/
 waste code/waste form combinations are not found  elsewhere in the Waste
 Characterizaton Data Base.
       codes listed in 40 CFR 261.31,  Hazardous wastes from non-specific
 sources. 48

cPhysical/chemical  waste forms are coded as follows:

 1XX = Inorganic solid                    4XX = Organic liquid
 2XX = Aqueous sludge                     5XX = Organic sludge
 3XX = Aqueous liquid                     6XX = Miscellaneous.
                                  D-47

-------
              TABLE D-9.   CONCENTRATION LIMITS ASSUMED IN SOURCE
       ASSESSMENT MODEL (SAM)  FOR ORGANIC CONCENTRATIONS IN WASTEWATERS
                             AND AQUEOUS SLUDGES9
Waste codeb
P 	 c
U 	 c
F001-F005
K 	 c,e
D001c.f
D002f
D003f
D004 and greater0 »f
Organic
Wastewaters
(waste form 3XX)
1%
1%
l%d
1%
5%
0.4%9
6%c
0.1%
concentration limit, %
Aqueous si
(waste form
1%
1%
l%c
1%
5%
0.4%c
6%9
0.1%

udges
2XX)








aThis table shows the maximum concentration the SAM assumes for organics
 when estimating emissions from wastewaters and aqueous sludges.  These
 assumptions are conditional  as described in the footnotes below and in
 Section D.2.2.11.

bWaste codes listed in 40 CFR 261,  Identification and Listing of Hazardous
 Waste, Subpart C, Characteristics  of Hazardous Waste,  and Subpart D,  Lists
 of Hazardous Wastes.50

cSource:  Best engineering judgment based on review of waste code descrip-
 tions.  (Nonconfidential Industry  Studies Data Base data are inadequate or
 do not exist.)

^Source:  Land disposal restrictions regulatory impact analysis.51

Concentration limits apply only to K waste codes that are organic by nature
 of their listing, e.g, organic still bottoms and organic liquids.  These
 limits do not apply to K waste codes that are listed as inorganic solids or
 aqueous sludges or liquids in 40 CFR 261.32.52

fConcentration limits apply only to aqueous liquids and sludges of RCRA D
 waste codes managed in open units, i.e., storage,  treatment, and disposal
 impoundments and open treatment tanks.

QSource:  EPA data analysis of nonconfidential Industry Studies Data Base
 data.
                                     D-48

-------
     Sections D.2.2.11.1 through D.2.2.11.4 discuss these limits on organic
content.
     D.2.2.11.1  F001 to F005 (spent solvent).  During the development of
the proposed land disposal restriction rules for solvents and dioxins,53
EPA/OSW analyzed waste composition data from a number of sources including
the ISDB.  The results of this analysis showed a median solvent concentra-
tion in wastewater (an aqueous liquid) of 0.05 percent and a mean of 0.3
percent.
     The 1981 Westat Survey54 identified greater than 99 percent of the
solvent waste treated in surface impoundments as a wastewater form of the
solvent.  The land disposal restriction Regulatory Impact Analysis did not
provide a typical waste composition of solvents in these wastewaters;
however, it did state that solvent constituent concentrations in F001 to
F005 wastes may be "as little as one percent or less (if present at
all)."55  por these reasons, a limit of 1 percent was set on solvents found
in wastewater.  The 1-percent limit was also assigned to aqueous sludges.
     D.2.2.11.2  Organic P, U, and K wastes.  It was also decided to assign
1-percent organic concentration limits to aqueous liquids and sludges of
organic P, U, and K wastes because of the decisionmaking used for solvents
F001 to F005.  Given that these P, U, and concentrated organic K wastes are
just as concentrated as solvent wastes (based on their normal listing as
organic liquids or sludges), their dilution to 1 percent or less in waste-
water or aqueous sludges should be comparable to the solvents in F001 to
F005.  Many of these organics also may be insoluble in water and are
decanted from the wastewater before it enters the open management unit.
Therefore, a 1-percent organic concentration limit was assigned to these
waste codes when they occur as wastewaters or aqueous sludges.
     D.2.2.11.3  D001.  This limit reflects the minimum concentration of an
ignitible organic in water that causes the water to exhibit an ignitible
characteristic.  Based on engineering judgment, the organic concentration
limit designated for D001 is 5 percent.  For example, an ignitible organic
liquid (about 100 percent organic) has a heat value of about 30,000 J/g; an
aqueous liquid containing 10 percent i-gnitible organic may have a heat
                                    D-49

-------
value of 3,000 J/g and thus still be burnable; however,  an aqueous liquid
with 1 percent ignitible organic will not be ignitible because the heat
value is 300 J/g.  As another example, ignitible methanol can have a
concentration in water between 2 and 10 percent and the water remains
ignitible.  Less than 1 percent would not be ignitible.   This range of 1 to
10 percent was used to arrive at an average minimum concentration of an
ignitible organic in wastewater that yields an ignitible aqueous liquid,
i.e., 5 percent.
     D.2.2.11.4  D002. D003. and D004 to D017  (DXXX).  Concentration limits
were established for these waste codes using the ISDB.  The  ISDB was
searched to identify D002, D003, and D004 to D017 waste codes that were
either aqueous liquids (wastewaters) or sludges and were managed in storage
surface impoundments, onsite wastewater impoundments, or onsite wastewater
tanks.  Each  of  these management devices was considered open to the atmos-
phere.  Once  these waste  compositions were  found, a weighted average was
taken for each waste code managed  in  these  open units based  on quantity
managed for each waste code/waste  form combination.  These weighted aver-
ages  serve  as organic concentration  limits  for the  open waste management
units.
D.2.3   Chemical  Properties
      0.2.3.1   Introduction.   Emission estimation on  a constituent  basis  for
each  of the more than  4,000  TSDF waste constituents  identified  in  the  data
bases was  not possible  because of  a lack  of constituent-specific  data  and
because of  the  large number of chemicals  involved.   Therefore,  to provide
the emission  models  with  relevant  physical, chemical,  and biological
properties  that  influence emissions and  still  maintain  a workable and
efficient method of  estimating emissions,  waste constituent  categorization
was required.  Waste constituent categorization .allows  the SAM to make
 emission estimates for all  constituents  by making  emission estimates  for a
 set of chemicals (surrogates) that represent the universe of organic
 chemicals that occur in hazardous waste  streams.
      D.2.3.2  Haste Characteristics Affecting Emissions.  In the develop-
 ment of air emission models for hazardous waste TSDF,  the means by which
 organic compounds escape to the environment from TSDF was determined.   It
                                     D-50

-------
 was  found  that  the  fate  of  organic  compounds  in  surface impoundments,  land
 treatment  facilities,  landfills,  wastepiles,  or  wastewater treatment  (WWT)
 plant  effluents  can  be affected  by  a  variety  of  pathway mechanisms,  includ-
 ing  volatilization,  biological decomposition,  adsorption,  photochemical
 reaction,  and hydrolysis.   The relative  importance  of  these pathways for
 TSDF waste management  processes  was evaluated  based  on  theoretical consid-
 erations,  data appearing  in the  literature, and  engineering judgment.  The
 predominant removal  pathways for  organic compounds  at  TSDF sites were  found
 to be  volatilization and  biodegradation.  For  this  reason,  the emission
 models used for  TSDF in the air  emission models  report5^ are all based on
 volatilization and/or  biodegradation  as the principal  pathways included in
 the  models.  Volatilization occurs when molecules of a  liquid or solid
 substance  escape to an adjacent gas phase.  Biodegradation  takes place when
 microbes break down organic compounds for metabolic processes.
     Several waste characteristics contribute to the potential for a waste
 constituent to be volatilized or released to the atmosphere.  Major factors
 include the types and number of hazardous constituents present,  the concen-
 trations of these constituents in the waste, and the chemical and physical
 characteristics  of the waste and  its constituents.  In conjunction with the
 type of management unit,  the physical  and chemical properties of the waste
constituents will affect  whether  there will  be pollutants released and  what
form the release will take (i.e., vapor,  particulate, or particulate-
associated).   Important physical/chemical  factors to consider when assess-
ing the volatilization  of a  waste constituent  include:
          Water  solubility.   The  solubility  in water indicates the maxi-
          mum  concentration  at  which a constituent can  dissolve  in water
          at  a given  temperature.   This  value  can be used to estimate the
          distribution  of  a  constituent  between the  dissolved aqueous
          phase  in the  unit  and the  undissolved solid or immiscible
          liquid  phase.   Considered  in combination with the constituent's
          vapor  pressure,  solubility can  provide  a relative assessment  of
          the potential for  volatilization of  a constituent from  an aque-
          ous environment.
          Vapor pressure.  This property  is  a  measure of the pressure of
         vapor  in equilibrium with  a  pure liquid.   It  is best used in  a
         relative sense as  a broad  indicator  of  volatility;  constituents
         with high vapor  pressures  are more likely  to  be released  than
         are those with low vapor pressures,  depending  on  other  factors
                                   D-51

-------
         such as relative solubility and concentration  (e.g.,  at  high
         concentrations,  release can occur even  though  a constituent's
         vapor pressure is relatively low).

    .    Octanol/water partition coefficient.   The octanol/water
         partition coefficient indicates the tendency of an organic
         constituent to absorb to organic components of soil  or waste
         matrices.  Constituents with high octanol/water partition coef-
         ficients tend to adsorb readily to organic carbon, rather than
         volatilize to the atmosphere.  This is particularly important
         in landfills and land treatment units,  where high organic car-
         bon content in soils or cover material  can significantly reduce
         the release potential of volatile constituents.

    •    Partial pressure.  A partial pressure measures the pressure
         that each component of a mixture of liquid or solid substances
         will exert to enter the gaseous phase.  The rate of volatiliza-
         tion of an organic chemical when either dissolved in water or
         present in a solid mixture is characterized by the partial
         pressure of that chemical.   In general, the greater the partial
         pressure, the greater the potential for release.  Partial
         pressure values  are unique for any given chemical in any given
         mixture and may  be difficult to obtain.

     •    Henry's  law constant.  Henry's law constant is the ratio of the
         vapor  pressure  of a constituent to its aqueous solubility  (at
         equilibrium).   This constant can  be used to assess the relative
         ease with which  the compound may  vaporize  from the aqueous
         phase.   It  is applicable  for low  concentration  (i.e., less than
         10  percent) wastes  in  aqueous  solution and will be most  useful
         when the  unit being  assessed is  a surface  impoundment or  tank
         containing  dilute wastewaters.   The potential  for significant
         vaporization  increases  as the  value for  Henry's  law  constant
         increases.

     •    Raoult's  law.   Raoult's  law accurately predicts  the  behavior  of
         most  concentrated mixtures of  water and  organic  solvents  (i.e.,
         solutions over  10 percent solute).  According to  Raoult's  law,
         the rate of volatilization of  each chemical  in a  mixture is
         proportional  to the product of its concentration  in  the  mixture
         and its vapor pressure.   Therefore, Raoult's  law can also be
         used to characterize volatilization potential.

     The air emission models  report provides the most up-to-date  guidance

on assessing the volatilization of waste constituents and contains a com-

pilation of chemical/physical properties for several hundred constituents.

     Through review of available literature relating to TSDF emission

modeling, it was judged that  volatility, which  is  an index of emission

potential,  can best be characterized across the entire waste population by
                                    D-52

-------
 either vapor pressure or Henry's law constant depending on the waste
 matrix.  One case accounts for chemical  compounds in situations in which
 Henry's law governs mass transfer from the waste (i.e., low organic concen-
 tration in aqueous solution),  and the other case accounts for chemical
 compounds in those situations  in which mass transfer is governed by vapor
 pressure (i.e.,  concentrated mixtures of organics).
      Three chemical  and biological  properties are therefore critical  in
 estimating TSDF  emissions:   vapor pressure,  Henry's  law constant,  and bio-
 degradation rate.   These were  selected as  the basis  for designating waste
 constituent and  surrogate categories.
      D.2.3.3  Waste  and Surrogate Categorization.
      D.2.3.3.1  Haste properties—physical  and chemical.   Efforts  to
 categorize the universe of  chemical  compounds found  at  hazardous waste
 sites  were based on  information  contained  in  the  CHEMDAT3  data  base.57 The
 60  chemicals and their properties  available  from  this data base, originally
 used  in predicting organic  emissions,  formed  the  basis  for both waste con-
 stituent  categorization  and  surrogate  properties  selection.   Table  D-10
 provides  the primary  data for  the 60  chemicals  used  in  developing  surrogate
 categories  and properties.
     D.2.3.3.1.1  Vapor  pressure  categories.   In  1985,  EPA published  a
 comprehensive  catalog  of physical and  chemical  properties  of  hazardous
 waste  in  relation to  potential air emissions  of wastes  from TSDF.   The
 waste  volatility categorization scheme presented  in the document5^  divided
 vapor  pressures  into  three useful categories:   high  (>1.33  kilopascals
 [kPa]), moderate  (1.33 x 10'4 to  1.33  kPa), and low  (<1.33 x  10'4 kPa).
 Sensitivity  analysis on the  impact of  vapor pressure on emissions pointed
 out that organics that are gases at standard temperature and pressure
 skewed  the average emission  rates for the high vapor pressure chemicals.
 Emission estimates for high vapor pressure chemicals were  dominated by the
 gases;  an average figure would overestimate emissions for most high vapor
 pressure chemicals because gases are relatively few in number among the
 high category chemicals.  Therefore,  compounds with vapor pressures greater
than  101.06 kPa  were segregated into their own "very high" category,
creating four categories of vapor pressure chemicals.  Vapor pressures for
                                    D-53

-------
                           TABLE 0-10.   DATA USED  FOR WASTE CONSTITUENT  CATEGORIZATION AND SURROGATE PROPERTY

                                               SELECTION  IN THE SOURCE ASSESSMENT MODEL".b
o
i
en

Compound name
Acetaldehyde
Methyl ethyl ketona
Toluene
Aery Ion itri le
Pyridine
Phenol
Butanol-1
Dichloroethane (1,2)
Formaldehyde
Cresol (-m)
Cresol (-p)
Cresols
Cresol (-0)
Methylene chloride
Isobutyl alcohol
Methyl acetate
Benzene
Benzyl chloride
Fthyl acetate
Cresy lie acid
Vapor
pressure,
kPa
122
13.3
3.99
15.2
2.02
0.045
0.864
10.6
465
0.01
0.015
0.019
0.032
58.2
1.33
31.2
12.7
0.0093
11.3
0.04
Henry's law
constant,
10 kPa»m /g
mol
9.6
4.4
675
8.9
2.4
0.0469
0.9
6.4
5.8
0.3
0.3
0.3
0.3
322
0.2
NA
556
52.7
12.9
0.2
Surrogate
cateoorv0
Biorate,
mg VO/g/h
82.4
73.8
73.5
44.30
35.03
33.6
32.4
0.302
20.91
23.2
23.2
23.2
22.8
22.00
21.2
19.9
19.00
17.8
17.6
15.00
Biodegradability
category
High
High
High
High
High
High
High
Low
High
High
High
High
High
High
High
High
High
High
High
High
Vapor
pressure
10
1
1
1
1
4
4
3
10

