EPA-450/3-77-028
September 1977
               DEVELOPMENT
          OF COMPUTERIZED
      EMISSION PROJECTION
   AND ALLOCATION SYSTEM
      PHASE II: COMPARISON
       OF EXISTING SYSTEMS
  V.S. ENVIRONMENTAL PROTECTION AGENCY
      Office of Air and Waste Management
    Office of Air Quality Planning and Standards
   Research Triangle Park, North Carolina 27711

-------
                                  EPA-450/3-77-028
DEVELOPMENT OF COMPUTERIZED
     EMISSION PROJECTION AND
         ALLOCATION SYSTEM

      PHASE II: COMPARISON OF
           EXISTING SYSTEMS
                       by

             Richard R. Cirillo and George A. Concaildi

           Energy Research and Development Administration
                Argonne National Laboratory
             Energy and Environmental Systems Division
                  9700 South Cass Avenue
                  Argonne, Illinois 60439
               Interagency Agreement No. D7-0077
              EPA Project Officer: Joseph Sableski
                    Prepared for

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

                    September 1977

-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers.  Copies are
available free of charge to Federal employees, current contractors and
grantees,  and nonprofit organizations  in limited quantities  from the
Library Services Office (MD35) , Research Triangle Park, North Carolina
27711;  or,  for a fee, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency by
Energy Research and Development Administration, Argonne National
Laboratory, Energy and Environmental Systems Division, 9700 South
Cass Avenue,  Argonne, Illinois, in fulfillment of  Interagency Agree-
ment No. D7-0077.  The contents of this report are reproduced herein
as received from Energy Research and Development Admini stration .
The opinions,  findings, and conclusions expressed are those of the
author and not necessarily those of the Environmental Protection Agency.
Mention of company or product names is not to be considered as an endorse-
ment by the Environmental Protection Agency.
                   Publication No. EPA-450/3-77-028
                                  11

-------
       ill
TABLE OF CONTENTS

1.
2.























3.




4.
5.






INTRODUCTION 	
ANALYTICAL REQUIREMENTS OF A CEPA SYSTEM 	

2.1.1 Residential Fuel Combustion 	
2.1.2 Commercial/Institutional and Industrial
Fuel Combustion 	
2.1.3 Electric Generation 	

2.2 Process Emissions 	
2.3 Solid Waste Disposal 	

2.4.1 Highway Vehicles 	
2.4.2 Other Vehicles 	
2.4.3 Gasoline Handling Evaporation Losses 	


2.5.2 Fires 	
2.5.3 Fugitive Dust 	
2.5.4 Other Sources 	 	 	
2.6 Gridding 	
2.6.1 Calculational Procedure 	
2.6.2 Master Grid Development 	
2.7 Growth 	

2.9 Growth Tracking System Capability 	
COMPUTER CONFIGURATION OF THE CEPA SYSTEM 	
3 .1 Constraints on Hardware and Software . 	
3.2 Hardware Alternatives 	
3.3 Software Considerations 	

COMPARISON PROCEDURE FOR ALTERNATIVE CEPA SYSTEMS 	
COMPARISON OF EXISTING SYSTEMS 	
5.1 The AQUIP System 	 	
5.1.1 System Description 	
5.1.2 System Use 	
5.1.3 Comparison with CEPA Requirements 	

Page
..... 1
.... 3
.... 4
.... 7

.... 12
.... 12
.... 16
.... 16
.... 21
.... 23
.... 23
.... 26
.... 27
.... 28
.... 28
. . . . 28
.... 28
. . . . 29
. . . . 29
. . . . 29
. . . . 31
. . . . 31
. . . . 33
.... 34
. . . . 37
. . . . 37
. . . . 38
. . . . 39
.... 41
.... 43
. . . . 47
.... 47
. . . . 47
.... 54
.... 54
.... 57

-------
     5.2   The  CAASE  System
          5.2.1   System Description
                                      iv


                          TABLE OF CONTENTS  (Cont'd)
                                                                         Pag

                                                                          57
          5.2.2   System Use .......................
          5.2.3   Comparison with CEPA Requirements ...........  6^
          5.2.4   Required Modifications .................  66

     5.3   The ESAQ  System  .......................  67
          5.3.1   System Description  ..................  67
          5.3.2   System Use  ......................  1}-
          5.3.3   Comparison with CEPA Requirements ...........  71
          5.3.4   Required Modifications  ..... . ...........  73

     5.4   The MWCOG System .......................  73
          5.4.1   System Description  ..................  73
          5.4.2   System Use. .... ..................  77
          5.4.3   Comparison with CEPA Requirements ...........  77
          5.4.4   Required Modifications .................  7^

6.  COST  ANALYSIS OF THE CEPA SYSTEM ..................  79
     6.1   System Development Costs ...................  '"
          6.1.1   Modification and Development Resource Requirements.  .  .  79

     6.2   System Installation and Application  .............  88
          6.2.1   Training   .......................  88
          6.2.2   System Support  ....................  90
          6.2.3   Potential System Use ..................  92

7.  CONCLUSIONS  AND RECOMMENDATIONS ................... 103
     7.1   System Evaluation ....................... 103
          7.1.1   AQUIP ......................... 103
          7.1.2   CAASE ......................... 104
          7.1.3   ESAQ  ......................... 104
          7.1.4   MWCOG ......................... 105
          7.1.5   New CEPA System .......... ' .......... 106
     7.2   System Use .......................... 106

     7.3   Alternative Courses of Action ................. 107
          7.3.1   No Further Action ................... 107
          7.3.2   Modify AQUIP,  CAASE, OR MWCOG ............. 107
          7.3.3   Initiate New System Development ............ 107
          7.3.4   Modify ESAQ .....................  '. 1Q8
          7.3.5   Proceed with Stepwise Modification of ESAQ ....... 108
     7.4   Summary ............................
APPENDIX A - Detailed Evaluation of the AQUIP System .......... Ill

APPENDIX 3 - Detailed Evaluation of the CAASE System .......... 130
A??i:7DIX C - Detailed Evaluation of the ESAQ System  .......... 149

-------
                         TABLE OF CONTENTS (Cont'd)
                                                                         Page
APPENDIX D - Detailed Evaluation of the MWCOG System	169
APPENDIX E - Development Effort of a New CEPA System	187
ACKNOWLEDGMENTS	189
REFERENCES	190

                              LIST OF FIGURES
2.1  Limits of CEPA System	    5
2.2a Computational Flow for Residential Fuel Combustion Sources
     (Levels 1 and 2)	    8
2.2b Computational Flow for Residential Fuel Combustion Sources
     (Level 3) .  . .	    9
2.3a Computational Flow for Commercial/Institutional and Industrial Fuel
     Combustion Sources (Levels 1 & 2)	   13
2.3b Computational Flow for Commercial/Institutional and Industrial
     Fuel Combustion  Sources  (Level 3)	   14
2.4  Computational Flow for Industrial Process Sources 	   17
2.5  NEDS Point Source Coding Form	   19
2.6  Computational Flow for Solid Waste Disposal Sources 	   22
2.7  Computational Flow for Transportation Sources  (Highway Vehicles).  .   24
                                                    9
4.1  Comparison Procedure for Existing Computer Systems	   44
5.1  Flowchart of AQUIP System	   48
5.2  Flowchart of CAASE System	   59
5.3  Flowchart of ESAQ System	   68
5.4  Flowchart of MWCOG System  	   75
6.1  Range of CEPA Cost Savings Estimates	   99
6.2  Analyses Required to Recover Investment 	  101

                                LIST OF TABLES
2-1  National Emissions Report Format	    6
2-2  Data Available for Residential Fuel Combustion Sources	   10
2-3  Data Available for Commercial/Institutional and Industrial
     Fuel Combustion Sources	   15
2-4  Data Available for Industrial Process Sources  	   18
2-5  Data Available for Transportation (Highway Vehicles) Sources.  ...   25
5-1  AQUIP System Elements	   50

-------
                                     vi
                           LIST OF TABLES (Cont'd)
                           	                                Page

5-2  Objective Apportioning Factors Current Area Source	    "^
5-3  Objective Apportioning Factors New Area Source 	    °2
5-4  Components of the MWCOG System	    ?6
6-1  Summary of Modification and Development Efforts for CEPA System.  .    80
6-2  Summary of Staff Costs	    85
6-3  Summary of Resource Requirements 	    86
6-4  Resource Requirements for Training Workshops 	    89
6-5  Resource Requirements for System Support 	    91
6-6  CEPA System Cost Summary	    93
6-7  Activities Required for Emission Projection and Allocation ....    94
6-8  Control Agency Personnel Costs 	    96
6-9  Cost of Emission Projection and Allocation	    97
6-10 Summary of CEPA Cost Savings	    98

-------
                               1  INTRODUCTION

       This report documents the second phase of a feasibility study to de-
termine the need for a computerized emission projection and allocation  (CEPA)
system to assist state and local air pollution control agencies in conducting
air quality analyses.
       The possible need for a CEPA system came as a result of informal dis-
cussions with agencies and individuals conducting analyses required to conform
to air quality maintenance planning regulations.  It appeared that the calcu-
lation procedures, although relatively straightforward, were long and tedious
and might be consuming an inordinate amount of resources to perform.  At the
same time, it was evident that such a system would have possible applications
in other types of air quality analyses.
       The determination of need for a CEPA system is being carried out in a
3-phase feasibility study.  The Phase I effort  focused on the potential demand
for a CEPA system based on a series of interviews with control agency staff.
The results were somewhat mixed in that there was no clear cut and definitive
demand on the part of the agencies for such a system.  At the same time, all of
the agencies surveyed expressed some interest in the system and indicated they
would consider using it to assist in their analyses.  On the basis of these in-
conclusive results, it was decided to proceed to the Phase II effort to review
existing systems to determine if any or all of the CEPA requirements could be
met without the need for an entirely new system development.  The results of
Phase II are reported here.
       The Phase II evaluation procedure is carried out on four existing air
quality analysis systems: the Air Quality for Urban and Industrial Planning
system (AQUIP), the Computer-Assisted Area Source Emission Gridding Procedure
(CAASE), the Engineering-Science Air Quality system (ESAQ), and the Metropoli-
tan Washington Council of Governments model (MWCOG).  The methodology involves
a description of the CEPA requirements without reference to  any  existing  systems,
a comparison of existing packages to those requirements, an identification of
deficiencies, an estimate of effort required to remove those deficiencies, an
evaluation of the effort needed to develop an entirely new system, and an
assessment of the potential savings to be realized by employing a CEPA system
in place of manual procedures.

-------
       Upon a decision to proceed with the acquisition  of  a  CEPA system,  the
Phase III effort will concentrate on the  preparation  of a  system specification
document for procurement  purposes.

-------
                2  ANALYTICAL REQUIREMENTS OF A CEPA SYSTEM

       The CEPA system must be designed to function in the analysis of a vari-
ety of air quality management problems.  The situations in which CEPA will be
operated include the following:
       1.  Periodic analyses of areas to determine whether air quality
           standards will be violated in the future due to growth in
           emissions and hence whether revisions are needed to the state
           air quality implementation plans.  These periodic analyses
           are required under 40 CFR 51.12(h)(2).
       2.  Evaluation of the impact on air pollutant emissions of strat-
           egies that are designed to control the magnitude, timing, or
           location of new emissions.  The results of this evaluation can
           be used in air quality dispersion models to estimate air pol-
           lutant concentrations and thus determine whether a strategy is
           adequate to attain and maintain a national ambient air quality
           standard.
       3.  Analysis of the air quality effect of different land use plans
           and system level transportation plans.
       4.  Assessment of the direct air quality impact of large scale projects
           such as the provision of sewers or highways.
       5.  Evaluation of the effect of new sources of air pollutant con-
           centrations to determine whether the new sources will violate
           an air quality standard or a significant deterioration increment.
       6.  Development of environmental impact statement assessments.
       7.  Assistance in the implementation of an emission offset policy
           in non-attainment areas.
       8.  Evaluation of air quality impact of alternative economic and
           energy policies.
       9.  Incorporation of air quality considerations in to other long-
           term planning efforts such as EPA's Section 208 waste water
           management planning, HUD's Section 701 comprehensive planning,
           and Coastal Zone Management planning.

-------
        Despite the rather large set of  applications,  the CEPA system will per-
 form a relatively limited set of tasks  that  are  crucial to  each of the anal-
 yses but are far from complete in terms of the entire scope of each effort.
 CEPA will be limited to three basic tasks:   (1)  receiving current information
 on emission sources, population and housing,  economic activity and employment,
 land use, transportation and other planning  data and  translating this into
 gridded point and area source emissions for  use  in  a  dispersion model, (2)apply-
 ing the results of a growth analysis to the  above information and generating
 gridded point and area source emission  for future years,  (3)  applying control
 strategies and generating emissions subject  to the  various  control scenarios.
 Figure 2.1 illustrates the extent of the CEPA system.
        It is evident from this structure that there are several things CEPA
 is not.  These include the following considerations:
        1.  CEPA is not a growth analysis package.   Studies  and pro-
            jections of growth are done  externally.  CEPA only applies
            these projections to the existing  information base.
        2.  CEPA is not a data management system  although it will, of
            necessity,  have to be designed for ease  of data  manipula-
            tion.   Long-term storage and access to data is externally
            provided.
        3.   CEPA is  not an  air pollutant dispersion  model.   It only
            generates output  in a format compatible  with input require-
            ments  of models.
        With these considerations,  it  is possible to outline the kinds of  com-
putations  CEPA must be  able  to  perform.  For  convenience, these are discussed
in  terms  of  the emission source  categories affected using the National Emis-
sions Report  (NER) format  shown  on  Table 2-1.

2.1  FUEL  COMBUSTION SOURCES
       Fuel combustion  sources  to be  handled  by  the CEPA system can be grouped
into 5 basic categories:  residential,  commercial/institutional,  industrial,
electric generation, and internal combustion.  All  of  these categories  can  be
made up of both point and area  sources,  each  of  which  is  handled differently
in both the computational and data  handling routines.  Point  sources will all

-------
                                                                            	 CEPA SYSTEM
EMISSION


r •
1
1
1
1
1
\PLANNING / 1
DATA /
/ 1


— 	

GROWT
APPLICA
j



hi
TION





	 ^
•»••«•• ^m^m
]
.
FMTSSTDN 1 ^
COMPUTATION '
i
^•™»» a^MM
1
1
1
1
STRATEGY
"* APPL1CAI10N 1
_J

                                       GROWTH
                                 _^1 ANALYSIS
STRATEGY
ANALYSIS
                                                                                                   GRIDDED
                                                                                                   POINT
                                                                                                   AND AREA
                                                                                                   SOURCE
                                                                                                   EMISSIONS
                                       Fig.  2.1.   Limits  of CEPA System

-------
                 Table 2-1.  National Emissions  Report Format
         Category
Subcategory
 Fuel Combustion: External
 Fuel Combustion: Internal
 Industrial Process
 Solid Waste Disposal
                                   Government
                                   Residential


                                   Commercial/Insti-
                                   tutional
                                   Industrial
Transportation
Miscellaneous
                                  Solvent Evapor-
                                  ation

                                  Fires
                                  Dust Caused By
                                  Human Agitation
                                  Of The Air
                                  Airborne Dust
                                  Caused By Nat-
                                  ural Winds
     Residential  Fuel
     Commercial/Institutional
     Industrial
     Electric  Generation

     Commercial/Institutional
     Industrial
     Electric  Generation
     Aircraft  Engine Testing

     Chemical  Manufacturing
     Food/Agriculture
     Primary Metals
     Secondary Metals
     Mineral Products
     Petroleum Industry
     Wood Products
     Process Evaporation
     Metal Fabrication
     Leather Products
     Textile Manufacturing
     Inprocess Fuel
     Other/Not Classified

    Municipal Incineration
     Open Burning
    Other

    On-Site Incineration
    Open Burning

    On-Site Incineration
    Open Burning
    Apartment
    Other

    On-Site Incineration
    Open Burning
    Auto Body Incineration

    Highway Vehicles - gasoline, diesel
    Off-Highway Vehicles - gasoline,
      diesel
    Aircraft
    Vessels - railroad, ship
    Gasoline Handling Evaporation Loss

    Industrial Sources
    Dry Cleaning

    Structural
    Frost control
    Slash Bruning
    Wild Forest
    Agricultural
    Unpaved Roads
    Unpaved Airstrips
    Paved Roads
    Mineral Processing
    Tilling Activities
    Loading Crushed Rock, Sand, Gravel
    Construction

    Storage Piles
    Tilled  Land
    Unfilled Land

-------
be handled through an interface with the emission inventory be it in National
Emission Data System (NEDS) or other format.  Area source information is much
more diverse and requires a greater range of computational alternatives.

2.1.1  Residential Fuel Combustion
       Figures 2.2a and 2»2b show the flow of the calculations necessary to
compute emissions from residential fuel combustion sources.  Table 2-2 lists
the potential sources of data available.  The two different approaches shown
result from two basically different types of information used as the starting
point.  The CEPA system must be able to handle both procedures.
       The Level 1,2 analyses (these are the same level designations used in
Ref. 2) of Fig. 2-2a start with two basic pieces of information: state or
county fuel consumption in the residential sector and a distribution of popu-
lation or dwelling units by state, county, and subcounty areas (e.g., census
tracts, municipalities, planning districts, etc.).  This distribution should
be a resident data file since it will be used for other portions of the analy-
sis.  The first step of the computation is to distribute the fuel consumption
to the subareas using the population or dwelling unit distribution.  The fuel
consumption is also calculated as a total heat energy (Btu) consumed for use
in growth and strategy analyses later.  Next, the fuel consumed by point
sources in the emission inventory is extracted since these sources are handled
separately.  The result of these steps is a data file containing residential
area source fuel consumption by subarea and by fuel type and heat energy total.
The next step is to map the fuel consumption by subarea into fuel consumption
by master grid (i.e., the grid network that is used as input into a dispersion
model).  For the basic CEPA system, it is only necessary to have the areal
fraction of the subarea in the master grid cell; that is, the master grid is
developed externally and only the mapping of subareas into grid cells is nec-
essary as input.  The implications of this are discussed later.  The master
grid fuel consumption file and a data file containing emission factors are
used to compute master grid residential fuel combustion emissions.  These cal-
culations, combined with other emissions in the master grids, are used as dis-
persion model input.
       There are two ancillary sets of calculational procedures that a CEPA
system must have available in addition to the basic calculational stream just

-------

                       COMPUTE
                       COUIITY 8,
                       SUBAREA
                       RESIDENTIAL
                       FUEL CONSUMP
                       TION, BTU
                       TOTAL
                                                            EXTRACT
                                                            POINT
                                                            SOURCES
                                                                             RESIDENTIAL
                                                                             FUEL CON-
                                                                             SUMPTION BY
                                                                             SUBAREA
  COUNTY t,
  SUBAREA POP
    D.U. FRAC
  TION
                        BTU,
                        % SULFUR,
                        VALUES OF
                        FUEL
STATE & CNTY
 SUBAREA
  POP. OR
   D.U.
                   POP., D.U.
                   GROWTH PRO-
                   JECTIONS, 'I.
                   GROWTH, ACTUAL
                   VALUES BY
                   STATE, CNTY,
                   SUBAREA
                                                                           LEGEND
                                                                           BASIC CALCULATIONS
                                                                                                                                     oo
                                                                           GROWTH  ANALYSIS  ----

                                                                           CONTROL  STRATEGY -- — • -

                                                                           INPUT                 \ J

                                                                           CALCULATION           |    \
                                                                            DATA FILE
F1S-  2.2a.   Computational  Flow for  Residential  Fuel  Combustion Sources
                (Levels  1 & 2)

-------
                                                                                                                                             STATE, CNTY, SUB
                                                                                                                                              AREA FUEL CON-
                                                                                                                                               SUMPTION
                     FUEL CONSUMP-
                     TION FACTORS
                     BY D.U., FUEL
                     SIZE. FLOOR
                     SPACE, LAND USE
STATE, CNTY, SUB-
 AREA O.U., FUEL
 USE DIST..SI2E
  OIST.,FLOOR
   SPACE, LAND
      USE
                   RESIDENTIAL
                   FUEL CON-
                   SUMPTION BY
                   SUBAREA
                                                                                  MASTER GRID
                                                                                  RESIDENTIAL
                                                                                  FUEL CON-
                                                                                  SUMPTION
                                                                                                                 MAPPING
                                                                                                                 FRACTIONS
                                                                                                                 OF SUBAREA
                                                                                                                 TO MASTER
                                                                                                                 GRID
BTU, -, ASH
", SULFUR,
VALUE OF
FUELS
WEATHER DATA
(DEGREE DAY)
       CONTROL STRATEGIES
       GROWTH i DEVELOPMENT
MASTER GRID
RESIDENTIAL
EMISSIONS

                                                                         LEGEND

                                                                         BASIC CALCULATIONS

                                                                         GROWTH  ANALYSIS

                                                                         CONTROL STRATEGY

                                                                         OPTIONAL FEATURE

                                                                         INPUT

                                                                         CALCULATION

                                                                         DATA FILE
                                                                                                                                                                   CU
                  Fig.   2.21),    Computational  Flow  for  Residential Fuel  Combustion  Sources   (Level  3)

-------
                            Table  2-2.   Data Available  for  Residential Fuel Combustion  Sources
                       I'OJJU I d t ioil
                       Number of dwelling units  (ii.u.)
                       Number of d.u. using

                        UliliLy gas
                        Fuel oil or kerosene
                        Coal or coke
                        Wood
                        Electricity
                        Bottled gas
                        Oilier fuel
                        No fuel

                    tor space heating
                        water heating
                        cooking

                       Number of d.u. in structure

                        1, 2. 3 and 4,  5-49, 50+
Spatial
Disaggregat ion
All data: census tract,
county, state, and other
Form
Available
Computer tape,
hard copy
Date of
Information
Every 10 years for
full set of data.
General
Availability
Virtually entire U.S.
with some exceptions
                                   miscellaneous  aggregations
                                   of the tract data
                                             Latest is 1970.
                                             Interim updates of
                                             selected data or
                                             areas sometimes
                                             available.
l\L-giundl  or Local
Planning  Agency
All or some of the above
informat ion
Floor space (sq.  ft.)  of
residential buildings
Land area devoted to
residential use
Growth projections

  population
  dwelling units
  land use
Regional planning  dis-
tricts.   May  or may not
be coincident with
tracts,  size  range
highly variable
Dependent on
agency.   May
or may not be
machine read-
able.
Latest planning
cycle.
Variable.   Most  likely
in major metropolitan
areas.
Kuel Dealers
                      All or some of the above
                      information
                      Actual fuel consumed  by
                      residential customers
                                                         Variable with dealer.
                                                              Generally hard     Latest data
                                                              copy but occa-     collection
                                                              sionally machine-  cycle
                                                              readable.
                                                                                                                             Variable
State Agencies
Fuel consumed  by  residential
customers
Generally by county
or for entire state
Hard copy
Latest year of
statewide data
collection plus
possibly pro-
jections
Generally available
       Agencies
Fuel consumed  by  residential
users
By state
                                                                                     Hard copy
                   Latest year  of
                   data collection
                   (usually 1-2
                   year lag)
                                                                                                                              Entire  U.S.

-------
                                      11
described.  These are a growth routine and a strategy analysis.  The growth
routines must be able to take input in the form of % growth or real growth
values of population and/or dwelling units from an exogenously performed
growth analysis.  This information will then be used to modify the data files
on state, county, and subarea population and dwelling unit distributions to
generate future distributions.  The future fuel mix must also be included as
part of the data the growth package must handle.  With the new distribution
and new fuel mix, the basic calculational stream can be repeated to generate
master grid emissions for a future year.
       The control strategy routines must be able to handle three types of
control options.  The traditional emission limit regulations are modeled by
changing the emission factors just prior to the master grid emission computa-
tion.  Fuel controls (such as sulfur content limitations or prohibition of
certain fuel types) are simulated by changing the fuel characteristics and
/or the future  fuel mix.  Growth and development controls are modeled by chang-
ing the population or dwelling unit distributions.
       The Level 3 analysis differs from Level 1,2 with respect to the infor-
mation used in  the initial calculations.  Instead of beginning with a state
or county fuel  consumption that is distributed to subareas, a set of surrogate
variables such  as dwelling units, floor space, residential land use, or others
is combined with fuel consumption factors in related units (e.g., fuel con-
sumed per dwelling unit, per acre of land used, etc.), and weather data to com-
pute fuel use by subarea.  The state or county totals computed this way can be
cross-checked against actual data and the fuel consumption factors adjusted
appropriately.  The remainder of the calculations including the growth and
strategy analyses are identical to Levels 1,2.  In this level of analysis, pro-
vision should also be made to input subarea fuel consumption totals directly
(collected, for example, from interviews with local fuel dealers).  The use of
the surrogate variables is, in this case, for growth and control strategy use
only.
       The need for the CEPA system to handle the Level 1, 2 and the Level 3
analyses illustrates an important design philosophy for the entire package.  The
simplified as well as the more sophisticated procedures must be built into the
system to enable a wide variety of users to operate it.  Likewise, the possi-
bility of using more than one type of surrogate variable  (e.g., population,
dwelling units, land use, etc.) is necessary to cover the range of data avail-
ability.

-------
                                      12
2.1.2  Commercial/Institutional and Industrial Fuel Combustion
       Figures 2.3a and b illustrate the computational scheme for commercial/
institutional and industrial fuel combustion.  Table 2-3 shows the sources of
input information.  The calculations are entirely analagous to those that are
performed for residential fuel combustion with some small differences.  In the
Level 1,2 analyses the statewide fuel consumption must be disaggregated into
commercial/institutional and industrial fuel use.  This information may already
be available from the basic information or some estimate may have to be made.
Also, the distribution function whereby the statewide fuel use is mapped into
subarea fuel use is made up of employment, land use, or other data as opposed
to population and dwelling units in the case of residential fuel combustion.
In the Level 3 analysis the different surrogate variables are the only point
of difference from the residential calculations.
       Because of the high degree of similarity between residential and commer-
cial/institutional and industrial sources, it is entirely possible that the
same computational modules can be used for all three source categories.

2.1.3  Electric Generation
       For the most part, emissions from the generation of electricity are
traceable to large centralized power plants.  These facilities are treated as
point sources and would be handled by the CEPA system in a manner entirely
analagous to the Industrial Process emissions discussed in the next section.
In the case of power plants, information on new facilities and on plant retire-
ments is available from numerous sources including the utilities, the Electric
Reliability Council, the National Coal Association, and others.   The CEPA sys-
tem should be able to process this data in the same way as data on industrial
facilities.
       There is one aspect of the electric generation that is slightly differ-
ent and that a CEPA system should be designed to handle:  that is,  the proiec-
tion of the demand for electricity.   This information may be available from
federal, state, or utility energy planning studies.  The demand  for electric-
ity must be translated into the load factors of the various power plants serv-
ing the study region.  It might also be useful to have the CEPA system check
the demand against the available capacity.   Excess demand will have to be
by the purchase of power from interconnected utility grids.   The CEPA svste

-------
                                      COMPUTE CNTV 8
                                      SUBAREA FUEL CON-
                                      SUMPTION IN COMM/
                                      INST. 8 INDUS.
                                      SECTORS.BTU TOTAL
  RESIDENTIAL
  COMM/INST.,
  INDUSTRIAL
  DISTRIBU-
  TION
                                                              COMM/INST,
                                                              INDUSTRIAL
                                                              FUEL CON-
                                                              SUMPTION
                                                              BY SUBAREA
COUNTY 8 SUB-
AREA DISTRIBU-
TION FUNCTION:
EMPLOYMENT,
LAND AREA,ETC.
, CNTY AND SUB-
  AREA SURRO-
   GATE VAR-
    IABLES
                                                                COMPUTE FUEL
                                                                CONSUMPTION
                                                                IN MASTER
                                                                GRIDS
                                        MASTER GRID
                                        COMM/INST 8
                                        INDUSTRIAL
                                        FUEL CON-
                                        SUMPTION
          EMPLOYMENT,
          LAND USE,
          FLOOR SPACE,
          GROWTH PROJ
          % GROWTH,
          ACTUAL VAL-
          UES BY
          STATE, CNTY,
          SUBAREA
 !
A
                                               GROWTH
                                               ANALYSIS
CONTROL STRATEGY

FUEL CONTROLS

  GROWTH 8 DEVEL ,
    PLANS

   EMISSION
   LIMITS
                                                                  MASTER GRID
                                                                  COMM/INST &
                                                                  INDUSTRIAL
                                                                  EMISSIONS
                                                                                                UJ
                                                         LEGEND

                                                         BASIC CALCULATIONS

                                                         GROWTH  ANALYSIS

                                                         CONTROL STRATEGY

                                                         INPUT

                                                         CALCULATION
                                                                                                                      DATA FILE
              Fig.  2.3a.   Computational  Flow for  Commercial/Institutional  and  Industrial
                                       Fuel  Combustion Sources  (Levels  1 & 2)

-------
                                                   COMPUTE CNTY
                                                   8 SUBAREA
                                                   FUEL CONSUMPTION
                                                   BUT TOTAL
                                    FUEL  CON-
                                    SUMPTION
                                    FACTORS
                                    BY SURRO-
                                    GATE  VAR-
                                    IABLE
                          EXTRACT
                          POINT
                          SOURCES
WEATHER  DATA
(DEGREE  DAY)
                                                    CROSS-CHECK
                                                    FUEL USE
f-NPLOYMENT,
1AND
Fl MR SPACE ,
GROUTII PRO-
JECT IONS,
GROWTH ACTUAL
BTU,   ASH
  SULFUR
VALUE OF
FUELS
                                STATE, CNTY,
                                SUBAREA FUEL
                                CONSUMPTION
MASTER GRID
COMM/INST 8
INDUSTRIAL
FUEL CON-
SUMPTION
                                                                           \   CONTROL STRATEGIES

                                                                                 :FUEL CONTROLS

                                                                                  EMISSION
                                                                                  LIMITS

                                                                                  GROWTH 8
                                                                                  DEVELOPMENT
                                                                                  PLANS
                                                                                 	I
                                                                                                          LEGEND
                                                                                                          BASIC CALCULATIONS

                                                                                                          GROWTH ANALYSIS

                                                                                                          CONTROL STRATEGY

                                                                                                          OPTIONAL FEATURE

                                                                                                           INPUT

                                                                                                           CALCULATION


                                                                                                           DATA FILE
                         Fig.  2.3b.   Computation  Flow  for  Commercial/Institutional  and  Industrial
                                                    Fuel  Combustion  Sources  (Level 3)

-------
                                   Table 2-3.  Data Available for Commercial/Institutional and Industrial Fuel Combustion Sources
Source
Census Bureau
Regional or Local
Planning Agency

Data
Employment
Number of establishments
Industry type (SIC) distribution
Economic data for manufacturing
All or some of the above informa-
tion
Floor space (sq. ft.) of comm/
inst and industrial buildings
Land area devoted to comm/ inst and
industrial uses
Projection parameters to convert
population to comm/inst uses
Growth projections
Spatial
Disaggregation
County
Some cities, SMSAs
Regional planning
districts. May or
may not be coincident
with other areas
Size range highly
variable

Form Date of General
Available Information Availability
Computer tape, Annual Entire U.S.
hard copy
Dependent on Latest planning Variable. Most likely in
agency. May or cycle. major metropolitan areas.
may not be
machine read-
able.

Fuel Dealers
                       All or some of the above informa-
                       tion
                       Actual fuel consumed by commer-
                       cial/institutional and industrial
                       customers
                                                           Variable with dealer.
                                                                Generally  hard
                                                                copy but occa-
                                                                sionally machine
                                                                readable.
                                              Latest data collec-
                                              tion  cycle.
                                                                                                                                 Variable
State Agencies
Fuel consumed by commercial/
institutional and industrial
users.
Generally by county
or for entire state.
                                                                                       Hard copy.
Latest year of
statewise data
collection plus
possibly projec-
tions.
Generally available.
Federal Agencies
Fuel consumed by commercial/
institutional and industrial
                                                           By state
                                                                                       Hard copy.
                                               Latest  year  of
                                               data  collection
                                               (usually  1-2 yr
                                                lag)
                                                                                                                                 Entire U.S.

-------
 should not be designed as an electrical load management program but should be
 able to supply some rudimentary information in this area.

 2.1.4  Internal Combustion
        Emissions from stationary internal combustion sources (e.g., gas turbines,
 diesels, gasoline generators, etc.) are generally only small contributors to
 emission levels.  Large electrical peaking units can be treated along with power
 plants while the smaller units at industrial facilities and aircraft engine
 testing facilities should be handled as individual sources in a point source
 inventory.  The CEPA system handling of these facilities is analagous to the
 Industrial Process emission sources .described next.

 2.2  PROCESS EMISSIONS
        The treatment of industrial process emissions primarily involves the
 handling of point source data; that is, the specific location and operational
 characteristics of Individual facilities are identified.  In some cases the
 nature of a particular process activity is such that'for air quality modeling
 purposes it will be treated as an area source (e.g., large open pit mining
 activities) but the specific operation is still handled as an individual facil-
 .ity in the calculation.  In other cases, the small and dispersive nature of a
 process activity may require treatment as an area source but this is not usu-
 ally encountered with great frequency.
        Figure 2 .4 shows the flow of computations and Table 2-4 shows the avail-
" able sources'of information.  Despite the relatively simple appearance of the
 basic calculation (solid lines in Fig. 2 ,4), this part of the CEPA system re-
 quires the greatest flexibility since it must be able to process information
 on a facility-by-facility basis.
t
        The initial source of information is the point source file obtained as
 part of an emission inventory process.  Current state files of this nature are
 in a variety of formats, the most frequent of which parallels EPA's National
 'Emission Data System (NEDS), the form of which is shown on Fig. 2 .5.  The basic
 operations that a CEPA system must be able to perform on this file include the
 retrieval of certain key items of,information (e.g., the process weight rates
 of all  sources  of a given type), the modification of the file with new  informa-
 tion on such things as  new  plant additions,  plant  retirements,  equipment

-------
=^
^=^
POINT
FILE
	 ^



i
\ SPECIFIC DATA ON /
\NEW PLANT, PLANT /
\ RETIREMENTS, /
\ PHASE OUT OF / 	 .
\ EQUIPMENT, / 1
VTC / i
1
GENERATE
MODIFIED
POINT
SOURCE
FILE
1
\

SUBTRACT KNOWN
POINT SOURCES
 GROWTH ANALYSIS
 USING EMPLOYMENT,
UAND USE, ETC.
                                                                                                             BASIC CALCULATIONS
                                                                                                             GROWTH ANALYSIS
                                                                                                             CONTROL STRATEGY
                                                                                                             INPUT
                                                                                                             CALCULATION

                                                                                                             DATA FILE
                               Fig. 2.4.   Computational Flow  for  Industrial Process  Sources

-------
                                                       Table  2-t\.   Data  Available  for  Industrial Process Sour
Suite Air Pol lutiou
Cunt ro1  Agency
                                    Data
Point source inventory
Interview results for new
plants, retirements, etc.
                                            Spatial
                                        Disaggregation
                                  Form
                                Available
                          Date  of
                        Information
                               General
                            Availability
Specific point sources
                            Machine-readable:   Latest  inventory  cycle.
                            either on state    Generally  2-3  year  lag.
                            system or NEDS.
                            Hard copy of new
                            plants, retirements,
                            etc.
                                             Inventory  for  entire
                                             U.S.   Interview  results
                                             varled.
Kcgioiutl or Local
Planning Agency
Growth projections:
  Population
  Land Use
  Industrial Output
Regional planning dis-
tricts.  May or may not
be coincident with other
subareas.
Dependent on
agency.  May or
may not be ma-
chine readable.
Latest planning cycle.     Variable.
Federal Agencies
Generalized growth pro-
jections (e.g., OBERS)
Generally state level or
industry level.
                                                                                        Most hard copy.
                   Latest year of planning   Entire U.S.
                   cycle.
                                                                                                                                                                       CD

-------
         *QC«
                PUnlU
                Nu*t |lt"»l fl  I  fluri Hilt ill1 am)   II til I
                      i' i  i  I   i  r t  j  t "i  I  i  i  i  i     L   r
                  3TTl31Tl-Tl;rr1--Tl41<-|"r'Uil II
                                                                                                               Onticl PcilOMl
                                                                       iJtOJi'l 6;[6J64l til 6<[67[«»[6al>t[>l|7l[;5
                                                                                                                                       J6J77
                                                                      mlh
                                                                  I   fomnHHi  t

                                                                  |_ "J". -I--,—
                                                                                                                              -l-l XEII
                                                                                                                                            7H79
 e«,l
            BotUi O't'l*
              Cipidlf
             IO'BIU iu
Lsiu
 16 t?
              ia|2o[2ib2
                 PfiMlf
                  Pill
                 ;ih4
  E
       i ANNUAL TNRUrUT
               JUM
      Fek   Kir  Au|  NM
:coadii|il Piiuiy IStcond.
 Ptll  j SO;  j SO;
'27[;8la!30|31l32[d
; inxi-Llii
                        CONTROL (QUIPXCNI
                          PliM'i  I Scc«n4j(y
                            NO,  I  NO,
                                          34 IS 36mba Wl40
Pnmaiy
  HC
                                                          4JU2 43
Sccondifyi  Piiajfy IStcaiiaiiy
  HC  I   CO  I   CO

44J4slj6|47J48J4aLj5![52
  (SIIMATCO CONTROL IFFICIENO (tl

Pul. I  SO] I  NO, I   HC  I  CO
                           j|54 65M57lS«ksteOtelbl|S3l§4b5
                                  HORBH
                                OPERATING
                                ViLnniuiu                 .

