United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA-450/2-90-012
July 1990
AIR
      EPA
          Designing and Implementing
         an Air Toxics Control Program:
        A Program Development Manual
          for State and Local Agencies

-------
   DESIGNING AND IMPLEMENTING
 AN AIR TOXICS CONTROL PROGRAM:
 A PROGRAM DEVELOPMENT MANUAL
  FOR STATE AND LOCAL AGENCIES
             Prepared by:

   Noncriteria Pollutant Programs Branch
Office of Air Quality Planning and Standards
   U. S. Environmental Protection Agency
    Research Triangle Park, NC 27711
               Juiy 199C

-------
                                   FOREWORD

      This program development manual was prepared by the U. S. Environmental
Protection Agency (EPA) at the request  of the State and Territorial Air Pollution
Program Administrators and the Association of Local Air Pollution Control Officials
(STAPPA/ALAPCO).  The manual is intended to assist State and local  (S/L) air
pollution control agencies in making informed decisions about the development of air
toxics control programs and to  enhance consistency among S/L program  approaches.
In particular, insight is provided into how to  make critical decisions regarding program
scope and stringency, evaluation of program impacts, and achievement of overall
program effectiveness.  While the information contained in this manual does not
represent EPA policy,  it should prove useful  to many S/L agencies now  actively
engaged in air  toxics program development.
      In order to  develop the manual, the advice and experience of S/L agencies at
various  stages in their own program  development were  sought and incorporated into
this manual.  The  State of Maryland, currently in the implementation phase of their
air toxics program, provided advice and information based on the State's own
experience with program design.  The States  of Colorado and Illinois, currently taking
on the challenges of planning air toxics programs,  helped identify issues and  decision
points faced by States  in the early stages  of program development. The information
in the manual attempts to address these  issues by illustrating the experiences of S/L
agencies and EPA. in making air toxics program decisions.
      The  control of toxic air pollutants is a high priority for both S/L  and  EPA.
Studies  have  revealed numerous sources of air toxics nationwide, including  both point
sources  (e.g., smokestacks and process vents)  and area sources (e.g., vehicles  and
other highly dispersed  emission points).   Data compiled under Section 313  of the
Superfund Amendments Reauthorization  Act  show that  2.7 billion pounds of toxic
chemicals are released to the air annually from industrial sources alone;  many of the
chemicals released are known or suspected cancer-causing agents.
      In response to the  air toxics problem,  EPA developed and implemented the
National Air  Toxics Strategy in 1985. This policy  calls  for action on high risk point
                                        111

-------
sources, the complex mixture of pollutants in urban areas, and mobile sources.  The
strategy also  strongly supports the development of S/L air toxic control  programs as
an important part of an overall national effort to reduce toxic air emissions.   In
particular, the strategy has a goal that each S/L agency develop a program capable of
addressing high risk point sources and carry  out  any federally adopted rules.   Support
by EPA for achieving this goal includes  development grants, technical assistance,
operation of  technical assistance  centers, air  quality monitoring programs,  and
guidance documents.
       Since  1985, a majority of S/L air pollution control agencies have established
programs to  address their air toxics problems.  A significant trend is a growing
similarity in S/L air toxic programs.  While  some programs give somewhat greater
emphasis to one  control approach over another,  virtually all programs today include a
combination  of at least two basic program elements:  use of control technology  and
ambient impacts  requirements.   In fact, use of these mutually reinforcing control
program elements has emerged as the predominant feature of S/L programs.  For this
reason, the manual uses this multi-element approach as the baseline for discussing
approaches to control of toxic  air emissions.
       To date, most  S/Ls have focused their regulatory efforts primarily on individual
high risk point sources; thus, high risk point  source control is the focus  of the manual.
However, both S/Ls and EPA  recognize the importance of addressing the  problems of
area sources, motor vehicles, and other clusters of sources.  Research and  regulatory
activities are ongoing in these  areas. The emphasis of this manual on high risk pomr
source control is not  intended  to  minimize the importance of these other  problem
areas.  .
       At the tune this manual was developed, efforts were underway to amend the
Clean  Air Act.  While  the nature of these amendments will become increasingly clear,
it is already  clear that  a substantial increase in  the number of sources regulated under
federal rule is expected.  S/L agencies currently engaged hi rule development should
consider designing their programs to complement the  anticipated control approach,
source/pollutant  applicability, and implementation framework set forth by the current
legislative proposals.  Accordingly, S/L agencies  with developing programs are further
                                         iv

-------
advised to consult with applicable EPA officials to incorporate the current best
understanding of the future federal program.
      The sections that follow include: Section  1.0, Designing the Basic Program
Approach; Section 2.0, Making Critical Design Choices: Scoping the Program; Section
3.0, Evaluating Program Impacts; and Section 4.0, Avoiding Pitfalls.

-------
TABLE OF CONTENTS                                          PAGE

     FOREWARD  	   iii
     LIST OF TABLES	ix
     LIST OF FIGURES	ri
     ACKNOWLEDGEMENTS	   xiii
1.0  DESIGNING THE BASIC PROGRAM APPROACH  	   1

     1.1   DEVELOPING AN AIR TOXICS CONTROL PROGRAM  	   1

          1.1.1  Why Have an Air Toxics Program?  	   2
          1.1.2  Defining the Problem and Designing the Program	   2
          1.13  Implementing, Enforcing and Evaluating the Program	   7

     1.2   .APPROACHING .AIR TOXICS CONTROL
          COMPREHENSIVELY	   8

          1.2.1  Converging  Program Approaches 	   8
          1.2.2  Putting the  Elements into a Comprehensive Framework ....  10


2.0  MAKING CRITICAL DESIGN CHOICES  	   1

     2.1  SELECTING POLLUTANTS OF CONCERN	   2

          2.1.1  General Approaches  	   2
          2.1.2  Steps for Selecting Pollutants  	   4

     2.2   SELECTING SOURCE CATEGORIES OF CONCERN	  13

     2.3   REGULATING EXISTING SOURCES	  16

     2.4   SELECTING SIZE AND EMISSION CUTOFFS/EXEMPTIONS . .  20

     2.5   TIMING/PHASING OF PROGRAM COVERAGE	  20

     2.6   AGENCY RESOURCE REQUIREMENTS 	  21


3.0  EVALUATING PROGRAM IMPACTS  	   1

     3.1   WHAT ARE THE COSTS?   	   1
                                Vll

-------
     3.2   SELECTING A METHODOLOGY: MARYLAND'S APPROACH  .   2

           32.1 Compliance Costs  	   3
           3.2.2 Administrative Costs	   7
     3.3   SELECTING A METHODOLOGY: NORTH CAROLINA'S
           APPROACH	   7

     3.4   PLANNING A COST OR ECONOMIC IMPACT EVALUATION .   11
4.0  AVOIDING PITFALLS 	    1

      4.1   DEFENDING THE PROGRAM ALONG THE WAY	    1

      4.2   ENSURING EFFECTIVE COMMUNICATION  	, .    2
           4.2.1 Communicating Early and Often ....................  2
           4.2.2 Understanding the Nature of
                  Environmental Communication ....... . ............  2
           4.2.3 Communicating Program Objectives ..................  6
           4.2.4 Meeting Your Communication Objectives  ..............  8
           4.2.5 Communicating About Risks ...... . ...............  12
           4.2.6 Working With the Press  .........................  14

      4.3   COORDINATING WITH  OTHER REGULATORY
           PROGRAMS  ....................................  15

           43.1 Recognizing the Commonality of Different Regulatory
                  Programs ..................................  16
           4.3.2 Coordinating with State Implementation Plans  fSIPV. ......  17
           4.3.3 Coordinating with Other Toxics-Related Programs ........  27
      4.4 MEASURING PROGRAM EFFECTIVENESS
           4.4.1  Need for an Air Toxics Program Performance Measurement
                 System  ....................................  34
           4.4.2  Performance Criteria ...........................  36
           4.43  Example Measurement Techniques  .................  .  38
                                   VIII

-------
LIST OF TABLES                                           PAGE
1-1   SUMMARY OF TECHNOLOGY ELEMENT	14/15
1-2   CITATIONS TO CONTROL TECHNOLOGY
     REQUIREMENTS	  19
1-3   SUMMARY OF AMBIENT IMPACT ELEMENT	22/23
1-4   EQUIVALENT UNCERTAINTY FACTOR FOR 24-HOUR
     AVERAGING TIME  	27
1-5   LIST OF POLLUTANTS FOR WHICH INHALATION
     REFERENCE CONCENTRATIONS (RfCs) HAVE BEEN
     DEVELOPED                                      29
1-6   EXAMPLE CALCULATION OF AAL FROM UNIT RISK
     FACTOR 	30
1-7   POLLUTANTS WITH EPA INHALATION UNIT RISK
     FACTORS  	31
1-8   CITATIONS TO AMBIENT IMPACT REQUIREMENTS	34
1-9   PRIMARY EPA SOURCES OF RISK ASSESSMENT
     INFORMATION  	39
2-1   CASE EXAMPLES:  BASIS FOR POLLUTANT
     SELECTION	  6/7
2-2   LISTS OF CARCINOGENIC COMPOUNDS 	8
2-3   SOURCE CATEGORIES S/L's REPORT EXEMPTED  	17/18
2-4   DE MINIMIS LEVELS REPORTED BY S/L AGENCIES  . . 22/23
3-1   CONSIDERATIONS IN THE PLANNING OF
     REGULATORY IMPACT EVALUATIONS 	  12
4-1   LIKELY ADVERSE COMMENTS AND SUGGESTED
     RESPONSES	3/4/5
4-2   TOXICS - RELATED PROGRAMS INFORMATION  	28
4-3   EXAMPLE PROGRESS MEASUREMENT TECHNIQUES .... 39
                             IX

-------
LIST OF FIGURES                                          PAGE
1-1   STEPS IN PROBLEM DEFINITION	4
1-2   DESIGNING AN AIR TOXICS PROGRAM	6
1-3   COMPREHENSIVE APPROACH TO CONTROLLING
     AIR TOXICS EMISSIONS	9
1-4   FACTORS TO CONSIDER IN DESIGNING A
     COMPREHENSIVE AIR TOXICS CONTROL PROGRAM .... 12
1-5   KEY DECISION POINTS FOR CONTROL TECHNOLOGY
     ELEMENT	 17
1-6   KEY DECIsioN POINTS FOR AMBIENT IMPACT
     ELEMENT	24
                              XI

-------
                            ACKNOWLEDGEMENTS

      This manual represents a collaborative effort between the U. S. Environmental
Protection Agency (EPA) and State and local (S/L) air pollution control agencies.
Information and case examples  presented in this manual were gathered from a variety
of sources,  including numerous  contracts with S/L agencies over a long period of
time.  The  EPA would like to specifically acknowledge the contributions of the State
of Maryland whose personnel provided extensive input  and assistance to EPA during
the preparation of the manual.  The EPA is also grateful  to personnel  from the
States of Colorado and Illinois  for reviewing and commenting on the manual several
times  prior to its publication.
      The  EPA's effort to provide ihis document was also ably supported by Frank
Sheffield of Radian  Corporation, Alice Pelland, of ABB Environmental Services, Inc.
and Jim Rickun of RMT, Inc.  The EPA authors were Martha  H. Keating and
Michael A. Truma, both of the Office of Air Quality Planning and Standards.

-------
    SECTION 1.0
DESIGNING THE BASIC
PROGRAM APPROACH

-------
              1.0  DESIGNING THE BASIC PROGRAM APPROACH
                            Highlights of This  Section
            Reasons for developing an air toxics program include prevention
            of future health problems, the need for a system  to handle high
            risk  sources,  and  a  national  policy  supporting  3/L program
            development.

            Air toxics program development begins with assessing the problem
            and  designing the program;  extensive  data  collection  is  not
            necessary to get started.

            Implementation, enforcement and evaluation follow.

            A comprehensive  approach,  combining control technology  and
            ambient impacts  requirements, is  a  good  starting  point  for
            development of a  long-term  program.

            A comprehensive  approach  capitalizes on the  best features of
            individual elements and provides flexibility for future changes.
1.1    DEVELOPING AN AIR TOXICS  CONTROL PROGRAM


      This subsection describes how State and local (S/L) agencies proceed from
recognizing the need for additional air toxics control to designing and implementing a

regulatory program.  This subsection will be most useful to agencies jusi beginning a
program.
                                       1-1

-------
       1.1.1    Why Have an Air Toxics Program?


       Understanding the need for  an air toxics program is an important issue that
should be addressed early.  Some of the main reasons for developing a program
include:
      o      Protection of Public Health - The primary goal of S/L pollution control
             agencies, and the basic argument in favor of an air toxics program, is
             the protection of public health.  Studies associating air toxics with
             elevated cancer levels  and other adverse health effects, as well as
             monitoring data and data from Title HI of the Superfund Amendments
             Reauthorization Act (SARA Title HI)  showing large quantities  of toxics
             being emitted, support the conclusion that there is an air toxics problem
             that needs  10 be  addressed.  The  public"'? demand  for clean air and
             protection from  exposure  to air toxics  is also a basis for  developing an
             air toxics program.   Some S/L agencies have developed programs  as a
             reactive or curative measure to address problems  that already exist.
             Many S/L  agencies now developing programs reason that the need to
             prevent future air toxics problems (and subsequent health risks) is a
             legitimate basis for justifying a program.  This  "preventative approach" to
             air toxics control allows agencies to act rather  than react.

      o      National policy supporting S/L air toxics control programs - As
             mentioned  in the Foreword, the National Air Toxics Strategy strongly
             supports  S/L program development. This  strategy was prompted  by
             EPA studies suggesting potentially significant health risks from  exposure
             to air toxics. The current  Federal program will be strengthened with
             adoption of Clean Air Act amendments.

      1.1.2  Defining the Problem  and Designing the Program
       In  the early stages of program  development, a S/L agency collects and
evaluates  information to define the  specific tasks  that must be  completed to develop a
program.  At this point, the agency's list of action items might include:


       o     Defining the basic steps   - The  S/L agency should define the steps
             required to  define the air toxics problem and design the air toxics
             program.  The steps might include completing a review of legal
             authority, preparing of draft regulations or guidelines, and developing
                                         1-2

-------
             a communication strategy to facilitate program approval and
             implementation.

      o     Problem definition.  A large data collection effort is not a prerequisite
             for defining the problem and,  in fact, can deplete agency resources
             without achieving progress toward program approval.  SARA Title HI
             data  and a point source screening survey may be enough to launch a
             program.  If a preventative approach is adopted, little effort needs to be
             expended on documenting existing toxics-related health problems.

      o     Identification of air toxics  of concern -  Some  S/L agencies use an open-
             ended  approach, regulating any  noncriteria pollutants: others use a
             specific list.  Identifying pollutants of concern can start with a review of
             various State and national lists (e.g., EPA's lists associated with proposed
             Clean Air Act amendments), or lists from  the National Institute of
             Occupational Safety  and  Health (NTOSH),  or  the American  Conference
             of Governmental Industrial Hygienists (ACGIH).  These lists can be
             reviewed in conjunction with emission inventory or SARA Title ffl data.
             Some S/L agencies have undertaken chemical-specific health effects
             reviews, and selected chemicals  based on these reviews.

      o     Identification of air toxics  sources - To  define what sources will  be
             included in the program, sources of  air toxics emissions  could  be
             identified using source surveys, reviewing emission inventory or SARA
             Title HI data, or reviewing various national databases that can help
             identify source  categories  of concern (see Section 2.2).

      o     Review of agency legal authority - Where any questions exist as to
             whether or not the agency has the authority to regulate  air toxics,
             confirmation of legal authority to regulate  air toxics is recommended.
             Most S/L agencies have  not needed additional legislative authority, and
             have undertaken policy or guidelines development or ruiemaking via
             tneir general air pollution  control authority.

      o     Consideration  of interim  measures -  Because program development
             requires a significant  period of time, several S/L agencies implemented
             interim control strategies such as informally reviewing new and existing
             sources for toxics as  they  apply  for construction and/or operating
             permits.  Others have screened  existing sources to help define the
             problem and encourage voluntary actions by problem sources.  These
             interim measures can mitigate air toxics problems until a full program is
             in place, and can also build a valuable  base of knowledge and operating
             experience.

Figure 1-1 summarizes key issues  in the problem definition step.
                                        1-3

-------
What are the objectives
for program development?
What are the air toxics
of concern?
Defining the air toxics problem
Review of legal authority
Preparation of draft policy
or regulations
Development of communication
strategy
Other States lists
National lists
Emission inventory data
Health effects reviews
of individual pollutants
What are the sources
of air toxics?
Is there adequate legal
authority for regulating
air toxics?
What interim measures can
be implemented during
program design/approval?
Source surveys
Emission inventories
SARA Title m data
Other State/national lists
Risk assessments
 General authority
 Authority to control
 existing sources
 Construction/Operating permit authority
 Source registration
 Authority for standards/guidelines
 Autnonry to regulate area sources
 Relationship to federal requirements
 Review vs. screening of sources
 Review of new sources only
 Review of high risk sources
 Source registration
                  Figure  1-1.  Steps in Problem Definition.

-------
      Once the S/L agency has defined the problem, the S/L is ready to  design an
air toxics control program.  As illustrated by Figure 1-2, there are several  key points
to consider:
      o      Development of short- and long-term program goals and  milestones -
             The S/L decides what its air toxics program will accomplish and when.
             Many S/L agencies have already accomplished this via EPA's Multiyear
             Development Plan (MYDP) process.

      o      Development of communications  strategy - Communications objectives
             are denned, target audiences identified, communication techniques
             selected and messages refined during this step.  For reasons discussed
             further in Sections 4.1 and 4.2, an effective communications strategy can
             be a key to successful program development.

      o      Development of criteria for program evaluation - Beginning with the list
             of program goals, the S/L decides how to  measure progress made
             toward each objective.  Selecting measurable objectives, such as the
             number of  sources reviewed and  controlled annually,  is essential for
             evaluating program progress and  results.

      o      Selection of control approach - The  key points to consider here are:
             Will the agency base  its program on promulgated rules or standards or
             will it opt for  a policy or informal guidelines?  What will be the basis of
             emission  control (e.g., pollution prevention, control technology  or
             ambient impacts)?  (See Section  1.2 for further discussion of control
             approaches).

      o      Definition of program scope - The S/L considers its program goals, and
             the information  it has gathered,  to determine which pollutants  and
             sources should  be addressee first.  Program scope includes issues  such as
             pollutant selection, source selection,  and de minimis levels or
             exemptions.  (See Section 2.1 and 2.2  for further discussion).

      o      Assessment of resource requirements - This point is closely related  to
             program scope  and stringency.  More extensive  source coverage and
             review requires correspondingly greater staff resources.  Phasing of
             implementation can help fit the  program to available resources.  A
             projection of necessary resources  will likely be needed for budget
             requestpurposes.

      o      Determination of program stringency - This includes establishing control
             technology and/or emission reduction requirements for each source or
             pollutant category.  For the ambient impact element, methods  for
                                        1-5

-------
                          PROGRAM GOALS
            COMMUNICATION
             STRATEGIES
                      PROGRAM
                    EFFECTIVENESS
                      MEASURES
                          CONTROL APPROACH
                         POLLUTION PREVENTION
                 CONTROL
               TECHNOLOGY
                       AMBIENT
                       IMPACTS
                    \
                           RISK ASSESSMENT/
                          RISK MANAGEMENT
PROGRAM SCOPE
         RESOURCES
     STAFF
PROGRAM
IMPACTS
 PROGRAM
COMPLIANCE
                  PROGRAM
                 STRINGENCY
  REPORTING
REQUIREMENTS
              Figure 1-2. Designing an Air Toxics Program

-------
             developing ambient guidelines are needed for each pollutant or category
             of pollutants covered.

      o      Definition of staff responsibilities - The agency must decide who will do
             what during  program implementation and how many and what type of
             personnel will be required for the job.

      o      Evaluation of program impacts - Agency procedures or  policies may
             require evaluation of the cost, economic, and  administrative impacts of
             the proposed program.  This  assessment  may, in turn, influence program
             design or implementation decisions.

      o      Determination of compliance  schedules - The agency may decide during
             program design that, while it  would be appropriate to have the air toxics
             control program broad in scope, it is not ready to tackle such  a broad
             scope immediately.  If this is  the case, the agency should  use  a phased
             approach.  Several agencies have adopted a phased approach  with
             defined compliance  schedules.  For  example, high  risk point sources
             could be  addressed  first on a prioritized  basis.

      o      Definition of reporting requirements - Reporting is often accomplished
             through the existing permitting process because States have the authority
             to request information from sources through permit applications and
             source registrations.
      1.13   Implementing, Enforcing and Evaluating the Program


      At this point the S/L agency begins to put its control program into operation

through:


      o      Adoption of rules or policies - The agency's program is officially
             launched with the approval of appropriate rules, guidelines, policies, or
             procedures.  Some S/L agencies have chosen to launch their initial
             program through New Source Review while  others have initiated more
             comprehensive programs, including incorporation into the  control
             program of existing  sources for several pollutants and source types.

      o      Resource acquisition - Resources include funding, equipment and staff.
             For example, many  S/L agencies have recruited staff with  specialties
             such as toxicology and chemistry.  Some States finance their programs
             through  permit fees or other dedicated sources of revenue (e.g., vehicle
             tag fees).
                                        1-7

-------
            Initiation of source-specific review - Depending on the program design,
            this may mean starting permit application review and establishing permit
            conditions and compliance schedules.  For programs  covering existing
            sources, this may mean source registration and screening.

            Initiation of information collection activities  - For example, an ambient
            monitoring network might be established to  gather information on
            conditions near high risk sources  or in certain urban areas. Or, a source
            survey (i.e., questionnaire) or review of SARA Title  HI data could be
            undertaken.

            Coordination with other programs - For efficiency and enhanced
            effectiveness, agencies coordinate  with  other programs such as ozone
            control measures, community right-to-know, paniculate matter (PM,0)
            control, etc.  Often strategies can be designed to utilize additional
            controls that  simultaneously attain the  goals  and objectives of both air
            toxics and criteria pollutant  programs.
12    APPROACHING AIR TOXICS  CONTROL COMPREHENSIVELY


      This subsection presents guidance on selection of an agency's basic air toxics

control approach.  This subsection will be useful to all S/L agencies, but especially

those in the process of designing their program.