4
4
4
1
4
1
1
4
1
4
Henry's
law
constant
4
4
1
4
7


7




7
1
1
A
7
	 (continued)

-------
                                                            TABLE D-10  (continued)
en
en



Compound name
Acetone
Methanol
Cyclohexanone
Dichl orobenzene (1,2) (-0)
Acrolein
Nitrobenzene
Maleic anhydride
Chloroform
Chi orobenzene
Ethy (ether
Methyl isobutyl ketone
Ally! alcohol
Carbon disulfide
Carbon tetrachlor ide
Chloroprene
Cumene (i sopropy (benzene)
Dichl orobenzene (1,4) (-p)
Dimethyl nitrosamine
Dioxin
Epichlorohydrin
Ethy 1 benzene

Vapor
pressure,
kPa
35.4
15.2
0.64
0.2
32.5
0.04
1.33xl0~5
27.7
1.67
69.1
0.097
3.098
48.7
15.03
36.3
0.612
0.16
NA
NA
2.26
1.33
Henry's law
constant,
•~3 3
Surrogate
cateaoryc

10 kPa»m /g Biorate,
mol
2.5
0.3
0.4
196
5.7
1.3
0.004
3.42
397
68.7
4.4
NA
1,212
3,030
NA
1,480
162
NA
NA
3.3
650
mg VO/g/h
14.6
12.00
11.5
10.00
7.80
0.302
4.08
0.302
1.46
0.77
0.74
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA

Biodegradabi 1 ity
category
High
High
High
High
Moderate
Low
Moderate
Low
Moderate
Low
Low
Moderate
Moderate
Low
Low
Moderate
Moderate
Moderate
Low
Low
High

Vapor
pressure
1
1
4
4
2
6
8
3
2
3
6
2
2
3
3
5
5
NA
NA
3
4
Henry's
law
constant
4
7
7
1
5
6
8
3
2
6
6
8
2
3
NA
2
2
NA
NA
6
1
See notes at end of table. , 	 «.:n...j>

-------
                                                         TABLE D-10  (continuad)
o
en
Compound nama
Ethylene oxida
Fraons
Haxach 1 orobutad i ana
Haxach 1 orocyc 1 opantad iana
Naphtha! ana
N i trosomorpho 1 i na
Phosgene
Phthalic anhydride
Pol /chlorinated biphenyl*
Proplyana oxida
Tatrachloroathana (1,1,2,2)
Tatrach 1 oroethy 1 ana
Vapor
pressure,
kPa
166
NA
0.02
0.0108
0.031
NA
185
0.0002
NA
69.2
0.864
2.63
Trichloro (1,1,2) tr if luoroathana
Trichloroathana (1,1,1)
Tr i ch 1 oroethy 1 ana
Trichlorof luoromathana
Vinyl chloride
Vinylidena chloride
Xylena (-0)
16.4
9.97
105.8
354
78.6
0.931
Henry's law
constant,
10 kPa'm /g
mol
3.7
NA
2,590
1,620
119
NA
17,300
0.1
20.2
19.4
38.4
NA
119
8,081
919
6,890
8,690
19,200
618
Surrogate
cateaoryc
Bi orate,
«ng VO/g/h
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Biodegradabi 1 ity
category
High
Low
Low
Low
Moderate
Low
Low
High
Low
High
Low
Low
Low
Low
Low
Low
Low
Low
High
Vapor
pressure
10
NA
6
6
6
NA
12
4
NA
1
6
3
3
3
3
12
12
3
4
Henry's
law
constant
4
NA
3
3
2
NA
3
7
NA
4
6
3
3
3
3
3
3
3
1
         NA = Not available.
         VO = Volatile  organics.

         •This table provides tha primary  data  for 60 chemicals usad  in developing surrogate categories to ba used  in the
          Source Assessment Modal.
         ^Source of data:   Research Triangle Institute.  CHEMDAT3 Database for Predicting VO Emissions from Hazardous Waste
          Facilities.   Developed  for Office of  Research and Development, U.S. Environmental Protection Agency.  Cincinnati,
          OH.  1986.58
         cRefers to surrogate categories defined in  Table D-ll.

-------
 the 60 reference chemicals were obtained from or estimated using methods
 commonly found in engineering and environmental science handbooks.60-61.62
      D.2.3.3.1.2  Henry's law categories.  The Henry's law constant is a
 measure of the diffusion of organics into air relative to diffusion through
 liquids.  Henry's law constants are generated using vapor pressure, molecu-
 lar weight, and solubility.  Henry's law is used in predicting emissions
 for aqueous systems.  An analysis to determine the effects of Henry's law
 constant on the organic fraction emitted to air,  using the TSDF air
 emission models,  was used in establishing Henry's law constant categories.
 Results showed discernible patterns in  the relationship between the organic
 fraction emitted  and Henry's law constant.   The fraction emitted begins  to
 drop sharply for  low values of Henry's  law constant (<10-3 kPa m3/g mol)  as
 the mass transfer becomes affected by both  gas and liquid phase control.
 When Henry's law  constant is greater than 10'1 kPa m3/g mol,  rapid  vola-
 tilization  will generally occur.   A number  of citations found in the
 literature  support  the Henry's law constant volatilization categories
 selected.63,64  Henry's  law constants were  grouped as  follows:
       •   High        >10-!,  kPa m3/g  mol
       •   Moderate    10'1  to  10-3,  |10 mg
 organics/g  of biomass/h,  moderate  =  1 to  10 mg organics/g/h,  and low =
 <1 mg organics/g/h.  This classification  follows the biorate  designation
 provided with the data base on the 60 chemicals.65  in  some cases, the
 biodegradation rate was inconsistent with values reported  elsewhere for
 measures such as BOD5, soil half-life, and ground-water degradation.  It is
 understood that biodegradability is variable and depends on the matrix,  the
 concentration of organics and microorganisms, and temperature.  However,  to
 provide an "average" biorate that represents all TSDF management processes,
biodegradation rates provided for many of the 60 chemicals were compared  to
other measures of  biodegradation and adjusted if appropriate.
                                    D-57

-------
     0.2.3.3.2  Surrogate categories.   With 4 categories  of vapor pressure,
3 of Henry's law constant,  and 3 of biodegradation,  a chemical  could fall
into one of 12 possible categories of  vapor pressure and  biodegradation
(4 x 3) and into one of 9 categories of Henry's law constant and biodegra-
dation.  These two surrogate groups (i.e.,  vapor pressure surrogates and
Henry's law surrogates) represent two volatility situations:  where vapor
pressure is the mass transfer driving force in one case and where Henry's
law constant best represents or governs mass transfer in  the other.  Table
D-ll provides the definition of surrogate categories.
     D.2.3.3.3  Surrogate properties—physical and chemical.  The chemical
and biological properties selected to represent each surrogate are, gen-
erally, averages for groupings of the 60 chemicals categorized by vapor
pressure/biodegradation and Henry's law constant/biodegradation.  It should
be noted that not all of the possible categories of vapor pressure/bio-
degradation and Henry's  law constant/biodegradation were unique.  The low
vapor  pressure categories were judged to be relatively equivalent; there-
fore,  the  low vapor pressure/moderate biorate  (LVMB) properties were used
for all low vapor pressure compounds.  The  low Henry's law  constant/low
biorate (LHLB) category  was judged  to be very  similar to the low Henry's
 law constant/moderate  biorate  (LHMB) category.  The high vapor pressure/
moderate biorate  (HVMB)  and the  high vapor  pressure/low biorate  (HVLB) were
 also found to be  similar in predicting emissions.   Property values  for all
 surrogate  categories are therefore  not presented.   Tables  D-12  and  D-13
 summarize  the surrogate  properties  for the  vapor  pressure  and  the  Henry's
 law constant  groupings,  respectively.66
     Emissions  for  waste management processes  that  are modeled  using vapor
 pressure draw their surrogate properties from vapor pressure  and  biodegra-
 dation group  averages.   Similarly,  processes  best modeled  by  Henry's  law
 constant draw surrogate  properties  from  the groupings  of  Henry's  law con-
 stant  and  biodegradation.   This is  because the SAM,  as designed,  handles
 only  a single set of emission factors  for  each waste management unit;  for
 example,  only Henry's  law constant  surrogates are used to calculate emis-
 sions  for  surface impoundment operations because emissions from surface
 impoundment wastes  are predominantly Henry's law controlled and because
                                     D-58

-------
   TABLE D-ll.
              DEFINITION OF WASTE CONSTITUENT CATEGORIES  (SURROGATES)
                APPLIED IN THE SOURCE ASSESSMENT MODEL3
Surrogate
category
Vapor Pressure
Surrogates










Henry's Law
Constant Surrogates







1
2
3
4
5
6
7
8
9
10
11
. 12
1
2
3
4
5
6
7
8
9
Constituent properties
Vpb
H
H
H
M
M
M
L
L
L
VH
VH
VH
NA
NA
NA
NA
NA
NA
NA
NA
NA
HLCC
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
H
H
H
M
M
M
L
L
L
Biod
H
M
L
H
M
L
H
M
L
H
M
- L
H
M
L
H
M
L
H
M
L
NA = Not applicable.

aThis table describes the volatility and biodegradation properties of each
 waste constituent (surrogate) category developed for use in the Source
 Assessment Model.

                                     dfiio = Biodegradation rates:

                                            High (>10 mg VO/g biomass/h)
  = Vapor pressure categories:

VH = Very high (>101.06 kPa) .
H  = High (1.33-101.06 kPa) .
M  = Moderate (1.33xlO-4-l .33 kPa) .
L  = Low (<1.33xlO-4 kPa).
                                        H =
                                        M =

                                        L =
                                            Moderate (1-10 mg VO/g
                                            biomass/h).
                                            Low (<1 mg VO/g biomass/h)
CHLC  = Henry's law constants.

       High (>10'1 kPa m3/g mol).
H
M
L
       Moderate (lO-O'  kPa m^/g mol)
       Low (<10'3 kPa m3/g mol).
                                   D-59

-------
                                     TABLE 0-12.   PROPERTIES  FOR  VAPOR PRESSURE AND BIOOEGRADATION GROUPINGS*
                                     AT 26 °C OF  WASTE  CONSTITUENT CATEGORIES  (SURROGATES)  SHOWN IN TABLE D-ll
o
 i
Surrogate Vapor pressure at 26 °C
categoryb M.W. kPa (10"3)
HVHB (1) 73.6 27.4
HVMB (2) 72.6 24.2
HVLB (3) 117.0 34
MVHB (4) 111.0 0.346
MVMB (6) 132.0 0.266
MVLB (6) 186.0 0.386
LVMB (8) 98.0 1.33 x 10~&
VHVHB (10) 39.3 261
VHVLB (12) 80.7 270
M.W. ~ Molecular weight.
VO = Volatile oftgantcs.
HVHB = Higb vapor pressure, high biorate.
HVMB = High vapor pressure, moderate biorate.
HVLB = High vapor pressure, low biorate.
MVHB s Moderate vapor pressure, high biorate.
MVMB '= Moderate vapor pressure, moderate biorate.
MVLB = Moderate vapor pressure, low biorate.
LVMB = Low vapor pressure, moderate biorate.
VHVHB = Very high vapor pressure, high biorate.
VHVLB = Very high vapor pressure, low biorate.
Oiffusivity in water,
em2/s (10-6)
10.6
10.7
9.63
9.02
7.60
7.32
11.1
14.6
11.8











Oiffusivity in air,
cn>2/s (10-3)
98.9
134
89.9
76.8
64.3
66.9
96
101
107











Biorate,
mg VO/g/h
34.30
6.97
0.30
22.60
3.02
0.39
4.08
47.60
0.30











          •Properties presented in this table are averages  for compounds found within a, given category.
           development of this table can be found in a  memorandum to the docket.®^
A detailed discussion on the
          t>Not all  of the 12 possible categories were unique.   The low vapor pressure categories (LVHB,  LVMB,  and LVLB)  were judged to
           be relatively equivalent.  Therefore, the LVMB group properties were used for all  low vapor pressure compounds.   The
           moderate and low biorate categories for the very  high vapor pressure group were also shown to result in similar  emissions;
           therefore, the VHVLB group properties were used for both categories.

-------
                      TABLE D-13.  PROPERTIES FOR HENRY'S LAW CONSTANT AND BIODECRADATION GROUPINGS OF WASTE CONSTITUENT

Surrogat, Diff. W.U|.(
c.Ugory M.W. Cm2/, (10-6)
MHLB (6) 112 0 R A«

HHLB (3) 144.0 9.39
LHMB (8) 78.4 11.3
"HUB (6) 67.0 H 8

HHMB (2) 117.0 8 24

LHHB (7) 97.3 9.54
MHHB (4) 69.9 11.6
HHHB (1) 98.4 9.40
Diff. air, Bior.t.,
cn.2/, (10-3) ,„„ V0/g/h

78.4 0.39
87.6 0.302
180 3.66

H6 11.2

7* 2.71
82.7 23.2
95. 6 40.1
87.3 29.2
XVQC
(10-3)

3.27
2.64
4.66

6.40

3.13
3.78
6.23
3.72
T*mp.ratur, adju.tm.nt «qu.tion°

H
H
H

H

H
H
H
H

= .[(-4879.12/T) *
= .[(-2276.36/T) »
= •[(-11662. 27/T) <

x ,[(-4090. 16/T) +

* .[(-6462.8.7/T) *
= •[(-11662. 27/T) +
= «[(-3266.36/T) +
= •[(-3180. 14/T) »

17.1726]/990
16.6418]/99fl
' «3-l-«]/.0099

16. 13143] /9M

23. 10247] /990
23.14]/" - •* p-vid.a b.c.u..  it ... juda.d ko b. v.Py
       (    Th. w.ight fr.ction  of t ...... og.U (g .Urrog.t./9 w.at.) . W5/W, ... ...ula.< bo b. ,w „ 10.2 f.p .„

              ''                               found in
                                                                                                               to th. LHM8
                                                                                                                           of this
            C°n't'nt unit«
                                   .3/9 «>l.  Th. .qu.tion pr.dict. H.nry.. ,.w COn.t.nt for .  r.ng. of t«,p.r.tur., for ..ch
CH.nry,  ,.w con.t.nt, .t 26 «C  (298 K) ,r. tho ..... d in .mi..ion »«,d.,.; .^ APP.ndix C.

