United States
           Environmental Protection
           Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/2-79-001a
OAQPS No. 1.2-120
April 1979
           Air
&EPA     Guidelines Series
           Example Control
           Strategy for Ozone
           Volume 1:
           General Guidance for
           Nonattainment Areas

-------
                           EPA-450/2-79-001a

                           OAQPS No. 1.2-120
Example Control Strategy
            for Ozone
            Volume 1:
    General  Guidance for
    Nonattainment Areas
                  by

          Association of Bay Area Governments
               Hotel Claremont
             Berkeley, California 94705
             Contract No. 68-02-3001
          Andrew T.Creekmore, Project Officer
                Prepared for

        U.S. ENVIRONMENTAL PROTECTION AGENCY
           Office of Air, Noise, and Radiation
        Office of Air Quality Planning and Standards
        Research Triangle Park, North Carolina 27711

                April 1979
                                 EPA [
                                   I',

-------
                                     OAQPS GUIDELINE SERIES

The guideline series of reports is being issued by the Office of Air Quality Planning and Standards (OAQPS) to
provide information to state and local air pollution control agencies; for example, to provide guidance on the
acquisition and processing of air quality data and on the planning and analysis requisite for the maintenance of
air quality. Reports published in this series will be available-as supplies perm it-from the Library Services Off ice
(MD35), U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711; or, fora nominal
fee, from the National Technical Information Service, 5285 Port  Royal Road, Springfield, Virginia 22161.
                                 Publication No. EPA-450/2-79-001a


                                        OAQPS No.  1.2-120
                                                ii

-------
                              TABLE OF CONTENTS

CHAPTER    TITLE                                                       PAGE
           LIST OF TABLES                                              v
           LIST OF FIGURES                                             vi
           GLOSSARY                                                    vii
  1.       INTRODUCTION                                                1-1
  2.       SPECIAL PROBLEMS OF PHOTOCHEMICAL OXIDANT CONTROL           2-1
                Dynamics of Oxidant Formation                          2-1
                Characterization of Sources                            2-2
                Multiple Scales of Effect                              2-3
                Institutional/Regulatory Implications                  2-3
                References                                             2-5
  3.       INTERGOVERNMENTAL COORDINATION AND PUBLIC INVOLVEMENT       3-1
           IN THE PLANNING PROCESS
                Requirements of the Clean Air Act                      3-1
                State and Local Consultation Process                   3-2
                Determination of Responsibilities
                Local Lead Agency Designation                          3-4
                Intergovernmental Coordination                         3-5
                Public Participation                                   3-6
                References                                             3-9
  4.       GENERAL TECHNICAL APPROACHES                                4-1
                Various Levels of Analysis and the Corresponding       4-2
                Analysis Elements
                Level 1 Analysis                                       4-2
                Level 2 Analysis                                       4-6
                Level 3 Analysis                                       4-10
                Summary of Various Approaches                          4-12

-------
                                                                     PAGE
 4.             General  Guidelines for Technical  Analysis              4-15
               References                                            4-20
 5.        DETERMINING BASEYEAR OXIDANT CONCENTRATION                  5-1
               Interpretation of Air Quality and Ozone Standard       5-1
               Determination of Design Value                          5-2
               References                                             5-3
 6.        EMISSION INVENTORY AND PROJECTION                           6-1
               Planning for Emission Inventory/Projection             6-2
               Overview of Emission  Inventory/Projection Process      6-5
               Evaluation of Existing Emission Inventories            6-9
               Determination of the Grid System                       6-10
               Data Collection                                        6-10
               Data Analysis and Emissions Calculations               6-11
               Spatial  Resolution of Emissions                        6-13
               Temporal Resolution of Emissions                       6-17
               Emissions (VOC and NOX) Species Allocation             6-18
               Emission Projections                                   6-18
                  Problems                                            6-19
                  Guiding Principles                   .               6-21
                  Overview of Projection  Approaches                    6-22
              References                                              6-24
7.        OXIDANT  PREDICTION METHODS                                   7-1
              Level 1:  Photochemical Dispersion                     7-1
              Models - Grid or Trajectory Models
                 General Description                                 7-1
                 Input Data Requirements                              7-4
                 Model  Validation                                    7-7

-------
                                                                     PAGE
7.               Baseline Oxidant Predictions and                    7-8
                 Development of Alternative Emission
                 Control Strategies
                 Sensitivity Analysis                                7-8
              Level 2:  EKMA                                         7-9
                 General Description                                 7-9
                 EKMA Using Standard Isopleths                       7-11
                 EKMA Using OZIPP-Generated Isopleths                7-12
              Level 3:  ROLLBACK                                     7-13
                 Application Procedure                               7-13
                 References                                          7-15
8.       DEVELOPMENT AND ASSESSMENT OF OXIDANT CONTROL STRATEGIES    8-1
              Determination of Requirements for Oxidant Control      8-3
              Inventory of Existing and Currently Scheduled          8-4
              Control Programs
              Development of Candidate Control Measures,             8-5
              Including RACTs
              Screening the Options                                  8-8
              Selection and Definition of Alternative Control        8-9
              Strategies
              Assessment of Alternative Control Strategies           8-10
              Technical Effectiveness                                8-11
              Economic Considerations                                8-11
              Institutional Impacts                                  8-13
              Overall Assessment                                     8-13
         Staff-Recommended Control Plan                              8-14
         References                                                  8-17

-------
                                                                       PAGE
 9.       PLAN REVIEW, ADOPTION AND APPROVAL PROCESS                   9-1
               Providing Adequate Time for Public Review               g_]
               Availability of Review Documents                        9-2
               Demonstration of Reasonable Further Progress            9-2
               Compliance with EPA RACT Measures                       9-5
               Demonstration of Legal, Financial and Manpower          9-6
               Commitment to Implementation
               References                                              9-7
10.       THE CONTINUING PLANNING PROCESS                              10-1
               Tracking Reasonable Further Progress                    10-1
               Work Program Submittals                                 10-4
          APPENDIX A - BIBLIOGRAPHY OF URBAN SYSTEM MODELS             A-l
          APPENDIX B - EPA CONTROL TECHNOLOGY GUIDELINE DOCUMENTS      B-l
                                       IV

-------
                               List of Tables


Table                                                                   PAGE


2-1        Summary of Oxides of Nitrogen (NOX) Control Issues 1n        2-4
           the Bay Area


3-1        Sample Division of Responsibilities for Non-Attainment       3-7
           Planning


4-1        Comparison of the Three Levels of Technical Analysis         4-13


6-1        Comparison of Three Different Levels of Emission             6-3
           Inventory


6-2        Comparison of Various Data Collection Techniques for         6-12
           Baseyear Emission Inventory


6-3        Alternative Methods for Spatial Distribution of Area         6-14
           Source Emissions


6-4        Comparison of Emissions Estimate Methods                     6-16


8-1        Forthcoming CTGs for Stationary VOC Sources                  8-6


8-2        List of Reasonably Available Transportation Control          8-7
           Measures


8-3        Washington Environmental Research Center (WERC) Matrix       8-15


8-4        Format of Decision Mjatrlx Suggested in EPA AQMP Guidelines   8-16

-------
                               List of Figures


Figure
                                                                        PAGE
4-1        A Simplified Flow Diagram for (h Non-attainment              4-4
           Planning Analysis:  Level  1  Analysis  Using  Complex
           Models


4-2a       A Simplified Flow Diagram for Oq Non-attainment              4-8
           Planning Analysis:  Level 3  Analysis  Using  EKMA
           Standard Curves


4-2b       A Simplified Flow Diagram for (h Non-attainment              4-9
           Planning Analysis:  Level 2  Analysis  Using  EKMA
           along with OZIPP Curves


4-3        A Simplified Flow Diagram for (h Non-attainment              4-11
           Planning Analysis:  Level 3  Analysis  Using
           Rollback Model
6-1        Block Diagram for Aggregated  Emission  Inventory/            6-6
           Projection Procedures


6-2        Block Diagram for Disaggregated  Emission  Inventory/         6-7
           Projection Procedures


7-1        Example Ozone Isopleths  Used  in  EKMA                        7-10


8-1        Overview of Procedures for  Developing  Oxidant Control       8-2
                                      VI

-------
GLOSSARY
      The definition of acronyms  and terms used in this  document are
 given below.
 ABAG:
 APCD:
 AQMD:
 AP-42:

 CARB or ARB:
 CFR:
 COG:
 EIS:

 EKMA:
 ELSTAR:

 EPA:
 HC:
 Level:

 LIRAQ:
 MTC:
 MPO:
 NEDS:
 NMHC:
 NOx:
 OPR:
 OZIPP:
Association  of Bay Area Governments
Air Pollution Control District
Air Quality  Management District
EPA document,  Compilation of Air Pollutant  Emission
Factors, including supplements
California Air Resources Board
Code of Federal Regulations
Council  of Governments
Emission Inventory System.  (This term  refers to the
system used  by the U.S. Environmental  Protection  Agency.)
Empirical Kinetic Modeling Approach
Environmental  Lagrangian Simulator of Transport and
Atmospheric  Reactions
U.S. Environmental Protection Agency
hydrocarbons
three levels of sophistication are given for the oxidant
modeling approaches; Levels 1, 2, and 3 are characterized
as complex,  intermediate, and simple, respectively.
Livermore Regional Air Quality Model
Metropolitan Transportation  Commission
Metropolitan Planning Organization
National Emission Data System
Non-methane hydrocarbons
Oxides of Nitrogen
Oxidant  Prediction Relationship
EPA's Kinetics Model  and Ozone Isopleth Plotting  Package
                                 VII

-------
ppm:          parts per million
RACM:         Reasonably available control  measure
RACT:         Reasonably available control  technology
RATCM:        Reasonably available transportation  control  measure
SIP:          State Implementation Plan
UTM:          Universal Transverse Mercator Coordinate System
VMT:          Vehicle miles travelled
VX:          Volatile organic compounds
                                 viii

-------
                            Chapter 1
                   INTRODUCTION
     The Clean Air Act Amendments  of 1977 continue the requirement for
State Implementation Plans (SIP) to  be prepared to  attain  and maintain
national  ambient  air  quality standards.  For a variety of reasons, the
deadlines  for submittal of the SIP's and the  dates by  which  the  plans
are  to  demonstrate meeting the various air quality standards have been
revised.  Wherever possible, SIP's are to  show meeting  all applicable
standards  no  later than  1982.  For  carbon monoxide and  photochemical
oxidants,  under specific conditions, a five-year extension to 1987  is
possible.
     An  important factor in determining what strategies are needed in an
air  quality plan  is the  level  at  which air quality standards  are set.
Recently,  for example,  the Environmental  Protection Agency changed  the
photochemical oxidant 0.08 parts per million (ppm) -  1-hour standard to
an ozone (0  ) 0.12 ppm  -  1-hour standard.   At  the time  the Bay Area
non-attainment plan was adopted locally, the oxidant standard was still
0.08 ppm;  therefore,  the  plan was developed  to meet  this  standard.
Although  the standard  has been changed to 0.12 ppm, the techniques used
to develop the control strategy  for  the  study are still valid.
Currently, the Bay Area  is  re-examining its adopted plan to  see what
changes  are appropriate.  Other non-attainment areas may prepare  plans
for the  new standard.
     These reports have  been prepared to  assist those  involved  in
preparing  non-attainment plans for use in SIP submittals.  Specifically,
the reports deal  with preparing photochemical oxidant or ozone control
                                1-1

-------
strategy  plans.  Depending on where  a particular region 1s  1n  the
development of Its plants),  these guidance materials should  be useful
for the 1982 SIP submittals.
     Volume I provides general  guidance  to non-attainment oxldant areas.
This guidance 1s sufficiently broad In  scope that all  areas experiencing
oxldant problems  should find the report to be useful.  Discussions are
presented on the technical  procedures  for analyzing the oxldant  problem
and alternative control  strategies.  The Intergovernmental coordination
and public Involvement required in the  planning process are similarly
described.  A  systematic  approach to plan development is given.  This
approach acknowledges the widespread differences experienced across the
nation in the  extent  and  severity of oxldant  air  pollution.  Three
different levels  of analysis are proposed depending  on  regional
availability  of data, staff and budgetary resources, and  overall
schedules for plan preparation.  These  three levels of analysis  vary  in
the degree of  sophistication of the models employed for land use and
transportation  simulation, emission Inventories,  and  oxidant
predictions.   The modeling  approaches can be broadly characterized  as
complex,  intermediate  and simple.   The confidence and accuracy of
technical  analysis  should be proportional  to the  degree of
sophistication  of  the models used  and the amount of  effort expended in
modeling.   As conclusions and  recommendations  derived from  the  technical
analysis are often important and costly, emphasis of this guideline is
placed  on  analyses using complex or intermediate models to obtain more
accurate results.  The simple  models,  e.g.,  rollback technique,  are
discussed  only for the purpose of preliminary assessment.
                                1 -2

-------
     Volume II  documents the results of a planning  program to develop an
oxidant  plan  for  the  San Francisco Bay Area during the  period from 1976
to 1978.   This  volume  is a detailed case study of the  planning process,
analysis  procedures  and development of final  plan recommendations.
     Volume II  is  intended primarily  for air  quality planners  and/or
technical  personnel.  As  other non-attainment  areas embark on similar
planning  programs,  the recent experiences of the Bay  Area should  be
instructive.   The  Bay Area  efforts  attempted  to maintain an open and
highly visible  process for developing  the plan.  At the  same time, the
technical  approach and analytical methodologies were as rigorous and
objective as possible  given the staff  and budgetary resources available
to the program.  Throughout  the plan  development as both process and
products  were balanced, lessons were learned for  conducting similar work
in the future.  The  documentation of these  lessons learned In the Bay
Area—what to do and what not to do--is a major purpose of this report.
                                 1-3

-------
                         Chapter 2
           SPECIAL PROBLEMS OF
PHOTOCHEMICAL OXIDANT  CONTROL
     The  special  problems of  oxldant control can be categorized under
 the following headings:
      0  Dynamics of Oxldant Formation
      t  Characterization of Sources
      •  Multiple Scales of Effect
      0  Institutional/Regulatory Implications

 DYNAMICS OF OXIDANT FORMATION
     Oxidants in urban areas result primarily from complex Interactions
 among reactive hydrocarbons and nitrogen oxides in the presence of
 sunlight.  The magnitude and distribution of oxidant in an urban area
 are affected by a number of factors, including:  1)  the amount and
 distribution  (spatial and temporal)  of  emissions of  reactive
 hydrocarbons  and  oxides of nitrogen; 2)  composition of reactive
 hydrocarbons; 3) meteorological and  topographical conditions; and 4)
 other factors, such as background conditions. A reasonable degree of
 technical  understanding of the dynamics  of oxidant formation  is
 considered crucial  to  establishing  the technical  credibility of the
 oxidant non-attainment planning effort.  Detailed discussions of the
 dynamics of oxldant formation may be found in References 2-1, 2-2, and
 2-3.  The  dynamics of oxidant formation can be simulated  by
 sophisticated, "state-of-the-art" photochemical dispersion models,  which
 are very data-intensive  and resource-intensive.  On the other hand,
 simplified oxidant  models are available  which are relatively easy to
                             2-1

-------
apply but  do not consider the complicated photochemical  reactions.   The
level of accuracy desired for oxidant modeling must be evaluated against
the resources  (money, skill, manpower and equipment) available,  overall
schedule for the planning effort,  and other factors.

CHARACTERIZATION OF  SOURCES
     The  characterizing of sources leading to oxidant formation  in  an
urban area has been an area of  research  for  many years.  Generally
speaking, oxidant  concentrations measured  in an urban area may be
attributable  to  the  following sources:
       0 emissions of man-made organic  and nitrogen compounds from
         the  study area
       0 emissions  of  natural  organic  and nitrogen compounds from
         the  study area
       0 oxidant (and/or precursors)   transported  from areas
         outside  the study  area.
     Relative levels of oxidant resulting from each of these three
sources will  significantly affect  the  planning strategy for attainment
and  maintenance of  oxidant  standards.  Unfortunately, detailed
information  on the  relative contributions of various sources to oxidant
formation is  not available for  most  non-attainment areas.  Based on the
most recent  studies of  this  issue, the U.S.  Environmental  Protection
Agency  has  suggested in a draft policy statement that a natural
background concentration of 0.04  ppm oxidant be  assumed for the  purpose
of control  strategy  planning  (2-4).   Further  investigations of the
sources of oxidant  precursors and their chemistry, meteorological
effects,  and  long-range transport of  oxidant  and precursors  are
necessary in order to better  understand and control  the  oxidant
problems.
                                 2-2

-------
MULTIPLE SCALES OF EFFECT
     The multiple geographical  scales of photochemical  oxidant impact
was an issue which surfaced in the oxidant modeling analysis  in  the San
Francisco  Bay Area (2-5).  The LIRAQ model  was used to  test the
sensitivity  of maximum  oxidant levels to changes in hydrocarbon and
nitrogen  oxides  emission  levels.  As an example, Table  2-1 summarizes
what is known and what is suspected  in regard to the effects  of  further
control of NOV emission on oxidant levels in the San Francisco Bay Area.
            A
     The LIRAQ sensitivity analysis  indicates that further NOx  control
will  result in higher oxidant levels within the region  than  would occur
with a "hydrocarbon only" control  strategy.  On the other hand, by not
controlling  nitrogen  oxides, it is suspected that the oxidant problem of
the Bay Area may be  transported to a neighboring airshed (e.g., to
Sacramento,  90 miles to the northeast, or to Monterey, 70 miles to the
south).  The implications to be drawn are that hydrocarbons should be
stringently  controlled and that care should be exercised in deciding the
appropriate  degree of nitrogen oxides emissions control.
     EPA  is continuing studies  of this 03/NOX  relationship and the
control  strategy implications  likely to  result  from additional
information.

INSTITUTIONAL/REGULATORY IMPLICATIONS
     A number  of agencies have direct control over  air pollution
sources.  Many other  agencies affect air quality indirectly through
agency policies.  A non-attainment area plan must be concerned with  both
direct and indirect causes of the  oxidant problem.  In other  words,  both
                                  2-3

-------
     THE IMPACTS OF ADDITIONAL NOX CONTROLS
                                THE  IMPACTS OF  NO ADDITIONAL NOX CONTROLS
              OXIDANT AIR QUALITY
 Within the Bay Area

 LIRAQ analysis indicates
 that  higher levels  of
 oxidant occur with NOX
 controls in the proposed
 comprehensive strategy.
Outside  the Bay Area

It  is  su s p e ct ed  that
downwind  areas where
transport may contribute
to  existing oxidant
problems  would  be
improved.
                                         OXIDANT AIR QUALITY
Within the Bay Area

LIRAQ  analysis indicates
that  lower levels  of
oxidant occur with m>  NOy
controls in the proposed
comprehensive strategy.
Outside the Bay Area

It  is  suspected  that
downwind  areas where
transport may contribute
to  existing problems
would experience worse
oxidant air quality.
       NITROGEN DIOXIDE (N02) AIR QUALITY
Within  the Bay Area

May reduce N0£ violations
jrf appropriate controls
can be identified, e.g.,
stationary vs.  mobile and
ground  level  vs. elevated
emissions.
Outside the Bay Area

Not  of  concern  (noNOx
violations  are recorded
in  neighboring  air
basins).
                                                               NITROGEN DIOXIDE  (N02) AIR QUALITY
Within the Bay Area

NOX   emissions  are
projected  to   be
relatively constant
between  1975-2000.
However,  the relative
contributions from mobile
sources and industry
change  substantially.
N0£  violations  may
decrease as the motor
vehicle NCL  emissions
decrease by  2QQQ.
Outside the Bay Area

Not of concern (no.NOp
violations  are recorded
in  neighboring air5
basins).

-------
control  strategies and policies influencing  air quality should  be dealt

with,  and all agencies  involved directly and indirectly  should be

coordinated and consulted in  the planning process.  The Clean Air Act is

explicit  in  the requirements that each plan provide for the exchange of

Information and identification of  intergovernmental responsibilities as

necessary to develop and implement the plan.  Unless  all  agencies

Involved are  well coordinated, any proposed  oxidant control program may

create  conflicts  between the various  agencies over responsibilities,

jurisdiction, control philosophy, etc.  Institutional coordination and

cooperation 1s an important part of the air  quality planning.   Chapter 3

of this guidance addresses the various aspects of this issue.




REFERENCES

2-1  Jerskey, T.N., and  J.H.  Seinfeld, "Continued Research in  Mesoscale
     Air Pollution Simulation Modeling:  Vol. IV - Examination  of the
     Feasibility  of Modeling  Photochemical  Aerosol  Dynamics,"
     EPA-600/4-76-016d, U.S.  Environmental Protection Agency, Office of
     Research and Development, Research Triangle Park, N.C., May, 1976.

2-2  Roth, P.M., et  al ,  "An Evaluation of Methodologies  for  Assessing
     the Impact of  Oxidant  Control  Strategies,"  prepared  for  the
     American Petroleum Institute, EF76-112R, Systems Applications,  Inc.
     San Rafael, California   August, 1976.

2-3  MacCracken, M.C., et al , "The Liver-more Regional Air Quality Model:
     Volume 1.  Concept and Development," Lawrence Livermore Laboratory,
     University of California, Livermore, California, Report No.  UCRL -
     77475, August, 1977.

2-4  "Uses,  Limitations and  Technical  Basis  of Procedures for
     Quantifying Relationships  Between Photchemical Oxidants  and
     Precursors," EPA 450/2-77-021a, U.S. EPA, Office of Air Quality
     Planning and Standards, Research Triangle Park,  N.C. November,
     1977-

2-5  DeMandel ,  R.E.,  L.H.  Robinson, J.S.C. Fong,  and R.Y.  Wada,
     "Comparisons  of  EPA  Rollback,  Empirical/Kinetic,   and
     Physicochemical  Oxidant  Prediction Relationships in the San
     Francisco Bay Area," J. Air Poll. Control Assoc., 29(4) ,  352
     (1979).


                                 2-5

-------
                       Chapter 3
INTERGOVERNMENT COORDINATION
       AND PUBLIC  INVOLVEMENT
       IN THE PLANNING  PROCESS
    Central  to  the success or  failure of air quality  planning in
non-attainment areas is the intergovernmental coordination and local and
public involvement necessary to begin and carry out the planning
program.  This chapter briefly describes key requirements  of the 1977
Clean  Air Act with respect to intergovernmental coordination and  public
involvement.  It also covers some important organizational aspects of an
air quality planning program in non-attainment regions.  For additional
information, earlier EPA guidance  materials on these topics are also
available (3-1, 3-2).

REQUIREMENTS OF THE CLEAN AIR ACT
    The  Clean Air  Act Amendments of 1977~as did the previous version
of the Act—require  the development of air quality plans in regions
where the national ambient air quality standards are not being achieved.
The 1977 Amendments, however, go much further than previous legislation
in specifying  the content and organizational requirements to preparing
such plans.  For example,  the 1977 Act also  specified  that,  where
possible, the  non-attainment plan is to be prepared by an organization
of elected officials of local governments.  Section  174  further
specified that, where feasible,  the  agency designated to prepare the
revised implementation plan shall be the metropolitan  planning
organization (transportation planning) or the organization responsible
for air quality maintenance planning,  or the organization with both
                            3-1

-------
responsibilities.  The language noted above contains a great deal  of
flexibility  for states and local  governments to determine how such plans
are to be  prepared.