                              0 <»|hr J»k r<      PulicuUlt    I       SO;
                              fejypkWw jV32'l/33ibfl6l3j]Mjj9|46l4l!4?J4l
                                            EMISSION ESTIMATES I Ion I (e>il

                                                       NO,     I       HC       I       CO
                                                                                    61r«.2!(3l««]6!
                                                                                    _LLLL

                            BCJ6? S8l69J7UJ7l[72'73 )4
JILU
76 7JJ7I ralap
  ilAfrn
                                                                                                                   ESTIMATION
                                                                                                IICICO
                                                                                                                                    \SpJK
 16|17
                                *LLO»*BLE CWISSIONS |ll»> >'<>)
                fulliulm    I
                2o|y 22J23 J4|2i
             SCO
                                  I                I                 ^  SCHEDULE
         |0j _            NO.      I       MC               CO       *  Uli |«onlh
          ztlzstMlai 3^3|34]Mh^7M33J4TOp2[43|4^i|4lT4>|4B[4>koril[i2 63  64J5C 56J67
                                                                                             :0:JPLIAHC£
                                                                                              SCHEDULE
                                                                             CCUPLIANCE STATUS
                                                                                 UPDATE!
                                                                                                LLl
                                                                                                                       CONTROL RECULATIOHS
                                                                                           R'll  I    «U»
                                                                                       ^|e«Tc;U~8]63T;ol7Tn>
                                                                                         rlTTT IT
                            Solid lull
                                           Mjilnum Oaf i(n
                                              Rjlt
                                                                 ±
                                                                             H»l Content
                                                                             IH'plU/uc
                                                                                4lt'40 il)
                                                                                 M-
                                                                                                         m
                                                                                                                                     UttJ
                                                                                                                                              ii  1^
                                     Fig.   2.5.    NEDS Point  Source  Coding  Form

-------
                                      20
 replacement,  process modifications,  etc., and the compiling of the  file  in
 mation in summary  formats  to allow for dispersion modeling, strategy analysis,
 and the like.   The system  should be  able to perform internal consistency checks
 on the information stored  in the file  (e.g., apply standard emission factors
 to see if the recorded  emissions are of the right order of magnitude)  and be
 able to call  out information that appears out of order.
        In this context  the CEPA system is functioning like a data base manage-
 ment system more than like a computational system.  There are some  valid argu-
 ments to be made that this type of operation does not fit into the  already
 defined concept of a CEPA  system but should be performed externally.   In any
 case, if these operations  become an  integral part of CEPA or if  CEPA is  de-
 signed to exclude  these operations,  some interface must be built so that these
 types of data  handling  can be done as part of the air quality analysis.
        In handling growth  projections, the CEPA system must be able to process
 both the plant specific information  and the generalized growth projections.
 Generalized growth projections are usually generated by a state  or  local plan-
 ning agency and would usually come in a form that specifies a percent  growth
 in  a given industry (e.g., output in secondary metals industry will grow by
 2%  between 1975 and 1980).  This information must be applied to  the industrial
 activity  already recorded  in the point source file and the resulting process
 activity must  be disaggregated into  growth that will occur at new sources,
 existing  sources,  and at sources whose location is presently unknown.  Any
 available  information on industry expansion plans must be included  to  separate
 the  growth  that has very definite locations identified and the growth  that  must
be distributed to  the most likely areas for industrial expansion.   The CEPA
 system must then be able to make this distribution on the basis  of  some  allo-
cation parameter (e.g.,  employment,  land use, etc.).  The end product  of these
calculations is a  projected point source inventory and a projected  industrial
process area source inventory made up of activity for which no specific  loca-
tional information  is available.
       Again,  despite the  relative ease with which the calculational procedure
can be described,  the manipulation of a significant amount of specific data
related  to  individual sources and the meshing of this information with gener-
alized growth data  is not  a trivial  task.   The CEPA system must provide  the
user with  enough flexibility to cover the most frequently encountered  situations

-------
                                      21
(e.g., knowledge of the startup or retirement of a specific facility) and be
capable of easy modification to handle the unusual situations.
       For control strategy testing the CEPA system must be able to treat three
basic types of procedures.  Emission limits are the most frequently used con-
trol techniques and are simulated by changing the emission factors applied to
the process activity.  In addition to the application of uncontrolled emission
factors (e.g., from Ref. 3) and the application of the regulatory emission lim-
its under consideration, the CEPA system should be able to perform calculations
assuming other basic emission limits: for example, New Source Performance Stan-
dards (NSPS), Resonably Available Control Technology (RACT), Best Available
Control Technology (BACT), etc.  To do this, the CEPA system would need a cat-
alogue of emission factors representing each control level and would apply the
user-selected emission rate.
       Control strategies involving growth and development plans would be sim-
ulated by the CEPA system by changing the growth rates or by regulating the
source-specific data.  Land use controls would be simulated by changing the
allocation parameters to allow or deny a desired level of growth in a given
location and then reapplying the emission factors.

2.3  SOLID WASTE DISPOSAL
       Emissions from solid waste disposal occur in several ways.  Large cen-
tralized municipal incinerators are the most obvious source but a significant
amount of refuse is incinerated on-site at large industrial facilities and
some is still incinerated in older apartment buildings.  Open burning of refuse
is prohibited in many areas although it still is practiced.  Refuse disposal
by land fill is not an emission source.
       The CEPA system should be able to treat solid waste data in two forms.
Figure 2 .6 shows the flow of the computations.  At the simplest level of anal-
ysis a surrogate variable (e.g., population) is input along with solid waste
generation factors.  The manner in which the refuse is disposed of (i.e., mu-
nicipal incineration, on-site incineration, open burning, landfill) is also
input and the quantity of solid waste is distributed accordingly.  Point
sources of waste disposal such as the municipal and large industrial or resi-
dential incinerators are separated from the totals and the remaining waste vol-
ume and disposal technique is allocated to subareas.  This is mapped into mas-
ter grids and emission factors are applied.

-------
                       SPECIFIC SOLID
                      tWASTE DATA
V'MII , 1 I1Y, SIIU-/ 1
\ AKI A, I'OI1, IN- / W
\riuvMi in / ^_ y
\ /

i ,
1
-J — . ^-J
I

COMPUTE SOLID
WASTE GENERATED
8 DISPOSAL
NIQUE


TECH-

      POPULATION,
      LAND USE,
      EMPLOYMENT
      GROWTH PRO-
      JECTIONS
SOLID WASTE
GENERATION
FACTORS
WASTE DIS-
POSAL TECH-
NIQUE DIS-
TRIBUTION
          I
          L. .
:                     GROWTH
                     ANALYSIS
L.
                                                      CONTROL STRATEGIES
                                                     DISPOSAL RESTRICTIONS
                                   GROWTH & DEVEL.
                                      PLANS

                                     EMISSION
                                     LIMITS
MASTER GRID
SOLID WASTE
DISPOSAL
EMISSIONS
                                                                           . J
                                                                                     LEGEND
                                                                                     BASIC CALCULATIONS 	

                                                                                     GROWTH ANALYSIS	

                                                                                     CONTROL STRATEGY 	
                                                                                     INPUT                 \—7
                                                                                     CALCULATION           |    [
                                                                                     DATA  FILE              f^>
                       Fig.  2.6.   Computational  Flow  for Solid Waste Disposal  Sources

-------
                                      23
       The second form of treatment involves the handling of specific solid
waste data obtained from local scavengers, incinerator operators, or indus-
trial facilities.  This data can then replace the computation of waste vol-
ume using the surrogate variables and the remainder of the calculation is
the same.
       The growth analysis provides input into three basic parameters.  It
first identifies the growth rate of the surrogate variable.  It also provides
an indication of the solid waste generation rate; that is, a determination of
whether the per capita generation rate will increase or decrease over time.
Finally, it identifies the future distribution of disposal techniques.
       Control strategies that CEPA must treat involve the application of
emission limits to the centralized incinerators, changes in growth and devel-
opment rates, and restrictions on disposal techniques.

2.4  TRANSPORTATION EMISSIONS
       Emissions from transportation sources can be grouped into 6 basic cate-
gories: highway vehicles, off-highway vehicles, aircraft, railroads, vessels,
and gasoline handling evaporation losses.

2.4.1  Highway Vehicles
       Highway vehicles generally represent the largest fraction of transpor-
tation-generated emissions.  For the purposes of air quality modeling the in-
formation will be handled as line sources (for major highway links) and area
sources.  The line source formulation is needed only if a model that specific-
ally simulates line emissions (e.g., HIWAY) is to be used.  In other cases the
CEPA system must be capable of mapping the line segments into appropriate area
sources for use with non-line-source models (e.g., CDM).
       Figure 2 .7 shows the flow of the highway vehicle computations and Table
2-5 gives the principal data sources.  The basic calculations can proceed in
one of two ways or in a combination of the two.  At the simplest level of
detail, the CEPA system takes countywide gasoline and  diesel fuel  sold from
tax records, transforms these to vehicle-miles-travelled  (VMT) using  an  aver-
age fuel economy (miles per gallon), and then maps  these to a  finer spatial
resolution using the population distribution.  For  the more sophisticated  levels

-------
       I VI IIICI I TYPL  ', AGL,
         UOAH MAPb, VH1 ,
         1KAFFIC COUNTS,
         MILLS Ul ROAD,   /
         iIRIP ORIGINS,
          M'LLD  BY
         \i\rn,  ZONL
 r'
t
w-


COMPUTE
SPEED, T
AGE, OKI
MASTER G
AND/OR 0

?~1
VMT,
VPE &
GINS
RID
N LINKS.

W

^
MASTER GRID
AND/OR LINK
TRAFFIC DATA
\ 	 ^
COUNTY GAS
AIID DIESEL
FUEL SOLD
ESTIMATE VMT
BY MPG.
DISTRIBUTE
TO SUBAREAS
           CROSS-CHECK
           FULL CON-
           SUMPTION
           NEW HIGHWAY
           CONSTRUCT ION
           DATA
                                                     A
                                                     j
                                                     i
             GROWTH
             ANALYSIS
                                                  POPULATION
                                                  GROWTH PRO-
                                                  JECTIONS.
                                                  VMT GROWTH
                                                  PROJECTIONS.
                                                                                    APPLY GROWTH
                                                                                    FACTORS.
                                                                                    SEPARATE
                                                                                    SPECIFIC LINKS.
                                                                                   LEGEND

                                                                                   BASIC CALCULATIONS

                                                                                   GROWTH  ANALYSIS

                                                                                   CONTROL STRATEGY

                                                                                   OPTIONAL  FEATURE

                                                                                   INPUT

                                                                                   CALCULATION

                                                                                   DATA  FILE
                                                                                                                                                                    NJ
                                                                                                                                                                    -P-
                     Fig.  2,7.   Computational Flow  for Transportation Sources  (Highway Vehicles)

-------
                     Table  2-5.   Data  Available for Transportation (Highway Vehicles)  Sources
Source
State
Highway
Dept.
Data
• Road Maps
• Traffic Counts
• Vehicle Speeds
• Vehicle Type and Age
Distribution
• New Highway Plans
Spatial
Disaggregation
Generally by link for
expressways, highways,
major arterials.
Form
Available
Hard copy
or machine
readable.
Date of
Information
Latest year of data
collection. (Usually
every two years.)
General Availability
Some of this information
is available throughout the
U.S. Detail level varies
with State.
Regional or
Local Planning
Agency
All or some of the
 above information
Origin-Destination
 studies
Traffic growth projections
Generally by  zone and
including local streets.
Hard copy
or machine
readable.
Latest year  of
planning plus pro-
jections.
                                                                                                                                               Ln

-------
                                     26
of analysis, the CEPA system must be capable of handling specific data from
regional planning, transportation, or highway agencies and process this infor-
mation into a format that is suitable for the application of emission factors.
For example, a highway department may supply a road map showing road segment
lengths and traffic counts on major highways.  The CEPA system should be able
to convert this into VMT; CEPA should also be able to read in the VMT directly
if the data is supplied in that format.  Other parameters that the CEPA system
must be able to deal with are vehicle speed, vehicle type (i.e., the five classes
in Ref. 3) and age distribution, and vehicle trip origins (for cold start calcu-
lations) .  The system should be able to read this information as direct input as
well as compute these parameters based on average or default values.
       The highway vehicle data may be presented in either roadway link format
or in traffic zone format and CEPA should be capable of processing both forms.
The process of mapping the parameters from zone and/or link into the master
grids and then applying emission factors is entirely analagous to the steps
carried out for the other source categories.
       Growth projections for highway vehicle activities must be handled by
                                                                         v
CEPA in two ways.  First, specific data on new highway construction and pro-
jected traffic levels must be one form of standard input.  Second, a general-
ized VMT growth projection based on population or other surrogate variable
growth projection must also be treated.  The two formats must be handled in a
consistent manner to avoid double counting.
       Control strategies for highway vehicles fall in three basic categories;
Emission limits, such as those achieved .through the Federal Motor Vehicle Pol-
lution Control Program or state inspection and maintenance programs, are treated
through changes in the emission factors.  Traffic controls, such as improved
traffic flow through intersections, are treated by changing vehicle speeds,
trip origins, or VMT ,in the controlled zone.  Growth and development controls
are simulated by changing the growth rates and/or "the operational dates of new
roadway segments.

2.4.2  Other Vehicles
       Emissions from other mobile sources generally represent a small portion
of an  emission inventory but they must be included nevertheless in the interest

-------
                                      27
of accounting for localized problems.  Off-highway vehicles (e.g., farm trac-
tors, construction equipment, etc.), aircraft, railroads, and vessels are
specialized both in terms of their operating characteristics and their spatial
distribution.  To treat these situations, the CEPA system must be able to han-
dle two different types of input.  First, it must be able to handle a basic
surrogate parameter to which an emission factor is applied.  For example, in
the case of aircraft, CEPA must be able to read in the number of landing-
takeoff (LTO) cycles and convert this to emissions by applying an LTO-based
emission factor.  Second, CEPA must be able to have as direct input the
emissions from each of vehicle sources.  This option would allow the user to
do a much more detailed emission computation (of an entire airport, for exam-
ple) , input the results directly into CEPA, and have the emissions remain as
an identifiable contribution from the specific source.
       The spatial distribution of emissions from other vehicles is highly
specific and does not lend itself to allocation by surrogate variables.
Instead, the CEPA system must allow for the input of specific locational
information about each source (e.g., the location of an airport, railroad yard,
or port facility).
       In a similar vein, growth projections and control strategies for these
sources are highly specific and would be difficult to simulate for all the
possible contingencies.  It is, therefore, more reasonable to require the user
to make the computations for these sources externally to the CEPA system and
input the final results for the application of emission factors and the allo-
cation to the master grids.

2.4.3  Gasoline Handling Evaporation Losses
       Emissions from gasoline handling lend themselves to relatively straight-
forward computation on the basis of either direct gasoline consumption data
(e.g., from tax records) or by using a per capita consumption rate.  The CEPA
system should be able to handle both of these contingencies relatively easily
if it is structured to do the calculations for other sources as previously
discussed.

-------
                                     28
2.5  MISCELLANEOUS SOURCES
       There are a number of miscellaneous sources that CEPA must handle whose
emission activity is not easy to compute but which may make sizeable contribu-
tions to the overall emission burden on the region under study.  From the com-
pilation of Table 2-1 these can be grouped into 3 basic areas: solvent evapora-
tion, fires, and fugitive dust.  A fourth area can be included that treats all
other sources that do not fall into any other category.

2.5.1  Solvent Evaporation
       The emissions from solvent: use in industrial processes (such as degreas-
ing) and in commercial operations (such as dry cleaning) come from a large num-
ber of relatively small sources.  These sources are too small to be included as
point sources under the Industrial Process category but their aggregate contri-
bution to the emissions can be significant.  These emissions can be estimated
in two basic ways.  If data on actual solvent use is available then it is pos-
sible to allocate this to master grid cells and compute emissions using an
emission factor.  If this information is not available then estimates of sol-
vent consumption are made using, for example, national average per capita use-
age.  Growth is handled as with other sources by applying growth factors to the
surrogate variable  (e.g., population) and applying the appropriate consumption
rates.  Control strategies in the form of emission limits are handled by CEPA
by changing the- emission factors; -growth and development controls are handled
by changing the growth factors; solvent use restrictions are handled by chang-
ing the consumption rates.

2.5.2  Fires
       Emissions from natural as well as man-set fires are extremely difficult
to estimate since the emission factors are not very well known.  In terms of
CEPA requirements,  the system only need have provision for inputting a basic
activity parameter, inputting an allocation parameter  (e.g., acres of forest
land), and applying an emission factor.

2.5.3  Fugitive Dust
       Recent studies have indicated that fugitive dust from both natural and
manmade sources represents a substantial portion of the particulate burden in

-------
                                      29
some areas.  The CEPA system must make provision for these calculations to be
carried out but only in the simplest of forms.  An input activity parameter
(e.g., miles of unpaved roads), an allocation parameter, and an emission fac-
tor are all that is needed.  Control strategies will operate on these basic
variables.

2.5.4  Other Sources
       The possibility always exists that an emission source that cannot be
classified elsewhere will need to be addressed in the air quality analysis
and the CEPA system will have to make provision for dealing with this situ-
ation.  The simplest way to treat this is to develop a generalized format
for these sources that specifies source activity, emission factor, alloca-
tion parameter, growth rate, and control level.  The CEPA system should per-
mit the user to input this data in point, area, or line source formulation.
It should also allow an abbreviation of all the needed information so that
the user need only input emissions.

2.6  GRIDDING
       All of the previous discussions of how the CEPA system should handle
emissions from the various source categories have followed a calculational
procedure that eventually led to the distribution of an activity from its
basic spatial resolution (e.g., census tracts, planning districts, highway
line segments, etc.) to a master grid network.  It is important to reempha-
size at this point some significant concepts of what a CEPA system should
and should not do with respect to this gridding procedure.

2.6.1  Calculational Procedure
       The master grid network, as defined in previous guidelines,   is
designed to display the emission data in a format that is compatible with
dispersion models.  In most cases, although certainly not all, the master
grid network is chosen once and all future modeling runs are made with this
network.  Under these conditions, the transformation from one set of data
(e.g., on census tracts)  to the master grid is made only once and the frac-
tional part of each district that resides in each master grid can become a
fixed data set to be used in a variety of situations.  For example, if it

-------
                                     30
is desired to distribute county wide fuel consumption on the basis of popula-
tion, the distribution is first made from county to census tracts using the
population and then from census tracts to master grid using the fixed mapping
fractions.  Likewise, anything else that is to be allocated on the basis of
population is first distributed to the census tracts and then from tracts to
master grid.
       The situation may also arise where several different data sets with
different spatial resolutions may be available.  For example, population may
be available on census tracts, employment on regional planning districts, and
VMT on traffic zones, and none of these areas are coincidental.  The CEPA
system should be able to handle all of these data sets individually and bring
them together in the master grid by mapping each separately.  In this
way the user can retain the identity of his basic data set until such time as
it is necessary to bring all together for a modeling run.
       Another important point to emphasize in this procedure is that the map-
ping should be done on the basis of process activity and not on the basis of
emissions.  For example, for residential fuel combustion, the fuel consumption
and not the emissions are transformed from the basic data set to the master
grid;  Emission factors are applied only after the master grid fuel consump-
tion is computed.  The reason for this procedure is to .minimize the errors
encountered by a mechanical transformation exercise with no interpretation of
the reasonability of the results on the part of the user.  By displaying the
process activity in each of the master grids, rather than just an emission
value, the CEPA system would give the user the opportunity to evaluate if the
mapping is reasonable.  The problems with the mechanical procedure was pointed
out in the Phase I feasibility study where one EPA Regional Office felt it
necessary to stipulate.in its contractural agreements that prior knowledge of
a given type of process activity in a given master grid cell was mandatory
before any allocation of emissions from that activity could be made to that
cell.
       This method, despite its obvious advantages, does present some problems
with regard to mapping certain parameters.  For example, it is easily under-
stood how fuel consumption instead of emissions can be allocated from the
basic data set to the master grid, but it is not obvious how an ancillary pa-
rameter, such as fuel sulfur content, can be allocated.  Careful weighting of

-------
                                      31
the sulfur content of fuels used in all data cells contributing to a given
master grid sell is necessary.  This procedure requires somewhat careful book-
keeping but the CEPA system should be able to handle this routinely.
       In some instances, the master grid network may not remain fixed for all
the calculations.  It may be desireable, for example, to have a fairly coarse
grid network for use in early analysis years and a more refined network when
growth and development make it necessary to have better spatial resolution.
Also, a changing land use plan may require a changing master grid network.  In
any case, the CEPA system should be able to process these varying grids in a
very simple fashion since only the mapping fractions changes; the calculation
procedure remains the same.  This feature need not consume excessive machine
core space since the mapping fractions can be stored off-line or read in as
part of the input stream.

2.6.2  Master Grid Development
       With regard to the development of the master grid, the CEPA system plays
no role.  This is done entirely external to CEPA and all that is needed by the
system is the set of mapping fractions that transforms the basic data set into
the master grid.
       The reason for this limitation on CEPA is that there already exists a
computerized procedure to develop master grids on the basis of population l
and there are numerous manual techniques to develop grids on the basis of
other considerations (e.g., land use, existing data bases, and others).  CEPA
need not duplicate these efforts but rather can make use of the grids developed
by these other procedures and enable the user to more efficiently distribute
activity to these grids.

2.7  GROWTH
       An important component of the air quality analysis to be assisted by
the CEPA system is the projection of growth in emission source activity.  It
has already been indicated that the CEPA system is not a growth analysis; that
is, it is not intended to be a tool for analyzing socioeconomic data and devel-
oping growth forecasts.  Rather, it is intended to take growth forecasts devel-
oped externally and translate those into emission forecasts.  To do this, the
CEPA system must be able to handle two types of growth projections.  First,  it

-------
                                      32
must be able to process information or specific growth plans. - If, for example,
it is known that a new manufacturing facility or a new highway link is sched-
uled to open on a specific date, the CEPA system should be able to develop an
emission forecast that accounts for that new activity.  Second, the CEPA must
be able to translate.this into an emission growth rate.  In addition to han-
dling these two types of growth separately, the CEPA system must be able to
coordinate the two to avoid double counting.  It must be able to identify and
separate growth at specific facilities from generalized growth.
       The growth routine should be able to accomodate linkages between growth
in different activities.  It should, for example, allow the user to couple
employment growth and population growth, VMT growth with residential land use
growth, and others.  In the purest sense, all of these linkages should be made
by the planning agency doing the overall growth analysis and the resulting pro-
jections should be entirely consistent.  In reality, the projections for var-
ious activities will come from different agencies and the air quality analysis
agency will have to make some attempts to coordinate and consolidate the data.
CEPA should provide an easy mechanism for doing this by allowing the agency to
input the linkages between the various activities and by cross-checking for
internal consistency.  An example of how this might work would be the follow-
ing: consider an agency receiving population growth projections on 3. census
tract basis from a local planning agency and VMT projections from the state
highway department on a traffic zone basis.  The CEPA system would take both
of these projections, process them to get growth projections on the master
grid network, and print out the population and VMT for each cell as well as
the growth ra'tes.  Wide discrepancies between these two would be immediately
obvious and alert the user to investigate for possible inconsistencies in the
data.  If it were desired to estimate VMT growth strictly on the basis of pop-
ulation growth, the CEPA system should allow the user to input this link with
a minimum of effort.
       Another feature of the growth analysis that would greatly enhance the
utility of the CEPA  system is the capability to process more than one. growth
scenario in the same computer run.  It should be possible for the user to
input a number of scenarios and have them for ready comparison.
       Although not  essential for a basic CEPA system, there is an operational
feature that could prove especially helpful to air quality analysts using the
system.  This feature would provide the user with information on what growth

-------
                                      33
is tolerable.  Under normal CEPA operation, growth is an input function and
if the user seeks to know, for example, how much growth can be absorbed under
certain constraints, he must proceed on a trial-and-error basis by inputting a
variety of scenarios.  To proceed in the opposite direction and have the user
specify a desired emission level with the CEPA system back-calculating the
allowable growth pattern, would involve non-trivial optimization routines, con-
straint specifications, and objective identification.  While this type of com-
putation is well beyond the design specifications of GEPA, the system can,
nevertheless, provide outputs that would allow the user to make a "better edu-
cated guess" on the next scenario to be tested.  These outputs would include,
for example, a summary table identifying the source categories experiencing
the most rapid growth, areas having greatest (and least) emission growth, and
the sources within each category and area that are making the biggest contri-
butions to emission growth.

2.8  CONTROL STRATEGIES
       The application of control strategies is a common item for all of the
calculations for the various emission sources.  In a sense it represents the
most significant procedure of the CEPA system in that it allows the user to
identify the effectiveness of various steps taken to minimize air quality im-
pacts.  There are several features of a control strategy routine that would
make CEPA a useful analytical tool.
       For the most part, many of the control strategies for the various emis-
sion source categories can be simulated by changing the basic data.  For exam-
ple, emission limits can be modeled by changing the catalogued emission fac-
tors.  To operate CEPA in this mode could easily become a tedious chore and
could minimize the utility of CEPA.  Instead, the strategy calculation should
be done as part of an individual routine that allows the user to specify the
control regulation and have the system change the emission factor for affected
sources.  This mode of operation gives the user the sense of inputing the con-
trol regulations in one place rather than attempting to pick out all the appro-
priate data values to change.
       The CEPA strategy package should allow the user to test more than one
regulation at a time to minimize the time required to conduct the analysis.
For example, the user should be able to specify an emission limit, a fuel

-------
                                      34
sulfur content restriction, and a coal ban strategy in the same run and have
the CEPA system display them in the same run.
       In applying control strategies, the CEPA system should address only
the source categories that are affected.  It should be capable of preserving"
the calculations for the base data sets that are not affected by the regula-
tion constraints.  The CEPA system should also provide the user with feedback
information to assist in the evaluation of the effectiveness of the regulation
tested and to help select the next regulation to be tested.  This information
would include the number of sources.affected, location of regulation's great-
est impact, percent emission reduction achieved, and others.  The extrapola-
tion of this feedback process to its ultimate conclusion would be equivalent
to having the CEPA system back-calculate what regulations would be necessary
to achieve a emission reduction.  This, however, runs into the same problems
as the growth analysis in that specification of objectives, constraints, and
optimization procedures that are well beyond the scope of a CEPA system, are
needed.  Nevertheless, the feedback of information from the strategy calcula-
tion should be as extensive as possible to minimize the effort required to
evaluate alternative strategy effectiveness and identify the best options.

2.9  GROWTH TRACKING SYSTEM COMPATIBILITY
       As part of the air quality analysis requirements issued by EPA under
Section 301(a) of the Clean Air Act, the states are required to assemble data
on growth in all areas of the state and conduct an analysis to identify those
portions that have indications of potential National Ambient Air Quality
Standard violations.  These areas would then be subject to more detailed
analysis for possible Implementation Plan revisions.
       EPA has recently issued a report  providing guidelines on tracking this
growth.  One of the principal concerns with CEPA system development is that it
be compatible with these guidelines.  The guidelines prescribe procedures for
four types of analysis conditions:  (1) areas with existing detailed projec-
tion of emissions and simulation of air quality, (2) areas with a less detailed
or condensed analysis of projected air  quality, (3) areas with no current
analysisjpf projected air quality but with air  quality monitoring data, and
(4) areas without an analysis and without monitoring data.  The guidelines out-
line a set of procedures to be followed in each of the four area-types.  In the

-------
                                       35
first type of situation (.i.e. with. a. detailed analysis already available) the
procedure calls for a collection of growth, information and a comparison to the
growth projections used in the analysis.  If any of the growth parameters ex-
ceeds the information used in the analysis, then a rough estimate of emissions
growth is made to determine the potential for NAAQS violations.  In the second
type of situation the process is basically the same only the parameter compari-
sons are made on a much less detailed basis.
       In the third type of situation the guidelines call for a linear "roll-
forward" of air quality data based on rates of growth in several basic parameters
(e.g. population, employment).  The fourth type of situation calls for an
emission projection and an air quality  estimate based on some very rough
approximations.
       It is evident that these guidelines are suggesting analyses that could
easily be handled by a CEPA system.  The suggestions generally amount to a
Level 1 analysis; the CEPA design calls for this capability to be built into
the  system.  It must be pointed out though, that while the CEPA design as
described here could do the growth tracking, it would be unlikely that a control
agency would install the CEPA system for that purpose alone.  The CEPA would
provide much more capability than is necessary.  The conclusion is that a state
agency that had a CEPA system operational would definitely use it in the growth
tracking analysis but a state would not proceed to install a CEPA to do that
task only.

-------
36

-------
                                      37
                3  COMPUTER CONFIGURATION OF THE CEPA SYSTEM

       The computer configuration to be used when implementing the CEPA sys-
tem consists of two distinct yet totally dependent parts.  First is the hard-
ware, namely the main computer memory, intermediate storage and the input-
output devices.  Second is the software which is the set of programs at both the
support and applications levels.  These two parts must be integrated in such
a fashion that it will provide a feasible tool for a maximum number of users
and yet be simple to use.  The applications software is more flexible in its
development than the hardware, hence the hardware requirements are more defin-
itive.

3.1  CONSTRAINTS ON HARDWARE AND SOFTWARE
       As a direct consequence of the Phase I Feasibility Study,1 a set of
objectives and constraints pertaining to the hardware and software configura-
tion of a CEPA system has been established.  These are as follows:
       1.  The CEPA system must be designed for operation on both
           UNIVAC and IBM equipment.  Users with other machines may
           have to modify their version of CEPA to use it on their
           facilities.  The system should be designed to facilitate
           conversion to other machines.
       2.  The system must be capable of installation on EPA's
           UNIVAC 1110 machine in Research Triangle Park, N.C.
       3.  The system software must use only FORTRAN and/or COBOL.
       4.  The system software, in either card, tape, or other for-
           mat, must be in a form that is easily duplicated for
           transmission to potential users.
       5.  The system must not operate exclusively in the inter-
           active mode.  Batch mode or a combination of batch and
           interactive should be employed.
       6.  The system must be capable of accepting machine readable
           input from EPA's National Emission Data System (NEDS) and
           Emission Inventory/Permits and Registration Subsystem

-------
                                     38
           (EIS/P&R).  The only other  existing machine-readable
           format that the CEPA system should be designed to ac-
           cept  (to  the  extent practicable)  is the Bureau of the
           Census data tape  format.
       7.   Users with only limited familiarity with  automatic  data
           processing should be able to use  the GEPA system with
           the help  of a user's manual.  Extensive machine  job con-
           trol language should be avoided for normal operation.
       8.   The system should provide output  that is  machine-readable
           for direct input  into  the AQDM, CDM, IPP, and Valley mod-
           els;  that is  suitable  for input into isopleth plotting
           routines; and that  is  in hard copy  (printed output) for
           the entire area or  for individual subareas.
       9.   The system should be modular in structure so that a user
           may choose to run a portion of the  system or the entire
           system.
      10.   The system should be designed for possible inclusion into
           EPA's Aerometric  and Emissions Reporting  System(AEROS) .

3.2  HARDWARE ALTERNATIVES
       As  was stated above,  the UNIVAC 1110  in Research Triangle  Park satis-
fies part  of the hardware  restrictions.  The IBM portion can be satisfied by
a moderate^to-large size computer in the range of an IBM 370/160  to 370/195.
The final  choice is dependent  on  how much software  is required and whether or
not the CEPA system will function independently of  other systems.
       An  intermediate  storage capability must be available to the user for
the transient files that will  be  created during an  analysis.   (Transient files
contain results  of  intermediate computations and require too much main memory
to be permitted  to  reside  there,  hence the name  transient.) The  varied forms
of intermediate  storage  can  be magnetic tape,  high  volume  disk, data cell,
high speed drum  or  cards.  No  permanent files  can be allocated by the user
because of the prohibitive cost required to  provide enough hardware to support
the large  number of potential  users.   (Put  simply,  if each user were to have
files reside ad  infiniturn, there  would be no more available space for new files.)
The magnetic tape then  is  the  only  cost effective and portable form of

-------
                                      39
intermediate storage for large permanent files.  Smaller files can be stored
on cards if necessary.  The remaining forms of data storage can be used for
the transient files.  Since CEPA must function as a viable user oriented com-
puter system, then these transient files can and should be purged from the
system automatically after being retained for a finite period of time.  This
retention period should be sufficiently long to allow analysis to be performed
without the extra cost in time and effort of recreating the transient files.
       Input and output (I/O) options must also be considered.  For input,
magnetic tape and cards are the only reasonable choices.  Cost and portabil-
ity are the key factors governing these choices.  These two modes of input
would then contain the machine-readable raw data necessary to begin the anal-
ysis.  The output can be in the form of magnetic tapes, punched cards, or
printed output.  These formats will not require extensive control language to
manipulate.
       The only other form of I/O that can be considered is an interactive
terminal; namely, a device by which communication with the central computing
facilities is made possible.  With this device and the proper software support,
the user can communicate with the central facilities via the telephone lines.
However, the interactive mode must not be the sole form of I/O available under
the restrictions stated above.

3.3  SOFTWARE CONSIDERATIONS
       The software requirements to handle the potentially large volumes of
data immediately indicate that some form of Data Base System (DBS) and Data
Base Management System (DBMS) be employed.  The DBS and DBMS required is the
Emissions Inventory/Permits and Registration Subsystem (EIS/P&R) for point
and area sources.  This system will permit NEDS data to be input and output.
Retrieval and updates of existing data as well as addition of new data is
made efficient and simple by this system.
       The CEPA system may become part of the EPA's Aerometric and Emissions
Reporting System (AEROS), which further heightens the desire to maintain com-
patability with the NEDS format.  The AEROS system is used to provide multi-
leveled reports of air quality and emissions for states, AQCRs and counties.
Other systems can be developed to handle the more general user input but this
would require a great deal of time and effort to implement.  An alternative

-------
                                      40
 to  the  development would be to purchase a proprietary package for DBS and
 DBMS  such  as  the DMS 1100 created by Sperry Univac Company or IMS created
 by  IBM.  Such packages can perform general DBMS functions but can represent
 a considerable  cost factor.
        The final output of the CEPA system will be in the form of gridded
 emissions  and have formats which are compatable with the AQDM, CDM, IPP,
 Valley  and isopleth plotting programs.  Intermediate output can be the  tran-
 sient files discussed earlier, tabulated printout for each module of the anal-
 ysis, magnetic  tape or cards containing either the tabulated results or the
 transient  files for future analysis.
        Some problems can always be expected when transportability of computer
 systems software or programs is required.  To relieve some of these difficult-
 ies the following restrictions are made: only ANSI (American National Stan-
 dards Institute) FORTRAN IV and ANSI COBOL will be used in applications pro-
 grams.   Whenever possible, the primary language should be FORTRAN IV and only
 where absolutely necessary should COBOL be used.  A minimum of interaction
 between these two languages is desired since the interfacing will vary  from
 computer installation to computer installation.  The FORTRAN language has
 good  computational capability while COBOL is good for file manipulation.
        The  prime mode of operation is to be batch, namely, an entry of  data
 into  the CEPA system and execution of CEPA modules without further interac-
 tion  by  the user until the results are compiled.  A secondary mode of opera-
 tion  can be the interactive mode whereby the user is in constant communica-
 tion  with the CEPA system, which would provide intermediate results for the
 capability of on-line supervision of the procedures used in producing the
 results.  Since this mode of operation, in the general case, is not required
 by the average user,  this system should be developed only if it does not de-
 grade the operation of the batch mode and if it can be developed in such a
 fashion  that it is separable from the batch system.  The interactive mode
 should be used only if intermediate results can change the path of a given
 strategy and the compilation of the intermediate results is relatively  short.
If these two conditions prevail, it is to the user's advantage to use the
 interactive mode since it will shorten the time required for the analysis.