      1.2.1   Converging  Program Approaches


      Although  some programs emphasize one  control approach more than  another.
the  majority of S/L programs use a combined strategy made up of two basic program

elements:  control technology and ambient impact requirements.  A 1989 survey of
S/L  programs by the State and Territorial Air Pollution Program Administrators  and

the  Association of Local  Air Pollution Control Officials (STAPPA/ALAPCO) showed
that  most  agencies have adopted this  more comprehensive approach to controlling air
toxics. Figure 1-3 illustrates the comprehensive approach.
      Two primary factors underlie this strategy.  First, each individual  approach  has
inherent limitations that can be offset by adoption of  a combined approach.   Second,
                                       1-8

-------
          • Technology-based

          • Applies to specific source
           categories and/or pollutants

          Specifies minimum control
           requirements

          Achieves relatively
           uniform reductions

          • Promotes reasonable certainty
           re: industry requirements

          • Promotes pollution
           prevention
- Health-based


-  Applies to specific pollutants
- Standardizes air quality
  levels

-  Inhalation reference doses


- Ambient limits based on
  on occupational levels

-  Pollutant-specific
  evaluations
                                  •.s  -V^ •.<
                                 "^ <;  '^VSX ^

                                Risk^e^nW^
                                _JX - S^-'^T^-"0^^^'^
                                RISK p!a»agein«f|i.t.
                           - Useful ior tailoring control
                             level to degree of health
                             concern

                           - Applicable to carcinogens
                             and noncarcinogens

                           - Based on target risk levels

                           - Can be used to establish
                             ambient limits

                            - Can be used to examine
                             post-control residual risks
Figure 1-3.  Comprehensive Approach to Controlling Air Toxics Emissions

-------
because of the complexity of toxic air pollutants and emission sources, no one
regulatory approach is likely to address fully the air toxics problem.
      Under a comprehensive approach, individual program  elements are combined
into an  integrated control strategy.  The individual elements  form a complementary
set of requirements.   In this way, the program can effectively address sources of air
toxics that pose significant public health problems.  For example, a program might
include  a Best Available Control Technology (BACT) requirement plus ambient
impact requirements for new sources.  In some cases, the optimum mix of controls is
achieved based on pollution prevention,  control technology and ambient air
considerations.  In others, a minimum  level of control technology can be assured
before addressing issues regarding the  acceptability of the remaining residual  risk.
      An advantage of a comprehensive approach  is that it  combines  the best
features  of the individual  control elements.  For example, the reasonable  certainty of
emissions reduction of a control technology approach is combined with the  site-specific
protection provided by an ambient  impact approach.  These  advantages can be further
enhanced by use of risk assessment/risk management techniques to evaluate individual
sources.
      A second  advantage is that a comprehensive approach provides a framework
within which to address new pollutants or sources as additional information or
resources become available.  Air toxics control is a long-term process.  A first
generation  program, for example, may focus only on certain  pollutants emitted by a
limited  number of sources.  As  a program evolves  and matures, a second or  third
generation program may cover a much broader array of pollutants  and sources,
including area, nontraditional, and mobile sources.  A comprehensive approach
provides a  framework for responding to these and  other problems as they are
identified.

      1.2.2  Putting  the Elements  into a Comprehensive Framework

      As mentioned  above, a comprehensive approach combines individual program
elements into an integrated control program.  To  accomplish this, several key
                                        1-10

-------
decisions are required  Figure 1-4 illustrates the key factors to consider for each step
of a comprehensive approach.
      For any type of program, whether it emphasizes control technology or ambient
impacts, one of the first major decision points is the selection of the basis for the
control program.  At this point, an agency is faced with the decision of how, or
whether or not to combine control technology and ambient  impacts  requirements,
whether to incorporate  risk assessment/risk management into the program,  and how
to incorporate pollution prevention strategies into the program.  It is important to
remember  that decisions about the basis of the control program are closely related to
decisions about program scope, compliance schedules, and resources. The number
and  type of pollutants and sources covered affect an agency's  ability to implement a
particular type of program within  available  resources.
      The following summarizes the  different control approaches that, in combination,
establish the basis of a comprehensive  air toxics control program.
              Pollution Prevention Initiatives
       Pollution prevention initiatives are those which reduce pollution at the source,
or which use environmentally sound recycling to prohibit releases of emissions,
discharges, and waste.  Pollution prevention is one of EPA Administrator Reilly's
major  environmental  themes,  signaling a change in policy from current EPA programs
whicn  stress treatment and disposal  after pollution has been generated.   In  the future,
EPA's emission regulations will be developed giving full  consideration to prevention
potential.
       Probably the most significant prevention opportunities in air programs involve
energy conservation and evaporation of volatile organic compounds (VOC).   Examples
include promoting the use of low solvent coatings, handling and storage of volatile
fuels, and regulations that stimulate recycling of wastes to combustion units. The
emission regulations EPA has set for over 40 source categories of VOC  emphasize
process modifications, reformulation, prevention measures, and recycling.
                                        1-11

-------
 Key Decisions
Factors to Consider
 PROGRAM GOALS
 AND OBJECTIVES
                                                           Reduce
  PROGRAM
 TYPE/BASIS
   CONTROL
   TECHNOLOGY
   ELEMENT
 AMBIENT IMPACT
    ELEMENT
 RISK ASSESSMENT/
 RISK MANAGEMENT
PROGRAM SCOPE
   REPORTING
 •REQUIREMENTS
   RESOURCE
REQUIREMENTS
    PROGRAM
   COMPLIANCE
Reduce cancer risks
Mitigation of fuuue neks
Control of specific sources/pollutants
Pollution prevention
Comprehensive versus limited
Guidelines versus pi'f^'Mgitrri
rales or standards
Preventanve vs. Curaave

Pollution prevention
New vs. easting sources
Carcinogens vs. noncsranogens
Major vs. minor sources
BACT vs. LAER (new)
RACT vs. BACT (existing)
Specific teccaoiogies vs.
percent reductions

Appropriateness of occupational limus
Inhalation Rfd'i
Risk-based AAL's
Multiple lources/poiluunts
vs. single source/single pollutant
 Accepubie risk level
 Identificitioo of bi£h "^ poflutmts
 Risk urget versus sundnds
 Residual risJa after control

 Specific source categories
 PoUutamx covered
 Point vctxus cxea sources
 Oenriniims size oaoff:
 New vcnus eusaoa source

 New and icnewal penmi applications
 Source mspec&oc

 Operating du
 Agency requests to sources
 SARA Tulc ffi

 PCZIXQl ICV1CW
 Digpcnrion modeling
 Rick ttKumcnt
 Ambient mwriionng
 Control technology cvtluaioa/coKting

       courcci
 New «nd modified cotaces
 Point sources
 Are* sources vs. aanor point sources
 Nootrtditionil sources
 Schedule for complimce

-------
      Pollution prevention is also an important facet of control in a comprehensive
S/L air toxics program.  In conjunction with considering "end-of-pipe" control
technologies, process modifications, reformulation, prevention measures, and recycling
can also be considered.  As discussed below, some S/L agencies (e.g.,  Maryland,
Nebraska, Kentucky) include pollution prevention in  their  definitions of BACT for
toxics.

      1222  Control Technology Element

      Under the control technology element, minimum control technology  or
emissions  reduction requirements are established. These may be applied across-the-
board or to specified  source categories and/or pollutants.  This element is
summarized in Table  1-1.
      Control technology requirements may be in the  form  of pollution prevention
initiatives, specified technologies or control methods, a generic percent reduction
requirement, an  emission limit, specified  operating practices, or a combination thereof.
      The label given to the control technology requirement differs from State-to-
State.  For example, Maryland  requires  that most new sources use best available
control  technology for toxics,  or "T-BACF.  Wisconsin has a "Lowest Hazardous
Emission Rate" (LHER) requirement for sources of known human carcinogens and
"Best Commercially Available Control Technology" (BCACT) requirement for  sources
of suspected  human carcinogens.  New York and Texas require new sources to meet a
BACT requirement.  Irrespective of the name, the most important ingredients of a
control  technology requirement are its stringency and the sources to which  it applies.
Stringency and applicability requirements vary  considerably among S/L programs, and
may vary according to the  toxicity of the pollutant.   Many S/L programs with a
control  technology emphasis apply the most stringent level of control to new sources,
and to the highest priority  pollutants, such  as carcinogens.  Maryland, on the other
hand, applies their top-down T-BACT requirement on  a case-by-case basis  to all new
sources  of toxics, taking into consideration  the toxicity  of the pollutants emitted and
the cost of control, among
                                       1-13

-------
      TABLE 1-1.  SUMMARY OF CONTROL TECHNOLOGY ELEMENT
Main Features:
o     Categorical control technology or emission reduction
      requirements, or
o     Case-by-case BACT requirement with or without a
      minimum  emission reduction level
o     Requirements  typically apply independent of ambient
      impacts or risks
Advantages:
o     Ensures minimum degree of control
o     Achieves equitable emission reductions within
      categories
o     Provides less ambiguous  requirements for affected
      facilities
o     Encourages pollution prevention
o     Is technology -forcing
Decision Points:
example i-anguage:
(1)   Rigor/stringency of control requirement
   _  (e.g.,definition of BACT)
(2)   Relationship  to other program elements (e.g.,
      residual risk  issues)
(3)   Process for making control technology decisions(i.e.,
      generic criteria or case-by-case)
(4)   Applicability  to new and/or existing sources
(5)   Applicability  to pollutant types  (e.g., carcinogens)

Wisconsin
* WI - NR 445.02 (4)

(4) "Best available control technology" means an emission
limit for a hazardous air contaminant based on the
maximum degree of reduction practically achievable as
specified by the  department on an individual case-by-case
basis taking into account, energy, economic and
environmental impacts  and other costs related to the
source.

Kentucky
* KY - KAR 63:022 Section 2(4)
(4) "Best available control technology" means emissions
limitation based  on the maximum degree

              1-14

-------
TABLE 1-1.  SUMMARY OF CONTROL TECHNOLOGY ELEMENT (CONTINUED)
                          of reduction for each pollutant listed in Appendix B to this
                          regulation which would be emitted from any proposed
                          affected facility which the cabinet, on a case-by-case basis,
                          taking into account energy, environmental and economic
                          impacts and other costs, determined to be achievable for
                          such facility through  application  of production processes or
                          available methods, systems, and techniques for control of
                          such pollutant  In no event shall application of best
                          available control technology  result in the emission of an air
                          pollutant which would exceed the emissions allowed by any
                          applicable standard under Title 401, Chapters 57  and 59, or
                          40 CFR pans  60 and 61.

                          Maryland
                          « MD - COMAR  10.18.15.01 B(3)

                          (3) "Best available control technology for toxics"  (T-BACT)
                          means that production technology, emissions control
                          technology, operation and maintenance procedures, other ~
                          measures, or combinations of them that results in the
                          maximum degree of emission reduction that  the
                          Department determines, on a case-by-case basis,  is
                          achievable by an  installation, for  each toxic air pollutant
                          discharged, taking into account the potency and toxicity of
                          each toxic air pollutant discharged as  well as technical and
                          economic feasibility.

                          Nebraska
                          * NE -  Title 129,  Ch. 1 013

                          013 "Best Available Control  Technology" means an emission
                          limitation  or a  design, equipment, work practice,
                          operational standard or combination thereof, which  results
                          in the greatest  degree of reduction of a pollutant, as basis,
                          taking into account energy, public health, environmental
                          and economic  impacts and other  costs.
                                        1-15

-------
other factors.  As in many other States, Maryland applies its control technology
requirement first, independent of the ambient impacts.
      Development of a program including a control technology element involves at
least three major decision points.  These are illustrated in Figure 1-5.   The three
decision points are:

      (1)    the rigor or  stringency of the control requirement,
      (2)    the relationship to other program elements,  (i.e., ambient impacts  and
             residual risk analyses) and
      (3)    the process for making control technology decisions (e.g., whether  to
             establish generic requirements/guidelines or  use a  case-by-case
             approach).

      1.     Rigor/Stringency of Technology Requirement.  The first  decision point
involves the rigor or  stringency of the requirement.  The  level  of stringency varies
among agencies and,  in fact, may  vary from one  source to another.  The level of
stringency does not ^vary so much  in the formal definition  of BACT.  Indeed, most
agencies seem to rely on  a fairly standard definition where BACT determinations are
made on a case-by-case basis and usually consider cost.  Instead, the variation is in
which sources and for what pollutants BACT is required.   It is  also likely that the
stringency of the case-by-case BACT determinations varies between agencies,  but that
possibility has not been investigated.
      Although formal definitions of BACT are  fairly consistent, they may be used
differently in practice.  For example, the BACT guidelines developed by  the Northeast
States for Coordinated Air  Use Management (NESCAUM), .and Maryland's T-BACT
approach  use top-down BACT  where a permit applicant must first identify the most
stringent control possible  (usually  referred to as the lowest achievable emission  rate  or
LAER). and then quantify emissions. At this step, no technically feasible alternative
is ruled out.  The starting assumption for the top-down BACT approach  is that the
most stringent control possible  is BACT.  The burden of proof for applying  a less
stringent control requirement rests with the applicant
                                        1-16

-------
 o    c
^—    D

^2   S
 r-   *^
.s    >^
 O    bfi
      O

 C   'O

 o    c
 V*    >-«
 on   ^*— i
 52    o
     r ->


1-1
o
~*~*
en
1
OX)
q
'3
CQ
!=
1 Decisioni

a
u
S
4>
h*
- Generic icqu
CO
O
en
y
•o
?
i

S
•a
to

- Phasing/prior


8
O
w
S
1 - Available res


to

c.
- Economic im
1
.3
£
_a" —
= ~
- Dcnionslialio
source or pre
by agency
                                                                             c
                                                                             »
 u^


Q
c
oo
•r-
^
O
00
5
c
•— ;
'G
0
G
•JS
i2
5
i
a
01
C
O
•a
§•
c
o
en
1
O
in
i
ng sources
t/i
s «
3 *
O O
50 z
_o
1
u
- Pollution pr
oo
O
o
"o

- Degree of If
forcing
                                                                             0)

-------
      Other S/L agencies set a percent emission reduction as BACT.  For example,
New York has a BACT requirement for all sources of pollutants classified as highly
toxic and  generally considers BACT to result in  99 percent control.  Similarly,
New Jersey  requires  BACT for sources that emit any of the toxic volatile organic
compounds that the  State regulates and considers this to be 99 percent control.  For
toxic particulates, 99 percent control is expected from existing sources and 99.9
percent for new sources.
      Related to the use of BACT are the requirements for reasonably available
control  technology (RACT)  and the lowest achievable emission rate (LAER).  A
RACT requirement is typically less stringent than BACT, while LAER is more
stringent.
      Table 1-2 lists citations regarding control  technology in several State air toxics
regulations.
  A brochure about  EPA's Control Technology Center (CTC),  which can provide
  information on control technology options and on how various S/L agencies have
  determined BACT  and LAER for various sources, can be  obtained by calling the
  CTC at (919)-541-0800.
       2.   Relationship to Other Program Elements.  A second decision point
involves determining the relationship between  the control technology requirement and
other reviews  and requirements included in the program.  Specifically,  linkage to the
ambient impacts element, both procedurally  and conceptually, must be  determined.
Will sources be required to meet an AAL requirement  in addition to the control
technology requirement?  Will sources be analyzed  for their residual risks and become
subject to additional controls or other  changes before approval  if they  do meet the
target  risk level after meeting the control requirement?   Most S/L agencies with a
control technology component to  their program couple it with an ambient impact
component or use risk assessment techniques to evaluate the residual risks posed by a
source after applying controls (e.g., New Jersey,  Maryland, Texas).
                                       1-18

-------
TABLE 1-2.  CITATIONS TO CONTROL TECHNOLOGY REQUIREMENTS
         Control Requirements: Control Technology-Based

               CT - Section 22a-174-29(d)
               KY - 401 DAR 63:021  Section 3(2)
               KY - 401 DAR 63:022  Section 3(2)
               MD - COMAR 10.18.15.05 A,B,C
               MD - COMAR 10.18.15.05 B
               NE - Title 129, Ch.6 007
               NV - 445.719
               NV - 445.7195
               NY - 6 NYCRR 212.2(c)
               NY - 6 NYCRR 212.4(d)
               NY - 6 NYCRR 212.9, Tables  1,2,3,4
               OK - 3.8.4(c)(l)
               OK - 3.8.4(d)
               VA - 120-04-0303
               VA - 120-05-0303
               WI - NR 445.04(3)(a) and (3)(b)
               WI - NR 445.04(4)
               WI - NR 445.04(5)(c)
               WI - NR 445.05(3)(a) and (3)(b)
               WI - NR 445.05.(5)(e)
         Definitions: Best Available Control Technology

               KY - 401 KAR 63:022 Section 2(4)
               MD - COMAR 10.18.15.01 B(3)
               NE - Title 129, Ch.l 013
               WI - NR 445.02 (4)
                                  1-19

-------
o     Relationship Between BACT and Ambient Requirements.  Several
      agencies require sources to apply BACT and then to meet acceptable
      ambient levels as well, while some either (1) allow sources to choose
      between a BACT requirement or an ambient level, or (2) require
      sources to apply BACT if the source cannot meet the agency's
      acceptable ambient level  In the first case, where BACT does not
      achieve ambient impact objectives, options that are technology-forcing,
      include pollution prevention or limiting operation are also considered.

o     BACT Requirement  for New Versus Existing Sources.  A BACT
      requirement for a  source not yet constructed is much different from an
      economic standpoint  than a BACT requirement for an existing source
      that has applied for  a renewal of an operating permit  Accordingly, few
      agencies apply rigorous top-down BACT requirements to existing sources,
      and those that  do  generally combine the technology requirements  with
      acceptable ambient levels and/or risk assessment requirement.

o     BACT Requirements for  Different Toxicity Classes.  Another issue
      relating to the stringency of BACT  requirements is the pollutants  that
      trigger application of BACT. Several agencies require BACT,  or  even
      the more stringent LAER only for emissions of carcinogens.  Some
      agencies require BACT on all new sources, but only  on sources of
      carcinogens for existing sources.

o     Emissions Sources.   At least one State (Wisconsin) makes a  distinction
      between which  emission units within a facility must install BACT  or
      meet LAER.  In  this case, the emission units at  the facility that emit the
      greatest amount of the pollutant must install BACT (or meet LAER in
      the case of known human carcinogens).  If the pound per hour emission
      rate given in the  regulation is not met for the  facility, then additional
      units must install BACT  (or meet LAER) until the emission rate  is met.

o     Residua] Risks.  The issue concerning residual risk is "What  additional
      source requirements  should  apply if an unacceptable  residual risk
      remains after applying BACT or other applicable controls?"  In this case,
      the agency must carefully weigh  post-BACT options.  A likely  concern in
      such  cases will  be the potential economic consequences of various post-
      BACT options.  As  a result, S/L agencies will need to evaluate the cost-
      benefit of reducing risks  to  target levels,  in the face of severe economic
      consequences, a compliance schedule may need to be adopted that
      would achieve risk targets within a  specified period of time in lieu of
      imposing unacceptably high  costs or economic consequences  (i.e.,  plant
      shutdown) in order to achieve immediate compliance.
                                 1-20

-------
3.   Process for Making Control Technology Decisions.  A third decision point
involves how control technology decisions are made.  The question is whether such a
requirement should apply to individual source categories or pollutants, or whether a
case-by-case approach should be used.  A categorical approach may require somewhat
more  time initially.  However,  once established, control technology requirements or
guidelines can facilitate the review of individual source applications.   On the other
hand, a case-by-case approach provides greater flexibility  to handle unique  situations.
However, it rnay involve more  engineering evaluation during the review of permit
applications.

       1.2.2.3   Ambient Impact Element

       Under the ambient impact  element, health-based ambient concentration limits
are used to establish control requirements for individual sources.   This element is
summarized in Table 1-3.
       The main feature of the ambient impact  element is that it relies on ambient
concentration values as a control mechanism and therefore, control is based on
possible health effects  rather than control technology.  The objective in setting an
acceptable ambient level (AAL) is that it will be stringent enough to protect the
public from exposure to concentrations that would  result  in unacceptable health
effects. Under this approach, AALs have been  established as  either screening levels,
standards, or guidelines. This approach is premised on the assumption that
"acceptable" levels of exposure  can be defined for individual toxic  compounds in the
ambient air.
       There are two major decision points in the design of a  program that includes
an ambient impact element.  The decision points are illustrated in Figure 1-6.  They
are related to:
      (1)  The use of occupational  levels
             (a)  which uncertainty  factors to apply, and
             (b)  which averaging times to use.
                                       1-21

-------
             TABLE 1-3.  SUMMARY OF AMBIENT IMPACT ELEMENT
Main Features:
Advantages:
Decision Points:
o     Health-based acceptable ambient levels  (AALs) are
      used as basis for control requirement
o     Occupational levels, inhalation reference doses, or
      risk-based AALs  commonly used
o     AALs may be ambient standards or  screening
      guidelines

o     Source controls based on health considerations
o     Considers characteristics/toxicity of individual
      pollutants
o     Considers ambient impacts of individual sources
o     Can use lexicological  research done  by  others

(1)   Whether to use  occupational levels or
      chemical-specific  values
(2)   What uncertainty factors to apply
(3)   What averaging  times to use
Example Language:
Connecticut
* CT - Section 22a-174-l

"Maximum allowable stack concentration" or "MASC" is the
maximum allowable concentration of a "hazardous air
pollutant" hi the exhaust gas stream of a "source" at the
"discharge point."  (MASC is based on a "Hazard Limiting
Value" which is  derived from an uncertainty factor applied
to an occupational standard.)
                         Kentucky
                         *KY - 401 KAR 63:021 Section 2(3)

                         (3) "Threshold ambient limit (TAL)" means the
                         concentration level in the ambient air of a toxic air
                         pollutant,  calculated pursuant to Appendix B  to this
                         regulation. (Appendix B lists 736 pollutants,  and gives the
                         averaging  time and a  value for each.  The Threshold
                         Ambient Level is calculated by dividing the given value by
                         the hours  of emission of the substance per week from the
                         source.)
                                       1-22

-------
TABLE 1-3.  SUMMARY OF AMBIENT IMPACT ELEMENT (Continued)
                    Maryland
                    «   MD - COMAR 26.11.15.01 B(l)

                    (1) "Acceptable ambient level" (AAL) means a
                    concentration of a toxic air pollutant in the  atmosphere
                    that the Department determines will provide a margin  of
                    safety to protect the public health from toxic,
                    noncarcinogenic effects  that  may be caused by the toxic air
                    polluiant and that is used to evaluate :he air quality
                    impacts of all premises  within a 5 kilometer  (3.1  mile)
                    radius. AALs are listed in Regulation ,13, below.
                    Oklahoma
                    *   OK - 3.8.2 (2)

                    (2)  Maximum Acceptable Ambient Concentration (MAAC)
                    means the maximum allowable twenty-four hour average
                    concentration, in ambient air,  of a toxic air contaminant.
                                 1-23

-------
C/5
       O


I    I
 00   . C
O    ^
D     52
     <
SO
co 5
t^
O
>
1 - Stajidartls

CA
0)
_0>
uniplions
S
Modeling
concentration
•o
3
So
a
ec
>
§
3
"op
CO
P
g
"3,
"3
S
va
<
2r
Defensibil
o.
c
- Economic
                                                'S
.M
•b

S  <  1     c I'
«^o   « c  I S
 c   *>  J. S   5 >
 O   «   ta •—   S *s
                          o.
        g 3  ^ P  w __   S>-±3
        C ••••  ^- ••  r" ^^   ^ ""*

.2 V5
1.1
5 CD
£~
"^^


"S
crt
«
"T
"w
S
1

available
P^
'•5
a
a>

i
»
es resourc
>
i>
vs
§
o
1

w

c
o
C/J
1
CA
o
Js
CB
S
12
'

Cfi
00
c

O

                                                                                                     O
                                                                                                    .1—(


                                                                                                    'o
                                                                                                     §
                                                                                                     So
                                                                                                    E

-------
       (2)  Alternative approaches
             (a)  inhalation reference doses concentrations (Rfcs),
             (b)  risk-based ambient levels, and/or
             (c)  chemical-specific evaluation.