STATE AND  LOCAL CONSULTATION PROCESS
     Section 121 of the 1977 Amendments ensures an orderly process  for
organizing  and carrying out the planning process. The section provides
that each  State shall provide a "satisfactory" process of consultation
with general purpose  local governments  and designated air  quality
planning organizations of local  elected  officials.  The consultation
process covers the  actual  process  of designation of agencies under S.
174 of the Act and the process of developing plans required by the Act.
     A  consultation process  for plan  preparation  is  defined in EPA
regulations  (40 CFR Part 51) as  satisfactory  if  it includes:
       t Information dissemination  and education  of relevant
         organizations and  individuals
       • Opportunity for  involvement of affected  governmental
         organizations and elected officials
       • Opportunity  for joint resolution by affected governmental
         organizations  and Individuals on  key issues (including
         selection  of control  strategies, especially  those
         requiring  local enforcement, implementation or commitment
         of  resources).
     The regulations on intergovernmental consultation also  provide for
coordination of other Federally  assisted regional planning programs,
State  and areawide clearinghouse review, summaries of plan development,
and public and agency participation.

DETERMINATION OF RESPONSIBILITIES
     Although it is readily apparent  from the 1977 amendments that  the
                                 3-2

-------
states  continue to  be responsible  for adoption of the  State
Implementation jjlans, the partnership emphasis of the 1977 Amendments
presents  the states and  local  agencies with a wide  array of choices
about assigning planning tasks.  Among the regulatory requirements  under
S. 121 1s that the governor  of each state discuss the  "alternatives
investigated  for consolidation of  environmental  and other planning
functions."
     In   determining  responsibilities to be divided,  a number of
different approaches can be taken.  The central  features of any planning
process as rigorous as  that required by the Clean Air Act are these:
      §  Technical analysis (including baseline  emission inventory,
         future emission inventory projections,  oxidant prediction,
         modeling)
      0  Control  strategy  analysis (including  assessment of
         effectiveness, costs, and  environmental,  social  and
         economic effects)
      •  Public participation
      •  Plan review and adoption
     The  more complicated the organizational  setting,  the  less readily
apparent will  be the  division of  responsibility.  For example,  in a
state where a branch of  state  government serves as  an  air pollution
control   agency  and has authority over a wide  variety of air pollution
sources,  but where there is no  comparable local  regulatory agency, the
state may play a more significant role in planning than in a state  where
local/regional  agencies are more capable of performing of the necessary
tasks.
     Another  example  of  the decisions to be  made in  the process of
"joint determination" is what pollutants to cover.  For example,  some
states have chosen to interpret various provisions  of the Act to mean
                                 3-3

-------
that they  (the states) have  responsibility  over  total suspended
participates  and sulfur dioxide—with local governments' tasks covering
the transportation-related pollutants (hydrocarbon, nitrogen oxides  and
carbon monoxide).  Other states have determined it appropriate to have
local  agencies deal with all pollutants  where a standard  is  violated.
As  long  as a plan is  prepared to meet  the  requirements of the law,
however,  how such  a plan is prepared will  depend on the desires of both
state  and local officials.  Wherever  a local  agency has demonstrated a
willingness or capability to address all pollutants,  it makes  sense that
they  do  so,  because  a  single, coordinated  planning  and  public
participation process  can be used to examine alternative  controls and
assess  their  costs and benefits.  It also tends  to  present information
in a standardized  fashion to Improve public understanding.

LOCAL LEAD AGENCY  DESIGNATION
     Although the  Act  itself does not use the  phrase  "lead agency" the
concept  of a  lead agency  is  used in  EPA's  S.  121 regulations.  The
responsibility for ensuring satisfactory intergovernmental  consultation
is assigned by the regulations to any  governmental organization that has
the "lead responsibility  for development"  of  ai.y of the  following
Implementation plan elements:
       • Procedures for pre-construction review of direct sources
       •  Transportation-related control  measures
       •  Control  measures other than  transportation-related
         measures.
       •  Measures for prevention of significant deterioration of
         air quality and protection  of visibility
       •  Air quality maintenance measures
                                 3-4

-------
       t Delayed compliance orders  under  S. 113 (d).
     The regulations, therefore, contemplate substantial  involvement in
the planning  process by many agencies.

INTERGOVERNMENTAL COORDINATION
     In many respects  the 1977 amendments and EPA's  implementing
regulations recognize the intergovernmental complexities  of most  of the
nation's  metropolitan areas, and  the varying degrees of organizational
levels in  the different states.
     In most  states there is no single statewide agency  responsible for
both air pollution controls and transportation.  Thus, just as  there is
a  national  interagency  agreement  on air quality/transportation
coordination, there may have to be state-level  agreements between  a
state  transportation agency and the state air pollution  control agency.
Similar agreements might be necessary or  desirable to coordinate air
quality activities with energy facility siting agencies.
     At the metropolitan level, similar coordination activities will  be
required.   For example,  at  least 10 of the states have designated air
quality planning  organizations  that  are not metropolitan planning
organizations in  their  respective  urban areas.  These states include
California, Illinois, Montana, Nevada, New  York, Ohio,  Pennsylvania,
South Carolina, Virginia and Washington.
     Another  example where coordination is needed is for  stationary and
mobile  sources.   In many  states--Cal ifornia  is  one example—mobile
source  controls are the sole responsibility of the state.  Regional  or
local  air quality regulatory agencies may exist independent  of areawide
                                 3-5

-------
 planning organizations.
      In  most areas, sucessful  air quality  planning  will  require
 extensive cooperation among local, regional  and state agencies.  The
 coordination and organizational  cooperation needed will vary from region
 to region, reflecting differing  organizational authority and air quality
 planning  capability.  Table 3-1  depicts  a sample  division of
 responsibilities for non-attainment planning purposes.  It is intended
 only to illustrate the kinds of determinations that should occur during
 the  initial planning period, before the technical analysis begins.

 PUBLIC PARTICIPATION
      Most  Federal  programs, including  those  of EPA, require  public
 participation  as an integral part of planning.  As noted previously  in
 this chapter,  the "intergovernmental  consultation"  requirements
 generally apply to parties interested in  or affected by the air quality
 planning provisions.
      Numerous  techniques  for  public  participation programs ar«
 available; most  are familiar  to organizations and agencies experienced
 with Federal programs.   What techniques  are chosen for the air quality
 planning programs should be an essential  feature of the designated
 agency's overall  public participation efforts.  Goals and objectives of
 the  public participation program should be defined in measureable terms,
 and  the overall  project budget should  adequately support a well-designed
 public  participation program.   A central  feature  of the  public
 participation   program  should  be  evaluation  of its  strengths and
weaknesses  during the planning project so that,  as problems occur, they
can be corrected.
                                3-6

-------
                            Table 3-1

                 Sample Division of Responsibilities
                    for Non-Attainment Planning
       Program Component

Planning Coordination/Program
Management

Provision of Policy Guidance Task
Force
Organization  of
Participation Program
Public
Organization of Joint Technical
Staff

Development of regional land use
and population activity forecasts

Analysis and monitoring of effects
of land uses on air quality

Monitoring and data  analysis of
ambient air quality and related
meteorology

Maintenance of emissions inventory
for modeling purposes, assessment
of emission trends and projection
of  future  emissions   for
non-vehicular sources

Assessment of air quality  trends
and  projections of  future air
quality

Development,  analysis  and
monitoring of stationary source
control measures
   «*
Preparation,  adoption  and
enforcement of stationary source
control measures
Air quality impact analysis
proposed projects
     for
     Responsible Agency

Council of governments (COG)


COG


COG


COG


COG


COG
              Local  air  pollution control
              district (APCD)
              APCD
Development  of  regional
transportation policies and plans
              APCD
              APCD
              APCD
APCD
              Metropolitan planning organization
              for transportation (MPO)
                              3-7

-------
       Program Component                   Responsible Agency

Development of regional travel       MPO
demand forecasts

Development, analysis and       MPO
monitoring  of transportation
control measures

Projection  of emissions  from       MPO
transportation sources

Adoption of Mobile  Source Control       State air pollution agency
Measures

Overall review and coordination of      COG
air quality planning effort (data
analysis,  modeling, control
strategy   development  and
assessment)

Analysis of plan's environmental,      COG
social  and economic impacts

Plan adoption and submittal to Air      COG
Resources Board
                              3-8

-------
REFERENCES
3-1  U.S.  Environmental  Protection Agency, Office of Air and Waste
     Management,  Office  of Air  Quality Planning and Standards,
     "Guidelines for  Air  Quality Maintenance Planning and Analysis,
     Volume  2:   Plan Preparation", Report No. EPA-450/4-74-002, July
     1974.

3-2  U.S.  Environmental  Protection Agency and U.S.  Department  of
     Transportation,  "Section  174 Guidelines", December  1977.

3-3  U.S. Environmental Protection  Agency, "Requirements  for
     Preparation, Adoption and Submittal  of Implementation Plans",  40
     CFR Part 51, Subpart M (Intergovernmental Cooperation), May 1978.
                               3-9

-------
                         Chapter 4
 GENERAL TECHNICAL APPROACHES

     This  chapter discusses planning of  the technical analysis  for
oxldant control  programs.   As  there 1s a great  variation among
non-attainment areas with respect  to the magnitude of the  oxldant
problem and resources available  for  plan  preparation, the  technical
approaches presented are deliberately  general and flexible.  In  order to
be useful to a  variety of non-attainment areas, three different levels
of technical analysis  are  discussed.  These three  levels of  analysis
vary In the degree  of  sophistication between  the  models used, and
involve different levels of available resources (manpower, money, time,
skill and equipment).  The accuracy of the technical  analysis  generally
increases as the level of  effort increases.  However, there are other
factors to be  considered  in selecting the level of  analysis for a
non-attainment area.   The following  section describes these  three
general  technical  approaches which  can  be used in non-attainment
planning and analysis,  The subsequent section summarizes and  compares
various technical approaches. The last section of this  chapter  presents
several  general  guidelines, which if followed, can  help improve  the
technical credibility of  a non-attainment planning analysis.  More
specific guidelines  to each technical  analysis element, e.g., compiling
emissions inventory,  are discussed in  other chapters  of  this guidance.
     It  should be pointed out that the U.S. EPA is currently  developing
a guidance specifying what levels  of  ozone modeling analysis  various
non-attainment areas should undertake for the 1982  SIP's. Five levels
of  analysis are under consideration.  These include:
                             4-1

-------
          Level  1 - Photochemical Grid Models
          Level  2 - Photochemical Trajectory Models
                    (Including  Intensive City-Specific EKMA)
          Level  3 - Application of Basic City-Specific EKMA
          Level  4 - Application of Standard EKMA,  and
          Level  5 - Application of Rollback Model.
     The U.S. EPA is also considering various criteria which should be
 used in selecting an appropriate level of analysis.

 VARIOUS LEVELS OF ANALYSIS AND  THE CORRESPONDING ANALYSIS ELEMENTS
     An  overview of the three  different technical approaches to oxidant
 planning and analysis are given in this section.  For each technical
 approach, a simplified  flow diagram  for major analysis  tasks is
 presented.  These flow diagrams illustrate  where to start a technical
 analysis for oxidants, what tasks  are  involved and the interrelationship
 and time sequence of those tasks.  The major  analysis elements  involved
 in each level  of analysis are also briefly  discussed.  The information
 provided is intended to give  readers a clear  understanding of  the
 analysis process involved.  They will be useful to the project manager
 in developing  project organization  and  in preparing a detailed
management plan  for conducting  the technical analysis.

LEVEL 1 ANALYSIS
     The  Level  1 analysis  involves  the  most detailed techniques
considered to be  available.   It  employs a  grid-based photochemical
dispersion  model  such as LIRAQ (4-1)  or SAI  (4-2), or a trajectory model
such as ELSTAR (4-3), as the primary  tool for  oxidant control analysis.
                                 4-2

-------
Because of the  extensive input data required by the  oxidant prediction
model, high levels of effort (manpower, time,  skill, etc.) are needed to
carry out this  level of analysis.  This analysis may  provide the most
accurate answers to a number  of  non-attainment planning  questions.
Figure  4-1 is  a simplified  flow  diagram illustrating the  analysis
process and specific tasks  involved.   As noted, there are five major
tasks involved  in one analysis cycle.  The initial cycle  begins with
land  use, economic, demographic and transportation analyses in order to
determine urban activity  levels.   Then the emission  inventories and
projections are made based  on regional  activity levels and emissions
data.  As required by the photochemical dispersion model,  spatial  and
temporal  resolution of emissions  and  species allocation of  organic
(and/or NOx) emissions  are conducted.   The  next  task  is to predict
oxidant concentrations  using a  selected photochemical  model.   The
additional inputs required (e.g., meteorological and topographical data,
and  initial and boundary  emission  concentrations)  are prepared.  By
comparing predicted oxidant concentrations with the oxidant standard,
the emission reductions necessary to attain the standard are determined.
     At the end of the initial  analysis cycle,  candidate oxidant control
strategies are developed.  The effectiveness  of emission reductions of
candidate control  strategies  are assessed.   Then,  a  new  set of
regionwide gridded emission data is developed by  considering the
emission reductions resulting from the selected control  strategies.
With  the  new set of emission  data, a second analysis cycle is initiated
to analyze the  effect of the selected control  strategies  on regional
oxidant concentrations.  This cycle  is  repeated to evaluate various
oxidant control  packages until an appropriate  control  package is
                                4-3

-------
                            Figure 4-1    A Simplified Flow Diagram for O3
                                          Non-attainment  Planning  Analysis:
                                          Level 1 Analysis Using  Complex Models*
                  1.1 Detailed land use,
                  employment, and
                  population analysis
                  (complex urban
                  system models)
i
-P.
2. Emission inventory/
projection (using
disaggregated
emission models to
generate regionwide
gridded emission data)
                             1.2 Detailed travel
                             demand and
                             network modeling
3. Oxidant modeling
(using a grid or a
trajectory model)
i
> i

fc/

                              Meteoro
                            /logical and
                             topographic/
                               data"
 ' compare
  withO,
 . standard y
Meet standard
4. Determination
of required HC
and/or NOx
emissions
reduction
    Baseline
    emission
    projections
                                                                               5. Development of
                                                                               alternative HC
                                                                               and/or NOx
                                                                               emissions control
                                                                               strategies
                                                               Recommended
                                                               oxidant control
                                                               program
                 •This is the level of analysis used In the non-attainment planning analysis In the San Francisco Bay Area.

-------
selected.

     The five major  tasks included in one technical  analysis cycle are

summarized  in terms of  their purposes and products  in the following

paragraphs.   More  detailed guidelines for  each  specific task are

discussed  in Chapters 5, 6, 7 and 8 of this guidance.


Task 1.1:   Land Use, Economic and Demographic  Analysis

           The  purpose of  this task  is  to  estimate present  and
           future levels of urban activity for traffic analysis  and
           emission inventory preparation. This  task is generally
           conducted by regional and local planning agencies.  To
           project regional  land  use,  employment, population  and
           housing, a number  of models may be used (EAAM,  PLUM,
           USM,  ALDM, LUAM, LUPDM, NBER-USM,  ITL4SMP, BEMOD, etc.,
           see  Appendix A).   Since a  disaggregated  emission
           inventory is required  by  the photochemical model,  the
           result   of this task is  a   set  of disaggregated
           employment, population, and land  use  data suitable  for
           use in preparing the emission inventory.

Task 1.2;   Travel Demand and Network Analysis

           The  purpose of  this task is to generate the travel data
           necessary  for   projecting transportation-related
           emissions.  Using the outputs of  Task  1.1, and present
           and  future transportation   networks  and travel
           assumptions, this task generates the traffic volumes  and
           average speed for each segment in the regional network.
           Trip generation  data are also generated in this task.   A
           number of common travel demand models (trip generation
           model,  trip distribution model,  modal choice  model,
           network assignment model, etc.),  may be used for the
           analysis.   (See Appendix A.)  In areas where there is  a
           high level  of air  travel activity,  air travel  may be
           considered separately from ground travel.

Task 2:     Emissions Inventory/Projections

           Based  on the urban activity data generated in Tasks  1.1
           and 1.2, this task projects emissions inventories for
           the  baseyear and  selected  future years.  The level of
           detail  of the emissions inventory must be consistent
           with  the requirement of the  photochemical model to be
           used.  At this level of analysis, a grid model or a less
           complex trajectory model may be  used.  A grid model
           requires a spatial  and temporal resolution of emissions
           inventory, while  a trajectory model  considers only  the
                                4-5

-------
          pollutants emitted  into the  air parcel  along  the
          trajectory path.  Both types of model require that the
          emissions be broken down into certain  species classes.

Task 3:    Prediction of Peak Oxidant Concentrations

          This task is to  estimate peak oxidant levels  in the
          non-attainment area  in order to calculate the  required
           emission  reductions.  This task includes  preparation of
           inputs,  execution of  computer modeling,  and
           interpretation  of modeling results.   In  addition to
           emissions  data, a grid or trajectory model  will  require
           several  types of  inputs.   This  may include:

             • Meteorological  data - wind speed, wind direction,
               mixing height, etc.

             t Sun  radiation data - Light  intensity, etc.

             0 Topographical data, and/or

             • Initial and  boundary ozone concentrations  (or
               natural  and transported ozone concentrations)

           Detailed guidance to  photochemical modeling is presented
           in  Chapter 7  of this document.   The peak oxidant
           concentrations  determined from  computer modeling are
           used  as a basis for  calculating emissions reductions
           required and developing oxidant  control  strategies in
           the following tasks.

 Task 4;    Determination of  Emissions Reductions

           By comparing  the estimated peak  oxidant level to the
           oxidant  standard, the emission reductions  required to
           attain  air  quality  standards can be calculated.  The
           required reductions establish the goal for developing
           various  oxidant control strategies.

 Task 5:    Development  of Oxidant Control  Strategies

           Alternative  oxidant  control strategies are developed and
           assessed in  this task.  The developed  control  strategies
           should  demonstrate  that the oxidant  standard  will be
           attained as  expeditiously  as practicable, but  no later
           than  1987.
LEVEL 2 ANALYSIS

     This  level  of  analysis employs a less complex model, i.e.,  EPA's
                                4-6

-------
Empirical  Kinetic Modeling  Approach (EKMA) (4-4),  to  estimate the
required oxidant  controls.  There  are two types  of EKMA which should be
distinguished.  The standard EKMA uses a standard set of ozone isopleths
to  calculate the required  ozone  control (4-4).  Another EKMA uses the
Kinetics Model  and Ozone Isopleth Plotting Package (OZIPP) to calculate
maximum one-hour average ozone concentrations for the non-attainment
area  (4-5).   The OZIPP is  a  "trajectory"  photochemical  model,  and
requires a number of city-specific  inputs,  including  initial precursor
concentrations, light intensity, dilution, diurnal  and  spatial emission
patterns,  transported ozone concentrations  and reactivity of the
precursor mix.  The results of OZIPP  simulations  are used to produce an
ozone isopleth diagram.  This diagram is then used in  EKMA procedure  to
calculate the necessary ozone controls.
     Figure  4-2a  shows a simplified flow diagram for  the standard  EKMA
procedure.  This procedure uses a  standard set  of ozone  isopleths to
determine  the HC and/or NOX  emission reduction required to attain the
ozone standard.  The  standard ozone isopleths  were developed by EPA
based on fixed assumptions  of sunlight intensity,  atmospheric dilution
rate, reactivity and diurnal  emission patterns.   For  this reason the
standard EKMA has relatively limited  flexibility in application and
cannot  treat city-specific factors  such as emission  patterns or
meteorological conditions.   Nevertheless, transported  ozone
concentration can  be crudely  treated by this method using a modified
procedure.
     Figure 4-2b shows a flow diagram for EKMA using OZIPP-  The initial
task is to collect data for baseyear emissions inventory and to estimate
future land use, economic,  demographic data and  future transportation
                                 4-7

-------
        Figure 4-2a A Simplified Flow Diagram for Oa
                    Non-attainment  Planning Analysis:
                    Level 2 Analysis Using  EKMA Standard Curves
  Determination of
  Design value for the Base Year
  and the corresponding
  6-9 am HC and NOx
  levels
00
    Establish annual total
    non-methane HC and
    NOx emissions for the
    base year (aggregated)
Use standard EKMA
curves to determine
required HC and/or
NOx reductions
 Adjustment for
 Transported Ozone
 Concentrations
Land use, employment,
popujation and
housing analysis
Development of
alternative HC/NOx
emission control
strategies
Recommended
oxidant control
program
Future year projected
annual (aggregated)
non-methane HC and
NOx emissions

-------
  Figure 4-2b A Simplified Flow Diagram for O3  Non-attainment
            Planning Analysis: Level 2 Analysis Using
            EKMA along with OZIPP
-pi

VO
Simulation day
selection; Daily
HC and NOx
emissions
i
^**~
\ 4
•^»^ ^~"
»-

Using OZIPP to
Simulate max. 1-hr
Ozone Concentrations
and to produce an
ozone isopleth
diagram.
I t
>"^ ^^
i i
m^. ' ^f
»-

Using the generated
ozone isopleths in
EKMA procedure to
calculate the required
HC and/or NOx controls
t
^^
•+

t
L
Development of
alternative HC/NOx,
emission
control strategies


Recommended
oxidant
control
program

                       'Mixing
                       height
                        light
                       intensity
  Base year emissions
  inventory
Land use, employment,
population and
housing analysis
Future year HC
and NOx emissions

-------
activity  for emission  projections.  As  required by the OZIPP,  the
diurnal  and spatial  distribution pattern  of emissions are estimated.
Other input data required by  OZIPP  are also prepared;  this includes
natural and transported  ozone  concentrations, light intensity, dilution,
reactivity of the precursor mix, and others.  The  next task is to  use
OZIPP to simulate  maximum 1-hour  average ozone  concentrations in the
non-attainment area.  The  results of OZIPP simulations  are subsequently
used  to produce an ozone isopleth diagram.  This  diagram  is then used in
the EKMA procedure to calculate the necessary HC  and/or NOX reductions.

LEVEL 3 ANALYSIS
     This level of analysis uses  a  simple  technique, i.e., rollback
technique  (4-4), to  estimate  the necessary  emissions reductions.
Rollback requires even less data than EKMA used in  the  Level 2 analysis.
A flow diagram for Level  3 analysis is shown in  Figure 4-3.  This level
of  analysis may also be  a  possible alternative for  estimating bounds on
control requirements for  hydrocarbon emissions.  The rollback technique
is  based on the assumption that a fractional  reduction in hydrocarbon
emissions will  result  in an  identical reduction in the maximum oxidant
concentration; it does not consider  the effect  of nitrogen oxides on
oxidant formation.   Thus the application of the  rollback technique Is
limited to a first order approximation, and 1s viewed  by  many to be of
questionable validity.
     Using the rollback  method, the HC  emission reduction necessary to
achieve the oxidant  standard can be calculated using  the  following
equations:
                                4-10

-------
Figure 4-3  A Simplified Flow Diagram for Oa
        Non-attainment Planning Analysis:
        Level 3 Analysis Using Rollback Model
Determination of
Design value for
the base year


Establish annual
total non-methane
HC emissions for
the base year


Use rollback model
to determine percent
HC reduction to
meet standard
i

Estimate transported
ozone concentrations
and the related
additivity factor


i

i
i

Development of
alternative HC
emission control
strategies

Future year
projected annual
non-methane
HC emissions

-------
               R    =      (03)adj  - X
                            (03)adj

               (03)adj  =  (03)design - A(To)
where
      (03)adj     = the adjusted  ozone concentration  to be used  in
                   estimating  control requirements;
      (03) design = the 03 design value for the base period;
      X           = a specified  air quality goal, usually the ozone
                   standard;
      A           = the additivity factor for transported ozone;
      To          = the concentration of transported ozone estimated
                   for the base  period;
      R           = the emission  reduction required.
 SUMMARY OF VARIOUS APPROACHES
      In  a  non-attainment  planning effort, technical  analysis  of  oxidant
 concentrations and estimated emissions reduction requirements serve as
 the basis for formulating needed oxidant  control strategies.   The
 adopted oxidant control  program  usually involves costly or controversial
 emission control  measures.  Thus, selection of an appropriate level of
technical  analysis is  important.  Table 4-1  summarizes the advantages
and disadvantages of the above-mentioned three  levels of technical
analysis.  The  major tasks involved in each  of the three  levels of
analysis  are  also  compared  in the table.
                                4-12

-------
                                                               Tiblt 4-1.   Comparison of thi Thrte Levels «f Technical Analysis*
Level of
Analysis
Land use. employment.
population and trans-
portation analysis
Emission
projection/Inventory
Oxldant
prediction Mthods
Oeter*1nat1on of
emission reductions;
Development of
control strategy
Advantages of
OPRUsed
Disadvantages of
On Used
        e Use computer models
          to estimate  size and
          distribution of urban
          activities.
o Use computer models
  to generate special
  and temporal distribu-
  tion of HC and NO
  emissions.  HC enfl/or
  NO, emissions tarn to
  be divided Into species
  classes.
e Use a grid-based or a
  trajectory photo-
  chemical dispersion
  model to estimate re-
  glonwlde grldded oxidant
        itratlons.
                   complex
          models to estimate
          levels of urban
          activities at county
          or county-equiva-
          lent  level.
e Use manual or computer
  models to estimate
  aggregated HC a NO,,
  emissions; relative
  distribution patterns
  of emission on
  simulation day are
  required.
                                                    o Consider a variety of
                                                      HC and  NO. emission control
                                                      strategies (e.g., spatially
                                                      or temporally variable
                                                      controls).
                                                            e Consider complex
                                                              topography, meteorology,
                                                              source distribution, and
                                                              tnuuimrhnl uione  '
                                                              concentrations.
                          e Use EPA's EMU to relate e Consider total HC and/or
                            oxidant levels to f         —    -   -
                            cursors concentrat
                                                     e Treat  limited HC and/or
                                                       NO, control measures.
                    pre-     NO  emission reductions.
                    lions.      *
e EKHA cannot predict ab-
  solute ambient oxidant
  concentrations; OZIPP
  may be used to estimate
  amb1ent_ozone concen-
  trations "

o Transported ozone can
  be crudely treated
                                                              e  The models have the highest
                                                                 potential  for validation
                                                                 Kith observed oildint
                                                                 concentrations.