-------
                                     41
3.4  DOCUMENTATION
       The documentation of the CEPA system should be developed at two lev-
els and in accordance with some predefined guidelines.  The first level is
the user guide and must contain at the very least a description of the theoret-
ical methods used in applications programs, a detailed description of how to
use each of the applications programs, and a set of comprehensive sample
problems.  In addition, any control procedures necessary to facilitate data
handling and linking one module of CEPA to another should appear as sample
problems.  The second level is a programmer's manual on the details involved
with applications program, so that the user may develop modifications of his
version of the system.  Flow charts, discussion of primary variables,
input parameters and formats, output parameters and formats, and linking of
one module to another are just a sample of the items to be discussed in the
programmer's manual.
       In addition to these basic requirements, one other documentation con-
sideration must be addressed.  If the CEPA system is to be included as part
of EPA's AEROS system and is to be maintained and supported by EPA, it must
meet certain documentation requirements that would not ordinarily be required.
These requirements are based on the concept that an EPA staff member who was
not involved in system development would be able to learn the structure and
operation of the system quickly and would be in a position to make updates
and changes that could be transmitted to all users.  The meeting of these
requirements could amount to a significant effort above and beyond that re-
quired to meet ordinary user needs.

-------
42

-------
                                      43
             4  COMPARISON PROCEDURE FOR ALTERNATIVE CEP'A SYSTEMS

       With the basic analytical and computer requirements laid out for the
CEPA system, it is now necessary to define a comparison procedure to determine
if any existing computerized air quality analysis packages are capable of
meeting most of these needs.  The systems reviewed here are the following:
       1.  Air Quality for Urban and Industrial Planning (AQUIP)
       2.  Computer-Assisted Area Source Emission Gridding pro-
           cedure (CAASE)
       3.  Engineering Science Air Quality System (ESAQ)
       4.  Metropolitan Washington Council of Governments
           Air Quality Analysis (MWCOG) Models

       In addition to matching each of these systems against the requirements,
a comparison will be made with two other calculational procedures that can be
used.  These are:
       5.  Manual calculations
       6.  Newly developed CEPA system.
These last two can, by definition, be made to meet the requirements and they
will serve to bound the evaluation by estimating the costs of doing the calcu-
lation by hand or developing an entirely new system to do the required calcula-
tions .
       The comparison procedure to be followed here involves the steps shown
on Fig. 4.1.  First, each of the existing systems will be briefly described
to give an overview of how each is designed and the major computational phil-
osophies of each system.  Next, each system will be compared to the analytical
requirements spelled out in Section 2.  If the system does not meet the analyt-
ical requirements, then the significance of the lack will be identified and an
estimate of the modifications necessary will be made.  If the system is capa-
ble of performing the required calculations, then a review of the data required
and the validity of the approach will be made.  This is to identify potential
problem areas where a system will perform a certain calculation but use diffi-
cult to obtain data or use a procedure that is of uncertain validity.  Despite
the answer to the analytical evaluation questions, each system will be reviewed
to determine if there are extra features that are not required as a part of
CEPA but which are especially useful to an air quality analysis.

-------
                                       44
   WHAT IS THE
   REASONABLENESS
   OF THE DATA?
   WHAT  IS THE
   ACCURACY OF THE
   CALCULATIONS?
ANALYTICAL EVALUATION:
DOES THE SYSTEM DO THE
REQUIRED CALCULATIONS?
HOW SIGNIFICANT
IS THE  LACK?
                                   WHAT CHANGES.
                                   ARE NECESSARY?
                                DOES THE  SYSTEM
                                HAVE EXTRA DE-
                                SIRABLE  FEATURES?
                               DOES THE SYSTEM  MEET
                               COMPUTER REQUIREMENTS?
                                   HOW SIGNIFICANT
                                   IS THE LACK?
                                           YES
                                                                 WHAT CHANGES
                                                                 ARE NECESSARY?
                                DOES THE SYSTEM
                                HAVE EXTRA DE-
                                SIRABLE FEATURES?
                               HOW DOES THE SYSTEM
                               COMPARE TO MANUAL
                               CALCULATIONS?  TO
                               A NEW CEPA DESIGN?
                                 SUMMARY AND
                                 RECOMMENDATIONS
Fig.  4.1.   Comparison  Procedure  for Existing  Computer  Systems

-------
                                      45
       The next step in the comparison procedure is to review each system to
determine if the computer requirements are satisfied.  If the answer is neg-
ative, then the significance of the lack and the modifications required will
be evaluated.  Again, any desirable extra features will be highlighted.
       Next, each system under consideration will be compared to a set of cri-
teria that will measure the capability of that system against a manual calcu-
lation procedure and against a new CEPA system developed from the ground up.
The criteria used for this evaluation are the following: (1)  effort required
to use the system - including getting the system operational, preparing the
data for input, and operating the system, (2) Level of expertise needed to
operate the system, and (3) cost of using the system - including cost to get
it operational, cost of preparing the data, and cost of operating the system.

       The  final step in the evaluation will be to summarize the assessments
and develop a  set of recommendations for future action.  No attempt will be
made  to reduce this summary to a single number for comparison as this would
tend  to obscure the details of the problem areas.
       The  following section presents the descriptions of the existing systems
and the evaluations.

-------
46

-------
                                      47
                       5  COMPARISON OF EXISTING SYSTEMS

       In this section four existing computerized air quality analysis systems
are described and evaluated against the CEPA system requirements.  System
descriptions are drawn primarily from documentation available, sometimes
verbatim.  Detailed evaluations against CEPA requirements are presented in the
appendices; only summary conclusions are discussed here.

5.1  THE AQUIP SYSTEM
       The evaluation of AQUIP was made on the basis of the documentation con-
tained in References 6-10 and on discussions with EPA staff using the system.

5.1.1  System Description
       The Air Quality for Urban and Industrial Planning (AQUIP)  System was
developed as a joint venture between the New Jersey Department of Environ-
mental Protection and the U.S. Environmental Protection Agency.  Environmental
Research and Technology, Inc. (ERT) of Lexington, Massachusetts was selected
as the contractor to build the system.
       The objective of the ERT work was to develop a methodology to assess
the air pollution impact of land use plans and to apply this methodology to a
test case in the New Jersey Hackensack Meadowlands.  Because of this objective,
the system carries a distinct orientation toward use by planners.  Much of the
input and output is structured around the variables and parameters normally
used by planners (as opposed to those used by air pollution control engineers).
As such, it is the only one of the systems evaluated as CEPA candidates that
allows for direct and straightforward treatment of land use plans.
       The AQUIP software system makes use of input data sets and model para-
meter data sets, performs computations using four basic computer programs,
and provides tabular and graphical outputs of the results.  The logical
relationships among these elements of the software system are shown in Fig. 5.1.
Data sets are shown as rectangles, computation steps as circles, and printed
output as document symbols.  In addition, each element is identified by a code
made up of a generic letter followed by a number.  The letter prefixes and
their meanings are:
       I  -  Input data set, prepared by the system user.

-------
Planning
Inputs
Air-Quality
Prediction Model
Air-Quality
Impact Model







Tabulated
Emiis-ons
Doio



Cumpuled
At Qyol.ty
by Rf cepto*

•

_^/\_^
	 »^ SYMAP 1 	 »•
^—^?4

                                 14
                                                                                                     Iioplrlhi
                                                                                                      of A,,
                                                                                                     Ouol.t,
                                                                                                                      oo
                           Fig.  5.1.   Flowchart  of AQUIP System

-------
                                     49

       M  - Model parameter data set, established initially for the study
            conditions, and modified only as necessary for updates to the
            model.
       P  - Computation step involving one of the four basic computer
            programs.
       C  - Computed data set formed as an output of one computation step
            and used as an input to another.
       T  - Tabulated outputs (or line printer graphics) delivered to the
            system user.
Table 5-1 gives a summary of the elements of the system.
       Of the four computer programs that comprise the AQUIP system, only the
LANTRAN routine is of direct relevance to the needs of a CEPA system.  The
MARTIK program is a dispersion model and, by definition, is excluded from the
CEPA consideration.  The SYMAP routine is a standard plotting package which
can be incorporated into a CEPA system, but its location in the AQUIP structure
(i.e., receiving output from the dispersion model) puts it beyond the bounds
of CEPA.  The IMPACT program is designed to determine various land use and
population exposures to air pollutant concentrations and it also is beyond
CEPA bounds.
       The purpose of the LANTRAN program is to convert land use data to a
rectangular grid system; to provide land use statistics; to provide certain
commonly used preprocessing procedures for land-use data; and to establish
data sets for use by other programs.  The program is organized around two basic
forms of data:  that related to land use activities and represented by a set
of geographically defined "figures," and that related to a grid system with
its associated "cells."  In LANTRAN the "figures" are the input and the grid
system the output, i.e., the result of an allocation of activities defined on
the figures to cells of the grid system.  Internally, the two forms of data are
represented by two large arrays.  The first enables up to 18 different sets of
data to be defined on up to 400 different figures, with each figure consisting
of either:  (1)  a single point, (2) a broken line of up to 50 vertices, or
(3) a polygon area of up to 50 vertices.  The 18 "variables" are assigned sym-
bolic names by the user at run time, making possible the manipulation of data
by reference to the symbolic name.  Examples of symbolic names which might be
useful in land use applications are "POP-DENS" for population density or
"DU/ACRE" for density of dwelling units.

-------
                                      50
                       Table 5-1  AQUIP System Elements
Element
Designation
       Element
         Description
Input Data Sets

      II.



      12.
Original Land-Use Data
Highway Emissions Data
      13.
      14.
      15.
Point Source Emissions
Data

Land Uses for
Correlation

Impact Criteria Data
      16.
Map Options
Model Parameter Data Sets
      Ml.
      M2.
Activity Indices


Fuel Use Data


Emission Factors


LANTRAN Program
Parameters


Background Emissions,
by Season
This data set is specified  as a set
of point, line or polygon "figures"
to which "values" representing plan-
ning variables are assigned.

This data set is specified  as a set
of "line" sources, to which emission
densities have been assigned by the
application of emission factors to
traffic data.

This data set is specified  as a set
of "point" sources to which emission
rates have been assigned.

Specified as a set of "figures" reprt
senting land uses to be correlated
with air quality predictions.

This data set is a set of operations
to be performed upon gridded air
quality data for comparison with
standards or correlation with
various land uses.
Which select variables for  isopleth
plotting and specify characteristics
of output maps.
To relate activities specified in the
given land use data to fuel demand.

To specify overall fuel availability
data.

To relate fuel use or process rate by
activity to emissions by pollutant.

To specify the grid properties, pro-
gram options and computation para-
meters.

A previously generated data set to
account for the contribution of all
point, line and area emissions
sources outside the study area to
computed concentrations at the
receptor sites.

-------
                                     51
                  Table 5-1  AQUIP System Elements (Cont'd)
Element
Designation
      Element
            Description
Model Parameter Data Sets (Cont'd)
      M3.
Meteorological Data
      M4.
      M5.
                   Meteorological
                   Parameters
                   MARTIK Program
                   Parameters
SYMAP Base Map
Allocation Options
Computer Programs

      PI.          LANTRAN - Land Use Data
                   Transformation Program
      P2.
MARTIK - Martin-Tikvart
Diffusion Modeling
Program
      P3.
IMPACT - Impact Analysis
and Display Program
The set of normalized weighting
factors to be assigned to each
of the 480 meteorological condi-
tions, based on the relative
frequency of occurrence of these
conditions.
To determine such model character-
istics as plume dispersion coef-
ficients, mixing layer depth and
vertical wind-velocity profile.
To specify receptor properties,
program options and computation
parameters.
The set of SYMAP input packages
which define the study region and
the coordinates of the data points.
The set of LANTRAN control options
required for allocation of computed
concentrations by receptor to the
chosen grid system.
The fundamental purpose of this
program is to convert data defined
on point, line, or irregular polygon
"figures" to a regular grid system.

Computes the arithmetic mean air
quality levels at designated
receptor locations for a given
distribution of emission sources
with meteorological data specified
for the averaging period of interest
and the climatology of the study
region.

This program performs arithmetic and
logical operations as specified at
run-time by a "user hyper-language"
on each element of a gridded system
of data, allowing cell-by-cell compar-
ison with user-specified criteria.

-------
                                       52
                  Table  5-1  AQUIP System Elements (Cont'd)
Element
Designation
         Element
              Description
Computer Programs  CCont'd')
    P4.
Computed Data  Sets
    Cl.
 SYMAP  -  Synagraphic
 Computer Mapping Program
 Point  and  Gridded Area
 Source Emissions
    C2.


    C3.


    C4.
Computed Air  Quality
Gridded Air  Quality
Correlation Data  Set
System Outputs

    11.
Tabulated Emissions
    T2.
Tabulated Air Quality
Predictions
A general-purpose graphics display
program presently implemented for
the display  of  isopleths of air
quality as computed by MARTIK.
Allocated by  pollutants to the
specified grid  system.   The point
sources in  the  data set represent
discrete sources  with emissions in
excess of a given threshold.   The
area sources  represent  the remainii
activities  distributed  to grid eel]
on the basis  of area overlap  or
"extent".

By pollutant  for  each of the  spec-
ified receptors.

By pollutant  converted  to mean con-
centration  for  each grid cell.

A gridded data  set representing
allocation  of specified land-uses
or their derivatives (e.g., popu-
lation density) selected for  cor-
relation with air-quality levels.
Projected emissions  as  computed by
LANTRAN for  the  given ensemble of
input data and model parameters,
given as a summary for  each constit
uent land use "figure",  with tables
and plots of resultant  emissions
presented for the specified grid
system..

For the given ensemble  of  planning
inputs, model parameters and meteo-
rological conditions.   Tabulated by
pollutant for each of a  specified
set of "receptor" locations within
the study region.

-------
                                      53
               Table 5-1  AQUIP System elements  (Cont'd)
Element
Designation
       Element
           Description
System Outputs  (Cont'd)
     T3.




     T4.



     T5.


     T6.
Isopleths of Predicted
Air Quality
Tables and Plots of
Predicted Total Air
Quality
Tables and Plots of
Land Use Data

Tables and Plots Pre-
senting the Results
of Impact Analyses
A graphical display of isopleths
of pollutant concentrations generated
by the line printer using an over-
print technique to produce "shading".

Expressed in absolute units of con-
centration for each cell of the
study region grid system

To be used for correlation with
gridded air quality data.
e.g., (1) statistics of compliance
with standards; (2) integrated
dosage by land use; and (3) overall
land use compatibility.

-------
                                      54

       The second array corresponds to the same 18 variables defined on a grid
system of up to 400 cells.  The grid system is specified by the horizontal
and vertical coordinates of its "origin," the cell count in the horizontal
and vertical directions, and the dimension of the grid cell in the horizontal
and vertical directions.  In addition, a scale parameter is specified to
enable a convenient set of units such as kilometers or miles to be used for
the coordinate system; the physical height of the grid system is specified
in meters.
       In summary, the use of LANTRAN consists of (1) defining the set of
FIGURES, (2) defining the variables associated with the figures and assigning
VALUES for these variables to ^the figures, (3) performing an ALLOCATION which
                              ^«
distributes selected variables among cells of the grid system, and (4) creating
an OUTPUT data set defined on the grid system, and putting this data set out
either in punched-card form or as card images on a specified file.  In
addition, the two basic forms of data represented by the figure-values or
"FV" array and the grid-values or "GV" array may be manipulated before or
after allocation using an application-specific subroutine (COMP) written by
the user.
5.1.2  System Use
       The AQUIP system has not been widely used.  Apart from the original
application to the New Jersey Hackensack Meadowlands, there have been only
limited attempts to use the system in air quality analyses.  The system has
not been used as part of any required air quality control plan (e.g. SIP
revision, AQMA analysis, etc.).
5.1.3  Comparison-with CEPA Requirements
       The details of the comparison of the AQUIP system against the CEPA
requirements are given in Appendix A.
       The strong point of the system is its ability to map emissions from sub-
area to master grids.  The LANTRAN routine allows the user to easily change
from one subarea set to another and have the program determine the appropriate
transformation from subarea to grid.  The routine is generalized enough to
handle areas, points, and lines and treats them all as generic "figures."
       The procedure whereby the figure is transformed to the grid can be
varied depending on the nature of the situation.  Allocations can be made by
extent (i.e. by the portion of a figure lying in a grid cell), by association
(i.e. by choosing the dominant value of a parameter from among all the values
on all the figures lying in a grid cell), by interpolation (i.e. by developing

-------
                                     55

a weighted average of the parameter values of all the figures lying in the
grid cell), or by proximity (i.e., by choosing the value of a parameter cor-
responding to the figure whose centroid lies closest to the centroid of the
grid cell).
       In making the transformation, the user also has the option of inter-
spersing a subroutine to do additional manipulations on the variables before
they are transformed.  This is a very desirable feature in that it gives the
user a great deal of flexibility with respect to the calculations that can be
performed.
       The structure of the LANTRAN routine meets the CEPA requirements of sur-
rogate variable input for the residential and commercial/institutional fuel
combustion, solid waste disposal, transportation, and miscellaneous sources.
The surrogate parameter (e.g., population density, housing units per acre, etc.)
can be defined as one of the 18 "variables" on each subarea or "figure."  The
translation from a surrogate variable to fuel consumption (or solid waste gener-
ated, or solvent used, or VMT, etc.) is made via a table look-up routine and the
calculation of emissions is done using emission factors.  The one weakness in
this procedure is that the translation tables and the emission factors are
strongly linked to land use parameters (e.g., acres of commercial land, emissions
per acre of commercial land used, vehicle density, etc.) and are not readily
adaptable to the use of direct information on fuel consumption, solid waste
generated, etc.  In this regard, AQUIP cannot handle the direct data input for
either residential and commercial/institutional fuel combustion, solid waste
disposal, or miscellaneous sources, and does not meet the CEPA requirements for
these types of analysis.  For highway vehicle transportation sources AQUIP can
accept VMT data by link or traffic zone but cannot treat vehicle fuel consump-
tion inputs.
       AQUIP is especially weak with regard to its treatment of point sources
(industrial process and electric generation).  The system simply reads in point
source data and cannot provide the user with any ability to manipulate, sum-
marize,  or evaluate the information.  Because of its orientation towards land
use planning applications this is not a serious problem with respect to these
uses.  It does, however, represent a significant deficiency with respect to
CEPA requirements.  The user would still need to process much of the point  source
data manually to get the information in the desired format.

-------
                                      56
     The growth analysis for all source categories is handled by AQUIP by in-
puting an entirely new data set representing the projected information.  This
means that the system can technically treat a growth projection, but that the
user must develop and apply the growth parameters externally.  There is no pro-
vision for inputing a base data set and growth factors and having the system
generate a new data set.  This is a significant flaw in AQUIP used as a CEPA
system since the user must still do a substantial amount of manual calculation.
     Analysis of alternative control strategies is done in the same way as is
the growth analysis; that is, the user must input a new data set representing
the effects of the controls.  The program has no provision for the user to input
a base data set and a control strategy and have the system recompute the impact
of that strategy on emissions.  Here again, AQUIP has significant deficiencies
relative to CEPA requirements.  The significant exception to this is the appli-
cation of land use control strategies.  In this case the user will input an
entirely new land use plan to represent the control and there is no need to have
the program operate from a base data set.  The ease with which AQUIP can treat
land use plans makes it especially useful for these applications.
     The CEPA computer requirements are only partially met by AQUIP.  The code
is written in FORTRAN, is modular in structure, does not have only interactive
processing requirements, and uses standard data transfer procedures (i.e., tape,
cards).  The system has only been run on IBM equipment although the translation
to UNIVAC equipment should not be-a major problem since there are no highly un-
usual features to the code.  The ease of portability is unknown since the system
has not been widely used.  There are two major deficiencies with respect to
AQUIP's use as a CEPA system.  First, the existing user's manual is not easily
understandable and does not adequately describe the way in which the system can
be used.  The attempt was made to keep the program descriptions very general and
to minimize the ties to specific examples.  The result is that the average user
cannot readily determine if the system can meet his calculational requirements
and what information is needed to operate the system.  Also, there is no program-
mer's manual and it is not possible to get into the details of the code very easily
     The second, and perhaps more significant, deficiency in the computer area
is the incompatibility of AQUIP with existing emission data systems.  The program
does not accept data in NEDS format and cannot at all interface with the EIS/P&R
system.  The reasons for this are obvious; EIS/P&R was not available at the time
AQUIP was being developed and the orientation towards land use planning did not
dictate any pressing need to interface with a large emission inventory system like

-------
                                   57

NEDS.  In any case, this leaves AQUIP as being basically separate and incom-
patible with systems that are in wide use today.

5.1.4  Required Modifications
       Based on the detailed evaluation of AQUIP in Appendix A, it is estimated
that about 4-7 person-years (51-85 person-months based on the sum of the efforts
for each task) of effort would be required to modify AQUIP to meet all the CEPA
requirements.  The largest single effort (14-23mm) would be spent on bringing
AQUIP into compatibility with the computer requirements.  Substantial effort
would be needed on new coding to make the system compatible with EIS/P&R, NEDS,
and Census data.
       Significant effort would also be needed on developing a control strategy
routine that could be used to eliminate the need for the user to manually compute
a new emission inventory reflecting the effects of each strategy.

5.2  THE CAASE SYSTEM
       The evaluation of CAASE was made primarily on the basis of the docu-
mentation contained in References 8-9 and on discussions with EPA staff re-
sponsible for system development.
5.2.1  System Description
       The Computer Assisted Area Source Emissions (CAASE) system is designed to
provide a method for allocating county area source emission data to grid squares
selected on the basis of demographic features and sized to give appropriate detail
for input into air quality modeling programs.  The Research Triangle Institute
(RTI) of Research Triangle Park, N.C. was selected by EPA to do the original de-
velopment of CAASE.  RTI is currently under contract to do some additional modi-
fications on and upgrading of the capability of CAASE.
       The principal objective of the development of CAASE is to improve the
characterization of emissions from area sources.  The development program is
based on the premise that substantial amounts of data needed for determination
of area source emission are available only on the countywide level.  Since
county sizes are generally too large for use in air pollutant dispersion models,
some means of allocating these data to smaller areas or grids is needed.  Popu-
lation, housing units, and land use are among the many criteria that have been
used to make this allocation.  The development of the CAASE system was  begun as
an effort to reduce the subjectivity in selecting the appropriate grid  sizes and
to reduce the time and effort required to carry out the allocation.

-------
                                      58
       The design of the CAASE system centers on the use of the Bureau of the
Census information contained in the Master Enumeration District Listing extended
with geographic coordinates (MED-X) tapes.  These tapes contain all of the data
compiled by the Bureau of the Census for each of the enumeration districts along
with the geographic coordinates of the center of area of the district.  This
information is used to both develop an appropriate grid system and make the
allocation of data to these grids.
       Figure 5.2 gives a flow chart of the current version of the CAASE system.
This is being modified by RTI but the basic flow through the system is not
significantly altered.  CAASE currently has five computer programs associated
with it and various subroutines called by these programs.  A sixth program
(CAASEQ) has been developed to generate the data file titled "'Fuels' Totals
from Stripped NEDS Files Area Source Category" and this will be incorporated
when  the revised version of CAASE is issued.  Off-line gridding is now done
in the procedure steps between the execution of the second and third programs.
One of the modifications underway is to eliminate the need for manual gridding
at this point.  The programs have been numbered CAASE 1 through CAASE 5 and
they perform the following functions:
       CAASE1 strips the MED-X census tape files for all of the enumeration
district population entries for all counties in the Air Quality Control Region
(AQCR) being processed.  CAASE1 also converts the coordinates of the center of
each enumeration district from latitude and longitude (in degrees) to Universal
Transverse Mercator (UTM) coordinates,- which are used in dispersion modeling
programs.  CAASE1 also writes tape files to be used as input to the CAASE2 and
the CAASE4 programs.
       The current format of the CAASE2 program, using edited tape files written
by CAASE1 and the line-drawing plotter (in this application a CALCOMP plotter),
plots circles with their radii proportional to the population counts.  When all
counties, for a particular AQCR have been processed through CAASE1 and CAASE2,
a grid for the entire AQCR must be determined using partly subjective means.
In order to make this determination a light-table is used; the population
plots are overlayed onto a USGS map(s) containing all counties for the AQCR,
and a grid is manually selected for the entire AQCR.  Because determining  the
sizes of the grid squares and where they whould be placed is partially subjective,
the technical personnel performing this step should have had some experience in
gridding area source emissions using other techniques or should have been

-------
                                                    59
                                        CAASE  SYSTEM DESCRIPTION
I       INPUT VARIABLES
     TO CONTROL PROGRAM
        AMD IDENTIFY
           OUTPUT
                                                  PROGRAM CAAStl
                                             STRIPS CENSUS'PIUS AND
                                              CONVERTS COORDS.  FROM
                                              LONGITUDE AMD LATITUDE
                                                     TO UTH
                                             ERflOH HESSACE5
                                               AND INPUT
                                             INFO. NZCtSSARY/
                                              FOR PROGRAM
                                                CAASEZ

- PLOTTED _» 	


//
ERROR MESSAGES /
Am /ej ,,,,«__
DIAGNOSTICS / "™

/

Jr , , , .
gMfe 1 INPUT VARIABLES
"S^g 1 TO CONTROL PPOf7HAM fa
1* "7 1 A.SD IDENTIFY **"
r*^^^ '

•^ „
r^~H>"«>- 1 i CRID
\ DRA"^ IDENTmE.S D
\ I GRUI 1 1
\- ' 	 	
\)
\7 r+J

^l';!?UT VARIABLES
TO CONTROL P30C7.AM

DATA rO» OVFtllDINC
IU« OBJ. APPORTIONING.
FACTORS

^

P^RpCRAH CAASH




I
1
l
i
v


DRAWS AREA SOURCE



1
1
1
1
v

PROGRAM CAASE4
ASSIGNS APPORTIONING


1
|




(INPUT VADIABLZS
TO CONTROL
PROGRAM AND
IDENTIFT OUTPUT


/ ERROR MESSAGES /
/ EXPEDITB /
, _ 	 . r*J CORRECTION / V
^701 ANY ERMONEOUS/
/ GRID COORDS. /


COKPUTER-
.s^ DRAWN
^ GKID OF
AQCR


w
/ APPORTIONING\
fj FACTORS TOR 1

V AQCR S /ER70R MKSSACES /
/ / PROGRAM CAASE5/
                     'CV'SMEAREO'
                   FUELS TOTALS,
                    "SMEARED"
                    EMISSIONS,
                    4 IPP  CARD
                      IMAGES
   --.-STATION
—JiNtSC PROGRAM
 CA.1D U2CX
    PROGRAM CAASES
APPORTION'! "FUELS" AND
  EMIS.'IONS INTO THZ
   IMDIVIDU.\L GRID
      SQUARES
                                                     "FUEtS"
                                                  ^TOTALS FROM
                                                 r STRIPPED NEDS
                                                    FILES AJUA
                                                  SOURCE CATEGORT
                                                \EPA(DUR) 219 3/7J/
                                                                                           INFJT VARIABLES TO
                                                                                           CONTROL PROCHAX ,«uVD
                                                                                             IDENTIFY OUTPUT
                          Fig.  5.2.    Flowchart   of  CAASE  System

-------
                                       60

trained to use this technique.  The modifications to CAASE2 currently underway
are designed to eliminate the manual gridding step and generate the grid
system entirely by computer.
       The CAASE3 program uses the input grid description cards and draws,  to
scale, a map of the entire AQCR.   The map drawn by CAASE3 portrays the  grid,
and it is helpful in isolating any errors which may have been introduced when
preparing the load sheets or in keypunching and verifying the cards.  All grid
elements must be square and errors of omission or the incorrect recording of a
coordinate(s) are quite obvious when this map is visually checked.  A symbol,
in this application an "X," is optionally plotted at the center of each grid
square to help in the location of errors.
       After the grid description cards have been corrected, if necessary,  for
any errors found by using the CAASE3 program, the next step in the procedure
is to use the CAASE4 program which assigns apportioning values to each  of the
grid squares.  For each area source emission category included on the area
source input form, an apportioning factor has been assigned using objective data
when possible.  Bureau of the Census MED-X data tapes contain a population  count,
a housing count, and a rural/urban classification for each enumeration  district.
Each grid description card includes the side length of the grid square  from which
the area is calculated.  County totals for most of the area source emissions
categories can be objectively apportioned using population, housing, area,  or
a combination of these three measurements.  One obvious exception is the appor-
tioning of emissions from aircraft operations which would require a knowledge
of airport locations and, if more than one airport was located within a county,
their relative operations activity.  Table 5-2 illustrates the apportioning
factors used in the current CAASE system and Table 5-3 illustrates those factors
that have been decided on for the new NEDS area source format.
       The CAASE4 program logic has been written to permit the user to  subjec-
tively override any of the objective apportioning factors.  The actual  appor-
tioning factor for each source category used within the program is the product
of a weighting factor and the assigned objective factor.  This allows the user
to override the programmed (or objective)  apportioning factor within any partic-
ular  county (or counties) if information to do so is available.  The output
of the CAASE4 program includes binary tape files which are used as input files
to the CAASE5 program.  CAASE4 output files contain, for each grid square and
source category combination for each county, a number which can be used to

-------
                                    61
  Table  5-2.   Objective Apportioning  Factors Current Area Source
Category
 Number
     Major
Classification
     Minor
Classification
      Objective
Apportioning Factor
   1      Residential Fuel
   2      Residential Fuel
   3      Residential Fuel
   4      Residential Fuel
   5      Residential Fuel
   6      Residential Fuel
   7      Comm'l & Institl Fuel
   8      Comm'l & Institl Fuel
   9      Conan'l & Institl Fuel
  10      Comm'l & Institl Fuel
  11      Comm'l & Institl Fuel
  12      Comm'l & Institl Fuel
  13      Industrial Fuel
  14      Industrial Fuel
  15      Industrial Fuel
  16      Industrial Fuel
  17      Industrial Fuel
  18      Industrial Fuel
  19      Industrial Fuel
  20      Industrial Fuel
  21      On-Site Incineration
  22      On-Site Incineration
  23      On-Site Incineration
  24      Open Burning
  25      Open Burning
  26      Open Burning
  27      Gasoline Fuel
  28      Gasoline Fuel
  29      Gasoline Fuel
  30      Diesel Fuel
  31      Diesel Fuel
  32      Diesel Fuel
  33      Aircraft
  34      Aircraft
  35      Aircraft
  36      Vessels
  37      Vessels
  38      Vessels
  39      Vessels
  40      Evaporation
  41      Evaporation
  42      Measured Veh Miles
  43      Measured Veh Miles
  44      Measured Veh Miles
  45      Measured Veh Miles
  46      Dirt Rds Traveled
  47      Dirt Airstrips
  48      Construct Land Area
  49      Rock Handlg & Storage
  50      Forest Fires
  51      Slash Burning
  52      Frost Control
  53      Structure Fires
  54      Coal Refuse Burning
                     Anth.  Coal
                     Bitum.  Coal
                     Dist.  Oil
                     Resid.  Oil
                     Nat.  Gas
                     Wood
                     Anth.  Coal
                     Bitum.  Coal
                     Dist.  Oil
                     Resid.  Oil
                     Nat.  Gas
                     Wood
                     Anth.  Coal
                     Bitum.  Coal
                     Coke
                     Dist.  Oil
                     Resid.  Oil
                     Nat.  Gas
                     Wood
                     Process Gas
                     Residential
                     Industrial
                     Comm'l & Institl
                     Residential
                     Industrial
                     Comm'l & Institl
                     Light Vehicle
                     Heavy Vehicle
                     Off Highway
                     Heavy Vehicle
                     Off Highway
                     Rail  Locomotive
                     Military
                     Civil
                     Commercial
                     Anth.  Coal
                     Diesel Oil
                     Resid.  Oil
                     Gasoline
                     Solvent Purchased
                     Gasoline Marketed
                     Limited Access Rds
                     Rural Roads
                     Suburban Rds
                     Urban Roads
                     Area-Acres
                     Area-Acres
                     Orchard Heaters
                     No.  Year
                     Size of Bank
                       Housing  Units
                       Housing  Units
                       Housing  Units
                       Housing  Units
                       Housing  Units
                       Housing  Units
                       Population
                       Population
                       Population
                       Population
                       Population
                       Population
                       Population
                       Population
                       Population
                       Population
                       Population
                       Population
                       Population
                       Population
                       Housing  Units
                       Population
                       Population
                       Housing  Units
                       Population
                       Population
                       Population
                       Population
                       I/Population Density
                       Population
                       I/Population Density
                       Grid Sq. Side  Length
                       Area
                       Area
                       Area
                       Grid Sq. Side  Length
                       Grid Sq. Side  Length
                       Grid Sq. Side  Length
                       Grid Sq. Side  Length
                       Population
                       Population
                       I/Population Density
                       I/Population Density
                       Population
                       Population
                       I/Population Density
                       I/Population Density
                       Area
                       Area
                       I/Population Density
                       I/Population Density
                       I/Population Density
                       Population
                       Area
Each of  the above apportioning factors is multiplied by a weighting factor where
some are initialized as zero for all grid squares and some are initialized as 1.0 for
all grid squares.  These initial weighting factors can be overridden with input
data  if  desired.

-------
                                      62
    Table 5-3.   Objective Apportioning Factors New Area Source
Category
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Major
Classification
Residential Fuel
Residential Fuel
Residential Fuel
Residential Fuel
Residential Fuel
Residential Fuel
Cotom'1 & Institl Fuel
Comm'l & Institl Fuel
Comm'l & Institl Fuel
Comm'l & Institl Fuel
Comm'l & Institl Fuel
Comm'l & Institl Fuel
Industrial Fuel
Industrial Fuel
Industrial Fuel
Industrial Fuel
Industrial Fuel
Industrial Fuel
Industrial Fuel
Industrial Fuel
On-site Incineration
On-site Incineration
On-site Incineration
Open Burning
Open Burning
Open Burning
Gasoline Fuel
Gasoline Fuel
Gasoline Fuel
Gasoline Fuel
Diesel Fuel
Diesel Fuel
Diesel Fuel
Aircraft
Aircraft
Aircraft
Vessels
Vessels
Vessels
Vessels
Evaporation
Evaporation
Measured Veh Miles
Measured Veh Miles
Measured Veh Miles
Measured Veh Miles
Dirt Rds Traveled
Dirt Airstrips
Construct Land Area
Misc. Wind Erosion
Land Tilling
Forest Wildfires
Managed Burning
Agri. Field Burning
Frost Control
Structure Fires
Minor
Classification
Anth. Coal
Bitum. Coal
Dist. Oil
Res id. Oil
Nat . Gas
Wood
Anth. Coal
Bitum. Coal
Dist. Oil
Res id. Oil
Nat. Gas
Wood
Anth. Coal
Bitum. Coal
Coke
Dist. Oil
Res id. Oil
Nat. Gas
Wood
Process Gas
Residential
Industrial
Comm'l & Institl
Residential
Industrial
Comm'l & Institl
Light Vehicle
Light Truck
Heavy Vehicle
Off Highway
Heavy Vehicle
Off Highway
Rail Locomotive
Military
Civil
Commercial
Coal
Diesel Oil
Res id. Oil
Gasoline
Solvent Purchased
Gasoline Marketed
Limited Access Rds
Rural Roads
Suburban Rds
Urban Roads

...