-------
dilute aqueous wastes are typically stored there.   In  the case of Henry's
law constants, surrogate values were not based on  group averages.  For the
surrogate's Henry's law constant,  a single constituent was selected to
represent the surrogate group;  all other surrogate properties are averages
of the group of constituents that fall into the particular surrogate cate-
gory.  This approach was selected in order to generate the temperature-
dependent Henry's law constant equations needed for each surrogate
category.
     D.2.3.4  Assigning Surrogates.  The TSDF Waste Characterization Data
Base  (see Section D.2.2) data sources often provided only generic descrip-
tions of waste constituents, e.g.,  "amino alkane."  Therefore, the first
requirement in assigning a surrogate to the more than 4,000 constituent
chemicals found  in the WCDB was the assignment of specific common chemicals
to  represent  the generic compounds. ' Next, all specific chemicals were
assigned physical, chemical, and  biodegradation values.  Vapor pressures
and Henry's law  constants were  estimated  for  25 °C, if possible.  Vapor
pressure values  were not available  for  a  large fraction of the chemicals.
Vapor pressure assignments were completed by  relating molecular  structure
and molecular weight to  similar chemicals with known  vapor pressures.
Specific solubility  values,  used  to estimate  Henry's  law  constants, were
assigned as  follows  when  qualitative  descriptions were  found  in  the  litera-
ture:69-70  insoluble—2  mg/L,  practically insoluble--10 mg/L,  slightly
 soluble--100  mg/L,  soluble—2,000 mg/L,  very  soluble—10,000  mg/L,  and
miscible—100,000  mg/L.   If no information  was  found  in  the  references,
 solubility values  were estimated  based  on molecular structure.   The molecu-
 lar weight of chemicals was readily available or  determinate,  although
 there was  some judgment required in assigning molecular weight for poly-
 mers.  Biodegradation assignments were  based on  quantitative measures,
 although largely unavailable,  or on a comparison  of molecular structure
 with chemicals well  characterized by biodegradation.71  The  approximate
 breakdown of biodegradation information is shown  in Table D-14.
      The biorate values used for predicting emissions were based on the
 biodegradation  rates for the "high" class of 60 chemicals.  The average
 biodegradation  for  the high category is approximately 30 mg  VO/g biomass/h.
                                     D-62

-------
              TABLE D-14.  CLASSIFICATION OF BIODEGRADATION
Parameter
BOD5
Soil half-life

High
<3 days
Classification
Moderate
1.0 to 0.25
3 days to 30 days

Low
<0.25
>30 days
      = 5-day biochemical oxygen demand.
 aThis table provides classification of biodegradation data so that waste
  constituents may be categorized for the Source Assessment Model based on
  biodegradability.
 A value of l/10th the average of the "high"  biorates was applied for those
 compounds judged to display "moderate" degradation,  and a value equal  to
 l/100th of the average of the "high" biorates was  applied for those com-
 pounds judged to display "low"  biodegradation.   The  low and moderate bio-
 degradation values (1/100 and 1/10 of "high," respectively) were consistent
 with group averages for the 60  chemicals.
      Once the complement of properties for all  chemicals was completed,
 then all  chemicals were grouped into appropriate surrogate categories  based
 on  their  vapor pressure,  Henry's  law constant,  and biodegradation  values.
 D.2.4  Emission  Factors
      D-2.4.1   Introduction.   A  major objective  of the SAM was  to develop
 nationwide estimates  of organic compound emissions to the atmosphere for
 the range of  organic  chemicals  found  at hazardous waste  sites.   Therefore,
 for each  of the  TSDF  chemical surrogate categories selected to  represent
 the organic chemicals that  occur  in  hazardous waste  streams, the emission
 models discussed  in Appendix  C  of  Reference 5 and the air emission models
 report72  were  used to estimate  organic losses to the atmosphere.  Emissions
 were  estimated for process  losses  and transfer and handling  losses  (i.e.,
 spills, loading  losses, and equipment leaks)  for each type  of TSDF manage-
ment  process.- Loss of organics from the waste stream through biodegrada-
tion was  also estimated for those management  processes having associated
biological activity.
     An important point concerning the emission factors  is that they are a
function of chemical surrogate properties,  air emission models, and TSDF
                                    D-63

-------
model unit parameters.  For each chemical  constituent,  the assigned surro-
gate's chemical, physical, and biological  properties are used in determin-
ing the fraction of incoming organics that are emitted or biodegraded.
Other input parameters to the emission models are provided by the TSDF
model units discussed in Chapter 5.  Once a surrogate is chosen, the TSDF
model unit selected, and the emission model determined, values for emission
factors can be estimated.
     D.2.4.2  Emission Models.  The emission factors used for estimating
TSDF emissions  in this document were calculated using the TSDF air emission
models as presented in the March 1987 draft of the Hazardous Waste Treat-
ment, Storage,  and Disposal Facilities:  Air Emission Models, Draft Report.
Since that time, certain  TSDF emission models have been revised and a new
edition of the  air emission models report was released  (December 1987).
The  principal changes to  the emission models involved  refining the biode-
gradation component of the models  to more accurately reflect biologically
active systems  handling  low organic concentration waste streams.  With
regard to emission model  outputs,  the changes from  the March 1987 draft  to
the  December  1987  version affected,  for the most part, only  aerated surface
impoundments  and result  in  a minor increase  in the  fraction  emitted for  the
chemical  surrogates  in  the  high biodegradation categories.   For  the other
air  emission  models,  such as  the land treatment model,  which were  also
 revised  to  incorporate  new biodegradation  rate data,  the  changes did  not
 result  in appreciable differences  in  the  emission  estimates.   Since  the
 December 1987 report  version  was issued,  new data  on biodegradation  rates
 have been obtained and  comments were received.73   Based on  these data and
 comments,  the biodegradation  model for  aerated wastewater treatment  systems
 was  further revised to incorporate Monod  kinetics.   Additional  investiga-
 tion and comments led to an evaluation  of changes  to the model  units used
 for aerated tanks and impoundments and  assumed  surrogate concentrations.
 These changes improve the technical basis for the biodegradation model.
 However, the combined effect of these changes did not significantly affect
 the estimated nationwide emissions and other impacts presented in this
 document.74  Therefore, the emission factors listed in this appendix remain
 based on the March 1987 draft  of  the air emission model report, the model
 unit definitions have not been changed, and the assumed surrogate
 concentrations  have not  been changed.

                                     D-64

-------
      These models  represent long-term steady-state emissions for land
 treatment,  first-year emissions  for landfills,  and emissions consistent
 with  residence times  identified  for the model  units for wastepiles,  surface
 impoundments,  containers,  and  tanks.   Inputs  to the models are those that
 are determined to  best predict average,  long-term emission characteristics
 rather than short-term peak concentrations.   Long-term emissions are judged
 to be more representative  of actual  TSDF emission patterns and best  charac-
 terize those management process  emissions that  are potentially controlled.
 Long-term emission estimates (i.e.,  annual  averages)  are also required  for
 impacts analysis;  costs, cancer  incidence,  and  ozone effects all  are based
 on long-term emissions.  Short-term emissions  such as those resulting from
 application of waste  to the soil  surface in land  treatment,  as opposed  to
 postapplication emissions,  and therefore are  not  included  in the emission
 estimates.
      Input  parameters  differ for  each  emission  model  and include such
 variables  as unit  size,  throughput,  and  retention time,  all  of which were
 selected  to be as  consistent and  representative as possible  across the
 management  processes.   A detailed breakdown of  the model unit  input  param-
 eters  by management process  is presented in Appendix  C,  Section  C.2,  of
 Reference  5.
      D.2.4.3  Emission  Factor  Files.   To determine TSDF  emission  factors
 for use in  the  SAM, information on process type,  design  characteristics,
 and operating  parameters was necessary.   Within each  waste management
 process  represented by  a process code  (e.g., SOI,  S02, T01,  or  T02),  there
 are in  most  cases  distinct process types.  For  example,  treatment tanks
 (T01)  can be quiescent  or aerated, and quiescent  tanks can be either
 covered or  uncovered.   Table D-15 presents the distribution  of waste
 management  process types used  in the SAM  to characterize the breakdown of
 waste management processes on a nationwide basis.
     For each waste management process type within  a  process code, multiple
model units  (described  in Appendix C of Reference  5) were developed  to span
the range of nationwide design characteristics and  operating parameters
 (i.e., surface area, waste throughputs, detention time, etc.).  Because
these particular characteristics  were generally not available for site-
specific estimates, it was necessary to develop a  "national average model
unit" to represent  each waste management process or process type.  This was

                                    D-65

-------
          TABLE D-15.   NATIONWIDE DISTRIBUTION OF WASTE MANAGEMENT
             PROCESS TYPES USED IN THE SOURCE ASSESSMENT MODEL3
   Process code
                                                             National
                                                           distribution,
      Process type
Container storage (SOI)
Tank storage (S02)
Tank treatment (T01)
Surface impoundment
 treatment  (T02)
Other treatment (T04)b

Landfill disposal  (D80)




Land treatment  (081)
Drum storage
Dumpsters

Covered storage
Uncovered storage

Quiescent covered treatment
Quiescent uncovered treatment
Aerated/agitated uncovered
 treatment
Quiescent treatment
Aerated/agitated treatment

Quiescent covered treatment

Onsite active landfill
Onsite closed landfill
Offsite active  landfill
Offsite closed  landfill

Surface application
Subsurface  application
 97
  3

 79
 21

 30
 20
 50
 29
 71

100

 14
 55
  6
 25

 93
  7
 aThis  table  presents  the  estimated  national distribution of waste management
  process  types within a process  code.   Those  process codes not  listed  are  not
  subdivided  within  the process code.

 bOther treatment  is not subdivided, but is  defined  as  a quiescent covered
  treatment tank for modeling  purposes.
                                     D-66

-------
 accomplished by generating a set of weighting factors for each TSDF waste
 management process or process type based on frequency distributions of
 quantity processed, unit size,  or unit area that were presented in results
 of the Westat Survey.  Each set of weighting factors (presented in Appendix
 C, Section C.2, of Reference 5) approximates a national  distribution of the
 model units defined for a particular TSDF waste management process or
 process type.
      An emission estimate was generated for each chemical  surrogate
 category for each management process or process type.  Process parameters
 and surrogate properties used to estimate emission factors are presented in
 Table D-16.  Emission estimates generally were calculated  on a mass-per-
 unit-time basis (i.e.,  grams per second)  and scaled by  the appropriate
 operating times to get  emissions in megagrams per year.   The emission
 values then were divided by the annual  organic input quantity for  the
 respective model  unit in megagrams  per year.   The emission factors for each
 model  unit, emissions per megagram  of throughput,  were then  multiplied by
 the appropriate weighting factor, and those products were  summed to get the
 weighted emission  factor for each waste management process.
      A set of weighted  emission factors was generated for  all  surrogate
 classes  and all  the SAM management  processes.   In  addition to emission
 factors  for process-related emissions,  emission  factors were developed for
 transfer and  handling related emissions.   Also calculated  were factors used
 to predict biodegradation  quantities.   These  TSDF  emission factors  were
 developed  to  be general  representations of  emissions  and biodegradation
 fractions  for all waste  types,  waste  concentrations,  and waste forms  as
 well as  management  process  combinations and process  unit sizes  on  a
 nationwide  basis.   As such, these emission  factors were incorporated  into
 the SAM  program file  that  is used to  generate  the  SAM nationwide emission
 estimates.  A listing of the TSDF emission  factor  files is included in
 Table D-17.  A separate block of numbers  is presented for  each management
 process with rows denoting  surrogate category  and columns  denoting:
 (1) surrogates,  (2) annual  fraction of surrogate emitted to  air as a
process emission, (3) annual fraction biodegraded, (4) annual fraction
emitted from handling and loading,  (5) annual fraction emitted from spills,
and (6) annual fraction  emitted  from equipment leaks.
                                    D-67

-------
   TABLE D-16   HAZARDOUS WASTE MANAGEMENT PROCESS PARAMETERS AND WASTE
       CONSTITUENT PROPERTIES USED TO ESTIMATE EMISSION FACTORS FOR
                         SOURCE ASSESSMENT MODEL3
Waste management
process
Covered tank storage
(S02)
Uncovered tank
storage (S02)
Storage impoundments
(S04)
Covered quiescent
treatment tanks (T01)
Uncovered quiescent
treatment tanks (T01)
Uncovered aerated
treatment tanks (T01)
Quiescent treatment
impoundments (T02)
Aerated treatment
impoundments (T02)
Disposal impoundments
(D83)
Terminal loading
impoundments and
tanks (L01)
Terminal loading
storage tanks (L03)
Wastepiles (503)
Landfills (D80)
Land treatment (D81)
Physical/chemical
waste form
Organic liquid
Aqueous liquid
Aqueous liquid
Aqueous liquid
Aqueous liquid
Aqueous liquid
Aqueous liquid
Aqueous liquid
Aqueous liquid
Aqueous liquid
Organic liquid
Organic/aqueous
liquid (2 phase)
Organic/aqueous
liquid (2 phase)
Organic liquid
Surrogate
group
Vapor pressure
Henry's law
Henry's law
Henry's law
Henry's law
Henry's law
Henry's law
Henry's law
Henry's law
Henry's law
Vapor pressure
Vapor pressure
Vapor pressure
Vapor pressure
Waste organic
concentration
Pure component
1,000 ppm
1,000 ppm
1,000 ppm
1,000 ppm
1,000 ppm
1,000 ppm
1,000 ppm
1,000 ppm
1,000 ppm
Pure component
5%
5%
--
ajhis table presents,  for those air emission models that require a waste
 concentration as input, necessary information to estimate organic emission
 factors from hazardous waste management facilities used in the Source
 Assessment Model.
                                     D-68

-------
                                                              TABLE 0-17.  EMISSION FACTOR FILES*.b
O
 I
CT>
Weighted emission factors
for container and drum storage
(S01) using vspor pressure surrogstes
Surrogate
1
2
3
4
6
6
7
8
9
10
11
12
• C..O
0
0
0
0
0
0
0
0
0
0
0
0
f(»p>
.0001
.0001
.0001
.0001
.0001
.0001
.0001
.0001
.0001
.0001
.0001
.0001
f(load)
0.0013
0.0011
0.0018
0.0000
0 .0000
0.0000
0.0000
0.0000
0.0000
0.0069
0.0140
0.0140
k(fug)
0.0164
0.0164
0.0164
0.0164
0.0164
0.0164
0.0164
0.0164
0.0164
0.0164
0.0164
0.0164
Weighted emission factors
for dumpster storage
(S01) using vapor pressure surrogates
f(alr)
1.0000
1.0000
1.0000
0.4780
0.4014
0.8269
0.0000
0 .0000
0.0000
1.0000
1.0000
1.0000
f(*p) f(load)
.0001
.0001
.0001
.0001
.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
Weighted emission factora
for covered tank storage
(S02) using vapor pressure surrogstes
k(fug) f(air)
0
0
0
0
0
0
0
0
0
0
0
0
.0012
.0011
.0017
tiUMUl
,WW
• 0000
. 0000
.0000
OUJU3UX
.Knew
AAAA
. WW
MdUUH
> KWW
AAAfl
,ww
.0000
'(•P)
0.0000
0.0000
0.0000
0 .0000
0.0000
0 .0000
0.0000
0.0000
0.0000
0.0000
0 .0000
0.0000
f(load)c
0.0000
0 • 0000
0.0000
0 40000
0.0000
0 • 0000
0.0000
00UM£U)I
. VWV
0 * 0000
aaHflUMI
• KnEncW
0.0000
0.0000
k(fug)
0.003034
0.003034
0.003034
0.003034
0.003034
0.003034
0.003034
0.003034
0.003034
0.003034
0.003034
0.003034
    See notes at end of table.
                                                                                                                                               (continued)