                                                              e  Can provide estimates of
                                                                 magnitude. Ijcatlon and  time
                                                                 period of  the peak  oxidant
                                                                 concentretions•

                                                              e  Consider composition of  HC
                                                                 and treat  NO. emissions;
                                                                 treat spatial and tem-
                                                                 poral  distributions and
                                                                 transport  of  constituents;
                                                                 consider background con-
                                                                 centrations

                                                              e  Capable of evaluating •
                                                                 wide range of oxidant con-
                                                                 trol  strategies such as
                                                                 emission reduction  at
                                                                 specific sources, emis-
                                                                 sion controls at certain'
                                                                 locations  or  time periods,
                                                                 HC  emissions  control  com-
                                                                 bined with NO, emission
                                                                 control, etc.
                                                              e Less extensive data re-
                                                                quirements

                                                              e Relatively easy to prepare
                                                                the Inputs and apply the
                                                                model

                                                              e Can evaluate certain types
                                                               . of oxidant. control .
                                                                strategies

                                                              e Consider the effect of both
                                                                HC and  NO,, and 03 levels;
                                                                allow for consideration of
                                                                background concentrations
o Data requirements are very ex-
  tensive; great amount of man-
  power, time and money re-
  quired to prepare the Inputs.

• Expensive to execute the model
  (cost Is on the order of 1500
  per run); computer with large
  storage capacity Is required

e Require  experienced pertcnmel
  to perform the modeling;
  difficult to Interpret the
  model's outputs
                                                                                                 o Model
                                                                                                              difficult to validate.
  Analysis agrees with smog
  chamber experiments; accuracy
  In atmosphere may not be good
  because of gross simplification
  of meteorology.
                                                                                                                                                              e Relatively limited application
                                                                                                                                                                for oxidant control evaluation

                                                                                                                                                              o Is not feasible to apply  the
                                                                                                                                                                Em  In areas where the observed
                                                                                                                                                                       ratio Is low    (<2:1)
        e Use simple tech-
          niques  to estimate
          levels  of activities
          at  county or county-
          equivalent level.
   for the study area are
   needed.
e Aggregated HC emissions e Cannot predict oxidant   o Consider total  HC
  	              concentrations.            emission reductions

                          e Use modified rollback      only-
                            technique to determine
                            emission reductions required.
                          e Transported ozone can be
                            crudely treated.
Technical analysis Includes 1) analysis ef urban activities, 2) emission Inventory and projection,  3)  oxidant  prediction
and 4) determination of oxtdant control requirements.   The results of technical  analysis provide a  basis  for developing
alternative control strategies.
                                                                                                                           e Easy to apply
                                                                                                 e  Does not consider NO, emissions;
                                                                                                    does not treat spatial or tem-
                                                                                                    poral distribution of HC  emis-
                                                                                                    sions

                                                                                                 e  Model Is not  verified

                                                                                                 e  Can only be used as a first order
                                                                                                    approximation

-------
     As noted, the advantages  of the Level 1 analysis  include a high



 level  of sophistication, completeness and  accuracy.   The oxidant



 analysis  results can be verified with observed concentrations and a wide



 range of HC and NOX emission  control strategies can  be  evaluated.   This



 level  of  analysis was selected and successfully carried  out in several



 air quality maintenance areas  such as San Francisco and  Denver. However,



 the emissions and meteorological data required are  very  extensive  and  a



 substantial amount of manpower, time and money is  needed to perform the



 analysis.



      Level  2 analysis  may  use either the standard EKMA or OZIPP-EKMA



 procedures to determine the  required emission  reductions.  Both EKMA



 procedures  are  the result  of several years of studies  conducted by  the



 U.S.  EPA.  The standard EKMA requires less extensive  data and is a less



 time-consuming  technique, but has  relatively limited flexibility in



 application.  The OZIPP-EKMA procedure is more complicated, and requires



 a computer facility to execute the OZIPP and  to plot an ozone isopleth



 diagram.   City-specific factors are considered  in the  OZIPP-EKMA



 procedure.   In general,  the Level 2 analysis, especially the OZIPP-EKMA



 procedure, may be a desirable  alternative to  the Level  1  analysis, if



 the available data  and  resources are limited.   However, the results of



 EKMA procedures  are generally  more difficult  to  verify  as  compared to



 Level 1 analysis.



     Level 3 analysis is a simple  technique  for  calculating the



 emissions reductions.   However, this  technique  does  not  consider  the



effect of  NOX emissions on oxidant levels.  The  assumption  used in the



model,  i.e., the maximum  one-hour average  oxidant  concentration is a



linear function of  hydrocarbon  emissions,  has  been  criticized  in  many
                                4-14

-------
reports (4-6, 4-7,  4-8).  This technique may be used for a  first order
approximation.  Use  of  this technique for non-attainment  planning and
analysis may require EPA's approval  in  advance.
     Strictly from the technical point of view, the Level 1 analysis is
the most desirable approach to a  non-attainment planning  analysis.
Level  2 can be considered  as  an acceptable  alternative  and  level 3
analysis may be used only for preliminary assessment.

GENERAL GUIDELINES  FOR TECHNICAL ANALYSIS
     Several general guidelines  for  technical analysis  are presented  in
this  section.  These guidelines,  if followed, could  help to  facilitate
future oxidant non-attainment analysis.  These guidelines stem primarily
from  the experiences and  lessons  gained in preparing the  1979 oxidant
non-attainment plan for  the San  Francisco Bay  Region  (4-9).
Recommendations are  presented for what to do and what not  to do when
selecting and conducting technical  analyses.   It is  hoped that other
                                                              •
non-attainment areas  may extrapolate from the Bay Area's experience and
gain useful  insight into how to  improve the technical  analysis  work  in
their non-attainment planning efforts.
     As noted previously, the oxidant control analysis  involves  various
kinds of  uncertainties,  assumptions,  and simplifications.  The
techniques  used are at best classified  as "projection", "estimate",  or
"prediction".   Such an analysis is  frequently frustrated by the lack  of
data, guidelines,  resources and experienced personnel.   Tradeoffs
between  degree of accuracy  desired and available resources  have to  be
made at most steps of the analysis.  Analyses and predictions,  whether
they be  land use,  emissions,  or  future  oxidant  levels,  will not  be
                                 4-15

-------
completely defensible.   Credibility of the technical  analysis  in  an

oxidant non-attainment  planning  effort is of importance as  it usually

leads to a costly and frequently controversial control  program.  The

following  basic considerations are suggested  for  improving  future

non-attainment analysis efforts.

       t Emphasize the Initial Planning  and Design of Technical
        Analysis

        A successful non-attainment  planning effort requires a
        sound  initial  planning  and  well-organized management
        approach.   Before  initiating  the actual technical
        analysis,  the  non-attainment lead agency must carefully
        define what objectives and  requirements  of the Clean Air
        Act the  non-attainment plan  has  to  meet.  Then an
        appropriate  level  of analysis  can  be  selected.   The
        selection must  be  based on a systematic evaluation of a
        number of factors including:

          • magnitude of oxidant problem

          • level  of sophistication and degree of accuracy
            desired

          • time schedule for the non-attainment planning

          • availability of  data, budget,  manpower, equipment
            and other resources

          • capability of staff

        Once  the  level  of analysis (either Level 1 or Level  2) is
        selected,  a sufficiently detailed  project  plan should be
        prepared.   As a minimum, the project plan should cover:

          • specific analysis  elements  and tasks which must
            be completed

          t time schedule and  interrelationship  of those
            tasks

          • approximate  manpower resources requirements for
            each task

        In addition to  these  technical  analysis elements, the
        public  and local government coordination should be
        initiated  into  the  project  plan   along  with a
        non-attainment plan  review, revision, adoption  and
        approval process.   The initial  planning effort should
                               4-16

-------
  cover  a sufficient  level  of detail to  ensure  the
  successful  completion  of the  entire non-attainment
  analysis.  The Guidance discussed throughout this document
  can  be used  in planning and design of a successful
  non-attainment planning program.

• Balance the Process and the Product

  Virtually all of  the  analysis elements, e.g., emission
  projection, oxidant prediction, etc.  previously discussed
  contain some  kind of  process.  Since  it is difficult at
  best  to verify the product  (an  emission inventory or an
  air quality  projection for some  future years), the  most
  that  can be expected is some general  agreement that the
  analysis is  as  good as can be performed at the moment.
  The process by which the analysis is  conducted is crucial
  to obtaining such an agreement.  Conversely, as the number
  of individuals  involved increases,  staff  resource
  requirements, review  time, and coordination requirements
  multiply.  It is  therefore necessary  to  strike a balance
  between process and product according to the circumstances
  under which a particular analysis is  being made.

• Keep  All Analysis Elements Consistent

  As noted  previously,  outputs  of one element (e.g.,
  emission inventory) are usually to  be used as inputs to a
  subsequent analysis element  (e.g., oxidant projection).
  The  consistency of  all  analysis elements must  be
  maintained.  A technically  sound and defensible air
  quality projection is all but  impossible without a
  technically  sound and defensible emission projection.
  Similarly, the  quality of an emission  projection is
  largely affected  by the  quality  of land use, economic,
  population and transportation projections.  The allocation
  of  resources  to  the analysis elements should be
  commensurate with  the estimated  significance  of  the
  element.   For example,  it is  not  prudent to spend 50
  percent of the resources on 10 percent of the problem.

t Maintain Flexibility in the Choice  of Methodology

  In an  oxidant control analysis,  techniques (especially
  projection techniques)  and available  data bases are still
  at a crude stage of development.   Maintaining flexibility
  in the  methodology used from one analysis to the next is
  important.  Both  methods and data  bases, as well  as  the
  resources used to prepare the analysis, will change.  The
  most appropriate  methodology at any given time should be
  selected on the basis of conditions and  constraints
  applicable to that time  frame, rather than using one set
  of methodologies.
                         4-17

-------
 •  Perform Sensitivity Analysis

   Projections, whether they be emissions  or  future oxidant
   levels,  always have a  number of inherent uncertainties,
   e.g., emission growth rate, meteorological  conditions,
   etc.  One  way to deal with  uncertainty is to conduct
   sensitivity analyses.  Rough sensitivity  analyses can be
   performed  to determine whether the  method(s) used is
   appropriate to the policy  decisions being made.   If
   significant uncertainty  exists, the sensitivity of the
   decisions to such  uncertainty should  be examined.  If the
   sensitivity is significant, then additional effort should
   be devoted to improving the methods and data  bases used.
   (Several  examples  of sensitivity analyses  are presented in
   Volume II of this  Guidance.;

 •  Technical Analysis Should be Objective and Open

   Because  a non-attainment planning program will  be highly
   visible (due to public participation  required by the  Clean
   Air  Act), objectivity is very  important in all aspects of
   the work.  Several  methods can  be used  to maintain the
   professional  integrity  of  the planning  effort.  An
   advisory committee may  be established to provide early
   review  and comment on the analysis.  This group may  meet
   on an as-needed  basis  for  all  the technical  work,
   including technical assumptions, projection methodology,
   emission  controls, etc.   The comments received from this
   group should  be  incorporated  into the  work of the
   technical staffs.  When agreement can not  be reached on  a
   particular assumption,  it  should be clearly stated and
   documented along with the rationale for each argument.

   Documentation of the key dssumptions is  also critical in
   projecting future  land use, population, emissions and air
   quality.   At a  minimum,  sufficient documentation should be
   provided  for others to be able  to reproduce the results.
   Where other reports are  used,  both the actual data and
  methodologies  used should be cited.   For the more complex
   analyses,  sample calculations  or interim data should be
  provided.  As  part of any non-attainment planning effort,
   time  and money should  be  allocated   in  advance for
  providing reasonably detailed documentation of references,
  assumptions,  projection  methodologies, sources of data and
  the  actual results.

• Be Consistent With Other Data

  Air  pollution  emissions and  concentrations  are directly
  related to urban activity  in many forms.  An air pollution
  emission-projection  is  generally based  upon the projected
  levels of urban activities.   In any given  area  where
  emission  projections  are  being conducted,  other
                         4-18

-------
projections are also  being undertaken for population,
employment,  land  use, transportation, water quality and
supply, energy supply and  demand, etc.  To the  extent
possible,  information from  this other work should be used
to ensure consistency of data from one program to another.
A questionable  assumption  can always be debated  and
defended if it has been clearly documented.  Very seldom,
if ever, can  inconsistencies (internal or external) in
projections data be defended.
                      4-lg

-------
                             REFERENCES


4-1  MacCracken, M.C., et al,  "The Livermore Regional Air Quality Model:
     Volume  1.  Concept and Development," Lawrence  Livermore Laboratory,
     California, Livermore, California, Report No.  LICRL-77475, August,
     1977.

4-2  Reynolds, S.D., "Further  Development and Evaluation of a Simulation
     Model for Estimating Ground Level Concentrations  of Photochemical
     Pollutants,"  Systems Applications,  Inc., San Rafael, California
     February, 1973.

4-3  Environmental  Research and  Technology,  Inc.  ERT  Newsletter No.
     1-79, Concord, Mass. 01742, May, 1979

4-4  "Uses, Limitations  and  Technical  Basis  of  Procedures for
     Quantifying  Relationships  Between Photochemical  Oxidants and
     Precursors,"  EPA-450/2-77-021a, U.S.  EPA Office of Air Quality
     Planning and  Standards, Research  Triangle Park,  N.C.  27711
     November, 1977.

4-5  "User's Manual  for Kinetics Model and  Ozone Isopleth Plotting
     Package", EPA-600/8-78-014a, U.S. Environmental Protection Agency,
     Environmental Sciences Research Laboratory. Research Triangle Park
     N.C. 27711, July, 1978.

4-6  Reynolds, S.D. and J.H.  Seinfeld, "Interim Evaluation of Strategies
     for  Meeting  Ambient  Air Quality Standard for Photochemical
     Oxidant," Environ. Sci. Techno!.. 9(5), 433 (1975).

4-7  National Academy of Sciences, "Air Quality and  Automobile Emission
     Control," Vol.  3.   "The Relationship of Emissions to Ambient Air
     Quality," a report by the Coordinating Committee  on  Air Quality
     Studies to the Committee on Public Works, U.S. Senate, Washington,
     D.C., 1974.

4-8  Schuck, E.A., and R.A. Papetti, "Examination of Photochemical Air
     Pollution Problem  in the Southern California  Area," Appendix D of
     "Technical  Support Document for the  Metropolitan Los Angeles
     Control Plan  Final  Promulgation,"U.S.  Environmental Protection
     Agency, Region IX, San Francisco, California,  1973.

4-9  Leong,  E.Y.  and  R.Y.  Wada,  "Emission  Inventory  Projections:
     Hindsight, Insight and  Foresight,"  paper  presented at the Air
     Pollution Control Association,  Specialty  Conference  on Emission
     Factors and Inventories,  Anaheim, California,  November, 1978.
                                 4-20

-------
                        Chapter 5
         DETERMINING  BASE  YEAR
         OXIDANT CONCENTRATION
     For ozone control planning and analysts, one of the  initial tasks
is to define  the baseline ozone  concentrations.   The baseline ozone
concentrations are to be used  to assess compliance with the ozone
standard; they also provide a  basis for determining an appropriate
design value  for use in developing control strategies.   In most
non-attainment areas, baseyear ozone concentrations are determined based
on ozone sampling  data obtained at existing monitoring stations.
     In February  1979, U.S. EPA promulgated a new ozone standard (5-1).
The new standard is defined as "the expected number of days  per calendar
year with maximum hourly average concentrations above 0.12 ppm must be
less  than or equal to 1".  This differs from the previous standard  for
photochemical oxidants, which simply states a particular standard not to
be exceeded more than once per year.  As  a  result, the approaches to
interpreting compliance with the ozone standard, and the procedures for
determining an appropriate  design value  have also changed.   The
following two subsections briefly discuss the procedures  for
interpretation of  the new ozone standard and for determination of a
design value.

INTERPRETATION OF AIR QUALITY AND THE OZONE STANDARD
     The  key  issue in interpreting air quality data and the ozone
standard is to determine when the  expected number of exceedances is
equal to or less than 1.   In the  areas where sampling is complete, i.e.,
a  valid daily maximum hourly average value is available for  each day of
                             5-1

-------
 the year, the  procedure for  determining  the expected  number  of
 exceedances is straight forward.  The expected number of  exceedances at
 a monitoring site is determined  by calculating the average of the number
 of  exceedances recorded  for each  of the past 3 calendar years.
 Compliance with the standard  is  determined based on whether this  average
 is less  than or equal to 1.
      In  reality, a valid daily maximum 1-hour average ozone value may
 not be  available for each day of the year.   Incomplete sampling must be
 accounted for in determining  expected number of exceedances.   This can
 be done  by using the following formula:
                       v
               e = v + - • (N-n-z)
                       n
     Where
          e = the estimated number of exceedances for the  year
          N = the number of required monitoring days in the year
          n = the number of valid daily maxima
          v = the number of daily values above 0.12 ppm, and
          z = the number of  days assumed to  be  less  than  the
              standard  level.

     More detailed  information on the procedures for analyzing  sampling
data and computing  the expected number of exceedances  can be founded in
References 5-1  and 5-2.

DETERMINATION OF  DESIGN  VALUE
     Conceptually, the  design value  for a non-attainment area is the
value that  should be  reduced sufficiently to ensure that  the area  will
                                5-2

-------
meet the  standard.  The design value is important  in determining
required  emission  controls and in  developing control strategies,
especially when  EKMA  (Level  2 analysis) or Rollback technique  (level  3
analysis)  is to be  used.   Under the previous  photochemical oxidants
standard,  a design  value is the second  highest measured value.   The
promulgation of  the new ozone standard  has resulted in  a  revision to the
definition of the design value.  With the wording of the new standard,
an appropriate design value is the concentration with expected number of
exceedances equal to  1.  In other words, the design value is chosen so
that the  probability of exceeding this concentration is 1 out of 365.
Four alternative methodologies which may  be  used to estimate a design
value are  discussed in detail in Reference 5-2.  They are: (1) fitting a
statistical  distribution  method; (2) table look-up procedure;  (3)
graphical  estimation;  and (4)  conditional probability approach.  The
first three methods group sampling data from several  years into a single
frequency distribution and estimate  the design value at the frequency
equal to 1/365.  The  last method treats  sampling data from each year
individually in determining a design  value.  For the  uses, limitations
and examples  of  these methods, the reader is referred to  Reference 5-2.