• • •
Area-Acres
Area-Acres
Area-Acres
Orchard Heaters
No. Year
Objective a
Apportioning Factor
Housing Units
Housing Units
Housing Units
Housing Units
Housing Units
Housing Units
Population
Population
Population
Population
Population
Population
Population
Population
Population
Population
Population
Population
Population
Population
Housing Units
Population
Population
Housing Units
Population
Population
Population
?
Population
I/Population Density
Population
I/Population Density
Grid Sq. Side Length
Area
Area
Area
Grid Sq. Side Length
Grid Sq. Side Length
Grid Sq. Side Length
Grid Sq. Side Length
Population
Population
I/Population Density
I/Population Density
Population
Population
1 /Population Density
I/Population Density
Area
7
7
I/Population Density
?
?
I/Population Density
Population
aEach of  the above apportioning factors is multiplied by a weighting factor where
 some are initialized as zero  for all grid squares and some are  initialized as 1.0
 for all  grid squares.  These  initial weighting  factors can be overridden with
 input data if desired.

-------
                                     63
apportion a fraction of the county total into each grid square within the
county.  Each county within the AQCR is processed separately through the CAASE4
program using the grid squares associated with the county, the MED-X census
data, and any overriding weighting factors provided as additional input data.
       The CAASE5 program, using "fuel" totals for each of the emission source
categories for area sources, apportions these "fuels" into the individual grid
squares.  CAASE5 uses the same methods as those used in standard EPA programs
to calculate the emissions using fuel totals and emission factors for each of
the source emissions categories.  The term "smear" has generally been used when
describing the process of apportioning the total emissions for a county into the
grid squares within a county.  The CAASE5 program does the "smearing" by using
apportioning factors assigned by CAASE4.  CAASE5 first "smears" the "fuel" for
each of the categories into each of the grid squares and outputs (prints) a
tabular listing (and writes a binary magnetic tape) for all grid squares within
the county for each emissions source category.  For each area source emissions
category, each grid square receives a fraction of the county total - that
fraction being the number associated with that particular grid square and "fuel"
category divided by the sum of all apportioning numbers for that "fuel" category
within the county.  For any area source category, the apportioning fractions
summed over' all grid squares for that county equals unity.
       Procedurely, the pollutant emissions are calculated for the county totals
and then "smeared."  This procedure is used, rather than calculating emissions
for each grid square using "smeared" fuels, because the calculations for
"smearing" do not require as much computer time as the calculations of the
emissions.  For each source category, emissions are calculated for the five
pollutants:  suspended particles (SP) , sulfur dioxide (SCO , oxides of nitrogen
(NO ), hydrocarbons (HC), and carbon monoxide (CO).  As emissions of each
   X
pollutant are calculated and "smeared," a tabular listing is output (printed)
of the "smeared" emissions for each pollutant as was done with the fuels.  The
county totals for each emissions source category are output to indicate the con-
tribution of each of them to the total emissions for each pollutant.  For each
grid square the "smeared" emissions from all source categories are summed for
each pollutant for output in the Implementation Planning Program (IPP) expanded
card format for area source inputs.  A binary magnetic tape is also written
containing all data items in the tabular listings and card decks.  The output
from CAASE5, then, includes tables of "smeared" fuel totals and "smeared"

-------
                                      64
emissions for each of the five pollutants of interest, where for each grid
square a separate value is printed for each source category.  Also, a card
deck is punched in the IPP format, containing,  for each, grid square,  th.a
total suspended particles, sulfur dioxide, oxides of nitrogen, hydrocarbon
and carbon monoxide emissions "smeared" into each grid square for all source
categories.
5.2.2  System Use
       The CAASE system has been used in a number of applications.  The docu-
mentation for CAASE   was issued as part of the guidelines on air quailty
maintenance planning and as a result, the applications of CAASE have focused
on its use as part of an AQMA analysis.  Of the 'Seven state agencies surveyed
in Phase I of this feasibility study, three had used CAASE, at least in part,
for their AQMA analysis.  A comprehensive survey of CAASE users was not con-
ducted, but informal contacts with state and local agencies indicates that
the system is widely recognized as an available tool for air quality analyses
and has been used in a number of situations.
       The Phase I report indicated that experience with the system was mixed.
The system was presenting more problems in its implementation than the standard
dispersion models had, but this is to be expected since the system is much
more complex.  The current modifications to CAASE designed to eliminate the
manual gridding process may eliminate some of this complexity.
       Also, some questions were raised as to the accuracy of the CAASE procedure
of allocating the countywide totals to the grids on the basis of the population,
housing unit, or area allocation parameters.  As is shown in the detailed
evaluation of CAASE against the CEPA requirements, this procedure corresponds
to the Level 1 and 2 analysis; there is no provision for surrogate variable
inputs to do the more detailed calculations.
5-2.3  Comparison with CEPA Requirements
       The details of the comparison of the CAASE system against the CEPA
requirements are given in Appendix B.
       The strong point of the CAASE system is its ability to generate a master
grid system on the basis of an objective measure of population distribution.
In all of the other systemsthe user must define the master grid manually, often
on the basis of subjective judgements.  This concept may be open to challenge
using the argument that a population based grid system will not necessarily
accurately reflect the emission distribution.  That is, emissions are not

-------
                                       65
 always  distributed  in the same ways  as  the  population.   Nevertheless,  the
 majority of air  quality analyses  that have  been done  use a grid  that  is
 population-oriented.   This is accepted  practice and also provides  a method
 for  focusing on  population exposures to air pollutants.   With  regard  to  the
 CEPA requirements of  being able  to  process  several subarea sets  into  the master
 grid and of being able to map activity  into a changing master  grid, CAASE
 cannot  meet either. CAASE-Starts  with the Census Master  Enumeration Districts
 and  maps into a  population-based  master grid only.
        Another strong point of the  CAASE  system is its ability to  process Bureau
 of the  Census tapes.   This is not a trivial problem because of the large amount
 of information to be  handled  and  because  of the geographical idiosyncrasies of
 the  county and subcounty boundaries.  This  capability is  a very  strong analytical
 tool for the air quality analyst  in that  it makes available to him the full
 extent  of the Census  data.
        The CAASE system meets the CEPA  requirements for  inputing fuel consumption
 in the  residential, commercial/institutional,  industrial,  and  transportation
 sectors,  solid waste  disposal, solvent  use, and futive-dust-generating activity.
 All  of  these are input in standard  NEDS countywide format  and  allocated  to the
 grid squares on  the basis of  the  allocation parameter shown on Table 5-2.  The
 system  does not, however,  have any  provision for dealing with  surrogate variables
 and  calculating  the emission  distribution from them.  (The surrogate parameters
 of population and housing units on  the  Census  tapes are  used to  determine the
 allocation proportions only and are not used for direct  emission computations.)
      This  situation  illustrates,  the basic  design philosophy of CAASE as it
 relates to  CEPA  requirements.  CAASE was  designed to  assist in the development
 of a grid and  the allocation  of emissions to that grid.  It was not intended to
 provide substantial assistance in emission  computations.   In this  light, the
 majority of  the CAASE system  is meant to  be run only  once.  Programs CAASE1
 through CAASE3 need not  be used after the master grid is set up.   CAASE4 will
 be used only infrequently  after the initial run and serves the function of
 changing any of th.e apportioning  factors.   CAASE5 is  the only program that
 needs to be run more  than  once as it operates  on the  emission inventory, which
will change as growth scenarios and  control strategies are applied.  The.
 entire CAASE system is,  therefore,  a tool that is used to  initiate an air quality
 analysis (by developing  the grid) but is  not used to  continue  the  analysis to
 study various management and  control options.

-------
                                      66
       This is further evidenced by the way in which growth and control
strategies are handled by CAASE.  The system treats these scenarios  as inputs
in the form of NEDS area source data.  It does not provide a means for computing
what the effect of a particular growth or control strategy is, but only  computes
emissions from a specified strategy. . In essence, the user must externally
determine how an emission-producing activity is affected by growth or controls,
input these into CAASE in the NEDS area source format, and then the  system will
take over to allocate these to the grid cells.  Therefore, although CAASE
technically meets the CEPA requirements of being able to process  data indicating
the effect of growth and controls, it still requires a great deal of user  manual
calculation to prepare the input data appropriately.
       The CAASE system does not treat point sources at all.   Its design was
intended to be oriented exclusively to area sources.  To meet'CEPA requirements,
entirely new coding would be needed.  This is  tantamount  to developing  the
entire CEPA system for point sources anew.
       The CAASE system meets virtually all of the CEPA computer  requirements.
The only significant requirement that the current version does not meet  is its
ability to operate on the EPA UNIVAC 1110 computer, but the current  modifications
underway call for the UNIVAC conversion to be made.
5.2.4  Required Modifications
       From the detailed evaluation of CAASE in Appendix  B, it is estimated
that to modify CAASE to meet all of the CEPA requirements would take 5-7.5
person-years (60-89 person-months using the sum of all the tasks) of effort.
The largest efforts involve the development of point source, growth, and
control strategy routines, the- upgrading of the gridding routines to handle
other than Census Districts and populatiOEfcgrien.ted__grids,. and the development
of surrogate variable input routines.  A number of small tasks needed to
upgrade the. transportation sources also add up to a significant, effort in
this sector.
       It is evident by reviewing  the extent of  the modifications needed for
CAASE  that  the efforts amount  to almost an entirely new  system development.
This is because  CAASE was  designed to do  a very  specific 'job  and  there  was never
any need  to generalize the routines  for other  applications.   This is not a
criticism of CAASE  for it  serves a useful  function  in  performing  its design
tasks  but it casts  significant  doubt  on  the  reasonability of  attempting to
modify it to fit  CEPA requirements.

-------
                                     67
5.3  THE ESAQ SYSTEM
       The evaluation of ESAQ was made on the basis of the information con-
tained in Refs. 13-14.  These materials do not constitute formal documentation
of the system but are only general descriptions used for overview information;
formal documentation does not now exist on the ESAQ system.  To further identify
the performance of ESAQ, discussions with Engineering-Science representatives
were held.  Most of the details of the evaluations were made on the basis of
these discussions.  For this reason, the comments made about the ESAQ system
must be offered with a caveat.  The information is based on the interpretation
of verbal communications and may be subject to inaccuracies typical of this
type of procedure.  Every effort was made to clarify any points of uncertainty;
nevertheless, it is possible that the results of some of these evaluations may
be erroneous or incomplete because of the unavailability of written
documentation.

5.3.1  System Description
       The Engineering-Science Air Quality (ESAQ) system was developed as a
result of air quality analyses performed by Engineering-Science (ES) of McLean,
Virginia.  The original impetus for the development of the system came from
some studies that ES performed in Fairfax County, Virginia.  Later studies
resulted in modifications and upgrading of the system.
       The ESAQ system consists of a number of computer programs, some of which
were developed by ES and some of which were modified from codes developed by
EPA and the National Climatic Center (NCC).  Figure 5.3 illustrates the struc-
ture of the code.  The system has five major subsystems:  (1) a "Land Use"
subsystem that processes data on residential and commercial/institutional fuel
combustion and allocates area source data to subcounty areas, (2) a "Traffic"
subsystem that handles all motor vehicle sources, (3) a "New Industry" sub-
system that processes point source information, (4) an air quality and meteor-
ological data subsystem, and (5) an air quality dispersion model subsystem.
These subsystems are not entirely discrete entities in that there is some over-
lap and sharing of functions.  Also, the titles of each subsystem do not com-
pletely reflect the functions performed.
       The air quality dispersion model subsystem consists of the Air Quality
Display Model (AQDM), the APMAX model for short term, point source analyses
and the AQHIWAY model for line source analyses.  AQDM and AQHIWAY are modifi-
cations of EPA programs.  These functions are outside the range of CEPA

-------
ENGINEERING-SCIENCE  AIR  QUALITY  MODEL
                (ESAQ)
                                                        oo
         Fig. 5.3.  Flowchart of ESAQ System

-------
                                      69
requirements and the only important point to emphasize is that the structure
of the ESAQ system allows for these to be easily replaced with other models
simply by changing the preprocessing programs to generate input decks in the
desired format.
       The air quality and meteorological data subsystem includes the AQDHS
system for storing and maintaining air quality data, the AQSTAR program for
generating statistical summaries of meteorological data, the AQPREAQ program
that preprocesses air quality data into the appropriate format for the dispersion
models, and the AQLCLSTR that generates the statistical meteorological summaries
for locally generated data.  The AQDHS system is an EPA code and the AQSTAR
program was developed by NCC.  This entire subsystem is also outside the scope
of a CEPA system and will not be discussed further.
       One of  the most significant components of the ESAQ system is the "New
Industry" subsystem.  The title is somewhat of a misnomer since the subsystem
handles all sources.  The core of the subsystem is the Emissions Inventory
System/ Permits & Registration (EIS/P&R).  EIS/P&R consists of approximately
15 programs and was developed by EPA.  It is a data management program designed
to edit, update, and calculate emissions for point and area source inventories;
select and retrieve specific information; prepare emission reports;  and process
data for creation of emission scenarios.  To the full capability of EIS/P&R,
ES has added several preprocessor programs AQPREPTE, AQPREAAE and AQCOMBIN
to translate EIS/P&R output into model-compatible form and another module,
AQUPNEDS, to update residential, commercial/institutional fuel, and VMT data
in the EIS/P&R files.
       The "Land Use" subsystem contains four important codes:  AQVOL13R,
AQVOL13C,  AQLNDUSE, and AQALLOC.  The first two compute residential and
commercial/institutional, respectively, fuel use by means of the surrogate
variable procedure.  That is, number of housing units in a subarea Cfor
residential) or floor space (for commercial/institutional), building size
distribution,  fuel use distribution, fuel consumption factors, and degree-
days are input and the fuel consumption for each subarea is computed.  This
corresponds to the Level 3 analysis described in Section 2.  The third is an
updating code to change the building size distribution used in the first two.
The fourth program, AQALLOC allocates area source emissions in NEDS format
to subareas on the basis of input allocation parameters.  This program is
almost identical in operation to the CAASE 4 and CAASE5 routines operated

-------
                                       70
 in the  full  override node  (i.e.,  where  the user specifies the allocation
 parameters rather  than using  the  population-based parameters generated by the
 program).  The AQALLOC program corresponds to  the Level 1 and 2 of Section 2
 and processes the  emissions from  all  area sources included in the NEDS
 structure.
        The "Traffic" subsystem centers  around  the AQTRFGEN program.  One
 function  of  AQTRFGEN is to calculate  emissions of carbon monoxide and hydro-
 carbons from each  link, after considering such factors as type of  road, speed,
 and vehicle  mix.   The  emission information is  reported on a link-by-link basis.
 In addition, carbon monoxide  emissions  are written to  a file named AQ.TRAFIC.
 MSTR, where  data concerning the link's  location and configuration are stored.
 This file is converted to HIWAY format  by AQHWYSRT for subsequent analysis of
 carbon  monoxide concentration.  Another basic  function is to read estimated
 traffic counts on  each segment  of the highway  network  and assign vehicle miles
 traveled  to  the proper subcounty  area.  The totals  for  each,  subcounty  area
 are sent  to  the EIS/P&R system  for calculation of emissions,  and subsequently
 to AQDM for  an area-wide analysis of particulate  and sulfur  dioxide concen-
 trations.
        The purpose of  the AQHWYADD program is  to  modify a file containing data
 concerning highway links or segments that are  not maintained by TRIMS (typical
 traffic model).  The format for this file is the  same  as  that for AQ.TRAFIC.
 MSTR, which  is updated with information supplied  by TRIMS (or other traffic
 models) each time  that AQTRFGEN is run.   The AQHWYADD  file,  AQ.HWY.AQDL. SEG,
 is  used in conjunction with AQ.TRAFIC.MSTR by  the AQHIWAY preprocessor AQHYWSRT.
 The outputs are a  modified file,  and a  formatted  listing  of  the file after all
 modifications have been  performed.

        The AQHWYSRT program accesses the  highway  link  files  maintained by
 AQTRFGEN and AQHWYADD,   selects those within a  certain  radius  from a selected
 center  point, converts  average daily traffic to 1-hour or 8-hour carbon monoxide
 emissions using emission factors, and reformats the data  for  use by AQHIWAY.
 The output consists of  a file containing  those highway links  within the selected
 area with emissions greater than  zero,  and  printed  messages  indicating  how
many links were selected.  The file may be used directly  by AQHIWAY.

-------
                                      71
5.3.2  System Use
       The ESAQ system has been used for air quality analyses in 5 areas.  The
system is currently operational only on ES in-house computer (an IBM 370/165)
and has not been used outside the company.  The lack of formal documentation
and the general unavailability of the code have precluded its use elsewhere.
In its current state, the system must be viewed as an in-house program that is
not available for use by air pollution control agencies except through
Engineering-Sc ienc e.

5.3.3  Comparison with CEPA Requirements
       The details of the comparison of the ESAQ system against the CEPA
requirements are given in Appendix C.
       The ESAQ system comes closest, of all the systems evaluated, to meeting
the CEPA requirements.  Its structure, designed to meet air quality maintenance
planning needs, parallels very closely the general analytical capability required
of CEPA.  One of its strongest features is its focus on the EIS/P&R system
as the core of its data management.  This makes the system very attractive in
that it is entirely compatible with the emission inventory routines that are
being more widely accepted for use in the states.
       The major weakness of the system is its lack of documentation, its
general unavailability for use in the states, and the lack of experience with
it outside of Engineering-Science.  These are not significant problems to
overcome but they are important in that the entire evaluation of the system
must be qualified by these considerations.
       The ESAQ system can meet virtually all of the CEPA requirements for
residential, commercial/institutional, and industrial fuel combustion sources.
The lack of the ability to extract point source fuel use from input fuel use
totals is relatively minor and would require only small programming changes.
Likewise the transportation source requirements are almost entirely met.  New
coding to allow a user to input generalized growth factors would not be
difficult to develop.  The treatment of solid waste disposal sources requires
a little extra effort to allow the waste generation to be calculated on  the
basis of a surrogate variable.  Miscellaneous source treatment also requires
only small modifications.  For industrial process sources, the system does

-------
                                       72
not disaggregate growth among existing, new, and unknown sources.  This would
require some more extensive effort to program but would still not be difficult
to achieve.
       In dealing with the mapping of emissions from subareas to master grids,
the system can only deal with one subarea set and one master grid network.  As
this is primarily a bookeeping problem, the development of new code to handle
several subarea sets and/or master grids would be straightforward.
       The manner in which the system deals with growth and control strategies
is one of its weak points.  This is a function of how the EIS/P&R system  is
used.  The in-line COBOL retrieval system is used to extract those sources for
which a growth rate or control strategy is to be applied.  The user must  then
program, in COBOL, the application of each scenario  to each source category
separately.  While this process does, in fact, allow the user to deal with a
wide variety of growth and control scenarios, there are two major problems
with it.  First, the coding must be done in COBOL.  This language was not
designed to handle extensive or complex computations and may prove difficult
to use in complicated conditions.  Also, Phase I of this feasibility study
indicated that COBOL was not as widely used in the state agencies as FORTRAN.
Of the seven states surveyed, one did not have COBOL capability at all and
two others had only limited experience with it.  It may be argued that any agency
using the EIS/P&R system would, of necessity, have to have COBOL capability and
this problem would not arise.  This a valid point but the use of the COBOL
language in a computational mode to apply growth or control strategies may be
beyond the capabilities of an agency or, at best, many not be the most efficient
way to do this type of analysis.
       The second problem with the ESAQ system's growth and control strategy
procedure is that the scenario must be programmed for each source category
separately.  Where only a few source categories are affected this is not  a
problem, but when a large number of categories are involved this may be a
tedious and time-consuming chore.  Also, this process does not aid the user
in doing standard types of analyses with minimum effort.  Every scenario  must
be programmed anew as opposed to just inputing data representing the desired
conditions.
       With regard to the CEPA computer requirements, the ESAQ system satisfies

-------
                                     73
most of the needs with the exception of the availability of documentation and
the use on other computers, especially the UNIVAC.  These problems have already
been addressed.

5.3.4  Required Modifications
       From the detailed evaluations of ESAQ in Appendix C, it is estimated
that the modifications necessary to meet all of the CEPA requirements would
take about 3-5 person-years (36-61 person-months using the sum of all the tasks)
of effort.  The largest efforts would involve the development of better growth
and control strategy routines, the preparation of documentation, and the testing
of the code on other computers.
       Review of the detailed evaluations also shows that a good deal of this
cost is taken up by making a large number of relatively small modifications.
Also, these small modifications, in many cases, represent desirable although
not essential features.  Significant cost savings could be effected by reducing
the CEPA requirements to the minimum acceptable level.

5.4  THE MWCOG SYSTEM
       The evaluation of the MWCOG system was made primarily on the basis of
the information contained in Refs. 15-17.  The materials do not constitute
formal documentation of the system and were supplemented with discussions with
MWCOG staff.  The comments made regarding the MWCOG system must be tempered
with the qualification that there is no documentation and the possibility of
misinterpretation of verbal communications is present.

5.4.1  System Description
       The Metropolitan Washington Council of Governments (MWCOG) system was
developed to assist the air quality planning efforts of the Council.  Its
design was based on making use of existing data and systems, particularly
transportation-oriented, that were available to the COG.  It was intended
primarily as an in-house analytical tool but has seen some applications out-
side the Council.  It was developed entirely by MWCOG staff.
       The MWCOG system can only loosely be described as a "system."  More
accurately, it is a set of computer programs, each of which generates a

-------
                                      74
 specific output.  These outputs can be  fed  into other  programs  to obtain
 additional results.  Figure 5-4 illustrates  the relationships between the
 different codes; Table 5-4 gives a brief summary of  the  programs.
       The system begins with the calculation of fuel  combustion emissions.
 The input data consists of a 1972 demographic data base  (primarily Census
 information aggregated to planning districts) and 1980 and  1985  projections
 of this data base.  The GROWTH routine  uses  this information  to  compute growth
 factors for households (H), employment  (E),  and a parameter called "activity"
 (A=&fE).  It was discovered by some statistical analyses  that the activity
 parameter sometimes gave a better growth projection  than  either  households
 or employment alone.
       The growth factors, a 1972 fuel  use survey, and assumptions about future
 fuel use patterns are used in FUELGR to develop growth factors  for fuel con-
 sumption.  The GROW routine then proceeds to compute future emissions from
 fuel combustion and tabulates this information by planning  district.   GROW
 also receives input in the form of an area source emission  inventory  and applies
 growth factors to generate an updated inventory.  The  update is  computed by
 applying either the household, employment, activity, or  fuel growth factors
 to the current emissions.  The user can input non-demographic growth  rates
 to handle special sources (e.g., airports).
       The GROW routine is also used to compute the  effect  of changes in
 emission rates due to regulations, changes in emission factors,  etc.   This is
 done by developing an effective growth  rate  that reflects both  growth and
 changes in emission rate.
       The output of GROW is emissions  by planning district.  The CONVRT routine
maps these into grid emissions using a  table look-up procedure.   The  mapping
 can be made on the basis of area, population, employment, or any other desired
 parameter.   CONVRT prepares the emissions for input  into  any one of a number
of dispersion models.
       The EMSTJM routine takes the district  emissions  and generates a summary
by ring,  political jurisdiction, and region.
       Transportation emissions are handled  by using a travel demand  model that
operates  from data on the present transportation system and on projected

-------


cnuvi.-r
DI. IT cum


                                                                                    | NON
                                                                                    I DEMOr.RAPIITC
                                                                                    I GKOUTII KATE
      PIlOCKr.S DIfFF.I!S DEPKNI1ING OH WHICH
	|  MODI::, ir. usro.
                                                                                                                                               01
                                        Fig.  5.4.   Flowchart of MWCOG System

-------
                                                76
                   Table  5-4.   Components of the  MWCOG System
    Program
      Name
Description
 1.  ALLOK3      Version of the HANNA Model  which iterates for point and area sources until air
                 quality standard is violated.

 2.  BIOMED      Statistical Package

 3.  CALIBRATE   Applies calibration factors.

 4.  CDM         The Climatological  Dispersion Model estimates long-term concentrations of non-
                 reactive pollutants due to  emissions from area and point sources in an urban area.

 5.  COMPCOa     Converts Hanna  CO concentrations in 16- x 15 grid matrix to a 12 x 12 grid matrix
                 for input to ICOM.

 6.  CONVRTa     Converts district emissions to grid emissions for a 5 km.  and/or 2.5 km. grid system.

 7.  EGAMA       Numerical simulation model  for non-reactive pollutant analysis.
 8.  EMISa       Computes auto emissions per AP-42 Supplement #5 for years  1974  through 1992 by
                 district given  the  number of trip starts and ends along with VMT and average speed
                 per district.

 9.  EMSUM3      Program compiles district area source inventory by jurisdiction and ring.

 10. EXPOSE11     Computes percent household, employment and activities over the  primary and sec-
                 ondary  standards.

 11. FUELGRa     Program will project fuel use by district given energy use assumptions and growth
                 factors from GROWTH.

 12. GROW        Program will project area source emissions inventory by district given future fuel
                 inventory from  FUELGR, growth factors from GROWTH, nondemographic growth factors
                 (airports,  etc.), and projected auto and truck inventories along with the base year
                 inventory.

 13. GROWTH3     Computes growth factors given base«and future year projections  of housing, employ-
                 ment and activities by district.

 14. HANNA       Box model used  to estimate long-term concentrations of non-reactive pollutants due
                 to  emissions from area and Gaussian model for point sources.

 15. HIWAY       Line source model used to simulate short term CO concentration  near a roadway.
 (batch  &  in-     The model assumes Gaussian plume dispersion.
 teractive)

 16.  HIWEMF       Computes  CO emission rates  (g/sec-m) using techniques described In AP-42, Supple-
                 ment 5.   Results are used as Input to HIWAY.

 17.  ICOM3       Program  incorporates the EPA-HIWAY model, Urban Street Canyon subroutine of APRAC-1A
                 and the  Hanna-Gifford area source model used to calculate  the CO urban background.

 18 .  INTRANS      Interactive program performs many statistical manipulations to  data sets then
                visually  displays the results as graphs, maps or list of statistics on CRT terminal.

 19.  LOADEM      Converts  output from the Travel Demand Model for autos in  168 districts to trip end
                VMT data  (by special categories) for 134 districts for input  to EMIS.

 20.  LOADTRK3    Same as LOADEM  except for trucks.

 21.  MDXY        Program  converts the longitude and latitude coordinates of a  geographical point to
                the X and Y coordinates of the Maryland Plane System.

22.  PLUME3      Program  calculates both the Brlggs and Holland plume rise  in  meters at  several
                downwind distances.

23  .  PSMAP3      Program draws the outline of the region and  plots data points on a  graph plotter.
^MWCOG-developed.  Available on request.

-------
                                      77
demographic forecasts.  The output of this model is a set of trips and VMT
on the traffic planning zones.  The model itself is not part of the MWCOG
system and is a standard transportation planning tool.  The EMIS routine
computes motor vehicle emissions and feeds this information into the GROW
program for assignment to the appropriate district.
       All of the other programs in the MWCOG system  (HANNA, CDM, BIOMED,
PLUME, CALIBRATE, SYMAP, EXPOSE) are out of the scope of the CEPA system.
       Point sources are handled in the MWCOG system as input data only.  No
attempt is made to do any calculations on these data other than air quality
computations.  Growth in point source activity is handled manually.

5.4.2  System Use
       The MWCOG system has been used extensively by the Council in its air
quality analysis programs.  Approximately 20-30 different growth scenarios
for the metropolitan Washington area have been tested with the system.
       The COG has offered to give the programs to any interested party. To
date, the EMIS routine has been most in demand since it handles the com-
plexity of applying motor vehicle emission factors.  The system as a whole has
not been used outside of the agency.

5.4.3  Comparison with. CEPA Requirements
       The details of the comparison of the MWCOG system with CEPA requirements
are given in Appendix D.
       The MWCOG system is attractive as a CEPA candidate from the standpoint
of its simplicity and ease of operation.  A number of simplifying assumptions
are made that reduce the generality of the system but also make it much easier
to understand and operate.
       The system meets the fuel combustion requirements for residential and
commercial/institutional sources reasonably well.   The GROWTH routine that
allows the user input a base and projected scenario and computes growth factors
in especially useful.  Likewise, the highway vehicle emission computations are
handled reasonably well, with the exception that the user cannot easily input
specific link data; all information is handled through the travel demand model.

-------
                                       78
        The  system  is weak  in  the manner  in which  it  handles area source emissions
 other  than  fuel  combustion and highway vehicles.   The data is input to the GROW
 routine where  growth factors  are applied.  These  growth factors account for both
 the  increase in  activity and  the change  in emission  rates.   This approach is a
 simplification that enables the  user  to  avoid  getting into the details of each
 source category.   At the same time  it reduces  the accuracy of the calculation
 and  does  not allow the user to simulate  growth and/or control strategies that
 cannot be represented by a simple growth rate.  If the user does wish to do
 a more detailed  calculation he must manually compute an "effective" growth
 rate for  input into the system.
        A  second  area where the system does not meet  the CEPA requirements is
 in the handling  of point sources.   The system  was never designed to treat point
 sources other  than as input to the  dispersion  models.   This leaves a significant
 gap  in the  needs as outlined  for CEPA.
        In terms  of computer requirements, the  system's simplicity assures that
 it can function  under most  of the requirements.   The lack  of documentation is
 the  most  severe  limitation at this  point.

 5.4.4   Required ModifIcations.
        From the  detailed evaluations  of  Appendix  D,  it is  estimated that modi-
 fications to the MWCOG system to meet CEPA requirements would thake about 4-6
person-years (51-79 person-months using  the sum of all the  tasks)  effort.   The
largest efforts would be in developing routines to handle the point sources
and  adding more detailed treatments of some of  the area  source  categories.  A
significant effort would also be spent on making modifications  to  the  trans-
portation routines to handle other  than highway vehicles in more  detail  and in
allowing user input of  specific highway link data.   The  development  of a  control
strategy routine to replace the "effective" growth rate procedure would  also be
an extensive task.

-------
                                       79
                      6  COST ANALYSIS OF THE CEPA SYSTEM
       The CEPA system concept is based on the consideration that the avail-
ability of such a system will save time, effort, and money in the conducting
of an air quality analysis.  This section will summarize the evaluation of
the existing systems as well as the development of a new system on this basis.
6.1  SYSTEM DEVELOPMENT COSTS
       There are two basic procedures that can be followed in developing the
CEPA system.  One of the existing systems just described can be modified to
meet the CEPA requirements or an entirely new system can be developed.  The
comparison of these two approaches must be made on the basis of several criteria.
First, the effort required to either modify an existing system or develop a new
system must be  estimated.   This effort must be described in terms of skills
required, extent of effort, (in person-months), and personnel required on the
part of EPA staff and assisting contractors.  Second, the time required to per-
form the modifications or develop the new system must be considered.  Third,
the cost of modifications or development must be estimated.
6.1.1  Modification and Development Resource Requirements
       The estimates of the effort required to make the modifications on each
of the existing systems has already been described  (Appendices A-D).  Appendix
E gives the estimates for the development of an entirely new CEPA system.  These
effort levels are consistent with those given on the modifications in that a
major coding effort for modifying an existing system is assumed to be equivalent
to developing that piece of the CEPA system anew.
       Effort Table 6~1 summarizes the technical effort required to modify the
existing systems and to develop a new CEPA system.  It is important to note
that these effort estimates vary by about a factor of two for the programming
associated with each task category and for the entire modification or develop-
ment.  This is because past experience with the development of large  scale
computerized systems has indicated that integrating a number of independent
programs into a unified whole and identifying and correcting coding errors can
easily consume substantial amounts of time above and beyond that required to
write the first version of the programs.  The lower effort numbers should, there-
fore, be taken to represent that which is required  if no substantial  problems

-------
Table 6-1  Summary of Modification and Development Efforts for CEPA System
Category
Residential Fuel Combustion
Commercial/Institutional and Industrial Fuel Combustion
Electric Generation and Internal Combustion
Industrial Process
Transportation
Solid Waste Disposal
Miscellaneous
Gridding
Growth
Control Strategies
Computer Requirements
Effort, person-months
AQUIP
5-10
2-4
2-3
5-8
4-6
3-4
1-2
5-9
4-6
6-10
14-23
CAASE
6-10
3-6
2-3
8-10
11-15
3-6
2-3
5-7
7-11
6-8
7-10
ESAQ
1-2
-
2-3
4-7
3-4
3-6
2-4
2-4
5-9
5-8
9-14
MWCOG
3-5
-
2-3
8-10
11-18
4-8
6-8
2-4
-
6-9
9-14
New CEPA
13-23
6-12
2-3
7-10
17-27
7-14
6-13
5-10
7-11
6-10
11-16
                                                                                                CD
                                                                                                O
                           TOTAL
51-85
60-89
36-61   51-79
87-149

-------
                                       81
are encountered and the coding proceeds without the need for significant back-
tracking to trace down errors.  The larger effort numbers reflect the resources
required if there are substantial difficulties in integrating the various pieces
of the system together and if errors that are difficult to find and correct
appear regularly.
       It is evident from the table that the effort required to either modify
an existing system or develop a new system is substantial.  Even the least
effort (i.e., modification of the ESAQ system) still requires about three
person-years of work to meet all of the CEPA requirements.  The ESAQ system
modification would require the least amount of effort since, as was indicated
in the detailed evaluations, it already meets many of the requirements.  The
AQUIP and MWCOG systems require about equal effort to modify and the CAASE system
requires slightly more effort.  The low effort estimates for developing an en-
tirely new CEPA system about match the high effort estimates of modifying AQUIP,
CAASE, and the MWCOG routines.  The indications are that if the modifications run
into difficulties in integrating the codes and tracking down errors, they could
end up consuming as much effort as developing a whole new system if the develop-
ment were done efficiently and without many problems arising.
       Personnel   The personnel required to carry out the modifications or new
system development would include both EPA and contractor staff.  The magnitude
of even the smallest effort indicates that EPA in-house staff would not be able
to carry out these tasks without a significant readjustment in their current
priorities.  The staffing of the group to perform these tasks would depend on
the path chosen.  At a minimum, EPA would need to assign a project officer to
monitor the work and to provide overall policy guidance.  In addition, staff from
several EPA divisions would have to participate in setting down specific needs and
constraints that the system would have to meet.  The project officer need not be
intimately involved with systems development and would probably spend only 1/4 -
1/2 time on this program.  The other EPA staff would be involved only intermi-
mitently.
       The contractor group would require a program manager to oversee the pro-
ject and to coordinate the efforts of other personnel.  An air quality analyst,
either an engineer or a meteorologist, would be required  to ensure  that  the
proper analysis procedures are being used and that the system will  provide the

-------
                                       82
most useful outputs to the ultimate users.  A senior systems programmer would
be needed to lay the system out in the most efficient fashion from a computa-
tional standpoint.  This is important in the design of a large and complex com-
puter package that will be processing a significant amount of information.
       It is possible that one individual can function in more than one of these
roles.  For example, the program manager and the air quality analyst could be
the same individual and the senior programmer could do some of the basic coding
and debugging.  In any case, the two minimum skills required to effectively
modify an existing system or develop a new system would be those of an ex-
perienced air quality analyst and a senior programmer.
       For the  purposes of this analysis it will be assumed that the contractor
staff will be composed of the following personnel:

         1  Program Manager/Air Quality Analyst - This individual will
            be responsible for coordinating the effort and for providing
            guidance on the air quality analysis procedures to be used.
            This person will be attached to the program on essentially
            a full-time basis.
         1  Senior Programmer - This individual will have responsibility
            for the computational design and structure of the system.
            This person will also be assigned on a full time basis.
       1-3  Junior Programmers - These people will have responsibility-for
            the coding and debugging of various portions of the system.
            They will be assigned as needed.
There are obviously numerous perturbations to this scheme that could be con-
sidered.  Nevertheless, this appears to be a reasonable structure possessing
the necessary skills to do the job.  The last requirement on the number of Junior
Programmers is based on some practical considerations.  The Senior Programmer
should have at least one assistant, even in the shortest efforts, to avoid hav-
ing to spend a great deal of time on simple coding and debugging.  This assis-
tance can be used  to shorten the time required to get the system operational.
The upper limit of  three is based on the maximum number of people that could
effectively contribute  to  the program without creating undue confusion and
problems of coordination.  System development or modification could probably
not be broken  into more than  three  discrete pieces  and still have  the  end pro-
 duct  remain a  coherent  whole.   If  this maximum  group  of  five staff  could  not