-------
                                                                         TABLE 0-17  (continued)
 I
-^J
o




weighted emission factors for uncovered
tank storage (S02)
using Henry's law constant surrogates
Surrogate
1
2
3
4
E
6
7
8
9
10
11
12
f(..r)
0.EE10
0.E480
0.SE10
0.6460
0.E410
0.6460
0.1180
0.1680
0.1680
'(«P)
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
f(load) k(fug)
0.00304
0.00304
0.00304
0.00304
0.00304
0.00304
0.00304
0.00304
0.00304


for wastepi les
(S03) using vapor pressure surrooatei
f(..r,
0.012S
0.0115
0.0176
0.0020
0.0020
0 . 0020
0.0000
0.0000
0.0000
0.0276
0.0276
0.0276
f(sp) f(load) k(fug)
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000 — — _—
0.0000
0.0000
0.0000



Weighted emission factors
for storage impoundments
f(air)
0.7460
0.7330
0.7470
0.7390
0.7280
0.6630
0.0690
0.0930
0.0930
f(sp) f(load) k(fug)
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.00068
0.00068
0.00068
0.00068
0.00058
0.00068
0.00068
0.00068
0.00068
See notes at end of table 	
                                                                                                                                                     (conti nu«d)

-------
TABLE 0-17 (continued)
Weighted emission factors for covered
Quiescent treatment tanks (T01)

Surrogate
1
2
3
4
5
8
7
g
9

f(air)
0.0113
0.0022
0.0422
0.0002
0.0002
0.0001

0 0000
0.0000

f(sp) f(load)c k(fuo)

0.0000
0.0000
0.0000
0.0000
0.0000

0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0 .0000
0.0000
0.0000
0.0000
0.000066
0.000066
0.000068
0.000066
0 . 000000
0.000086
0.000068
0.000066
0 . 000066
Weighted emission factors for uncovered
quiescent treatment tanks (T01)
using Henry's law constant surrooatea •
f(.ir)
0.1120
0.1060
0.1120
0.0990
0.0910
0.0840
0.0010
0.0030
0.0030
f(sp) f(load)
0 . 0000 •*"
0.0000
0 .0000 ~"~
0 . 0000 — —
0.0000
0.0000
0.0000
0.0000
0.0000
k(fug)
0.000066
0.000066
0 . 000068
0.000066
0 . 000000
0.000066
0.000066
0.000066
0.000066
Weighted •mission factors for uncovered
aerated treatment tanks (T01)
using Henry's
f(air)
0.8780
0.7790
0.9550
0.1460
0.1810
0.0940
0.0005
0.0020
0.0020
f (bio)
0.0510
0.0110
0.0000
0.4200
0.0020
0.0060
0.3100
0.0550
0.0550
law constant surrogates
f(.p)
0.0000
0 .0000
0 > 0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
f(load) k(fug)
0.000013
0.000013
0.000013
0.000013
0.000013
0.000013
0.000013
0.000013
0.000013
	 	 ' 	 ~~ ' (continued)

-------
-vl
rv>
— 	 — 	 	 — 	 '
Weighted emission factors for quiescent
treatment impoundment* (T02)
using Henry's law constant surrnfl»f..
Surrogate
1
2
3
4
5
6
7
8
9
10
11
12
See note* at
fOr)
0.6180
0.6000
0.6190
0.6060
0.4910
0.4090
0.0170
0.0260
0.0260
*(»P) f(load)
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
k(fug)
0.000009
0.000009
0.000009
0.000009
0.000009
0.000009
0.000009
0.000009
0.000009
Weigh
f(air)
0.7120
0.9780
0.9900
0.3290
0.8330
0.7470
0.0040
0.0480
0.0480
impoundments (T02) using Henry's
law constant murron.f..
f (bio)
0.0630
0.0010
0.0000
0.7700
0.0060
0.0040
0.9160
0.3180
0.3180
f(sp) f(load)
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
k(fug)
0 • 000009
0.000009
0.000009
0.000009
0.000009
0.000009
0.000009
0.000009
0.000009
/T0|O\
f (a I r)
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
Weighted emission factors
for incineration
f(»p)
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
pressure
f (load)
—
end of table. 	 — — 	 	 	
.surrogates
k(fug)
0.00046
0.00045
0.00045
0.00045
0.00045
0.00045
0.00045
0.00046
0.00045
0.00045
0.00045
0.0004E

                                                                                                                                                          (continued)

-------
                                                                     TABLE D-17  (continued)
o



CO



Weighted emissions factors
for injection wells
Surrogate f(air) f(sp) f(lo.d) Mfug)




6 —

7 —

9 —
10
11
12
0.0000

0 0000
0.0000
0.0000
0 0000
0.0000
0 0000
0.0000
0.0000
0.0000
0.0000
0.00009
0.00009
0.00009
0.00009
0.00009
0.00009
0.00009
0.00009
0.00009
0.00009
0.00009
0.00009


Weighted emission factors for onsite
active landfills (D80)
using vapor pressure surrogates.
f(air) f(sp) f(load) k(fug)
0.2230
0.2070
0.3110
0.0300
0.0300
0.0410
0.0002
0.0002
0.0002
0.4870
0.7000
0.7000
0.0000
0.0000 —i" ™~
0.0000
00UMUN — -~ — —
,ww
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000 ~"— *"~
0.0000
0.0000



Emission factors for onsite
closed landfills (080)
using vaoor pressure surrogates
f(air)
0.0091
0.0087
0.0171
0.0002
0.0001
0.0003
0.0000
0.0000
0.0000
0.043S
0.0951
0.0951
f(sp) f(load)
0.0000
0 . 0000 ~™
0.0000
0.0000
0.0000
0.0000
0.0000
0 . 0000 ~" ""
0 «0000 *~~
0.0000
0.0000
0.0000
k(fug)
—
~~
~~
""
~~
~~
~~
"
— _

""

	 — 	 - (continued)
     See notes at end of table.

-------
                                                                              TABLE 0-17  (continued)
O
-vl
Weighted emission factors for commercial
active landfills (080)
	 using vapor pressure surrogates
Surrogate
1
2
3
4
E
e
7

8
9
10
11
12
n»ir)
0.1110
0.1030
0.1EE0
0.0160
0.01E0
0.0210
0.0001

0.0001
0.0001
0.2420
0.3S60
0.3560
f(»p) f(load) fc(fug)
0.0000
0.0000
0.0000 — — __
0.0000
0.0000
0.0000
0.0000

0.0000
0.0000
0.0000
0.0000
0.0000
Weighted emission factors for closed
commercial landfills (080) using vapor
pressure surrogate*
f(«ir)
0.0076
0.0070
0.0146
0.0001
0.0001
0.0002


0.0000
0.0000
0.0367
0.0798
0.0798
f(sp) f(load) k(fug)
0.0000
0.0000
0.0000 — — — —
0.0000
0.0000
0.0000


0.0000
0.0000
0.0000
0.0000
0.0000
Weighted emission factors for land
treatment surface application (081)
us i no vaoor oressure murranmt.**
f(.ir)
1.0000
1.0000
1.0000
0.2663
0.3943
0.8EE1

0.0020
0.0020
0.0020
1.0000
1 0000
1.0000
f(sp) f(load) k(fug)
0 0AA9A
0.0000
0.0000

0.0000
0.0000

0 0000 — —
0.0000
0.0000
0.0000
0 0000 	
0.0000
See notes at end of tahla. 	 	
                                                                                                                                                         (continued)

-------
                                                                TABLE 0-17 (continued)
o



en
Weighted emission factors
for lend treatment
subsurface injection (081)

Surrogate
1
2
3
4
6
6
7
8
9
10
11
12
using
f(air)
0.8480
0.9640
0.9960
0.1610
0.3310
0.8320
0.0020
0.0020
0.0020
0.9550
0.9990
0.9990
vapor pressure surrogates
f(sp) f(load) k(fug)
0.0000
0.0000
0.0000
0.0000
0.0000
v * 0000 ~ ~ ~ ~
0>0000 ~~ ~~
0.0000
0 • 0000 ~ •• ~ *•
0 . 0000 •• — ~ ~
0.0000
0.0000
Weighted emission factors for
disposal impoundments (083)
us i ng
f(air)
1.0000
1.0000
1.0000
1.0000
1 .0000
1.0000
0.4700
0.6300
0.6300



Henry's
f(»p)
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000



law surrogates
f(load) k(fug)
0.00011
0.00011
0.00011
0.00011
0.00011
0.00011
0.00011
0.00011
0.00011



Weighted
emission factors
for terminal loading
of containers
f(sp) f(load)
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
(L01)
k(fug)
0 • 0000
0.0000
0 .0000
0.0000
0 (0000
0.0000
0.0000
0.0000
0.0000
0 . 0000
0.0000
0.0000
                See notes at end  of  table.
(continued)

-------
                               Weighted emission factors
                               for terminal  loading from
                                 impoundments and tanks
                                (L02)  using Henry's law
                                	    surrogates
                                                               TABLE D-17  (continued)
                               Weighted emission  factors
                               for  terminal  loading from
                               storage tanks  (L03)  using
                               vapor  pressure  surrogates
                                               Weighted emission factors for
                                                waste fixation using vapor
                                                   pressure Au
                Surrogate
f(»p)    f(load)
k(fug)
f(sp)
f(load)    k(fug)
                                                               f(air)
                                                                                                         f(sp)     f(load)    k(fug)
a
--j
1
2
3
4
E
e
7
8
9
10
11
12
0
0.
0,
0.
0
0
0
0
0



.00001
.00001
.00001
.00001
. 00001.
.00001
.00001
.00001
.00001



0.0013
0.0011
0.0018
0 . 0000
0.0000
0.0000
0.0000
0.0000
0.0000



0.00799
0.00799
0.00799
0.00799
0.00799
0.00799
0.00799
0.00799
0.00799



0.00001
0.00001
0.00001
0.00001
0 . 00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.0013
0.0013
0.0018
0.0000
0.0000
0 . 0000
0.0000
0.0000
0.0000
0.0069
0.0140
0.0140
0.00799
0.00799
0.00799
0.00799
0.00799
0.00799
0.00799
0.00799
0.00799
0.00799
0.00799
0.00799
0.6800
0.6800
0.6800
0 . 5300
0.5300
0.E300
0.0600
0.0E00
0.0E00
0.6800
0.6800
0.6800
0
0
0
0
0
0
0
0
0
0
0
0
MOUJUH
. KWKW ~~
OUH0UH
• WW — •"
aujunn
, VWv " ~
.0000
.0000
.0000
.0000
. 0000 ™~
. 0000 ~ ~
. 0000 "••"
.0000
MMUJt
. WW ~~
	
—
—
—
—
—
--
—
—
—
"
	
                Note:  Oash indicates emission factors not applicable.

                •Some waste management processes, such as S01,  S02,  and S03,  lack a column for biodegradation fraction.  They have
                 no biodegradation component, or biodegradation has  been considered in the air emission factor determination, and
                 they are read in SAU as zeros.

                 Values of 0.0000 do not imply that the emission factors equal  zero.  Rather, 0.000 reflects values less than
                 0.001 percent.

                bjhe f(	) in the column headings represent fractions  emitted  or degraded.  The k(f) in the last column represents
                 fractions emitted from equipment (e.g., pumps, valves, open-end line, etc.).

                     f(air)  =  process emissions fraction leaks associated with the waste management process.
                     f(bio)  =  biodegradation fraction
                     f(sp)   =  spills fraction
                     k(f)    s  equipment leak fraction.

                'Loading emissions included  in f(air).

                ^Emission factors for waste  fixation are based on the information and data contained in a report prepared by Acurex
                 Corp. for the U.S. EPA titled 'Volatile Emissions from Stabilized Waste in Hazardous Waste Landfills," Project
                 8186, Contract 68-02-3993,  January 23, 1987.

-------
D.2.5  Control Technology and Cost File
     A file was developed for the SAM that provides control  device effi-
ciencies for each emission control alternative (see Chapter 4.0 of
Reference 5) that is applicable to each waste management process.  Certain
control options are specific to waste form.  The control technology file
provides control efficiencies for land treatment alternatives and add-on
control alternatives among others.  The control file is a combined file
that includes control costs as well as control efficiencies.  Model waste
compositions (defined in Appendix C, Section C.2.2, of Reference 5), pro-
vided the bases for estimating control costs and control efficiencies by
waste form.
     Tables D-18 and D-19 present the control cost file broken down by
emission source and control option.  A key is provided at the bottom of the
table that explains the columns and how they are used in the SAM.
     One important note is that the control cost profile requires that
controls and costs be developed for all physical/chemical waste forms even
though certain forms and management processes are incompatible or improb-
able (e.g., storage of a solid hazardous waste in a closed storage tank or
storage of an organic liquid waste in an open impoundment).   The SAM
dilutes incompatible waste forms, when necessary, but cannot redefine the
waste form.  Therefore, the cost/control file was modified to estimate
emission reductions and costs for all waste forms.  The SAM will substitute
the control costs for a similar waste form if there are no cost factors for
a particular (incompatible) form.  For example, cost factors for control of
dilute aqueous wastes will be used for estimating control costs of a
(diluted) aqueous sludge slurry because this waste form did not have
control costs developed specifically.
     Costs were developed in a way that allows one to estimate capital and
annual costs based on total volume waste throughput.  Within each manage-
ment process, total capital investment and annual operating costs were
determined for a range of model units and the appropriate add-on control
technologies applicable to these processes.  The same waste management
process weighting factors used to develop emission factors were used to
develop weighted cost factors.  Estimation of the costs for applying
emission controls to TSDF waste management units would  ideally be done
using specific information about the characteristics of the waste

                                    D-77

-------
                                                   TABLE 0-ia.  SUPPRESSION AND AOO-ON CONTROL COST FILE USED BY  THE SOURCE ASSESSMENT MODEL'.*1
C3

—I
CO
TSOF
process
code
(I)
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOI
SOS
SOS
SOS
SOS
sos
SOS
SOS
'sos
sos
SOS
sos
TSDF

Drui Storage
Drm Storage
Orin Storage
Drm Storage
Orm Storage
Drui Storage
Dmpster
Duipster
Duipster
Duipster
Dunpstrr
Dmpster
Fugitives- Drm load
Fugitives- Drm Load
Fugitives- Drm load
Fugitives- Drm Load
Fugitives- Drin Load
Drui Loading
Drui Loading
> Drui Loading
Drui Loading
Drm Loading
Drm Loading
link Storage
Tank Storage
•"ink Storage
Tank Storage
lank Storage
Tank Storage
Tank Storage
Tank Storage
Tank Storage
Tank Storage
Tank Storage