REFERENCES
5-1  U.S. EPA,"  National  Primary  and  Secondary Ambient  Air  Quality
     Standards, Revisions to the National Ambient Air Quality  Standards
     for Photochemical  Oxidants." Federal  Register,  pp 8202 to 8237,
     February 8,  1979.
5-2  U.S. EPA,"  Guideline for Interpretation of Ozone Air Quality
     Standards," Office of Air Quality  Planning and Standards, Research
     Triangle Park, North Carolina  27711, January, 1979.
                                5-3

-------
                             Chapter 6
EMISSION INVENTORY AND PROJECTIONS

       In  an  oxidant non-attainment planning  analysis, the emissions
   inventory and projections are used as the basis  for predicting oxidant
   concentrations  and subsequently for developing control strategies.
   Additionally, the emissions inventory and projections can be used to
   evaluate "reasonable further progress" toward attainment of the oxidant
   standard.  A sound non-attainment plan relies heavily upon developing a
   detailed and objective emissions inventory.
       As  required  by  the Clean Air Act, an emissions inventory must be
   accurate, comprehensive,  and current.   The  inventory is an accurate
   accounting and characterization of all emissions and their sources in
   the non-attainment area.   To  compile an inventory, the following
   information is  needed:  location of emission,  type of process,  type and
   quantity of emissions,  type and efficiency of control  system  and
   stack/effluent  parameters.  Additionally, significant sources affecting
   non-attainment outside the non-attainment area need to be included.
   Because the emissions inventory for an oxidant control program should be
   as current as possible, a recent baseyear must be selected.  Consistency
   with air quality data should also be considered  in selecting a baseyear
   for the emissions inventory.  To determine whether the oxidant standard
                                 -j
   is to be achieved in future years, emissions projections for those years
   need to be developed.  These projections are based on assumptions for
   future emission source growth and development,  technical developments,
  and various emission limitations.  This chapter  presents  guiding
  principles for preparing the emission inventory and projections.
                                6-1

-------
PLANNING FOR  EMISSIONS INVENTORY/PROJECTIONS
     Before  compiling an emission  inventory, some initial planning with
a sufficient  level  of  detail  must be carried  out in order to  ensure
successful  completion of  the project.   During the planning,  the
objectives and  specific needs  of the emission inventory with respect to
the  oxidant  control  program  must be defined first.  Consistency among
the oxidant analysis elements,  e.g., urban  activities  projection  and
oxidant modeling, needs to be determined.  Appropriate methodology is
then developed  to  meet the  objectives.  Levels  of detail, manpower,
time,  equipment  and expertise requirements of  each methodology must be
evaluated against  the availability of source data,  resources, and other
constraints.  General Guidelines  for  planning  the technical  analysis
have been discussed  in  Chapter 4.   The considerations specifically
related to emission  inventory and  projections  are discussed in the
following paragraphs.
     As  noted previously,  one of the primary  uses  of an emission
inventory is  as input to an  oxidant model that relates source emissions
to ambient oxidant concentrations.   The  emission inventory must be
designed so that the results can be interfaced with  control  strategy
development  activities.   Before  planning an  emission inventory,  the
oxidant prediction relationship (OPR) to be  used should  be determined.
Then  an  emission inventory  with  a similar level  of  detail can be
designed.  Three levels of OPR  are discussed in Chapter 4.  The  level  of
detail  for the  emission inventory  and projections required  by these
three  levels  of analysis  are different.   Table 6-1  summarizes the three
levels  of  emission inventories  corresponding to each level  of analysis.
                                6-2

-------
                                              Table 6-1.  Comparison of Three Different Levels of Emission Inventory
Level
of
Analysis
1








2









3

OPR
Used
photo-
chemical
disper-
sion
model





EKMA









Rollback

Details of Emission Inventory Required
Precursor
Emissions
Considered
reactive
volatile
organic
compounds*
NO , CO.
anfi some-
times SO*
and particu-
lates.

reactive
volatile
organic
compounds
and NO,





reactive
volatile
organic
compounds
Geographical
Area
entire non-
attainment
area (signif-
icant sources
affecting non-
attainment
outside the
non-attain-
ment area also
considered)
entire non-
attainment
area (signif-
icant sources
affecting non-
attainment
outside the
non-attain-
ment area also
considered)
entire non-
attainment
area
Spatial
Resolution
grid cell
level
(1 .2 or
5 Km grid
cell)





county or
county
equivalent
level






county or
county-
equivalent
level
Temporal
Distribution
hourly








annual (or
seasonal —
sunnier
months)






annual (or
seasonal —
summer
months)
Consideration
of Species and
Reactivity
Classification
VOC and NO
species allo-
cation are
required






aggregated
VOC emissions
and NOX







aggregated
VOC emis-
sions
Information
on Indivi-
dual Source
required








desired
but not
required




.


desired
but not
required
Use of Exist-
ing County
Inventory
(NEDS or EIS)
substantial
modifications and
dlsaggregatlon
of existing
county Inven-
tory are
necessary.



may be used
directly or
with minor
modifications






may be used
directly or
with minor
modifications

Requirement
for Computer
to Process
Data
computer with
large storage
capacity Is
required






not required
but desirable








not required

CO

-------
     As can  be seen, the Level 3 analysis (rollback method)  requires an
aggregated,  annual HC emissions  inventory for the baseyear.   The Level 2
analysis (EKMA) requires an emission inventory with a  degree of  detail
similar to  that of Level  3;  the only difference is that a NOx emission
inventory is also required by the Level 2 analysis.   In  the Level 2
analysis,  the relative  temporal  (posts  a.m. period)  and spatial
distribution patterns of  HC  and NOX  emissions  for  the selected
simulation  day are needed if  the city-specific (instead of standard)
EKMA is to be used.  In both Levels 2 and 3,  the emission inventory can
be  compiled at county or  county-equivalent levels and then aggregated
for the entire  non-attainment  area.
     The  Level  1 analysis (photochemical dispersion model) requires a
much more detailed emission inventory for HC,  NO  and other pollutants.
                                             rt
A disaggregated emission  inventory with  spatial and temporal details,
i.e., hourly emissions  for each grid  cell  (1, 2, or 5  km) level, is
required by photochemical dispersion  models.   Additionally, organic
emissions must be broken  down  by species classes;  ground-level  and
elevated emissions can also be treated separately.   For this level of
analysis, a  significant portion of the effort will  be expended  on the
preparation  of the emission inventories.
     Based on the above discussions, an emission inventory can  be
broadly grouped into two  categories—aggregated and disaggregated.  An
aggregated emission inventory  can be used   in the Level  3 (rollback
model)  and  Level  2  (EKMA) analyses.   The disaggregated (spatially and
temporally  distributed)  emission  inventory is  used for Level  1
(photochemical  dispersion model) analysis.  The basic  differences
between  the  aggregated and disaggregated  emission inventory  and
                                 6-4

-------
projections are already  summarized  in Table 6-1.  The  information
provided can be used for  planning and designing  the various levels of
emission inventory required for the desired level of analysis in order
to meet the oxidant control objectives.

OVERVIEW OF EMISSION INVENTORY AND PROJECTIONS PROCESS
     To  gain a  perspective of the overall  emission  inventory and
projections process, Figures 6-1  and 6-2 present block diagrams for the
tasks  involved in preparing  the aggregated and disaggregated emission
inventory and projections.
     The process  for  compiling an aggregated emission inventory and
projections begins with planning  and design of the work tasks.   A review
and  evaluation of existing emission inventories is the next task  to be
conducted.  The guiding principle for this task is to utilize  the
existing emissions data and inventories to the fullest extent possible.
Various techniques can  then be used to collect missing data such as
activity levels, process and operating variables of major sources, etc.
Once the data collection  phase is completed,  the  collected data are
analyzed and compiled, and  emissions  are calculated.  The  baseyear
emissions inventory is then established.  The  next  step  is  to project
the  baseyear emission  inventory to future  years.  Both  the annual
emissions  inventory  and  projections are  made at the  county  or
county-equivalent levels, and then  summed  together  for  the entire
non-attainment area.
     Compared to an aggregated emissions  inventory,  a disaggregated
emissions inventory involves many more tasks and a much greater level of
effort.   After completing the initial planning  and  evaluation of
                                 6-5

-------
            Figure 6-1  Block Diagram for Aggregated Emission
                         Inventory/Projection Procedures
    Inventory
    planning
    and design
Evaluation
of existing
inventories
CTi
                                      Data
                                      collection
                                      point source
Data collection
area source
                                     Data collection
                                     mobile source
                                 Data analysis
                                 and emission
                                 calculations
                                 point source
Data analysis
and emission
calculations
area source
                                Data analysis
                                and emission
                                calculations
                                mobile source
                                                       Base year
                                                       emission
                                                       inventory
                                 Emission
                                 projection
                                 point source
Emission
projection
area source
                                 Emission
                                 projection
                                 mobile source
Interfacing
with OPR's;
reporting

-------
Figure 6-2   Block Diagram for Disaggregated Emission
           Inventory/Projection Procedures
          Point source
          data collection
          Point source
          spatial
          resolution
          Point source
         , temporal
         ' resolution
Highway motor
vehicle temporal
resolution
                              6-7

-------
existing  emission  inventories, a grid  system upon which the  emission
inventory will  be  based is determined.   The next step is to  collect data
on urban  activity  levels, major point source operation,  and others.
Then,  substantial amounts of effort are  required to translate the data
into emissions  and to spatially and temporally distribute the emissions.
From  this the baseyear  emissions inventory is established.  A similar
process using areal and source growth projections, legal requirements
and technological advancements is then used to prepare the emissions
projections for the future years.  Another major task in the preparation
of disaggregated inventories  is the  splitting of the HC and NOx
emissions into the various species  classes required by the photochemical
model.  This  effort requires the  expertise of an organic chemist and
detailed knowledge of the  model's photochemistry and the  region's
emission inventories.
     Some  of the  tasks,  e.g., "Evaluation of  Existing Emission
 Inventories"  and "Data  Collection",  are common to both aggregated and
disaggregated inventories.  In these cases, the data and information as
well  as  level  of  detail  required by the same tasks are more substantial
in the disaggregated inventories than for  the aggregated inventories.
However, the techniques  employed to carry out the tasks may be similar.
For this reason, the guiding principles for those tasks  common to both
levels of inventory compilation are discussed jointly in the following
task description  sections.   Tasks  unique  to each type of inventory
(either aggregated or disaggregated) are noted.  The planning and design
of emission inventory and projections have been addressed  in previous
sections.
                                6-8

-------
EVALUATION OF EXISTING EMISSION  INVENTORIES
     Following completion  of all preliminary  tasks,  a review and
evaluation of the existing emission  Inventories should be made.  In  most
non-attainment areas,  a  conventional annual,  count/wide emission
Inventory, I.e., EPA's NEDS  and/or EIS (6-1, 6-2, 6-3)  already exists.
Additional  permit  and  compliance related  emissions  data are also
maintained by local  air pollution control agencies.   More  importantly,
an inventory  for  oxidant precursor emissions has been prepared  and is
generally available in most  designated non-attainment  areas.   Many of
these data may be used directly  in the oxidant  control program.  In many
cases,  the existing  Inventories  can be updated or upgraded through
limited  contact with the sources or through  the use of an abbreviated
questionnaire.  Attempts should also  be  made to identify the data
collection,  data processing  systems, and  other  framework  already
established for the inventory.   Thus, 1f the existing inventory  does not
meet the  current  needs,  data collection and updating procedures  can be
built upon the existing framework. In principle, an emission inventory
should make use of the existing data and system to the fullest  extent
possible.
     The following considerations are suggested for the evaluation of
any existing emission inventory:
       0  Representative of current conditions -  is  the existing
         inventory  current or out-of-date?
       •  An accurate accounting of all  appropriate sources  - as
        most existing inventories were not originally prepared  for
         oxidant control  programs, there may  be a number of major
        organic and NOX emitters that may either  be  omitted from
         the inventory or treated  collectively as area  sources
        because their emissions of  some other pollutants  are
         small;  look for any obvious errors or omissions of large
         point sources.
                                6-9

-------
       • Consistency  with the  needs  of the non-attainment plan -
         level  of detail of the  Inventory should be consistent with
         the OPR to be  used.  In general, the existing  NEDS or EIS
         may be suitable for Level 2 (EKMA) and Level  3  (Rollback)
         analyses.  Substantial modification and disaggregation of
         existing  NEDS  and/or  EIS  are required if  the Level  1
         analysis  (Involving a grid-based photochemical  model) is
         to be used.
       • Consistency in  institutional  boundaries  - most existing
         emission inventories are at county levels.  If  a  county is
         not entirely  located within the non-attainment  area, extra
         effort is needed  to apportion  that county's  total
         emissions to  the  non-attainment area.
DETERMINATION OF THE  GRID SYSTEM
     This task is needed only when a grid-based photochemical dispersion
model  is used as a primary tool in oxidant control  analysis  (see Figure
6-2).  A grid-based photochemical model requires distribution  of  total
emissions to the grid cell level.  Therefore,  developing  an  appropriate
grid system at the start of the emission inventory  effort  is  important.
Generally, rectangular grid cells of equal  size are chosen.   Selection
of  the cell  size and the  overall grid  system is affected by  the
resources available for the detailed emission inventory and the computer
model to be used. More detailed guidelines can be  found in  references
6-4 and 6-5.

DATA COLLECTION
     As shown  in Figures 6-1  and 6-2, data for emission estimates are
generally collected and compiled in three separate  categories:   point
source, area  source,  and mobile source.  This is because the methodology
for preparing emission estimates and the required data are different for
each category-   For  each  of  the three  categories, there are several
                                 6-10

-------
techniques which can be used  to obtain the necessary data.   In  general ,
the necessary  data for large  and small point  sources are the process and
operating variables, and test emissions data  if available.  For mobile
and area sources,  activity levels  and  emission factors are needed.
Selection of particular techniques or combination of techniques must  be
based  on the  guiding principles discussed previously (objectives, level
of accuracy and detail desired,  consistency with the requirements  of
OPR,  overall  time  schedule,  resources  available, etc.).  The most
commonly used  data collection techniques are summarized in Table 6-2.
The mail survey approach and  plant inspection are most useful for
gathering data  from point  sources.   Field surveys  and review of
published data sources are  useful for collecting  area  source
information.  The  urban systems modeling is  a  primary  tool  for
estimating the  size and distribution of urban activities.  It is  the
best way to produce  the  detailed urban  activity data necessary for
spatial  and temporal distribution of emissions.  A number  of references
for urban systems modeling  is  given in  Appendix A.  Table 6-2 also
briefly  describes  the advantages and disadvantages of each technique.
The "how to" procedures for these techniques  can be found in References
6-6 and 6-7.

DATA ANALYSIS AND EMISSIONS CALCULATIONS
    Once  the necessary data  have  been obtained, they must  be
interpreted, compiled and entered into the data base. With these data,
various  methods  can be used to estimate emissions for area and mobile
sources.   The  "emission  factors  method" is most commonly used.   An
emission  factor  for a particular source  is  defined  as  the amount of
                                6-11

-------
                            TABLE 6-2

           COMPARISON OF VARIOUS DATA COLLECTION TECHNIQUES
                  FOR BASEYEAR EMISSION INVENTORY
    Technique

Mall Survey Approach
  (Questionnaire)
Plant Inspection
Field Surveys
Published Data Source
Urban Systems Modeling
         Characteristics

• Most  common  technique  for
  collecting point source data.

• Also useful  for collecting snail
  point sources to  be  reported
  collectively  as area source.

• Response of questionnaire varies.

•Errors  may  result  from
  Interpretation  of questionnaire, or
  from Incomplete questionnaire.

• Provide most  complete and accurate
  Information.

• Resource-Intensive; usually
  performed only at  Important or
  unique large  point sources.

• A primary  tool  for  directly
  gathering Information  on  certain
  area sources, such as agricultural
  burning.

•Errors  may  result  from
  extrapolating the survey results to
  the  entire area.

• Costly method;  used only when  It Is
  absolutely necessary.

• Primary tool  for obtaining  activity
  level  Information  on  most  area
  sources.

• Supplemental method for gathering
  point   source   data  1f  a
  questionnaire Is not returned.

• Published data  are generally not as
  current as required.

• Primary tool for estimating size
  and distribution  of  urban
  activities.

• Best way to produce urban  activity
  data necessary  for spatial  and
  temporal distribution of  area
  source emissions.

• Re source-1ntens1ve;  only
  experienced staff can perform the
  modeling.
                          6-12

-------
pollutant released  to the air from that  source divided by the  activity
level  of that source;  for  example, grams of hydrocarbons  per vehicle
mile  traveled.  The emission  factor method calculates  the total
emissions simply  by applying  a representative  activity level to an
emission factor.  EPA has published numerous techniques for  estimating
emissions for many  pollutant sources,  e.g., AP-42 and its  Supplements.
     For point sources, attempts  should be  made to collect emissions
data directly from  the owners (see previous section on Data Collection).
If they are not collected or not available, the "source test" method  or
"material balance method" may  be used  to determine the emissions.   The
source test method monitors  the  source  to determine the actual
emissions.   The  material  balance method  calculates emissions  by
analyzing the material composition and balance for a process.   It should
be noted that the emission factor method is also frequently used for
estimating point  source emissions.
     Table 6-3 compares the above-mentioned three emission estimation
methods  in terms of their application, resource  requirements  and
limitations.  More detailed information on the procedures can be found
in references 6-6 and 6-7.

SPATIAL RESOLUTION  OF EMISSIONS
     If a grid-based photochemical  dispersion model(s)  is  to be used  in
an oxidant control program,  emissions data must be resolved  to the  grid
cell  level.  The effort  needed to spatially resolve emissions  data
varies, depending on the type of emissions involved.   For  point sources,
it is  easy to assign the emissions to  the appropriate grid cell as one
can pinpoint their locations  with a street map and a USGS map of the
                                 6-13

-------
     Method

Source  Test
              TABLE 6-3

COMPARISON OF EMISSION ESTIMATE METHODS



                           Characteristics

                  • Most   accurate
                    resource-intensive method.
Material  Balance
Emission Factors
but
                  • Particularly useful  for quantifying
                    major point sources;  also useful
                    for developing emission factors for
                    area sources.

                  • Statistical validity  of  testing
                    results greatly affected  by the
                    number of tests, time period and
                    duration of testing,  and other
                    factors.

                  • Useful method to determine  certain
                    types of large  and  small  sources,
                    e.g.,  evaporative,  low  level,
                    intermittent or fugitive  organic
                    emissions, for which  source testing
                    method cannot be used.

                  •Needs  detailed   process  and
                    operational information;  only
                    experienced engineers can do the
                    calculations.

                  • Most commonly used method  for area
                    sources.  EPA AP-42  has  developed
                    emission factors for  most pollutant
                    sources.

                  • Most suitable for estimating mobile
                    source emissions, petroleum storage
                    and handling emissions,  and other
                    area sources.
                               6-14

-------
same scale.
     For area sources,  spatial  distribution of emissions may require
substantial effort.  In most cases, area  source emissions are  reported
at county or county-equivalent levels.  Various alternatives may be used
in distributing the countywide emissions  to the grid cell level  (6-4,
6-5).  Table  6-4  compares four commonly-used techniques for  spatially
distributing  area source emissions.   Reference  6-4 has  detailed
discussions  on spatial  resolution  of  various types of area source
emissions (petroleum storage and marketing,  solvent evaporation,
residential  fuel combustion,  etc).   More information on spatial
distribution of emissions may be found in Reference 6-5.
     Mobile  source  emissions  represent a challenge for spatial
resolution.  In conventional countywide  emission inventories, mobile
source emission estimates are usually  based  on countywide gasoline sales
or total VMT.  This kind of inventory  does  not provide the information
necessary for spatial distribution.   Since mobile sources generally
represent a large fraction of total organic  emissions, a more  detailed
procedure is  needed to estimate and distribute mobile source emissions.
Procedures  for estimating and distributing mobile source emissions
generally include:
       • obtaining  detailed  traffic data from transportation
        planning agencies (link-by-link  VMT, trip generation data,
        etc.),  or by using travel demand  forecasting models  (see
        Appendix A),
       • allocating link-by-link VMT and  trip generation data to
        the grid cell  level, and
       • estimating emissions at  a grid  cell  level using EPA's
        methodology  for calculating mobile  source emission
        factors.
                               6-15

-------
                            TABLE 6-4

                  ALTERNATIVE METHODS FOR SPATIAL
               DISTRIBUTION OF AREA SOURCE EMISSIONS
1.
    Method

Uniform distribution  of
emissions.
2.
Distribution of emissions
as  a  function  of
population.
3.
Distribution of emissions
as  a  function  of
surrogate Indicators.
4.
Emission  1nventory
directly for each grid
cell.
        Characteristics

• This 1s the  simplest method,
  and might be  accepted  for
  fairly uniform residential or
  agricultural  area.

• Cannot be used  In areas with
  diverse land  uses.

• Useful technique to distribute
  residential areas' emissions.

• Most census data concerns
  residential  population only,
  and would thus  displace many
  non-major point sources which
  operate  1n  Industrial  and
  commercial areas.

• Spatial detail  may be limited
  by  census  tract  size,
  especially  1n  sparsely
  populated areas.

• Most commonly used technique;
  a  great Improvement  over
  methods 1 and 2.*

•Require Information on
  population,  land  uses,
  employment  and others;  a
  resource-Intensive method.

• Most accurate method.

• Most  resource-Intensive;
  computer 1s required.
*Used 1n the non-attainment planning
discussed 1n Volume II of this Guidance.
                                In the Bay Area.  Details are
                              6-16

-------
     Allocation  of link-by-link VMT and trip generation data to  the grid
cell  level is  usually  a simple matter.   It can be done by  using a
network map and  a same  scale USGS map  and  then determining  the UTM
coordinates for each segment of traffic network and each traffic zone.
More detailed  discussions on spatial  resolution of mobile sources can be
found in References 6-4 and 6-5.
     Errors in spatial distribution  do  not  change the  total  mass  of
pollutant  emissions.  In this case,  a photochemical  dispersion  model is
intended for regional photochemical  modeling (as opposed  to local  plume
modeling)  so small  distribution errors should not be  critical to the
oxidant analysis  result.

TEMPORAL RESOLUTION OF EMISSIONS
     In order  to  be used in photochemical dispersion  modeling, emissions
must be resolved to the  hourly level.   For major point sources, the
hourly  and seasonal  process  variations are generally known.  This
temporal  resolution of  point source emissions constitutes no  problem.
For mobile  source and other area sources,  the most  accurate and  most
resource-intensive  approach is to  determine the  emissions  of  each
individual  source for each hour of the time  period  being  modeled.   A
less resource-intensive  and  more commonly used approach is to develop
typical  hourly patterns of activity levels for each source category, and
then to apply these hourly  factors to annual  or  daily  emissions to
estimate hourly  emissions.  In developing hourly factors, considerations
should  be  given to seasonal, weekend/weekday,  and peak  and average
condition variations.  The factors  important to temporal  resolution of
point,  area,  and mobile  sources are discussed in  detail in Reference
6-4.
                                6-17

-------
EMISSIONS (HC AND NOX) SPECIES ALLOCATION
     Since a photochemical dispersion model simulates the process  of
photochemistry,  precursor emissions must be  distributed into the various
species  classes required by the model.  Generally speaking, hydrocarbon
(HC) emissions  are  divided into  several  species classes that behave
similarly in photochemical reactions.   One  classification scheme groups
HC into four  classes:   paraffins,  olefins, aromatics  and aldehydes;
another scheme  divides HC by the number  of certain types of carbon bonds
present in each HC  molecule.  Some models also  require breakdown of NOX
into  NO  and  N02-   Actual  species classification required by various
photochemical models can be found in Reference  6-4.
     The  HC and NOX species allocation  scheme  used in the photochemical
model  must be known before compiling an  emission inventory because the
classification scheme influences the information requirements for each
source category and thus the data handling  process.   If  the emissions
data  collected are not broken down into  species classes, assumptions
about HC  and NOX species  distribution must be made for each source
category in  order to allocate emissions  to  various species classes.
Techniques for  HC and NOX species allocation  are discussed in detail  in
Reference 6-4.

EMISSIONS PROJECTIONS
     Emissions  projections include not  only forecasting changes  in
activity  levels, but also involve estimating the effects  of emission
control  regulations to be implemented.   Estimates of changes in activity
levels are usually based on  projections of  land  use,  population,
housing,  employment  and travel  demands.  These forecasts  are generally
                               6-18

-------
made by regional,  local and/or state planning agencies (see Appendix A.)
Additionally, construction of new  sources, expansion of existing
facilities, closing  of old plants  and  other similar  information on
individual  sources must also be included.
     Estimating  the effect of emission control regulations  is also a
complicated task.   It requires an evaluation of the  effect of the
promulgated "reasonable available  control  technologies", compliance
schedules,  new  source emission standards  and change in emission  factors.
     There  are  two types of emissions projections:  baseline emission
projections and  control  strategy  emissions  projections.   Baseline
emissions projections are prepared  by considering the projected  changes
in  activity levels  and  the effects  of control  regulations  already
scheduled or in  effect.  Baseline emissions projections serve  as a basis
for determining  the emissions reductions  necessary to attain the oxidant
standard.
     Control strategy emissions projections are emissions estimates
which incorporate the effects of new, changed, and/or additional control
strategies not included  in  the baseline projection.   Control  strategy
emissions projections are used as inputs  to a photochemical dispersion
model  to predict  the ambient oxidant  concentrations resulting  from the
proposed control strategies.
     Problems
     Emissions  projections, by their very nature, are speculative and
represent nothing more than a "best  guess" of future conditions.   They
frequently  involve a number of assumptions, judgments, and other similar
uncertainties.   There are generally  inherent  problems  in  any emission
                                6-19

-------
projections.  These may  include the following (6-8):

     Defensibility - Projections,  whether they be population,
     unemployment, or emissions, will  always be attacked because of
     their speculative nature.   The technical  credibility  of
     emissions  projections  will  be  a  function of their
     reasonableness (i.e., is it plausible), the amount of research
     and  documentation  of  assumptions, and the  procedures  or
     methodologies  used to make the projections.  In the  final
     analysis, no  projections are completely defensible; however,
     there is no reason  why they should not be clearly documented,
     reasonable and technically credible.

     Objectivity - Because of the importance emission inventory
     projections play in  urban activity planning—e.g., industrial
     siting decisions, air quality control strategies, highway  and
     sewage  treatment construction programs—the motives  of
     projecting emission inventories are always questioned.
     Depending on  the perspective being pursued the forecast
     emissions are  "too high,",  "too  low,"  "surprisingly
     conservative," or "unduly optimistic".  It is not uncommon  for
     a particular set of projections  to be a\± of the above at  the
     same time!  Projections  will  be perceived very differently
     depending on the eye  of the beholder and what must be  strived
     for is technical  objectivity in  the process.

     Openness - Because  of the previous two problems--defensibility
     and  objectivity--openness and candor in making emission
     projections is  needed  to  gain  any credibility.   Many a
     projection has  been made by junior staff in a  "closet"  which
     upon  closer examination was not  documented,  and  thus seemingly
     arbitrary,   or which though  documented was clearly
     unsupportable.  Internal  and  external  review of emission
     inventory projections  will  almost always  improve  their
     technical quality and enhance their credibility.