       1.  Occupational Levels.  The first decision point is whether or not to use
occupational levels  as a  basis for  ambient limits.  Acceptable  ambient levels have
been established  by some S/L agencies by  applying an uncertainty factor to  an
occupational standard for a particular pollutant.   This approach has typically been
chosen because of two main advantages.  First, for many  pollutants, occupational
standards are the only health-based standards that are readily available.  Secondly,
resource limitations  have not allowed  chemical-specific studies to be undertaken.
       On the other hand, there are several disadvantages to using occupational  levels
to set ambient limits.  For  example, Threshold Limit Values (TLVs) are occupational
levels established by ACGIH.  They are based on a variety of health effects including
cancer and acute and chronic noncancer effects.   However, the TLV is set to protect
a healthy worker population, and  as such is based on an  8 hour/day, 40 hour/week
exposure (which allows for  recovery between exposures).  The TLV does not take into
account exposures that may be  of a prolonged duration (such as  ambient exposure),
or consider  sensitive populations, such as children or the  elderly.  The  application of
an uncertainty factor to  the TLV  is intended to lower the TLV to a level where
prolonged exposure  and  sensitive populations are  considered.
       Exposure time-adjustment factors are typically applied to convert  an
occupational level from  an  8-hour exposure to a 24-hour  exposure.  A second
uncertainty factor is also applied to account for the  sensitivity of the general
population as compared  to  the  healthy worker assumption used for the occupational
limit,  and to account for uncertainties in the scientific basis of the
standards/guidelines. Different uncertainty factors may also be applied  to different
groups of chemicals in recognition of their  relative toxicity.  For  example, one factor
might be used for known human carcinogens, another for probable human
                                        1-25

-------
carcinogens, a separate factor for chronic systemic toxicants, and another for acute
irritants.
      Averaging times, or the length of time over which a source is allowed to
average its contribution to the ambient concentration, also have a significant impact
on the stringency of the AAL.  For  example, an ambient level of  10 ug/m3 with an
averaging time of 8 hours is  more stringent than an ambient level of 10 ug/m3 with
an annual averaging period.
      Many agencies choose averaging periods based on the health effects of
particular pollutants and, consequently, use more than one averaging period in their
program.  For example, an annual averaging period  is frequently used for carcinogens
and other pollutants with effects associated with long-term, chronic exposure, while a
1-hour averaging time is often used  for pollutants  associated with  health effects
occurring from short-term, acute exposures.  Averaging periods reported in the
STAPPA/ALAPCO survey ranged from 3 minutes to annual.  The most commonly
used averaging times are annual, 24  hours, 8 hours,  and  1 hour.
      The relationship between AALs and averaging times  is important to the overall
stringency of the control requirement  An averaging time says, in effect, that
excursions above the AAL are permissible  provided  that they are  compensated for by
equivalent excursions below the  AAL over the  period of the averaging  time.  In
determining the stringency of the AALs and their  associated averaging  times,  an
agency has to address whether  an AAL may be exceeded for some period of time.  In
the case  of some very  acutely toxic pollutants,  a short averaging time (e.g.,
15 minutes) is necessary to ensure that short-term releases do not go uncontrolled.
      One question frequently asked is how to compare ambient  levels that have the
same uncertainty factor, but  different averaging times. To perform this comparison,
certain emission characteristics  and dispersion assumptions must be made.  In a study
done  by  the Commonwealth  of Virginia,  various safety factors were converted to an
equivalent basis for direct comparison of safety factors and  averaging times.  The
results of this analysis  are summarized in Table 1-4.  As shown in the  table, the
uncertainty factor and  averaging time both affect the stringency of the AAL.  For
example, a TLV/100 level averaged  over an 8-hour  period is more stringent than a
TLV/300 level averaged on an annual basis.
                                       1-26

-------
TABLE 1-4.
EQUIVALENT UNCERTAINTY FACTOR FOR 24-HOUR
         AVERAGING TIME
AVERAGING
  TIME
    UNCERTAINTY
   APPLIED TO TLV
EQUIVALENT UNCERTANITY
   FACTOR FOR 24-HOUR
    AVERAGING TIME  a
8 hour
24 hour
1 hour
24 hour
24 hour
24 hour
annual
8 hour
24 hour
annual
24 hour
30 minutes
8 hour
24 hour
8 hour
24 hour
1 hour
24 hour
annual
24 hour
annual
24 hour
annual
10
10
40
40
30
42
42
50
50
50
60
100
100
100
200
200
300
300
300
420
420
1000
1000
20
10
200
40
30
42
10
100
50
10
60
1000
200
100
400
200
1500
300
60
420
84
1000
200
  Based on calculations by the Commonwealth of Virginia, under typical conditions and non-varying
emission rates:

The highest 1-hour concentration of a pollutant = 25 X ANNUAL MEAN
                                    5 X 24-HR MEAN
                                    2JX 8-HR MEAN

For example, if the annual mean concentration is 10, the 24-hour mean would be 50, the 8-hour
concentration would be 100, and the highest 1-hour concentration would be 250.
                               1-27

-------
      2. Alternative Approaches.  There are alternatives to using the occupational
level/uncertainty factor approach. These include using EPA's reference concentrations
(RfCs),  developing risk-based AALs for carcinogens using EPA's unit risk factors, or
evaluating health effects on a chemical-specific basis and basing the AAL on that
evaluation.
      RfCs, developed by EPA,  are based on a review of health effects data and
represent an annual average concentration below which  a long-term exposure is unlikely
to result in adverse health effects.  Consequently, they are an alternative to using
occupational standards as screening levels for ambient impacts. As of June 30,  1990,
RfCs have been developed and verified for 19 chemicals.  Many more (primarily for
chemicals listed in Clean Air Act amendment legislation) are currently under review by
EPA's Reference Concentration Verification  Work Group. The pollutants for  which
RfCs have been developed are listed in Table 1-5.
  For more information on inhalation RfCs, call the Air Risk Information Support
  Center (Air RISC) hotline at (919) 541-0888.
      A number of S/L agencies (e.g., Connecticut, Rhode Island) have developed
AALs for carcinogens using unit risk factors developed by EPA's Human Health
Assessment Group (formerly the Carcinogen Assessment Group). To convert a unit risk
factor to an AAL, the agency first needs to determine a target risk level that will be the
basis of evaluation.  The most commonly used measure  of acceptable risk is the
maximum individual risk (MIR) of the most exposed individual (MEI). Risk to the MEI
means the risk of contracting cancer associated with the highest ambient concentration
to which anyone would be continuously exposed over a 70-year lifetime.  According to
the 1989 STAPPA/ALAPCO survey of S/L agencies, most agencies that use risk
assessment have an acceptable risk level of either one in one million (i x 10"6), or one
in one hundred thousand (1 x 10'5). An example calculation showing how to convert a
unit risk factor into an AAL is shown hi Table 1-6.  Pollutants for which  inhalation unit
risk factors have been developed are listed in Table 1-7.
                                       1-28

-------
 TABLE 1-5.    LIST OF POLLUTANTS FOR WHICH INHALATION
REFERENCE CONCENTRATIONS (RfCs) HAVE BEEN DEVELOPED
                    (AS OF JUNE 30, 1990)
           CHEMICALS WITH VERIFIED RfCs

           Acetaldehyde
           Acrolein
           Bromomethane
           Carbon disulfide
           Cumene
           1,4-Dichiorobenzene
           Dichloromethane (methyiene chloride)
           Dimethylamine
           Epichlorohydrin
           Hydrogen chloride
           Hydrogen suifide
           Hexane
           Mercury (inorganic)
           Nitrobenzene
           Propylene glycol monoethyl ether
           Tetrahydroniran
           Toluene
           Xylene
                             1-29

-------
TABLE 1-6. EXAMPLE CALCULATION OF AAL FROM UNIT RISK FACTOR
1.     URF =  (4.8 x 10"3) = the probability of dying of cancer following
      70-years of continuous exposure to 1 ug/nf  of a particular chemical in
      the ambient air.

2.     State's ARL = 1 x 1(T5

3.     URF x AAL =  ARL

4.     AAL = ARL
              URF
5.     AAL = 1 x 10 "5
             4.8 x 10
6.     AAL = 0.00208 ug/m3
Where:     URF = unit risk factor
           ARL = acceptable risk level
           AAL = acceptable ambient level
                                    1-30

-------
TABLE 1-7.  POLLUTANTS WITH EPA INHALATION UNIT RISK FACTORS
          Acetaldehyde
          Arsenic
          Benzene
          Beryllium
          Cadmium
          Chloroform
          Coke oven emissions
          1,1-Dichloroethylene
          Ethylene dibromide
          Formaldehyde
          Hexachlorobenzene
          Nickel
                - Refinery dust
                - Subsulfide
          Tetrachloroethlyene
          Vinyl chloride
Acrylonitrile
Asbestos
Benzo(a)pyrene
1,3-butadiene
Carbon Tetrachloride
Chromium (VI)
1,2-Dichloroethane
Epichlorohydrin
Ethylene oxide
Gasoline (marketing)
Methylene chloride
Propylene oxide
Styrene
2,3,7,8-TCDD
Trichloroethylene
                                   1-31

-------
  For  more  information  on unit risk factors, call  the Air RISC hotline at (919)
  541-0888.
      Evaluating health effects on a chemical-specific basis and deriving AALs from
the evaluation is  a resource-intensive effort  The  advantage of this  type of study is
that the AALs  that are derived are health-based numbers that can consider  cancer
and noncancer  effects and  are protective of public health (including sensitive
populations). The Massachusetts Department of Environmental Quality Engineering
has completed  such a study (a 5-year effort) and compiled a document entitled The
Chemical  Health  Effects Assessment Methodology and the Method to Derive
Acceptable Ambient Limits (CHEM and AAL).  The  document describes  the
methodology used to review the scientific data for four categories of health  effects for
each chemical.  The health effects review results in  a "hazard score" being assigned to
each chemical for each of  the four health endpoints.  The hazard scores are used to
adjust the occupational level downward to account for factors that include chronic
exposure,  sensitive populations, and noncancer effects.  Because the hazard scores are
different for  each chemical, the occupational levels are "customized" to form the basis
for the AAL,
      As of this  writing, Massachusetts has assigned AALs  to approximately 105
chemicals and expects  their chemical-by-chemical evaluation to be an ongoing  process.
      Another issue related  to the stringency of any ambient-based control
requirement  is  whether the source must take  into account background concentrations
of the pollutant in question,  or if the source must simply be concerned  with its own
contribution  to the ambient concentration.   From the  standpoint of  public health
protection, background  concentrations are very important.  From a source's  point of
view, however,  if background concentrations are taken into  account, the source is
forced to  use tighter control measures when the level of the pollutant in the ambient
air may not be attributable entirely to the source. Probably because of the difficulty
involved in measuring  background levels, very few agencies consider background
concentrations  in evaluating whether or not acceptable ambient levels are  being met.
                                        1-32

-------
Because  of the impact that the consideration of background concentrations can have
on control decisions, this may be a factor to include in a "later generation" of a
control program.
      Table  1-8 provides citations to S/L regulation containing ambient impact

requirements.


      1.2.2.3  Risk Assessment/Risk Management


      1.  Overview.  Risk assessment  is the process of determining the potential

adverse health  effects of human exposure to environmental hazards.  Risk assessment

includes  four steps:
      (1)  Hazard  identification.  A qualitative risk assessment, hazard identification
      attempts to answer the question: Does exposure to a specified hazard cause
      adverse health effects (such as cancer) in humans?  A qualitative determination
      of risk may entail some interpretation, such as whether animal carcinogens are
      presumed to be human carcinogens, or whether benign tumors in animals
      represent potential human carcinogenicity.

      (2)  Dose-response assessment.  This is  the process of characterizing the
      relationship between the dose of a toxicant received, and the incidence  of
      adverse health effects in exposed populations.  This step  usually requires
      extrapolation of data from animal studies to humans and from high doses to
      low doses.

      (3)  Exposure assessment.  This is the process of measuring or estimating the
      intensity, duration, and frequency  of human exposure to a toxicant.

      (4)  Risk characterization.  Tnis is the final step of risk assessment and is the
      quantitative process  of estimating  the magnitude of the public health problem
      that results from the hazard.  It involves the combination of the hazard
      identification, dose-response assessment,  and the exposure assessment.  Thus,
      the final expressions  of health risks are  derived in this step.


      Risk management is  the decision-making process in which an action is taken or

a policy  developed  once a  risk has been determined to exist.  It integrates the risk
assessment with technical, political, social, and  economic issues.  Risk management is

also  a means of setting priorities among possible actions for a toxicant or  source.

Risk management must take into account the uncertainties associated with various

assumptions and judgments  made in  each step  of the risk assessment process.

                                       1-33

-------
    TABLE 1-8.     CITATIONS TO AMBIENT IMPACT REQUIREMENTS
Control Requirements, Ambient Concentrations-Based

      CT - Section 22a-174-29(b)(3) and (b)(4)
      KY - 401  KAR 63:021  Section 3 and Appendix B
      KY - 401  KAR 63:022  Section 3 and Appendix B
      MD - COMAR 10.18.15.06
      MD - COMAR 10.18.15.07 B(2)
      MD - COMAR 10.18.15.07 C(2)
      NV - 445:7195
      NV - 445:720
      OK - 3.8.4(b),(c),(d),(e)
      RI - 22.5.4
      VA -  120-04-0304
      VA -  120-05-0304
      WI - NR  445.04
      WI - NR  445.05 (l)(a),(b);(2)(a),(b)
      WI - NR  445.05 (4)(a),(b)

Control Requirements: Ambient Concentrations - Setting
Acceptable Concentrations

      CT - Section 22a-174-29(a)(2) and (a)(3)
      CT - Section 22a-174-29(c)
      KY - 401  KAR 63:021  Appendix A
      KY - 401  KAR 63:022  Appendix A
      MD - COMAR 10.18.15.08 A
      MD - COMAR 10.18.15.09 D(l),(2)
      NV - 445.718
      NV - 445.7185
      OK - 3.8.4(f)
      VA -  120-04-0304
      VA -  120-05-0304
      WI - 445.04 (l)(a) and (l)(b)
      WI - 445.04 (2)(a) and (2)(b)
      WI - 445.05 (l)(a) and (l)(b)
      WI - 445.05 (2)(a) and (2)(b)
      WI - 445.05 (4)(a) and (4)(b)
                                1-34

-------
      There are no fundamental differences between cancer and noncancer risk
assessment/risk management issues.  The major difference between the two is that
unit risk factors are available for carcinogens, thus giving the S/L agency a method to
derive a single number at the acceptable risk level.  This makes cancer effects easier
to deal  with  compared to noncancer effects. The development of inhalation Rfcs
(discussed in Section 1.2.2.2), will  make noncancer risk assessment much more
accessible to S/L agencies.
      The main  advantage  of using risk assessment/risk management is that the end
result is an  indication of relative health risks and a uniform threshold of protection
for the  public.  Risk assessment/risk management also allows the S/L agency to take
into account  variations in the economics associated with individual sources in  setting
control  requirements.
      The main  disadvantage of using  risk assessment/risk management is the
difficulty in  determining an  "acceptable" risk level.  This difficulty is  due to the
uncertainties associated with the many  assumptions that must be made regarding
population exposure, dose-response, hazard identification, etc.  The lexicological
resources required to evaluate individual facilities is  another drawback.  As described
in Section 1.2.2.2, according to the 1989 STAPPA/ALAPCO survey of S/L agencies,
most  agencies that use risk  assessment/risk management have an acceptable risk level
of either one in one million (1 x 10 "*), or one in  one hundred thousand  (1 x 10"5).
      A decision closely associated with the determination of the  target or acceptable
risk level is  whether the risk level  should be a screening tool or a regulatory
requirement.  Several S/L agencies use nsk targets as a screening value for triggering
a more  detailed evaluation  of a source. Others use  risk levels as  an "action"  level.   If
the risk level is used as an  "action" level (i.e., plant  shutdowns  are possible), an
appeal process will  probably be necessary to allow the source the chance  to, for
example, provide better modeling  data  or initiate actions to resolve the problem.
        An issue here is whether the burden to undertake risk assessments should be
placed primarily on  industry or whether the agency will conduct source-specific risk
assessments.  The State of Maryland requires the source to perform  the risk
assessment, an approach that Maryland feels saves  the agency significant resources.
                                        1-35

-------
      Risk assessment/risk management comes into play with both control technology
requirements and ambient impact requirements.  For example, residual risk after the
application of controls is an issue addressed by risk assessment/risk management, as is
the use of inhalation Rfcs or unit risk factors as a basis for setting AALs.
      2.     Risk Assessment/Risk Management Guidance.  Risk assessment
guidelines relating to environmental carcinogens, mutagens, developmental toxicants,
chemical mixtures and estimating exposure were issued by EPA on
September 24, 1986  (51 FR 33992).  The guidelines were developed to promote high
technical quality and Agency-wide consistency in the risk assessment process.
      The guidelines were developed partly in  response to a 1983 National Academy
of Sciences publication entitled "Risk Assessment in the Federal Government:
Managing the Process," which recommended that Federal regulatory agencies establish
risk assessment guidelines.  An EPA task force, convened by then EPA Administrator
William D. Ruckelshaus to study ways to improve the scientific foundation for Agency
regulatory decisions, accepted the recommendation, and work on the guidelines began
early in 1984. As announced on August 26, 1988,  some of the  guidelines are
currently being revised (53 FR  32656). Additional guidelines are also under
development for various topics, including noncancer effects, reproductive  toxicity, and
exposure-related measurements.
      A second source  of information is the NATICH report Qualitative  and
Quantitative  Carcinogenic Risk Assessment (EPA 450/5-87-003), which describes
EPA's  risk assessment procedures for carcinogens.  The Air Risk Information Support
Center (Air RISC) is  also developing a guidance document on risk assessment that
will include noncancer risk assessment.
      Another important source of assistance for risk assessment/risk management is
EPA's  Integrated Risk Information System (IRIS).  IRIS provides information  on how
chemicals affect human health and is a primary source of risk assessment information
on chemicals of environmental  concern.  IRIS  makes chemical-specific risk information
readily available to  those who must perform risk assessments and also increases
consistency in risk management decisions.  The health  assessment information
contained in IRIS has been reviewed and  agreed upon by two interdisciplinary review
                                       1-36

-------
groups of EPA scientists who have extensive experience in risk assessment.  This
Agency-wide agreement on risk information is one of the most valuable  aspects of
IRIS.
      The IRIS is an online database that may be accessed via EPA's electronic mail
(E-mail) system (available through DIALCOM, Inc.).  The information in IRIS is
divided  into two major components:  (1) chemical-specific toxicity and regulatory
information, and (2) documentation providing instruction and explanation in support of
the system and the chemical files.
      Each chemical file contains:
      o    verified reference doses  based on noncancer health  effects  data
            associated with chronic exposures,
      o    verified risk estimates for carcinogenicity,
      o    drinking water health  advisories,
      o    summaries of selected EPA regulations,  and
      o    supplementary data  (e.g., acute toxicity information  and  physical-chemical
            properties)
The database can be searched by chemical name or Chemical Abstracts  Service (CAS)
number.  Over 260 chemicals are represented.
  For  more information on IRIS, call  IRIS User Support  at (513) 569-7254 (EPA
  Cincinnati).
      A fourth source of assistance is EPA's Air RISC.  One of Air RISC's goals is
to serve as a focal point for disseminating information on health, exposure, and risk
assessments for toxic pollutants.
                                       1-37

-------
  For a brochure describing Air RISC activities call the Air RISC hotline at (919)
  541-0888.
      A more complete list of primary sources of EPA information on risk
assessment is provided in Table 1-9.
                                      1-38

-------
Z
w

CO
CO
ffl
oo
CO
CO
5
8
a
u
§
O
CO
w
Contact
 CO
'S

<

 o
*T"5
 CB
nfo
                        «.2
     £C 3
     O 0) «J
     •3. O y
     •3 g-'S
     I So

     •2 if'3
     S

     o
Infor
relative to heal
assessment of
 va
 t/5
6
                   ce
                   £

                   I

                   a
                   p
                   o
                           «?
                        £ ea
                                on
                          .     S3
                        co 3W a,

     CO J< '= ^

     PH P*!  F Q
               .
             £
          £ CO
541-08
S)629-08
19
o
FT
9
                  O  C >,
                  Q..O oa
                  C^*-*2 O
t3 u ^2
 S ^ °
 C3 Q (jj

 §Ii
Engineering guida
on air emissions c
and pollution cont
                     o
                     c
                                                       i
                                               con
                                                 »-  o
                            13 li? i
                             05 T; f-« H
3 05 ra •—
S O.— >-
C "O O Si
Agency consensus sui
mation on hazard and
assessment; reference
EPA contacts for furtl
                            u.    e

Cally
Air T
ADP
Cleari
(919
                                                                 0-2
Loca
cies;
i
ona
el
g
&
e
re
n
e
g
State
air a
EPA
pers
                 U
                                      - o
                                       2"2-
                                                   4J
                                                      o,
                                                     §
                                                  "8
                                  («
                                *-* *'—
Federal, Stale
& Local air
agencies

-------
          SECTION 2.0
MAKING CRITICAL DESIGN CHOICES

-------
                   2.0  MAKING CRITICAL DESIGN CHOICES
                             Highlights of This Section

             Defining the  scope of an air toxics program is a critical step
             in program development which is  closely related to program
             stringency and other design and implementation issues.

             Program scope includes pollutant selection,  source category
             selection, existing source coverage, size and emission cutoffs.
             program phasing, and resource allocation issues.

             A variety of  resources are available to  assist S/L agencies with
             program scope decisions.
      Defining the scope of a program is a critical step in  the program development

process.  Program scope decisions are critical for several reasons,  including the

following:


      o      Program scope decisions affect the ability of an agencv to address air
             toxics goals.  For a program to be successful, the number of pollutants
             and sources selected for regulation should be realistic and in keeping
             with the program's goals.  Defining program scope too narrowly,
             however, may result  in less public health protection than desired.

      o      Program scope decisions influence whether  program goal." can be
             achieved within available resources. An overly expansive program, for
             example, may subject more sources to regulation than an agency  can
             handle within available  resources.   If staff and other resources  are
             overloaded, the chance  of achieving the  program's goals is likely  to
             diminish.

      o      Program scope decisions determine the size and nature of the regulated
             community. Program scope decisions can greatly influence the politics of
             program approval by defining the  nature and extent  of the potential
             opposition  to a program.

      There are  at least six main topics to  consider in defining the scope of

an air toxics program.  They are:

                                        2-1

-------
      (1)    selection  of pollutants,
      (2)    selection  of source categories,
      (3)    regulation of existing sources,
      (4)    selection  of size and emission  cutoffs/exemptions,
      (5)    timing/phasing  of program coverage, and
      (6)    resources required.

The  following subsections discuss each of these program scope decision points and the
approaches used by S/L agencies with operating  programs.