(3)
VOC-cont Solid
Aq Sldg/Slurry
Dilute Aq
Ore. Liquid
Org Sldg/Slurry
2-fhase Aq/Org
VOC-coitt Solid
Aq Sldg/Slurry
Dilute Aq
Org Liquid
Org Sldg/Slurry
2-fhase Aq/Org
Aq Sldg/Slurry
Dilute Aq
2-Wiase Aq/Ory
Org Liquid
Org Sldg/Slurry
VOC-Cont Soli.!
Aq Sldg/Siw
Dil Aqueous
Org liquid
Org Sldg/Slurry
2-fhase Aq/Org
Sub 2n> for Ini
Aq Sldg/Slurry
Dilute Aq
Org Liquid
Org Sldg/Slurry
2-Phase Aq/Org
Sub 2n> for Inn
flq Sldg/Slurry
Dilute Aq
Org Liquid
Org Sldg/Slurry
VoU-
tlltty

All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
mi
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
Con-
trol
(S)






2
2
2
2
2
2
3
3
3
3
3
4
4
t
4
4
4
1
1
1
1
1
1
2
2
2
2
2

Control
•ff IcUncv

(8)
Vent to Car Ads
Vent to Car Ads
Vent to Car Ads
Vent to Car Ads
Vent to Car Ads
Vent to Car Ads
Dmpster Cover
Dmpster Cover
Duipster Cover
Duipster Cover
Duipster Cover
Duipster Cover











IFR, Cflds,Vent to CD
IFR, CAds,Vent to CD
IFR, CAds,Vent to CD
IFR, CAds,Vent to CD
IFR, CAds, Vent to CD
IFR, CAds,Verit to CD
Fined Roof
Fined Roof
Fined Roof
Fined Roof
Fined Roof
(7) (B)
95.00
95.00
95.00
95.00
95.00
95.00
99.00
99.00
99.00
99.00
99.00
99.00











84.5
88.5
vVS
91.75
91.5
86.5
86.4
98.70
86.40
93.90
93.85
Tnnx-

C
C
C
C
C
c






c
c
c
c
c
c
c
c
c
c
c
D
D
D
0
D
D
D
D
D
D
D

Ing
(IB)
A
A
A
A
A
A











A
A
A
A
A
A











S«r-

20
20
20
20
20
20
20
20
20
20
20
20











10
10
10
10
10
10
20
20
20
20
20
Totll
opt til
(IS)
liwar
Liwar
Li rear
Liwar
Liwar
linear
liwar
liwar
Liwar
Liwar
liwar
Liwar











Liwar
liwar
Liwar
liwar
Liwar
liwar
Linear
liwar
Liwar
liwar
Liwar
(13)
0.00
0.00
0.0V
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00











0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.09
0.00
o.oo
b « Q
490M1
72.000
89.000
91.000
64.000
91.000
6.250
9.375
25.030
25.640
17.860
25.610









i

9.740
9.7*0
13.210
12.360
11.080
10. 740
14.660
14.660
18.470
18.530
15.310
Co.t
func-
tion
(IB)
liwar
liwar
Liwar
Linear
Liwar
Linear
Liwar
liwar
Liwar
linear
Liwar
linear











linear
Linear
liwar
liwar
Linear
Liwar
Liwar
Liwar
Liwar
Liwar
liwar
Annua 1
operating
<1«)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00











0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
b » q
(17)
15.006
21.000
27.006
27.000
20.000
27.000
1.670
2.500
6.670
6.840
4.980
6.830









.

3.290
3.290
3.450
4. 720
5.710
4.870
.070
.070
.360
.370
.ISO
       See notes »t end of table.
                                                                                                                                                                               (continued)

-------



TSOF
process
cod*
(D
502
502
502
502
502
502
502
503
503
S02
503
503
0 S03
^, 503
i£> 503
503
503
504
504
504
504
504
504
S04
S04
504
504
504
S04
504
S(>4
504




TSDF

(2)
— 	 — —
Tank Storage
Tank Storage
lank Storage
Tank Storage
Tank Storage
Tar* Storage
Tank Storage
Fugitives- Tank Load
Fugitives- Tank Load
Fugitives- Tank Load
Fugitives- Tank Load
Fugitives- Tank Load
Haste Pile
Waste Pile
Haste Pile
Haste Pile
Haste Pile
Uaste Pile
Stor Upd Surface
Stor l«pd Surface
Stor lupd Surface
Stor Upd Surface
Stor Upd Surface
Stor lupd Surface
Stor Upd Surface
Stor Upd Surface
Stor Upd Surface
Stor Upd Surface
Star I«pd Surface
Stor Upd Surface
Fugitives- Up Load
Fugitives- Up Load
Fugitives- Up load


	 —



(3)
_^ —
2-Phase Aq/Org
Sub 2«« for Imi
Aq Sldg/Slurry
Dilute Aq
Drg liquid
Org Sldg/Slurry
2-Phase Aq/Org
Aq Sldg/Slurry
Dilut Aq
2-tttase Aq/Org
Org Liquid
Org Sldg/Slurry
Aq Sldg/Slurry
Sub 2» for 3»K
Sub 7n« for 4«»
Sub 7»« for 5«»
2-Phase Aq/Org
VOC-cont Solid
Sub 2«« for i«»
Aq Sldg/Slurry
Dilute Aq
Sub 2» for 4««
Sub 2»» for 5«»
2-Phase Aq/Org
ISub 2»« for l»
Aq Sld?/Slurry
'Dilute Aq
Sub 2«« for 4»«
Sub t»* for 5»i
2-Phase Aq/Org
Aq Sldg/Slurry
Dilut Aq
2-Phase Aq/Org

TM


~
Vols- Con-
til Ity trol Emission S

(8)

All 2 Fi«ed Roof
All 3 Roof,IFR,CAds,Vent
All 3 Roof ,lfR,CAds, Vent
All 3 Roof,IFR,l»ls,VeiA
All 3 Roof, IFR,CAds, Vent
All 3 Roof, IFR,CAds, Vent
All 3 Roof, IFR,Cads, Vent
All

All
All
All
All
All
All
All
All
All
All
All
All
1 All
All
All
All
All
All
All
All
All



HD Cover 30 nil
HD Cover 30»il
HD Cover 30 eil
HD Cover 30 mil
HD Cover 30 »il
HD Cover 30e.il
Syn Nenbrane
SynHesbrane
Syn HenVaM
Syn Nesbrare
Syn NesbraM
Syn Kesbrane
2 Struct H Car Adsorp
2 Struct M Car Adsorp
2 Struct M Car Adsorp
2 Struct M Car Adsorp
2 Struct M Car Adsorp
2 Struct » Car Adsorp
All •>
All •*
All J
	 _ 	 —
E 0-18 (continued]
.
Control
.fficUno
uppr«s- Emission
•Ion control
(7) <•)

90.00
97.9
99.65
97.90
99.99
99.99
96.70



99.70
99.70
49.30
49.30
49.30
0.00
65.00
65.00
65.00
65.00
65.00
65.00
95.00
95.00
95.00
95. 00
95.00
95.00



	 • ' 	 —

Trsns- S«r-
f.r Losd- vie*
cod* Ing IH.
(») (!•) (It)

D ev
D 10
6 10
D 10
D 10
D 10
g 10
D
D
D
D
D


i
5
S
S
S
S
10
10
10
10
10
10
10
10
10
10
10
10



	 — — 	 —
	 1 	 . — — —

Cost
function
(12)

Linear
Linear
Li Mar
Li Mar
linear
llr«ar



LiMar
LiMar
Linear
linear
LiMar
linear
LiMar
LiMar
LiMar
LiMar
LiMar
Linear
linear
Linear
lirear
LiMar
linear
LiMar




.^ 	 —
Tots!
c.pltsl
• b x q
(13) OP
0.00 15.720
0.00 20.960
0.00 20.960
0.00 27.660
0.00 26.640
0.00 22.780
0.00 22.550



0.00 0.310
0.00 0.310
0.00 0.310
0.00 0.310
0.00 0.310
' 0.00 0.310
0.00 .650
0.00 .650
0.00 .650
0.00 .650
0.00 .650
0.00 .650
0.00 10.760
0.00 10.760
0.00 9.430
0.00 10.760
0.00 tO. 760
0.00 9.430





Cost
tion
(16)

Linear
LiMar
Linear
LiMar
linear
Linear
LiMar



linear
LiMar
LiMar
LiMar
LiMar
Linear
LiMar
lirear
Linear
Lirear
Lirear
Lirear
linear
Lirear
Lirear
lirear
Lirear
Lirear





Annu* 1
op*rstlng
cost, S
• b » q
(18) (17)

O.«0 1.160
0.00 3.960
0.00 3.560
0.00 10.500
0.00 6. 150
0.00 6.390
0.00 6.69(>



0.00 O.OBO
0.00 0.060
0.00 0.060
0.00 0.030
0.00 0.080
0.00 O.OSO
0.00 0.240
0.00 0.340
0.00 0.240
0.00 0.240
0.00 0.24(i
0.00 O.J40
0.00 3.030
0.00 3.030
0.00 2.270
0.00 3.030
0.00 3.030
0.00 2.270



(continued)
Se« not«J st and of tsbl*.

-------
                                                                                      TAKE D-18  (continued)
O

OO
O
TSOF
process
CD
S04
S04
im
TOI
TOI
TOI
TOI
TOI
TOI
TOI
TOI
TOI
101
TOI
101
TOI
TOI
TOI
101
TOI
TOI
TOI
TOI
TOI
TOI
TOI
TOI
TOI
101
TOI
TOI
TOI
TOI
TOI
TSOF
(2)
Fugitives- lap Load
Fugitives- l«p Load
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Tar* Surface
Tank Surface
Tank Surface
Tank Surface
Fugitives- 0 Tank Ld
Fugitives- 0 Tank Ld
Fugitives- 0 Tank Ld
Fugitives- 0 Tank Ld
Fugitives- 0 Tank Ld
Fugitives- A Tar* Id
Fugitives- A Tank Ld
' Fugitives- A Tank Ld

(3,
Orj Liquid
Org Sldg/Slurry
Sub 2« for 1>»
Aq Sldg/Slurry
Dilute Aq
Org Liquid
Drg Sldg/Slurry
2-Phase Aq/Org
Sub 2>x for l<«
Aq Sldg/Slurry
Dilute Aq
Org Liquid
Org Sldg/Slurry
2-Phase Aq/Org
Sub 2>» for Ixi
Aq Sldg/Slurry
Dilute Aq
Org Liquid
Drg Sldg/Slurry
2-Phase Aq/Org
Sub 2x> for IxK
Aq Sldg/Slurry
Dilute Aq
Sub 2xx for 4x>
Sub 2>x for 5xx
Sub 2xx for 7sx
Aq Sldg/Slurry
Dilut Aq
Org Sldg/Slurry
Org Liquid
Org Sldg/Slurry
Aq Sldg/Slurry
Dilut Aq
2-Phase Aq/Org
Vola-
tility
(0
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
fill-
Con-
trol
ln q

0.030
0.030
0.030
0.030
0.030
0.030
0.130
0.130
0.390
0.280
0.300
0.360
0.10
0.10
0.37
0.25
0.27
0.36
0.19
0.19
0.30
0.19
0.19
0.19








        See notes »t end of t»bt«.
                                                                                                                                                                               (continued)

-------
TABLE O-l« (continued) 	 . 	 	
TSDF
process
cod*
(1)
101
101
102
102
102
102
102
102
102
102
102
102
102
102
C_J 102
1 J(£
2 T02
102
102
102
102
102
102
ice
102
TO;
103
103
103
103
103
103
104
104

TSDF
•mission sourc*
(2)
Fugitives- A land Ld
Fugitives- A Tank Ld
Treat Upd Surface
Treat Upd Surface
Treat Upd Surface
Treat l«pd Surface
Treat Upd Surface
Treat Upd Surface
Treat Upd Surface
Treat lupd Surface
Treat Upd Surface
Treat Upd Surface
Treat Upd Surface
Treat Upd Surface
Fugitives- Up Load
Fugitives- Up Load
Fugitives- Up Load
Fugitives- lip Load
Fugitives- Up Load
Irt Upd Surface
Trt lupd Surface
Trt Upd Surface
Irt Upd Surface
Irt Upd Surface
ITrt Upd Surface





lank Surface
Tank Surface

	 	
Wist* fore
t31
Org Liquid
Org Sldg/SUrry
Sub 2«« for l»»
Aq Sldg/Slurry
Dilute Aq
Sub 2»» for 4»«
Sub 2«« for 5»
2-Phase Aq/Org
Sub 2»« for l»>
Aq Sldg/Slurry
Dilute Aq
Sub 2<» for 4m
Sub 2«« for 5«»
2-Phase Aq/Org
Aq Sldg/Slurry
Dilut Aq
2-Phase Aq/Org
Org Liquid
Org Sldg/Slurry
Sib 2>i for lull
Aq Sldg/Slurry
Dilute Aq
Sub 2«« for 4x»
Sub 2»« for 5»»
2-Phase Aq/Org
VOC-cont Solid
Aq Sldg/Slurry
Dilute Aq
Org Liquid
Org Sldg/Slurry
2-Phase Aq/Org
Sub 2» for l»«
Aq Sldg/Slurry

Vol»- Con-
tlllty trol
class lnd»«
(4) (6)
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All

All
All
All
All
All
All
All
All
All





All
All

6
6
2
2
2
I
2
2
3

3
3
3
5
5
5
5
5
I
1


1
I
1
3
3


Control
•fflcUncv
• 	 • — " 	 Tr«n«-
S.r-
e-lssion Suppr.,- E-ls.l.o f-r L.sd- ^c.
control option slon control cod. I»B ' f»
(«) m (8) (») (»•> <»»

Stuct I Car Adsorp
Stuct 1 Car Adsorp
Stuct 1 Car Adsorp
Stuct 1 Car Adsorp
Stuct 1 Car Adsorp
Stuct 1 Car Adsorp
Stuct 1 Car Adsorp
Stuct 1 Car Adsorp
Stuct 1 Car Adsorp
Stuct 1 Car Adsorp
Stuct 1 Car Adsorp
Sluct 1 Car Adsorp



Syn Henbrane
Syn Hwbrane
Syn Heubrane
SynMeihrane
Syn Henbraw
Syn Mestoraw





lFR,CAQS,Vent to CO
|FR,CMs,Vent to CD

E
G
95. M J
95.00 J
95.00 *
95.00 1
95.00 J
95.00 J
95.00 J
95.00 J
95.00 J
95.00 J
95.00 J
95.00 J
J
J
1
j
J
15.00 1
65.00 J
85.00 J
65.00 " J
85.00 J
65.00 J
E

E
E
E
E
E
64.50 M
M.50 H


10
10
10
10
10
10
10
10
10
10
10
10



10
10
10
10
10
10





10
10

Totsl
cspltll
Cost
function
(12)

Linear
Linear
Linear
Linear
Linear
Linear
Linear
linear
Linear
Linear
linear
Linear



Li war
linear
linear
linear
liwar
Linear





Liwar
liwar

. b > a
(IS) (")

0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00



0.00
0.00
0.0(1
0.00
0.00
0.00





0.00
(.00


2.600
2.600
2.300
2.600
2.600
2.300
2.900
2.900
2.500
2.900
2.900
2.500



0.460
0.460
0.460
0.460
0.460
0.460





0.220
0.220

Cost
func-
tion
(16)

Linear
Linear
Linear
Linear
Linear
Liwar
Linear
Li war
Liwar
liwar
liwar
Liwar



Liwar
Linear
liwar
Liwar
liwar
Linear





Liwar
liwar

Annual
op.rsttng
cost. S
• b « Q
(te) (17)

0.00 0.600
0.00 0.800
0.00 0.500
0.00 O.BOO
0.00 0.600
0.00 0.500
0.00 1.200
0.00 1.200
0.00 0.700
0.00 1.200
0.00 1.200
0.0(1 0.700



0.00 0.060
0.00 0.060
0.00 0.060
0.00 0.060
0.00 O.W.O
0.00 0.060





0.00 0.10
0.00 0.10
(continued)
£•• not** »t •nd of t»bl«.