     Uncertainty - Simply  stated, projecting emission inventories
     is fraught with  uncertainty.  Some  uncertainty can  be
     accommodated by projecting in ranges or conducting sensitivity
     analyses with different assumptions.  However,  some degree of
     uncertainty will  always accompany emission projections.  This
     fact should always  be  acknowledged immediately.  The art of
     projecting emission  inventories  is  not eliminating
     uncertainty, but learning how to minimize  it.

     Consistency - Emission inventories are seldom compiled in a
     vacuum.Tn any given  area  where  emission inventories  are
     being projected,  other projections are also being undertaken
     for population, employment, land use, transportation,  water
     quality and supply, energy supply and demand, etc.   To the
     extent possible,  information  from this  other  work should be
     used  to  ensure consistency  of  data from  one program  to
     another.   A questionable assumption can always  be debated  and
                                6-20

-------
     defended  if 1t has been  clearly documented.   Very seldom, if
     ever,  can  inconsistencies (internal or  external)  of
     projections data be defended.

     Resources - Very substantial amounts of time and money can be
     spent on emission inventory projections.  Generally though,
     since neither a lot of time or money  are available, common
     sense should dictate the resources  to  be allocated to this
     task.  Seldom are  too  many resources devoted  to  this
     assignment (if anything  the converse  is  true).  As a rule of
     thumb, sufficient resources should be committed to conducting
     emission inventory projections to circumvent many of the
     problems cited above.   Also, the allocation  of  resources
     should be commensurate with the estimated significance of a
     particular source  category.  For example,  it is not prudent to
     spend 50 percent of  the resources  on 10  percent of the
     problem.


     Guiding Principles

     The general  guidelines  for technical analysis discussed in Chapter

4, if kept in mind, will help to minimize the above-mentioned problems.

Additionally, the following guidance  is  also  important to improving

emission projections.
       t Methodology consistent  with that used  in baseyear
        emissions inventory - it is essential to project emissions
        based on methodology and computation  principles consistent
        with  those used  in  the  emission  inventory for  the
        baseyear.  This  procedure,  which should be applied  to each
        emission category, will  insure  consistency between
        baseyear emissions inventory and the future year emission
        projections.  For example,  if a predictive land use model
        (PLUM) is used to estimate  land use activity levels  in the
        baseyear, the same model should be used to project  future
        land  use activity levels.   Otherwise, spurious emission
        differences may  result from changes in methodology  and the
        associated assumptions.

       • Close coordination with  planning  agencies - as  noted
        previously,  forecasts of  future  urban  activities (land
        use, population, employment, transportation, etc.)  are
        generally conducted by agencies other than air pollution
        control  agencies.  The benefits of close coordination with
        planning agencies are two-fold:  1) consistency with other
        planning programs -  in a  non-attainment  area  where
        emission projections are  being made, there are probably
        several  other projections also being  undertaken for land
        use,  population,  employment, etc.  Emission projections
                                6-21

-------
         should  be  based on the information from this other  work,
         thus,  the  consistency of  data from the  oxidant control
         programs to others can be established;  and 2) optimizing
         the usefulness of planning data already developed for the
         non-attainment area - this may  substantially reduce the
         effort  required for emission projections.
     Overview of Projection Approaches
     There  is  no one correct way of projecting  emissions.  Many factors
influence the procedures to be used.  These factors  include availability
of  data and resources, personnel, and level of  sophistication or degree
of  detail  required.  Since  so many air  pollution  emission  source
categories exist,  it is  prudent to be cognizant of  as many  projection
techniques available as possible  so  that different procedures can be
used as appropriate and as differing circumstances dictate.
     Perhaps the simplest and most frequently used technique  is linear
extrapolation or regression where a series  of historical data  points are
projected into the  future in some simple linear  fashion.  This procedure
has obvious advantages and disadvantages  which will  not be articulated
here.  A more rigorous extension of the  linear regression techniques is
the time series multiple  regression  procedure.  Here the  historical
emission estimates  are correlated to any number  of independent variables
(e.g.,  population,  automobiles, income, fuel use).   What results is an
equation which represents the  "best"  correlation  cf the independent
variables with emissions  for use in projecting future emissions.  The
theory of this  technique is that  the  historical  factors which  have
apparently accounted  for emissions will  continue to do so in the future
in  the  same manner.  To accurately forecast  emissions  using  this
technique requires  an  accurate forecast  of  all the independent variables
assumed  to  be  important in accounting for historical emission trends.
                                6-22

-------
     Frequently, very  little data are available  to assist in making
future  projections.  In these cases, engineering or scientific judgment
must be  used.  This  can  be  done in one step  in what amounts to be  a
"best guess"  or in more systematic group techniques  such as the Delphi
method.   Other gaming procedures  are also possible.
     Another emission  projection methodology is  the use of surrogate
variables.  This  procedure  assumes  that a given pollution  source
category can  be  accurately projected by  projecting some related
variable. For example, increases  in aircraft emissions  might be based
on  projected increases in passenger air travel. Similarly, hydrocarbon
emissions from  service  stations might be  projected on the basis of
forecasts of gasoline sales.
     The EPA and other  agencies have published  several  reports and
guidelines addressing emission projections.   Detailed "how to"
procedures for the  above-mentioned approaches can be found in References
6-7, 6-9, 6-10  and 6-11.  The surrogate indicators  methodology,  used in
the oxidant  planning program in the San Francisco Bay  Area, is detailed
in  Volume II of this Guidance.   Selection of particular techniques or  a
combination  of techniques  should be based  on  the  guiding principles
discussed above.
                                6-23

-------
6-1
                             REFERENCES
U.S.  EPA,  "AEROS Manual  Series, Volume  I:   AEROS Overview," U.S.
EPA, Office of Air Quality  Planning and Standards, publication  no.
EPA-450/2-76-001, February.  1976.
6-2  	, "AEROS Manual Series, Volume II:  AEROS User's Manual,"
     publication no. EPA-450/2-76-029, December, 1976.

6-3  	, "AEROS  Manual  Series,  Volume  III:   Summary  and
     Retrieval," publication  no. EPA-450/2-76-009a, July, 1977.

6-4  Drivas,  P.J.,  L.G.  Wayne,  and K.W. Wilson,  revised draft final
     report,  "Procedures for  the Preparation  of Emission Inventories for
     Volatile Organic Compounds,  Volume II:  Emission  Inventory
     Requirements for  Photochemical  Air Quality Simulation  Models,"
     prepared for  U.S.  EPA, Office of Air  Quality Planning  Standards,
     Pacific  Environmental  Services,  Inc., Santa Monica,  California,
     September, 1978.

6-5  U.S. EPA, "Guidelines  for  Air Quality Maintenance Planning and
     Analysis, Volume 8:  Computer-Assisted Area  Source Emissions
     Gridding Procedure." U.S. EPA, Office  of Air Quality Planning and
     Standards, Research Triangle Park, N.C.  27711, September,  1974.

6-6  U.S.  EPA, "Procedures for the Preparation of Emission  Inventories
     for Volatile Organic Compounds, Volume  I," U.S. EPA,  Office of Air
     Quality Planning  and Standards, publication  no. 450/2-77-028,
     December, 1977.

6-7  U.S. Environmental  Protection Agency,  "Guidelines for Air Quality
     Maintenance Planning and Analysis - Volume 7:   Projecting  County
     Emissions," EPA-450/4-74-008,  U.S. EPA, Office  of Air Quality
     Planning  and Standards,  January, 1975.

6-8  Leong, E.Y.,  and R.  Wada,  "Emission  Inventory Projections:
     Hindsight, Insight and Foresight," paper presented at the APCA
     Specialty Conference on  Emission Factors and Inventories,  Anaheim,
     California, November,  1978.

6-9  TRW, Inc., "Development  of a Sample Air Quality Maintenance Plan
     for San  Diego," prepared for U.S. Environmental  Protection  Agency,
     Report No. EPA-450/3-74-051,  1974.

6-10 Booz-Allen and Hamilton,  "Regional Emission Projection System,"
     prepared  for  U.S.  Environmental  Protection  Agency,
     EPA-450/3-75-037,  1975.

6-11 Bay Area Air  Pollution Control District, "Method of Projection,"
     San Francisco,  California, May 31, 1977.
                                6-24

-------
                          Chapter 7
    OXIDANT PREDICTION  METHODS

     Three levels of sophistication in oxldant prediction  methods are
presented In this guidance.   The most sophisticated Is the Level  1
analysis.  At this level of analysis, a dynamic grid-based photochemical
model or a trajectory model  may be used for predicting oxidant levels
resulting from proposed  emission control  strategies  of  oxidant
precursors.  The Empirical-Kinetic Modeling Approach, EKMA,  is the Level
2 analysis.  A standard set of ozone isopleths can be used  in the EKMA
procedure to assess emission control strategies.  Another option of EKMA
is to use a trajectory photochemical  model,  i.e.,  the  Ozone Isopleth
Plotting Package (OZIPP),  to simulate maximum 1-hour average ozone
concentrations.  The results of OZIPP simulations are used  to generate
an  ozone  Isopleth diagram.  This diagram  is then used  in the EKMA
procedure to calculate the  required ozone controls.  Last  is the
rollback technique (I.e.,  Level  3 analysis).  The rollback model is
based on the assumption  that for a  given  region, a reduction  in
hydrocarbon emissions will produce a proportional  reduction in maximum
oxidant levels.  This chapter discusses these three  levels of analysis.

LEVEL 1 -  PHOTOCHEMICAL DISPERSION MODELS: GRID OR  TRAJECTORY MODELS
     General  Description
     Photochemical dispersion models attempt to describe in  mathematical
terms the various physical  and chemical processes that  are  responsible
for ozone formation.   They  can be used  to quantify the relationships
between emission sources and ozone concentrations at various receptor
                              7-1

-------
locations, and similarly  to evaluate the  effectiveness of proposed
emission  control  strategies  on  oxidant precursors.   At  this level of
analysis,  a  grid (Eulerian) or a  trajectory  (Lagrangian)  model  may be
used.   The  following paragraphs briefly discuss  these two types of
models.  For more detailed information on application of photochemical
dispersion models, the reader is  referred to Reference  7-1.
     A grid-based photochemical dispersion  model  simulates ozone
formation and transportation on a grid over the region of  interest by
considering  diffusion and chemical reactions of  oxidant precursors as
they  are  passed from grid  to grid according to the wind  field defined
for each grid.  The atmospheric diffusion of pollutants is  treated in
three  or  two  spatial dimensions.  Among various photochemical models,
the grid-based models are the most sophisticated  and  offer the highest
accuracy  of ozone prediction.  However, the level  of effort  required is
the greatest and the data and resources needed are the  most extensive.
     A trajectory model  may be a  compromised alternative to a grid-based
model  if  available  data  and  resources  are relatively   limited.   A
trajectory  model  considers oxidant precursors as they  enter into a air
parcel moving  along a trajectory  path according to local mean wind speed
and direction.   Ozone  formation and  transportation  is  simulated by
solving equations  of chemical  reactions and vertical  mixing  of
pollutants in  the air parcel.
     The following is a  list of example photochemical dispersion models
available:
       Grid  or Eulerian  models  -
       t LIRAQ-2, developed by Lawrence Livermore Laboratory.  This
        was the model  used for  the San  Francisco Bay Area
        non-attainment  analysir  (7-2)*.
                                 7-2

-------
       t SAI  'Urban Airshed Model,  developed  by  Systems
         Applications, Inc.  The model  has been used for  several
         urban studies (7-3)*.

       t IMPACT,  developed by Science Applications,  Inc.  The model
         was developed for the CARB  (7-4)*.

       • MADCAP,  developed by Form and Substance, Inc.  San  Diego
         is using this  model  for its non-attainment planning
         analysis (7-5)*.

       Trajectory or Lagrangian models -

       • ELSTAR, developed  by  Environmental  Research  and
         Technology,  Inc.   This  model  was developed  for  the
         Coordinating Research  Council  (7-6, 7-7)*; an official
         document is to be published in  the near future.

       • OZIPP,  developed by  the U.S. EPA.  This model is usually
         used  along with EKMA and  will be discussed  in the  Level 2
         analysis of this guidance (7-8)*.

     Because of their sophistication, photochemical  dispersion models

 have  the highest potential for answering a number  of strategic oxidant

 control issues.   These issues would  be of importance in  evaluating the

 effectiveness of alternative  emission control strategies in  reducing

 ambient oxidant levels.   Among  these issues are the  following:

       • Hydrocarbon versus NCX control - These pollutants  are  the
         primary  ingredients  in  the  photochemical reaction  for
         ozone  formation, and changes in the hydrocarbon/NOx ratio
         due to various  emission control strategies would change
         the resultant  oxidant  concentration.   Eulerian  or
         Lagrangian models can be used to provide estimates of  the
         degree of both  hydrocarbon and NOX  control  required  to
         meet  the  standard.

       • Assessing attainment of oxidant versus  short-term  N02
         standard - A short term N02 standard  is called for in  the
         1977 Clean Air  Act Amendments.  Eulerian and Lagrangian
         models can  theoretically  perform both oxidant  and  NOg
         predictions.   The validity of the latter prediction still
         needs  to be field-verified.
*References  cited are the user's manual(s)  for  the model.

                                 7-3

-------
       •  Control of low-level  vs.  elevated sources - This issue
         weighs the cost-benefit  ratio of one  type of source
         control policy vs.  another.  Vertical  resolution  is needed
         to  address this issue.  All  but the LIRAQ-2 models  are
         multi-layer models  in the vertical and  can be used for
         such  a  policy evaluation.

       •  Selective control of  hydrocarbon compounds  -  It  is
         possible that  in  the  preparation of alternative emission
         control strategies, sources emitting a  single hydrocarbon
         class might be singled out for control.  The  effect of
         such  a  policy on ambient ozone levels can be assessed by
         the complex models  to a degree.  The number of hydrocarbon
         classes ranges from three to five and these  specific
         hydrocarbons can be used for control strategy testing.

       •  Long-range transport  - Oxidants  and its  precursors are
         known to be capable of  long-range atmospheric transport
         from ojie air region  to another.  In such an instance,  the
         boundary and initial  concentrations of  the  air region into
         which the pollutants  are carried would  reflect this effect
         as  would the ambient oxidant  levels.   The magnitude of
         such  an effect has  not  yet  been assessed.

       •  Effect of spatial  and temporal  variation - Selective
         siting and defined  operating schedules  for projected
         sources  would affect  ambient  oxidant  levels.   As  the
         photochemical  dispersion models are  space  and time
         dependent, they are most  suited for such  a growth policy
         study.


     Input Data  Requirement

     Input  data for a photochemical dispersion model can be broken down

into three categories: 1) emissions, 2) meteorology,  and 3) initial and

boundary ozone and precursors  levels (see Figure 4-1).   For a grid

model, these data need to be spatially and temporally distributed over  a

grid  over the region.   In  other words, data are needed for each of the

grid cells in  the region. Efforts required for  preparing such detailed

inputs could  be substantial.  However, it should be  pointed out that it

is not necessary  to observe all   of  the  required data  directly.

Reasonable  compromises can  be  made;  such compromises were made with

LIR'AQ  as  described in Chapter  5  of the Volume II of  this guidance.
                                7-4

-------
     Additionally, each  of the  above-mentioned  grid models has

preprocessor  programs for manipulating raw  input  data into computer

readable codes.   This would minimize computation  time  and data

preparation effort.   Most grid models have preprocessor routines  for

each of  the three input data categories.  These are described below:

      "(Emissions:   A complete,  chemically disaggregated,
         spatially and temporally varying emission inventory must
        be  developed for  the study  area.   The amount of
         disaggregation necessary  is dependent  on  the  specific
        model.  At the least,  the hydrocarbon sources must be
         divided  into  those  classes used  in the photochemical
         simulation.   N0y emissions need to be divided among  NO and
        N02-

        Because the  models are space dependent, the emission
         sources must be  located to the resolution of  the  grid
        except for point sources.   These are specifically located
        at  a  grid  point.   For  those  models  incorporating
        multi-vertical layers, a vertical  resolution of the point
         source  is also required.   This is  accomplished via  plume
        rise  calculations so that the stack and plume design
        conditions are required input.

        The temporal  variation  of each  emission source must be
        stated.  Usually,  a one hour  resolution is  adequate.

        Area and mobile sources can be  defined on the  grid by  a
        complex urban system model  and a detailed traffic demand
        and network model.

        The labor requirement  to produce this inventory is
        dependent  on  the existing  data.  However, most air quality
        regions either do not have  a complete inventory and/or it
        is in  the  wrong format.  Several  man-months are normally
        required.

      t Meteorology:   A mass  consistent wind field is used by all
        the  models for pollutant transport.   All the models except
        LIRAQ-2 use  a  three  dimensional field.  LIRAQ is single
        layered.  Input  to  generate the wind  field  are  wind
        direction  and speed at different heights (except for  LIRAQ
        which  requires only surface measurements); wind shear
        between layers; inversion  height;  insolation; topography;
        latitude and  longitude and  date.   Except for the last
        three  inputs, all  the  date needed to generate  the  wind
        field  should ideally be  available  for each grid.  This  not
        being  the case usually,  interpolation between  sites of
        actual measured data must be carried  out.  Because of the
                                7-5

-------
         lack of data and interpolation, the resultant  field must
         be  scrutinized by an experienced meteorologist  to  appraise
         its consistency, especially in complex terrain.   This
         process needs  to  be repeated for each simulation  day as
         they are selected.   This  is  nontrivial task  in  itself,
         requiring several person-days  per simulation day.
       t Initial and boundary ozone  and precursor levels:   These
         specify the concentration  of  ozone and  precursors at the
         edge  of  the study  area  and at the start  of the model
         simulation for each grid.   Most of the time, little or no
         actual data are available and professional  judgment is
         applied.
     Generally, the types of input data required for a  trajectory model
are similar  to  those for a grid model,  but the level of detail may be
less.   Theoretically,  a trajectory model requires  input  data only for
the corridor  encompassing  the  trajectory path.   However, general
application of a trajectory model to a non-attainment  area usually
requires that  emissions and meteorological  data  be assembled  for a
significant  portion of the non-attainment area.
     The most advanced trajectory model,  i.e., ELSTAR,  requires  input
data with a  level of detail  almost  identical  to that  required for a grid
model.  The  ELSTAR requires the following inputs:
       t Area Source Emissions
       • Point  Source Emissions
       t Diffusivity Coefficients and Spatial  Mesh
       • Initial Concentrations of  Species
       • Latitude, Longitude, and Date
    The  area and  point  source  emissions  including NOX, CO, and five
separate classes of hydrocarbons  must be prepared as  a function of  time
(e.g.,  hourly).   The emissions  data are created by  an  emissions source
flux generation program.   This preprocessor program  requires  a gridded
emission inventory  to compute  schedule  fluxes for a  specified
                                7-6

-------
trajectory.  The fluxes  are computed by summing  up  all of the emissions
from all  sources types  in the inventory that exist in a given grid cell
at a given time.
     The data  measured at meteorological  stations (surface wind speed,
surface temperature  data and radiosonde data) are  required by another
preprocessor program  to  determine the  diffusivity coefficients.  The
output from the program  also supplies information necessary to the  user
to specify mesh geometry.  The last two items of the required inputs can
be readily obtained  or estimated from existing sources.
     Model  Validation
     Because  a photochemical dispersion model attempts to duplicate  a
real world situation mathematically, its predictions should be verified
by  comparison  with  observed  data.  If all  the input data were correct
and the input  assumptions were within reason, the modeling results
should parallel the measured data.  If possible,  input information may
be adjusted within their range of  uncertainty  to  see  if  agreement is
improved.
     A model  which is sufficiently complex  to simulate chemical-physical
processes  should not have to be calibrated.  There are two options for
dealing with  the problem of imperfect model performance.  The  first
option is  to  accept  the model  results  as  they are, making no attempt to
"correct"  or  "calibrate" the results according to observed data.   This
may  be done  with confidence  if more  than  one prototype  day is used
during the  model  evaluation and validation  process to  eliminate  any
systematic  bias in model  results.   However, it may  not always be
possible to eliminate all  systematic  bias or  to  obtain  satisfactory
results for  several  different  prototype days, particularly if schedules
                                7-7

-------
are constrained by regulatory deadlines.   In  such cases, some method
should  be  considered  for  appropriately compensating  the  results to
ensure  that control requirements  are not  either  understated or
overstated. Issues relating  to  model  validation  or calibration need
further investigation.
     Baseline Oxidant  Predictions  and Development  of Alternative
     Emission Control Strategies'
     The baseline oxidant prediction  is defined as the  oxidant level  for
some future year  if  only current regulations for emission control were
applied.   Calculation of this  baseline prediction involves developing  a
future emission  inventory  based on regional  growth  and development
policies;  regional  transportation  plans and policies;  and regional
stationary  source development.   The meteorological  scenarios for  the
simulation days  are used.   Boundary  and initial  conditions used for  the
simulation  day can be used  or others  can be input.   Once the baseline
oxidant predictions  are made, various emission  control  strategies  can be
applied to  the future year  inventory  and the effect on oxidant formation
can  be  discerned.   However, because  of the large  computational costs of
running these models, the labor requirements  to  develop the emission
inventory  and to interpret  the output, the number of control  strategies
should not  be excessive. A  prior  screening is  advisable.
     Sensitivity Analysis
     The model  predictions should  be used with  discretion because  the
results are only  as  good as the  input.   Many assumptions  and
simplifications  have gone  into  the model  formulation and input  due to
inadequate  knowledge of certain  actual  phenomena, limited computer
capability  and lack of good  data.
                                7-8

-------
     The  initial  and boundary  levels are an important  assumption,
especially in a short  (e.g., 12 hours)  simulation.  The time required to
derive input values  from the simulation is  equal to or greater than the
time needed to perform the short simulation.
     Some sensitivity  studies have been performed on these models.  A
common one is the effect of  varying  precursor levels on oxidant
formation, which is  another way  of  looking at  emission control
strategies.  LIRAQ has been used in San Francisco for such  a study.  SAI
AIRSHED addressed this in the Denver area.  Results for both' studies are
applicable only to these specific sites.

LEVEL 2 -  EKMA
     General Description
     EKMA  was developed by the EPA as an oxidant prediction  relationship
a step beyond ROLLBACK (7-9, 7-10).   It is  based on a series  of  curves
relating  observed early morning NOX and non-methane hydrocarbon  (NMHC)
levels with  the predicted maximum  oxidant  level downwind  of  a city
center  (see Figure 7-1).  With the ozone isopleths, relative  changes in
ozone levels can be  assessed as  a  function  of changing NOX  and NMHC
levels.
            »
     There are  two ways to use EKMA.  One is to use the standard  set of
ozone  isopleths which are non-site specific and which reflect only one
meteorological  scenario to  assess alternative  control  strategies.
Another way is  to make the set of isopleths city-specific.   This  is done
by using the Kinetics Model  and Ozone Isopleth Plotting Package (OZIPP)
(7-8).  By using OZIPP, the  city specific factors  including  morning
concentration of precursors; post 8  a.m. precursor emissions;  mixing
                                7-9

-------
!