2.1   SELECTING  POLLUTANTS OF CONCERN

      2.1.1   General Approaches

      Two general approaches have been taken  by  S/L agencies in selecting toxic air
pollutants for review or regulation.  Under  one approach, a specific list of toxic air
pollutants is  prepared.  In the second approach,  the agency addresses any substance  of
concern.  Each of these approaches has certain advantages  and disadvantages.
      One point to  keep in  mind when considering these two approaches is that,
under either  approach, once  the total number  of pollutants regulated  reaches a certain
level (e.g., 100-300), the number of additional  sources covered may decline. Tnat is,
sources  typically emit mixtures of pollutants.  As a result, regulation of a specific
compound often results in simultaneous control of other compounds emitted by a
source.   Consequently, listing hundreds  of pollutants may not necessarily  expand the
universe of regulated sources.

      1.     Specific List of Pollutants.  One advantage of defining a specific list  of
toxic air pollutants is that an agency may scope  or limit their air toxics program
efforts and focus their resources on the selected pollutants.  The amount of research
                                        2-2

-------
required for identifying pollutant health effects associated with exposure to a specific
pollutant and  emissions estimates would likely be less than for the open-ended
approach.  Also, by defining a specific list of pollutants, the agency may, in effect, let
industry and the public know what to expect, and  thereby  create a more cooperative
atmosphere.  Also, as mentioned above, control of one toxic pollutant is likely to
control several other types simultaneously without  their specific regulation.
      A disadvantage of defining a specific list of pollutants is that it may be  difficult
to add new pollutants to  the list  Depending  on legal and regulatory constraints,  the
additions may have to be approved by an administrative board or by  the legislature.
Some pollutants may inadvertently be left off the list, even though they represent a
significant potential for exposure and/or risk to the  population.  Conversely, the
longer the  list becomes,  the  more the resources that must  be devoted  to data
collection and management.  Finally, there may be public  controversy over which
pollutants are selected and the rationale for their  selection.
      Regardless of the number of pollutants covered, it is important to incorporate
some degree of flexibility into the procedure for listing pollutants.  Such flexibility
includes the ability to add or delete pollutants  as new information becomes available,
and  to prioritize in some manner regulatory efforts among various categories of
pollutants.   Generic language can be included in the regulation (or policy) that
requires public notice, but not legislative action, before adding to or deleting from the
list.  Some agencies note that such flexible procedures make the list  of pollutants a
"living list" that  may change  considerably over time.   For example, a flexible list would
make it easier to incorporate the approximately 200 pollutants contemplated for
regulation under proposed Clean Air Act  amendments.  Wisconsin is  an  example  of a
State with a flexible list  of regulated pollutants.
      Generally .speaking, S/Ls have listed more than 100 pollutants  in their
programs.  There is widespread consistency in listing carcinogens; the choice and
number of noncarcinogens vary widely.

      2.     Open-ended Approach. Because of the disadvantages mentioned  above,
some agencies have  found it overly constraining to be limited to a specific list  of
                                        2-3

-------
pollutants.  An advantage offered by an open-ended approach is that the agency can
regulate emissions of any toxic substance without undertaking a formal pollutant
selection process.  Texas is  an example of a State  that has adopted this  approach.
Texas regulates all noncriteria pollutants  under their air toxics program.
      One disadvantage of the  open-ended approach is that the number of pollutants
addressed may become very large, such that research into health  effects  and emission
rates for such a large  number of pollutants may have to be rather cursory.  Data
collection and management may become  resource-intensive.  To address  this concern,
Texas uses screening values based on available health effects information and on  their
own staff evaluations when  reviewing sources.  Also, industry may not know what is
expected and may be  concerned that an  open-ended approach does not define a
specific control plan.
      A combination  of these two approaches has also been used.  Typically, a
specific, yet flexible list is used  for existing sources, and an open-ended approach  for
new sources.

      2.12   Steps for Selecting  Pollutants.

      To select pollutants,  S/L agencies have proceeded through one or more  of the
following  steps:

       1.     Review chemical lists,  such  as Superfund Amendments Reauthorization
             Act Title IQ  (SARA Title  HI), American Conference  of Governmental
             Industrial Hygienists (ACGIH), or  lists from States with operating
             programs,
      2.     Evaluate of the  health effects of individual chemicals,
      3.     Evaluate of the  expected occurrence  of emissions of individual chemicals.
      4.     Seek  internal  assistance from divisions within the agency, and
      5.     Seek  external  assistance from such sources as Air RISC, IRIS, special
             task forces or work  groups.
                                        2-4

-------
The methods are discussed independently of one another in this section.  However, it
should be emphasized that pollutant selection  actually involves completion of some or
preferably each of these steps.  Table 2-1 summarizes the basis for selecting pollutants
by nine S/L agencies.

      Step  1:  Chemical Lists.  There are many pollutant Lists that can be reviewed.
These include the list of 400+ hazardous chemicals listed under SARA Title HI
(listed  in the Toxic Chemical Release Inventory reporting Form R - EPA-560/4-88-
005), and EPA's list of pollutants with unit risk factors (see Table 1-7), or for which
EPA has developed inhalation reference  concentrations (see Table 1-5).  Table 2-2
summarizes  several groups that publish lists of carcinogens.  In  addition,  as shown in
Table 2-1. S/Ls also  incorporate pollutants from lists  published  by ACGIH, OSHA,
and  NIOSH.  The Data Base Report on State. Local, and EPA Air Toxics Activities
(EPA-450/3-89-29)  published annually by the National Air Toxics Information
Clearinghouse (NATTCH), also  has  pollutant lists for  the many  States that set ambient
limits. In many  cases, S/L agencies  have used one or more of these lists in developing
their own list  of toxic air pollutants.  For example, North Carolina used  an initial
screening process in which they identified pollutants that  appeared on the lists of at
least five other  States.  This list, coupled with a review of North  Carolina's emissions
inventory data, resulted in an initial list of 67  pollutants of concern.  About 30 more
chemicals were  subsequently added  to North Carolina's list.
      While lists of pollutants may  be appropriate for the purpose  for which they
were prepared,  they should not necessarily be  adopted m toto by another agency
because the types of  industries and  chemicals  used differ from area to  area.  Other
States' lists  can, however, serve as a good starting point for identifying pollutants of
initial concern.  The  derivation of any other State's list under consideration should be
checked by  contacting the particular State to help determine the similarity between
industry and emission characteristics.  Discussions with the appropriate  personnel in a
State or  local  agency may also shed light on the rationale for pollutant selection used
by that agency.  In other words, any list must  be "customized" to  meet a State or
local agency's  own needs.
                                        2-5

-------






z
o
JJ
i
H
w

^
z
£
w.
•*
•?
/«*
a:
i
WJ
1
MJ
w^
<

^"
«*
•*s
a
H



























M
O

1


l

il
EC*
EC
-1
;z
•§.S
- <
.3 a
3
s o
8. "2-
III

o

r
1
;

,






























5 §
jl
£^





•
e
<_ O
0 'i
(0
V P
hi c
II

V)
e
CQ
•sl
«. o
% Q.



r
£3



ul
— -3

1 1
73 « u.
o i, c
.fs s
Carcinogens
Suspect carcinogens
Substances wilh eslabi
maximum allowable ex
exposure limits
NY's 16 high loxicity
o o o o

p."
Z
^
z_i
III
U a; u
-. ^
§ 1 J
.5^ S-
T3 g M
= ^ 2
„ 1
•= C"S 8 S
2ci'lll -
xg£-ffg^- 3
h8iltfs 1
ii;Ni« J
35ll2*g f
^Is^li i
— •=• i ^% ._ v> .s («
SO^TSruiec -a
•BO^S = -S =
^ti'^Sl |,
ao-ii .•= C U > .= ^ &
XI5.S2SS Jf -

0 0






« aa s
"2 "r i
5 !«$ r
-i s^J -I H
iT-s-sl 1 |l
51|=1 5 *~
*\ 1i f I?
•S g c '5 Z -a s s
sfiil s^li
og-^eS oz-6«
o, S = t. 5, &..£<<
o o o o o o


Wl
.2 AC
i c c
ni •=•
ail g u
s -1 u
5± i
5 B> r
Z ^ c ^ c
tf *lil
a; — — 9 —
< -"Izl
— — a- S ft.
0 COO



— . ^o



IS
f 1
1 2
k C
u O
z 2
g
.i!
1
£••
aj) hoc adviso
recommended
guidelines.

^i5
Is3-
S i-
>*• S'S

0









Pollutant einilled in
Philadelphia
Human carcinogen
Suspected carcinogen
o o o





p.
Z
-
r".
d



s;


i
{*
5 ^
i«i
e «
c ™
ot£
2-6

-------




2
0
i
3
V2
'•^
^
*"i
*^
^
£
«>
sc
g
(g
i
g
i
?
a
—^
B~

S
~i !
2
1













U)
a
: u
e
e
-9


1'
_






5.1
•3*
eo 7>
Oc/5






a
u* O
O —
<0
§1
II


^5
c
CD
•81
*£

t-
1
ll
I c
s!
"2 1
15
-3
SB'S
1J-2
Z'S s

o
^
*1
,V J=
o •"
S c
-. * V«
•— ** 1
C ^"* O
i = *,
r; -.s ^*^ C
h% all
ri *~*
!l t ti
«3|? t2 ^ge
000



(0
v e
^ _, ~ tS
-2 S •» 5 « «5 «
1 i? &>7 3i |-i
"^^.i 25US" - *
J5 ^^ « 2 =J=S"_g ^ -
— — o uisS^S — u
]j '•§ 5 g iS SJS<"°o ^a
^ s^_ c u •-= £ ^-o ^ =
So BO »> _ sei- •— c 2
SS.£S^ .S.2g3« -5 f
5 s S — S Sc-sffl2 'C3
scusS S5s=o "2
•T'crC"1-^ C — •— P — "S'6-
il^g.0 l<2elj ii
o o o o o o c


(0
'C 2
i! c
*i- A
is
3? —  > a.














M
I
Jr £
™" "S
— • 4>
« T«
'5 c~
w w 0*0
O 3 ' C C *«
ii 2 f<
<3 1 1 11 =
s .S "s O
§ c E $ '5
» •c S
•= S < "S -C
si s So
« "~ •* ^ ^
^ P}l
5 * S e-s S s
u v ^H a c
>S U 'S A U
ll-af^l
<|8|^||
1 0 g .0 1 | <
§ = y ^ .2 ~ -o
1 1 1 § i s 1
1 g -2 -1 § .2 S
£!i2£2Jr
«9 M M M &0 M »
e e S e c e e
a <( S « so a m
iiigiii
j>
ua =P
o * <• z ...
l^iiill
2-1

-------
     TABLE 2-2.    LISTS OF CARCINOGENIC COMPOUNDS
      Group
      Special Features
National Toxicology Program
   (NtP)
Also publishes results of Mutagenicity
and Teratogenicity testing.
International Agency for
Research on Cancer
   (IARC)
Uses weight of evidence classification
      scheme.
Occupational Safety and Health
   Administration (OSHA)
Publishes list of carcinogenic
compounds for hazard communication
and regulatory purposes.
American Conference of Govern-
mental Industrial Hygienists
  -(ACGIH)
Has proposed system to indicate
carcinogenic potential and potency.
Environmental Protection Agency
   (EPA)
Maintains list of unit risk factors
for certain carcinogens  and has
prepared a weight of evidence system.
Agency for Toxic Substances
and Disease Registry
   (ATSDR)
Prepares risk assessments of hazardous
compounds at Superfund sites.
                                2-8

-------
      Step 2:  Expected Occurrence.  A second step in selecting toxic air pollutants
is to evaluate the likelihood of a substance's use and emissions within a State or
region.  This step identifies which pollutants are actually emitted or anticipated to be
emitted, but does not provide an indication of the pollutant's toxicity.
      Several data sources are available to determine which pollutants are
used/emitted.  These include literature reviews, questionnaires and  SARA Title HI
reports, monitoring data, source testing/stack sampling, and  the criteria pollutant
inventory.

      Literature reviews.
      Literature reviews are less expensive than some other sources of data.  An
excellent starting point for a literature review is the Bibliography of Selected Reports
and Federal Register Notices Related to Air Toxics, published annually by NATICH.
The Bibliography is  a three-volume set (plus an index) containing over  1850
bibliographic citations to reports and Federal Register notices related to air toxics
(EPA-450/5-87-005,  EPA-450/5-88-005, EPA-450/3-89-25).  To facilitate the use of the
Bibliography, citations are  indexed by document type, pollutant  name or class,
Chemical Abstract Service  (CAS) number, source category and  Standard Industrial
Classification (SIC) code, and sponsoring agency.  (The  index is EPA report number
EPA-450/3-89-25a.)  The information contained in the bibliography is also available
through the NATICH on-line database.
  For information on accessing the NATICH database online,  call  the Clearinghouse
  Staff at (919) 541-0850.
      Another important source of information on emissions sources for individual
pollutants and source categories is the Locating and Estimating Emissions series of
documents published by  EPA  This series of documents, known as the "L&E
documents", compiles available information on sources and emissions of various air
                                        2-9

-------
toxics.  The series includes documents on almost two  dozen of the highest priority
pollutants,  such  as acrylonitrile, chloroform, formaldehyde, chromium, ethylene oxide,
and others. The documents  contain information on emissions from production of each
chemical, emissions from its  major uses, emissions  from miscellaneous sources, and
source test procedures.
  For information  about the  L&E documents,  call (919) 541-5522, and request the
  brochure, 'Tools  for  Estimating Emissions of Air Toxics".
        Questionnaires to sources  and SARA Title III reports.
       Information from questionnaires or surveys on a facility's production, use and
storage of chemicals can indicate  the types of pollutants  expected to be emitted.
Surveys might be sent to all sources  or specific surveys might be developed for
selected source categories.  The EPA documents Compiling Air Toxics Emission
Inventories (EPA-450/4-86-010, currently  being updated)- and Compilation of Air
Toxics Emission Inventory Questionnaires (EPA-450/4-88-008) provide useful
information on identifying potential sources of air toxics  and designing questionnaires
for industry surveys.
       Toxic  chemical  data from SARA Title HI Sections 312  and 313 inventories also
can be used  for this purpose.  By serving as  a screening tool, a
proauction/use/storage evaluation may besi fit the needs of an agency just beginning
to develop an air toxics program.   The results of the production/use/storage
evaluation could  focus the efforts  of  follow-up, detailed data gathering. The use of
SARA Title  III data or questionnaires  is  relatively less expensive than other data
gathering methods.  Such  data could  also serve as the  cornerstone for ac air toxics
emission inventory.
       SARA Title III Section 313 requires EPA to establish a toxic release inventory
and to make the toxic  emissions data available to the public through a national Toxic
Chemical Release Inventory (TRI) data base. The  TRI  data base is accessible by
                                        2-10

-------
Chemical Release Inventory (TRI) data base.  The TRI data base is accessible by
S/L agencies as well as the general public.  See subsection 4.33.1 for further
discussion of SARA Title lU and how to access the TRI database.

       Monitoring data.
       A third way to determine which pollutants are emitted is to conduct
monitoring.  Ambient monitoring can provide  definitive indications of the presence of
a toxic pollutant  in ambient air.  However, proven ambient monitoring and  analytical
methods have not yet been developed  for all toxic pollutants.  Naturally  occurring and
secondary pollutants also need to be considered.   Monitoring is very expensive and
may not be a necessary step in pollutant selection in the beginning stages of an air
toxics program when less expensive methods can be used.  Existing monitoring
programs for criteria pollutants can also be used for air toxics identification.  This
topic is discussed in more detail in Section 4.3.2.2.

       Source  testing/stack  sampling.
       Source  testing/stack  sampling can determine which pollutants are actually
emitted.   Modeling of the measured emission  rates can -be used to estimate ambient
concentrations.  However, source testing is very expensive and is usually  not an option
that is available to most S/L agencies.

       Criteria pollutant inventories.
       Criteria pollutant inventories offer another source of data.  This topic is
discussed further in  Section 4.3.2.2.

       Step 3:  Health Effects Evaluation.  Another step in selecting toxic air
pollutants is to examine the known or expected health effects resulting from exposure
to the  pollutant.   The primary advantage of this method is  that the pollutants that
have the greatest potential  to cause adverse health effects are identified.   However,
the major disadvantage to this method is that  it does not identify pollutants actually
emitted in the area, and hence, resources  could be devoted to gathering  information
                                        2-11

-------
on pollutants that are not truly a problem in the State or region. Therefore it is
recommended that the expected occurrence of pollutants be evaluated prior to a
health effects evaluation.
      The major issue when selecting pollutants on the basis of health effects is how
to define measures  of toxicity.  In other words,  what toxic effects will warrant listing
and how much experimental evidence of those  effects is necessary?  Many  agencies
agree that  all carcinogens should be listed.  As  shown in Table 2-2, lists of known
and suspected carcinogens are published by various groups.
      The National Institute for Occupational Safety and Health (NIOSH) maintains
the Registry of Toxic Effects of Chemical Substances (RTECS) database.  RTECS is
an online interactive version of  the NIOSH publication of the  same name.   It
provides data for 90,000+ potentially toxic chemicals,  including toxiciry data, chemical
identifiers,  exposure standards, National Toxicology Program test status, and status
under various Federal regulations and programs.
       Many  S/L agencies have  frequently used  the ACGIH list of chemicals as a
basis  for pollutant selection, particularly for noncarcinogens.  The ACGIH list is
especially useful for identifying substances used  in the workplace that are toxic and
have  the potential to be emitted into the ambient air.
       Finally, EPA's Integrated Risk Information System (IRIS) is an important
source of information on the health effects of over 260 chemicals.  The IRIS was
described above, in Section 1.2.2.3.

       Step 4:  Internal Assistance From Divisions Within the  Agency. Another
method of choosing pollutants to list is requesting recommendations from  agency
divisions outside the air pollution group.  The public health  department or division
can identify toxics associated with public health problems in the State or locale.   The
water pollution department or division may be able to provide information on the
most  frequently discharged water pollutants.  The air pollution group  can then
evaluate how those substances are used  in industry and whether the substances may
be emitted and become airborne pollutants.  The same approach may be applied to
information obtained by the State hazardous waste management staff.  Permits issued
                                       2-12

-------
under the Resource Conservation and Recovery Act (RCRA) contain lists of the
constituents of hazardous waste handled by a facility.  The  air pollution group could
evaluate chemicals on those lists for the potential for the pollutants to be emitted to
the atmosphere, based on their use in industry.  The overlap of toxics programs and
RCRA is discussed further in Section 43.2.2.

      Step 5:  External  Assistance.  Several agencies have  requested  assistance from
individuals or groups  outside the  agency during  development of a list  of pollutants or
after pollutant selection,  as an external review.  Toxicologists and physicians from
local medical and research institutions can assist in evaluating health effects of toxic
pollutants. Consultants and members of the academic community can perform
technical review of potential  sources and quantities of the toxic pollutants listed as
well as  review of the  policy implications in listing pollutants.  Personnel from outside
of the agency can  give different perspectives on policy issues  and  supplement the
knowledge of agency  staff.
      Many agencies have requested industry participation  in development of their
entire air toxics program, including a review of  the list of pollutants to be covered.
These agencies report that the success of their air toxics programs was due in part to
industry involvement that led  to increased  awareness and cooperation.  Participation
by public groups, citizens' committees, and environmental groups can yield the same
benefits.

22   SELECTING SOURCE CATEGORIES OF  CONCERN

      There are three main  issues with respect to source selection that S/L  agencies
encounter when designing an  air  toxics  control program:
      (1)    What types of air toxic-emitting sources, if any, should be excluded from
             coverage under the  program?
      (2)    Should the program apply to existing as well as new sources?
                                       2-13

-------
      (3)    Should sources with very low levels of toxic emissions (de minimis
             levels) be exempted?

      Similar to pollutant selection decisions, resolution of these issues is critical in
determining program  effectiveness,  and the  size and nature of the regulated
community.

      1.     Types of Sources to be Regulated.  Air toxics control programs open  up
a whole new universe of sources that need  to be considered and evaluated for
possible regulation.  The focus will not automatically be on large volume paniculate
matter (PM) or volatile organic compounds (VOC)  sources as it has been
traditionally.  Sources that were previously viewed as being too small of an emitter
(e.g., auto body painters)  and/or as having  configurations not easily amenable to
control (e.g., small volume printers) now are being examined under toxics programs
because they emit one or more chemicals deemed to present some measure of health
risk.  Area  sources, whether large or small, that have been unregulated in  many State
VOC programs, must now be better identified and characterized because of the
aggregate impact of their emissions.  Programs based on risk levels or ambient
concentration limits may bring into consideration some very small sources because,
even though emission quantity is low, pollutant  toxicity may be quite high.   The high
toxicity may trigger the review of the source under the air toxics program.
      For any air toxics program,  there is a strong  presumption in favor of regulating
all  sources that emit  air toxics in amounts that  ma}' pose health problems to the
public.  However,  such a broad regulatory scope may not be practical  in terms of
agency resources  and priorities.  Therefore, the issue usually becomes, "Which sources
or source categories can justifiably  be initially excluded from coverage without
seriously jeopardizing the achievement  of the program's long-term goals?"
      State  and local agencies have  adopted a variety of approaches to resolve this
issue.  There is no one correct or  necessarily best approach.  The resolution of the
issue will depend to a large extent on  the number and mix of sources in a given State
or locale, the degree of risk posed by air toxics sources,  the availability of controls,
                                        2-14

-------
the resources of the agency, and numerous other agency-specific factors.  Some of the
more common source  exemptions are discussed below.
       Mobile Sources.  With the exception of California, no S/L agency currently
includes  mobile sources  in their  air toxics program because of the dominance
(although not preemption) of Federal regulation in this area, and the resources
required for  S/L agency regulation of mobile sources. This is not to say that mobile
sources are not a  major contributor of air toxics exposure — they  are!  However, S/L
agencies  have tended to focus on stationary sources for the reasons  mentioned above.
The only air toxics-related mobile source regulations implemented by S/L agencies  are
carbon monoxide and  ozone State Implementation Plan measures.  These measures
can indirectly control air toxics through the co-control of criteria and noncriteria
pollutants.

       Area Sources.   Area sources  (e.g., dry  cleaners, woodstoves, fireplaces and
other small, but numerous sources)  are regulated in some jurisdictions, either under
air toxics regulations or  separate regulations.  Collectively, such sources can emit
significant amounts of toxic air pollutants. Because of their large number,  these
sources are usually too numerous to review on  an individual basis.  As such,  S/L
agencies  should carefully evaluate their ability and need  to  cover such sources initially.
      Nontraditional Sources.  A difficult issue for many agencies is the regulation of
various  "nontraditional" sources, such as wastewater treatment plants and hazardous
waste landfills.  These have been shown to be significant air toxics sources in  several
studies.   There is  a strong presumption in favor of ultimately regulating such sources;
however, an agency should carefully evaluate the number, size, and type of such
sources  in their area to decide whether regulation would be  practicable in the
beginning.

      Sources Regulated Under the Clean Air Act. Some States have included
provisions in their regulations that exclude  from State regulations pollutants  regulated
                                        2-15

-------
under Sections  lll(d) or 112 of the Clean Air Act  The Virginia regulations (Section
120-05-0301) state, "The provisions of this rule do not apply to the following:

      Hazardous air pollutants regulated under Section 112 of the Federal  Clean Air
      Act, except to the extent such pollutants are emitted from facilities which  are
      not subject to emissions standards in Rule 6-1."   (Rule  6-1 includes Section 112
      standards enforced by Virginia).