-------
                                                                                   TABLE 0-18 (continued)
O

00
ro
TSDF
process
(1)
TO*
104
TO*
TO*
104
Ten
TO*
TO*
TO*
D79
D79
D73
D73
D79
D73
D79
D90
080
DBO
D90
; DBO
DSO
D60
DSO
DBO
DBO
DBO
DBO
DBO
DBO
DSC
DBO
DBO
DBO
TSDF
(2)
Tank Surface
Tank Surface
Tank Surface
Tank Surface
Fugitives- 0 Tar* Id
Fugitives- 0 Tank Id
Fugitives- 0 Tank Id
Fugitives- 0 Tank Ld
Fugitives- 0 Tank Ld







Landfill (Open)
Landfill (Open)
Landfill (Open)
Landfill (Open)
Landfill (Open)
Landfill (Open)
Landfill (Closed)
Landfill (Closed)
Landfill (Closed)
la.dfill (Closed)
Landfill (Closed)
Landfill (Closed)
Landfill (Closed)
Landfill (Closed)
Landfill (Closed)
Landfill (Closed)
Landfill (Closed)
Landfill (Closed)

O)
Dilute Aq
Org Liquid
Orj Sldg/Slurry
2-ftiase Aq/Org
Aq Sldj/Slurry
Dilut Aq
Org Sldg/Slurry
Org Liquid
Org Sldg/Slurry
VOC-cont Solid
Aq Sldg/Slurry
Dilute Aq '
Org Liquid
Org Sldg/Slurry

2-Phase Aq/Org
Aq Sldg/Slurry
Sub 7x» for 3««
Sub 7» for *»
Sub 7>» for 5«»
2-fhase Aq/Org
VOC-cont Solid
VX-cont Solid
Aq Sldg/Slurry
Sub 7«» for 3««
Sjb 7«« for *««
Sub 7i> for 5««
2-ttiase Aq/Org
VOC-cw.t Solid
Aq Sldg/Slurry
Sub 7»x for 3»>
Sub 7»» for *»«
Sub 7« for 5««
a-Phase Aq/Org
Vol«- Con-
tllUy trol Emission
(«) (6) («)
All 3 IFR.CAds.Vent to CD
All 3 lFR,CAds,Vent to CO
All 3 IFB,Cads,Vent to CD
All 3 IFR, Cads, Vent to CO
All 5
All 5
All 5
All 5
All S







All
All
All
All
All
All
All







Earth Covtr
Earth Cover
Earth Cover
Earth Cover
Earth Cover
Earth Cover
HD Cover 30 ill
All 3 HD Cover 30 oil
All 3 HD Cover 30 «il
All 3 HD Cover 30 nil
All 3 HD Cover 30 nil
All 3 HD Cover 30 »il
All HD Cover 100 nil
All HD Cover 100 «il
All HD Cover 100 «il
All HO Cover 100 nil
All HO Cover 100 sil
Al 1 HD Cover 100 KI 1
Control
officUnc*
Trsns-
Suppros- Emission for L
Sor-
oso*- vleo
(7) (8) (») (!•) (11)
84.50
91.75
91.50
86.50












11.00
11.00
11.00
11.00
11.00
11.00
0.00
99.70
49.30
49.30
49.30
43.30
0.00
99.90
B4.BO
14. SO
B4.BO
64.80
H
H
H
H
H
H
H
H
H





F
F


















10
10
10
10












20
20
20
20
20
0
30
30
30
30
30
30
30
30
30
30
30
30
Cost
function
(12)
Lircar
linear
Lircar
Lircar












Linear
Li war
Li Mar
Linear
Lircar
Linear
linear
Linear
Linear
Li rear
Lircar
Lircar
Lircar
Linear
Linear
linear
Lircar
linear
Tot si
cap Its!
Invostowit. S
(13)
0.00
0.00
0.00
0.00












0.00
0.00
0.00
0.00
0.00
•0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
b > 0
O«)
0.620
0.360
0.360
0.600












0.000
0.000
0.000
0.000
0.000
0.000
0.760
0.760
0.760
0.760
0.760
0.760
.960
.960
.960
.960
.960
.960
Cost
func-
tion
(16)
Linear
lircar
I i rear
Linear












Linear
Linear
linear
Linear
LI war
lircar .
Lircar
Lircar
Lircar
lircar
lircar
lircar
Lircar
Lircar
Lircar
Lircar
Lircar
Lircar
Annus 1
oporstlng
cost. S
(16)
0.00
0.00
0.00
0.00












0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
b > 4
(17)
0.37
0.25
0.27
0.36












2.690
£.690
2.690
2.690
2.690
2.690
0.03'1
0.030
0.030
0.030
0.030
0.030
O.OBO
O.MO
O.OBO
0.060
0.060
o.oao
      Sea notes  at end of  table.
                                                                                                                                                                             (continued)

-------










1
00



TSDF
proce. .
code
O)
DB3
DB3
D83
083
D83
FXP
FXP
FXP
FXP
FXP
FXP
LWI
IDA
LM
IDA
LM
•This

TSOF
emission .ource
Fugitives- lep Load
Fugitives- Inp Load
Fugitives- lip Load
Fugitives- lup Load
Fugitives- lip Load
Filiation Pit
Fuation Pit
Filiation Pit
Final ion Pit
Filiation Pit
Fiiation Pit
Fugitives- LDB Ircin
Fugitives- LOB Intin
Fugitive?- 11* Incin
Fugitive- LM Incin
Fugitive?- IM Incin


W..te form
(3)
Aq Sldg/Slurry
Sub 2»» for *«»
2-Phase Aq/Org
Sub 2«» for *,»»
Sub 2« for Sm
Aq Sldg/Slurry
Sub 7«» for 3«»
Sub 7»» for *»«
Sub 7«« for 5»
2-Phast Aq/Org
VOC-cont Solid
Aq Sldg/Slurry
Dilute Aq
Org Liquid
Org Sldg/Slurry
2-Phase Aq/Org

TABLE D-18 (continued)
Vol.- Con
tllity tro
cl.ss Ind
(<> <«


All
All
All
All
All
All
All
All
All
All
Control
efficiency
	 ™ 	 Tr.n.-
I Emission Suppre.- EeiL.lon fer
.. control option .ion control cod*
) («) (7) (•) (»)
K
K
K
X
K
95.00
95.00
OR AA
95.00
95 00

E
E

t
All


Ser-
Lo.d- vice
(1C) (>»)


20
20
20
20
20






Mnt Model.



Tot.l
c.plt.l
(12)


Linear
linear
linear
linear
Li rear
linear







. b > Q


0.00
0.00
0.00
0.00
0.00
0.00









12.030
12.030
12.030
12.030
12.030
12.030








Co.t
func-
tion


linear
Liwar
Linear
linear
Linear
liwar









Annu. 1
oper.ting



0.00
0.00
0.00
0.00
0.00
0.00







b x q


3.720
3.720
3.720
3.720
3.720
3.720







dof.nition. of column, for th» TSOF Proco.. Control F1U «r«:

 =  U.n.gomont proc«*» cod*.
 -  M»n»gement proco** definition.
 =  W.st. form definition.
 =  Volatility definition.
 =  Emiijion control numeric Indlcetor.
 =  Emission control definition.
 =:  Suppression control efficiency.
 =  Control efficiency.
                                                            to T.bl* D-19,'column 1, THL proce.. Indicator.
II
12
13
14
IE
16
U
    Service  life of control equipment  (yr) .
    Cost  function description, for  c.plt.l  lnve.twe.it.
    Fi««d control cost  for c.plt.l  Investment.
    Throughput multiplier for c.pit.l  Investment.
    Cost  function description for annu.l  oper.ting co.t.
    Fixed .nnu.l oper.tlng co.t.
    Ihrovjghput multiplier for .nnu.l oper.tmg  cost.

-------
                                                        TABtE 0-19.   TRANSFER, HANDLING, AND LOAD CONTROL COST FILE USED BY THE SOURCE ASSESSMENT MODEL*."
 I
oo
THL
procoss
indicator Emission sourco
O> (2)
A
n
A
n
A
A
B
B
B
8
B
C
C
C
C
C
D
D
D
0
0
E
E
E
E
C
F
F
F
F
F
6
G
G
"run Loading
DrtM Loading
Drim Loading
DriM Loading
Driin loading
On* Loading
Truck loading
Truck Loading
Truck Loading
Truck Loading
Truck Loading
Fugitives- Drus Loading
Fugitivti- Drui loading
Fugitives- Oru* Loading
Fugitives- Driw Loading
Fugitives- DriM Loading
Fugitives- Sto Tank loading
Fugitives- Sto Tank loading
Fugitives- Sto Tank loading
Fugitives- Sto Tank Loading
Fugitives- Sto Tank. Leading
Fugitives- 'icin LoadlT03)
Fugitives- trcin load)T03l
Fugitives- Irclr, Ic.adlWI
Fugitives- Ircin LoadH03t
Fugitives- Iron Lwd(T03>
Fugltlves-Inj Uell Load(D79>
Fugitives-lnj Uell LoadllW)
Fugitives-Inj Uell Load
FugitivH-ln; UeJl LoartlOTI
Fugitives- Rertd Treat Tank LMding
Fugitives- Aertd Treat Tank Loading
Fugitives- Aertd Treat Tank Loading
W»U for.
(»
VOC-Cont Solid
Aq Sldg/Slur
Dil Aqueous
Org Liquid
Org Sldg/Slunry
2-flua Aq/Org
Ag Sldg/Slur
Dil Aqueous
Org Liquid
Org Sldg/Slunry
2-Phase Aq/Org
Aq Sldg/Slurry
Dil Aqueous
Org liquid
Drg Sldg/Slurry
2-Ptiase Aq/Org
Aq Sldg/Slur
Dil Aqueous
Drg liquid
Org Sldj/Slurry
2-Fhase Aq/Org
Aq Sldg/Slur
Dil Aqueous
Org Liquid
Drg Sldg/Slurry
2-ttuse Aq/Org
Aq Sldg/Slur
Dil Aqueous
Org Liquid
Org Sldg/Slurry
2-Hiast Aq/Org
Aq Sldg/Slur
Dil Aqueous
2-Miase Aq/Org
Vol.- Con- Control
tlllty trol E.I. .Ion control offlcl.ncv
clsss lnd*K option (suppression]
<«> (S) («) (7)
All
All
All
All
All
All
All
All
All
All
All
nn
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
Subnerged loading
Sutwrged Loading
Submerged loading
Submerged Loading
Submerged loading
Sutmerged Loading
Submerged loading
Submerged Loading
Subierged Loading
Submerged Loading
Subwrged Loading
Monthly Inspt/Repalr
Monthly In>pt/Repilr
Monthly Inspt /Repair
Monthly Inspt/Repair
'Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repalr
Monthly Inspt/Repair
Monthly Inspt/Repalr
Monthly Inspt/Repair
Monthly Inspt/lepalr
Monthly Inspt/Repalr
Honthly Inspt/Repair
Ncothly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
. Monthly Jnspt/Scpair
Monthly Irispt/Repair
Monthly Inspt/Repair
Mc*nthly Inspt/Repalr
65.00
65.00
65.00
65.00
65.00
65.00
65.00
65.00
65.00
65.00
65.00
70.29
70.23
70.29
70.29
70.29
70.21
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
S.r-
vlco
II (o
(8)
15
IS
15
IS
IS
IS
IS
IS
IS
IS
IS
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
Cost
function
(•)
linear
linear
Li war
linear
Linear
11 mar
Linear
Li war
linear
linear
Linear
Linear
Linear
Linear
linear
Linear
Linear
Li war
Li r*ar
Li Mar
linear
Linear
Li rear
Linear
LiMar
liMar
Linear
LiMar
liMar
Linear
Li war
LiMar
Linear
Ll Mar
Tot.l cspltil
lflv««tJH*nt. •
(»•)
0.00
0.00
0.00
0.00
0.00
0.0(1
0.00
0.00
0.00
0.00
0.00
uia.oo
6318. (0
6318 00
6311.00
£311.00
631 S. 00
(311.00
(318.00
(311.00
(311.00
(311.00
(318.00
(318.00
(318.00
(318.00
(318.00
(318.00
(318.00
(318.00
6318.00
(318.00
(311.00
(318.00
b > q
(11)
O.*9000
0.70000
0.87000
0.19000
0.64000
0.89000
0.75000
0.92000
0.94000
0.78000
0.79M*
3.5700n
3.57ftf
.28nOO
I. I*OV>
2. 14040
2. 14040
2. 14040
2. 14040
0.11300
0.11300
O.IIW
n.HJOO
0. IUVK1
0.01(30
0.01(10
1.0I6JO
Cost
function
<1»)
LiMar
linear
LiMar
Linear
LiMar
Linear
LiMar
Lircar
LiMar
LiMar
Lircar
LiMar
Linear
liMar
Linear
LiMar
linear
LiMar
liMar
Linear
linear
linear
Lircar
Linear
Lircar
LiMar
Linear
LiMar
Liwar
Linear
Linear
Ll Mar
Linear
Linear
Annus 1
oporstlno cost
s b . Q
(13) (H>
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
o.oo
0.00
918.00
911.00
911.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918. CO
918.00
918.09
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
0.03000
0.04000
0.04000
0.05000
0.03000
0.05000
0.04000
0.05000
0.05000
0.04000
0.04000
0.14IOH
0.14100
n.'HIflri
0.94100
0.94100
«. 33710
0. 337
n,ft?9*>
(l. or-TXi
ft. W9 VI
O.IVWi
n. 004.10
0.00430
0.00470
             $•• not** «t «nd of tabl*
                                                                                                                                                                                          (continued)