I—'

C
         ri
        o
                                                                                                          1J
             Figure 7-1.  Examole ozone Isopleths used 1n EKMA

-------
height; latitude,  longitude and  date;  reactivity of  precursor mix;
transported ozone from  upwind areas; and  ozone and precursors from aloft
are reflected in the  generated ozone isopleth diagram.  In other words,
the city-sped f ic  factors can  be  considered in developing ozone
controls.
     EKMA 1s  particularly useful for addressing the "hydrocarbon  vs NOx
control" strategic  oxidant control  issue.   It can do this  because EKMA
is  basically a photochemical simulation  model.  Various  combinations of
hydrocarbon  and NOX control can be evaluated of their effectiveness in
oxidant control.
     The two EKMA procedures using either the standard ozone  isopleths
or  OZIPP-generated  isopleths  are illustrated in Figures  4-2a and 4-2b
respectively.   EKMA application procedures  are briefly discussed  below.
For more detailed information, the reader is referred to  references 7-8,
7-9, and 7-10.
     EKMA  Using Standard Isopleths
     The data  required for this  analysis are:   ozone design value for
the baseyear and the  corresponding  6-9 am  measured  non-methane
hydrocarbon (NMHO/NO   ratio,  and the  baseyear and future year
                        A
hydrocarbon  and  NOX regional  emission inventory.   Care must  be  exerted
in determining  a  design value for the baseyear (details are discussed in
Chapter 5).  A  selected design value for  the baseyear is  located  on the
set of  isopleths (see Figure  7-1).   Changes in hydrocarbon and/or NOX
levels  needed to meet the oxidant standard  can  then be determined.  A
direct relationship  is assumed between  the hydrocarbon  and NOX
measurements and  emissions.
     For each control strategy,  a "target"  emission rate  of hydrocarbons
                                 7-11

-------
and/or NO  is  determined.  This is an absolute quantity.  Comparing  this
         n
with  the  projected  future year  emissions, the relative  reduction
required for each pollutant can be determined and the  various control
strategies can be developed.
     EKMA Using OZIPP-Generated Isopleths
     This approach  tunes the model to a particular site.  In addition  to
the data required by the standard method, the  following is also needed:
       • hydrocarbons split into propylene, butane,  and aldehyde
         fractions
       • NOX/N02 ratio
       • concentration of transported ozone  and precursors
       t mixing heights at 8 a.m. and 3 p.m.
       • longitude, latitude and date
     With  this information, the  computer program OZIPP  is used to
calculate maximum 1-hr  average ozone  concentrations.  The  results of
OZIPP simulation are then used to generate an ozone isopleth diagram.
The  procedure for  the use of this city specific isopleth diagram is the
same as for the standard isopleth diagram previously described.
     The additional  labor costs associated  with  generating the
city-specific  isopleth are minimal.  Given the total  hydrocarbon  and NOX
emissions, only relative ratios are needed to  separate these two  sets of
compounds into their respective  species classes.  The  meteorological
data  is easily taken from published  sources.   Transported ozone and
precursor concentrations can be  obtained  from upwind measurements.
Also,  there are default values for all  these inputs.   Computation  time
to generate a  set  of  city-specific  isopleths is insignificant at 3
minutes on an  IBM 370/168 computer.
                                 7-12

-------
     EKMA cannot be  used to predict  absolute ambient  oxidant values;
rather,  it is meant  to  be  used for evaluating the effects of relative
changes in precursor levels on oxidant formation on a  regional basis.
The EKMA model hasn't been validated or verified, though its chemistry
has been verified  in the sense that it is based on smog chamber tests.

LEVEL 3 - ROLLBACK MODEL
     In the brief  history of oxidant modeling for control  strategy
evaluation since  the  1970 Clean Air  Act, only rollback models have been
sanctioned by  the  U.S. Environmental  Protection Agency.   These are the
"Appendix J" rollback  model and the rollback model.  However,  recent EPA
announcements  imply  that Appendix J is no longer an acceptable method
(7-9).  The rollback model is based on the assumption that for any given
region, reduction  in hydrocarbon emissions will produce  a  proportional
reduction in oxidant levels (7-11).  Thus by using the rollback equation
and knowing the ozone  design value  for the  baseyear  and transported
ozone concentration, one could  compute the percentage control  for
hydrocarbon emissions  necessary to achieve the ozone standard.
     Application Procedure
     Figure 4-3 is  a  flow diagram  of the application sequence  for  a
rollback calculation.  The first  step is to  determine  an appropriate
design value for the baseyear.   A design value is the concentration with
expected  number of exceedances  equal  to  1.    The procedures  for
estimating a  design value from sampling data are discussed in Chapter  5
of this  guidance.  Once  the  design value  is  selected, the  total
non-methane hydrocarbon inventory for the region can be assembled  on an
annual  basis.   This  inventory is represented by an annual total tonnage
                                 7-13

-------
of emissions which  becomes the foundation  for hydrocarbon  rollback
calculations.
     The rollback model also requires estimates of transported ozone
concentrations for the baseyear.   Transported ozone concentrations in
the baseyear can be obtained from  upwind measurements or estimated from
published  sources.  Additionally,  the addivity factors for  transported
ozone are required.  Addivity factors for  transported ozone generally
range from 0.2 to 0.7;  more detailed  information can be  found in
Reference 7-9.
     With the  above-mentioned information, the relative  hydrocarbon
reduction,  i.e., % reduction, necessary to meet the oxidant standard can
be  calculated using  the  rollback  equations  (see Chapter 4 of this
guidance).   This number applied to  the baseyear emission inventory gives
the  allowable amount of  NMHC that can  be  emitted and still meet the
ozone standard for  the  baseyear and for  any  future year.  The NHMC
reduction  for the future  year is  simply the projected emissions minus
the  allowable emissions  divided  by the projected  emissions.   The
emissions control for the future year is thus identified given the total
NMHC cannot  exceed a stated quantity.
     The rollback model  is the cheapest and easiest to  use.  It is also
generally acknowledged to  be  the  least precise-   No computer is
necessary and only aggregated emissions data for the region are needed.
                                7-14

-------
                             REFERENCES


7-1  Association of Bay Area Governments,"  Application of Photochemical
     Models  in the  Development  of  State Implementation  Plans,"  3
     volumes, being prepared for  the U.S. Environmental Protection
     Agency, Office of  Air Quality Planning and Standard,  Research
     Triangle Park, N.C., 27711.

7-2  MacCraken, M.C. and G.D.  Santer, eds, "Development of an Air
     Pollution Model  for the San Francisco Bay Area," prepared for U.S.
     Energy Research and  Development Administration, Lawrence  Livermore
     Laboratory, University of California, Livermore California, Report
     No. UCRL 51983, December, 1975.

7-3  Reynolds, S.D. and  L.E. Reid,"  An  Introduction to the SAI Airshed
     Model  and Its Usage," System Applications,  Inc., San Rafael,
     California, December, 1978.

7-4  Fabrick, A., R.C.  Sklarew; et.  al., "Point Source Model Evaluation
     and Development Study," prepared for the California Air Resources
     Board and the California Energy Commission, Contract No.  A5-058087,
     Science Applications Inc., Westlake Village,  California, March,
     1977.

7-5  Sklarew, R.C., M.A.  Joncich, and  K.J. Iran, "An Operational Air
     Quality Modeling  System for Photochemical Smog in the San Diego Air
     Basin,"  paper presented at the  Annual  American  Meteorological
     Society, Reno, Nevada, January, 1979.

7-b  Environmental  Research and Technology, Inc., ERT Newsletter, No.
     1-79, Concord, Mass. 01742,  May, 1979.

7-7  Personal Communication with  Alan C. Lloyd of Environmental Research
     and Technology, Inc., Santa  Barbara, California.

7-8  "User's  Manual for Kinetics  Model  and Ozone  Isopleth Plotting
     Package," EPA-600/8-78-014a U.S.  Environmental  Protection Agency,
     Research Triangle Park, N.C. 27711, July, 1978.

7-9  "Uses,  Limitations  and  Technical  Basis  of Procedures for
     Quantifying  Relationships  between Photochemical Oxidant  and
     Precursors," EPA-450/2-77-021a,  U.S. Environmental Protection
     Agency, Office of  Air Quality Planning and Standards, Research
     Triangle Park, N.C.  27711, November, 1977.

7-10 "Procedures for Quantifying  Relationships between Photochemical
     Oxidants  and Precursors:  Supporting Documentation,"
     EPA-450/2-77-021b,  U.S. Environmental  Protection Agency,  Office  of
     Air Quality  Planning and  Standards, Research Triangle  Park, N.C.
     27711, February,  1978.
                                7-15

-------
7-11 Roth, P.M.,  S.D.  Reynolds, G.E. Anderson, R.J. Polack,  M.A.  Yucke,
     and J.P.  Killus,  "An  Evaluation of Methodologies for Assessing the
     Impacts  of Oxidant  Control Strategies," prepared for the American
     Petroleum Institute,  Report No.  EF76-112R, Systems  Applications,
     Inc., San Rafael,  California, August, 1976.
                                 7-16

-------
                        Chapter 8
  DEVELOPMENT AND ASSESSMENT
  OF OXIDANT CONTROL STRATEGIES
    This chapter discusses the procedure for developing and evaluating
alternative oxidant control strategies as part of a  non-attainment area
plan.  As required by the  Clean Air Act, an acceptable  oxidant control
plan must demonstrate that the oxidant standard will be attained and
maintained as expeditiously as possible, but no later  than 1987.  To
achieve this air quality goal, a wide range of control strategies must
be considered.    In addition to air quality,  the health, welfare,
economic, energy, and social impacts of alternatives considered must be
analyzed.  Formulation and assessment of alternative control strategies
is one of the most complex tasks in a non-attainment planning effort.
    To gain a perspective of the overall process, an overview of the
procedure for developing oxidant control strategies is shown in Figure
8-1. This overview illustrates the time sequence of several major tasks.
They are:
      • Determination of  requirements for oxidant control
      • Inventory of existing and currently scheduled control
       programs
      t Development of candidate control measures,  including
       RACM's
      • Screening the options
      • Selection and definition of alternative control strategies
      • Assessment of alternative control strategies, and
      • Recommendation of an oxidant control plan.
                             8-1

-------
 Figure s-1 Overview of Procedures for
          Developing Oxidant Control Strategy

1. Requirements
for oxidant
control



2. Inventory
of existing and
currently
scheduled
programs



3. Development
of candidate
control
measures
including
RACM



4. Screen
the options


i
r
5. Selection
and definition
of alternative •
control
strategies


i
k
6. Assessment
of alternative
control
strategies
®
               (Recycle as necessary)
o
o
D_,

i
P
7. Staff
recommended
control plan




Non-attainment
plan assembly
h


Local
adoption



State approval
and adoption




Submittal
to EPA
^

Approved
non-attainment
plan and SIP
   t
Additional
requirements of
1977 Clean Air
Act

-------
     Each of these tasks is discussed  in  the  following sections.   The
remaining tasks  shown in the Figure  (local adoption, state  approval and
adoption, etc.) will be discussed in  the  next chapter.

DETERMINATION OF  REQUIREMENTS FOR OXIDANT CONTROL
     The  initial  step in developing an oxidant control strategy is to
determine how  much control  is  needed.   The  control  requirements are
determined by a series of technical analyses including determining the
baseline  oxidant concentrations and  emission inventories  for the
baseyear  and future years,  and relating these oxidant concentrations to
the  precursor  emissions.   Control  requirements have been  discussed in
Chapters  4 through 7 of  this guidance.  Overviews of the  technical
analyses  which lead to the  determination of requirements for  oxidant
control have been shown in Figure 4-1, 4-2, and 4-3,  for the Level  1, 2,
and 3 analyses, respectively.
     Briefly, in  the Level  1  analysis, a photochemical  dispersion model
is used to relate precursor emissions to  ambient oxidant concentrations.
The requirements  for emissions (HC and/or NOX) control  are  determined by
comparing  the  predicted  oxidant concentrations with  the  oxidant
standard.   The  disaggregated emission  inventory used in the  analysis has
detailed  information for  identifying the major emission  sources,  with
spatial and temporal details as the primary targets of control.
     In the Level  2  analysis,  EKMA  is  used to determine  the necessary
emissions  (HC  and/or  NOX) reductions for  attainment of  the  oxidant
standard.  In the Level  3  analysis, the percentage reduction  of HC
emission  necessary  to attain the oxidant  standard is determined by the
rollback calculations.  The rollback model can be used to  determine the

                                 8-3

-------
control  requirement for HC emissions  only; NOX emissions are  ignored by
this method.

INVENTORY OF  EXISTING AND CURRENTLY SCHEDULED CONTROL PROGRAMS
     Prior to  actual  development of  alternative control measures, an
accurate inventory of existing and planned  programs which  affect the
ability of a non-attainment area  to attain  the oxidant standard is
necessary. Therefore, an assessment  of the effectiveness of  ongoing and
currently scheduled controls is made.   This  will  provide a  starting
point  for developing additional  measures for  further air  quality
improvement.
     The inventory should be a complete accounting and characterization
of  all  existing programs  directly  or indirectly affecting  the oxidant
concentrations  in a non-attainment area, e.g., air pollution controls,
land use policies and transportation  programs.  One direct way to obtain
the necessary information is to survey all cities, counties,  districts,
regional  agencies,  and  other political  jurisdictions  wholly or partly
within the non-attainment area.  Additionally, any applicable State and
Federal  air pollution control  rules and regulations should be included.
     In the  inventory,  a brief definition  of the existing programs
should  be provided.   Additionally, information on the program's current
and future applications,  estimated potential  effectiveness, responsible
agency,  and  implementation status and schedules should be included, if
possible.  It  is  also suggested  that all  the  existing and  planned
programs be  classified  into appropriate categories,  such as
technological control, land use management, transportation control, etc.
A complete and well-documented inventory of the existing and planned
                                 8-4

-------
programs provides a good starting  point for  development of additional
measures for consideration.

DEVELOPMENT OF  CANDIDATE CONTROL MEASURES INCLUDING  RACT'S
     After  the  Initial  Inventory of the  existing and planned control
programs 1s completed, the next task 1s to Inventory many of the still
remaining  control  measures  which might be considered for further  air
quality Improvement.  Any measures which have a potential  for reducing
or  preventing VOC and/or NOX emissions  should  be Included.
Additionally, the measures which will  result in spatial and/or temporal
redistribution  of emissions within a non-attainment area should also be
considered.  In  some cases, the measures which  are considered may in
fact already be in existence, e.g., transit service and vehicle exhaust
emission standards.  What is considered then is a further strengthening
or expansion of  the program 1n place, e.g., more  transit service or more
stringent vehicle exhaust emission standards.
     All the  reasonably  available  control  technologies (RACTs)
identified  by EPA must also be Included.  The Clean  Air Act Amendments
of 1977  require the use of  RACTs—at  a minimum—in all areas of  the
country  where  the  oxldant standard  is  being exceeded.  Up  to   the
present, EPA  has  Identified  RACTs,  also known as  Control Technology
Guidelines  (CTGs)  for 12  categories of sources.   Additional  source
categories of VOC for which EPA will  identify RACTs are listed in Table
8-1.
     In addition to  RACTs, EPA has published  a list of 19 reasonably
available transportation  control measures (RATCMs)  (see Table 8-2).  A
non-attainment plan should consider all  of the  19  RATCMs for adoption,

                                 8-5

-------
        TABLE 8-1.   FORTHCOMING CTGs FOR STATIONARY VOC SOURCES
GROUP  I*

Large  Appliance  Manufacture
Magnet Wire  Insulation
Gasoline  Bulk  Plants
Metal  Furniture  Manufacture
Petroleum Liquid Storage,
   Fixed Roof Tanks
Degreasing
Bulk Gasoline  Terminals
Petroleum Refinery  Vacuum  Systems,
   Waste Water  Separators and
   Process Unit Turnaround
Service Stations, Stage  I
 GROUP  II

 Petroleum  Refinery  Fugitive
   Emissions  (leaks)
 Surface Coating  of  Other  Metal
   Products - Industrial
 Pharmaceutical Manufacture
 Rubber Products  Manufacture
 Paint  Manufacture
 Vegetable  Oil  Processing
 Graphic Arts (Printing)
 Flat Wood  Products
 Service Stations, Stage  II
 Petroleum  Liquid Storage
   Floating Roof  Tanks
GROUP III

Ship and Barge  Transport  of
  Gasoline and  Crude  Oil
Organic Chemical  Manufacture
  Process Streams
  Fugitive (Leaks)
Dry Cleaning
Wood Furniture  Manufacture
Architectural and Miscellaneous
  Coatings
GROUP IV

Organic Chemical Manufacture
  Waste Disposal
  Storage and Handling
OTHERS

Natural Gas and Crude Oil
  Production

Adhesives

Other Industrial Surface
  Coatings

Auto Refinishing

Other Solvent Usage
*CTGs have been published  for  this  group.

Source:  Reference 8-1.
                                  8-6

-------
TABLE 8-2.  LIST OF REASONABLY AVAILABLE TRANSPORTATION CONTROL MEASURES

                           MEASURES
                    Inspection/maintenance
                    Vapor recovery
                    Improved public transit
                    Exclusive bus and carpool  lanes
                    Areawide carpool  programs
                    Private car restrictions
                    Long-range transit Improvements
                    On-street parking controls
                    Park and ride and fringe parking  lots
                    Pedestrian malls
                    Employer programs to encourage car and van pooling,
                    mass transit, bicycling and  walking
                    Bicycle lanes and storage  facilities
                    Staggered work hours
                    Road pricing to  discourage  single  occupancy  auto
                    trips
                    Controls on extended vehicle idling
                    Traffic flow improvements
                    Alternative fuels or engines and  other  fleet  vehicle
                    controls
                    Other than light duty vehicle retrofit
                    Extreme cold start emission  reduction  programs
Source:Reference «-l.
                                  8-7

-------
or docunentation must be made for not adopting  the  RATCMs.
     In short,  the guiding  principle for this  task is to develop as
complete an inventory of candidate  control measures as possible.   A
suggested  approach is  to have each of the agencies participating in the
plan development prepare a list of options for  their area of expertise
and/or responsibility. For example, the transportation  agency prepares
transportation control options, the  air pollution  control  agency
prepares  stationary  source control  options, etc.   Properly executed,
this task will  result  in a wide range of options for  the  staff and other
decision-making bodies to choose from in developing  alternative control
strategies.

SCREENING THE OPTIONS
     The purpose of this task is to screen the  inventory  of options down
to a more manageable size  for  further evaluation.   This initial
screening should be based on technical effectiveness,  applicability, and
other preliminary  criteria. For example, RACTs  for  all 12 CTG source
categories  of VOC  are  supposed to be considered in all non-attainment
plans, however, RACTs  for the source categories which  do not exist in  a
particular  non-attainment area should be eliminated from  consideration.
Information on the effectiveness of  various measures  can be found  in
EPA's CTG document series (see Appendix B). The screening primarily on
the  basis  of technical effectiveness  will  avoid the  technical  staff
making political judgments regarding an option's political  merits and
implementabil ity.  Elected officials  and  the public  could judge its
political merits and public acceptability in the later  phase of the plan
development.  This task will result in a refined  list of options  which
                                 8-8

-------
 are considered as technically effective  and applicable  for the
 non-attainment area.

 SELECTION AND DEFINITION OF ALTERNATIVE CONTROL STRATEGIES
     Up  to  this  stage of the plan  development process, the measures
.represent nothing more than "ideas".  For example,  "Ramp metering" is
 listed only as an  "idea" without any detailed information regarding the
 implementation. This task will  define the measures  in  a more detailed
 fashion.  For example, where to  apply ramp meters and when to start the
 implementation will be defined and  the effectiveness of that particular
 application will   be estimated.  The  control measures for mobile and
 stationary sources  have traditionally been direct  controls.  As such
 they  should be specified in  precise terms.  Many  of the land use
 planning and transportation measures  are Indirect controls.  Thus, they
 may be described in more general  terms.
     In the  process of defining  the detailed measures the possibility of
 establishing and implementing them is  concurrently evaluated.
 Availability of resources  necessary  to  implement and enforce the
 measure, time requirements,  and  legal authority are among the factors
 that should  be considered.  Additionally, compatability of the measures
 should also  be considered.  For example,  a  successful  vanpool and
 carpool program may  not  be considered as compatible with a measure
 calling for capital  improvement  of a mass transit system.  During the
 evaluation process, the number of feasible measures will  be reduced, and
 it is conceivable that  it may clearly Indicate  the superiority of
 certain measures over  others.  As  a result,  this  task will develop
 several preliminary packages  of alternative  control  strategies for

                                  8-9

-------
further assessment.
     Public and  local  government and  intergovernmental  coordination  is
especially important during this grouping  process.  Understanding and
agreements should be obtained regarding those measures to be eliminated
from further consideration and those measures to be included for further
assessment.  Public participation and  intergovernmental  cooperation are
discussed in detail in Chapter 3 of this guidance.

ASSESSMENT OF  ALTERNATIVE CONTROL STRATEGIES
     This task is to  conduct a  systematic  assessment  of  the various
packages of control strategies selected  in  the previous  task.  The
objective is to identify an optimum package  of control  strategies for
inclusion in  the non-attainment plan.   The  control strategies  still
under consideration at this point are effective in emission  reduction.
The next step  is  to quantify their effectiveness.  From this  a technical
ranking of the control strategies can be made.
     In  addition to  technical  effectiveness, the overall  economic and
institutional  impacts of the control  strategies must be analyzed.   Each
control  strategy will  impact the community  in  various ways.  These
impacts may be beneficial to one segment of the cormmity and adverse to
another.   As mandated by the Clean Air  Act Amendments  of 1977,  a
non-attainment plan should  include identification  and analysis  of
economic,  social,  welfare,  energy, health, and air quality effects of
the control strategies.   The final selection of control  strategies
should  be based an on overall  evaluation of all  appropriate factors.
     In the following paragraphs,  the above-mentioned technical,
economic,  institutional and  other effects of control  strategies are
                                 8-10

-------
discussed and an overall  procedure leading to an assessment of these
effects is recommended.
     Technical Effectiveness
     In  the  previous  tasks,  the  technical  effectiveness of the
individual  control strategies has been identified and  analyzed.  Based
on  that  information,  the purpose of this task  is  to  analyze the
composite effectiveness  of  various packages of control strategies.  The
estimated effectiveness of a control  package  must then be evaluated
against the requirements  for oxidant control  previously  determined.  A
selected  package of control  strategies should meet the necessary
requirements and provide  for attainment of the standard as expeditiously
as  practicable.  In the  Level 2 (EKMA) and Level 3 (Rollback) analyses,
attainment of the oxidant standard is determined by calculating whether
the  required emission  reduction is met.  In the Level 1 analysis,  the
effectiveness of a control  package is translated into the appropriate
variable or  into an adjustment of  the  emission inventory.  Then,  the
adjusted emission inventory is input to a photochemical dispersion model
to  project ambient oxidant concentrations to  determine  whether the
oxidant standard will be  attained.
     Economic Considerations
     Economic  considerations  of control strategies  should  include
cost-benefit and cost-effectiveness analyses.  There are direct  and
indirect costs of oxidant  control strategies. The direct costs include
all expenditures  required of  a  source, including capital  construction
costs, operational  and maintenance  costs, and administrative/regulatory
costs.  The indirect costs include the incremental  costs to  sectors of
economy other  than an emission  source,  such  as  price changes  for
                                 8-11

-------
materials,  changes in property value,  etc.
     The benefits of control strategies include better  protection of
health, vegetation  and materials due  to cleaner air.   Some control
strategies may generate revenues, e.g.,  increasing tolls  as  a means to
discourage  auto  driving.  Cost-effectiveness can be measured in terms of
the  cost ($)  per unit mass  (ton)  of emission  reduced or  prevented.
Since  various control strategies  will have  the same benefits—the
control of  oxidant pollution—the cost-effectiveness (dollars  per ton of
emission reduced) may serve as a  common  indicator for  ranking various
control measures.
     Not all of  the  above mentioned costs  and benefits  can  be measured
quantitatively.   The costs of many technological control measures, e.g.,
a device applied to an emission source,  a  fuel  switch,  etc. may be
quantified quite precisely.   However, for those  control  strategies
described only in general terms,  e.g., land use planning  and emission
density zoning, the cost  estimate  can only  be  qualitative.  In this
case, the assessment may have to  rely  on subjective judgment.   Cost data
sources for a number  of selected  control strategies can be found in
References  8-2 and 8-3.
     Since most control strategies will be implemented  over a period of
time, the associated costs should be computed in terms of present value.
The  resultant present values  should be converted into equivalent annual
costs  so that all  costs can  be compared on  a  common basis.   The
procedure  for cost analysis  can be found  in many engineering economics
textbooks.   The  factors (interest rate, rate  of return, etc.)  used  in
computing  the costs should  be  consistent for all  control strategies.
The total  costs of  a control  strategy should be distributed to  the
                                 8-12

-------
appropriate recipients  (consumer, owner,  government, etc.).
     Institutional  Impacts
     Institutional  impacts of control strategies should be assessed  in
terms of implementability, public acceptance (social impact), welfare,
legal  authority,  social effects and others.  These institutional  impacts
are difficult to measure in quantitative  terms.
     Public hearings  and review meetings are probably the best way  to
measure  some of  the institutional  impacts.  Public acceptability,
political merits and implementability of each control  strategy  are  best
judged by elected  officials  and  the  public whom  are likely  to be
affected by the  proposed strategies.  Adequate time should be  provided
for elected officials and the  public to review the plan or  alternative
controls.   In  the  evaluation  of public opinions  on  control  strategies,
care should be taken  to ensure that  undue weight is not given to
opinions  of special  interest groups.
     Other institutional factors  such as  the welfare costs  and social
effects  of control  strategies  are  even more difficult to assess  than
political merits  and public  acceptability.   The EPA AQMP Guideline
Series (8-2) suggests  using social indicators as  criteria  to assess the
social  effects of control strategies.  Ranking the strategies in terms
of their social  effects can  be done  by comparing the impact  of  each
strategy  upon each of the social  indicators.  The  procedures for  ranking
strategies  in  terms of their social effects are  discussed in detail  in
Reference 8-4.
     Overall Assessment
     To facilitate an  overall  assessment of various control  strategies,
a decision matrix which includes  all  the  above-mentioned technical,

                                 8-13

-------
economic,  and institutional factors  is  suggested.   Table 8-3 is an
example evaluation  matrix developed by the Washington Environmental
Research Center.  Table 8-4  is the evaluation  format  suggested in the
EPA AQMP  document series (8-2).  As a rule, the first column of the
evaluation matrix should include all the control  strategies to be
assessed;  the  evaluation  factors (cost, benefit,  social effects,  etc.)
are listed in columns 2 and  so on.
     For each control strategy, a relative score with  respect to each of
the evaluation factors is  given  in an appropriate  column.  For the
factors  which can be quantified, (costs, emission reduction, cost
effectiveness, etc.), ranking the strategies is  relatively easy.  For
the factors which cannot  be quantified, (social  effects, welfare costs,
etc.), ranking must rely upon  subjective judgment.  An  appropriate
methodology  should  be  used to weigh each of the evaluation factors;
thus,  an overall ranking of  the strategies based upon  weighed evaluation
factors can be determined.  The completed assessment  matrix can then be
carried forward  for further  evaluation.  The  public  and  various
governing bodies  of  the  political  jurisdictions  within  the
non-attainment area will probably make  the final judgment of the control
strategies.