-------
                                       83
perform the required tasks in the desired time it will be assumed that the
time frame will be extended rather than add additional staff.
       With regard to the four existing systems, the AQUIP, CAASE, and ESAQ
packages were developed by private contractors who could conceivably put together
the appropriate technical staff to  carry out the modifications without much dif-
ficulty.  The MWCOG system was developed by a regional planning agency for its
own needs and it is not clear that there would be any interest there in diverting
resources away from their prime mission  (i.e., planning for the Metropolitan
Washington area) into the activity required for a major modification of the system.
If the decision was made to use that system as the basis for CEPA, it might be
necessary to bring in another contractor to assist in the work.
       Time.  The effort required for either modification of an existing system
or development of a. new system has already been shown to be substantial.  Al-
though all of the effort estimates are shown in person-months, there is a limit
to how much time savings can be achieved by increasing the staffing.
       For the smallest effort  (i.e., modifying the ESAQ system) it appears rea-
sonable to assume that the basic tasks will take at least 9 months to complete.
A shorter period would probably not be reasonable in light of the fact that
about 3 person-years of work are required to make the necessary changes.  At
the upper end of the spectrum, the development of an entirely new CEPA system
that runs into significant difficulties could easily consume in excess of two
years to complete.  As an upper limit of time it will be assumed that a maximum
of 24 months will be allowed for new CEPA system development.
       It is evident from the Phase I feasibility study that time is a critical
element in making the CEPA system a useful tool for on-going air quality analyses.
On this basis, a decision to proceed with a modification of an existing system or
a new system development would probably be made with the intent of keeping the
developmental period as short as possible.
       Costs.   With the effort,  skill, and time figures given above, it is now
possible to estimate the cost of development of a CEPA system.  For the purpose
of this computation the following assumptions are made:   (1) the contractor
group will consist of one program manager/air quality analyst, one senior pro-
grammer,  and one - three junior programmers, (2) in the interest of minimizing
time for system development,  three junior programmers will be used where needed,
(3)  development time will not be less than nine months nor greater than 24 months,

-------
                                       84
(4) the lower estimate on resource requirements will assume no major problems
in programming, debugging, and testing the system while the upper estimate
represents substantial difficulties and problems encountered.
       Table 6-2 gives the cost figures to be used for each of the cost items
to be considered.  The personnel charges are averages taken from a review of
a number of proposals submitted by contractors with skills similar to those
required to do the CEPA development (no information from the four organizations
with existing systems was used in developing these data).  The monthly costs in-
clude all labor overhead and general administrative expenses.
       The program manager is assumed to be a senior staff member of the organi-
zation but not a principal.  The junior programmers are considered to be at a
level higher than technicians.  Secretarial time is computed on the basis of
I/5th of program manager and senior programmer time.  Travel expenses are com-
puted on the basis of program manager time only,  Graphics and printing is a
one time charge.  Computer use is assumed to be at the rate of seven hours per
month for each junior programmer.  Finally, a 12% profit is added to the total
cost.
       Table 6-3 summarizes the resource requirements for each of the four
modifications and for new system development.  It is evident that any approach
Caken will involve the commitment of a significant amount of resources.  The
least cost option is the modification of the ESAQ system, which is expected
given the evaluations of its performance as compared to CEPA requirements.
The most expensive option is the development of an entirely new CEPA system,
which is more than twice as costly as the ESAQ modification.  The modifications
to AQUIP, CAASE, and MWCOG are roughly comparable and are about 30-60% more
expensive than the ESAQ modification.
       All of the options can be completed within the established maximum of
24 months.  For the new system development, however, encountering significant
problems could extend the time for completion to 30 months.
                                                                     f
       One very important observation must be made from the data on the table.
With the exception of the ESAQ modification, the upper estimates for modifying
existing systems comes very close to the lower estimate of developing a new
system.  This same point was made with respect to the technical effort estimates
on Table 6-1.  The indication is that an attempt to modify an existing system
that runs into substantial problems could conceivably end up costing as much as

-------
                                       85
                     Table 6-2  Summary of Staff Costs
          Personnel
   Charge
Program Manager/Air Quality
   Analyst

Senior Programmer

Junior Programmer

Secretary
  (Time computed as 1/5 of Program
   Manager and Senior Programmer time)

Materials and Services

  Travel (per person-year of program manager
    time)

  Graphics and Printing

  Computer (time computed as 7 hours per
    person-month of Junior Programmer time)

Profit  (computed on total cost)
$5000/montha

$3900/montha

$3100/montha
           a
$2500/month
$1000

$1500

$150/hour


12%
 Includes all overhead charges

-------
                                      Table 6-3.   Summary  of Resource Requirements
AQUIP Modification
ITEM Effort
(person-
months)
Program Manager/Air
Quality Analyst 10-17
Senior Programmer 10-17
Junior Programmers 31-51
Secretary 2.0-3.4
Travel
Graphics
Computer
TOTAL 53-88.4
TOTAL with 122 profit
Minimum Time for Completion
Cost3
(1000$)
50-85
39-66
96-158
5.0-8.5
.8-1.4
1.5
33-54
225-374
$252-419
10 mons
CAASE Modification ESAQ Modification MWCOG Modification
Effort Effort
(person- Cost3 (person-
months) (1000$) months)
12-18 60-90 9-12
12-18 47-70 9-12
36-53 112-164 18-37
2.4-3.6 6.0-9.0 1.8-2.4
1 -1.5
1.5
38-56
62.4-92.6 265-392 37.8-63.4
$296-439
12 mons
Effort
Cost3 (person-
(1000$) months)
45-60 10-16
35-47 10-16
56-115 31-47
4.5-6.0 2.0-3.2
0.8-1.0
1.5
19-39
162-269 53.0-82.2
$181-301
9 mons
Cost3
(1000$)
50-80
39-62
96-146
5.0-8.0
0.8-1.3
1.5
33-49
224-348 90
$252-390c
10 mons
New CEPA Development
Effort
(person- Cost3
months) (1000$)
17-30b 85-150
17-30b 66-117
53-89 164-275
3.4-6.0 8.5-15.0
1.4-3.0
1.5
56-93
.4-155.0 383-656 O%
$429-735
17 mons
 Cost figures may not add due to roundoff.
b
 Under the assumptions of effort, the 24-month maximum  time for development may not be met  if significant problems are encountered.

 Assuming private contractor called into  assist.

-------
                                       87
a well-developed new system.  It can reasonably be argued that a modification
to an existing code is more prone to encounter problems since that code is being
changed to do tasks that are outside its initial design considerations.  The
costs in this case are more likely to tend toward the higher side of the esti-
mate.  For a new system, substantial problems requiring significant effort to
trace down are less likely to occur.  The costs, therefore, will tend to the
lower side of the estimate.  This should be given careful consideration in the
final decision on the best path to proceed.
       Sensitivity Considerations.  Given the range of estimates for CEPA
development, it is important to consider the parameters of most significance
in determining the total cost estimates.
       Starting with the smaller items on Table 6-3, the travel and graphics
charges are insignificant and make no impact on the relative merits of any
option.  The computer costs are significant but account for only 12-15% of the
total cost.  The assumed rate of computer useage of seven hours per month for
each of the junior programmers would probably not vary by more than a factor of
two and the assumed computer charge of $150 per hour might go as low as $100
per hour or as high as $300 per hour.  Despite the fact that the computer cost
could double or be halved}the resulting total cost would not change by more
than about 15%.  Thus the cost estimate is not especially sensitive to assump-
tion about computer costs.
       The effort component makes up about 85-90% of the total costs.  The charge
rates given on Table 6-2 were based on current figures quoted by contractors and
as such, would not be expected to vary by more than 15-20%.  Since the overall
cost is almost directly related to the charge rates and since expected deviations
from the assumptions used would not be more than 20%, it can be said that the
charge rate assumptions are important but will not cause a change on the order
of a factor of two in the total cost estimates.
       The only other parameter that is of significance is the amount of effort
required to carry out the tasks.  These already have a factor of two variation
in them and they strongly influence the total cost estimate.  The distribution
of effort among the skill types is not as important as the total effort since
it would be difficult to conceive of a radically different project team  (i.e.
project manager, senior programmer, three junior programmers) that could

-------
                                       88
accomplish the tasks in as efficient a manner.  The total effort required is,
therefore, the most sensitive variable in the estimate.
       In reviewing the source of the effort estimates (task-by-task estimates
in Appendices A-E) it must be emphasized that there is no way to precise in
obtaining these data.  They are based on considered judgement using past exper-
ience with computer system development.  As such, they are open to question and
revision.

6.2  SYSTEM INSTALLATION AND APPLICATION
       After a CEPA system has been developed it will be important to provide
the potential users  (i.e., state and local agencies) with instructions  in its
use and with support to resolve any problems that might occur with its  imple-
mentation.  The question of what kind of savings can be expected from CEPA
system use must also be addressed.
6.2.1  Training
       One concept that has been successfully used in introducing a new computa-
tional system to potential users is that of periodic workshops.  Users  would be
assembled for a one or two day session and given basic introductory information
on system use and potential applications.  The objective of  the sessions would
be to aid the users in getting started with the system and providing motivation
for further study.
       This procedure has been used extensively by computer  manufacturers in
getting people familiar with their hardware and software packages.  Training
sessions have been held by EPA on the use of the EIS/P&R system with a  great
deal of success.  The number of these activities indicates that the process
serves a useful function that cannot be met by providing users with written
manuals only.
       The cost of conducting such workshops should rightfully be considered
as part of the overall CEPA costs.  Table 6~4 summarizes the resources  required
to develop and conduct a series of such workshops.  The first part of the table
shows the cost of workshop development.  It assumes that the same contractor who
developed the CEPA system would also prepare the workshop materials.  The
instructional materials would include descriptions of the CEPA system,  problems
and test cases to be run for demonstration purposes, visual  aids (e.g., slides,

-------
                                        89
            Table 6-4  Resource Requirements  for Training  Workshops

 Workshop  Development
   Period  of Performance:    3 months
                                                        Effort         Cost
   Contractor Requirements:                           (Person-Months)    (1000$)
     Program Manager/Air Quality Analyst                   3            15,000
     Senior Programmer                                     3            11,700
     Secretary                                             1            2,500
     Materials and Services
       Travel                                                             300
       Graphics and Printing                                           3,000
       Computer                                                        1,000
                                        Total                          29,900
       Total with 12% Profit                                          $33,500
EPA Requirements:
  Project  Officer -                               1/2 time for 3 months
Workshop Presentation
  Duration:   2 days
  Contractor Requirements:
    Program Manager/Air Quality Analyst                 0.15              750
    Senior Programmer                                   0,15              585
    Materials and Services
      Travel                                                             500
      Computer                                                           500
                                       Total                           2,335
      Total with 12% Profit                                            $2,615
  EPA Requirements:
    Project Officer to attend each workshop -            .15  person-months per
                                                          workshop

-------
                                       90
overheads), and the use of remote terminal demonstrations of CEPA capabilities.
The workshop could be prepared in a three month period and would require  about
$33,500 of contractor assistance and about 1/2 time of an EPA project officer
to monitor the work and arrange for the logistics of conducting the workshops
themselves.
       The second part of the table indicates the cost of presenting  each work-
shop.  It assumes a two-day session and would involve two staff from  the  con-
tractor and the EPA project officer to attend.  It could reasonably be  expected
that 5-10 of these workshops would be held around the country  shortly after
CEPA becomes available.  Thereafter, additional workshops could be held every
three-six months for new users.  These follow-up sessions are  necessary because
the personnel turnover rate in control agencies requires that  new staff be
trained in the basic tasks that the agency performs.
6.2.2  System Support
       In  addition to the basic training program presented  through the  work-
shop program, it would be desirable to provide the users of the CEPA  system
with technical support and advice on any problems that arise with applications
of  the package.  This activity serves two useful functions  that will  affect the
ultimate  success of the system.  First, it provides users with expert capability
to  quickly resolve any problems that might otherwise discourage them  from fully
exploiting the capabilities of the system.   Second, it provides a mechanism to
identify  and correct problems with the system that were not uncovered in  develop-
ment and  only show up in the course of wide  application.  The  issuance  of updates
and modifications will help keep the CEPA system viable.
       Table 6"5 indicates the resource requirements for system support.   It
assumes that the contractor will provide staff to visit the state and local
agencies  that experience difficulties with the system and that these  staff mem-
bers will  assist the agency to correct the problems.  These staff will  also
prepare updates to the system reflecting deficiencies corrected and/or  caution
to be exercised when using the system for unusual applications.  The  EPA  project
officer will need to monitor the work and to arrange for publication  of the up-
dates.

-------
                                        91
              Table 6-5  Resource Requirements for System Support
Contractor Requirements;
                                               Effort              Cost
                                         (Person-Months/Year)     ($/year)

  Program Manager/Air Quality Analyst             2             10,000

  Senior Programmer                               3             11,700

  Secretary                                       1              2,500

  Materials and Services
    Travel                                                       5,000
    Computer                                                     1,000
                             Total                              30,200

    Total with 12% Profit                                      $33,800 per year

EPA Requirements;

  Project Officer - 1/4 time per year.

-------
                                       92
6.2.3  Potential System Use
       Table 6-6 summarizes the resource requirements for making a CEPA system
available to state and local control agencies.  With this considerable resource
investment it is imperative to ask what the potential savings might be to  state
and local control agencies that use a CEPA system instead of a manual procedure
to do their air quality analyses.
       To make this type of assessment it is necessary to specify the tasks in
a typical air quality analysis and to identify those areas where the avail-
ability of a CEPA system would enable a savings in time and effort to be realized,
Table 6-7 is a tabulation of those parts of an air quality analysis that relate
to emission projections and allocations.  Tasks involving dispersion modeling,
processing of air quality and meteorological data, and the like are not included
since, by definition, they are outside the scope of a CEPA system.
       To make a meaningful comparison of the merits of a CEPA system vs.  a
manual calculation procedure it is necessary to have a basis for comparing the
two methods.  Given the wide variety of circumstances for which an air quality
analysis must be performed, it appears reasonable to postulate three scenarios
to which each method may be applied.  These are:
       1.  Small data analysis - Under this scenario there are only a
           few point sources (less than about 50), a relatively small
           number of subareas and master grid cells (less than about
           100), only one or two growth scenarios that will be evaluated,
           and only one or two control strategies to be considered.
       2.  Moderate data analysis - In this instance the number of
           sources would number 100-200, the number of subareas and
           grid cells would be in the range of 200-400, and about
           4 or 5 growth scenarios and 5-10 control strategies would
           be evaluated,
       3.  Large data analysis - This situation would involve in excess
           of 400-500 sources, 800 or more subareas and master grid
           cells, more than 8-10 growth scenarios, and more than 15-20
           control strategies.
       Six people were asked to independently estimate the effort required on
the part of a control agency doing each of the three types of analyses using a
manual procedure and an automated CEPA system.  Two of the six were EPA staff
who have served as project officers on programs dealing with air quality analy-
ses,  three were Argonne staff, and one was a private contractor.  These people

-------
                                   Table 6-6   CEPA System Cost Summary

System Development
Time (months)
AQUIP CAASE
Modification Modification

10-17 12-18
EPA Staff (person-months) 2.5-8.5 3.0-9.0
Contractor Funds
r\
Training Workshops
Time (months)
EPA Staff (person
Contractor Funds
Total Investment
Time (months)
(1000$) 252-419 296-439


(1000$)
14-21 16-22
EPA Staff (person-months) 5.2-11.2 5.7-11.7
Contractor Funds
(1000$) 306-473 350-493
ESAQ MWCOG New System
Modification Modification Development

9-12 10-16 17-30
2.3-6.0 2.5-8.0 4.3-15
181-301 252-390 429-735

4
27 . i
54
13-16 14-20 21-34
5.0-8.7 5.2-16.7 7.0-17.7
235-355 306-444 483-789
System Support (annual)
Time (months)
EPA Staff (person
Contractor Funds

(1000$)
12
3.
34
                                                                                                                  vo
                                                                                                                  u>
n

 Assumes three months of preparation and one month of presentations.

-------
                      Table 6-7.   Activities  Required  for  Emission  Projection  and Allocation
1.   Mount and operate system
    Obtaining copy of code.
    Loading on computer.
    Identifying computer  bugs  -  i.e.,  incompatibilities between
      received version of code and  computer installation.
    Resolving hardware and software problems.
    Testing of system with simple  test  cases.

2.   Assemble basic data - all  systems  must have  this step.
    It is almost independent of  system used although the
    availability of a given  type of system (e.g., CAASE)
    might require getting data that would ordinarily not be
    used (e.g., Census tapes).

3.   Prepara data for analysis  -  some information must be
    processed manually; other  may be processed by machine,
    either as part of a CEPA system or  externally.
      Develop grid system.
      Select sources to be considered  as points, areas, lines.
      Identify source characteristics  needed - e.g., stack
        height, VMT on link, etc.
      Select calculation  procedure  to  be used -  Level 1, 2, 3,
        or Order 1, 2, 3.
      Determine variables needed -  surrogate parameters, fuel
        characteristics,  etc.
      Assemble or estimate needed variables.

 4.  Process  data - perform the calculations with an eventual output
    of  a point, area, and line source  emission inventory suitable
    for use  in a model.
      Identify and/or calculate  activity parameters.
      Apply emission factors to  activity parameters.
      Apply control efficiencies based  on existing regulations, com-
        pliance information, etc.
      Allocate activity and emissions  to grid cells.
      Review and correct  anomalous  data.
      Generate input  file for  dispersion model.
5.   Assemble  growth data - all systems must have this  step.  Again, depending
    on the system used, certain types of data would  be used  that would
    ordinarily  not be.

6.   Develop growth factors - using the assembled growth information,  trans-
    form these  into growth factors, specific levels  of growth,  etc.  (This
    is not a  growth analysis, but a conversion from  the planning version of
    growth to the version that can be applied to the basic data set.)

7.   Apply growth factors
      Apply the growth  factors to the base data set.
      Determine growth  at new, modified, existing sources.
      Distribute growth to known sites, projected growth sites.
      Apply emission  factors representing NSPS, SIP and other regulations-
      Allocate activity and emissions to grid cells-
      Generate projected  input file for dispersion model.

8.   Assemble control  strategy  information -  Parts of  this step are common to
    all systems, but  the  availability of a CEPA system might encourage the us
    of more detailed  Information.
      Identify types  of control  strategies - emission limits, fuel controls,
        land use controls,  traffic controls,  etc.
      Determine control level  (i.e.,  the controlled emission factor) - NSPS,
        LAFR, etc.
      Determine sources to  be  affected.

9.  Apply control strategies
      Calculate controlled  growth and development rate.
      Calculate controlled  emission  ratea  for affected sources.
      Distribute controlled activity and emissions  to known sites,
        projected sites.
      Allocate activity and emissions to grid cells.
      Review and correct  anomalous data.
      Generate controlled input file for dispersion model.
      Repeat for additional strategies  based on modeling results*

-------
                                         95
were asked to estimate effort of a senior engineer, junior engineer, junior
engineer/programmer, programmer, and data clerk(s) required to do each of the
nine tasks on Table 6-7 in each of the  three scenarios given above.  The Phase
I feasibility study indicated^ that some or all of  these personnel types would
be responsible in a control agency for  doing the air quality analyses.
       The effort estimated by each of  the six people was converted to a cost
estimate using a salary survey of air pollution control agencies plus an average
82% overhead charge.  Table 6-8 gives the cost used for each of the skill cate-
gories.  The cost difference between using the manual procedure and the CEPA
system was then computed.  The estimates varied widely, most probably because of
differing interpretations on what constituted each of the nine tasks.  For this
reason, and because of the small sample size, it was decided to reject data that
was more than one standard deviation from the mean of the six estimates and to
recompute the mean on the basis of the  remaining data.  In virtually all cases
this led to the rejection of only one or two very  high or very low estimates,
which would have reasonably been considered out of line.  Table 6-9 summarizes
the cost estimates and Table  6-10 shows the cost savings of the CEPA system.
Figure 6.1 shows the range of estimates of the  cost savings.
       The data show that some of the tasks will result in a cost penalty if a
CEPA system is used.  The cost of mounting and operating the system is an obvious
one, but some of the data collection tasks might also incur a cost penalty since
additional information might have to be collected  and coded into appropriate
formats for CEPA use.  The biggest cost savings are in the data processing,
application of growth factors, and application of  control strategies tasks.
Overall, the use of a CEPA system is estimated to save the control agencies money
under all three scenarios.  In the case of the large data analysis the savings
are in excess of $40,000 for each analysis done.
       These cost savings results must  be interpreted cautiously in light of  the
wide range of estimates shown in Figure 6.1.  In all three scenarios it was esti-
mated by at least two of the six people that the use of a CEPA system could
actually cost more.  This resulted from a change in the distribution of skills
and not from increased total effort.  CEPA would,  as estimated by these people,
require more skilled people and the time and effort savings would not be enough
offset the higher charge rate.

-------
                                     96
                  Table 6-8.  Control Agency Personnel Costs
                                                     Costa
    Skill Classification                       ($/person-months)


Senior Enginner                                     3185

Junior Engineer                                     2584

Junior Engineer/Programmer                          2584

Programmer                                          1984

Data Clerk                                          1456
a
 Includes average salary plus 82% overhead charge for fringe benefits and
 administrative expenses.

-------
                        Table 6-9.  Cost of Emission Projection and Allocation
Cost3, $1000

1.
2.
3.
4.
5.
6.
7.
8.
9.
Task
Mount and operate system
Assemble basic data
Prepare data for analysis
Process data
Assemble growth data
Develop growth factors
Apply growth factors
Assemble control strategy
information
Apply control strategies
TOTAL
Small
Manual
2.1
14.7
7.3
7.2
8.1
3.1
6.8
7.5
9.3
66.1
Analysis
With CEPA
8.2
14.9
7.9
2.8
8.1
2.4
2.2
7.5
4.4
58.4
Moderate Analysis
Manual
2.1
33.5
10.6
10.0
9.7
4.8
11.8
13.2
18.4
114.1
With CEPA
8.2
33.8
10.7
2.5
9.7
2.8
3.7
13.2
8.4
93.0
Large
Manual
2.1
52.6
18.5
25.3
20.1
9.9
21.1
16.7
28.8
195.1
Analysis
With CEPA
8.2
53.7
18.7
7.5
21.0
4.2
6.2
16.6
14.9
151.0
Computed as the mean of six estimates with data >1(J discarded.

-------
                               Table  6-10.  Summary of CEPA Cost Savings

1.
2.
3.
4.
5.
6.
7.
8.
9.
Task
Mount and operate system
Assemble basic data
Prepare data for analysis
Process data
Assemble growth data
Develop growth factors
Apply growth factors
Assemble control strategy
information
Apply control strategies
TOTAL
Cost
Small Analysis
-6.1
-0.2
-0.6
4.4
0.
0.7
4.6
0.
4.9
7.7
Savings3 »b of CEPA System,
Moderate Analysis
-6.1
-0.3
-0.1
7.5
0.
2.0
8.1
0.
10.0
21.0
1000$
Large Analysis
-6.1
-1.2
-0.2
17.8
-0.9
5.7
14.9
0.1
13.9
44.0
 Negative numbers indicate use of CEPA would be more costly.

b.T  ,
 Numbers may not add due  to rounding.
                                                                                                                 oo

-------
                              99
    150
    125
CO

CO
oc
UJ
(X,
o
o
100
CO
o
CO
CO
o
o
Q.
UJ
o
     75
 50
 25
    -25
    -50
                                                      [262
     L_   NOTE
MAXIMUM AND MINIMUM

VALUES  INCLUDE ALL
DATA.   MEANS ARE
COMPUTED WITH  DATA

> I 
-------
                                       100
       The final point of comparison is the question of whether these anti-
cipated cost savings are sufficient to offset the investment required to make
the CEPA system available.  To make this assessment it would be necessary to
project the extent of CEPA system use in air quality analyses.  This is highly
speculative in that it would require estimating the decisions that would be
made on an agency by agency basis to use CEPA or a manual procedure.  The indi-
cations from the Phase I study are that a CEPA system would at least be given
consideration in many agencies (all agencies surveyed indicated they would
consider it) and that it would be used in-house.  Nevertheless, for the pur-
poses of this study it will be more instructive to estimate how many applica-
tions the system would need to have in order for the investment costs to be
recovered.
       Figure 6.2 shows the total investment cost plotted against the number
of analyses that the CEPA system would be required to be applied to in order
to recover the cost.  The three bounding lines correspond to the cost savings
from Table 6-9 realized when doing small, moderate, and large analyses.  The
horizontal lines indicate the range of costs for the modification of the ESAQ,
AQUIP, CAASE, and MWCOG systems and for new system development.
       Using the moderate analysis line as an average indicator, it can be seen
that the investment in modifying ESAQ could be recovered if the system were
used on 10-15 applications.  In the worst case of maximum cost to develop a new
system, the investment would be recovered in about 38 moderate applications.
These values appear entirely reasonable when it is considered that there are
161 designated Air Quality Maintenance Areas all of which will have to have some
analysis done on them.  When the number of potential other applications of CEPA,
as outlined in Section 2, is considered it appears reasonable to expect that the
investment in CEPA could be recovered through the cost savings to the states.
The potential for cost savings in large analyses lends further weight to this
conclusion.  Considering the highest investment cost, a use in only 18 applica-
tions would still recover the investment.
       For small analyses, the lower limit for investment recovery is about 30
applications; the upper limit is about 102.  Although this is still less than
the total number of AQMAs needing analysis it may represent an unreasonably
high expectation for system use.  It appears that the cost savings must be
accrued through some combination of small, moderate, and large analyses.

-------
800  _
                      AQUIR\C/CASE\MWCOG
                                                                     90
100
110
120
                                 NUMBER  OF ANALYSES NEEDED
                      Fig. 6.2.   Analyses Required to Recover Investment

-------
                                      102
       For the annual cost of system support ($34,000 from Table 6"6)  only
1-4 analyses per year would be needed to recover that investment.  It  appears
that this would be easily attainable.

-------
                                      103
                      7  CONCLUSIONS AND RECOMMENDATIONS

       This Phase II feasibility study on the development of a computerized
emission projection and allocation system has focused on an evaluation of
existing computer systems in an attempt to determine if any of them could be
used to satisfy the CEPA requirements.  The evaluation methodology consisted
of (1) the definition of the CEPA requirements in general, without reference
to any existing system, (3) the comparison of each existing system to those
requirements, (3) an identification of deficiencies in existing systems, (4)
an estimation of the effort and cost required to remove those deficiences,
(5) an estimation of the effort and cost of developing an entirely new CEPA
system, and  (6) an estimation of the expected cost savings that would be
realized through use of a CEPA system.  The results of applying this methodo-
logy are given in this section.
7.1  SYSTEM EVALUATION
       Four existing computer systems were evaluated as part of this study
along with an evaluation of the development of a new CEPA system.
       The AQUIP system does not now satisfy many of the CEPA requirements.
It was designed primarily as a tool to evaluate land use plans and, as such,
it has an orientation that does not cover all of the aspects of an air quality
analysis that would be required of CEPA.
       The principal component of AQUIP that is of interest is the LANTRAN
routine.  This program provides a method of mapping arbitrarily shaped areas
into a rectangular grid.  Apart from any application as part of a CEPA system,
this routine has value in and of itself.  This mapping process can be extremely
tedious and prone to error if done manually.  LANTRAN has the potential for
providing the air quality analyst with an easier way of carrying this out.
       The primary weaknesses in the AQUIP system are its treatment of point
sources, which are handled as input with little opportunity for processing new
information, and its land use orientation.  The latter problem makes it diffi-
cult to treat non-land-use-related problems, which constitute the majority of
air quality analysis situations being dealt with today.

-------
                                      104
       The cost of modifying the AQUIP system to meet CEPA requirements was
estimated to be between $306,000 - $473,000.  Considering the extent of the
changes necessary, it is quite possible that the modification could encounter
substantial difficulties and the overall cost could approach that of develop-
ing an entirely new CEPA system.  On this basis it is recommended that AQUIP
not be considered as the foundation of the CEPA system.  However, some con-
sideration may be given to improving the documentation on the LANTRAN routine
and providing it to the states as an analytical tool to assist  in the gridding
process.
7.1.2  CAASE
       The CAASE program also does not now satisfy many of the CEPA require-
ments.  It was designed to perform the specific task of developing a master
grid based on population distribution and mapping countywide area source data
into these grids.  As such, it serves a valuable function and experience shows
that some states are attempting to take advantage of its capability.
       The principal deficiences of CAASE with regard to CEPA requirements are
its lack of treatment of point sources and its focus on the Level 1 and 2
types of analyses in distributing emissions.  Its strong points  are the ability
to handle Bureau of the Census information and the capacity to generate a master
grid network.
       The cost of modifying CAASE to meet CEPA requirements was estimated to
be $350,000 - $493,000.  As with the AQUIP system, the upper end of this estimate
approaches the cost of developing an entirely new system.  This, combined with
the fact that the bulk of the CAASE system centers around developing the master
grid, which is only peripheral to the CEPA requirements, leads  to the recommen-
dation that CAASE not be considered as the basis for a CEPA system.  CAASE
appears to have benefits by itself that do not entirely overlap CEPA require-
ments and there is little need to force-fit it into CEPA needs.

7.1.3  ESAQ
       The ESAQ system comes the closest to meeting the CEPA requirements.  It
is built around the EIS/P&R system and can perform many of the CEPA tasks.  It
was originally designed to handle a general air quality analysis and so does
not suffer from the limited objectives of AQUIP and CAASE.
       The principal deficiencies of the ESAQ system are in two areas.   First,
there is no formal documentation available.  The remarks made about the system

-------
                                      105
must be qualified by this consideration.  The system has never been used out-
side of Engineering-Science and must, therefore, be considered as an in-house pro-
gram that is not generally available.  Although this is not a serious problem
to resolve, it does cast a measure of uncertainty on the potential utility of
the system.
       Second, the system treats growth and control strategies through the use
of the in-line COBOL retrieval portion of the EIS/P&R system.  While the use of
the COBOL retrieval program is desireable from a data handling perspective, it
may become cumbersome in the application of complex growth and control strategy
scenarios.  What would be more desireable is to couple the retrieval code with
a user-oriented growth and control strategy package.  This would greatly enhance
the potential for system use.
       The cost of modifying ESAQ to meet all the CEPA requirements was estimated
to be $235,000-355,000.  This is the lowest cost of all the options considered.
Reviewing the areas on which the effort would be spent indicates that many of
the tasks are related to CEPA requirements that may not be absolutely essential.
A reduction of these cost estimates may be achieved through a scaling down of
CEPA requirements.
       On this basis it is recommended that the ESAQ system be given serious
consideration as the basic component of a CEPA system.
7.1.4  MWCOG
       The MWCOG system is actually a set of individual programs, each of
which generates a data set that is input to another program.  The principal
advantage of the system is its relative simplicity and ease of operation.
There is little reason to suspect that most of the control agencies would have
difficulty using it.
       This ease of use, however, has also lead to the biggest deficiency of  the
system.  The program makes several simplifying assumptions in the course of  the
analysis.  These assumptions are well within the requirements of EPA guidelines
but the resulting computer codes do not allow the user to do a more sophisti-
cated analysis.  Also the system does not treat point sources with any detail.
       It was estimated that $306,000-$444,000  would be required to modify
the MWCOG system to bring it up to CEPA requirements.  As with AQUIP and CAASE,

-------
                                      106
the upper limit is about equal to that required to develop an entirely new
system.  It is, therefore, recommended that the MWCOG system not be used as
the foundation for CEPA.
7.1.5  New CEPA System
       The development of an entirely new CEPA system was the most expensive
option of the five considered.  It could cost from $483,000-$789,000. .The
lower end of this cost is about the same as that required to do major modifi-
cations on AQUIP, CAASE, or MWCOG.
       In the light  of the fact that the ESAQ system already meets most  of the
CEPA requirements, it is recommended that the development of an entirely new
CEPA system not be considered as it would be duplicating existing capability.
However, should the ESAQ system be ruled as inappropriate either through the
discovery of system problems through documentation and field use or through
the consideration of its general unavailability, then the development of a
new CEPA system would be the only reasonable choice to provide the states with
the desired analytical tools.  As with ESAQ, a reduction in the CEPA require-
ments might also be used to cut the overall cost of system development.
7.2  SYSTEM USE
       For the sake of determining the extent to which a CEPA system would
have to be used to recover the investment of making the system available, the
cost savings of using a CEPA were estimated.  These estimates can also be
used to evaluate the option of not pursuing the development of any CEPA sys-
tem.
       The estimates of cost savings varied widely and it must be emphasized
that these are estimates.  There is little hard data that can be used to
firmly determine the time and money to be saved in using a CEPA system as
compared to a manual procedure.  Nevertheless, the assessments do indicate that
the potential for cost savings is significant, ranging from over $7000 per
study for small analyses to $44,000 per study for large analyses.  At this
rate it appears that even the highest cost investment in a CEPA system  (i.e.,
development of an entirely new system that runs into significant problems and
overruns) has a reasonable chance of being recovered through savings in state
and local control agency use.

-------
                                      107
       It can reasonably be argued that the indications of savings, although
admittedly of a high-risk nature, are significant enough to warrant considera-
tion of making some type of CEPA system available to the states.  The savings
in cost to the states might, in reality, be a savings in cost to EPA because
of the extensive Federal support given to air quality analyses.  The benefits
that may be accrued in improving the quality of the analyses are not included
in this assessment.
7.3  ALTERNATIVE COURSES OF ACTION
       The development of a CEPA system will involve the commitment of a
significant level of resources both in time, effort, and money.  The nature of
the information presented here indicates that this is a high-risk situation in
which the actual effort required to do the job may vary considerably (nominally
by a factor of two in all cases) and the actual savings to be realized are
based on considered judgements rather than hard data.  Nevertheless, there are
indications of considerable savings in making a CEPA system, in some form,
available.  The following options are presented as alternatives for consideration
along with the authors' recommendations on each.
7,3.1  No Further Action
       This option would cease any further consideration of developing a CEPA
system and would rely on the states and their contractors to make their own
provisions.  This is not recommended because of the potential for savings and
because the Phase I feasibility study indicated that the states would, in fact,
consider using a federally-developed system.
7.3.2  Modify AQUIP. CAASE, or MWCOG
       This option is not recommended because of the aforementioned considera-
tion that problems may be encountered that will drive the cost to that of new
system development.  Also, these systems serve purposes unique to their specific
design and need not become part of a CEPA system to see further use,
7,3.3  Initiate New System Development
       This option is not recommended at this time because of the potentially
high cost involved and because of the availability of an existing system that
meets many CEPA requirements.  This option may be the recommended  course  of
action at a later date.

-------
                                      108
7.3.4  Modify ESAQ
       This option would involve the initiation of ESAQ modifications to meet
all of the CEPA requirements.  This option is not recommended at this time
because of the uncertainty of the availability and ease of use of ESAQ  (i.e.,
resulting from the lack of documentation).  It is the authors1 feeling  that,
although the system looks extremely promising, these are too many unknowns to
commit a significant amount of resources to system development.  This option
may become the recommended course of action at some later date.
7.3.5  Proceed with Stepwise Modification of ESAQ
       This option would proceed to modify the ESAQ system in a stepwise fashion
with decision points at various milestones to determine if the system use justi-
fies continuing further.  Based on the review of the system in Section  5 it is
felt that the system has a sufficient amount of capability in its current form
to warrant issuance of it for general use by the states.  The first step in
this option would, therefore, be to prepare adequate documentation for  it to be
used by a control agency.  This step could also include the funding of  several
test applications of the system on various types of air quality analyses.  These
applications would be carried out by control agency staff.
       The second step would be to develop growth and control strategy  routines.
As these routines will interact with the EIS/P&R system, their development could
be considered as part of an EIS/P&R update and could be initiated before the
first step is completed.  Subsequent steps would add successively more  of the
CEPA requirements to the system as need demands.
       This option is recommended since it allows significant decision  points
to be incorporated into system development and also makes a useful tool avail-
able to the states in the shortest time period.  The incremental improvement
of the system should not be anymore costly than the modifications made  all at
once because of the modular nature of the system.
7.4  SUMMARY
       The results of this Phase II feasibility study has made some very
distinct points.  The development of a CEPA system will be expensive.   There
are no low cost options apart from discarding some of the requirements. The
potential cost savings also appear to be  substantial, but this is based on

-------
                                      109
estimates of effort and is not easily substantiated by actual data.  The
resulting decision on how to proceed from this point must be made with careful
consideration of all the possible outcomes.  For this reason it is recommended
that CEPA system development proceed in a stepwise fashion with adequate
decision points built into the process.