-------
                                                                                                 TABLE 0-19  (continued)
00
cn
TML
process
indicator EM! a* Ion source
(» (2)
C Fugitive- Aertd Trut Tar* Loading
G fugitives- Aertd Treat Tank loading
H fugitives- Osct Treat Tank Loadirig
H Fugitives- Out Treat Tank loading
N Fugitives- Osct Treat Tank Loading
H Fugitives- Osct Treat TarA loading
H Fugitives- Ofct Treat Tank Loading
Fugitives- Storage I«p Loading
Fugitives- Storage tup Loading
Fugitives- Storage lap Leading
Fugitives- Storage l»p Loadirig
Fugitives- Storage lip Loading
fugitives- Treat lip loading
Fugitives- Treat lap Loading
fugitives- Treat lip Loadirig
fugitives- Treat lap loading
Fugitives- Treat lip Loading
Fugitives- Disp lip Loading
K Fugitives- Disp lip Loading
K Fugitives- Disp Uploading
K Fugitives* Disp lip Loading
C Fugitives- Dicp I«p loading
M Fugitives- TerMinal lr«*dtrq Sldg/Slur
Dil Aqitpwis
Org liquid
Org Sldg/Slurry
?-Ph»e Aq/Org
Vol.- Con-
til Ity irol Ei.ll.lon control
etmmm tnd*n option
(«)  («)
All
an
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
All
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Hor.thly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly tupt/lepair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Mgpthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Hor.thly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Monthly Inspt/Repair
Control
•fMcioncy
(aupprosalon)
(')
70.29
70.23
70.29
70.29
70.29
70.29
70.29
70.29
70,29
70.29
70.29
70.29
70.29
70.29
70. »
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
70.29
Sor-
vlc.
llfo
(•)
10
10
10
10
10
10
10
10
10
10
10
to
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
Coat
function
(•)
Linear
Linear
Linear
Linear
Linear
Linear
linear
Linear
Li war
linear
Linear
linear
linear
Linear
linear
Linear
linear
Linear
Linear
Linear
linear
Linear
Li Mar
Linear
Li war
LI war
Li war
Tot.l c.plt.l
invoatnont. S
(!•)
£318.00
£318.00
631). 00
£311.00
£318.00
1316.00
6311.00
£318. 00
6318.00
(318.00
£318.00
£318.00
£318.00
£318.00
£318.00
£318.00
£318.00
£318.00
£318.00
£318.00
£318.00
£318.00
£318.00
£318.00
£318.00
£318.00
£318.00
b x «
<»O
O.OIHO
0.01630
0.0(1340
0.08340
O.OIWO
O.OMW
0.08340
0.71400
o.7:»oo
0.73400
0.73400
0.73400
0.01130
0.01130
0.01130
0.01130
0.01130
O.I3MK)
0.13*00
0. 13(00
0. 13WO
0. IWO
0.8'VOfl
O.BWOO
0.8-WOO
ft.e'Vfin
o.sisoo
Coat
function
<«2)
Linear
11 war
Liwar
linear
Li war
Liwar
Liwar
liwar
Liwar
Liwar
Liwar
Linear
linear
Liwar
Liwar
Liwar
Liwar
liwar
Liwar
liwar
Liwar
liwar
Liwar
Liwar
Liwar
Linear
liwar
Annu. 1
operating coit
• b « Q
(13) (14)
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918. 0»
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
918.00
0.00430
0.00430
0.0??00
».«??««
0.0??00
n.n??oo
O.ft'IW
A. i
0.03?dO
O.n3r«o
n. 03^*0
A.?WO
0.23500
«.?3Vn
o.'rw
'i.?33
                 S««  not»»  at «nd of  t»bl».
                                                                                                                                                                                            (continued)

-------

proce..
indic.tor
(1)
Emli.lon .ourco
(2)
W..te form
(3)
Vole-
tillty
cla«
Con-
trol
Index
(5)
Eml..lon control
option
Control
efficiency
(auppreaeion)
(7)
Ser-
vice
life
Coat
function
Total c.plt.l
Inve.tment. *
a b > q
Coat
function
Annua 1
operating co.t
a b
> fl
•Thi. t.bl. confln. .11 co.t-r.l.tod d.t. n.c....r, to ..tl-t. control ,c..t l^.ct. .Itn th. Sourc

kTho d.finltlon. of column, for th. TSOF Proco.. Control /IU .ro:

   1  =  Tr.n.for, h.ndllng, .nd lo.dlng (THL)  proco..  Indlc.lor.
   t  -  cm)..Ion .ourco,
   3  =  W.,t. for. definition.
   4  =  Vol.til It, definition.
   E  =  E»i»ion control numorlc Indie.tor
   8  =  Emi.tion control definition.
   7  »  Suppro.iion control offlcloncy.
   8  =  Service life of control equipment (/r).
   9  »  Cost function de.cription.  for c.plt.l  lnv.itm.nt
  10  =  Fixed control co.t for c.pit.l Inve.tment.
  11  =  Throughput multiplier for  c.plt.l  Inve.tment.
  12  =  Co.t function de.cription  for .nnu.l operating co.t.
  13  a  Fixed .nnu.l  operating co.t.
  14  =  Throughput multiplier for  .nnu.l  operating  co.t.

-------
management unit, such as the surface area and waste retention time for
surface impoundments.  In general,  information at that level  of detail is
not available for all the TSDF.  For most TSDF,  only the total  throughput
of the waste management units is known.   Therefore, to estimate costs of
emission control, it was necessary to derive cost functions that estimate
control costs as a function of the waste management unit throughput as was
done for the TSDF emission factors.  The throughput data available for the
TSDF waste management units are total values.  For instance,  for treatment
surface impoundments, a particular facility may have a million gallons per
day throughput; however, that could be in one large impoundment or three
smaller impoundments.  This lack of unit-specific information prevents
rigorous determination of facility-specific emission and control cost
estimates.
     Although the information about the characteristics of specific waste
management units is limited, there are statistical data available with
which it is possible to describe certain characteristics of the units on a
national basis.  The Westat Survey conducted in 1981, for instance,
provides considerable statistical data useful for determining the national
distribution of sizes of storage tanks (storage volume), surface impound-
ments  (surface area), and landfills  (surface areas and depth).  With these
statistical data.it is 'possible to generate cumulative frequency distribu-
tions of unit size characteristics.  Much of these data, in fact, were the
bases for the selection of the model unit sizes.  Each model  unit has a
certain waste throughput and other design and operating characteristics;
multiple model units were selected for each waste management process to
represent the range of sizes nationally.  These model units served as the
basis for the development of emission estimates as well as control costs.
     The costs for controls applied to the model units were developed and
the relationship of control cost to throughput was computed for each of the
model units.  Because there are no data to determine which of the model
unit sizes most closely matches a management process in a particular
facility, a method of assigning the model unit costs (and emissions) to
each waste management unit  in each TSDF, nationally, was needed.  To this
end, a national average model unit was defined from the statistical
                                    D-87

-------
information on TSDF management units.  Each model unit size was assumed to
represent a certain portion of the nationwide cumulative frequency distri-
bution curve for that particular management process.  The weighting factor
for each management process model unit is the percentage of the cumulative
frequency for that model unit.  The weighted costs per megagram of waste
throughput were then determined by multiplying the weighting factor by the
total capital investment and annual operating cost for the corresponding
model unit.  These weighted costs were compiled for each management process
to constitute the control cost file used as input to the SAM.  This
methodology for developing weighted control cost factors is the same as
that used for emission factor determinations and is ah approximation of the
effects of economy-of-scale on nationwide control cost estimates.
0.2.6  Test Method Conversion Factor File
     An important aspect of any pollution control strategy applied to TSDF
involves identifying those hazardous waste streams that require control.
One means of accomplishing this is to establish control levels based on the
emission potential of the waste entering a particular management process.
Several test methods have been evaluated to quantify emission potential.
The test method selected to measure the waste stream emission potential,
which has been defined as the VO content of the waste, is steam distilla-
tion with 20 percent (by volume) of the waste distilled for analysis.  In
general, the VO test method results are a function of the volatility of
individual compounds because the amount of a particular waste constituent
removed from the waste sample and recovered for analysis depends largely on
volatility.  The test method results in essentially 100 percent removal and
a high distillate recovery for the most volatile compounds in the waste;
the removal and recovery of less volatile and more water soluble compounds
are less than 100 percent.  With a VO test method established, the VO
content of a hazardous waste can be measured and then compared to the
limits on VO content,  established as part of a control strategy, to
determine if emission controls are required for the specific waste stream.
     Test method conversion factors were developed,  based on laboratory
test data, to allow the SAM to simulate the VO test method numerically to
obtain VO measurements similar to those found in the laboratory.  In this
                                    D-88

-------
way the SAM can determine what waste streams in the data base would be
controlled for different VO action levels (waste VO concentration above
which controls must be applied to units managing that waste)  and, as a
result, define the affected population of wastes for a given  control
strategy.  For example, the waste data base used in the SAM contains
concentrations of specific compounds in specific waste streams.   These
compounds are assigned a surrogate designation on the basis of their
volatility.  The test method conversion factors are applied to each type of
surrogate to estimate how much of the surrogate would be removed by the
test method and contribute to the total measured VO.  The contribution of
each surrogate is then summed for the waste to estimate the VO content that
the test method would measure.  The only use of the test method conversion
factors is to estimate (from the data base on waste compositions) what the
test method would measure as the VO content of a waste stream.  This
estimated VO content is compared to the VO concentration limits to deter-
mine whether a specific waste stream would be controlled under a given VO
action level.  The regulated wastes that are identified for control are
used in the SAM to determine the nationwide impacts of the given VO action
level within a control strategy.
     In the development of the conversion factors, several synthetic wastes
containing nine select compounds, which represent a wide range of volatili-
ties, were evaluated for percent recovery using the test method.  The com-
pounds were present in different types of waste matrices that included
aqueous, organic, solids, and combinations of the three.  The recovery of
these different compounds in different synthetic waste matrices forms the
basis for the test method conversion factors.
     The approach was to assign each of the nine synthetic waste compounds
to its corresponding SAM volatility class based on vapor pressure and
Henry's law constant.  The normalized percent recovery was used to adjust
for recoveries that were either greater than or less than 100 percent.  The
normalized recovery for each compound in a given volatility class was aver-
aged to provide a single conversion factor for each class.  The results are
summarized in Table D-20 for each volatility class and type of waste
matrix.  The results indicate that the method should remove all of the
highly volatile compounds from the waste.  All of the moderately volatile
                                    D-89

-------
      TABLE  D-20.   SUMMARY  OF  TEST  METHOD  CONVERSION  FACTORS3
Volatility class
Very high
High
Moderate
Low

Aqueous
NA
1.0
1.0
0.2
Waste matrix
Organic
1.0b
1.0
0.3
QC

Solid
1.0b
1.0
0.5
Oc
NA * Not applicable.

aThis table presents factors that, when multiplied by the con-
 centration of a specific volatility class in the waste, provide
 an estimate of the volatile organic content that the test method
 would measure for the waste.

^Assumes that the test method will remove all of the highly
 volatile gases from the waste.

cAssumes that because of the very low vapor pressure for this
 category (<1.33 x 10~4 kPa) the test method will remove very
 little from the waste.
                             D-90

-------
compounds in an aqueous matrix are expected to be removed;  however,  only 30
to 50 percent of the moderately volatile compounds (conversion factors of
0.3 to 0.5) in an organic or solid matrix are expected to be recovered by
the method.
     A headspace analysis was also investigated as an alternative procedure
for covered tanks because emissions from this source are more directly
related to the vapor phase concentration than to the total  VO content
measured by steam distillation.  For the headspace analysis, a conversion
factor was also necessary to estimate the vapor phase concentration that
the headspace method would measure from a known waste composition.  The
vapor phase concentration is to be expressed  in kilopascals for comparison
with existing regulations for storage tanks.
     The conversion factors for the headspace method are given in
Table D-21.  When these  factors are multiplied by the concentration in  the
waste (expressed as weight  fraction) for each volatility class, the sum of
the  results  for each class  is  an  estimate of  what the headspace methods
would measure.  These  factors were derived  from  the  synthetic waste stud-
 ies,  and each  factor is  the average from all  compounds  that  are grouped in
 a given  volatility  class and  waste matrix.
      The headspace  conversion factors  are  used with  the waste compositions
 in the  SAM's data  base to estimate what the headspace method  would  measure
 for a given waste  stream.  The predicted method  results are then  compared
 to VO concentration limits  for storage tanks to  determine  whether controls
 are required.   This approach defines  the population  of  controlled wastes,
 which is used in  the SAM to determine the  nationwide impacts for control-
 ling covered tanks.
 D.2.7  Incidence and Risk File
      Health risks posed by  exposure to TSDF air emissions  typically are
 presented  in two forms:  annual cancer incidence (incidents per year
 nationwide resulting from exposure to TSDF air emissions)  and maximum
 lifetime risk (the highest risk of contracting cancer that any individual
 could have from exposure to TSDF emissions over a 70-year lifetime).   These
 two health risk forms are  used as an index to quantify health impacts
 related to TSDF emission controls.  Detailed discussions on the development
 of health  impacts data  are found in Appendixes  E and J of Reference  5.
                                     D-91

-------
       TABLE  D-21.   SUMMARY OF HEADSPACE CONVERSION  FACTORS
                  TO OBTAIN KILOPASCALS (kPa)a

                          	  Waste matrix
Volatility class          Aqueousb         Organic          Solid
High
Medium
441
26.2
24.8
5.10
3.93
0.09
       Low                  3.520           0               0
aThis table presents conversion factors that are multiplied by the
 concentration (as weight fraction) of the volatility class in a
 waste to estimate what the headspace method would measure for
 that class.  For example, with an organic waste containing only
 medium volatiles at a level of 0.1 weight fraction (10 percent),
 the headspace method results are estimated as 0.1 x 5.1 = 0.51
 kPa.

bThe results for aqueous wastes are capped by the vapor pressure
 of the waste constituent surrogate compound (i.e., if the
 predicted method results exceed the surrogates'  vapor pressure,
 then the vapor pressure should be used as the method
 measurement).
                             D-92

-------
     The Human Exposure Model  (HEM)  provided the basis in the SAM for
estimating annual  cancer incidence and risk to the maximum exposed indi-
vidual due to TSDF-generated airborne hazardous wastes.   The HEM is a com-
puter model that calculates exposure levels for a population within 50 km
of a facility using 1980 census population distributions and local (site-
specific) meteorological data.  The HEM was run for each TSDF using a unit
risk factor of 1 and a facility emission rate of 10,000 kg/yr.  The HEM
results were then compiled into risk and incidence files that can be ad-
justed to reflect the level of actual emissions resulting from implementa-
tion of a particular control strategy.  The site-specific HEM incidence and
risk values are adjusted within the SAM by the ratio of annual facility
emissions to 10,000 kg and by the TSDF unit risk factor to give facility-
specific estimates for the control strategy under consideration.  Individ-
ual dual facility incidences are summed to give the nationwide TSDF
incidence value.
D.3  OUTPUT FILES
     The SAM was developed to generate data necessary for comparison of
various TSDF control options in terms of their nationwide environmental,
health, economic, and energy impacts.  Therefore, emissions  (controlled and
uncontrolled), costs (capital, annual operating, and annualized), and
health impacts  (annual cancer incidence and maximum risk) that represent
impacts on a national scale are the primary outputs of  interest.  In
addition, the SAM was designed to provide data that could be  stored and
summarized in a number of ways.
     Through manipulation of the SAM  post-processor, emissions can be
summed and presented by management process  (e.g., nationwide  emissions  for
all open storage impoundments), and by source  (e.g., nationwide emissions
from process  losses, spills, or transfer  and handling).  On  a nationwide
basis, the emission and cost data are also  available  for each waste code,
for each waste  form, and  for each constituent  within  a  waste.   Emission  and
cost  data  are required  at  this level  of detail  for comparison and
evaluation of the  various  control strategies being examined.  Health
impacts, however,  are  expressed  in  terms  of overall nationwide  risk or
cancer  incidences.
                                     D-93

-------
 D.4  REFERENCES

  1.    Memorandum from Maclntyre, Lisa, RTI, to Docket.  November 4, 1987.
        Data from the 1986 National Screening Survey of Hazardous Waste
        Treatment, Storage, Disposal, and Recycling Facilities used to
        develop the Industry Profile.