STAFF-RECOMMENDED CONTROL PLAN
     After completion of  the overall  assessment process, an optimum
package of control  strategies  is identified  and  recommended  for
inclusion in  the  non-attainment  plan.   In  addition to other
requirements,  the recommended control plan must demonstrate  numerically
that  the  oxidant standard  would be achieved  as  expeditiously as
                                8-14

-------
                         TABLE 8-3.'      Washington Environmental Research Center (KERC) Matrix












Implemen-
tation
Tool




in
in
«
*
*j
u
o
«+H
U4
(U
•N»





g
*J
u
1
•»

ECOI
1 - J
5:
Direct
Costs

t.
0)
1-1
1— 1
o
9U
£


*+
y
•f.
§
^



1
&



0

OMIC CONSIDERATIONS
•tost Expensive
foderately Expensive
.east Expensive
Admin.
Costs
4-»
io
i
o
o.


4-1
u
J-
5
tS


00
Monitori



1
0

Social
Costs

i
u
c
a
1

I
0
i— i
f-
C3
1

8
•rt
U
C.
Regional



1

Social
Responsiveness

•
3
0.
u
4-1
i

















*



f%
Ij
O.
O
U
s
4-1
2
o

Administrative
Considerations
Control
Application

4-1
«-> in
in rt
X i
l-t •
•i.CM
*,=?
o
•a en
o» i
VI
1-4 4J
« -y.
00—

P
i o
w 3
l!-S
.Slg
4-»j U
«S'2



Selective



Uniform

00
I
_J

en
  Strategy 1


  Strategy 2


  Strategy 3
 Source:  Reference 8-1.

-------
  TABLE  8-4.   FORMAT  OF  DECISION MATRIX SUGGESTED  IN  EPA AQMP GUIDELINES
                             Arti of KJNA nhere
                             Pollutant
                                                     Deduction or redistribution required.
Measure
Land Use and Planning
1. [Mission allocation
2. Regional devtlO»»*nl
planning
J. Eaittston density lontm
4. Ion ing approval
5. Transportation control
6. E«
-------
practicable, but  no  later than 1987.  A  summary of the recommended
control  plan should be prepared, and supporting technical data and the
analysis and decision processes  should be documented and retained  for
future reference.
     In the recommended  plan,  a concise  description of the actions
should be  provided.  In addition, the plan should include the agencies
responsible  for implementation,  the implementation  schedule, the costs
of the plan,  sources of  finances, the direct benefits of the plan  (in
terms  of emission reductions), and  other relevant  environmental,
institutional, financial,  economic, and social  information.
     The recommended control  plan and the other additional requirements
of the Clean Air Act will  then be assembled  as  a complete non-attainment
plan.  This  plan will subsequently  go through  local  adoption, state
approval  and adoption, and EPA approval processes.

REFERENCES
8-1  U.S.  Environmental  Protection Agency, "Workshop on Requirements for
     Non-Attainment Area Plans  - Compilation of  Presentations,"  workshop
     held  in San Francisco,  California, March, 1978.
8-2  U.S.  EPA "Guidelines  for Air Quality Maintenance  Planning and
     Analysis,  Volume 2:   Plan Preparation,"  publication  no.
     450-4/74-002.
8-3  U.S.  EPA, "Guidelines  for Air Quality Maintenance  Planning and
     Analysis,  Volume 3:   Control  Strategies,"  publication no.
     450-4/74-003.
8-4  Klee, A.J., "The  Role  of Decision Models  in the  Evaluation of
     Competing Environmental  Health Alternatives."  Management Science,
     Journal  of the  Institute  of Management  Sciences,  18, No. Z,
     October, 1971.
                                8-17

-------
                         Chapter 9
         PLAN REVIEW, ADOPTION
         AND APPROVAL PROCESS
    The Clean Air Act requires that the  State Implementation Plan be
adopted by  a  state after a public hearing.  The Act does not specify a
local process for adoption of non-attainment plans.  As with other parts
of the  air  quality planning  program,  the nature  and schedule  of the
local approval process may vary from region to region, depending on the
geographic size of the non-attainment planning area,  the number of local
jurisdictions  in the area, and the number and  authority of  other
agencies involved in the planning program.

PROVIDING ADEQUATE TIME FCR PUBLIC REVIEW
    EPA  regulations (4C CFR Part 61)  require that any draft of a major
implementation plan revision (including any of six elements listed in
51.244  of the regulations) shall be submitted to the  state  and areawide
clearinghouses for review and comment.  The review-and-comment period is
established  at 45 days, and comments received during  that time are to be
considered before an SIP revision is adopted.
    As  a practical matter,  experience  in preparing  the 1979 plan
submissions  suggests that a 45-day  review-end-comment  period  is not
adequate.   To ensure effective and informed decision-making,  review
periods  of  60-120  days or  longer are likely.   If control  programs
recommended are likely to be highly  controversial,  requests will
probably be  made for more public review time.   Schedules  for 1982 plan
revisions should provide  for  some margin toward  the end of the plan
                              9-1

-------
revision process for schedule changes to  accommodate, where appropriate,
public requests for extensions in the review period.

AVAILABILITY  OF REVIEW DOCUMENTS
     Plan revision documents should be made available well  in advance of
public hearings.  Where  possible,  non-technical summaries should be
prepared.  All material  should  be  readily  available to interested
parties, and  local libraries and  other  public locations can be used
effectively  in ensuring access to information about what is recommended
and why.

DEMONSTRATION OF REASONABLE FURTHER PROGRESS
     The 1977 Clean Air  Act Amendments  provide an important concept
missing from the 1970 Act—the requirement  for  "reasonable further
progress." Section 171 defines the term  "reasonable further progress"
as annual  incremental reductions in emissions of the applicable air
pollutant--(including substantial reductions in the  early years
following approval  or promulgation of plan provisions under Part D and
Section 110(a)(2)(I) and regular reductions thereafter) —sufficient,  in
the judgment of the EPA Administrator, to provide for attainment of the
applicable national ambient air quality standard by  the date required in
Section 172(a).
     Section  172(b)(3) provides that the  "non-attainment" SIP submittal
shall  require,  in the interim,  reasonable further progress  (RFP) (as
defined above), including  such reduction  in  emissions from  existing
sources in the area  as may be obtained  through  the adoption,  at  a
                                 9-2

-------
 minimum of, reasonably  available control  technology.
     Non-attainment plans should include an RFP schedule.   The  purpose
 of an RFP schedule  is to verify that the  emission reductions  obtained
 are  being accomplished at a reasonable and  efficient rate so that
 attainment by the  prescribed time  will  take place.   Where only
 violations of the annual primary  NAAQS are involved,  RFP is intended to
 represent  the  application  of  effective controls as  expeditiously as
 practicable.   RFP  schedules  to  correct violations  of the secondary
 standards  are  again  intended to depict a timetable for  compliance within
 "a reasonable time."   Where  attainment of the  oxidant standard can not
 be accomplished by 1982, then  an RFP schedule reflecting  application of
 reasonably available  control measures  will  be acceptable as long as
 attainment is  accomplished  by  the agreed upon date (not to exceed 1987).
     For the majority  of  cases, EPA expects that RFP  schedules will be
 linear.  A linear RFP schedule  is thought by  EPA to be  completely within
 the intent of the  Section 171 definition.   That is,  such  a line would
effect substantial  reductions in  the  early going and not  put  off
 accomplishing most of the required reductions until the  last year(s).  A
linear schedule  is also thought  to  be a  stringent timetable  to
 accomplish emission reductions in  light of  the lag time for installing
control  technology.
     In terms of answering the 1979 plan requirement to show RFP,  the
submission of  a linear  RFP  schedule  will be acceptable  even  if  initially
 a linear  timetable  would be thought inappropriate.  Unless  the state or
region  on  its  own chooses to submit  a detailed nonlinear RFP,  a  linear
 schedule  for  1982  attainment  SIPs will be acceptable  to  EPA.  Such a
                                 9-3

-------
schedule Is thought to reflect a reasonable rate of progress considering
the relatively  short time frame allotted for attainment (maximum of 4 to
5 years).  To  routinely require a detailed  RFP may direct critical
resources from accomplishing the actual attainment.  Only in certain
cases where the controls necessary for attainment cannot be reasonably
applied as quickly as required by  a linear RFP would it be in the best
interest of a state or region to develop and submit a "less-than-linear"
RFP schedule.
     The linear assumption is initially acceptable for all 1979  plan
submittals.   Of course, some of the plans, particularly  those attaining
standards after 1982, may have to amend their  schedules over the  year
following the 1979 submittal  to make  RFP more meaningful.  Where the
linear  assumption is thought  to be  inappropriate, an  acceptable
alternative RFP  schedule must be  submitted  to  the Regional
Administrator.   Such nonlinear schedules should be developed over  the
year following  the  1979  plan submission.   EPA acknowledges  that
developing such RFPs is more resource intensive and should be attempted
only when absolutely necessary.
     A non-linear  RFP will  typically be developed in the  case of oxidant
plans that  fail to  attain by 1982 but do  so not later than 1987.
Herein,  the  aggregate nature  of the  emission  inventory must be
considered.   Individual  RFP schedules for each gener-1 component of the
1977 emissions  inventory (i.e.,  mobile, major stationary, and minor/area
sources)  must be developed.   For post 1982 oxidant SIPs,  reductions  from
mobile sources  until  1983 are expected to follow as a  minimum a  schedule
consistent with the  Federal Motor Vehicle Control Program (FMVCP)  and
                                 9-4

-------
the application of RACMs.  Beyond 1982 until  attainment, the reductions
from mobile sources will reflect the imposition of other measures (as
necessary)  as well as the continuing cleanup  accomplished by the FMVCP.
Major  stationary sources correspondingly must be controlled until  1982
at least  as  effectively as they would under an  RFP schedule exhibiting
the application of RACT.   After 1982 and until  attainment (no later  than
1987)  the schedule for reductions will  reflect  the application of other
control  measures.   A linear assumption  can be made for the portion of
the RFP  schedule beyond 1982.   Finally,  a linear assumption Is
appropriate throughout  the RFP schedule  for controlling minor and  area
(other  than mobile)  source  growth.

COMPLIANCE WITH EPA  RACT  MEASURES
    The  Clean Air Act Amendments of 1977  require the use of reasonably
available control technology—at a  minimum—1n all  areas of the country
where a standard  is  being exceeded.   EPA has identified reasonably
available control  technologies (RACTs),  also known as control  technology
guidelines  (CTGs),  for  12 categories  of  sources.  Controls  for
additional source  categories are forthcoming.
    Local  districts and  air quality  planning organizations should
review their regulations and, by comparing the regulations with the CTG
requirements, determine  if all  RACT  measures are  being  implemented.
Where  they  are  not,  local agencies should proceed  promptly  with
development of regulations to comply with  the 1977 Clean  A1 r  Act
requirements  for implementation  of  all  reasonably available control
technologies.
                                 9-5

-------
DEMONSTRATION  OF LEGAL, FINANCIAL AND MANPOWER COMMITMENT TO
IMPLEMENTATION
     This requirement  is  established in  S. 172  of the  act.  The
requirement  itself is a principal  reason why  implementing agencies
should be directly involved in the planning  process and adoption  of the
non-attainment plan.   The Act requires that the SIP  be legally
enforceable.  For example,  stationary  source controls are primarily
implemented  by local  or regional districts by regulation.   A
demonstration of commitment  to  implement  stationary source control
measures occurs only when the local district agrees to carry out the
measures and it has the authority to do  so.   Similar  demonstrations for
other  control strategies can only occur as  part of the plan  development
and approval process.   For  actions assigned  to the state, demonstrated
commitment to  implement control measures—as opposed to expressions of
willingness to carry out control  measures made by state agencies prior
to  local  adoption--can only be reasonably  expected to occur after the
plan is approved locally and submitted to the state.
     For actions for which  there does  not yet exist legal authority,
demonstrated commitment  can be reasonably expected to consist of:  a)  a
willingness to carry  out a reasonably available control measure,  and b)
a willingness to seek legal authority.  Actions assigned by locally
adopted plans to the State  but for which no  legal  authority to implement
the measures  exist require that  the State seek legal  authority, or
provide  adequate  documentation of the infeasibility of the proposed
action.  Preferably  this  documentation  should  occur  during  the
review-and-comment period prior to  local   adoption,  so the plan  may be
                                9-6

-------
 revised.  Otherwise  the  locally adopted plan must  be  revised at the
 state level  or  by a local-state  negotiation process.
     Financial  and manpower commitments are dealt with using a similar
 line of reasoning.  Unless an agency demonstrates lack of financial and
manpower commitments,  it can  reasonably be expected  that  such
commitments  are  available if an agency  agrees to carry out a control
measure.

REFERENCES

9-1   Environmental  Protection Agency,  Workshop on Requirements for
     Nonattainment Area Plans, revised  edition, April
 Requii
,  1978.
                                9-7

-------
                          Chapter 10
THE CONTINUING  PLANNING PROCESS

     The continuing air quality planning process should logically follow
 from the process  and work accomplished  during the initial planning
 period.  Two key tasks to be carried out during the  continuing  planning
 process are to:   1) respond to  issues raised during state and federal
 review of the  1979  plan submittal  when  such issues cannot be
 satisfactorily  resolved prior to plan approval at  the state or federal
 level, and 2) monitor implementation of initial plan control measures to
 ensure that reasonable further  progress (described in  Chapter 9) is
 achieved. The first key task is self-explanatory; the second  deserves
 further explanation.

 TRACKING REASONABLE FURTHER PROGRESS
     RFP is to be  tracked primarily by a yearly assessment of the net
 reductions to actual emissions (including growth) that have taken place.
 A summary of  these reductions will  be submitted as an annual report.
 The S.  171 definition of RFP is non-specific  as to what  type of
 emissions  inventory should be tracked and reported on an  annual basis.
 For several reasons, however, EPA  believes  that the  intent of  Congress
 was to monitor  yearly trends in actual  emissions to assess  if RFP is
 being accomplished.  Section  172(b)(4) requires  emission inventory
 development on an actual emissions basis  and  sufficient revisions
 thereafter to  enable a determination regarding  RFP  to be made.  The
 tracking of actual emissions also  seems to be reasonable in that air
                               10-1

-------
quality trends  are best linked to changes in  this emissions inventory.
Finally,  actual  emissions are also important since they are most likely
the basis  for determining what controls are needed for cleanup.
     The annual  report should summarize how total actual emissions  have
been  reduced  relative to  the applicable RFP milestones.  Only where
special non-linear RFPs are applicable will  the annual  report be  more
specific  than  reporting  total  actual  emissions (i .e., mobile, major
stationary,  and minor/area).  Although not formally required, a state or
local agency might also report or at least keep track of how maximum and
allowable  emissions  have changed over the same time period.
     Section 171 defines RFP in terms of annual emission reductions and
does not mention changes in air  quality directly.   EPA's position  is
that air  quality  changes should  be taken  into  account, at least
qualitatively,  in determining if RFP is being  accomplished.  That is,
the intent of reviewing air quality changes with the associated emission
reductions is to ensure that the reductions accomplished are meaningful
in terms of actual air quality improvement.
     The annual  report for RFP or the air  quality  trends data are
expected  to identify  from  time to  time  the existence of some control
strategy or implementation problems.   In fact, this  is  precisely the
intended  function of RFP.  That is, the RFP  report attempts to identify
problems so  that they can be resolved without jeopardizing attainment by
the prescribed  date.
     Three types  of  problems can occur.  The  first is simply failing  to
obtain  sufficient  actual  emission reductions to meet the specific RFP
milestone.  Such a  problem could  be the result of excessive growth
                                 10-2

-------
 (major and/or minor area  growth), enforcement  problems, or the lag time
 to install control  equipment.
     The  second  problem identified  is  poor air quality  improvement
 despite the accomplishing of all the emission reductions called  for
 under RFP.  Conceivably, this problem  could be caused by the persistence
 of unusual  meteorological  conditions, overestimating the value  for
 cleaning up  certain emissions sources  (bad emission factors),  or the
 omission  of key  sources from the  inventory altogether.   The latter more
 specifically refers to the lack of control on  key sources which have not
 been  included in the initial  emissions inventory  developed  for the
 non-attainment area.
     The third and final  potential problem is the  inability of maximum
 emissions  to  decline at an acceptable rate even though actual emissions
 have been reduced  in accordance with the RFP schedule.   This problem
 will likely be noticed by  only those states  or local  areas tracking
 trends  in maximum  emissions.  The  problem  may occur  due  to delays in
 converting SIP allowable emission  limits that  far exceed what a  source
 could physically emit to more realistic emission rates or  the failure to
 constrain  the operating hours and/or capacity  use  of  key sources  as
 appropriate.
     In general, no sanctions will be imposed  when  problems such  as
 those mentioned  are  identified  if  a justification is  presented that
demonstrates  that attainment by the prescribed date  is not jeopardized.
 However,  in  many cases some  additional  corrective action(s)  may be
 necessary.
                                 10-3

-------
WORK PROGRAM SUBMITTALS
     The 1979 plan  submlttals,  plus  issues raised in  the  state/federal
approval  process, should provide  adequate task and schedule detail  for
submittal  of  work programs for  the first round of S.  175  grants.  Each
year's previous work,  plus information gained from tracking  RFP  and
changes  in EPA requirements, should be used  in  the succeeding work
programs. The  key to the work,  however, will be the 1982  plan  submittal
and the  necessity  to keep  the program  on  track to meet the  plan
submittal deadline.
                                10-4

-------
                              Appendix A
                   Bibliography of Urban System Models

     Urban system modeling  1s a primary tool  for predicting the size and
distribution of land use activities 1n an  urban area;  the modeling
results  provide .necessary Information  for  estimating  mobile  and
area-source emissions.   If  a photochemical  dispersion model  Is  to  be  used
as a primary tool  1n  a non-attainment planning  analysis, a  region wide,
grldded  emission Inventory Is  needed.   In  this case,  urban  system
modeling  Is the  best  way to generate the urban  activity data  necessary
for  spatial  and  temporal  resolution of emissions.  In  a  non-attainment
planning  effort, urban system modeling, If required, 1s usually conducted
by regional planning  agencies.
     Urban activity data required to project  emissions Include  the levels
and distribution of land use, population, employment,  housing,  and travel
activities.  A system capable of producing the  necessary Information Is
generally composed of the following types of  models:
       •  Regional  Projection Models
           -  Regional Demographic Model
           -  Regional Econometric Model, etc.
       •  Activity Allocation (Zone Level) Models
           -  Basic Employment Model
           -  Protective Land Use Model (PLUM)
           -  Urban System Model  (USM)
           -  Empiric  Model, etc.
       •  Travel  Demand Forecasting Models
           -  Trip  Generation Model
           -  Trip  Distribution Model
           -  Modal  Split Model
           -  Network  Assignment Model, etc.
                                 A-l

-------
     The  following is a compilation of several  of the most  widely used

models.   A number of special-case models  (i.e., the models  which  cannot

be easily applied to other non-attainment areas) are not included in the

list.
                      REGIONAL PROJECTION MODELS
ABA6 Regional  Demographic Model (APPLE);  ABAG Regional Econometric  Model:

AQMP/Tech Memo 2, "Projections/Forecasting:   System Description and
Technical  Assumptions," Association of Bay Area Governments,  Berkeley,
December,  1976.

Simple Economic  Model:

Kogiku, K.C.,  "An  Introduction  to  Macroeconomic Models,"  New  York,
McGraw-Hill, 1968.

Input-Output Models:

Miernyk,  W.H.,  "The Elements of Input-Output Analysis," New York,  Random
House, 1967.

Other Models:

Comprehensive Planning  Organization,  "IPEF73,  Interactive  Population
Employment Forecasting  Model, Technical User's  Manual," San Diego,
California, March, 1974.

Puget Sound Governmental Conference, "Population and Employment Forecasts
and  Distributions  for  the Central  Puget  Sound Region,  1975-1990,"
Seattle, Washington, 1972.

Lockfeld, F.M., "Short Term Projection of House Values, Rents  and  Income
for Small  Areas:  The  Santa Clara  County Housing Projection Model,"
Monterey,  California, 1973.

Southwestern Pennsylvania Regional  Planning  Commission,  "Forecasting
Framework:  JOBS, PEOPLE, and LAND," Pittsburgh, PA, 1974.

Metropolitan Council, "Urban Activity Forecast Study Design,"  St.  Paul,
Minnesota, 1971.
                                 A-2

-------
 ,_____, A Summary  of the Urban Systems Model. McLean,  Virginia,
 January, 1974.   	"

 Access and Land Development Model (ALP):

 Creighton, Hamburg,  and Associates,  Inc., et. al.. Summary Final Report -
 Access and Land  Development  (ALP), prepared  for Chicago  Area
 Iransportation Study, September, 19/4., Praft CATS Publication.