-------
110

-------
                                      Ill


                                  APPENDIX A
                    Detailed Evaluation of the AQUIP System
       The tables included in this Appendix compare the AQUIP system with
the CEPA requirements described in Sections 2 and 3.  The evaluation is based
on whether the system will do the required calculation.  If it does not, the
significance of the lack of this capability is given along with the changes
that would be necessary to enable the system to perform as desired.  An esti-
mate of the effort, in man-months (mm), of making the modification is also
given.  If the system does the required calculation, the reasohability of the
data requirements and the accuracy of the calculation are evaluated.  Finally,
any extra features of the system are identified.

-------
                                               Table A-l.    Computation  Comparison  for  Residential  Fuel
                                                                    Combustion Sources  -  AQUIP
       Calculation
.aa 1 un Update

 Fuel Use Input (Level 1,  2)

 1.  Input  state/county fuel con-
     sumption  in residentlal sector

 2.   Distribute fuel to county/sub-
     area by surrogate variable (e.g.
     d.u.,  population) distribution

 3.   Extract point sources

 it.   Go to  C.

 Surrogati-Jjariabla Input  (Level 3)

 1.   Input  state,  county,  subarea,
     surrogate variable (e.g.,
     population, d.u., floor area,
     land use)


 2.   Input  fuel consumption factors
   3.  Compute subarea fuel use
       Extract  point sources
    5.  Co to C.
                                            Dots the- system do
                                                                       Significance of
                                                                          the Lack
                                                                                                 Changes  Necessary
                                                                                 Reasonable Data
                                                                                  Requirements
                                                           Accuracy of
                                                           Calculation
                                                                                                                                                                              Extra Features
                                                                   Would be difficult for
                                                                   agency seeking  to do a
                                                                   simplified analysis.
                                          Yes.  Uses  du/acre only
                                          as surrogate
Yes.   Factors based on
Btu/d.u.-hr.  Seasonally
variable.

Yes.   Need to know
residential acreage.
                                          No.  Must  start with
                                          separate point and area
                        A small inconvenience.
                        User must manually
                        separate point and are
                        source totals.
                                                  Modification of code to
                                                  accept data on a wide area
                                                  (e.g., county ) and distri-
                                                  bute it to "figures"

                                                                2-4 mm
Modification of code to
scan point source inven-
tory to determine If any
large sources need to be
extracted from area
source totals

                1-2 mm
                                                                            Keyed almost exclusive-
                                                                            ly to data obtained from
                                                                            a land use plan.  Diffi-
                                                                            cult to use other types
                                                                            of data.

                                                                            Yes.
                                                                                                                       Yes.  Linked to  land us
                                                                                                                       plan as above but pro-
                                                                                                                       gram computes area of
                                                                                                                       "figures"
                                                       Basically the same as
                                                       Level 3.
                                                                                                                                                   More detailed since
                                                                                                                                                   factors  vary by season.
                                                                                                                                                   Same as Level 3.
C.   Emission Computation and Mapping

    1.  Map fuel  consumption to inaste
       t;rids


    «?.  Apply ^mission factors
                                          Yes

                                          Yea, |,ut  for MART IK
                                          model only.  Slight
                                          change for  use with
                                          other models.
                                                                            Yes.  Must  specify
                                                                            "figures" and master grid


                                                                            Yes.

                                                                            Yes.
                                                       Probably much less
                                                       prone to error than
                                                       nanual system.
                                                       Standard procedure.
                                                       Standard.
                                                                                                                                                                         Can deal with arbitrary
                                                                                                                                                                         and changing subareas
                                                                                                                                                                         easily.  Very desirable.

-------
                                           Table  A-l.    Computation  Comparison for  Residential  Fuel
                                                              Combustion Sources  -  AQUIP  (Contd.)

Calculation
II, Growth Analysis
A. Input Growth Data

2. Future fuel mix
B. Apply Growth Factors


Does the system do
the calculation?



Yes.
No. Must input new
actual values.
No Yes
Significance of Reasonable Data Accuracy of
the Lack Changes Necessary Requirements Calculation Extra Features



actual values.
Yes. Input future Standard
fuel mix.

                                                               values by hand.  May limit  growth factor to be
                                                               the number of growth       applied to base data.
                                                               scenarios that can be con-              „ ,
                                                                . ,   ,                              2-4 mm
                                                               sidered.


III. Strategy Analysis

    A.  Emission Limits

       1. Change emission factors            Yes,                                                                 Yes, but must change        Standard
                                                                                                             input data set.

    B.  Fuel Controls

       1. Change fuel mix                  Yes.                                                                 Yes, but must change        Standard.
                                                                                                             input data set.

       2. Change fuel characteristics        Yes.                                                                 Yes( but must change        Standard
                                                                                                             Input data set.

    C.  Growth and Development Plans

       1. Change surrogate variable          Yes.                                                                 Can change population       Basically the same
          distribution                                                                                        densities and dwelling       as Level 3.
                                                                                                             unit densities.  Again
                                                                                                             keyed almost exclusively
                                                                                                             to a land use plan.

-------
                                             Table  A-2.    Computation   Comparison  for   Commercial/Institutional
                                                                   and  Industrial  Sources  -  AQUIP
                                       Uoea Che  system do
                                        the calculation?
                                                             Significan
                                                                the Lac
                                                                                          Changes Necessary
                               Reasonable Da
                               Requirements
Accuracy of
Calculation
                                                                                                                                                                                  Extra Features
kni l_s;, i u 11 Update

A  j-ui-l Use Inuu^ (Level  i, 2)

   I.  ln|'ul  sidle/county  fuel  con

       nector
       (e.g.,  employment . land  area)
       distribution for  C/I, I  sector
                                                                  difficult  for ,
                                                                 eeking to do a
                                                                 ed analysis.
Modification of code to
accept data on a wide area
(e.g., county), separate
residential and commercial/
inst itut ional/Indus trial,  ani
distribute to figures.
                 1-2 mm
                              el 3)
       Input stJte, county, subareJ
       surrogate variable  (e.g..  popu
       latlon, J.u. floor  area,  land
       use)
    1.   Inpu
    3.  Coop LI
    -..   Extract  poit\t sources
                                  Yes.  Uses % sq.  ft.,
                                  and X coverage,  pupils/
                                  class.
                                  Btu/sq.[L-hr, Btu/class-

                                  able.

                                  Yes. Need to know  total
                                  sq. footage, coverage,



                                  No.  Must start  uitlt
                                  separate  point and area
                                  source  totals.
                                                                                                                from a Ijnd u^e  plan.
                                                                                                                Difficult Co use
                                                                                                                other types of data.
                               om fuel require-
                               nts may be diffi-
                               ]t to obtain.
                               s.  Linked to  land
                               c plan as above but
                               >gram computes a red
                                                                                                                    of "figures"
                                                              An  inconvenience.   User
                                                              musL manually separate
                                                              point and area source
                                                              totals.  May be a  signi-
                                                              ficant effort for  indus-
                                                              trial sources.
 Modification of code to scan
 point source Inventory to de-
 termine if any large sources
 need to be extracted from
 area source totals.
                                                                                                                                             Basically the same as Level  3.
                                                                                                                                         More detailed el
                                                                                                                                         vary by season.
                                                                                                                                             Same as Level 3.
Ealssion Computation j_n_J Mapping


   grids
    2.   Apply emission factors
                                                                                                                    Yes. Must  specify         Probably much  leas prone  to error    Can deal with arbitrary
                                                                                                                    "figures"  and master       than manual system.                and changing aubareas
                                                                                                                    grid.                                                     easily.  Very desirable
                                       Yes.

                                       Yes, but  for MAK'l IK
                                       model only.  Slight
                                       change for use with
                                       Othur models.

-------
Table A-2.    Computation  Comparison for Commercial/Institutional
                   and  Industrial  Sources - AQUIP  (Contd.)

Does the system do
Calculation the calculation?
II. Growth Analysis
A. Input Growth Data

2. Future fuel mix Yes
B. Apply Growth Factors No. Must input new
actual values.
No Yea
Significance of Reasonable Data Accuracy of
the Lack Changes Necessary Requirements Calculation


_ . ,
values.
Yes. Input new actual Standard.
values.
Inconvenience to user who Modification of code and
must generate new growth input data set to permit

Extra
Features





                           values by hand.  May limit  growth factor to be applied
                           the number of growth
                           scenarios that can be con-
                           sidered.
                                                  to base data.
III. Strategy Analysis
    A,  Emission Limits
       1.  Change emission factors.    Yes.


    B.  Fuel Controls
       1.  Change fuel mix           Yes


       2.  Change fuel characteris-    Yes.
          tics.

    C.  Growth and Development Plans
       1.  Change surrogate variable   Yes
          distribution
                                                                         Yes. but must change
                                                                         input data set.
                                                                         Yes,  but must change
                                                                         input data set.
                                                                         Yes,  but must change
                                                                         input data set.
                                                                         Can change population
                                                                         densities, % sq.  ft.,
                                                                         % coverage, pupils/
                                                                         classroom. Again,
                                                                         keyed almost exclusively
                                                                         to a land use plan.
Standard.


Standard.
                                                                                                  Basically the same as Level 3.

-------
                              Table  A-3.    Computation  Comparison  for  Electric  Generation
                                                  and  Internal  Combustion Sources  -  AQU1P
                                    Does the system do
                                     the calculation?
Significance of
  the Lack
                                                                                 Changes Necessary
Reasonable Data
 Requirements
Accuracy of
Calculation
                                                                                                                                                               Extra Features
Electric Generation

I.  Treat power plants similar to
                                                                                                      See comments on Indus-
                                                                                                      trial Process sources.
                                                                    See comments on Industrial
                                                                    Process sources.
                                                         Dots not give the user      Modification to code to
                                                         Information on electrical    determine electrical
                                                         requirements.
                                                                                demand from surrogate
                                                                                variable (e.g.,  population)
Internal Combustion

I.  Treat similar to industrial
   process sources
                                           See comments on Indus-
                                           trial Process sources.
                     See comments on Industrial
                     Process sources.

-------
                                Table  A-4.    Computation  Comparison  for  Industrial  Process  Sources  -  AQUIP
       Calculation
                                        Does the ayatea do
                                         the calculation?
                             Significance of
                                the Lack
                                                                                            Changes Necessary
Reasonable Data
 Requirements
Accuracy of
Calculation
                                                                                                                                                                                    Extra Features
I.   Emission Update
    A.  Receive emission Inventory
       Input.

       1.   NEDS
                                     No.  Cannot  read NEDS
                                     point source data.
                          Problem for user who must    Straight-forward  pre-

                          NEDS format.               format  or modification
                                                    of code Input require-
                                                    ments.
       2.  Other syste
       Retrieve and summarize
       inventory data.
   C.  Modify inventory with source
       specific data.
   D.  Perform internal consistency
       checks.
                                     Yes.  Point sources
                                     input: in generalized  form.
No.  Must input  entire
data  file.
                                                                Major inconvenien
                                                                MaJ or inconvcn ience.
                                                    Write entirely separate
                                                    data manipulation routines
                                                    or tie in with existing
                                                    system (e.g., EIS/P&R).

                                                                 2-3 mm

                                                    Same comment as B.
                          Inconvenience to user who   Same comment as
                          must check data separately.
       compatable form.

       1.   Point sources
       2.   Area sources
                                      Yes, buc for  HARTIK model
                                      only.  Slight change for
                                      use uich other models.

                                      Same comment  as 1.
   rt.  Input  source specific
       In format Ion.
   B.  Apply  generalized growth
       factors.
                                                                             Yes, but must input
                                                                             entire new data set.
                          Inconvenience to user who   Modification of code and
                          must generate new growth    input data set to  permit
                          manually.  May limit the    growth factor to be applied
                          number of growth scenarios   to base data.  (Same as
                          considered.                residential and commercial/
                                                    institutional).

-------
                            Table  A-4.    Computation  Comparison  for  Industrial  Process  Sources  -  AQUIP   (Contd.)
                                        Does the system do
                                         the calculation?
                                                             Significance of
                                                               the Lack
                                                                                        Changes Necessary
Reasonable Data
 Requirements
Accuracy of
Calculation
                                                                                                                                                                            Extra Features
                                                              Significant inconvenienc
                                                                                       Write new code Lo make
                                                                                       this disaggregation.
                                                          disaggregation manually.
                                                          Could present major problems
                                                          for large  inventories.

i).  Allot-jti- growth at unknown       No.  Allocation muse be     Inconvenience to  user who    Some modification of
   sources by surrogate parameter.   defined off-line and input,  must do the allocation       code.
                                 Several mud^s  are avail-    manually.
                                 able.
                                                                                                   ,  .
                                                                                                                                                                        Several methods of allo-
III. b ir^tt.!tsy__Analjs_ls_
    11.  Apply emission limits.
        Apply growth and development      Yea.
        controls.
    C.  Apply land use controls.
                                                                                                           Yes, but must change
                                                                                                           emission factors on
                                                                                                           input stream.

                                                                                                           Yes, but must change
                                                                                                           input data set.

                                                                                                           Yes.  Closely tied to
                                                                                                           land use plans.
                                                                                                                                                                                                           00
                      Better than other procedures
                      because of direct connection to
                      land use plans.

-------
                                      Table  A-5.    Computation  Comparison  for  Transportation  Sources   -  AQUIP
Calculation
Does the ayacea do
the calculation?
Significance of
the Lack
Changes Necessary
Reasonable Data
Requirements
Accuracy of
Calculation
Extra
Features
Highway Vehicles

I.   Emission Update

    A.  Fuel Consumption Input          No                        Some inconvenience  to an    Modification of code  to
       (Level 1)                                                agency seeking to do a      accept data on a uide area
                                                               simplified analysis although (e.g., county) and  distri-
                                                               this procedure is not used   bute to "figures".  (Simi-
                                                               very frequently.            lar to residential  sources.)
                                                                                                         1-2  mm
       1.  Input state/county fuel
           sold.

       2.  Estimate VMT.
       3.  Distribute VMT to sub-
           areas by surrogate
           variable (e.g., populatIon)
       It.  Go to C.

    B.  Specific Daca Input (Level  2, 3)

       1.  Input VMT, vehicle type dis-
           tribution, speed, etc.  data.

           a.  Link                   Yes                                                                            Yes.  Inputs major           Standard.
                                                                                                                    highway links and
                                                                                                                    traffic volume.

           b.  Traffic zone            Yes                                                                            Yes, although required        Vehicle density approach
                                                                                                                    input of vehicle den-         may not be as accurate as
                                                                                                                    sity in zones may make        direct zone VMT estijnates.
                                                                                                                    it difficult to use
                                                                                                                    other types of data.

       2.  Co to C.

    C.  Emission Computation and Happing
       1.  Map traffic data to master   Yes                                                                            Yes.  Must specify           Probably much less  prone        Can deal with arbitrary
           grid and/or links.                                                                                         "figures", links and          to error than annual system.     and changing subareas
                                                                                                                    master grid.                                               easily.  Very deslreable.
       2.  Apply emission factors.      Yes


       3.  Generate output  in raodel-
           compatable form.
           a.   Line sources            No                         Line source format is        Minor ch
                                                                not generated but data       format.
                                                                is available in code.

           D.   Area sources            Yes, buL  for MART1K model
                                     only.  Slight change for
                                     use with  other models.

-------
                         Table  A-5.     Computation  Comparison  for  Transportation  Sources  -  AQUIP   (Contd.)
                                     Dues  the system do
                                                              Slgnlflc
                                                                 the L
                                                                                       Changes Necessary
                                                                                                             keasonable Data
                                                                                                              Requirements
                                                  Accuracy of
                                                  Calculation
                                                                                                               Yes.  Must input as
                                                                                                               part of  a new data
                                                            growth externally.   May
                                                            ILmic the number of  growth
                                                                                  Inpu
                                                                                  grow
                                                                                  appl
                                                                                  (Sim
                                                                                                      -. and
 data sec  to allow
h factors  to be
ed to base data sets
lar to residential
es.)

          2-3 mm
111.  Stt.iLL.-gy  An.ilyais

     A.  Ap|ily emission Itmitt,
                                                                                                               Input data set.
                                                                                                                                        Simulation of traffic controls
                                                                                                                                        Is not very detailed because of
                                                                                                                                                                                                   ro
                                                                                                                                                                                                   o
                                                                                                               Yes, but must Input
                                                                                                               entirely new data set
                                                                                                                                     t e rms.
                                                                                                                                     Standard.
                                                                                                                                                                      :an simulate land
                                                                                                                                                                      ise changes very
                                                                                                                                                                      Lastly.
 I.  Act ivliy Parjmc[er Input

    *.  Input vehicle activity.
                                                                                                               Yes, but may need to
                                                                                                               tie input data  to land

                                                                                                               density function.
                                                                                                                                     Some loss of detail becaus
                                                                                                                                     land use tie.
                         11V -        Yes .
                                                                                                               lation of input data
                                                                                                               is required to have
                                                                                                               the code treat the
                                                                                                               data as emissions and
                                                                                                               not activity.
oling Handling Evaporation Loa^b

 Gasoline Marketed Input

 n.  Input gasoline sold.
II.   Surrogate Vari^blt- Input
     A.  Input per capita gas
                                                                                                               manipulation of Input
                                                                                                               data.
                                                                                                                                        Standard.

                                                                                                                                        Standard.
                                                                                                                                        Standard.
                                                                                                                                        Standard.

-------
                             Table  A-6.    Computation  Comparison  for  Solid  Waste  Disposal  Sources  -  AQUIP
      Calculation
                                        Does Che system do
                                         the calculation?
  Significance of
     the Lack
                                                                                              Changes Necessary
                              Reasonable Data
                               Requirements
                                                                                                                                                      Accuracy of
                                                                                                                                                      Calculation
 Extra
Features
I.   Emission Update

    A.  Surrogate Variable Input  (Level 1,2)

       1.  Input surrogate variable     Yes.
           to be used.


       2.  Input solid waste           Yes,
           generation factors.

       3.  Input disposal technique     Yes.
           distribution.
       4.  Compute solid waste
           generation and disposal
           technique in subareas.

       5.  Extract point sources.
                                      Ho, must  start with
                                      separate  point and area
                                      source totals.
                                                       Yes.  Uses  standard         Standard.
                                                       input to LANTRAN
                                                       routine.
                                                       Yes.                       Standard.
                                                       Yes, but  must treat         Standard.
                                                       each disposal techni-
                                                       que as a  separate
                                                       variable.
                                                       Yes.                       Ves.
A small  inconvenience.       Modification of code to scan
User must manually separate  point source inventory to
point and area sources.      determine  if any large point
                           sources  need to be extracted
                           from area  source totals.
                           (Similar to residential sources).
       6.  Go to  C.
   B.  Solid Was_te_Data Input
       (Level 3)
    1.   Input solid waste  generation
        and disposal data  from local
        sources.

    2.   Extract point sources.

    3.   Go to C.

C.   Emission Computation and Mapping
    1,   Map solid waste generation
        and disposal technique to
        master grids.

    2.   Apply emission factors.

    3.   Generate output In model-
        compa table form.
                                      Yes.

                                      Yes, but  for MARTIK
                                      model only.  Slight
                                      change for use with
                                      other models.
Inconvenience co user.
May not  be many users who
exercise this option except
for point source municipal
incinerators.
Modification of code.
                 2-3 mm
                                                       Yes.   Must specify
                                                       "figures" and master
                                                       grid.

                                                       Yes.
                                                       Yes.
                                                       Probably much less
                                                       prone to error than
                                                       manual method.

                                                       Standard.
                                                       Standard.
                                                                                                                                                                         Can deal with arbitrary
                                                                                                                                                                         and changing subareas
                                                                                                                                                                         easily.  Very desireable.

-------
Table A-6.  Computation Comparison for Solid Waste
            Disposal Sources - AQUIP (Contd.)

Does the system do
Calculation the calculation?
II. Growth Analysis
A. Input Growth Data
1. Surrogate variable Yes.
projections.
2 . Sol Id waste generation Yes .
rates .
3. Accept local solid waste No.
proj ections .
4 . Disposal techniques. Yes .



III. Strategy Analysis
A. Emission limits. Yes.

B. Growth and development Yes.
controls .


No Yes
Significance of Reasonable Data Accuracy of Extra
the Lack Changes Necessary Requirements Calculation Features


Yes. Input new Standard,
actual values.
Yea. Input new Standard.
generation rates.
Same comment as I, B Same modification as I, B.
above.
Yes, but must treat Standard.
each disposal techni-
que as a separate
variable.

Yes, but must change Standard.
input data set .
Yes, but must change Standard.
input data set .
the distribution to
new disposal processes
prior to input.
                                                                                        N>
                                                                                        N>

-------
                                    Table  A-7.    Computation  Comparison  for  Miscellaneous  Sources  -  AQUIP
         Calculation
                                          Does Che  system do
                                           the calculation?
                                                                    Significance of
                                                                       the Lack
                                                                                                Changes Necessary
Reasonable Data
 Requirements
Accuracy of
Calculation
 Extra
Feature
  Solvent Evaporation
  I.  Emission Update

      A.   Direct Data Input

          1.  Input actual solvent use.    No.



      B.   Surrogate Data Input

          1.  Input solvent consumption    Yes.
             factors

      C.   Emission Computation and^ Happing

          1.  Map solvent use to master    Yes.
             grids.


          2.  Apply emission factors.      Yes.


          3.  Generate output in model-    Yes, but for MARTIK
             compatable form.            model only.  Slight
                                        change for use with
                                        other models,

 II.   Growth Analysis

      A.   Apply growth factors.           No.
Itl.   Strategy Analysis

      A.  Emission limits.


      B.  Solvent use restrictions.
      C.  Growth and development
         controls.
                                                                   Would prevent use of actual  Modification of  code to
                                                                   data.                       accept this data.

                                                                                                           1-2 nun
                                                                   User roust apply growth
                                                                   factors externally.  May
                                                                   limit the number of growth
                                                                   scenarios considered.
                                                                                                                         Yes. Standard Input.
                                                                                                                         Yes.  Must specify
                                                                                                                         "figures" and master
                                                                                                                         grid.

                                                                                                                         Yes.  Standard
                                                                                                                         procedure.
                                                                                             Modification  of code (similar
                                                                                             to residential sources).
                                                                                                                         Yes, but must change
                                                                                                                         Input data set.

                                                                                                                         Yes, but Bust specify
                                                                                                                         this through input
                                                                                                                         data set.

                                                                                                                         Yes, but must change
                                                                                                                         input data set.
                                                                                                                                                   Standard.
                        Probably much less prone to
                        error than manual system.
                        Standard.
                        Standard.


                        Standard.
                                                                                                                                                    Standard.
 Fires

 I.   Input basic activity factor.


II.   Input a Hoc at ion  parameter.
                                                                   Must distribute to the
                                                                   "figures"  manually.
                                                                                                                          Yes,  can use standard
                                                                                                                          information.
                                                                                             Mod ifleat ion to code.
                                                                                             Similar to residential
                                                                                             sources,  Fuel Use Input.
                                                                                                                                                     Standard.

-------
Table A-7.  Computation Comparison for Miscellaneous Sources - AQU1P (Contd.)

Calculation
III. Apply emission factor.
Fugitive Dust
I . Input basic activity factor .

II . Input allocation parameter .


III. Apply emission factor.


Other Sources




II. Emission Input


No Yes
Does the system do Significance of Reasonable Data Accuracy of Extra
the calculation? the Lack Changes Necessary Requirements Calculation Features
Yes . Yes . Standard .

Yes. Yes. Can use standard Standard. JjJ
information.
No. Must distribute to Modification to code.
"figures" manually . Similar to residential
sources, Fuel Use Input.
Yes. Yes. Can use standard Standard,
information.
input data set.


Yes. Depends on the source. Standard.
Keyed heavily to land
use informat ion .
Yes. Yes. but must raani- Standard.
pulate the input data
carefully .

-------
Table A-8.  Computation Comparison for Grlddlng  -  AQUIP
Nc
Does the system do Significance of
Calculation Che calculation? the Lack
A. Map from subarea to master grid No. Computes the mapping No inconvenience to user
fractions. time if the "figures" and
grids do not change from one
run to another.
grid.
on to grid.
D. Map into changing master geld. Yes.



, Ves
Reasonable Data Accuracy of
Changes Necessary Requirements Calculation

run to another. May not be
a simple task.
2-4 ram
with care. than manual system.
3-5 mm
Yes. Easy to do since Probably less prone to error
grid is input on each than manual system.
run. Grid flexibility
points of model.

Extra
Features
One of the biggest benefits
deal with arbitrary land use
conf Igura t ions .







                                                                                          NJ
                                                                                          Ul

-------
                                            Table  A-9.    Computation  Comparison  for  Growth  -  AQUIP
Calculat ion


Does
the
Yes.
the syatet


a do


No
Significance of
the Lack Changes Necessary

Yes
Reasonable Data
Requirements


Accuracy of
Calculation
Standard .
Extra
Features

 II.  Determine  growth from gener-
     alized growth factors.
                           Inconvenience to user who
                           must  apply growth factors
                                                                 externally.   May  limit the   categories.
                                                                 number of growth  scenarios
                                                                 considered.
                           Modification of  code.
                           Discussed under  source
                                                                                                                        an entirely new set
                                                                                                                        of data for each run.
III.  Link growth  between activities.

     A.  Provide  linkages
     B.  Provide  output for data
         consistency checks.
 IV.  Process more than one growth
      scenario per run.
No.  Each scenario is
treated separately.
                           Does not supply user with
                           interpretive information.
Requires user to run each
scenario separately.
                                                                                                                        Yes.  Can do with
                                                                                                                        user-generated  sub-
                                                                                                                        routines .
Major change to code  to
identify growth in the
data set rather than  as a
separate input on each run.

          3-5 mm

Small change to code  to begin
reading new data input.
                                                                                                              As accurate as user  pro-
                                                                                                              vides  in new subroutines.
                                                                                                                                                           ro
  V.  Provide summary  tables of
      emission and activity growth.
No.   Result of treatment
of growth separately.
Does not supply user with
interpretive information.
                                                                                            Same as III,  B above.

-------
                                    Table  A-10.    Computation  Comparison  for  Control  Strategies  -  AQUIP
Calculation
Does the system do
the calculation?
Significance of
the Lack
Changes Necessary
Reasonable Data
Requirements
Accuracy of
Calculation
Extra
Features
  1.  Separate control strategy routine.   No.
 II.  Process more than one control
     strategy per computer tun.
Inconvenience to user who
must  identify controls on
each  source in the input
stream separately.  May
limit the number of strate-
gies  considered.

Requires user to run each
control strategy separately.
                                                                                         Major rework of the code to
                                                                                         have input control strategies
                                                                                         applied to input data set.

                                                                                                         3-5 mm
                           Small change  to code to begin
                           reading new data set.
                                                                                                                                                                                             to
III.  Apply  regulations only to
     affected sources.
 IV.  Provide suonary tables  for
     regulation evaluation.
Does not provide user with
interpretive  information.
                           Major change to code to
                           identify base case and
                           regulated case.

                                            3-5 m
                                                     Yes, but must Input
                                                     entire data set  re-
                                                     flecting control.

-------
                                       Table A-ll.    Evaluation of Computer Requirements  - AQUIP
          Requirement
                                                                                               No
                                   Does the System Meet
                                     the requirement?
                                                                     Significance of  the  Lack
                                Modifications Necessary
                                                                                                                                 Extra Features
      Computer System

      A.   UN I VAC 1110



      B.   IBM
                            No.
                                  Yes.  Was developed on
                                  IBM OS facilities.
Cannot be  run  on EPA facility.   Must be  converted to UNIVAC
                                form.                , _
                                                    1-2 mm
11.    Programming  Language

      A.   FORTRAN  and/or COBOL     Yes.

      B.   ANSI  standard            Unknown.
Ill.   Mode of  Operation

      A.   Batch and  interactive

      B.   Interactive  only.


IV.   Program  Structure

      A.   Modular
                            Yes.

                            No.  There is  no  Interactive
                            component.
                            Yes, but major component of
                            interest to CEPA (LANTRAN) is
                            a single routine.
                                                                                                                                 Has option to call
                                                                                                                                 user-prepared
                                                                                                                                 subroutine.
                                                                                              to
                                                                                              oo
 V.
B.  Complete or single
    module  run capability.

Off-Line Storage
      A.
          Permanent-tape,
          cards .
                                  Yes.
                                  Yes.
       B.  Transient - tape,  disk,  Yes.
          data cell. drum.
  VI.    Input  Format

        A.   NEDS  compatible.
                              No.  Must input point and
                              area sources in specified
                              form.
                                                              Cannot use NEDS data directly.
                                  Must  prepare a preprocessing
                                  routine  to convert NEDS format
                                  to AQUIP format.
        B.   EIS/P&R compatible.
        C.   Census  tapes.
                                    No.
                              No.   Does not handle census
                              at all.
  Does not  allow user to take
  advantage of EIS/P&R features.

  Census  information must be
  processed manually.
                                                                                                    Rewrite Input format  to accept
                                                                                                    EIS/P&R information.
                                                                                                                           2-3 ram
Major new code to  process
census data.
                                                                                                                            3-4 mm

-------
                                Table  A-ll.   Evaluation  of  Computer Requirements - AQUIP (Contd.)
          Requirement
                                         Does the System Meet
                                           the requirement?
                                                                    Significance of the Lack      Modifications Necessary
                                                               Extra Features
VII.  Output Format

      A.   Models


          1.  AQDM

          2.  CDM

          3.  IFF

          4.  VALLEY

      B.   Isopleth programs


      C.   Hard copy by area  or
          subarea.

VII.   Documentation

      A.   User's guide


      B.   Programmer's manual
                                 Compatible with MART1K model
                                 only.
                                 Yes.  Available for a variety
                                 of plotting uses.

                                 Yes.
                                  Yes, but is difficult  to
                                  interpret.

                                  No.
Cannot use other models
directly.
Difficult  to make program-
ming changes.
Relatively small change to
output.
                     1-3  mm
                                                                                                Manual must be prepared.
                                                                                                                     2-4 mm
                                                                                                                                                            to
                                                                                                                                                            vO
IX.    Portability

      A.  Easily transferable.


      B.  Transferred by cards,
         tape (binary or source
         form batch process).
                                  Uncertain.  Has not been
                                  widely used.

                                  Yes.  Cards and tape.
X.
      Compatibility

      A.  AEROS
                                  No.  Does not have proper
                                  documentation.
Cannot be  supported by AEROS
system.
Documentation must be
prepared.
                                                                                                                     4-6 mm

-------
                                     130

                                 APPENDIX B

                   Detailed Evaluation of the CAASE System

       The tables included in this Appendix compare the CAASE System with the
CEPA requirements described in Sections 2 and 3.  The evaluation is based on
whether the system will do the required calculation.  If it does not, the
significance of the lack of this capability is given along with the changes
that would be necessary to enable the system to perform as desired.  An esti-
mate of the effort, in man-months (mm), of making the modification is also
given.  If the system does the required calculation, the reasonability of the
data requirements and the accuracy of the calculation are evaluated.  Finally,
any extra features of the system are identified.

-------
                                                Table  B-l.    Computation  Comparison  for  Residential  Fuel
                                                                      Combustion  Sources  - CAASE

Does Che system do Significance of

I. Emission Update
A. Fuel Use Input (Level 1. 2)


Ho Yes
Reasonable Data
Changes ecessary Requirement B


Yes. Takes count/wide
input from NEDS.

Accuracy of



Standard.

        2.  Distribute  fuel to county/    Yes.  Does  this sijnul-
            subarea by  surrogate vari-    taneously with step C.I.
            able (e.g., d.u., population)  (CAASE 4 &  5)
            distribution.
        3.   Extract point sources
                                        No.  Must start with
                                        area source  totals.
        4.   Go to C.

        Surrogate Variable Input  (Level 3)  No.  There is no pro-
                                        vision for input ing
                                        any surrogate variables
                                        and making emission
                                        calculations.

        1.   Input state, county,  subarea
            surrogate variable (e.g..
            population, d.u., floor  area,
            land use)
        2.   Input fuel consumption factors
        3.   Compute subarea fuel  use

        4.   Extract point sources
        5.   Go to C.
                        A small Inconvenience.      Major new  coding effort.
                        User must manually  separ-   Program does not treat
                        ate point and area  source   point sources at all.
                        totals.                   (See Industrial Process
                                                    sources.)
                        Significant.  Does  not
                        allow user  to make  use
                        of more detailed data.
Moderate modifications and
new coding.  Must write new
emission computation sub-
routines.
                                                                                                        Allocation much less prone  to
                                                                                                        error than manual method but must
                                                                                                        start with countywlde data.  Cannot
                                                                                                        start with more detailed information.
        Emission Computation and  Happing

        1.   Map fuel consumption  to
            master grids

        2.   Apply emission factors
        3.   Generate output in raodel-
            compatable form.
Yes.  Does this simul-
taneously with step A.2.

Yes
Yes.  In IPP format.
                           Yes.
                           Yes.
                                                     Nuch  less prone  to error than
                                                     manual system.
Standard.
Standard.
II.  Growth Analysis
    A.  IjipuC Growth Data
        1.  % growth or ac
                                                                                                                     Yes.  Must inpuc new
                                                                                                                     actual values in NEDS
                                                                                                                     format.

-------
                                              Table  B-l.    Computation  Comparison  for  Residential  Fuel
                                                                   Combustion  Sources  -  CAASE   (Contd.)

Does the system do Significance of
e c


No Y
Reasonable Data
cnanges Necessary Requirements

fuel mix in NEDS
CB
Accuracy of



     B.  Apply  Grj
                                       No.  Must input ne1
                                       actual values.
Inconvenience to user.
Must do growth projections
manually.  May limit the
number of  growth scenarios
considered.
Modification of code and
input data set to apply
growth factors to base
data.
              2-4 mm
III.  Strategy Analysis

     A.  Emission Limits
        1.  Change emission factors
     B.   Fuel Controls

         i.  Change fuel mix
         2.  Change fuel cha
C.   Growth and Development Controjs

    1.  Change surrogate variable
       distribution
                                      No, since there is not
                                      provision foe  surrogate
                                      variable manipulation.
User must determine effect
of controls externally.
Hay limit the number of
controls considered.
Modification of code to
handle surrogate variables.
Similar to modifications
needed in steps I, B and
II,  B above.

                1-2 mm
                                                                                                                Yes.  Must change
                                                                                                                DATA  statements in
                                                                                                                code.
                                                   Yes.  Must change
                                                   Input data set.

                                                   Yes.  Must change
                                                   input data set.
                                                                            Somewhat prone  to error
                                                                            since appropriate factor
                                                                            to change must  be located.
                                                                                                                                          Standard.