  2.    Office of Solid Waste.  National Screening Survey of Hazardous
        Waste Treatment, Storage, Disposal,  and Recycling Facilities.  U.S.
        Environmental  Protection Agency.  Washington,  DC.  June 1987.

  3.    Memorandum from Maclntyre, Lisa, RTI, to Docket.  November 4, 1987.
        Data from the  National Hazardous Waste Data Management System used
        to develop the Industry Profile.

  4.    Westat,  Incorporated.   National  Survey of Hazardous  Waste
        Generators and Treatment, Storage and Disposal  Facilities Regulated
        Under RCRA in  1981.   Prepared for U.S. Environmental  Protection
        Agency.   Office of Solid Waste.   September 25,  1985.

  5.    Office  of Air  Quality  Planning and Standards.   Hazardous  Waste
        TSDF--Background Information  for Proposed RCRA  Air Emission
        Standards,  Draft EIS.   U.S. Environmental  Protection  Agency
        Research  Triangle Park,  NC.   EPA Publication No.  EPA-450/3-89-23
        March  1988.

  6.     Office of Water  and  Hazardous  Waste.   Application  for Hazardous
        Waste Permit-Consolidated Permits Program.  EPA Forms 3510-1  and
        3510-3.7.   U.S.  Environmental  Protection  Agency.   Washinqton   DC
        June  1980.

  7.     Code of Federal  Regulations.   Title 40, Part 261.  Identification
        and Listing of Hazardous  Waste.  U.S.  Government  Printing Office
        Washington, DC.   July  1,  1986.   p. 359-408.

 8.     Code of Federal  Regulations.   Title 40, Part 262.34(a).  Accumula-
        tion Time.  U.S.  Government Printing  Office.  Washington, DC.
        July 1, 1986.  p. 411.

 9.     Reference 4, p.  17.

10.     U.S. Office of Management and  Budget.  Standard  Industrial
        Classification Manual.   Executive Office of the President
       Washington, DC.  1987.

11.    Moody1s Investors Service, Inc.  Moody's Industrial Manual.  New
       York.   1982.

12.    North Carolina  Department of Commerce.  Directory of North Carolina
       Manufacturing Firms.   Industrial  Development Division.  Raleiah
       NC.  1984.  1985-1986.                                        y '
                                    D-94

-------
13.     Environmental  Information Ltd.   Industrial  and Hazardous Waste Man-
       agement Firms.  Minneapolis,  MM.   1986.

14.     U.S. Department of Commerce.   Census of Manufactures.   Bureau of
       the Census.  Washington,  DC.   1982.

15.     U.S. Department of Commerce.   Census of Mineral Industries.  Bureau
       of the Census.  Washington, DC.  1982.

16.     U.S. Department of Commerce.   Census of Retail Trade.   Bureau of
       the Census.  Washington,  DC.   1982.

17.     U.S. Department of Commerce.   Census of Service Industries.  Bureau
       of the Census.  Washington, DC.  1982.

18.     U.S. Department of Commerce.  Census of Wholesale Trade.   Bureau of
       the Census.   Washington, DC.  1982.

19.    Reference  14.

20.    Memorandum from Deerhake,  M.E.,  RTI,  to Docket.  November  20,  1987.
       RTI use  of the  1981  National Survey  of Hazardous Waste  Generators
       and Treatment,  Storage,  and  Disposal  Facilities Data  Base  (Westat
       Survey).

21     Memorandum from Deerhake,  M.E.,  RTI,  to Docket.  November  20,  1987.
       SAIC  nonconfidential  printouts of  the Industry Studies  Data  Base.

22.    Memorandum from Deerhake,  M.E.,  RTI  to  Docket. November 20,  1987.
       Printout of RCRA  K waste code  data base.

23.     ICF,  Incorporated.  The  RCRA Risk-Cost  Analysis Model.   Phase III
        Report and Appendices.   Prepared for the  U.S. Environmental
        Protection Agency.  Office of  Solid Waste.  Washington, DC.
       March 1, 1984.

 24.     Memorandum from Deerhake,  M.E.,  RTI, to Docket.  November 20, 1987.
        RTI use of the WET Model Hazardous Waste data base.

 25      Computer printout from the Illinois Environmental  Protection
        Agency.  Data Base of Special  Waste Streams.  Division of Land
        Pollution Control.  August 1986.

 26     Code of Federal Regulations.   Title 40,   Part 261.33(f).  Discarded
        Commercial Chemical Products,  Off-Specification Species,  Container
        Residues,  and Spill Residues Thereof.  U.S.  Government  Printing
        Office.   Washington, DC.  July  1, 1986.   p.  382-386.

 27.    Reference 7.
                                      D-95

-------
 28.     Hazardous  Waste  TSDF  Waste  Process  Sampling.   Volumes  I-IV.
        Prepared by  GCA  Corporation  for  U.S.  Environmental  Protection'
        Agency/Office  of Air  Quality Planning and  Standards, RTP,  NC.
        October  1985.

 29.     Memorandum from  Deerhake, M.  E.,  RTI,  to Docket.   December 30,
        1987.  U.S.  Environmental Protection  Agency.   Petroleum Refining
        Test  Data  from the  OSW  Listing Program.

 30.     Reference  27,  Volume  I,  p. 4-1 through 4-22.

 31.     Reference  27,  Volume  III, p.7-1  through 7-12.

 32.     Letter from  Deerhake, M.E.,  RTI,  to McDonald,  R.,  EPA/OAQPS.
        August 15, 1986.  Review of  Volumes I-IV of  "Hazardous  Waste
        Process Sampling" for test data  and OSW data on the petroleum
        refining industry.

 33.     Letter from  Deerhake, M.E.,  RTI,  to McDonald,  R.,  EPA/OAQPS.
        September  19,  1986.   Waste compositions found  in  review of field
        test  results.

 34.     Code  of Federal  Regulations.  Title 40, Part 261,  Subpart  C -
        Characteristics  of  Hazardous Waste and Subpart D  - Lists of
        Hazardous  Wastes.   U.S.  Government Printing Office.  Washington,
        DC.  July  1,  1986.  p. 373-386.

 35.     Letter from  Deerhake, M.E.,  RTI,  to McDonald,  R.,  EPA/OAQPS.
        October 1,  1986.  Approach for incorporating field test  data into
        the Waste  Characterization Data Base.

 36.     Dun and Bradstreet.   Million Dollar Directory.  Parsippany, NJ.
        1986.

 37.     Environ Corporation.  Characterization of Waste Streams  Listed in
        40 CFR 261  Waste Profiles,  Volumes 1 and 2.  Prepared for  U.S.
        Environmental  Protection Agency, Office of Solid Waste  Character-
        ization and Assessment Division.  Washington, DC.  August  1985.

 38.     Radian Corporation.   Characterization  of Transfer, Storage, and
        Handling of Waste with High  Emissions  Potential,  Phase  1.  Final
        Report.  Prepared for U.S.  Environmental  Protection Agency.
        Thermal Destruction Branch.   Cincinnati,  OH.  July 1985.

 39.     U.S.  Environmental Protection Agency.   Supporting Documents for the
       Regulatory  Analysis of the Part 264 Land Disposal  Regulations.
       Volumes I-III.   Docket Report.  Washington, DC.  August 24, 1982.
       Volume I,  p.  VIII-3.

40.    Reference 22, p.  2-17.
                                    D-96

-------
41.    Code of Federal Regulations.  Title 40, "Parts 261.31 and 32.
       Hazardous Wastes from Non-Specific Sources and Hazardous Waste from
       Specific Sources.  U.S. Government Printing Office.  Washington,
       DC.  July 1, 1986.  p. 377-379.

42.    Reference 34, p. 7.

43.    Reference 34, p. 35.

44.    Code of Federal Regulations.  Title 40,  Part 261.31.  Office of the
       Federal Register.  Washington, DC.  July 1, 1986.

45.    Metcalf and Eddy,  Inc.  Wastewater Engineering.  New York, McGraw-
       Hill Book Company.  1972.  pp. 231 and 304.

46.    Office of Solid Waste.  RCRA  Land Disposal Restrictions Background
       Document on the Comparative Risk Assessment.  U.S.  Environmental
       Protection Agency.  Washington, DC.  December 27,  1985.   170 pp.

47.    Reference 46.

48..    Reference 44.

49.    Code of Federal  Regulations.   Title 40, Part 261.33.  Office of the
       Federal Register.  Washington, DC.  July  1,  1986.

50.    U.S. Environmental Protection Agency.  Hazardous Waste  Management
       System:  General.  45 FR  33115.   May 19,  1980.

51.    Reference 34.

52.    Industrial  Economics, Inc.  Regulatory Analysis  of Proposed
       Restrictions  on Land  Disposal of  Certain  Solvent Wastes.   Draft.
       Prepared  for U.S.  Environmental Protection Agency, Office of  Solid
       Waste.   Washington,  DC.   September 30,  1985.   p. 3-15.

53.    Code of Federal Regulations.   Title 40,  Part 261.32.   Office  of the
       Federal  Register.   Washington,  DC.   July  1,  1986.

54.    U.S.  Environmental Protection Agency.   Hazardous Waste Management
       System;  Land Disposal Restrictions:   Final Rule.   51  FR 40572.
       November 7,  1987.

55.    Reference 2,  Exhibit A-9.

 56.     Reference 52, p. 3-15.

 57.     Research Triangle Institute.   Hazardous Waste Treatment,  Storage,
        and Disposal Facilities:  Air Emission  Models, Draft Report.
        Prepared for U.S. Environmental  Protection Agency.  Office of Air
        Quality Planning and Standards.   Research Triangle Park,  NC.   March
        1987.
                                     D-97

-------
 58.    Reference 57.

 59.    U.S. Environmental Protection Agency, OAQPS.  Physical-Chemical
        Properties and Categorization of RCRA Wastes According to Volatil-
        ity.  U.S. Environmental Protection Agency.  Research Triangle
        Park,  NC.  Publication No. EPA-450/3-85-007.  February 1985.  p.
        1D •

 60.    Reference 59.

 61.    Merck  Index.   Ninth Edition.   Merck and Co., Inc. Rahway, NJ. 1976.

 62.    Verschueren,  K.   Handbook on  Environmental  Data and Organic Chemi-
        cals.   New York,  Van Nostrand Reinhold Company.  1983.

 63.    Environ Corp.  Characterization  of Waste Streams Listed in 40 CFR
        Section 261:   Waste Profiles. Volumes I and II.  Prepared for U.S.
        Environmental Protection Agency.   Washington,  DC.  August 1985.

 64.    University of Arkansas.   Emission  of Hazardous  Chemicals  from
        Surface and Near-Surface Impoundments to Air.   Draft Final  Report
        EPA  Project No. 808161-02.  December 1984.

 65.    Reference 59.

 66.    Memorandum from Zerbonia,  R.,  RTI,  to Hustvedt,  K.  C.,  EPA/OAQPS.
        Development of waste  constituent categories'  (surrogates)  proper-
        ties for  the  Source Assessment Model.   December 30,  1987.

 67.     Reference 66.

 68.     Reference 66.

 69.     Reference 61.

 70.     Reference  62.

 71.     Reference  62.

 72.     Reference  56.

 73.    Chemical Manufacturers Association.  Comments of  the Chemical
       Manufacturers Association on the Environmental Protection Agency
       Document  "Hazardous Waste TSDF-Background Information for Proposed
       RCRA  Air Emission  Standards -  Volumes  I and  II."  Washinqton  DC
       July  11, 1988.  105 p.

74.    Memorandum from Coy, D.,  RTI,  to  Docket.  January 1989.  Investi-
       gation  of and  Recommendation for  Revisions to Aerated Model Unit
       Parameters Used in the Source  Assessment Model.
                                    D-98

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
II. REPORT NO.
 :PA-450/3-89--009
                              2.
                                                            3. RECIPIENT'S ACCESSION NO.
 4. TITLE AND SUBTITLE
 Hazardous  Waste Treatment,  Storage,  and Disposal
 Facilities—Background  Information  for Promulgated
 Organic Emission Standards  for  Process Vents and Equip-
           ment Leaks	
                                      5 REPORT DATE
                                          June 1990
                                     6. PERFORMING ORGANIZATION CODE
                                                            8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Air Quality  Planning  and Standards
Environmental Protection Agency
Research Triangle Park, North Carolina  27711
                                                            10. PROGRAM ELEMENT NO.
                                      11. CONTRACT/GRANT NO.
                                         68-02-4326
 12. SPONSORING AGENCY NAME AND ADDRESS
 DAA for Air Quality Planning  and  Standards
 Office of Air, Noise, and  Radiation
 U.S.  Environmental Protection Agency
 Research Triangle Park,  North Carolina  27711
                                                            13. TYPE OF REPORT AND PERIOD COVERED
                                      14 SPONSORING AGENCY CODE
                                         EPA/200/04
 15. SUPPLEMENTARY NOTES
 16. ABSTRACT
       Standards for the  control  of organic air emissions  from hazardous waste
 treatment,  storage, and  disposal  facilities (TSDF) and  waste solvent treatment
 facilities  (WSTF) are  promulgated under the authority of  Section 3004(n) of the
 1976 Resource Conservation  and  Recovery Act (RCRA).  These standards would apply to
 certain process vents  associated with distillation and  stripping equipment at
 WSTF (and at TSDF, if  applicable) and to fugitive emissions from equipment leaks
 at  TSDF where the waste  stream  (or its derivatives)  contain 10 percent or more
 total  organics.  This  document  contains summaries of public comments received on
 the proposed rule (February 5,  1987), EPA responses, and  a discussion of differences
 between the proposed and final  standards.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
 Air  pollution
                                               b.lDENTIFIERS/OPEN ENDED TERMS
                                                   c.  COSA'I I Field/Group
 Benzene
 Carcinogenic
 Equipment leaks
 Hazardous waste
 National  emission standards
  for hazardous air pollutants
 Process vents
Recycling
Treatment, storage,  and
 disposal facilities
Volatile organics
Waste solvent  treatment
            facilities
Air pollution control
13 B
 18. DISTRIBUTION STATEMENT

    Unlimited
                         19. SECURITY CLASS (Tins Report)
                            Unclassified
                                                                          21. NO. OF PAGES
L
                                               20. SECURITY CLASS (Thispage)
                                                  Unclassified
                                                   22. PRICE
 EPA Form 2220-1 (R«v. 4-77)   PREVIOUS EDITION is OBSOLETE

-------
U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Boulevard, 12th Floor
Chicago,  IL  60604-3590

-------