 ,^__^	,  The Access and Land Development Model Technical Report,
 Volume I  -"Computer Software Documentation and Base Year Data  Assembly,
 Draft CATS  Publication.	   	~~~

               .  The Access and  Land Development Model Technical Report,
 YoVume  II  -  TheTalibration of  the  ALP Travel  Function,  Draft
 Publication.

               , The Access  and  Land Development Model  Technical Report,
Volume III - Calibration and Evaluation of the ALP  Equilibrium Model,
Praft CATS Publication.

              , The Access and Land Development Model Technical Report,
Volume IV  - Development and Testing of the ALP  Dynamic Model, Draft  CATS
Publication.

Land Use Allocation Model  (LUAM):

Computer Sciences Corporation, LUAM - Land Use  Allocation Model for Urban
and Transportation Systems Simulations.

Mahonlng - Trumben Counties Comprehensive Transportation and Development
Study Land Use  Allocation Model, prepared for Mahonlng - Trumbell
Counties  Comprehensive Transportation  and Development Study of Ohio,
reprinted  for Eastgate Development  and Transportation  Agency, by  E.S.
Preston  Associates, Inc.,  July, 1973.

Land Use Plan Design Model:

Southeastern  Wisconsin Regional  Planning Commission, A Mathematical
Approach to Urban Design,  Technical Report No. 3, Waukesha, Wisconsin,
1966.

              , A Land Use Plan Design Model - Model Development. Volume
            Report No. 8, Waukesha, Wisconsin, 1968.	

	   A Land Use Plan Design Model_ -  Model  Test. Volume  2,
Technical  Report No. 8, wauicesha, Wisconsin, 1969.

          	, A Land Use Plan Design Model - Final  Report. Volume  3.
Technical  Report No. 8, wauicesha, Wisconsin, 1973.

validation and Test of a Land Use Plan Design Model.  Department of Civil
Engineering,  Marquette University,  Milwaukee, Wisconsin, May, 1973.
1,  Technica
                                 A-3

-------
                ACTIVITY ALLOCATION (ZONE LEVEL)  MODELS
EMPIRIC Activity Allocation Model;

Peat,  Marwick,  Mitchell, & Co.,  EMPIRIC Activity Allocation Model:
Summary.  Washington, D.C., April 1971.

	.  EMPIRIC Activity  Allocation Model:   Final  Report.
prepared  for the  Metropolitan  Washington  Council  of Governments.
Washington, D.C., May 1972.
	        ,  EMPIRIC Activity  Allocation Model  User's  Manual,
prepared  for the  Federal Highway  Administration  (FHWA).Washington,
D.C., January,  1974.

Protective Land Use Model (PLUM):

Goldner,  W., Protective Land Use  Model, Institute  of Traffic and
Transportation  Engineering,  University of California,  Berkeley, 1968.

Goldner, W.,  et al, Protective Land Use Model:  Volume 1  - Plan Making
With A Computer Model,  Volume 2  - Theory and Application, Volume 3 -
Computer Systems GuTae, prepared by University of  California for  FHWA,
Urban Planning  Division, 1972.

   	, Economic and Spatial Impacts of Alternative Airport
Sites and Locations In the San Francisco Bay Region,  Institute of Traffic
and Transportation Engineering, University of California, Berkeley,  1971.

                 PLUM/SD The Urban Development  Model:   Volume  1 - A
Regional Planning Tool, Volume 2 - Technical  Manual,  Volume 3 - Computer
Systems Guide,  prepared by San Diego Comprehensive  Planning Organization
and  Institute  of  Traffic and Transportation Engineering, University of
California, 1973.

Urban Systems Model (USM);

Turner, C.G., The Development of  an Activity Allocation  Model for the
Bristol Sub-ReigToh, Urban Systems  Research Unit, University of Reading,
England, 1970.

Alan M. Voorhees & Associates, Inc.,  Application  o.  the  Urban Systems
Model (USM) To  A Region  - North  Central  Texas, Volume 1;  Conceptual
Basis of the  USM And Application to The Planning Process, prepared for
the  North  Central Texas Council  of Governments,  McLean, Virginia,
October, 1972.

               , Application  of the Urban Systems Model  (USM) to a Region
-  North Central Texas,  Volume II:  Data Base,  Calibration  Procedure and
Results, prepared for  the North Central  Texas Council  of Governments,
McLean, Virginia, December, 1972.
                                A-4

-------
 NBER Urban Simulation Model:

 Ginn,  J.  Royce, "The NBER Prototype  Urban Simulation Model,"  paper
 presented at  the  Convention of the American  Institute of Planners,
 October, 1971.

 Ingram, G.K., Ka1n, J.F., and 61nn, J.R.,  The Detroit Prototype of the
 NBER Urban Simulation Model, New  York:   National Bureau  of  Economic
 Research, 1972.	

 Integrated Transportation and Land Use Models Package (ITLUP):

 Putman, Stephen H.,  University of Pennsylvania, The Interrelationships of
 Transportation Development  and Land Development,  volumes I and  II,
 prepared for the U.S. Department of Transportation, FHWA, Urban Planning
 Division, June, 1973.
                     „-*»>,'•« • •
                   .i_ -J^""^i-1fc-'"
               , "Further Results From, and Prospects for Future Research
 With,  the Integrated Transportation and Land Use Model  Package  (ITLUP),"
 presented at the Annual  Conference of the  Southern Regional  Science
 Association,  April 3-4, 1975,  Atlanta, Georgia.

               , "Calibrating a Disaggregated  Residential Allocation
 Model  -  DRAM:  The Residential  Submodel of the Integrated Transportation
 and  Land Use Package - ITLUP,"  presented  at  the  Regional  Science
 Association  - British Section, Eighth Annual Conference, September 4-5,
 1975, London,  England.
                     TRANSPORTATION SYSTEM MODELS
Trip Generation Models;

"Guidelines  for  Trip  Generation  Analysis," Federal  Highway
Administration, Report No.  HHP-22, Washington,  D.C.,  June, 1967.

"Trip Generation Analysis," Federal Highway  Administration, Report No.
HHP-20,  Washington, D.C., August, 1975.

"Trip Generation by Land Use," Maricopa Association of  Governments,
Marlcopa County, Arizona, April, 1974.

"Trip Generation," Institute  of Transportation  Engineers, 1976.

Trip Distribution Models;

"Calibrating  and Testing a Gravity  Model  for any Sized  Urban Area,"
Federal  Highway Administration, Report No. HHP-20, Washington  D.C '
October, 1965.
                                 A-5

-------
"Urban  Trip  Distribution  Friction  Factors,"  Federal  Highway
Administration, Report No.  HHP-20, Washington,  D.C., November, 1974.

"Computer  Programs for Transportation Planning  -  PLANPAC/BACKPAC General
Information,"  Federal Highway Administration,  Report No. HHP-20,
Washington,  D.C., 1977.

Model  Split  Models;

"Modal  Split, Documentation of Nine Methods  for  Estimating Transit
Usage,"  Federal Highway Administration, 1966.

"Introduction  to  Urban Travel Demand Forecasting,"  Urban Mass
Transportation Administration, NTIS-236-848/AS, March, 1974.

"Applications  of  New Travel Demand  Forecasting  Techniques to
Transportation Planning - A Study of Individual Choice Models," Federal
Highway  Administration, March, 1979.

"Estimating  Auto Occupancy,  A Review of Methodology,"  Federal Highway
Administration, 1972.

Network  Assignment Models:

"UTPS Network  Development Manual," Federal  Highway Administration,  Covers
Network  Assignment Computer Package  Including ULOAD  (Transit), LOADVN
(Highway), CAPRES (Highway  Capacity Restraint), and others.

"Traffic Assignment," Federal  Highway Administration, Report No. HHP-20,
August,  1973.

"Computer  Programs for Transportation Planning  -  PLANPAC/BACKPAC General
Information Manual," Federal  Highway Administration, Washington, D.C.,
1977.
                          GENERAL REFERENCES
Wilson, A.G., Urban and Regional Models in Geography  and  Planning.  New
York:  John  Wiley and Sons,  1974.

Yupo Chan, et al, A Review of Operational  Urban Transportation Models,
prepared  for  IT.S.  DOT, Transportation  System Center, Peat,  Marwick,
Mitchell  and Co., Washington, D.C., 1973.

Peat,  Marwick,  Mitchell  &  Co., _A_R_eview  of Operational  Urban
Transportation Models, Report No.  DOT-TSC-496, prepared for U.S. DOT,
Transportation System Center.   Cambridge,  Massachusetts, April,  1973.

U.S. EPA,  "A Guide to Models  in Governmental  Planning & Operations," U.S.
EPA, Office  of Research and Development, Washington, D.C., 1974.
                                A-6

-------
 U.S.  DOT,  An  Introduction to  Urban Development Models  and Guidelines for

 ineir  Use in  Urban  Transportation  Planning,"  Federal  Highway

 Administration, October, 1975.



 fl'iT* I.n*roduction  to Urban Travel Demand Forecasting," Federal Highway

 Aomimstration and Urban Mass Transportation Administration, 1977.



 piffivi MV, "Urban Modellin9 - Algorithms, Calibrations, Predictions,"
 Cambridge University  Press, Cambridge,  1976.



c?fnk«   Jn R<>  "ur?an Models:  Diffusion  and Policy Application,"  Regional
Science Research Institute, Philadelphia,  1978.                   "»•«"«,
                               A-7

-------
                               APPENDIX B

              EPA CONTROL TECHNOLOGY GUIDELINE DOCUMENTS*
 Control of Volatile Organic  Emissions from Existing  Stationary  Sources
 Volume V:  Surface Coating of Large Appliances - EPA-450/2-77-034.

 Control of Volatile Organic  Emissions from Existing  Stationary  Sources
 Volume  IV:   Surface  Coating  for  Insulation of  Magnetic  Wire -
 EPA-450/2-77-033.

 Control of  Volatile Organic  Emissions from  Bulk  Gasoline Plants -
 EPA-450/2-77-035.

 Control  of Volatile  Organic Emissions from Existing Stationary Sources
 Volume III:  Surface Coating of Metal  Furniture  -  EPA-450/2-77-032.

 Control  of Volatile Organic Emissions from  Storage of Petroleum Liquids
 In Fixed Roof Tanks -  EPA-450/2-77-036.

 Control  of Volatile  Organic  Emissions  from Solvent Metal Cleaning -
 EPA-450/2-77-022.

 Control of Hydrocarbons  from Tank Truck Terminals  - EPA-450/2-77-026.

 Control of Refinery Vacuum Producing Systems,  Wastewater Separators, and
 Process Unit  Turnarounds - EPA-450/2-77-025.

 Control of Volatile  Organic Compounds  from Use of  Cutback Asphalt.

 Control  Volatile Organic  Emissions  from  Existing Stationary Sources
 Volime II:  Surface  Coating of Cans, Colls,  Paper, Fabrics, Automobiles,
 and Light-Duty Trucks  - EPA-450/2-77-008.

 Design Criteria  for  State I Vapor Control Systems - Gasoline Service
 Stations.

 Control  of Volatile  Organic Emissions  from Existing Stationary Sources
 Volume  I:   Control  Methods  for Surface-Coating  Operations
 EPA-450/2-76-028.
*Reproduced from  "workshop on Requirements for Non-Attainment Area
 Plans—Compilation of  Presentations," U.S. Environmental  Protection
 Agency, March,  1978.
                                 B-l

-------
EPA 450/2-73-002    A  Technique  for  Calculating  Overall Efficiencies of
                   Particulate Control Devices.   8/73.

EPA 450/2-74-002A   Background  Information for New Source Performance
                   Standards:  Primary Copper, Zinc, and Lead Smelters.
                   Vol. 1 - Proposed Standards.   10/74.

EPA 450/2-74-008    Air Pollution Control  Engineering and Cost Study of
                   the Ferroalloy Industry.  5/74.

EPA 450/2-74-009A   Background  Information on  National Emission Standards
                   for Hazardous Air Pollutants,  Proposed Amendments to
                   Standards for Asbestos and Mercury.   10/74.

EPA 450/2-74-017A   Background Information for Standards of Performance:
                   Electric Arc Furnaces  in the Steel Industry.  Vol.  1
                   -  Proposed  Standards.   10/74.

EPA 450/2-74-017B   Background Information for Standards of Performance:
                   Electric Arc Furnaces  in the Steel Industry.

EPA 450/2-74-018A   Background  Information for Standards of Performance:
                   Electric  Submerged Arc  Furnaces  Producing
                   Ferroalloys.  Vol. 1:   Proposed  Standards.  10/74.

EPA 450/2-74-018B   Background  Information for Standards of Performance:
                   Electric Submerged Arc Furnaces for  Production of
                   Ferroalloys.  Vol. 2:   Test Data Summary.  10/74.

EPA 450/2-74-018C   Background  Information for Standards of Performance -
                   Electric Submerged Arc Furnaces for  Production of
                   Ferroalloys.  Vol. 3 - Supplemental  Information.
                   4/75.

EPA 450/2-74-019A   Background Information for Standards of Performance:
                   Phosphate Fertilizer  Industry.   Vol.  1 - Proposed
                   Standards.  10/74.

EPA 450/2-74-019B   Background Information for Standards of Performance:
                   Phosphate Fertilizer Industry.   Vol.  2:   Summary of
                   Test Data.  10/74.

EPA 450/2-74-020A   Background Information for Standards of Performance:
                   Primary  Aluminum Plants.  Vol.   1  -  Proposed
                   Standards.  10/74.

EPA 450/2-74-020B   Background Information for Standards of Performance:
                   Primary Aluminum Plants.  Vol.  2:   Summary of  Test
                   Data.  10/74.

EPA 450/2-74-020C   Background Information for Standards of Performance:
                   Primary Al urn in urn  Industry.  Vol. 3  -  Supplemental
                   Information.  1/76.
                                  B-2

-------
 EPA 450/2-74-021A   Background Information  for Standards of Performance:
                    Coal  Preparation  Plants.   Vol.  1:    Proposed
                    Standards.  10/74.

 EPA 450/2-74-021B   Background Information  for Standards of Performance:
                    Coal Preparation  Plants.  Vol. 2 -  Summary of Test
                    Data.   10/74.

 EPA 450/2-74-021C   Background Information  for Standards of Performance:
                   Coal Preparation Plants.  Vol.  3 - Supplemental
                    Information.   1/76.

 EPA 450/2-75-009   Standard Support  and Environmental  Impact Statement -
                   Emission Standard for Vinyl Chloride.  10/75.

 EPA 450/2-75-009B  Standard Support  and Environmental  Impact Statement:
                   Volume  2  Promulgated Emission  Standard  for Vinyl
                   Chloride.   1976.

 EPA 450/2-76-002   State Implementation Plan  Emission Regulations for
                   Sulfur Oxides - Fuel  Combustion.  3/76.

 EPA 450/2-76-012   Field Evaluation of Reid Jacket Vapor Control  System.
                   1976.

 EPA 450/2-76-014A  Standards  Support and Environmental  Impact  Statement:
                   Yol. 1  Proposed  Standards of Performance  for Kraft
                   Pulp Mil Is.  1976.

 EPA 450/2-76-016A  Standards  Support and Environmental  Impact  Statement:
                   Vol  . 1  Proposed Standards  of  Performance  for
                   Petroleum  Refinery Sulfur Recovery  Plants.   1976.

 EPA  450/2-76-028   Control  of Volatile Organic Emissions from  Existing
                   Stationary Sources:   Vol. 1  - Control  Methods for
                   Furnace Coating Operation.   1976.

 EPA 450/2-76-030A  Standards  Support and Environmental  Impact Statement:
                   Vol. 1  -  Proposed  Standard of  Performance  for
                   L1gn1te-F1red Steam Generators. EPA:  OAQPS 1976.

 EPA 450/2-77-001A  Standards  Support and Environmental  Impact Statement:
                   Vol. 1  - Proposed  Standards  of Performance for the
                   Grain Elevator Industry.  EPA:  OAQPS  1977.

EPA 450/2-77-005   Control  of Fluoride Emissions from Existing Phosphate
                   Fertilizer Plants:   Final  Guideline  Document.  EPA:
                   OAQPS 1977.

EPA 450/2-77-007A  standards Support and Environmental  Impact Statement:
                   Vol. 1  - Proposed  Standards of Performance for Lime
                   Manufacturing Plants.   EPA:  OAQPS 1977.
                                 B-3

-------
EPA 450/2-77-008    Control  of Volatile Organic Emissions  from Existing
                   Stationary Sources:  Vol.  II  - Surface Coating  of
                   Cans,  Coils, Paper, Fabric,  Automobiles and Light
                   Duty Trucks.  EPA:  OAQPS 1977.

EPA 450/2-77-019    Final Guideline  Document:  Control  of Sulfuric  Acid
                   Mist Emissions from Existing Sulfuric  Add Production
                   Units.  EPA: OAQPS 1977.

EPA 450/2-77-022    Control of Volatile Organic  Emissions  from Solvent
                   Metal Cleaning.

EPA 450/2-77-025    Control  of Refinery Vacuum Producing Systems -
                   Wastewater Separators:  Process Unit Turnarounds.

EPA 450/2-77-026    Control  of Hydrocarbons from Tank  Truck Gasoline
                   Loading Terminals.

EPA 450/2-77-032    Control  of Volatile  Organic Emissions - Surface
                   Coating Vol. III.

EPA 450/2-77-033    Control of Volatile Organic  Emissions  -  Magnetic Wire
                   Vol. 4.

EPA 450/3-73-003A   Emissions Control  in  the Grain  and  Feed Industry,
                   Vol. 1. Engineering and Cost Study.  12/73.  Midwest
                   Research Inst. 1973.

EPA 450/3-73-003B   Emissions Control  in  the Grain  and  Feed Industry.
                   Vol. 2.  Emission Inventory.   9/74.  Midwest Research
                   Inst. 1974.

EPA 450/3-73-004A   Air Pollution Control 1n  the Primary Aluminum
                   Industry.   Vol. 1  of  2 (Sections  1  through 10).
                   7/73.  Singmaster and Breyer.   1973.

EPA 450/3-73-004B   Air Pollution Control in  the Primary Aluminum
                   Industry.  Vol.  2 of 2  (appendices).   7/73.

EPA 450/3-74-002    Evaluation of the Controllability  of  Power  Plants
                   Having a Significant Impact  on  Air Quality Standards.
                   2/74.

EPA 450/3-74-015    Factors  Affecting  Ability  to Retrofit Flue Gas
                   Desulfurization  Systems.  12/73.   Radian Corp.  1973.

EPA 450/3-74-036A   Investigation of Fugitive Dust:  Vol. I - Sources,
                   Emissions, and  Control.  PEDCo  Env. Specialists.
                   1974.

EPA 450/3-74-036B   Investigation of Fugitive Dust:  Vol. II - Control
                   Strategy  and Regulatory Approach.  PEDCo Env.
                   Specialists.   1974.
                                 B-4

-------
EPA 450/3-74-060   Air Pollution  Control  technology and  Costs - Seven
                  Selected Emission  Factors.   12/74.   Indust. Gas
                  Cleaning  Inst.  1974.

EPA 450/3-74-063   Participate  Emission  Control  Systems  for Oil-Fired
                  Boilers.  12/74.  Geomet. 1975.

EPA 450/3-75-046A  A Study of Vapor Control  Methods for  Gasoline
                  Marketing Operations.  Vol. 1 - Industry Survey and
                  Control Techniques.  4/75.  Radian Corp.  1975.

EPA 450/3-75-046B  A Study of Vapor Control  Methods for  Gasoline
                  Marketing Operations.   Vol. 2 - Appendix.  4/75.
                  Radian Corp.  1975.

EPA 450/3-75-047   Comparison of  Flue Gas  Desul furl zatlon  Coal
                  Liquefaction  and  Coal  Gasification for  Use  at
                  Coal-Fired Power Plants.  Kellogg MS Co. 1975.

EPA 450/3-76-005   Control of Partlculate  Matter from 011 Burners  and
                  Boilers.  Aerotherm  Corp.  1976.

EPA 450/3-76-013   Cost of Retrofitting Coke Oven Partlculate Controls.
                  Vulcan Cincinnati.  1974.

EPA 450/3-76-036   Evaluation of Methods for Measuring and Controlling
                  Hydrocarbon  Emissions from Petroleum Storage Tanks.
                  Battelle Memorial Inst.  1976.

EPA 450/3-76-038A  Background Information on Hydrocarbon Emissions from
                  Marine Terminal  Operations:   Vol. I  - Discussion.
                  Radian Corp.  1976.

EPA 450/3-76-038B  Background Information on Hydrocarbon Emissions from
                  Marine Terminal  Operations:  Volume II - Appendices
                  Radian Corp.  1976.

EPA 450/3-76-042   Economic  Impact of Stage  II  Vapor  Recovery
                  Regulations:  Working Memoranda.  Little Ad.  1976.

EPA 450/3-77/-010  Technical Guidance  for Control of Industrial Process
                  Fugitive Partlculate Emissions.   PEDCo  Env.
                  Specialists.  1977.

EPA 450/3-77-026   Atmospheric Emissions  from Offshore Oil  and  Gas
                  Development and  Production,  Energy Resources  Co.
                  1977.

EPA 450/3-77-046   Screening Study to Determine  Need for  SOx  and
                  Hydrocarbon NSPS for FCC Regenerators.
                                 B-5

-------
                                    TECHNICAL REPORT DATA
                             (Please read Inslructions on the reverse before completing)
 1. REPORT NO.

  FPA-45n/?-7Q-nnia
                              2.
 4. TITLE AND SUBTITLE
  Example Control  Strategy for Ozone,  Volume I:
  General Guidance to Nonattainment  Areas
                                                            3. RECIPIENT'S ACCESSIOr»NO.
              5. REPORT DATE
               April,  1979
              6. PERFORMING ORGANIZATION CODE
 7. AUTHOR(S)
                                                            8. PERFORMING ORGANIZATION REPORT NO
                                                              OAQPS No. 1.2-120
 9. PERFORMING ORGANIZATION NAME AND ADDRESS
                                                            10. PROGRAM ELEMENT NO.
  Association of Bay Area Governments
  Environmental  Quality and Energy Resources Department
  Hotel  Claremont
  Berkeley,  California  94705
              11. CONTRACT/GRANT NO.
               68-02-3001
 12. SPONSORING AGENCY NAME AND ADDRESS

  U.S.  Environmental Protection  Agency
  Office  of Air Quality Planning and Standards
  Research  Triangle Park, North  Ca/olina  27711
              13. TYPE OF REPORT AND PERIOD COVERED

               Final	
              14. SPONSORING AGENCY CODE
 15. SUPPLEMENTARY NOTES
  Association of Bay Area Governments Project Manager:
  EPA  Project Officer:  Andrew  Creekmore
            Eugene  Leong
 16. ABSTRACT
      This  guideline presents  information to assist States  and local agencies  in
 preparing  ozone control strategies  for nonattainment areas.   The guidance should
 be most  useful  in preparing 1982  State Implementation  Plan revisions.

      The guideline covers the following topics:  Intergovernmental Cooperation,
 Development and Assessment of Air Quality and Emissions  Data, Modeling to Related
 Air  Quality to Emissions, Control  Strategy Analysis and  Assessment, Plan Adoption,
 and  the  Continuing Planning Process.   This volume covers general guidance to
 nonattainment areas.
 7.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b IDENTIFIERS/OPEN ENDED TERMS
                           c. COSATI Field/Group
  Air  Pollution
  Atmospheric  Contamination Control
  Ozone
 State implementation
 plan control strategy

 National ambient  air
 quality standard
 8. DISTRIBUTION STATEMENT
                                               19. SECURITY CLASS (This Report)
                                               Unclassified
                           21. NO. OF PAGES
                               139
           Unlimited
20. SECURITY CLASS (Thispage)

 Unclassified
                                                                         22. PRICE
EPA Form 2220-1 (9-73)

-------