      By including such "generic" exemptions, an agency would not have to amend its
regulations when additional 112  or  lll(d) standards are promulgated or when Clean
Air Act amendments are passed.
      In the 1989 STAPPA/ALAPCO survey, about 17 S/L agencies said that they
exempt  or plan to exempt  certain source categories. Table 2-3 summarizes  some of
the source category exemptions noted in the  questionnaire responses.  Several
agencies noted  that they exempt mobile sources and/or area sources; however, the
fact  that a State  did not mention area and mobile sources as  exempted does not
necessarily mean that the State regulates those sources.  Most State responses focused
on point source exemptions.  Before interpreting any of these exemptions, it is
important to  read the State's own regulatory language to avoid misconstruing the
context  of the exemptions.

2.3   REGULATING EXISTING SOURCES

      The second major source selection issue is whether  to regulate existing as well
as new  sources.  New and  modified sources are now addressed for their air toxics
impacts in virtually all S/L air toxics programs.  Starting with new and modified
sources  is a natural step because of the required review and  generally more favorable
control  economics for these sources.  Because existing  sources represent the origin of
all existing toxic  air pollutant emissions, the question should really be viewed as,
"When?" - not whether.  Ultimately, existing as well as new and modified sources
should be regulated.
                                       2-16

-------
        TABLE 2-3.  SOURCE CATEGORIES S/Us REPORT EXEMPTED
California



Iowa

Kentucky



Maryland


New Mexico
New York
Wisconsin
CA-Ventura



KS-Kansas City

KY-Louisville

NV-Las Vegas
Sources whose emissions are small are deferred so
resources  can be spent on source  categories  that pose a
higher risk.

Mobile sources.

NESHAP sources, laboratory equipment, dry cleaners.
indirect heat exchangers  using fossil fuel only, gas stations,
agricultural  operations.

Fuel burning equipment, char-broilers, gas stations,
NESHAP or potential NESHAP sources.

Gasoline service stations; automotive repair shops; laundry,
cleaning and garment services;  domestic wood stoves and
fireplaces; oil and gas production  facilities; agricultural
production and services;  containers,  tanks,  etc. without
vents;  nonprocess fugitive emissions  from  stationary sources
such as construction sites and mine  tailings.

17 specific source  exemptions including certain petroleum
liquid  storage and transfer facilities, certain solvent  metal
cleaning processes, certain surface coating  operations,
perchioroethyiene  dry cleaning  facilities.

Depends on  pollutant. Exemptions  include combustion of
certain fuels, laboratories, certain  gasoline dispensing
facilities, incinerators, indoor fugitive emissions,  and certain
wood combustion.

Gasoline dispensing operations using Phase I and II vapor
recovery system, small perchioroethyiene dry cleaning
facilities.

Existing sources.

Service stations, dry cleaners, fossil fuel burners.

Dry  cleaners, gas stations, businesses consuming less than
1000 gallons of solvent per year.
                                       2-17

-------
  TABLE 2-3.  SOURCE CATEGORIES S/L's REPORT EXEMPTED (Continued)
PA-Philadelphia           Commercial fuel combustion sources, retail dry cleaners,
                         gasoline service stations, incineration other than by-product
                         industrial work, and a list of "incidental" sources exempted
                         from emission  reporting, but not from "no health hazard"
                         requirement.

WA-Seattle               29 specific source exemptions including certain types of the
                         following:  fuel burning equipment, internal combustion
                         engines, laboratory equipment, storage tanks, water cooling
                         towers  and ponds, restaurants,  cold solvent cleaners, gas
                         stations, and dry cleaners.
                                       2-18

-------
      One possibility is to establish more flexible requirements for existing sources
such as process changes, raw material substitution, and other pollution prevention
initiatives.  Another approach for States with a smaller population of high risk point
sources is to address each problem source on a case-by-case basis using general air
pollution control/health protection authority.
      One concern  in the regulation of existing sources is how to cover sources that
typically do  not have to obtain air quality permits, but that are toxics emitters.  The
coverage of existing sources,  either immediate  or in a phased  approach, presents
additional implementation problems in terms of applicability determination.  If the
coverage is immediate,  all sources emitting chemicals  of concern will need to be
identified  and evaluated according to the basis of the program (i.e.,  emission limits,
acceptable ambient levels, etc.).  Many emitters of toxics will  not be traditionally-
regulated  PM and VOC sources, and inventories of such sources will likely not exist.
Whether the program has immediate or phased coverage for existing sources, the
identification of sources  releasing the regulated chemicals  may present problems for
the source types not traditionally  regulated or permitted.
      The inclusion of existing sources at some point should definitely be an element
in the  design of any air toxics program.  In the STAPPA/ALAPCO  survey, 20 States
and 7 local  agencies reported that they regulate or have definite near-term plans to
regulate emissions from  existing sources.  Many agencies that  have a system of
operating  permit  renewal in place begin to regulate existing sources  on this cycle of
permit renewal.   In  these cases, existing sources are exempt, or control is deferred,
until their operating permit comes up for renewal.  For programs covering  existing
sources in a phased approach under their criteria pollutant program, attention needs
to be paid to  the frequency and  requirements of the permit renewal system.  The
review cycle should be  reasonably short (i.e., no more than 5  years)  and capable of
imposing substantive new control  requirements in  order  to be  effective.  In addition,
Clean Air Act amendments that  have been proposed include operating permit
programs  capable of assessing and collecting a fee which would compensate the
reviewing  agency for its  time and expense.
                                       2-19

-------
2.4   SELECTING SIZE AND EMISSION CUTOFFS/EXEMPTIONS

      A fourth major decision point in determining the scope of an air toxics
program is the selection of size and emissions cutoffs/exemptions. Such
cutoffs/exemptions are a standard feature of most programs either as source screening
levels or regulatory exclusions.  Several of these agencies have stressed that this is
one of the most critical decisions associated with launching a new program.  There
are at least two  ways to structure such cutoffs/exemptions.

      1.     Lower Size Cutoff.  One approach is to apply lower size cutoffs.  These
can either be source-specific (e.g., production rate) or tied to the definition of a
major source/major modification (e.g., regulations  apply only to  sources classified as
major under criteria pollutant  programs).

      2.     Emissions Cutoffs/Exemptions.  Another approach  is to apply ^§ mim'mis
emission cutoffs  or exemptions.  Some agencies use stack concentrations while others
use emission rates. Some cutoffs/exemptions are pollutant specific (e.g., Maryland,
North Carolina).   Others are generic  (e.g., Pennsylvania exempts any source whose
stack concentration is no greater than the applicable AAL).  Table  2-4  summarizes
the jig minimis emission levels reported by S/L agencies responding to  the
STAPPA/ALAPCO survey.  For a complete understanding of these exemptions,
readers  should consult each agency's regulations.

2.5   TIMING/PHASING OF PROGRAM COVERAGE

      The development of an air toxics program is clearly not  a one-time effort, but
rather an ongoing task that requires multiple iterations  (e.g., evaluation of how
current  practices  are meeting agency  objectives, revision of practices to  meet
objectives,  and development of new objectives).  Some agencies  have planned their air
toxics control program to be implemented in several phases or  generations.  For
                                       2-20

-------
instance, they may target certain high risk point sources in the first generation and
then look at "urban  soup" in the second or third generation of their program.

2.6   AGENCY RESOURCE REQUIREMENTS

      As mentioned previously, agency resources are a key factor to consider before
the scope of the program is defined.  Some agencies have deferred control of certain
sources  or pollutants due to their limited available resources.  Another approach is  to
require  industry to provide information such as emissions data, monitoring data, and
modeling results to the agency, thus saving the agency's resources  that would be
needed  to perform these tasks. The Maryland Air Management Administration used
such an approach  during the development of their  program.  Maryland felt that
significant agency  resources were conserved this way.
      The  1989 STAPPA/ALAPCO  survey asked for dollar and work hour  allocations
devoted to  air toxics programs  in the  past year for specific work areas, such as
permitting,  program  development, and monitoring.  The STAPPA/ALAPCO document
can be referenced for  program allocations for individual S/L agencies.
  For  information  on  obtaining  the  compilation  of  the  results  of  the  1989
  STAPPA/ALAPCO survey, contact STAPPA/ALAPCO at (202) 624-7864.
                                     2-21

-------
          TABLE 2-4.  DE MINIMIS LEVELS REPORTED BY S/L AGENCffiS
California-Bay Area AQMD
Kentucky


Maryland
Michigan-Detroit

Nebraska

Nevada

New Mexico



New York


Oklahoma
Oregon


Pennsylvania
Noncarcinogens:  acceptable exposure value/4
expressed as an ambient concentration (g/nr) and  as
emission rates (Ib/day).  No exemptions for
carcinogens.

Pollutant-specific emission levels are given in Ibs/hr
in regulations  for new and existing sources

Noncarcinogens:  <5 Ibs./hour if screening level is
>200 ug/rn3'  Carcinogens:  same as above, but  also
<350 Ibs/yr if screening level for  carcinogenic  effect
is  > lug/nT.

Case-by-case

2.5 tons/year

0.25 lb/8 hours

Applicability thresholds are pollutant-specific and
presented in the regulation as emissions in Ibs/hr
and Ibs/yr.

Emissions potential of <3.5 Ibs/hr of VOC for
sources not given an "A" environmental rating.

Low toxicity pollutants:  6 tons/yr, not  to exceed 5.6
Ibs/hr Moderate toxiciiy pollutants: 1.2  tons/yr,  not
to exceed 1.1  Ib/hr High toxicity pollutants:  1,200
Ibs/yr, not to  exceed 0.57 Ibs/hr.
Compound-specific
pollutants.
                      minimjs limits  for list of
No definite cut-off established yet for many smaller
sources.  Their exempted status is based more
                                       2-22

-------
TABLE 2-4. DE MINIMIS LEVELS REPORTED BY S/L AGENCIES (Continued)
Pennsylvania-Pittsburgh


Rhode Island


South Carolina


Tennessee


Vermont -


Virginia
on resource limitations than on a defined de minimis
level.

Pollutant-specific, less than air toxic ambient
guideline in the  stack or emission stream.

Regulations give emission levels  in !bs/hr for 40
pollutants.

< 1.0 Ib/hr of PM uncontrolled
< 1,000 Ib/mo VOC uncontrolled

Ambient levels listed for pollutants specified in PSD
regulations.

Compound-specific levels given in pollutant  list,
expressed in lb/8 hours.

Emission rates in Ib/hr  depend on ACGIH  TLVs:
                                1-2
                                3-5
                                6-12
                               13-25
                               26-50
                               51-250
                               251-500
                               501 or greater
                                 0.
                                 0.13
                                 0.76
                                 1.52
                                 3.29
                                 6.58
                                12.90
                                63.51
                               126.77
Washington-Seattle

Wisconsin
Case-by-case consideration.

Regulation gives emission rate hi Ib/hr for listed
pollutants.
Wyoming
Determined on the basis of impact and exposure.
                                      2-23

-------
        SECTION 3.0
EVALUATING PROGRAM IMPACTS

-------
                     3.0  EVALUATING PROGRAM IMPACTS
                          Highlights of This Section
             Evaluation  of program  impacts,  at  least  to  some degree,  may  be
             necessary for either legal or practical reasons.
             Agencies should plan such evaluations carefully to ensure results are
             credible and program approval is facilitated.
3.1   WHAT ARE THE COSTS?

      Documentation of the estimated  costs of proposed air toxics regulations is  often
an important issue.  Estimates  of agency implementation costs may be needed for
administrative purposes, and estimates of the cost to affected industries may also  be
needed.  Effected industries and  the public at large have legitimate needs for
information concerning the likely consequences of proposed State or  local regulations.
For this reason, laws in many States and locales  require disclosure of the estimated
impacts of a rule prior to its adoption.
      In the case of air toxics programs, documentation of program  impacts  may be
necessary for strategic as well as  legal reasons.  Some S/L agencies have found it
necessary to prepare such evaluations simply to respond to industry claims that
businesses would experience unreasonable economic impacts.  Regardless of the
reason, as a practical matter agencies may find it difficult to convince their governing
boards or commissions of the reasonableness of the proposed rules without such an
evaluation.  Consequently, at least some cost and/or economic impact  evaluation  may
be a prerequisite for obtaining program approval.
      Agencies should plan such evaluations carefully before starting.  Estimating the
costs to industry of proposed air  toxics  controls, especially for existing sources, can be
expensive  and time-consuming.  If economic impacts are included, the task becomes

                                       3-1

-------
even more challenging.  Because  of the many assumptions required, the validity of
cost or economic evaluations is easily challenged.  Consequently,  such evaluations
should be carefully designed to ensure that the results are credible and the evaluation
will facilitate the program approval process.
      The following two subsections describe the economic impact analyses
undertaken by the States of Maryland and North Carolina, respectively.  The
approaches taken by the two States differ considerably. Maryland chose not to
estimate  emission reductions, costs, or risk reduction on a source-specific basis.
Instead, a cost-screening analysis  of four types of facilities was performed and control
costs were assessed qualitatively.  North Carolina chose a more quantitative approach
by performing source-specific dispersion modeling, calculating  emission control costs,
and estimating  exposure and risk reductions.

32   SELECTING A METHODOLOGY: MARYLAND'S  APPROACH

      Maryland's Air Management Administration prepared a cost analysis of their
proposed  regulatory program in 1986.  Maryland is now in the implementation phase
of their program.  The analysis by Maryland had two primary components: the cost to
businesses to comply and the administrative cost of implementing the regulations.
      The Department estimated that about 800 existing  installations at about 320
existing premises would be subject to the proposed regulations, in addition to about
140 to 150 new installations per year.  Existing premises would be  reviewed gradually
over the first four years, and then reanalyzed on approximately a three-year cycle.
      These estimates were based on a registration system that tracks new and
existing installations that discharge air emissions.  To estimate the number of effected
facilities, the Department counted all installations required to obtain an operating
permit that is renewed  annually,  and subtracted installations for fuel burning
equipment because they are exempt from the regulations.  The number of new
installations per year was based on a review of permit data from 1984-86.   The
Department did not attempt to estimate the  effect of small source  exemptions in the
regulations, or  related tasks companies  are required to perform to  comply with the
                                        3-2

-------
Superfund Amendments Reauthorization Act (SARA).  As a result, their estimates
were considered conservative.

      3.2.1  Compliance Costs

      For the analysis of the compliance costs to business, three categories  of costs
were estimated:

      o      emissions estimation costs,
      o      air quality modeling costs, and
      o      emissions control costs.

The Department did not attempt to quantify certain other possible costs to business,
such as  monitoring, risk assessment, and health effects data collection costs.

      Emissions Estimation Costs. Emissions estimation costs were divided  into two
categories:   (1) costs of performing stack tests, and (2) costs of estimating air toxics
emissions using other methods.

      1.  Stack Tests.  To estimate stack testing costs, the  Department worked
through the following calculations:

      (1)    (Total number of installations reviewed  annually)
      (2)    X  (Percentage of installations requiring tests)
      (3)    =  Number of installations requiring tests
      (4)    (-) Number of tests performed by the Department
      (5)    =  Total industry tests

      Then the Department estimated the average costs per stack test.  They
estimated a cost of $70,000 to $100,000 per test for complex installations.  An
estimate  of $20,000 to  $40,000 per test was used for less complex installations.  Based
                                        3-3

-------
on these estimated stack test costs and the  estimated number of installations tested
each year, the Department estimated the total cost of emissions estimation for
businesses for the first four years of the program.  The range in annual costs was
$75,000 to $260,000.

      2.  Other Methods.  The Department used a similar procedure to calculate the
costs of estimating emissions using methods other than stack tests.  All installations
were assumed 10 perform this type of estimation. Installations  were categorized as
either complex, moderate, or simple to reflect the varying costs for estimating
emissions.  Then the Department approximated costs for both  high and low cost
scenarios for each category.
      Using this procedure, the Department estimated that annual costs for  non-stack
testing methods  ranged from approximately 596,000  to $398,000 per year.
      In response to industry concerns about the cost of emissions estimation using
these other  methods, the Department  took  several steps to niinimize cost  impacts.
They included:

      o     Allow existing installations plenty of time to estimate emissions by
             phasing in the program over four years,
      o     Require existing premises to quantify emissions only for specifically listed
             compounds,
      o     Provide de minimis emission cutoffs,
      o     Provide "back calculation" procedures for estimating emissions from
             physical source data and screening  levels,
      o     Conduct a series  of workshops to familiarize businesses with
             emission estimation and  modeling methods, and
      o     Provide assistance to businesses unable to develop estimates.
                                        3-4

-------
                Costs.  For this estimate, it was assumed that modeling would be
done for entire premises, since the ambient impact requirements of the program apply
to entire premises  rather than individual installations.
      Based on a  screening study of approximately 50 premises, the Department
assumed that about 70 percent of the regulated premises would be able to
demonstrate compliance using the Department's conservative screening dispersion
model (TM  86-02). This would involve only minimal  costs to those businesses.
      To estimate the number of businesses required  to perform more rigorous
modeling, the Department assumed that about 50 percent of new installations will be
constructed at existing premises.  The Department further assumed that about  70
percent  of new installations located at existing premises  would be part of premises
that do  not  have operating permits.   These assumptions  were based on department
permit data.  For purposes of making the  analysis conservative, the  Department
assumed that each new installation locating at a new premise and each new
installation locating at an existing premise  without an  operating permit would be
required to perform the more rigorous modeling.
      Using these procedures and assumptions, the Department estimated that about
570 premises would incur minimal modeling costs by using TM 86-02 or other low-
cost methods. Nearly 250 premises would require additional modeling.  The
Department expected to perform the modeling for about 220  of these.  Thus, only
about 4 percent of the regulated premises  were assumed to incur modeling costs.
      For more complex premises, the Department  assumed average modeling costs
of from $15,000 to 530,000.   Very complex premises were estimated to incur average
costs of from $50,000 to $70,000.  By multiplying  the average modeling costs by the
aggregate number  of premises estimated to incur  such costs, aggregate modeling costs
ranging  from about $195,000  to $460,000 annually were  estimated.

      Emission  Reduction Costs.  For this category of cost impacts, the Department
chose not to provide detailed source-specific and aggregate estimates of emission
control  costs because of the expense  of such an analysis for the  Department,
confidential  plant data, and because of the heterogeneous nature of the different
                                       3-5

-------
facilities involved.  In lieu of a source-specific analysis,  the Department performed a
cost screening analysis of selected existing facilities and qualitatively assessed the
control  cost implications of the proposed program.
        The Department adopted a source-specific regulatory approach using a
conditional standard to create the most cost-effective regulation.  This  approach allows
a case-by-case appraisal  of the costs  and benefits of controlling each source.  Also,
sources  are  allowed to choose the method of compliance  which is least costly given
their individual characteristics.  Options include  choice  of control technology,
substituting  different raw materials, making process changes,  proposing higher
acceptable levels, and, in some cases, increasing stack heights.  This flexibility was
another reason why the  Department  found it difficult to estimate the  statewide
impacts of their proposed program.
      To estimate  the impacts on new installations, Department staff,  with the
assistance of a contractor, evaluated  the cost of meeting the  States proposed T-BACT
and ambient impact requirements for four types of facilities,  including:

      o     dry cleaners  discharging perchloroethylene,
      o     a refractory storage pile discharging chromium  and  aluminum oxide,
      o     an adhesive application facility discharging either toluene or methylene
             chloride, and
      o     a specialty steel facility discharging chromium and nickel.

      The analysis showed that new commercial and industrial dry cleaners would
require  no additional  control beyond current regulations.  T-BACT was assumed to
include  a leak detection and repair  system for controlling fugitive perchloroethylene
emissions.  This cost was estimated  to range from $200 to $5,000 for  the detector and
$50 to  53,000 for subsequent annual  maintenance.  Costs for the refractory storage
pile facility  were estimated to $180,000 per year for a building enclosure with a fabric
filter.  No additional control costs were estimated for the adhesive application facility.
                                         3-6

-------
Because of New Source Performance Standards requirements, the specialty steel
manufacturer was also estimated to incur no additional costs.

      3.2.2  Administrative Costs

      The Department evaluated administrative costs according to five categories:

      (1)    Personnel
      (2)    Laboratory equipment
      (3)    Data processing system
      (4)    Other equipment
      (5)    Operating expenses

      Costs were estimated  to range from  a first year cost of about $400,000 to about
5250,000 per year for subsequent years. Personnel costs included salaries and fringe
benefits for three engineers, a toxicologist,  a lab scientist,  a data device operator, and
a secretary.  The Department also assumed the purchase of a gas chromatograph-mass
spectrometer, sampling trains, and  a sample concentrator in the first year.

3.3    SELECTING A METHODOLOGY: NORTH CAROLINA'S APPROACH

      The approach used in  North Carolina's  assessment included an examination of
325 of the approximately 3,000 permitted sources in  North Carolina.  The facilities
selected by the State  for the  study represented a variety of major and minor potential
sources  of air toxics in all parts  of North Carolina.   Based on information supplied by
facility owners/operators, a list of 85 facilities  with emissions of air toxics exceeding
trace  amounts was compiled.  These 85 facilities were then screened by the Division
of Environmental Management (DEM) using accepted dispersion modeling techniques
to determine whether ambient air concentrations exceeded  the levels specified in the
States proposed regulation.  Facilities whose stack emissions exceeded acceptable
levels under current operating conditions were modeled to  determine whether ambient
                                       3-7

-------
air concentrations resulting from their emissions would be exceed the ambient limit if
their exhaust stacks were increased to 1.5 times the building height or,  failing that,  to
the maximum height allowed under federal  Good Engineering Practice  (GEP) exhaust
stack guidelines.
       Using the results of the dispersion modeling analysis, DEM estimated the costs
of complying with the proposed regulation for each facility. For those facilities
determined to be capable of complying with the  regulation through use of dispersion
techniques alone, the cost of an adequately-sized exhaust stack was calculated.  For
facilities determined to require add-on control technology, such as fabric filters or
carbon absorbers, the costs of applicable add-on  control systems were calculated.
Cost estimates did not take into account any existing emission controls  at the
facilities.
       Using these cost results, the potential economic effect on each facility  was
examined in terms of estimated product price increases and/or profitability decreases.
If a facility's products could potentially experience a price increase of more than
5 percent, assuming no absorption of cost increases, the facility was predicted to
experience a significant economic impact.  Similarly, if a facility's profit margin could
potentially drop  to zero as a result of compliance with the regulation,  the facility was
predicted to experience a significant economic impact.
       The benefits of the proposed regulation were assessed by analyzing selected
facilities representing the potential  range of benefit impacts and related costs resulting
from the  regulation.  Benefits were measured in  terms of the  reduced human
exposure  to ambient concentrations of toxic chemicals  in the vicinity of the example
facilities.   For carcinogens, benefits were expressed in  terms of the reduced annual
incidence of cancer for exposed populations resulting from the regulation.  For
noncarcinogens, benefits were expressed in  terms  of the number of individuals whose
exposure  to air toxics would  be reduced to concentrations below proposed ambient
thresholds.  The benefit estimates were then coupled with the estimated costs of
compliance for each example facility  to obtain an overall indication of  benefits
relative to costs.
                                         3-8

-------
The primary findings and conclusions of the study were as follows:

o     A majority of existing sources included in the  survey used to develop the
      findings in this report (216 of 325) would not  likely be affected by die
      proposed regulation because they do not emit  any of the 116 listed toxic
      air pollutants in more than trace  amounts. Eighty-five of the 325
      sources surveyed emit more than  trace amounts  of one or more listed
      pollutant

o     Fifteen of the 85  facilities emitting more than  trace amounts of air toxics
      would comply with proposed acceptable ambient levels under existing
      operating conditions.

o     Seventy of the 85 facilities  emitting air toxics in more  than trace
      amounts showed  ground level concentrations exceeding acceptable
      ambient levels for one or more of the 116 listed pollutants under current
      conditions.  This finding is  based  on conservative dispersion modeling
      assumptions  (i.e., the dispersion modeling analysis may predict higher
      concentrations than  actually exist). In some cases, facilities did  not
      report stack  parameter data and conservative assumptions were used to
      assign model inputs for these emission points.   More detailed dispersion
      modeling on a small number of facilities for the benefits analysis showed
      that many of these same facilities would be able to comply with  the
      proposed regulation without further control.

o     Forty-two of the  70 sources modeled hi the stack height analysis could
      comply by raising their stack height,  thereby reducing ambient impacts to
      acceptable levels.  However, most of these would need to raise  their
      stack to the  maximum height allowed by GEP to comply.

o     None of the 42 facilities required to  increase their stack height  to
      comply with  the proposed  regulation would likely experience  significant
      economic impacts.

o     Twenty-eight of the  70 facilities modeled would need to upgrade existing
      or install new add-on control systems in order to reduce ambient impacts
      to acceptable levels.  Ten  of these could potentially experience
      significant economic impacts as  a result.

o     The analysis of benefits of  the proposed regulation showed that cancer
      cases among persons living in the vicinity of the 4 example facilities
      analyzed for carcinogenic pollutant emissions would be
      reduced from 1 case every  20 years to 1 case  about every 60 years for
      two example facilities.  Reductions in annual cancer cases for another
      example facility would be from  1  case every 25 years to 1 case about
      every 40 years.