                                                                                                                                          Standard.
                                                                                                                                                                                                      U)
                                                                                                                                                                                                      N)

-------
                          Table  B-2.   Computation Comparison  for  Commercial/Institutional
                                         and Industrial  Fuel  Combustion  Sources - CAASE

Calculation
I. Emission Update
A. Fuel Use Input (Level 1, 2)
/
sumption In comm/inst/indus sector
2. Distribute fuel to county/sub-
area by surrogate variable (e.g.,
employment, land area) distribu-
tion for comm/Inst /Indus sector.
4 . Co to C .
B. Surrogate Variable Input (Level 3)
1 . Input state , councy , subarea
surrogate variable (e.g.. popu-
lation, d.u., floor area, land use)
P P
4. Extract point sources
5. Go to C.
C. Emission Computation and Mapping
1. Hap fuel consumption to master
2. Apply emission factors
3. Generate output in mod el -
compatible form
II. Growth Analysis
A Input Growth Data
No Yes
Does the system do Significance of Reasonable Data Accuracy of
the calculation? the Lack Changes Necessary Requirements Calculation Extra Features


„ dard
input from NEDS.
Yes. Does this simul-
taneously with step C.I. Yes. Uses surrogate of Wot as good as other possible
(CAASE 4 & 5). population only. variables (e.g. employment,
land use) .
Ho. Must start with area A small inconvenience. Major new coding effort.
source totals. User must manually sep- Program does not Creac
arate point and area point sources at all. (See
source totals. Industrial Process sources).
No. There is oo provision Significant. Does not Moderate BOO if ications and
for imputing any surrogate allow user Co make use new coding. (Similar Co
variables and computing of aore detailed data. residential sources).
emissions . j .


Yes. Does this siinulta-
•anual system.
"" "" Standard.
Yes. In IPP format. Yes Standard.

1. % growth or actual values
2. Future fuel mix
Yes.  Must input new actual    Standard.
values in NEDS format.

Yes.  Must input new fuel     Standard.
mix In NEDS format.
                                                                                                                                            H
                                                                                                                                            to
                                                                                                                                            OJ

-------
                                 Table  B-2.   Computation  Comparison  for  Commercial/Institutional
                                                   and  Industrial  Fuel  Combustion  Sources  - CAASE   (Contd.)
Does che system do


,.lue,.

Significance of


manually. Hay limit the
considered .
Reasonable Data Accuracy of

Modification of code and
growth factors to basic
entlal sources) . l-2mm.





A.  Emission Li.lt.                                                                                     Vea  Must change DATA          s
-------
                                         Table B-3.   Computation  Comparison for  Electric Generation
                                                          and Internal  Combustion  Sources - CAASE
Calculation
Does the system do
the calculation?
Significance of
the Lack
Changes Necessary
Reasonable Data
Requirements
Accuracy of
Calculation
Extra
Features
Electric Generation

I. Treat power plants similar to
   industrial process sources.

II. Project electrical demand and
   load factors.
See comments on Industrial Process Sources.
Ddes not give the user
information on electrical
requirements.
Major modification of code to
handle surrogate variables and
point sources (see Residential
and Industrial Process) and
determine electrical demand.

                2-3 mm
                                                                                                                                                                             U)
                                                                                                                                                                             LA
Internal Combustion

I,  Treat similar to industrial
   process sources
See comments on Industrial Process Sources.

-------
                                   Table  B-4.    Computation  Comparison  for  Industrial  Process  Sources  -  CAASE
Does
Calculation the
the syi
calculi
stem do
ition?
Significance
the Lack
of
Changes
Necessary
Reasonable Data
Requirements
Accuracy of
Calculation
Extra
Features
 I .  EmlasIon Update
    A.  Receive emission  inventory       N
        input.
        1.   NEDS
        2.   Other Systems
    B,  Ret rieve and summarize
        inventory data.

    C,  Modify inventory with source
        specific data.

    D.  Perform internal consistency
        checks.

    E,  Generate output in model-
        compatable form.
        1.   Pa Int sources.
        2.   Area sources

II. Growth Analysis
    A.  Input source specific growth
         information.

     B.  Apply generalized growth
         factors.

     C.   Disaggregate growth to existing,
         new, and unknown  sources.

     D.   Allocate growth at unknown sources
         by  surrogate parameter.

Ill-  Strategy Analysis
     A.   Apply emission limits.

     B.   Apply growth and  development
         controls.
CAASE was never designed to handle point sources.  It is
strictly an area source computation procedure.  Modifica-
tion for it to handle point sources would be equivalent
to writing an entirely new code in the  form of  a new CEPA
system.
                                              8-10 mm
                                                                                                                               U)
     C.  Apply land use controls.

-------
Table B-5.  Computation Comparison for Transportation Sources - CAASE
Calculation
Highway Vehicles
I. Emission Update
A. Fuel Consumption Input (Level)
/
sold.
2. Estimate VHT.

3. Distribute VMT to subareas
by surrogate variable
(e.g. , population) .
t. Go to C.
B. Specific Data Input (Level 2,3)
1. Input VMT, vehicle type dis-
tribution, speed, etc. data.


a. Link
b. Traffic zone
2. Co to C.
C. Emission Computation and Happing
1. Hap traffic data to toaster
grid and/or links.

2. Apply emission factors.
3. Generate output in model-
compatible form.
a. Line sources


b. Area sources

II. Growth Analysis
I
data.

B. Input generalized growth pro-
jections.



No Yes '
Does the system do Significance of Reasonable Data Accuracy of
the calculation? the Lack Changes Necessary Requirements Calculation Extra Feat




wide Input from NEDS.
Yea . Yes . Takes aunty- Standard .
wide input f om NEDS.
Yes. Does this Yea. Uses s rrogate Standard at this
simultaneously of populatio or level of analysis.
with Step C.I. population d naity.


No. Significant lack. Pre- Significant new code de-
eludes the use of local velopment to handle other
VMT data. than countywlde data.
4-6 mm




Yes. Does this Yes. Standard.
simultaneously
with Step A. 3.
Yes- Yes" Standard.


Ho. Does not Cannot run certain Significant new code de-
treat line sources models (e.g. liltMY). velopment in connection
at all. with I.B.
Yes. In IPP for- Yes. Standard.
mat .

Yes, but can only Yea. Must input new Losa o£ spatlal
deal with county county total VMT. resolution of new
Cocals- facilities.
Must do the growth and input variables to
May limit the number to base data set.
of grouch scenarios (Similar to residential
considered. sources), - 2 mm
                                                                                              co

-------
                        Table  B-5.    Computation  Comparison  for  Transportation  Sources  -  CAASE  (Contd.)

Ill .
OTHER
II.
No *ea
Does Che system do Significance of Reasonable Data Accuracy of
atlon he caicu a lonr g y q Extra Featured
S L rat «ay__An_aly_a I a
A. Apply L-mlMBiun limits. Yes. Yes. Must change Standard.
DATA statements in
code .
to these controls class of control strate- handle other than county-
since data Is on gles that could be con- wide data. (Similar to
county level. sidered. I.B.l)
2-3 aim
C. Apply growth and development Yee. Yes. Must input new Standard.
controls data set In NEDS
format .
VEHICLES
Activ tty Parameter Input
A 
-------
Table B-6.  Computation for Solid Waste Disposal  Sources  -  CAASE

Calculation
I. Emission Update
1. Input surrogate variable to
be used.
2. Input solid waste gener-
ation factors.
3. Input disposal technique
distribution.
ated and disposal technique
in subareas.
6. Go to C.
B. Solid Waste Data Input (Level 3)
IT
cion and disposal data from
local sources.
3. Go to C.
1. Map solid waste generation
and disposal technique to
master grids.
2. Apply emission factors.
compatible form.
11. Growth Analysis
A. Input Growth Data

tlons.
No Yes
Does the system do Significance of Reasonable Data Accuracy of
the calculation? the Lack Changes Necessary Requirements Calculation Extra Features

No. There is no Significant. Does not Moderate modifications and
provision for in- allow user to make use new coding. (Similar to
putting any surro- of different estiraat- residential sources.)
gate variables and ing procedures. 1-2 mm
making emission
calculations.



from NEDS.
with area source Program does not treat
totals. point sources at all.
(See Industrial Process
Sources.)
Yea. Does ibis Yes. Much less prone fcfl
simultaneously error than aamal
with Step B.I. system.
*«• Ves' Standard.
U>m firm" Standard.


new actual values. Must do growth projec- accept surrogate varl-
tions manually. May ables and apply growtli
limit the number of factors.
growth scenarios con- 2-4 mm
sidered.
                                                                                              U>
                                                                                              vO

-------
Table B-6.  Computation for Solid Waste Disposal Emissions - CAASE (Contd.)
No Yes
Does the system do Significance of Reasonable Data
Calculation the calculation? the Lack Changee Necessary Requirements
eol id waste gentM.it Ion No. Modification of code in
i,Ufd. conjunction with II, A, 1
above.
protections. actual data in NEDS
format .
A . Ills [nit, al techniques . Yes. Yes. Each technique
must be identified in
A. Kuiiti.s IDU 1 1ml ts. Yes . Yes. Must change
DATA statements in
code.
B. (ViowLh and development Yes. Yea. Must change In-
.onirols. put data set.
put data set.

Accuracy of

Standard.
Sotoehwac prone to
error since appro-
priate factor Co
change must be
located.
Standard.


-------
          Calculation
Sol vi-ni Kvjporac too
                                      Table  B-7.    Computation  Comparison  for  Miscellaneous  Sources  -  CAASE
                                           Doe a i he sy^tcia Ju
                                            thu calculation?
                                                                     SlBnifIcunce of
                                                                                                                          RequtreuenCu
                                                          Accuracy of
                                                          Calculation
                                                                                                                                                                                    Extra Features
        l.  Input  actual solvent


    ** •   Surrogate j)jtg
                                                                                                                             Standard NEDS
       factors.


C.  Ealaa tun Cuuipiiuij jqii and^jj

   1,  Map £iulvi.-nl  u=>e to BUS
        2.  Apply emission  t-it
                                           Yes for  1PE\ AQUM
                                           cUangt:  for ubti
                                           with other models
                                                                  Inability Co manipulate
                                                                  tliis type of data.
HoJiflcatton of coUe ,
Input data.
1-2 on
                                                                                                                       Yea.   Staitilard
                                                                                                                       cedure.
                                                                                                                                                   Probably nuch  ICH
                                                                                                                                                   prone Co error Ch
                                                                                                                                                   •onual ayecen.
                                                                                                                                                   Standard.
    A.  Apply
                                                                  l)SK=r uiuit apply growt
                                                                  factors externally.
                                                                  Hay limit th£ number
                                                                  of Drouth scenarios
HodIficaiIon of code.
(Similar to resident Ul
sources.)
    u.  l.rowth and Jeve Iu
                                                                                                                       Yea,  but must  change
                                                                                                                       factors in DATA
                                                                                                                       stateoent.
                            Yort. but anit spue 1fy
                            tlil.s tlirou^li Input
                            data att.
                            Yer.. but must change
                            Input dara .set .
                                                        Hay be prone to none
                                                        error nince appro-
                                                        priate data tu change
                                                        must be located.

                                                        Standard.

-------
                   Table B-7.   Computation Comparison  for  Miscellaneous Sources  -  CAASE  (Contd.)
No Yea
Uoutt Lhv ayatta do S Ijjntf Icaocfc of Reasonable Data
Calculation the calculation? the Lack Changea Nt=c«uaary Requlruyeuta
El-



met era or Input an
I Input b^slL activity factor. *ca. Yea> Cun UMe Htund_
ard NEDS Information.
acd allocation para-
meters or Input an
overriding parameter.


Input data set.

Accuracy of

Standard.

Standard.

Standard,
Standard.
Standard.
Standard.


1. CcnetalIzed formal.
      on  Input.
                              No.  Usts NtUS toc-
                              uai only.
                              No.  Usud NtOS for-
                              uut only.
Minor nodtfIcdtiuii
code to accept aJdl
t Ion ill input.


Minor modification
above.

-------
Table B-8.  Computation Comparison for Grlddlng - CAASE
Ho Yea
Uotiu the system du Significance of Heattonable Data
Calculation Che calculation? Che Lack Change* Necessary Requirements
A. Map Iron uubdfea to utidter grid using Yen. Yea. Uaua geucoded
standard tapes.

source. sources. Equivalent Co
writing new package.
2-3 uu
emissions.
ter grid based on grid system. Cannot j_^ aju
population density. use any oilier than
populat Ion oriented
grid.
Accuracy of
Calculation Ektra Feat urea
Much leaa prone co Helpa develop the Master grid baaed
ayscea. feature buC aonewhat external Co
CEPA.








                                                                                        OJ

-------
111.  Link growLh bclu
                                               Table  B-9.    Computation  Comparison   for  Growth  -  CAASE
                                            Doctt lite eiy«tea do          S ignif Icdticn of                                       Kiintjunab) e Dattt               Accuracy of
          C>lcul*t Ion                         lh«c*lcolai loo?             the Lack                Chung tin Nee tmuncy             fieijulreoient a                 Calculac Ion
                                                                  Inconvenience to user.       HodlfUaclou
                                                                  rtjbl  do growth piojuc-       Input,  (See
                                                                  tlon* BdQudlly.  Hay          sources.)
                                                                  Halt the nuub^r of

                                                                  alderea.
                                            Ho.  DJta Input in      Inconvenience to ut.er.       Mudltlcation of  Input
                                            NEDS fa nut only.        Hit it  «akt the tltikages       fornac and codlitg.
                                                                  lunually.                                   2-3 am
                      lor dtfL
-------
Table B-10.  Computation Comparison for Control Strategies - CAASE

Doea the uy«tea do Significance of

Must apply strategies
separately.
egy per computer run.
sources.
pretlve Information.
Ho Ic*
KeMganable Data Accuracy of

change Input data reflect-
ing control.
StCB.

neat regulated and un~
                                                                                              Ul

-------
                                      Table B-ll.   Evaluation  of  Computer  Requirements  -  CAASE
         Requirement
                                                                                             No
     Does  Che  System Meet
       the requirement?
                                                                    Significance  of  the Lack
Modifications Necessary
                                                                                                                                Extra Features
1 .    Computer System

     A.   UNIVAC 1110


     B.   IBM
No.
                                    Yes.  Was developed on
                                    IBM OS System.
                                                            Is currently undergoing modi-
                                                            fications  for use on UNIVAC 1110.
11.   Programming Lan

     A.   FORTRAN and/or COBOL        Yes.

     B.   ANSI  standard               Yes.
HI.  Mode ot Operation

     A.  Batch and interactive.

     B.  Interactive only.
Yes.

No.  There is  no  inter-
active component.
 IV.  Program Structure

     A.  Modular
                                   Yes, but output of some
                                   modules have relevance to
                                   other modules only.
     B.   Complete or single module   No.  Must run in 5 steps.
         run capability              (See Growth)
                              Not significant for CAASE 1-3  Major  effort for CAASE 1-3 since
                              since these  are generally run  an intermediate step of hand-
                                                            plotted  grids  Is necessary.  Minor
                                                            coding change, or JCL change, for
                                                            CAASE  4  and 5.  Currently under
                                                            extensive modification to elimi-
                                                            nate hand-plotted grids.
                                                                 only once.  An inconvenience
                                                                 to user for CAASE 4 and  5.

-------
                          Table B-ll.   Evaluation of Computer  Requirements - GAASE  (Contd.)
           Requirement
    DOCK  the System Meet
      the requirement?
                                                                                             No
  Significance of  the Lack      Modifications  Necessary
                                                                                                                              Extra Features
 V.    Office Storage

      A.  Permanent - tape,  cards.
      B.  Transient - tape,  disk,
          data cell, drum.
Yes.  Everything is tape.

Yes.  Everything is tape.
VI.   Input  Format

      A.   NEDS compatible.
      B.   EIS/PiR compatible.
      C.   Census tapes.
Yes.  Takes  input from NEDS.
No.
Yes.   Takes  input from
Census tapes.
                             Does not allow user  to take    Moderate  effort Co change
                             advantage of EIS/P&R          input  format to accept EIS/P&R
                             capabilities.                 Information.
                                                                               2-3 mm
                                                                                                                             Significant capa-
                                                                                                                             bility to process
                                                                                                                             Census tapes.
VII.   Output Format

      A.  Models

         1.  AQDM

         2.  COM

         3.  IPP

         4.  VALLEY

      B.  Isopleth programs
      C.  Hard copy by area
         or subarea
Yes.

Yes.

Yes.

Yes.

No.
Yes.  Whole area comes out
broken down by grid.
Does not allow user to map
emission densities.
                                                          Relatively minor modification
                                                          to add standard plotting
                                                          package.
                                                                               1 mm

-------
                                        Table  B-ll.   Evaluation of Computer  Requirements  -  CAASE  (Contd.)
           Requirement
       Does  Che System Meet
         the  requirement?
                                                                                              No
                                                                     Significance  of  the Lack
Modifications Necessary
                                                                                                                                Extra Features
V11I.   Documental ion

       A.  User's guide

       B.  Programmer's  manual
Yes.  New one being prepared.

Ves.  Only in Air Quality
Maintenance Guideline (Vol.  8).
IX.     Portability

       A.  Easily transferable
       B.  Transferred by cards,
          tape (binary or  source
          form, batch process) .
                                                                                                                        oo
Yes, but  some plotting CALLS are
specific  to University of N.C.
(system developer), and not easy
to get.

Yes.  Cards and tape.
 Is  being changed to
 identify IBM plotting
 packages.
X.
       Compatibility

       A.  AEROS
                                   No.  Does not have proper
                                   documentation.
                              Cannot be  supported by
                              AEROS system.
Documentation must be
prepared.
              4-6 mm

-------
                                      149

                                 APPENDIX C
                  Detailed Evaluation of the ESAQ System
       The tables included in this Appendix compare the ESAQ system with
the CEPA requirements described in Sections 2 and 3.  The evaluation is based
on whether the system will do the required calculation.  If it does not, the
significance of the lack of this capability is given along with the changes
that would be necessary to enable the system to perform as desired.  An esti-
mate of the effort, in man-months (mm), of making the modification is also
given.  If the system does the required calculation, the reasonability of the
data requirements and the accuracy of the calculation are evaluated.  Finally,
any extra features of the system are identified.

-------
                                             Table  G-l.    Computation  Comparison   for  Residential   Fuel
                                                                   Combustion  Sources  -  ESAQ
                                      Uoeu the eytttea Jo          Significance of                                       lieaaonuble Data                 Accuracy of
                                       ttui calculation?             the Lack                Chuitgec. Mectbaary            Requirements                  Calculation
A.   t-uuj Una  Input (Lave! 1.2)
    J.  Input aiatWcuimiy lucl         Yes.                                                                           *es-  lnVut in NKUS        Standard.
       con-.^.tlon in resided-                                                                                       format.
       t lal  acctor .

    2.  Ul.irlbut. fU«l 10             »«.                                                                           «"•  C"°  1"'~I           Standard.
       cou,Uy/.Ub.r«. by .urto-                                                                                      .llocatlon factor..
       gate  vttriablt: (c.a- > **•".
       populac Ion) dime tbut Ion .
    J   Kktrdci. point a our eta          Ho.  Must atatt utih      A small Inconvenience.  Uuei    Hoiierate coding effort to
                                    • fca source totals.        nutit nanually  suparate  point    process point eource duta
                                                            and areu source totals.         and retrieve fuel coa~
                                                                                         tiustlon.
                                                                                                          1-2 oa

    * .  Go tu C.

    Surru^dtu Variable Input (Level 3)
                                                                                                                   Yeb-  Uscfl  buUlj»n
                                                                                                                   ulze and fuel use
                                                                                                                   YeB'  Use" 8tailjard
                                                                                                                   factor* of Btu/Jeg-
                                                                                                                   day/houdlng unit.
    J .   Cunuute sub are

    .              for AQDH and III WAY .
                                                                                                                   sabarea in grid.

                                                                                                                   Yeli-
II . t^rumb _Aiiaiy_sl5

   « .   Input Crowtti  Uat j
                T         II         » •   u  » SIS/H.K
            """' '  "' """  " ""        "
                                                                                                                                          Standard.
                                            »      .                                                                "*"•  but •"""• "ru-
                                           COBOL rttrl«val.                                                          «"'  (ln CO'OU  """"
                                                                                                                   caCt^uocy Beparnt ely .
                                                                                                                   Couple* growth scenar-
                                                                                                                   los not easily
                                                                                                                   handled .

-------
Table C-l.  Computation Comparison for Residential Fuel

            Combustion Sources - ESAQ (Contd.)
No
Does Che sytiLeu do SlguU tcimce of
2 tutors fuel iul« »". Usca CIS/FIR lli-llne
CUUOL. recrlev.il.
III. Strategy Analysis
A. Emission l.imltb

&. fuel Controls
1. Change fuel ml* Ye*. Us«=s EIS/PiR.
2 . Change fuel Yes, Uses ElS/f&R.
C. Growth and Development Controls


Yea
Reasonable Data
Requirements
(in COBOL) each source
category separately.
Yes. but must program
each source separately.

Yea.

Yes.
Yea.

Yea, but must change
Input data set.

Accuracy At
Calculation Extra Features
Standard .
Standard.


Standard.

Standard.
Standard.

Standard.

                                                                                        H
                                                                                        U
                                                                                        H

-------
                                  Table  C-2.    Computation  Comparison  for  Commercial/Institutional
                                                      and  Industrial  Fuel  Combustion  Sources -  ESAQ
                                Uuub I lit uyal^o do
                                                       SIBnlf Uunce of
                                                          the l-ack
                                                                                                        Requirement*
                                                       Accuracy of
                                                       Calculutlim
                                                                                                                                                                Extra Ptttuur«u
                                                                                                     Yea.  Input in NEDS
                                                                                                     format.
   (e.g. , eni|iloyiki:iil ,  1 jiiil attu
   dlntrltiul Ion for cuouu/iiibt/
   Indus decIor.

)   Ex I rani  point  auuicub
4.  Co to C.
                                                    A t>mall Inconvenience.
                                                    User Buut manually
Moderate coding eftore tu
proceaa point source data
and retrieve fuel coa-
buslIon.  (Sinliar to
residential souicee.)
                                                                                                     Yea.  Can Input
                                                                                                     allocation facto
                        Li 3)
   surrogate  var lablc (« .g. ,  pouu-
   lalIon, d.u. floor area,
   land u»e) .
5.  liu 10 L.


I,  Hdp fuel  con^uupLlon
                         Yea.  Cua uae floor
                         apace or other
                         parameter.
                                                                                                     Yee.  Can Input
                                                                                                     standard factoru.
                          Yea.  Must  Input
                          portion of  each aub-
                          area  la grid.  .
                                                     Standard.

                                                     Standard.
                                                                                                                                                                                        Ln
                                                                                                                                                                                        to
    conpdtIblc  turn.
                              foe AQDH and HIUAV.
                                   Uses KIS/fUt i.L-
                                  COBOL retrieval.
                          Yen, but oust program
                          (In COBOU) each sourc
                          category separately.
                          Complex growth ace-
                          narloa not easily
                          handled.

-------
Table C-2,  Computation Comparison for Commercial/Institutional and
            Industrial Fuel Combustion Sources - ESAQ (Contd.)
No
Ltaeg the aystuB du Significance of
B 1
2. Future fuel mix Yes. Usea E1S/P&K In-line
COBOL retrieval.
B. Apply Growth taccar* Yet*. Uses eiS/?&8.
[IE. Strategy Analysis
A. Ealaslon Units
u tis/ft.k
b. Kut I Control a
1. Clunge fuel mix Ye*. Uses ClS/Ft-K.
£. GrowLl> dnd Development Planu

distribution
Yea
Ucaaonabte Data
Itequlreoeuts

(In COBOL) each source
Yea, but »u*t prograa
cacti source category
separately.
Yea.
Yea.
Yea.
Yea, but auat change
Input data aet.

Accuracy of
Calculation Extra Featurea
Standard .
Standard.
Standard.
Standard.
Standard.
Standard.
                                                                                            Ln

-------
                                      Table C-3.    Computation  Comparison  for Electric Generation
                                                       and  Internal Combustion Sources  -  ESAQ
                                   tile uyaleiB Jo
                                                   Significance of
                                                     the L.ck
Accuracy of
Calculation
t Ice i lie Ocuer.tt.lu
1.  Tieitt pttwur pliuiit* tiiulluc to
   tiiduairUl process aourcea.
                                                 Ooeu not glv*: user
                                                 litfon&at ion ou elec-
                                                                           electrical deaaad
                                                                           tate variable
                                                                                 2-3
                                                                                                                                                                 Ui
                                                                                                ts ou InduatildL

-------
Table C-4.  Computation Comparison for Industrial Process Sources - ESAQ

Calculation
i. Emission Update
A. Receive emission Inventory
input.
1. NEDS
2. Other Systems
n. Retrieve and summarize

specific data.
u. Perform Internal con-
sistency checks.
1. Point uources.


11. Cruwth Aiiulysiu
A. Input source specific growth
Information.
fucturs.


exlutlng, new. and unknown


eource* by HUI- rotate parameter
Ho
fKiea the ayaten do Slftiilf Icaoce of
the calculation? the Lack Changes Necessary



Yes
Yea. EIS/P&K
Yes. Dues full EIS/P&H

file management.
Yea. Uses EIS/P&K
checks
Yes. Set up for
AQDM.


Yea, Uses EIS/P6K
file management.
In-line COBOL retrieval.


to user who must make dls- make this die-
Could present major pio-

Input . Htunual ly . 1 «u
Yea
Reasonable Data Accuracy of
ftequiremenca Calculation Extra Features



Yea Standard
ye0 Standard
yes Standard
ves Standard

Ves Standard

Vea Standard

YuB - Standard

Yt!B Standard

must program
(In COBOL) each
separately.






-------
Table C-4.    Computation  Comparison  for  Industrial  Process Sources  -  ESAQ  (Contd.)
               Uoca chu syuica do        Significance of                                  Reasonable Uata               Accuracy of
                cite calcuUt Ion?            Che Lack              Cl^ngta Neceaudry           Requjrenenta                Calculation                   Extra Featurva



               Yea. Ua** EIS/t'tK                                                        **a                  Standard

               Yet,. Uaci* tli/^iLK                                                        Ye"' but •ual          Standard
                                                                                    pro^raai (  in                                                                 t~t
                                                                                    COBOL) each                                                                  (^
                                                                                    source caCeBory                                                               Q^

                                                                                    CoBplex concrula
                                                                                    •ay be difficult


               Ho. II^a nu ruudno        U«*er cdiinot alnuldte       Urlte nt=u code.
               to Interface with         a lanJ use control dir-                 Z-luo
               a Und uat plan.          ectly.

-------
                                     Table  C-5.    Computation  Comparison  for  Transportation   Sources  -  ESAQ
        Calculation
                                          Docs tli* uyattM Jo
                                           tlta calculation?
                                                                                                Changes Neceeaary
                                                                                                                    lieuaonable Data
                                                                                                                     Req,uir««entit
   Accuracy of
   Calculation
  1.   E«la«lou Update
       A.  Fuel Cooauu.pt Ion laput 
-------
                  Table  C-5.    Computation  Comparison  for  Transportation  Sources  -  ESAQ   (Contd.)
                                 Uotiu elm syutctD do
                                  llie calculation?
                           Significance! of
                              the Lack
                                                                                    Changes Neceaaary
                                                                                                             Reasonable Data
                                                                                                              Requirements
Accuracy oi
Calculation
D.   Input gen
    projectlo
Nu.  Hunt Input entirely   Inconvenience to usur utto    Modification of code to
iifcw  data set.             must develop VHT due to      allow growth factor to be
                        growth externally.          applied to base data set.
                        Current system uses TRIMS                 2-3 UUB
                        model to do thin.
A. Apply ualuuluo llulta.
D. Apply tiattlc controls.
C.  Apply growth and d«vtslu|)iucni
    controls.
                                                                             new data  set.

                                                                             Yea, but  must
                                                                             new data  set.
                                                                                                                                      Standard.

                                                                                                                                      Standard.

                                                                                                                                      Standard.
A.  Input vehicle activity.
fi.  Apply
                                                                             Ifeu,  Hust input
                                                                             activity In NEDS
                                                                             format.
                                                                                                                                                                                                     Ui
                                                                                                                                                                                                     00
A.  Inpu
                Input
*.   Input


B.   Apply
 A.  lupui per capita g
    consumption  rate.
 C.  A,>ply
                                                        Inconvenience to user.
                                                        InconvenIcnce to uaer.
                                                                             Yea, but  bouie careful
                                                                             manipulation of
                                                                             EIS/P&R required.
                                                                                                            Yea.  Input county
                                                                                                            data  In NEDS format.
                                                  Minor modification since
                                                  population  Is already treated.
                                                                     1 mm

                                                  Same us above.
                                                                                                        Standard.


                                                                                                        Standard.

-------
                                                  Table  C-6.    Computation  Comparison  for  Solid  Waste
                                                                        Disposal   Sources  -  ESAQ
                                        Does lite syateiB do          Significance of
        Calculation                       the calculation?              the Lack                Changes Necessary


I,  Enlasign Update
   A.  Surrogate Variable Input {Level 1,2)

       1.  Input surrogate variable to be   No.  There la no pro-    Significant.  Dues oat allow  Moderate nod Ificatlons
          uaed.                         vUioo for Inputting     user to oak* use of different  and new  coding.

                                        solid waste disposal.
       2.  Input aolid waste generation
          factors.

       3,  Input disposal technique
          distribution.

       4.  Compute solid  waste geuerated
          and disposal technique In
          aubareaa.

       5.  Extract point  sources.

       ft.  Go to C.

   B.  Solid Waste Data Input (Level  3)

       I.  Input solid waste generation     Yes.
          and disposal data from local
          sources.

       2.  Extract point  sources.          No.  Must start with      Inconvenience to user who     Moderate coding  effort
                                        area source totals.       must separate point and       to process point source
                                                                area source totals           data and retrieve solid
                                                                manually.                    waste disposed of.
                                                                                                        1-2 uu».

       3.  Co to C.
                                                                               Reasonable Ddt*
                                                                                                              Accuracy of
                                                                                                              Calculation
                                                                                                                                             Extra Features
       1.  Hap solid waste generation
          and disposal technique to
          master grids.

       2.  Apply emission factors.
   tiroutti Analysis

   A-   Input Growth Data

       1.  Surrogate variable  projec-
          tluna.
                                        Yes.  Currently  set up
                                        for AQDH.
Ho.  Same as I.A.I above.
                                                                             Yes.   inputs data in
                                                                             HEDS  forut.
                                                                             Yes.  Allocation dune
                                                                             along with other
                                                                             variables.

                                                                             Yea.

                                                                             Yes.
                                                                                                       Standard,
Standard.

Standard.
                                                                     Ui
                                                                     VO

-------
                                            Table  C-6.    Computation Comparison  for  Solid Waste
                                                                Disposal  Sources  -  ESAQ  (Contd.)
                                 Dot:a clid ayutua Jo         Significance of                                    Keasonable Data                Accuracy of
    Calculation                     th«i calculationt            tti« Lack               Changub Necesuary           Requirements                 Calculation                   Extra Features
                 gt>ni:iat ion      Ho.  Saute ati
3.  Accept
       projections.
       ULupoaal t«ctml4uea.        Y*=s.                                                                    V*a.  Input in MEDS        Standard.
                                                                                                      format.
                                                                                                       *«" •  <"!>"* In NKOS        Standard.                                                     
-------
Table C-7.  Computation Comparison for Miscellaneous Sources - ESAQ
No *es
Does the ayatuui do Significance of Reasonable Data
Calculation the calculation? the Lack Changes Htctaaary Requirements
Solvent Evaporation
I. fcaisbioo Update
A Direct Data Input
I. Input actual solvent use. Yes. **"• Standard NEDS
format.
6. Surrogate Data
Input
factors. Inability to manipulate code since basic surro-
Input .
C . Eiiiaa lull Cowputat Ion and Happing .
1. Kap solvent „«, to Ma«, »«. »«• U"'"8 »tDS """
rids to start troa.
2, Apply emission factoru. Yes. *es>
compatible torn. up for AQUH.
11. Growth Analysis
A. Apply growth factor*,. Ye*. Using CIS/mi. Ye9-


late EIS/P&R data
carefully.
1-2 uai.

Accuracy of
Calculation Extra Feature



Standard



Standard
Standard
Standard

Standard


Standard


-------
Table C-7.  Computation Comparison for Miscellaneous Sources - ESAQ  (Contd.)
No Yet*
Uoeit che uyt>t.iia da Significance of tteauonable Data

flies
standard NEDS
Information.
standard allocation
parameter or Input

Fugitive Dual
C d d
BEDS Informatlun.
11. Input allocution parameter. Yes. Yen. Can use standard
allocation parameter
or Input overriding
parameter .
IV. Apply tontcol «tr«l«8y- *". Yea- but BUB(: cl""»ae
input data set .
Other Souri-gg
ulate EIS/F&R
carefully.
II. talaal.jn Input. Ytt.. Yea, but muaL nanlp-
ulate eiS/P&R
carefully.

Accuracy of







Standard
Standard
Standard
Standard
                                                                                                    K)

-------
                                                        Table  C-8.    Computation  Comparison  for  Gridding  -  ESAQ
                                        Uous the syuteat do         Significance of                                       Reasonable Data                 Accuracy of
         Calculation                       the calculation!             the Lack               Changes Neceaaary             Requirements                   Calculation                     Emtra Features


A.  Hap frou aubarea to ataater grid using    Yeb.                                                                         Yea.                             Standard
   previously determined fractions.
                                       No.  Can uuc only one      Inconvenience to user.       Some reprugrumIng  to keep
                                       grid ttyuten.              Limits the different        accurate bookkeeping of                                                                                                  L_A
                                                               diitd ftlea that can bu       various Bubareas to gelds.                                                                                                Q^
                                                               used.                                      1-2 asm,                                                                                               t.\
C.  Hap procuaa activity Inutend of         Yea.
   umlssluua.

b.  nap into chunglng aastut grid.          No.   Uue one iBddter       Incuiivunlunce to user.      Some reprogrumulng  us
                                       grid  only.                                        above.

-------
                                                    Table  C-9.    Computation  Comparison  for  Growth  -  ESAQ
                                       Urn calculation?
                                                              Significance of
                                                                 the Luck
                                                                                             Neceusary
                                                  Reasonable Quta
                                                  ftequirenenta
Accuracy of
Calculation
II.    UuliMMilite gruwLll I roai
      gcuci-illzeJ growth
      facto™.


III.   Lluk growth btitwuen act
      IvltU-b

      A.   I'rowlJe llukaiicS
                                      Yeti. Uueu tIS/
                                                                                      HoJlfIcaclun of
                                                                                      code
                                                     , but soust
                                                     ut new
                                                     a aet.
                                                     . but aunt
                                                       » (In
                                                  COBOL) tot cac
                                                  source cdtegor
Does not  supply user
with Interpretive


Inconvenience to user
          cto pe
                                                                                      Hodiflotion of
                                                                                      output formata.
                                                                                      Minor uoJ-
                                                                                      1 f Icut |on of code.
                                                                                       utput fora j i a   „
                                                                                         '          l-2mi

-------
                                       Table C-10.   Computation Comparison  for  Control  Strategies  -  ESAQ

Does the eyateu do
1, Separate control »crategy routine. No. Uses EIS/P&R
t<
Significance of
the Lack
nmst program (In COBOL)
the particular control
to Ye«
Reasonable Data Accuracy of
ChungoH Necessary Requirements Calculation
common strategies.
3-5 «•

Extra F«atur«a

II.  Pro
           per computer ru».

III. Apply reguUtlo»B only to ttffecced
    aourcea

 IV. Provide euuunry tabled tor rejju-
    latlon evaluation.
                                                       Requites user to run each   Saall change to code.
                                                       control strategy separately.               1 •"
                                                       t>«eB not ptovldu user with  Moderate output  changes.
                                                                                            -
                                                                                                              «M/P«.
                                                                                                                            Huch «n convenient and l.sa
                                                                                                                            prone to error than input lag
                                                                                                                            entirely new data net.
                                                       Interpretive Information.

-------
                                         Table  C-ll.   Evaluation  of  Computer Requirements - ESAQ
           Requirement
      Does the system meet
       the requirement?
                            Significance of
                               the Lack
                                                                                                No
                                                                                                      Changes Necessary
                                                                                                                                   Extra Features
  1 .   Computer  System

      A.   UN1VAC  1110


      B.   IBM


 II .   Programming Language

      A.   FORTRAN and/or COBOL

      B.   ANSI  standard

III.   Mode of Operation

      A.   Batch and Interactive.

      B.   Interactive only.


 IV.   Program Structure

      A.   Modular

      B.   Complete or single [nodule
          run capability


  V.   Off-Line  Storage

      A.   Permanent - tape, cards.

      B.   Transient - tape, disk,
          data  cell, drum.

 VI.   Input Format

      A.   NEDS  compatible.

      B.   EIS/P&R compatible.

      C.   Census tapes
No.  Has never been run  on
other than ES System.

Yes.  Was developed on IBM
OS System.
Yes.
Yes.