                                  3-9

-------
      o     The cost/benefit ratio for control of sources of carcinogens ranged from
            a low of $90,000 per cancer case reduced to a high of $300 million per
            cancer case reduced.  The cost/benefit ratio for the other 2 example
            facilities analyzed were $2 million and $33 million per cancer case
            reduced, respectively.
      o     In the case of noncartinogens, the number of exposed individuals would
            be  reduced by 10 near 1 of the 4 example facilities analyzed and by
            1,960 near another example facility.  The number of exposed individuals
            would be reduced by 229 and  1,760 near the other 2  example facilities,
            respectively.
      o     The cost/benefit ratio of control for sources of noncarcinogens ranged
            from a low of S9 per exposed  individual  to a high of $160 per exposed
            individual.

      North Carolina's experience with attempting  to estimate program cost and
economic impacts was  similar to  Maryland's.  The North Carolina DEM completed
their analysis after more than two years  work by agency and contractor personnel.
The report,  published in April 1987, subsequently underwent intense scrutiny by
several industry groups and individual companies.  Industry representatives criticized
the report on several grounds, most especially because the cost estimates did not
include any  direct cost input  from individual plants.
      Subsequently, the  North Carolina  Department of Commerce commissioned a
professor in the  business school  of a local university to critique the DEM report.  The
critique contended that the analysis was  unreliable  because of several  factors  not
considered in conducting the analysis.  The  critique concluded by stating that a
reliable study would be very  time-consuming and expensive to carry out.
      North Carolina's effort to  analyze the cost and  economic impacts of their
proposed air toxics regulation raised the same basic issue as Maryland's effort:   how
to accurately estimate  a  program's impacts within available resources.  North
Carolina's study, which included dispersion modeling, control  cost  estimation,  economic
impact analysis,  and cost/benefit  analysis for several dozen sources, cost approximately
$65,000.   At a minimum, a truly  comprehensive assessment of cost and economic
impacts would require several hundred thousand dollars and the collection of detailed
                                       3-10

-------
plant-specific data from numerous sources.  As of the drafting of this manual, a bill
was pending in the North Carolina General Assembly to more  clearly define what
constitutes an "adequate" cost and economic impact evaluation for a proposed state
regulation.

3.4   PLANNING A COST OR ECONOMIC IMPACT EVALUATION

      Where preparation of a cost or economic impact evaluation is necessary either
for legal or practical reasons, plans for the  analysis should be carefully  considered
before commencing work. A list of considerations to aid the planning of such  an
analysis aooears in Table 3-1.
                                      3-11

-------
      TABLE 3-1. CONSIDERATIONS IN THE PLANNING OF
                REGULATORY IMPACT EVALUATIONS
Carefully consider the goals and uses of the analysis.

Consider the types of impacts that are most relevant to program approval and
focus  the analysis on those impacts.

Evaluate the acceptability or credibility of the results if the scope of the
analysis must necessarily be limited.

Use accepted and defensible methodologies, especially if major simplying
assumptions are  necessary.

Involve industry  and other outside advisors in the design of the  analysis.

Use actual plant data to the fullest extent possible.

Avoid reporting  aggregate statewide costs unless confident  of the accuracy of
the estimate.

Carefully consider the representativeness of the sources analyzed.

Report the results in a manner easily understood by the public.
                                3-12

-------
   SECTION 4.0
AVOIDING PITFALLS

-------
                            4.0  AVOIDING PITFALLS
                            Highlights of This Section
      o     Challenges  to  a proposed  program should be  anticipated and  sound
            responses developed.
      o     An effective communications strategy can be essential to the success of
            a program.
      o     Duplication  of effort  can be  avoided  and resources conserved  by
            coordinating with other toxics-related regulatory  programs.
      o     Administrative  progress and control effectiveness  should be measured
            botn  qualitatively  and  quantitatively  :o  ensure   program  goals  are
            achieved.
4.1    DEFENDING THE PROGRAM ALONG THE WAY

      This subsection summarizes some of the most common charges leveled against
proposed S/L air toxics programs by those opposed to  their approval.  Air toxics
programs have been controversial in some States and their  approval may require a
strong defense of the program's need and features.  Opposition by parties affected by
the  regulations should be anticipated. Understanding the likely adverse comments  and
developing  sound responses to those comments early in the process may be one  of the
most important tasks in the development of a  program.
      A valuable piece of advice from an experienced  former air toxics program
official is to build a network  of S/L agency staff contacts to counsel you, especially
during the program design and approval stage.  Conferring  with others who have
already encountered similar opposition can be  extremely helpful in answering
questions, testing out responses, and building confidence.  The Data Base  Report on
State. Local,  and EPA  Air Toxics Activities  (EPA-450/3-89-29) published annually by
                                       4-1

-------
the National Air Toxics Information Clearinghouse (NATICH), includes an extensive
list of S/L contacts in various program areas.
      Table 4-1 highlights some of the most likely comments and offers some
suggested  responses, based on the experience of S/L agencies who have successfully
defended their programs.

42   ENSURING EFFECTIVE COMMUNICATION

      This subsection discusses ways to communicate effectively with interested parties
during program development  and implementation.

      4.2.1    Communicating Sariy and  Often

      At  EPA workshops on air toxics program development held across the country,
one of the topics for group discussion was the nature of communication needs in air
toxics program development and implementation.  The consensus of those discussions
was "Communicate early and  often! The success of your program will  depend on it!"
            Understanding the Nature of Environmental Communication
      Communicating successfully requires an appreciation of the unique nature of
environmental communication.  Communication on environmental issues is distinctive
in four ways.  The first  is distinctive feature is the complexity of information.
Environmental communication frequently involves complicated subject  matter such as
science, economics, law, business management and human behavior, with their trade-
offs and interactions.  The second way is the gap in technical knowledge of the
general public.  Technical information is presented about problems and  solutions, but
the public may well not understand  all of the terms, technical processes, etc.  The
third distinctive feature  of environmental communication is the personal impact of the
issues at stake. The air people breathe is, of course, a deeply personal concern.  The
fourth way is the
                                       4-2

-------
TABLE 4-1.   LIKELY ADVERSE COMMENTS AND SUGGESTED RESPONSES
   Comment
   Potential Response (Aa Applicable)
   There is little or no scientific
   documentation that an air toxics
   problem exists.
o  Several EPA studies have, shown that significant risks are
   associated with emissions of air toxics.

o  SARA Title HI data show that large quantities of toxics are
   released into the atmosphere - 2.7 billion pounds annually
   from manufacturing plants alone.

o  Air toxics programs are designed to prevent puoiic health
   problems from arising.

o  Citizen complaints and reported illnesses near industrial
   plants often can be traced to toxic air emissions.
   Pollutant selection is arbitrary,
   as evidenced by the wide range in
   the number of pollutants regulated
   by States.
o  Evidence is available to support regulation of specific
   compounds, including categorical findingsr case-by-case
   determinations, extensive screening of specific compounds,
   emission inventories,  and  recommendations  by  other
   organizations (e.g^ ACGffl, HIOSH< CAG, OSHA).

o  The list includes only those of local concern.

o  Federally regulated pollutants are exempt - thus there is no
   duplication.

o  The list is subject  to revision and is a "living list."

o  Industries can demonstrate  that emission  of a particular
   pollutant from their  plant does not pose a  problem.
                                             4-3

-------
  TABLE 4-1.   LIKELY ADVERSE COMMENTS AND SUGGESTED RESPONSES
                                         (continued)
     Comment
   Potential Response (As Applicable)
3.    Air toxics are already regulated
     under other programs.
o  SARA Title HI data show air toxics are not well controlled
   generally.

o  Criteria pollutant programs don't focus on toxics.

|)  Many significant existing sources of .iir toxics are oooriy
   controlled.

o  Some air toxics sources are not covered by criteria pollutant
   programs (e.g., POTWs, dry cleaners).
4.    Program costs are unreasonable
o  Demonstrated controls at a reasonable cost are available
   for most sources.

o  Sources can select among several control alternatives.

o  Case-by-case procedures are  available to consider any
   unreasonable impacts.

o  Most existing sources would not be affected; most new
   sources require stringent controls anyway.
                                              4-4

-------
            TABLE 4-1.  LIKELY ADVERSE COMMENTS AND SUGGESTED RESPONSES
                                           (continued)
     Comment
   Potential Response (As Applicable)
     Legal authority is inadequate
o  General enabling authority has proven sufficient in most
   states.

o  The State Attorney General has concluded that  adequate
   authority exists.

o  Federal law may soon reauire regulation of air toxics sources
   in any case.
6.    Local industry will be disadvantaged
     and experience unacceptable delays
     by complicated reviews.
o  Air toxics control is not the most important factor in
   competiveness on an international level (e.g., value of S).

o  All local industries would be subject to the same degree of
   control

o  New source reviews requirements tends to equalize interstate
   effects for all new sources.

o  EPA resources are  available to support implementation.

o  Some  additional costs due in part to criteria  pollutant
   programs.

-------
involvement of relative risks.  The distinction between voluntary and involuntary risks
is frequently a factor in environmental communication.

      4.2.3   Communicating Program Objectives

      In the course of planning, developing and implementing an air toxics program,
S/L's will need to communicate and coordinate with industry, the public, and a variety
of public interest groups.  The information needs of these  groups differ, both in
content and timing.  Industry participation in the program  development process is
particularly crucial because the regulatory agency will not only need help in emissions
data collection, monitoring  and/or modeling efforts, but will  need industry's
cooperation to minimize opposition to  the program.  Typical industry  communication
objectives during the problem definition stage  of the program development process  are
to:

      o     inform owners/operators of potentially affected  sources  of the reasons
             for data collection,
      o     explain how data will be used, and
      o     explain how to generate the information requested (e.g., how  to estimate
             emissions).
      In some cases, industries are skeptical of data requests by S/L agencies.
Reasons for resistance to supply information have  included the attempt to gather too
much information initially  (e.g., requesting information about emissions for  a very long
list of pollutants), and the failure  to adequately explain the reasons for data
collection.
      As the program is designed and implemented, various  industry groups may have
different information needs.  It will then  be necessary  to:

      o     document the impacts of  the program,
      o     explain the rationale for pollutants  and sources  covered and for the
             standards selected,
                                        4-6

-------
      o      explain compliance requirements, including inspection, stack testing,
             monitoring, and reporting requirements, etc., and
      o      negotiate in cases of potential violation.
      The information needs of the public or public interest groups may be somewhat
different from those of industry.  Communication with these groups typically involves
describing the proposed program, explaining general progress in controlling  air toxics,
explaining actions  and seeking public opinion regarding a specific case.
Communication with the public about a particular source is  likely to be one of the
most common communication needs after an agency's air toxics program is  in place.
      .Another important consideration is presenting the  impacts of the proposed
regulation.  These imDacis may include economic impacts  to industry as well as health
impacts to the general public.  Such  information should be presented as soon as
possible, with emphasis on any reservations about reliability.  Impacts should be
neither  exaggerated nor minimized, and worst-case estimates should be identified as
such.  Maryland has found that presenting impacts information to the public about
health issues is their most effective communication approach.  In most cases Maryland
has been able to state that predicted ambient concentrations are below levels
considered protective.  Keeping information simple and straightforward is beneficial.
It is also useful  to anticipate questions the  public will ask and to prepare general
responses.   Maryland stresses that it  is very important for the communicators to
understand  that  people are truly  concerned about their own health  and that of their
children.
      In summary, throughout the  air toxics program development  process,  the agency
will need to work closely with all interested groups  in order to:

      o     explain problem definition measures underway  and the reason  air toxics
             control is being investigated,
      o     solicit program  design assistance, and most importantly,
      o     achieve support from  and consensus among various interests that will
             help in achieving compliance with proposed control measures.
                                        4-7

-------
      4.2.4  Meeting Your Communication Objectives

      S/L agencies have used a variety of communication techniques to convey the
objectives of their program.  These include such commonly used techniques as
committees, public hearings, public and industry seminars, and newsletters.

      Committees.  One of the most successful and widely used techniques is forming
a committee or task force.  According to several agencies who have  sought  committee
advice and comment, there are two keys to successful involvement of outside groups.
First, the  representatives chosen to speak for a  group must have the authority to
speak on  the group's behalf.   Second, each representative must take  information back
to their respective group ;o keep that group well informed.  Also heipfui, if possible
is to have the committee or task force  complete its work during the  same
administration to minimize the turnover in committee or task force membership.
Having to educate new members is time-consuming and can undermine  momentum.
In selecting committee members, it may be helpful to have the Governor, or other
leading official, request directly that certain groups or individuals participate on the
committee.  This may  ensure  fair representation of all groups.
      Michigan's air board appointed a Special Advisory  Committee to  determine if a
problem existed  with toxic  emissions, and if so,  how such emissions should  be
controlled.  Several disciplines and interests, including industry, were  represented  on
the committee and were involved throughout program development.  All meetings
were open to the public and invitations were sent to  many  groups and individuals to
attend each meeting.   The Air Quality  Division staff  attributed a large measure of the
program's success to such public participation.
      Philadelphia's Health Commissioner appointed  an ad hoc advisory committee to
advise the air  agency in setting ambient guidelines.  Industry, academia,  government,
and the community were all represented on this ten-member committee  of  experts in
the fields  of toxicology, industrial hygiene,  and  occupational medicine. At the
beginning of the guideline development process, each member assumed  responsibility
for  researching and evaluating available information on several of the 99 substances
                                        4-8

-------
proposed for regulation.  Final guidelines were set by a consensus of the group.
Philadelphia's Air Management Services (AMS) noted that it was sometimes difficult

for the committee to reach agreement on a particular guideline or procedure, but

added that use of the ad hoc advisory committee composed of different interest

groups was important.  The AMS said it was important that all members of the

committee were willing to work toward acceptable solutions.  Connecticut has also

had experience in forming  a working group to assist in program planning and
development.  The Connecticut groups  are described in the case example highlighted

on the following pages.

      At EPA's air toxics workshops, staff members from Connecticut's Department

of Environmental Protection (BE?) Air Compliance Section described the various
groups involved In that agency's decision-making process in program planning and

.development.  In addition to  the Department staff, nine other  groups are involved:


      o      the State Implementation  Plan  Revision Advisory Committee (SIPRAC),
             a permanent steering committee [not formed specifically for the
             hazardous air pollutant (HAP)  program]  of approximately 50 members
             representing a broad  range of interests;

      o      the HAP Study  Subcommittee,  a seven-member SIPRAC subcommittee
             formed for  discussions of  detailed  issues, consisting of two attorneys  (one
             representing business and  industry and one representing an
             environmental advocacy group),  an industrial hygienist,  a State
             lexicologist, an industry engineer whose company does  not have
             manufacturing facilities in Connecticut, and two DEP engineers [The
             HAP subcommittee discussed and  resolved several issues and presented a
             final version of  die program description to die mil SIPRAC];

      o      specific interest  groups, such as  the Connecticut  Business  and Industry
             Association, who gave "reasonable and prudent" recommendations on
             several issues such as the setting of acceptable ambient levels;

      o      environmental activists including the Environmental Caucus, an umbrella
             organization comprising members of several groups [These groups were
             instrumental in the reversal of  the policy  to allow dispersion as a means
             of compliance];

      o      the State's Department of Health Services (DHS), who had had
             experience in working with an advisory board and developing health
             standards for drinking water [Through a  memorandum  of understanding,


                                       4-9

-------
             the two departments agreed upon a procedure through which DHS
             would provide expertise to DEP];
       o     U.S. EPA Regional office staff, who reviewed the program descriptions;
       o     the State Attorney General's office, who gave legal advice;
       o     the general public, who commented in public hearings after the
             regulations were drafted and reviewed by SIPRAC; and
       o     the State General Assembly, who mandated a study of the feasibility of
             an air toxics program and passed the regulations for  the HAP program.

       Public Meetings and  Public Hearings.  Some States, such as Maryland,  are
required by law to give citizens  an opportunity to request a public hearing when
certain sources have applied for a construction permit.  In Maryland, of the permit
applications for which public hearings may be  requested, about half result in a
hearing. Maryland's Air Management Administration recognized a need to discuss  the
permit applications and air  pollution issues with the public in a more casual
atmosphere than offered by a public hearing.   Thus, a less formal  public meeting is
usually held followed by a formal  public hearing.
       In explaining their experiences for the report National Air Toxics Information
Clearinghouse:  Case Studies in Risk Communication (EPA-450/5-88-003), Maryland
noted  that in  a few instances, a brief discussion of the source  and  permit application
was sufficient to allay public concerns, and that little opposition had been expressed at
the public  hearing even though,  in most cases, both the public  meetings  and  the
public hearings have been well attended.   The public meetings  helped focus attention
on air pollution control issues.  It was Maryland's experience that some people wanted
to discuss issues related  to the facility over which the air management staff has no
jurisdiction (e.g., traffic  problems).
       In preparing for public meetings and hearings, Maryland judges the level of
detail needed for each meeting,  based on  their familiarity with  interested citizens'
groups, and business and industry  representatives likely to be present.  If necessary,
the Department's Assistant Secretary for Toxics, Environmental Science  and Health
can be  called  on to describe risks.
                                       4-iO

-------
      In Maryland, the opportunity to request a public hearing is announced in
newspapers.  Citizens' groups in the affected area  are  contacted, but the responsibility
for becoming aware of potential hearings lies with the interested parties.  Maryland
feels that the public is pleased  that toxic emissions have been evaluated and that
efforts have been made to solve any problems.  Simple, straightforward verbal and
written presentations are  the most effective, with more detailed material available on
request.  They have distributed fact sheets  as people enter the public meetings and
hearings.  The fact sheets describe, in laymen's  terms,  the facility or plant, the
products it manufactures, the processes it uses, pollutants that will be emitted,
modeling results  showing the  concentrations predicted  for each of the pollutants,  and
an  acceptable ambient level for each  pollutant.

      Special Industry Meetings or Seminars. Some agencies have targeted certain
industries common in their area for special meetings or seminars aimed  at the toxics
issues of the particular industry.  For instance, North Carolina and  Virginia have both
worked with the  furniture manufacturing industry.  Through EPA's Control Technology
Center, Virginia  investigated and  met with  furniture manufacturers on the control of
formaldehyde emissions.  This same type of meeting with a special  interest group has
also been used with citizen and environmental groups, as well as with groups of
people who are interested hi emissions from a particular source.
      Similarly,  Maryland held discussions  on proposed regulations with
representatives of the Baltimore Chamber  of Commerce, the  Chemical Industry
Council, and individual businesses.  These  meetings  were held to ensure that effected
industries understood  how the proposed regulations  would apply to  their facilities  and
to give  them an  opportunity to explain any specific problems  they foresaw in
implementing the regulations.

      Newsletters.  Some agencies publish newsletters to inform all concerned groups
on issues the  agencies are addressing.  Newsletters are prepared for a  wide audience
with varied interests  and concerns rather targeted  at only one interest group.
Wisconsin's Department of Natural Resources, Bureau of Air Management,  publishes
                                        4-11

-------
a free quarterly newsletter called "Air Matters."  Similarly, the Northeast States for
Coordinated Air Use  Management (NESCAUM) also publishes a quarterly newsletter
called "Northeast ATReport"  The purposes of this newsletter are to improve the
exchange of information among the eight NESCAUM States  and to share the
northeast region's concerns, proposals, and  programs with air quality officials and
other interested parties throughout the country.

      42.5    Communicating About Risks

      As S/L agencies design and implement air toxics control requirements, risk
communication, or explanation of the health risks posed by toxic emissions, will be
required.  In 1987,  EPA's Office of Policy, Planning, and Evaluation estabiisnea  che
Risk Communication Program (RCP).  The goal of the RC? is to  provide technical
assistance in communicating environmental  risks to the public.  In  addition to assisting
EPA staff with their risk communication efforts, the  RCP sponsors its own projects 10
advance the state-of-the-art of environmental risk  communication.  Projects sponsored
by the RCP are selected after receiving input from the Risk  Communication  Work
Group, comprised of representatives  appointed by each EPA Office.
      Currently, the RCP is  concentrating  on risk communication training, radon,  and
Title III of the Supernind Amendments Reauthorization Act  (SARA Title HI).  Risk
communication training has been selected  as a priority by  the RCP in order to
expedite the transfer of risk communication skills  to  EPA and S/L agency staff.
SARA Title HI is an  area  of emphasis for  the RCP  because  of its potential
importance at  the community level.  Radon risk communication is  an area of
concentration by the RCP because of the large need for it perceived by the Office of
Radiation Programs.  Several projects form an integrated examination of how to
develop  an effective radon risk communication program.
      Also,  an information hotline is operated by the staff of EPA's RCP.  The
hotline gives program offices, regions, and S/L agencies quick access to research
results, provides the names of experts, and  provides review of proposals and draft
materials.
                                       4-12

-------
  For comments or questions about the RCP, contact Deny Allen at (202) 382-2747,
  or Ann Fisher at (202) 382-5500. The RCP hotline number is (202)  382-5606.
      Two pressing reasons exist for improving an agency's ability to communicate
with the public about risks. The Srst is that the most  serious remaining
environmental problems are not the most obvious ones.  Because the public ultimately
determines the regulatory agenda, it is important to put the  risks in  context for them.
Second, many current environmental hazards are beyond the traditional tools of
regulation and  require individuals to take action themselves (e.g., radon levels in
houses).
      The EPA shares these concerns and  has  identified four steps  of risk
communication through its RCP.  These steps, applicable to  ail communication
objectives mentioned hi this section are:

      Step 1.      Determine  the purpose  and audience of the
                   communication (this will guide the  factual content of
                   the message).
      Step 2.      Develop strategies for getting the message across
                   (this includes determining which groups and media
                   would  provide the best  mechanism  for reaching the
                   audience and establishing the credibility  of the
                   information source).
      Step 3.      Implement the strategies,
      Step 4.      Evaluate the effectiveness of the  risk communication
                   effort.
      The  EPA's Office of Policy Analysis  has published a brochure titled, "Seven
Cardinal Rules of Risk Communication".  While aimed at making and explaining
decisions that involve risk to the public, these guidelines  give sound advice to agencies
developing and implementing an air toxics program.
                                       4-13

-------
  To obtain the brochure "Seven Cardinal Rules of Risk Communication", contact the
  Risk Communication Program at (202)382-5606.
       Case Studies  in Risk Communication (EPA-450/5-88-003) published by
NATICH, is a further source  of risk communication aid.  This document profiles
three case studies, and includes the agencies' assessments of what was done well,  what
could have been done better, and what advice they would offer to other agencies
undertaking a risk communication program.  The three  agencies emphasized that  it
was very important to be open  and straightforward  with  ail  interested  parties, :o be
willing to answer questions  and investigate all  issues.
       Summary advice for other agencies includes putting risk estimates into
perspective  for the public, thoroughly evaluating the facility  in question, anticipating
the public's questions in order to present technical information clearly, and being  sure
of the facts and confident of explanations of agency policy before  going to the public.

       42.6  Working With the  Press

       When addressing air toxics  issues (whether the initial adoption of a program or
a particular permit application)  agencies will undoubtedly have contact with the press.
How  the press portrays the issue may greatly affect the  public perception or reaction.
Agencies may have news  conferences to formally pass information to local newspapers,
and to television and radio  stations.  In addition to large news conferences, individual
or small group briefings with reporters  may  be  useful.  Such small group  sessions  are
more  relaxed, allowing the agency to elaborate  on complex technical concepts.
       Of major interest to  the press recently are reports generated by SARA Title III
of toxic chemical releases.  These reports of millions of pounds of toxic air
contaminates released  annually have  generated considerable  public support for air
toxics controls.  S/L agencies should be familiar with SARA Title III and be  prepared

                                        4-14

-------
to respond to press and public inquiries about efforts to control these releases.
Agencies can expect continued press attention on Title HI data and its implications
for public health.
      From SARA Title in coverage and other experiences, agencies have learned
some important lessons.  La dealing with a source of inorganic arsenic emissions in
Tacoma, Washington, EPA and the Puget Sound Air Pollution Control Agency
(APCA) had almost daily contact with the press.  Both agencies felt they were  able to
build a  good relationship with reporters.   As described in Case Studies in  Risk
Communication, the Puget Sound APCA felt the agency had learned some  important
techniques for dealing with the press.  They found that the main concern among the
newspapers concerning the story was to be the first to publish a particular item.
regardless  of accuracy.  For this reason, Puget Sound APCA learned the importance
of communicating facts to the press understandably and accurately.  The agency found
that the press was interested in covering events  rather than explaining the process  to
be  used in decision-making, or explaining a  control alternative.  When the agency  staff
and reporters came to understand each others' concerns, they were  able to help each
other and work well together.

4.3   COORDINATING WITH OTHER REGULATORY  PROGRAMS

      This subsection describes how State Implementation Plans (SIPs) and other
toxics-related programs can complement an  air toxics control program. Existing
regulatory programs can  provide a wealth of data useful to air toxics control
programs.  For  example, if the air agency is trying to quantify evaporative toxic
emissions (e.g., waste solvents) from settling tanks, the agency's waste management
staff may be able to provide data on tonnages of waste present, composition, tank
sizes, and  tank operating characteristics that would facilitate the emissions calculations.
The waste management staff may routinely require reports that provide data needed
for air  toxics emissions evaluations  and control.   Planning air toxics programs in
conjunction with other programs conserves S/L resources and avoids the pitfall  of
duplicating work already done.
                                       4-15

-------
      43.1   Recognizing the Commonality of Different Regulatory Programs

      Seemingly different regulatory programs share a common environmental goal.
Each is charged with protecting public health.  State criteria air pollutant programs,
for example, limit  criteria emissions to levels that will  not exceed  health-based
national ambient air quality standards (NAAQS).  Air toxics programs aim  to lower
toxic emissions  to  levels whereby public health  risks are  minimized or kept  below
levels deemed to be unacceptable.
      Coordination with existing criteria air pollutant control programs is a
prerequisite for any  successful air toxics program.  In addition, other existing
environmental and health protection programs also  provide benefits  to an air  toxics
program even though they may not be specifically aimed at controlling air emissions.
Some of the more prominent programs  include:

      o     Superfund Amendments and Reauthorization Act (SARA)
      o     Resource Conservation and Recovery Act (RCRA),
      o     Occupational Safety and Health Administration (OSHA), and  the
      o     Clean Water Act  (CWA).

      Programs such as these cover a diverse set of topics and provide multimedia
toxics control opportunities.  Focused primarily on hazardous waste treatment, storage,
and disposal, RCRA has components that are applicable to air toxics. Activities
under SARA Title HI cover all media and have direct implications for air toxics
control. Although OSHA does not address air toxics  releases to ambient air,  it does
identify industrial and commercial activities that produce airborne  toxics of  concern.
The  CWA program can serve as  a good source of information to  evaluate potential
air toxics  emission sources from the standpoint  of identifying chemicals in a process
                                       4-16

-------
and its wastewaters.  More detailed information on the coordination possibilities
between these programs and a developing air toxics control program is given in
Subsection 433.
      Air toxics programs can benefit from coordination with these toxic-related
programs.  As discussed in Section 2.0, a S/L agency's initial goal is usually to
determine the sources emitting pollutants deemed to pose  toxic concerns.  Each of the
four companion programs listed above can provide useful information  on sources
known to be handling/producing chemicals that may be  viewed  by the air agency as a
toxic of concern.   Some programs may have  developed  emissions information that
could be  used to help bound the problem.  The air agency can investigate the
potential  for air emissions.  In addition, the  air toxics program  may benefit from
requirements imposed by these companion programs (e.g.,  a source is  prohibited  from
putting  waste solvents in an open-air  lagoon).  As the  S/L program evolves, reporting
requirements of other programs may  have enforcement  and source monitoring benefits
for the  air toxics control effort.
      As a result of these commonalities, S/L agencies  should  take advantage  of
every opportunity  to use information  developed by other programs and to optimize air
toxics control via co-control efforts.  Recognizing and taking advantage of these
opportunities will  save agency time, money,  and waste  from duplicated efforts.

      432        Coordinating with State  Implementation Plans (SIPs)

                   SIP Activities and Their Relationship  to Air Toxics
       Although the primary goals of SIPs are attaining and maintaining the National
Ambient Air Quality Standards (NAAQS) for criteria air pollutants, an important
secondary goal can be enhancing the control of toxic air  pollutants.  Reductions in
criteria pollutants through  SIP activities, in addition to Federal regulatory  efforts
focused on criteria pollutant control, have been shown to produce  substantial
reductions in specific air toxics that exist predominantly as  a form  of either PM or
VOCs.
                                       4-17

-------
      The air toxics problem overlaps with PM and VOC problems in two ways.
First,  many areas that are nonattainment for participate matter nominally
10 micrometers or less (PM^) and virtually all ozone nonattainment areas to be
addressed in upcoming SIP revisions involve large urban areas that are believed to
have air  toxics problems.  Second, the vast majority of toxic substances belong  to the
general categories  of ?M and VOCs.
      Major SIP changes  are under development in response to a revision in  the  PM
NAAQS  and  the passing of the attainment date for the ozone NAAQS.  Additional
controls designed to reduce criteria pollutants  can be justified, in part, by their
simultaneous payoff in air toxics coniroi.  Moreover, these SIP control efforts,
properly  designed,  can do mucn  more  to reduce ihe air toxics concerns associated
with increased cancer incidence and risks  to the maximum exposed individual.
      In 1987, as  part of the Clean Air Act's  requirement that EPA periodically
review and, if appropriate,  revise the  NAAQS,  EPA revised  the  primary  and
secondary NAAQS for total suspended paniculates (TSP) to apply to PMj0. In the
case of PMjg, agencies were to submit a plan  by August 1988.  All areas within a
State  are classified as Group  1, II, or HI, based on the probability of attainment of
the new PMj0  standards.
      In May 1988, EPA issued "SIP calls" for 66  areas  that failed to meet the
December 31, 1987, Act deadline for  attainment of the ozone NAAQS.   Revised
ozone SIPs will be developed in two phases.  The  first phase will focus on upgrading
emission  inventories and correcting loopholes and deficiencies in existing SIP's.  The
second phase, and its timing, will be contingent on EPA's final post-1987 Policy (or
Clean Air Act amendments).  Regardless  of the ultimate schedule for revising existing
ozone SIP's, it is clear that many State and local agencies will need to consider
additional VOC control measures in the coming months.
      Both the ozone and PM.0 SIP  actions offer several opportunities for enhancing
State  and local air toxics programs.  First, S/L agencies are  encouraged  to develop an
air toxics component for their ozone and PMj0  SIPs.  In this way, air toxics and SIP
goals  can be explicitly addressed simultaneously hi  the  context of SIP themselves.
                                       4-18

-------
One benefit of this integrated approach could be the enhanced acceptability of SIP
control measures that can be credited with both criteria pollutant and air toxics
reductions.  In addition, SIP control measures that may be somewhat marginal  from a
purely VOC perspective may be regarded more highly if they are demonstrated as
having significant health risk reduction benefits.
      In short, significant cost saving opportunities appear possible for S/L  agencies
by promoting  better coordination vith air toxics programs in the  development of SIPs.
Many of the same point and area sources are likely to be arfected by both programs.
             Coordination Opportunities
       Updating and upgrading emissions inventories for ?Mj0 and/or ozone SIP
purposes is  a good opportunity for S/L's to enhance the utility of these inventories for
air toxics  applications.  Several suggestions follow on how  to optimize inventory and
air monitoring  efforts, so  that the data  can be  used for multiple  purposes.  If
implemented, these ideas  are likely to reduce total resource requirements compared to
carrying out separate programs to accomplish toxic and criteria pollutant goals.

       1.   Source Category Coverage.

       It is usually beneficial to inventory the sources and/or use the results of
monitoring programs to identify specific air toxics of concern.  The S/L agency is
already aware of major point sources included  in criteria pollutant  inventories.  When
updating these  inventories for SIP purposes,  the agency can also  review these sources
to determine if they could be sources of air  toxics  of concern.  The EPA's  Air
Emissions Species Manual (EPA-450/2-88-003a and b) and the Toxic Air Pollutant
Source Crosswalk (EPA-450/4-87-023a) are good references to  use  to identify
potentially toxic constituents of paniculate  matter (PM)  or volatile  organic  compounds
(VOC) emissions, and the sources that  emit  them.  The Crosswalk associates emitting
source categories with specific pollutants and as such is  an indication of the potential
for a given  substance to be emitted by a particular source category.  The Ajr
                                       4-19

-------
Emissions Species Manual is a two-volume document that contains speciation factors
for PM and VOC that can be used to estimate the specific compounds that  comprise
PM and VOC emissions.
      Many of the  toxics-emitting sources have typically been included in past
emissions  inventories for PMj0  or VOC (ozone).  Some newly-implicated source
categories such as treatment, storage, and disposal facilities  (TSDFs), leaking
underground storage tanks, and publicly-owned treatment works (POTW) are
important sources of toxic emissions, but have not typically  been included in past
criteria  pollutant  emission inventories.  Including  these sources in updates to the
inventory  should be required because opportunities for coordination certainly exist
with water, hazardous waste, and solid waste program offices.

      2.    Pollutant Coverage.

      Air toxics may be in  the form of PM, volatiles, or semivolatiles. Metals and
certain polycyclic organics [e.g., benzo-a-pyrene (B(a)P)] will likely exist in the
ambient air in particulate form, almost always concentrated  in the smaller size ranges
(<  10 micrometers).  The VOCs of concern for ozone formation also are usually
composed of multiple chemicals, many of which may  be  considered toxic.
      Ideally,  the VOC and PM^ inventory should be completely speciated  for each
toxic of concern.  Where this is not reasonable, a smaller list of compounds  of
interest can be developed by first scrutinizing the applicable point sources.  Using
data on hand such as permit files, tests/analyses,  the  Air Emissions  Species Manual.
and "right-to-know"  information from sources as a result  of  SARA Title HI
requirements, point sources  can be  evaluated with respect to their emissions  of air
toxics of potential concern.
      Area sources, such as motor vehicles and wood combustion, have been shown
to be a major concern in many areas.  This concern  can usually be  traced to a
number of pollutants which  include formaldehyde, benzene,  1,3-butadiene, chromium,
B(a)P, and others.
                                       4-20

-------
      3.    Monitorin  Considerations in SIPs for Ozone and
      Ambient monitoring is usually an important aspect of preparation and execution
of a SIP.  The ambient monitoring program for PMy, and ozone will result in a
collection of particulate samples and some measurement of the nonmethane organic
compounds.  While these results are of little use  directly in the  assessment of air
toxics "potential," some supplementary analyses can be useful.  The paniculate samples
can be inexpensively analyzed for total metals (e.g., arsenic, beryllium, cadmium,
chromium), and it can be a reasonably minor effort to supplement VOC programs by
speciating total VOC samples using EPA-supported programs.  The EPA supports a
sampling program by which S/L's agencies can obtain ambient nonmethane  organic
compound/nitrogen oxides  (NMOC/NOX) ratio data for 6 to 9 a.m. summer weekday
situations.  Through this program, S/L agencies provide contract funding, sampling
sites, and personnel.  The EPA provides sampling equipment,  transportation, and
contractual analytical capabilities. A number of the VOC samples collected under the
program in past years have  been analyzed for organic species. This support
mechanism continues to be  available to S/L agencies that want to obtain both VOC
and air toxics  data for a preliminary look  at what air toxics might exist in the ambient
air.
      Arrangements can also be made for analysis of particulate filters for metals and
B(a)P and for sampling and analysis of cartridges for aldehydes,  ketones, and other
oxygenates.  The EPA will facilitate S/L efforts  to evaluate these samples.  A broader
but similar program also exists to provide for sample and analyze of 24-hour samples
taken every 12th day throughout the year, if desired.  However,  this program is  almost
exclusively in the interest of air  toxics screening  and does not contribute greatly to
PMj0  or ozone SIP development. The S/L  agencies should keep this information in
mind when defining the nature and extent to which their ambient monitoring program
for criteria pollutants is augmented  and/or sensitized to air toxics concerns.
                                       4-21

-------
   For more information on urban air toxics monitoring, contact EPA's Monitoring and
   Reports Branch at (919)541-5652.
       4.3.2.3   Co-Control Opportunities

       Once specific air toxics concerns have been identified, to the  extent time and
resources allow, S/L agencies are encouraged to develop and implement appropriate
co-control measures.  In urban areas these  measures might emDhasize  the  control of
specific source categories  and ooilutants. and  particularly area sources.  Outside urban
areas the focus of co-control might be more on the control  of specific large point
sources.  Several types of co-control measures to  address this range  of air  toxics
concerns are discussed below.
       Within existing legislative authorities, ongoing EPA regulatory efforts for
criteria and noncriteria pollutants  will contribute toward the future reduction of air
toxics levels.  Continued implementation of existing S/L and Federal requirements will
also reduce PM and VOC emissions which  are likely to contribute to the current  air
toxics problem.  Despite these  efforts, inadequate progress on  the air toxics problem
will, in many cases, lead some  S/L agencies to adopt significant additional mitigation
measures for toxics.  Any co-control measures  adopted by a S/L agency for mitigating
PM  and VOC air toxics for any given area  will depend on such factors as  the number
and  type of sources, the types and quantities of pollutants emitted, ambient
concentrations of pollutants, and the costs of control (as well as local political,
economic, and legal  considerations).   The S/L agency will need  to make these
decisions based on their own individual policies regarding  acceptable ambient levels
and  risk management,  as well as what otherwise will happen under Federal programs.
       Following is a discussion of several candidate co-control  measures for
consideration by S/L agencies.  These measures represent a starting  point for the
consideration of approaches to  mitigating risks from exposure to air  toxics. Where a
measure might involve a source category which is under consideration for a national

                                       4-22

-------
measure might involve a source category which is under consideration for a national
standard, coordination with EPA is recommended to understand the most recent
information regarding possible overlap prior to the adoption of this measure.
      For the most part, the candidate co-control measures described below do not
include  new or necessarily innovative ideas.  The list does,  however, identify the
mitigation measures that appear to offer the best prospect for reducing  the  cancer  and
noncancer  risks of exposure to PM and VOC air toxics, particularly in urban settings,
As the understanding of the  air toxics problem improves, new approaches may
emerge.  Additional ideas  for co-control should come  from future field  and research
experience.  For  example,  the control of condensible paniculate matter  (now
implemented by several S/L  agencies and being  researched by EPA) may well
accomplish co-control objectives for certain source categories.

      1.     More Intense Audit and Enforcement  of Existing SIP Requirements.

      Analyses show that  considerable indirect control of air toxics can be  achieved
through intensified efforts to enforce existing SIP requirements.  Aggressive  auditing
and  enforcement of regulations for VOC and PM sources, which also  emit toxic
pollutants,  can be a cost-effective means of minimizing emissions of air  toxics from
new and existing sources.  An intensified  enforcement effort aimed at sources of air
toxics could include:

      (1)    compliance audits of selected high priority sources (e.g.,
             those emitting significant amounts of high priority
             pollutants, sources with a history of poor compliance, etc.);
      (2)    increasing the number of enforcement  source  inspections;
      (3)    step-up of enforcement actions  against noncomplying
             sources which emit  air toxics;
      (4)    evaluation of facilities most likely to experience process
             upsets or control equipment malfunctions;
      (5)    strengthened  source monitoring requirements;

                                       4-23

-------
      (6)    greater use of opacity monitoring as an enforcement tool; and
      (7)    expanded coverage of ambiguous VOC regulations to sources not
             previously judged to be  covered by those regulations.
      2.     Selection of Additional PM and VOC SIP Co-Control
             Measures.

      Revision of existing PM and VOC  SIP's will be necessary in many areas to
attain and/or maintain the PMjQ  and ozone  NAAQS.  Control strategies that offer
significant air toxics benefits  and, at the same time, satisfy PM,0 or ozone plans
should be given serious consideration over those  that do not enhance air toxics
control.  Also,  there may be  measures that are marginal from an ozone or PM,0
perspective,  but which can be further supported on the basis of the payoff of
additional toxics control.  For example, it has been shown that  VOC can degrade in
the atmosphere to form formaldehyde; therefore, any additional control of VOC from
stationary or mobile sources  could result hi lower ambient levels of formaldehyde.

      3,     Adoption of Control Measures  in SIP's for Sources/Pollutants Not
             Included.

      Consideration should be given to identifying and adopting control measures for
sources or pollutants not currently covered under SIPs. Numerous small sources
which are potential emitters  of toxics are  often exempt from S/L permit requirements
(e.g., degreasers or miscellaneous metal or plastic coating operations).  Also, sources
not presently required to apply Reasonably Available Control Technology (RACT)
could be controlled to RACT levels if they are sources of air toxics (e.g., spray
painting bridges with  chromate paints).  Thirdly, consideration should be given to the
control of VOCs which could be  exempt,  under current regulations, because of
negligible reactivity (e.g., small furniture refinishing businesses using methylene
chloride as a paint stripper).   Finally, some sources and/or source-related operations
                                       4-24

-------
are exempted within SIPs (such as peaking power units and malfunctions), and these
exemptions should be reexamined in light of their toxics implications.  Control of each
of these sources may not have been justifiable previously on the basis  of criteria
pollutants control, but may be worthy of consideration under a co-control program to
address air toxics.

      4.    Nontraditional Squrces and Control Strategies.

      An examination of the itrban air toxics problem reveals that, in many instances,
nontraditional  sources of air pollution may be major contributors to  the problem.  In
turn, this  conclusion suggests chat solving  the problem may call for the adoption of
some less traditional control strategies.
      Acting to increase public awareness can be a cost-effective means of minimizing
PM and VOC and toxic emissions from area sources such as woodstoves and solvent
use and disposal.  Woodstove  emissions are a major source of toxic  air contaminants
and PMj0 in many urban areas.  High concentrations of polycyclic organic matter
(POM) emissions resulting from woodstove and fireplace emissions have  been shown
to result in significant seasonal cancer risks hi some areas.  Also, potential
nonattainment areas for PMj0  may  need to achieve reductions in wood smoke
emissions to demonstrate  attainment of the PMj0  NAAQS.
       A cost-effective means of minimizing risks from wood smoke may be education
of woodstove owners about proper  operation of woodstoves.  Such a public education
effort has been implemented in some areas, and EPA is considering a public
education campaign in conjunction with the NSPS for woodstoves. Using materials
presently  available from EPA  (e.g., woodstove pamphlet included in  the back of this
document), S/L agencies could, for example, advise  woodstove owners of ways to
reduce woodstove emissions (e.g., small, hot fires; seasoned  wood; appropriate size
stoves operated  at efficient levels; catalytic combustors) or eliminate woodstove  use  at
certain times,  such as during inversions.
       Another area where public education can play a role in minimizing air toxics
and VOC  emissions is in the handling and disposal  of solvents.  A number of toxic
                                       4-25

-------
organics  (such as trichloroethylene, chlorobenzene,  and nitrobenzene)  escape  into the
atmosphere in the form of fugitive emissions as a result of improper solvents disposal
by individuals and small businesses. Because the use  of solvents is widespread in
urban areas, improper handling and disposal practices can be a significant contributor
to air toxics and VOC loadings.  A reduction in such  VOC emissions may help  areas
that are  currently nonattainment  for ozone.
      State and local  agencies may also want to consider a program to identify small
businesses  that commonly use solvents and  educate them, as well  as  the general
public, about good solvent handling and disposal practices.  In some areas, local
solvent recycling programs and recycling cooperatives have been initiated to  address
this  need.  The effectiveness of such measures will depend on the  amount of effort
and  resources applied, the attitudes of the public and  small business community
concerning air pollution and environmental  protection, and the  expense  and
inconvenience of recommended disposal methods versus traditional methods.  An
effective  public education campaign about solvent usage and disposal  has the potential
to reduce risks from air toxics  exposure for a relatively small investment of public and
private resources.
      Because health  risks  to the public 'arise from a variety of sources and  through
a multitude of pathways and media, a  more integrated approach to controlling urban
air toxics may be warranted.  For example, the  influents and  effluents of public
wastewater treatment plants are indicated as a sizeable source of air toxics, especially
those that  receive  discharges from chemical plants.  Rather than directly controlling
emissions at the wastewater treatment  plan  itself, pollution prevention measures such
as industrial pretreatment programs aimed at reducing the discharge of organics into
public sewer systems, and the subsequent intermedia transfer  of organics to the
atmosphere, may be more appropriate.   Such  an integrated, multimedia  approach can
target the pollutants and sources  posing the greatest overall risks to the public,
regardless of the particular  environmental medium  involved.  Good communication
and  coordination between Federal, State and local  agencies is crucial  to the success of
an integrated, multimedia approach  such as this.
                                       4-26

-------
      43-3  Coordinating with Other Toxics-Related Programs

      Coordination opportunities in the establishment of air toxics programs exist
with several other program areas in addition to SEP programs  for PM and VOC.  The
most prevalent programs offering air toxics benefits through coordination are the
SARA Title HI program, the RCRA/CERCLA program, OSHA programs, and the
Clean Water Act program.  The generic opportunities for coordination within each of
these programs are  described in the following  sections, with examples given where
specific States have successfully implemented such  coordination.  Table 4-2  provides
contacts for information related  to other toxics-related programs.