Yes.
No.  There Is no Interactive
component.



Yes.
Yes, but package Is  not
likely to be run straight
through.



Yes.

Yes.
Yes.
Yes.
Built around  EIS/P&R.
No.  Processes  census  infor-
mation input  on cards.
                        Cannot be run on EPA
                        facility.
Must be converted to  UNIVAC
form.                 ,
                     1-2 mm
                                                                                                                 
-------
                                  Table  C-ll.   Evaluation  of  Computer  Requirements  *  ESAQ  (Contd.)
           Requirement
      Does  the  system meet
       the  requirement?
                                                                         Significance of
                                                                            the Lack
                                                                                               No
                                Changes Necessary
                                                                                                                                  Extra Features
VII.  Output  Format

      A.  Models

          1.  AQDM

          2.  COM


          3.  IPP


          4.  VALLEY


      B.  Isopleth programs
      C.   Hard copy by area
          or  subarea

VIII. Documentation

      A.   User's guide

      B.   Programmer's manual

 IX.  Portability

      A.   Easily transferable


      B.   Transferred by cards,
          tape  (binary or source
          form, batch process).

  X..  Compatibility

      A.   AEROS
Yes.

No.


No.


No.


Yes.  Isopleths  used in air
quality packages.  May need
some generalization.

Yes.
There is no documentation
available for general use.
Uncertain.  Has  not been
used outside ES.

Yes.
No.  Does not  have proper
documentation.
Currently not  possible
for anyone outside of ES
to use system.
                            Minor modification.
                                            <  1 mm.
                            Minor modification.
                                            <  1 mm.
                            Minor modification.
                                            <  1 mm.
Significant  effort to
document work.
               3-4 mn
Cannot  be  supported by
AEROS systems.
                                                                                                  Documentation must be
                                                                                                  prepared.
                                                                                                                4-6 mm.

-------
168

-------
                                    169
                                 APPENDIX D
                   Detailed Evaluation of the MWCOG System
       The tables included in this Appendix compare the MWCOG system with
the CEPA requirements described in Sections 2 and 3.  The evaluation is based
on whether the system will do the required calculation.  If it does not, the
significance of the lack of this capability is given along with  the changes
that would  be  necessary to enable the system to perform as desired.  An esti-
mate of the effort, in man-months  (ma.) t of making the modification is also
given.   If the system does the required calculation, the reasonability of  the
data requirements and the accuracy of the calculation are evaluated.  Finally,
any extra features of the system are identified.

-------
                                               Table  D-l.    Computation  Comparison  for  Residential   Fuel
                                                                     Combustion  Sources  - MWCOG
                                      l>oes the ayytem do          Significance of                                        Reasonable Data                  Accuracy of
                                       tlie calculation!              the Lack                Changes Neceauary             Requirements                    Calculation
A.   Fuul UMO Input (Level  •.*)

    i .  Input atata/cuunty luul       Yuu, bu

    i.  uUtrlbul* luel 10           Nu.  Hunt input Juta      Inconvenience to user seeking   Hlnor coding change since
       Lounty/aubaiea by aurro-      already disaggregated     to do a simplified analysis.    all necessary Information
       gate variable (e.g..          to subarcas.                                           is there.

       Jlwtr ibut ton.
    J.  Eitruct  point nuuicen         No.  Husc =,lart with      A snail Inconvenience.  User    Moderate coding effort
                                   urea source totdlu.       nust manually separate point    in addition to new code
                                                           and area  source totals.        required to handle point
                                                                                        sources (see Industrial
                                                                                        Process Sources).
                                                                                                        1-2 ua.

    4,  Co to C.

tt.   Suiro&ate Variable Im.ut  (Level j)

    1.  input state, county, subtree  Ye,.                                                                             J«-  BotU.Ceo.-.
       »utcoftatc variable (e.g.,                                                                                     tnforaatlon  la Input.
       population, d.u., I lour area,
       land uc.e).
    2.  Input fuel consumption       No,  Baseline fuel  con-    Us,er is confined  to one baue-   Moderate coding effort.
       factom.                    sumption by iubart^ Is     line daca set.  New updated    All the basic Information.
                                  an input data :»«-•!.         data for  baseline Bust be      except the fuel consumption
                                                           computed  manually and Input.    factors, are available.
                                                                                                        1-2 a*

    3.  Compute  »ut«rca fuel use.     Ho.                      Sd»e us above.

    4.  Extract  point sutiices.        No.                      Same as 1.A.3 above.

    >.  Go to C.
       coapdtibl? fora.
    1   hup fuel .-onbuB.pt ion to       VCM  but applies ^l.slon                                                           *«•  Cao -peclfy                Leaii desirable than
       ^tergvid,.                facers (.tep 2) fir.t.                                                             alloca. loa on the                «tivltyflr.t.  Could Uad
                                                                                                                    basis of area,  popu-             to unusual  results.
                                                                                                                    Utlon, eaployoent


    2.  Apply .ai^Jon factors        Ye.;.  Ooes  this before                                                             *es.                            Standard.
                                   step I.

-------
                                          Table  D-l,    Computation  Comparison  for  Residential  Fuel
                                                              Combustion Sources  - MWCOG  (Contd.)



II. Crowtli Analysis
A. Input Growth Data

2. Future fuel •!*
fl. Apply Growth Factors
Nu
Does thu eyatoB Jo Significance of




values.
Yes.
Yea, UKOUTH routine
Yea
Reasonable Data Accuracy of
H««iry Requirements Calculation Extra features



planning data.
Yea. Standard.
Yes. Standard.
. Strategy Analysis

 A. Emission Limits

    1. Change eniatilan factors
    fuel Coot ro la

    1. Change fuel aix

    <.. Change fuel
       characterlstlcs.
 C.  Growth anJ
    1.  Change surrogate variable
       distribution.
                               Can override thc^e with
                               input to GBOU.
Yea, but inust
data carefully aa
ealaaion factor change
ia interpreted aa a
change in th* effec-
tive growth rate.
Yea.

Yea, buc tlie change
la  Interpreted aa an
effective growth
change aa above.
Yea.  £4ally done
alnce growth scenario
ia a direct input.
                                                                                                                            Hot aa arcurat* a procedure aa
                                                                                                                            could b« done.  Ptone to aoae
                                                                                                                            clerical errora.
Standard.

Sana problcsi aa III.A.I above.

-------
                                Table  D-2.    Computation  Comparison  for  Commercial/Institutional
                                                    and  Industrial  Fuel  Combustion  - MWCOG
                                                       Significance of
                                                         the Lack
                                                                              Chaniteu He
                                                                                                                                     Accuracy of
                                                                                                                                     Calculation
                                                                                                                                                              Extru Feature*
A.  fruet U»r  Input (Level 1. 2)

   1.   Input uLaie/cu

       •ector

   2.   UUlilbnta fue

       (e.g.. eiapltiyn
               ju/lnui/lnJu
J.  tkl i.icl |>ulltt
                  (e.g.,
                               No.  Hubt ulart wltli     A aaial 1 Incoikvenlence.        Moderate coding tiftort In
                               area source totals.      Uaer must manually separate    odd it ion to new code re-

                                                                             dources.  Similar CO resi-
                                                                             dential sources.
                                                                                                     Infortuat Ion la
                                                                                                     Input.
                                                                                                                                                                                      N>
2.  lm»ut futl cona>iui()[ lull
   t del or a.
                                  Nu.  Buicllni; tuel      User la cuntlned 10 one      Moderate coding effort.
                                  conbunptluu by ±>tib-     baseline data utt.  New up-   All tne Ladle  infortuatlun,
                                  area  La an Input daia    dait4 data for baseline      except tlie fuel cunsutup-
                                  t>et.                  must be computed manually     lion factors,  are auail-
                                                       and Input.                 able.  Same aa residen-
                                                                                tial sources.

                                  No.                  Same a±> above.

                                  Nu.                  Sdme at. i.rt. 3 abuve.

-------
                                          Table  D-2    Computation  Comparison  for  Commercial/Institutional
                                                             and  Industrial  Fuel  Combustion  -  MWCOG   (Contd.)
          Calcutatlun
                                        Does the ay at eta do         Significance of
                                         the calculation?             the Lack
                                                                                          Chaiigeti Necessary
                                                                           Reasonable Data
                                                                            Requirements
                               Accuracy of
                               Calculation
                                                                                                                                                                              Extra Feature*
 I.  Emission Uuoate (Cont'd)

    C.  Emission Computation and Happing

        1.  Hap fuel consumption to
           mauler gelds.
        ta
       2.  futuit fuel mix

    B.  Ap^iy Cruwtli b'accuca
Yea, bul applied tantd-
alon factors (step 2)
first.
                                        seep 1.
Yea.  Can nodlty foe u^
ultli  stwtral models.
                                          a.  Can Input actual
                                         cd.  CKOUTtl ruutlnu cal-
                                         ulates growth fuctorti.
                                         luput to CROW,
Yen.  Can aptcliy
allocation on Cite
baala of «rc2D pop-
ulacIon, «up Joyment
or  other parameter.

Yea.
                                                                                                                  Yea.  Operatea trun
                                                                                                                  planning data.
Less desirable than
•applng activity first.
Could lead  to unusual
results.
                                                                                                         Standard.


                                                                                                         Standard.

                                                                                                         Standard.
                                                                                                                  Yifti, but iHuut apply
                                                                                                                  data carefully at>
                                                                                                                  emission factor
                                                                                                                  cliange i» Inturpiated
                                                                                                                  an a change In the
                                                                                                                  effective growth rate.
                                                                                                         Not as accurate a pro
                                                                                                         cedure as should be
                                                                                                         done.  Prone to borne
                                                                                                         clerical errors.
                                                                                                                     t but (lit change
                                                                                                                     Interpreted as
                                                                                                         Sane problem as Hle
                                                                                                         A. 1 above.
    C. Gruuj,U._aud_D«rVulupi>i.;ni Plans.

       1.  Change surcu^ain variable
           distribution
                                                                            s.  Easily dune
                                                                                                                  scenario Is d direct
                                                                                                                  Input .

-------
      Table  D-3.    Computation Comparison  for Electric Generation and Internal Combustion  Sources - MWCOG
Docs the uyutwu Jo
 th* calcuUi io»?
Significance of
  the Lack
                                                                         Changes Necessary
                                             Keu sun able Uatct
                                              Requirements
                                                                                                                       Accuracy of
                                                                                                                       Calculation
                                                                                                                                                Extra Features
DotB oot «lve user
Information on electrical
requirements.
                                                                        Modification Co code to
                                                                        dettralnt nloctrlcal
                                                                        demanJ from surrogate
                                                                        variable (e.g. population).
Internal Cuabuat ion

I.  Treat Hl«lldr la

-------
                              Table  D-4.    Computation  Comparison  for  Industrial  Process  Sources  - MWCOG
                                          Doen the t>yuiutt do
                                           Che calculation?
                                                                Significance of
                                                                   the Lack
                                                                                                                    Reasonable Data
                                                                                           Changes Necet>»ary
Accuracy of
Calculation
                                                                                                                                                                              Extra Features
I.  Eatsalon
      Input.

      1.  NKDS
      2.  Other systems

   a.  Retrieve and suBBuciie
      inventory data.
   C.  HudIfy inventory with source
      specific data.
     D.
                            enty
        checks.

     E,  Generate output In modeI-
        compatible fora.

        1.  Point KouEces

        2.  Area sources
 II.  Ctowtti Analyais
     A.  Input auutcc bpeclflc growth
     D.  Apply
        factor
     0.  Allocate growth at unknown i
III.
     A.  Apply toilaaton limits.
     b.  Apply growth and development
        controls.
     C.  Apply land uue controls.
                                          No.
                                                               The HUCOC aybtta wua nuc dtsigned  Co do any cowyutatIons
                                                               with point sources.  The point source Information Is all
                                                               bandied manually and la uGfcd only  aa Input Into the
                                                               dispersion nodelu.                          8-10 mn

-------
                                 Table  D-5.    Computation  Comparison  for Transportation  Sources  -  MWCOG
                                                                                                                                       Accuracy ol
                                                                                                                                       Calculation
A.  fuel CoiifauBujL loti  Input  (Luyyj 1) Ho.  Syuten Udca a traiiu-  User cannot do simplified    Hud
.'.  Eat
1   l> U
       («.
   ...  Co
              VHf.
             Ld VHT to
   2.  Apply (y»l;
                                                                              all ntctuuary d
-------
                                                 Table  D-5.    Computation  Comparison  for  Transportation
                                                                      Sources  -  MWCOG  (Contd.)

Does the system ut

      A.   Input i±ulaaion!t directly

Cddoline Handling Evaporation Loaaea

   I.  Gaaolitte Harketed Input

      &.   Input gaaol Ine aolj .



      B.   Apply ealsaion factors.

  11.  Surrp^atj! Vdrlflblg InpuC
Ha.  Hust laput data  in    Uscx muat uanually calcu-
atfca  iiouEc« t^iuel
Inventory farnat.
late the emission rate for
Input.

Sane as above.
                                                          Qi» at  input and
                                                 addition of calculation
                                                 routines-        . _ „
                                                                            Yes. Input aa p&jC of
                                                                            area source inventory
                                                                                                          Leaa accurate than !•
                                                                                                          poaalble when doing, growth
                                                                                                          analysis.
Ho.  Muat Input data in    Uuer uust uanuully calcu-
area source aiiuel
inventory format.
      A.  Input  pet capUa gaaoLU
         conauiapt Ion tatt,

      B.  Commute gasoline udrkeL&

      C.  A^fly  eialntiiun toctors.
                        late the euiusion rate for
                        Input.

                        Same sa above.

                        User nuut make the Burro-
                        gate variable calcula-
                        tloaa externally.
                          Modification of input and
                          addition of calculation
                          routines.        ,  _ __
                                                 Hew codiiuj effort..

-------
                     Table  D-6.    Computation  Comparison  for  Solid  Waste  Disposal  Sources  -  MWCOG
                                                   Significance of                                  Hcttsoudblt Uuca               Accuracy of
                                                      the Luck              Uiangca Ntcct.bjcy           ftttqulreaeni a                 Calculation
                             tiuErogdti: vdrljblct* fur  f cfeiit eat liuiil Ing procedure.
                             build wusie Jl ^pOaal .
2.  Input «olU w.ttte ttim:i,
   and disposal technique In

3,  L.MH4CE point »ou(Ci:t>.
b.  Co Lo C.                                                                                                                                                                    |—I

"'""""Tu^T'.!!"':^     *.                  u ....  i. co.,,      Him «                                                                                            *
                             slon inventory furnat.                                         1-2 mm
         poliu bouicca.         Ho.  Huit  sCatl with    Inconvenience to uacr who      Hew coding effott la con-
                             **rea source totals.     must separate point and a ecu    ncccion with addlttoual

                                                                         (See Industrial process
                                                                                                Yea.  Can
                                                                                                spec It y *1 lo-
                                                                                                cation para-
;.  At)i>ly oBiaalou lacLois         No.  Map*  liipui ianj*-   Hone.  1 !«: owpping ul activity can be done ulicn


J.  Belief Jlc output In wuJtl-       VeS
   vonfMliklt: for*.

-------
                     Table  D-6.    Computation Comparison  for  Solid  Waste  Disposal  Sources   -  MWCOG    (Contd.)
A.  Injiut (jroMjh Data
                                  Doe- th« flyaten do         Significance of                                     Hcu^oiiabU Oat-                Accuracy of
                                   thi calculation?             the Lack               Cluinget. Necessary           Bettulreaeiita                  Calculation                   Extra Features
             : variable projtc-       No.  Same au l.A.
                                  above.

   2.  So I iii waste (ientiratlon          No.  Sane aa abuvt
      tatta.

   3.  Accept loc^l t.oliJ wuatt
      project lona.
                                  No,  Sane as abuvu.
A.  tmitii,ioii limit a.                  Yea.                                                                     Vca. but uuett sljnu-            Not *u accurate
                                                                                                          late as an effec-             Hracedure.
                                                                                                          ilv« grow tit rate.

b.  Cruutli JiiJ duvulujjiBcnt             No,  Utfcaube of lack    U^er caunoc easily aliuu-
   coiiitol^.                       u[ ducruttaLe variable.   late the effect of this
                                                       User oiuut manually Jultr-       Ncu coding ellort {•aw
                                                       oilde tlic effect of this        as l.U.l)

-------
                                      Table  D-7.    Computation   Comparison   for  Miscellaneous  Sources  - MWCOG
                                        l*oL-U the ttyuteu Jo          Significance of                                        Keadunable Data                  Accuracy of
                                         the calculation?             the Lack                 Changes Necebsaty             Requirements                    Calculation
    A.   t)Utu;L Uattt_ Injiut
                                                               User must manually compute    Modification to Input  and
                                                               emissions.                  neu code.
                                                                                                         1 mm.



        i.  Input  solvent consumption    No.  No  provision  tor      Inconvenience to ut,er.        Moderate modification  of
           fueionj.                    handling a surrogate       Inability to manipulate       code la connection ulth
                                      variable for this          this type of data.           I.A.I above.
                                      source.
    C.  Emission Computation  and Happing
        I.  Hap solvent use to master    Yea, but maps eaUalons.                                                         Yes.  Can specify               »•"• desirable thaa i
           urM.                                                                                                      allocation parameter.           activity first.
                                      No.  Haps Input           None.  The mapping of  activity can be done when
                                      illusions.                the computation procedure (I.A and B above) Is
                                                               changed.
                                                               User tnuut manually deter-     Hew coding effort (same
                                                               mine the effect of chls on    aa 1.A and I.B above).
         routh and development           No.                      User cannot  easily simulate   New coding effort.
         anliula.                                               the effect of this type of                    1-2 mm.
        J.  Generate output  in tuoJel-    Yet*.                                                                           Y*8.                           ataiiuoiu.
           compatible form.

11 .  lifowth Ajialygla
    A.  Apply groutl. factors.            Ye:,.                                                                           Yeb.  Can uae  either            Not fln accurate proce.
                                                                                                                     population or  a non-            *Wl* Bt°wth rate to  .
                                                                                                                     demographic growth rate.         directly.
                                                                                                                      \
                                                                                                                      ea  but mubt  aluuiatt:          H°t *« accurate procedure.


-------
                           Table  D-7.    Computation  Comparison  for  Miscellaneous  Sources  -  MWCOG  (Contd.)
                                                                                                                                     _*£!_
                                       Does the system do         Significance of                                       Reasonable Data                 Accuracy of
         Calculation                      the calculation?             the Luck                Changes Necessary            Requirements                  Calculation                     Extra Features

 Ptrea
  1.   Input basic activity factor.        No.  Uaer must  Input      Uaer must manually compute    New coding.
                                     emissions.               emi&alona.                                  1 ma.

 II.   Input allocation parameter.         Yes.                                                                        Yea.  Can use alloca-             Standard.
                                                                                                                 tton routine in GROW.

 III.   Apply emission factor.             No.                     Same aa I above.

 fugitive Dust
  I.   Input basic activity factor.        No.  User must  Input      User must manually compute    New coding.


 II.   Input allocation parameter.         Yes.                                                                        Ve».  Can use alloca-             Standard.
                                                                                                                 tlon routine in GROW.

 111.   Apply emission factor.             No.                     Same uu I above.

 IV.   Apply control strategy.             No.                     User must manually compute    New coding.
                                                             the effect of strategy on
                                                             emissions.

Other Sources
 I.  Generalized format.                 No.  User must Input      User uuet manually compute   New coding.
                                     ealshlons.               emissions.

II.  Emission Input.                    Yea.
                                                                                                                                                                                                        00

-------
                                           Table  D-8.    Computation  Comparison  for  Gridding -  MWCOG
                                    Owen clie uyaiciu da         Significance of                                    Keaionublt Data                Accuracy ol
                                    the calculation!             tlie Lack               Changes Nuceasaiy            Requltkitentti                  Calculation
t>ievloutily Jet c i ml MC J ft act loitu,
                                                   i lii   Inconvenience to user.       Some icprograiuiilng to keep
                                                        Llnitu chfc different        dccucace Lookkee^lnB of
                                                        data files that can be       various aubareaa to ucldu
                                                        uaed.                                  I - 2 «...
tt-ip Into ctiunBJiiH (ue>iui HTiJ.           No. Uttes o
                                   grid only.

-------
                            Table D-9.  Computation Comparison  for Growth - MWCOG


U OH
1. DfcLtunaini: gcowih tfom (.pcciflc
data.
grouth fattocd
A, Provide linlLagetf

consistency checks.
Etcenario pet run.

Doea the aytitea do Slgnlflcani:e of fteattonublfi D»c«

Kea. Done in GKUUTH Yea.
routine.
factors
Yes. Can compute Yen.
grouch OD lite basis,
of cuntbiodt tons uf
parameters.
Yea Yes
Yea YKS
Yes. Yes

fea
Accuracy of
Calculation
Standard.

Standard


Standard.


Extra Features



00
u>


unJ activity

-------
                                       Table  D-10.    Computation  Comparison  for  Control  Strategies  -  MWCOG
                                                               Significance ul                                     Kedttoiiuli U Data                 Accuracy of
                                                                  the Lack               Clian*eb Necessary            ftequirementb                   Calculation                    Ext CM Feature
                                                                    Intel [tret control     Nuw code .
                                                                   ifianually or by
 II.  f n»Ltii mo i u Llidii  uuc coin col           No.                   KtquirebUicrto run «acli       LUaoge to code In CUD-                                                                                              J—*
     sir-tvt) yer toiu^utei mil.                                  control  strategy separately.    nectlun ulth I above.                                                                                               Q£>


111.  AJJJ.IX icyuldllona  only lo ailecLed       No.                   Ubcr must i>t>Ler uitt.      Output clidns* i" cuuntc-
                                                            luc«r(>rel lue Infuraation.       t ion utcti I above.
                                                                                                         1 ODD

-------
                                         Table  D-ll.   Evaluation  of  Computer Requirements  -  MWCOG
           Requirement
                                              Does Che system do
                                               the calculation?
                                                                                                 No
                             Significance of
                                the  Lack
                                 Changes Necessary
                                                                                                                                    Extra Features
  I.  Computer System
     A.  UN1VAC 1110
      B.   IBM
No.  lias been run on
IBM 370/168.
                                              Yes.
Cannoc be run on EPA
facility.
Must be converted to
UNIVAC form.
                 1 mm
 II.   Programming Language
      A.  FORTRAN and/or COBOL
      B.  ANSI standard

III.   Mode of Operation

      A.  Batch and interactive

      B.  Interactive
 IV.   Program Structure
      A.  Modular

      B.  Complete or single module
         run capability
  V.  Off-Line Storage
     A.  Permanent - tape,  cards.

     B.  Transient - tape,  disk,
         data cell, drum.

 VI.  Input Format
     A.  NEDS compatible.
     B.  EIS/P&R compatible
     C.  Census tapes
Yea.  FORTRAN only.
Yes.
Yes.
No.  The interactive
component has been
eliminated.
Yes.

Yes, but would require
appropriate JCL to run
straight through;  not
likely to be used  in
this manner.
Yes.

Yes.
No,  Receives point
source information
in state-supplied
format.
No.
No.  Processes aggre-
gated Census
information.
                                                                      User cannot use NEDS data
                                                                      directly.
User cannot use EIS/P&R
system.

Inconvenience to user.  Must
create an aggregated  census
tape first.
                                Modification of input.
                                                 1-2 mo
Major new coding  effort.
(See industrial process
sources).
Minor modification.
                 1 mm
                                                                                                                  £
                                                                                                                  in

-------
                               Table  D-ll.   Evaluation of Computer Requirements  - MWCOG  (Contd.)
                                                                                                 No
             Requirement
                                               Does Che system do
                                                Che calculation?
                            Significance of
                               Che Lack
                                                        Changes Necessary
                                                             Extra Features
 VII.   Output Format

       A.  Models

          1.  AQUM

          2.  CUM

          3.  IFF


          4.  VALLEY


       B.  Isopleth programs
       C.   Hard copy by area
           or subarea

VIII.   Documentation

       A.   User's guide

       B.   Programmer's manual


  IX.   Portability

       A.   Easily transferable
       B.   Transferred by cards,
           tape  (binary or source
           form, batch process).

   X.   Compatablllty

       A.  AEROS
No.

Yes.

No.


No.


Yes.  SYMAP isopleths
used In air quality
packages.   May  need
some generalization.

Yes.  Special rou-
tine (EMSUM).

There is no documen-
tation available  for
general use although
the programs have ex-
tensive comments.
Yes.  Select programs
have been used  else-
where.
Yes.  Cards.
                                               No.  Does not have
                                               proper documentation.
A user must  interpret the
programs himself.  Not
especially difficult since
the codes are  short and
straightforward.
                        Cannot be supported  by
                        AEROS system.
                                Minor modification
                                                < 1

                                Minor modification
                                                < 1
                                Minor modification
                                                < 1
Prepare documentation

                 2-4 nun
                                Documentation must  be
                                prepared.         ,
                                                 4-6 o
                                                                                                                                                            co

-------
                                      187
                                 APPENDIX E
                 Development Effort of a New CEPA System
       The table contained in this Appendix gives an estimate of the effort
required to develop an entirely new CEPA system.  These estimates are given
for each task involved in an air quality analysis.  They are consistent with
the estimates of modifications to the existing systems in that the effort re-
quired to make a major modification is assumed to be equivalent to developing
that component of a CEPA system anew.

-------
Table E-l.  Development Effort of New CEPA System
Source Categoiy
He tj iilon I i a 1 Fuel Comlmsl ion
1. Lmlsslon Update
A. Fuel Use Input
U. Surrogate Variable Input
C. Eml:>tiiun Computation and
Mapping
11. Growth Analysis
A. Input Growth Data
111. Strategy Analysis
A. Emission Limits
b. Fuel Controls
C. Growth and Development
Conimerc tal / liibl 1 tut ional and Industrial
Fuel Combustion
1. Emission Update
A. Fuel Use Input
B. Surrogate Variable Input
C. Emission Computation and
Mapping
11. Growth Analysib
A. Input Growth Data
B. Apply Growth Factors
111. Strategy Analysis
A. Emission Limits
b. Surrogate Variable Input
Mapping
Electric Generation
1. Ireat Power Plants
Internal Combustion
Industrial Process Sources
[. Emission Update
11. Growth Analysis
111. Strategy Analysis
Effort Required Total Effort For
to Program Source Category
(man-months) (man-months)
2-4
3-4
1-2
1-2
2-4
2-3
1-2
1-2
13-23
1-2
1-2
(Same as Residential)
1-2
1-2
1-2
(Same as Residential)
1-2
6-12
(See Industrial Process Sources)
2-3
2-3
(See Industrial Process Sources)
3-4
2-3
2-3
Source Category
Highway Vehicles
I. Emission Update
A. Fuel Consumption Input
B. Specific Data Input
C. Emission Computation and
Happing
11. Growth Analysis
111. Strategy Analysis
Other Vehicles
Gasoline Handling Evaporation Losses
Solid Uaste Disposal
I. Emission Update
A. Surrogate Variable Input
B. Solid Waste Data Input
C. Emission Computation and
Mapping
II. Growth Analysis
III. Strategy Analysis
Miscellaneous Sources
Solvent Evaporation
Fires
Fugitive Dust
Other Sources
Crldding
Growth
Control Strategies
Computer
Census Tapes
Documentat ion
Users Guide
Programmers Guide
AEROS Requirements
TOTAL EFFORT
Effort Required
to program
(man-months)
1-2
4-6
2-3
2-4
3-5
3-4
2-1
1-2
2-4
1-2
2-4
1-2
2-5
1-2
2-4
1-2
5-10
7-11
6-10
1-4
2-1
2-3
4-6

Total Effort For
Source Category
(man-months)
12-20
3-4
2-3
7-14
6-13
5-10
7-11
6-10
11-16
87-149
                                                                                     CO
                                                                                     CO

-------
                                     189
                                ACKNOWLEDGMENTS

       The authors wish to adknowledge the cooperation of the following  people
without whose help the evaluation of the computer systems would not have been
possible:
       Lloyd Hedgepeth   - EPA/Monitoring and Data Analysis Division
       Thomas McCurdy    - EPA/Land Use Planning Office
       Jerome Mersch     - EPA/Monitoring and Data Analysis Division
       Vernon A. Krause  - Metropolitan Washington Council of Governments
       Michael Lukey     - Engineering-Science
       Terry LiPuma      - Engineering-Science
       Thanks are also due to the following EPA staff who provided guidance on
the project objectives:
       John Bosch        - EPA/Monitoring and Data Analysis Division
       Martha Burke      - EPA/Office of Transportation and Land Use Planning
       Curtis Devereux   - EPA/Monitoring and Data Analysis Division
       John Robson       - EPA/Land Use Planning Office
       David Sanchez     - EPA/Control Programs Development Division
       James Southerland - EPA/Monitoring and Data Analysis Division
       James Wilson      - EPA/Monitoring and Data Analysis Division
       Special thanks are due to Joseph Sableski and John Silvasi who provided'
overall guidance to this work.

-------
                                      190
1.   Cirillo, R.R. and M.J. Senew, Development of Computerized Emission Pro-
     jection and Allocation System — Phase I;  Preliminary Feasibility Study.
     Report No. EPA-450/3-77-001, U.S. Environmental Protection Agency, Research
     Triangle Park, N.C. (December 1976).
2.   Guidelines for Air Quality Maintenance Planning and Analysis,  Volume 7;
     Projecting County Emissions, Second Edition:.   Report No.  EPA-450/4-74-008
     U.S.  Environmental Protection Agency,  Research Triangle Park,  N.C. 27711
     (January 1975).

3.   Compilation of Air Pollutant Emission  Factors,  Second Edition  (with
     Supplements).  Report  No.  AP-42,  U.S.  Environmental Protection Agency,
     Research Triangle Park,  N.C. 27711 (April 1973).

4.   Guidelines for Air Quality Maintenance Planning and Analysis,  Volume 13;
     Allocating Projected Emissions to Subcounty Areas,  Report No.  EPA-450/4-
     74-014, U.S.  Environmental Protection  Agency,  Research Triangle Park, N.C.
     27711 (November 1974).
5.   Benesh, F.H. and P.D.  McLellan, Guidelines for State/Local Information
     Systems for Monitoring Emissions Growth and Air Quality Maintenance,
     Report No. GCA-TR-77-20-G(a), GCA Corp., Bedford,  Mass. (July  1977).
6.   Goodrich, J.C., Task 1 - Emissions Projection Methodology and  Its Appli-
     cation to the Hackensack Meadowlands Development Plans, Part I;  Emission
     Projection Methodology.  ERT Report No. P-24-1,  Environmental Research and
     Technology, Lexington, Mass. (May 1972).
7.   Goodrich, J.C., Task 1 - Emissions Projection Methodology and  Its Appli-
     cation to the Hackensack Meadowlands Development Plant, Part II:  Dis-
     cussion of Emission Inventories, ERT Report No. P-244-1,  Environmental
     Research and Technology, Lexington, Mass. (May 1972)-
8,   Reifenstein, E.G. and  M.J. Keefe, Task 5 Study Report, the AQUI? Software
     System User's Mamiqlt  ERT Report No. P-244-5,  Enviromental Research and
     Technology, Lexington, Mass. (May 1972).
9.   Willis, B.H., J.R. Mahoney, J.C. Goodrich, Task 4  - Guidelines for the
     Consideration of Air Pollution in Urban Planning,  ERT Report No. P-244-5,
     Environmental Research and Technology, Lexington,  'Mass. (August 1972).
10.  Willis, B.H., J.R. Mahoney, J.C. Goodrich, Hackensack Meadowlands Air
     Pollution Study - Air  Quality, Impact of Land Use Planning, Report No.
     EPA-450/3-74-056e, U.S.  Environmental  Protection Agency,  Research Triangle
     Park, N.C. (July 1973).
11.  Guidelines for Air Quality Maintenance Planning and Analysis,  Volume 8.
     Computer Assisted Area Source Emissions Gridding Procedure, Report No.
     EPA-450/4-74-009, U.S. Environmental Protection Agency, Research Triangle
     Park, N.C. (September  1974).
12.  IBM to UNIVAC Conversion and Gridding  Insertion for the Computer Assisted
     Area Source Emissions  Program (CAASE), Progress Reports 1-11,  Research
     Triangle Institute, Research Triangle  Park, N.C.  (August 1976  - June 1977).
13 .  Users Manual for ESVOL 13, A Computer  Program for  Allocation and Pro-
     jection of Residential Emissions, Engineering-Science, McLean, VA.
     (March 1975).

-------
                                     191
                            REFERENCES CCont'd)



14 •  Engineering-Science Air Qualify System (ESAQ) ,  Engineering-Science,
     McLean, VA.

15 *  Impact Assessment;  1980. 1985, 1995,  Air Quality Implications of  Growth
     Forecasts T  Metropolitan Washington Council of  Governments,  Washington,
     D.C. (March 1977).
16 •  Air Quality Maintenance Planning, Technical Analysis;   Projection Process.
     Metropolitan Washington Council of Government,  Washington,  D.C.  (October
     1976).

17.  Private communication from V. Krause, Metropolitan Washington Council of
     Governments, Washington, D.C. (June 1977).

-------
                                           192
TECHNICAL REPORT DATA
(Please read Ittsaucrions on the reverse before completing}
1. REPORT NO. 2.
EPA 450/3-77-028
4. TITLE AND SUBTITLE
Development of Computerized Emission Projection and
Allocation System — Phase II: Comparison of Existing
Systems
7. AUTHOR(S)
Richard R. Cirillo and George A. Concaildi
9. PERFORMING ORGANIZATION NAME ANO ADDRESS
Energy Research and Development Administration
Argonne National Laboratory
Energy and Environmental Systems Division
9700 South Cass Ave., Argonne, IL 60439
12. SPONSORING AGENCY NAME ANO ADDRESS
U.S. Environmental Protection Agency
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT N<
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
Interagency Agreement No.
D7-0077
13. TYPE OF REPORT ANO PERIOD COVEREC
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
Another report may follow if decision is made to continue study on the CEPA system.
Phase III would cover the development of the detailed system specification.
 16. ABSTRACT
   This report documents the second phase of a feasibility study  to determine  the need
   for a computerized emission projection and allocation  (CEPA) system  to  assist  State
   and local air pollution control agencies in conducting air quality analyses.   This
   phase entailed the review and evaluation of four existing emission analysis systems:
   the Air Quality for Urban and Industrial Planning  (AQUIP) system, the Computer-
   Assisted Area Source Emission (CAASE) gridding procedure, the  Engineering-Science
  -Air Quality (ESAQ) system, and the Metropolitan Washington Council of Governments
   (MWCOG) model.  The evaluation consisted of a description of the CEPA requirements
   without reference -to any existing systems, a comparison of the existing packages
   to those requirements, an identification of deficiencies, an estimate of effort re-
   quired to remove those deficiencies, an evaluation of the effort needed to  develop
   an entirely new system, and an assessment of the potential savings to be realized
   by employing a CEPA system in place of manual procedures.
   The report recommends that EPA proceed with stepwise modification of the Engineering-
   Science model by first documenting the model and making it available.   After  that,
   EPA could then begin to modify the model to correct deficiencies uncovered  by  the
   contractor.  The contractor estimates that the cost of modifying the Engineering-
   Science model ranges from $235,000 to $355,000.  This was the  lowest cost of  all
   the orations considered.
17. KEY WORDS ANO DOCUMENT ANALYSIS
a. DESCRIPTORS
Air Pollution
Atmosphere Contamination Control
Regional Planning
Release unlimited
tUOENT!FIERS/OP=N SNDEO TERMS
National Ambient Air Qual
Standards
Air Quality Maintenance A
Feasibility Study
Automatic Data Processing
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS /This page)
Unclassified
c. COSATl Field/Group
iC7 13-B
lalysis
21. NO. OF PAGES
22. PRICE
SPA Form 2220-1 ;S-73)

-------