      4JJ.1 SARA Title III Regulations

      The SARA Title HI provisions, also known  as the "Emergency Planning and
Community  Right-to-Know Act," can provide a great deal  of information that is
pertinent to the  air toxics control program.  Successfully coordinating the data
available from SARA Title HI with the information needs for  an air toxics  program
can lower the resources otherwise required for the air toxics initiative.
      Several portions of the  SARA Title ffl  program have relevance to an air toxics
initiative.  The sections that can provide the most  data and have the greatest impacts
on air toxics program development are Sections 304, 311,  312, and 313.
      The foundation  of the SARA program is to provide EPA,  S/L agencies, and
the public with information regarding the storage, use, manufacture, or release of
certain chemicals designated as hazardous.  The EPA has established a  list of
hazardous chemicals for which reporting must  be done.  The  list  currently numbers
over 400 chemicals  and is expected to grow over time.  The 400+ hazardous
chemicals are characterized as being toxic and/or extremely hazardous, which roughly
                                       4-27

-------
52      a
O
O
35
»— <
Q
CO
U
8
W
5


Contact




Infonnation Available


c«
W3
O
U
4—1
33
o
Ct,

V3
C
0




2
o
c

Ml
Asm
C1K4J

Manufacturers;
general public,
government
agencies


HH
< °
< o
co K

^
(2 v}
II
o *-•
00



P u
Data listing of Toxic Relea;
Inventory database availabi
tlirough National Library ol
Medicine (TOXNET)
T3 73
D O
=5 &
"5 .3
« -3

Tfi
FlI
Ul
y^j

Manufacturers;
general public,
government
agencies
&
«feo
Contact:
Regional TRI
Account Man
NATICH StaJ
(919)541-085




Database available through
National Computer Centei
amlNATICfl




.

id
1 1
|y|
i^ji

S/L agencies
t


^H
1— ( c«
&1 5
co Q


o
(202)554-14

W)
oS O
3^^ ^.
— « C
Infonnation on TSCA regs, rc(
for publications re; managemc
of toxic substances and referia
appropriate HP A personnel

O
c
o


M
LvuQ
UK4J
'^ -o
Chemical industj
labor and trade
! organizations;
environmental
groups; the publi
c
(U O
o • —
S c u
u -a o 2
CO «« *e w
E-i < £ co


-------
corresponds to chronic and/or acute health hazards, respectively.
      Reporting under the SARA Title HI provisions is due annually by July 1,
beginning in 1988.  The EPA is currently evaluating responses received, performing
quality assurance  checks, and developing a computerized data base of information that
is accessible by S/Ls through the National Computer Center in Research Triangle
Park, North Carolina.  In addition, States are also receiving the SARA Title m
submittals and data compilations.  In some cases, private groups such as the Natural
Resources Defense Council, the American. Lang Association, and the Sierra Club have
analyzed and summarized State  data.  Such analyses will likely be available to S/L
agencies and could provide information that could save agency resources for toxics
program development.  The SARA Title HI data are also available through the
National Library of Medicine and tne  NATICH database (see Table  4-2.)
     Perhaps the  most important part  of SARA Title HI from  an  air toxics program
standpoint is Section 313.  This section pertains to  uniform  toxic chemical release
reporting that must be done by industrial facilities meeting certain criteria.  The
reports  are filed with EPA and  State authorities. The Section 313 requirements apply
to facilities that:

      o     employ 10 or more persons full time,
      o     are  in Standard Industrial Classification (SIC) Codes 20 through 39 (as
             in effect on July 1, 1985), and
      o     in the calendar year for which a release form  is reauired. manufactured,
             processed, or otherwise used a toxic chemical on the list described below
             in amounts greater than  those specified.

      The toxic chemicals subject to the release reporting requirements are given in
the EPA publication 'Toxic Chemical  Release Inventory Reporting Form R (EPA-
560/4-88-005).  The EPA is continually evaluating and updating this list such  that
chemicals may have been added or deleted since the publication of this report. The
final determination on whether  a chemical release must  be reported is  based  on
established reporting thresholds.  For toxic chemical use at a facility, all uses  of
                                       4-29

-------
10,000 pounds per year or greater require Section 313 release reports.  For toxic
chemical manufacture or processing, the release reporting thresholds are
75,000 pounds per year for all forms due on or before July  1, 1988, 50,000 pounds
per year for forms due  on or before July 1,  1989, and 25,000 pounds per year for all
forms due on  or before July 1, 1990 and thereafter.
      The Section 313  chemical  release form (Form R) provides the name, location,
and primary business activities of a  subject facility.  For each listed toxic  chemical
emitted by the facility, the following data must be provided:

      o    Whether the  toxic chemical at the facility is manufactured, processed, or
            otherwise used, and the general category or categories of use of the
            chemical.
      o    An estimate of the  maximum amounts, in ranges, of the toxic chemical
            present at the facility at any time during the preceding calendar year.
      o    For each waste stream, the waste treatment or disposal methods used,
            and an estimate of  the treatment efficiency  typically achieved by such
            methods for that waste stream.
      o ~  The annual quantity of the toxic chemical entering  each environmental
            medium.
      The usefulness of the Section 313 data in the  problem definition and program
implementation phases of an air toxics control program are readily apparent.  Not
only are potentially significant sources identified by the pollutants emitted,  but
emission quantities are also provided.  Indications of the  potential for improvements
in control device use  and efficiency are provided.  The data can also be used as a
parallel enforcement tracking measure.  Similarly,  Section 313  reports can be used to
potentially measure the effectiveness and progress of the  air toxics control  program in
reducing overall airborne emissions.
                                       4-30

-------
      43.3.2 RCRA/CERCLA Regulations

      The Resource Conservation and Recovery Act (RCRA) and the Comprehensive
Environmental Responsibility, Compensation, and Liability Act (CERCLA) are two
Federal control programs that can provide information useful to air toxics control
efforts.  Both RCRA and CERCLA were originally designed as control programs for
hazardous wastes, particularly hazardous wastes that were being disposed  of by land
treatment means (e.g., landfills, deep well injection, etc.).  However, through the
Hazardous and Solid Waste Amendments of 1984  (HSWA) and SARA, both control
programs  have expanded their scope in addressing hazardous  waste management and
disposal.  Toxic air  emissions from hazardous waste incinerators are covered.
Multiple toxics emission points from comprehensive hazardous waste treatment,
storage, and disposal facilities (TSDFs) have been investigated for regulation jointly by
EPA's  air and hazardous waste offices. Regulations limiting volatile organics  releases
have been proposed (52 FR 3748) for TSDFs.  Potential toxic air emissions occurring
during site clean-up operations are being estimated and potential risks determined
under the CERCLA Superfund program.  A significant amount of information may be
available from these programs that could aid S/L  agencies in the problem definition
and implementation phases of an air toxics  control program.
      One of the primary benefits of  the RCRA/CERCLA programs  is to help
identify what sources handle substances deemed to be toxic.   The RCRA regulations
(40 CFR, Part 261,  Subpart D) contain six categories  used to classify  chemical wastes.
These six are toxic,  acutely hazardous, EP toxic (Extraction procedure-Method 1310),
reactive, corrosive, and ignitable.  Though the lists are dynamic and continually under
review,  the toxic and acutely hazardous lists combined currently include over
600 substances.  Sources handling RCRA-listed wastes must conform with specified
treatment, disposal,  and reporting requirements. The  data produced by these
reporting  requirements would be useful in denoting facilities handling  substances of
concern and flagging potential air toxics  emitters.  Information  may also be provided
that could be used  to calculate emission estimates.
                                      4-31

-------
      Other work done by hazardous waste agencies may provide similar types of
information that would be beneficial to air agencies. Inventories of TSDFs have  been
prepared by EPA and States for various purposes including the development of
regulations for TSDF air emissions.  These inventories  can provide much useful
information for problem definition purposes.  Similarly, permit information, namely
RCRA Part B permits, can provide useful data on what industrial sources (not just
TSDFs) handle RCRA-listed wastes, and  therefore may be sources of air toxics.   In
most cases, Part B permits would provide sufficient data to calculate toxics emission
estimates.
      The Superfund National Priorities  list (NPL) program can also provide a data
on of air toxics emissions  from  hazardous waste clean-up sites.  The  provides Hazards
Ranking  System (HRS) used for NPL listing data on sites that may present significant
risks until clean up.  It does not provide  much data on currently active TSDFs as air
toxics sources.  Through the HRS, Superfund sites are  evaluated and ranked according
to their overall health risk potential.  Several hundred  chemicals are listed by
CERCLA (40 CFR Part 302) as being hazardous  and are  evaluated by the system.
The overall evaluation includes  multimedia exposure routes and exposures from the
site during clean-up operations.   The rankings are used to prioritize clean up initiation
among sites nationwide.   The HRS provides a substantial amount of information that
effectively defines the extent of the health risk at a particular site.  This information
can be useful to a S/L agency in defining regulatory and control strategies for all
such sites in their area.
      In terms of program implementation, routine source reporting  requirements
under RCRA may have  some benefits and use for tracking and monitoring air toxics
emissions and compliance  with air toxics  regulations. It is possible the RCRA reports
could be used as one means of enforcement.   An air agency may be able  to make
use of the  reported data either  as is or by adding minimal additional questions to the
routine reports, and accomplish its monitoring/enforcement goals.  As with  any
coordination activity, extensive communication and initial planning will be necessary to
gauge feasibility and ensure  smooth operation.
                                      4-32

-------
      43.3.3 OSHA Regulations

      Regulations of the Occupational Safety and Health Administration (OSHA) can
produce information  beneficial to air toxics control program development The
primary benefits of OSHA are the identification of sources  potentially emitting toxic
compounds, the estimation of emissions rates and related process information, and the
identification of control options for various processes and pollutants.  In some  cases
air toxics control  efforts may be able to piggyback on OSHA regulatory initiatives.
      The primary objective of OSHA programs is worker  protection from unsafe
activities, including exposure to toxic chemicals.  By limiting the amounts of a
chemical that may be released into the workplace, some OSHA regulations may
indirectly benefit  air toxics control.  For example, OSHA places very strict limits on
the amount of toxic  arsenic trioxide (As^ Q) that arsenic acid production workers may
be exposed  to.  As a result, all domestic manufacturers of arsenic acid have installed
sophisticated capture and collection systems for arsenic trioxide dust.   The benefit of
these systems has  been to greatly lower both workplace and ambient air  arsenic
trioxide releases.
      Over 500 chemicals are regulated by OSHA in the workplace  (29  CFR,
Part 1910,  Subpart 2).  As a part of  its Hazards Communication System, OSHA
requires facilities to  make available to  employees and others, Material  Safety Data
Sheets (MSDS) for the chemicals present in the workplace  environment.  An
assessment of facility MSDSs is a quick and easy  way to determine what chemicals
are present at a plant  and which have  potential for emission  into ambient air. The
OSHA also conducts special studies of particular processes/source categories to
estimate chemical release rates and possible control measures. For example, studies
have been conducted to estimate hexavalent chromium releases from electroplating
processes  and possible  control options.  Studies such as this and  others, by groups
such as the National Institute for Occupational Safety and Health (NIOSH), may be
able to provide S/L air agencies with insightful information for problem  definition and
control strategy development phases of air  toxics programs.
                                       4-33

-------
      4.3.3.4 Clean Water Act Regulations

      The principal value of coordination with the Clean Water Act (CWA) program
is the potential identification and association of toxic chemicals  of concern with
specific  plants and general industrial processes.  Such associations may be indicative of
potential air toxics emissions sources.  The CWA program can also indicate what
facilities have wastewater streams containing  toxic  chemicals of concern.  Some  of
these chemicals may be ones that are readily volatilized and present air  toxics   .
problems.  Sufficient CWA information may  be available on toxic chemical loadings
and  discharge volumes to  quantify air emissions.
      The utility of CWA program information on air toxics control program
development  is probably much less  than SARA Title HI or RCRA, but the  potential
information overlap and coordination opportunities should  not be overlooked.

4.4 MEASURING PROGRAM EFFECTIVENESS

      The purpose of this subsection is to describe  various techniques for measuring
air toxics control'program performance.  The goal of this  subsection is to present
general  criteria for  the  design of an effective program performance measurement
system and to provide some example measurement techniques.

      4.4.1  Need for an  Air Toxics Program Performance Measurement System

      Like criteria  pollutant control programs,  air toxics control programs have a
need to assess their overall progress and effectiveness relative to specific program
goals.  For reasons  of internal agency accountability, accountability to the legislature
or local board/commission, and  accountability to the general public,  S/L agencies are
commonly required  to document the progress and  effectiveness of their air pollution
control programs.  In addition, agencies are required to document program activities
to qualify for grant funds from EPA (e.g., Clean Air Act Section 105 grants).
                                       4-34

-------
Program effectiveness measures also need to be considered right from the start during
program design, and not as an afterthought
      Progress toward achievement of program goals can be categorized as
administrative progress and control effectiveness. Administrative progress includes
development of regulations, policy, or guidelines, the number or percent of sources
inventoried, the number  of multiyear development plan (MYDP) milestones  met,  the
number of hiring and resource acquisition objectives achieved, or the number/percent
of source reviews completed.
      Control  effectiveness refers to the results or "environmental payoff of program
implementation.  This may include such items as the reduction in complaints, actual
emission reductions, the  amount of reduction in ambient concentrations  of air  toxics,
or the reduction in public  health risks.
         Care needs to be  taken  to ensure that the measure being used  to evaluate
effectiveness is consistent with the goal being measured.  For example, the
effectiveness of a program that has a goal of developing and adopting an air toxics
regulation is best measured through administrative progress and not control
effectiveness.  Obviously, since in this case the regulation has only been adopted  and
not yet implemented, one would  expect little progress  in control effectiveness
measures.  Measuring the  effectiveness of this program through a control effectiveness
measure (e.g., risk  reduction)  would indicate that the program was  not effective even
though  it was successful  in meeting the stated goal (i.e., adopting a  regulation).
      In developing a system for measuring the performance of an  air toxics program,
a S/L agency can and should draw upon the considerable  experience it  has  acquired
in tracking criteria  air pollutant control programs.  However, because  of the human
health ramifications of air  toxics  program, the measures of administrative progress and
control effectiveness and the implications  of the lack of progress and control are
unique  for air  toxics programs in several respects.  First, the pollutants involved are
different and more numerous than in criteria programs. Second, programs that
include a goal of reducing public health risks to a specified level may need  to
measure decreases  in exposure and risk levels.   Third, programs including acceptable
ambient guidelines  or standards need to include measures of progress in reducing air
                                       4-35

-------
toxics concentrations below those levels.  Finally, S/L agencies receiving EPA grant
funds for air toxics program development are required to report their progress
according to special EPA program oversight procedures (ie., MYDP  milestones).
      The  ultimate goal of a performance measurement system is to measure
program results.  Needless to say, it is possible to expend a great deal  of time,
energy,  and resources on an air toxics program with few or no measurable results.
Therefore,  S/L agencies and EPA need to be able to determine if any real progress
in regulating air toxics is occurring  as a result of S/L agency  air toxics  program
activities.
      Measurement of  environmental payoff is a difficult (and expensive) aspect of an
air toxics program to measure.   For example, measurement of reductions in air toxics
emissions actually achieved in  practice can be either speculative or expensive to
determine.   Consequently, tracking systems designed to measure program performance
may  focus  initially on administrative progress or on more qualitative  measures
performance, rather than on control effectiveness measures. Accordingly, a primary
characteristic of systems to measure the performance of air toxics  programs is a
graduated approach to measuring administrative progress and  control effectiveness.
This graduated approach to tracking program performance is based on  measures that
range from the easiest to measure and most qualitative hi nature to the most difficult
and  quantitative.  As a  program is implemented and matures, an agency should be
able  to  measure program performance through  use of increasingly more quantitative
and  results-oriented indices.

      4.4.2 Performance Criteria

      Design of performance criteria should  consider the following:

             (1)    Measure progress/effectiveness relative to program  goals.
                   Progress/effectiveness measurement is a function  of the primary
                   program goals and the basic design of the program. As such,
                   program performance measurement should provide  information
                   related to  the achievement  of program goals.  As such,  program
                   performance measurement should provide information related to

                                       4-36

-------
      the achievement of program goals. For example, some programs
      will want to show quantitative reductions in emissions to specified
      levels.  The system should also be flexible enough to
      accommodate new goals and changes in program design that occur
      over time.

(2)    Measure progress/effectiveness both qualitatively and
      Q\13fltit3flv?]y-  Both  qualitative and quantitative measures  are
      necessary to fully evaluate a program's performance relative to its
      goals.  Qualitative measures might include, for example, an
      evaluation of the program's overall progress in meeting program
      goals.  Quantitative measures might include tracking  the number
      of sources reviewed for air toxics each year, the number of
      sources inventoried, or reductions in the maximum individual risk
      (MIR) or incidence of cancer near high risk point sources
      (HRPS).

(3)    Focus  on the four National Air Toxics Strategy themes.
      The four themes of EPA's  1985  National Air  Toxics
      Strategy provide a good focus  for a performance
      measurement system.  The  themes include:  (1) acceptance
      of NESHAP delegation, (2)addressing HRPS,  (3) addressing
      high risk urban areas, and (4) enhancing S/L agency
      capability.  Thus, for example, the system should measure
      progress in reducing  emissions and risks associated with
      HRPS or the progress in accepting delegation of authority
      for enforcing NESHAP regulations.

(4)    Use available sources of data. Several sources of information are
      available that may be useful in designing an effective program
      performance measurement system. These include NATTCH, the
      1989 STAPPA/ALAPCO survey  of air toxics programs, and
      various EPA and State monitoring programs such as  the Non-
      methane Organic Compound (NMOC) Monitoring Program.
      These  and  other sources of information can provide useful
      information that can be used as  baseline measures.

(5)    Include measures of  agency capabilities to compare with oiher
      agencies.  To allow an assessment of  the investment  in air toxics
      control over time and relative to the magnitude of the problem,
      measures of agency capabilities should be incorporated into any
      program tracking system.  Such information can be used by S/L
      agencies for direct comparison to the  investment levels of agencies
      in other States  with comparable  air quality programs, and  by State
      agencies for comparison of investments in local agencies in their
      jurisdiction.
                            4*
                           -

-------
             (6)    Provide for flexibility and feedback.  The performance tracking
                   system should be flexible enough to accommodate new sources of
                   information, new data collection techniques, and new goals.  In
                   addition, the system should provide feedback from the evaluation
                   of program performance so that modifications in program goals or
                   implementation can be made.
      4.4.3  Example Measurement Techniques

      No one measure  of program progress can provide an accurate and
comprehensive indication of results.  Consequently, a combination  or series of
measures is  necessary to comprise an effective performance  tracking system.
      Table 4-3 is a checklist of possible program progress  and effectiveness
measures. The checklist is grouped according to administrative progress (or program
activity-oriented) measures and control effectiveness measures.  Several  of the
measures expressed as "number of could be expressed as percentages.   The list is not
intended to be exhaustive, but may serve to illustrate  a number of program
performance measures.                       __
                                       4-38

-------
      TABLE 4-3. EXAMPLE PROGRESS MEASUREMENT TECHNIQUES
Administrative Progress Measures
1.     Number of MYDP milestones achieved.
2.     Number of sources inventoried/surveyed.
3.     Number of sources reviewed for air toxics.
4.     Number of sources modeled (screening and/or detailed).
5.     Achievement of major program goals  (e.g., completion of draft regulations,
      approval of  regulations, hiring of budgeted personnel, etc.)
Control Effectiveness Measures
1.    Number of sources meeting AAL, target risk level, or control technology
      requirements.
2.    Number of sources installing controls or making process changes for air toxics
      beyond what would otherwise~be required for criteria pollutants.
3.    Number of listed pollutants for which all sources are in compliance.
4.    Number of HRPS in compliance.
5.    Measurable risk  reductions for HRPS (MIR and/or incidence).
6,    Reductions in  the volume of SARA Title HI toxic  air releases.
7.    Improvements  in monitored ambient conditions in high risk urban areas, (not
      otherwise  due  to industry shutdowns,curtailments, or reductions  hi mobile
      sources activity)
8.    Achievement of  major program goals related to  results.
9.    Number of air toxics-related complaints reviewed.
                                      4-39

-------

-------
                                    TECHNICAL REPORT DA-TA
                             (Please read Instructions on the reverse before completing)
  REPORT NO.
   EPA-450/2-90-012
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
   Designing and Implementing an Air Toxics  Control
   Program:  A Program Development Manual for  State and
   Local Agencies	
7. AUTHOR(S)
5. REPORT DATE
          JUILI  1990
6. PERFORMING ORGANIZATION COOE

        OAQPS
                                                            8. PERFORMING ORGANIZATION REPORT NO
   Martha H. Keating and  Michael A. Trutna  (U.S.  EPA)
 I. PERFORMING ORGANIZATION NAME AND ADDRESS
   Radian  Corporation
   P.O.  Box 13000
   Research Triangle ?ark,NC  27711
                                                             10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
 12. SPONSORING AGENCY NAME ANO ADDRESS
   Noncriteria Pollutant Programs  Branch
   Air  Quality Management Division
   Office  of  Air Quality Planning  and Standards
13. TYPE OF REPORT ANO PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
      This  manual is intended  to  assist State and  Local air pollution  control agencies
   in making informed decisions  about the development  of air toxics  control programs
   and to  enhance consistency among the State and local program approaches.  In
   particular,  insight is provided into how to make critical decisions  regarding
   program scope and stringency, evaluation of program impacts, and  achievement of
   overall program goals.  The manual addresses these  issues by illustrating the
   experiences  of State and local  agencies and EPA  in  making air toxics  program decision:
   The manual was developed by  seeking the advice and  experiences of  State  and local
   agencies  in  various stages of their own program  development.  While  this information
   does not  represent EPA policy,  it should prove useful to many State  and  local
   agencies  now actively engaged in air toxics program development.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                               b.IDENTIFIERS/OPEN ENDED TERMS  C. COSATI Field/Group
   Air toxics
   State and  local programs  (air  toxics)
   Program  development (air  toxics)
   Air toxics  strategies
   Control  of  air toxics
18. DISTRIBUTION STATEMENT
                                               19. SECURITY CLASS (Tins Report/
              21. NO. OF PAGES
                  144
                                              ! 20. SECURITY CLASS (This page I
                                                                          22. PRICE
EPA form 2270-1 (R«». 4-77)   PREVIOUS COITION is OBSOLETE

-------