Skip to Content Skip to Search Skip to Left Navigation U.S. Department of Transportation (US DOT) Logo Research and Innovative Technology Administration (RITA) Logo National Transportation Library
  ABOUT RITA | CONTACT US | PRESS ROOM | CAREERS | SITE MAP
 


Procedures for Emission Inventory Preparation - Vol IV: Mobile Sources





Click HERE for graphic.





                          ACKNOWLEDGMENT


     Several people in EPA's Emission Planning and Strategies
Division have contributed to this document.  Their names and
contributions are listed below.


     Chapter 1      Natalie Dobie       Introduction

     Chapter 2      Natalie Dobie       Overview

     Chapter 3      Terry Newell        MOBILE4.1
                    Natalie Dobie       Vehicle Miles Traveled

     Chapter 4      Greg Janssen        Nonroad Sources
                    Joe Somers

     Chapter 5      Richard Wilcox      Aircraft

     Chapter 6      Peter Okurowski     Locomotives








                              NOTICE
     The Procedures For Emission Inventory Preparation consists of
these five volumes:

               Volume   I -  Emission Inventory Fundamentals
               Volume  II -  Point Sources
               Volume III -  Area Sources
               Volume  IV -  Mobile Sources
               Volume   V -  Bibliography

     Volume I is a guide to the managerial and technical aspects of
the emission inventory.  It outlines the information sources
available, methods of estimating emissions, data validation and
quality assurance techniques, as well as procedures to maintain and
update the inventory.  Also included are a detailed analysis of the
manpower and resources required to derive each component of an
emission inventory and a comprehensive glossary.

     Volume II discusses point sources identification, data
collection, emissions calculation, and data presentation.  It
establishes standardized methods and procedures to develop a point
source data base.

     Volume III outlines the methods of collecting and handling
emission data from sources too small and/or too numerous to be
surveyed individually.  Collectively, these sources are known as
area sources.  Procedures are presented to identify area source
categories.  Important reference material that can be used to
determine the activity levels associated with area source
categories are also listed.  Finally, emission factors, emission
calculations, pollutant allocation and projection techniques, and
methods of data presentation are included to assist in the
preparation and maintenance of the area source emission
inventories.

     Volume IV presents an overview of the mobile source category
as a whole and identifies specific methods that can be used to
identify and inventory sources, estimate emissions, and establish
and maintain a useful, current mobile source emissions inventory.

     Volume V presents an extensive listing of currently available
reference material designed to assist in the development of an
emission inventory.  A concise abstract is provided for each
reference cited, outlining the pertinent emission inventory
information.

     These volumes are intended to present emission inventory
procedures and techniques applicable to state and local air
programs.  Please forward comments and suggestions for improvement
to the U. S. Environmental Protection Agency, Monitoring And
Reports Branch (MD- 14), Research Triangle Park, North Carolina
27711.







Other U.S. EPA emission inventory procedures publications include:

Emission Inventory Requirements For Ozone State Implementation
Plans, EPA-450/4-91-010, U. S. Environmental Protection Agency,
Office of Air Quality Planning and Standards, Research Triangle
Park, North Carolina, March 1991.

Procedures for the Preparation of Emission Inventories for Carbon
Monoxide and Precursors of Ozone, Volume 1: General Guidance for
Stationary Sources EPA-450/4-91-016, U. S. Environmental Protection
Agency, Office of Air Quality Planning and Standards, Research,
Triangle Park, North Carolina, May 1991.

Procedures for the Preparation of Emission Inventories for Carbon
Monoxide and Precursors of Ozone, Volume II: Emission Inventory
Requirements for Photochemical Air Quality Simulation Models, EPA-
45014-9-014, U. S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Research Triangle Park, North
Carolina, May 1991.

Emission Inventory Requirements for Carbon Monoxide State
Implementation Plans, EPA-450/4-91-011, U. S. Environmental
Protection Agency, Office of Air Quality Planning and Standards,
Research Triangle Park, North Carolina, March 1991.

Example Documentation Report For 1990 Base Year Ozone and Carbon
Monoxide State Implementation Plan Emission Inventories, EPA-450/4-
92-007, U. S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Research Triangle Park, North
Carolina, March 1992.

AIRS Facility Subsystem Source Classification Codes (SCCs) and
Emission Factor Listing for Criteria Pollutants, EPA-450/4-90-003,
U. S. Environmental Protection Agency, Office of Air Quality
Planning and Standards, Research Triangle Park, North Carolina,
March 1990.  Revised edition to be issued Summer 1992.

Guidance for the Preparation of Quality Assurance Plans O3/CO SIP
Emission Inventories, EPA-450/4-88-023, U. S. Environmental
Protection Agency, Office of Air Quality Planning And Standards,
Research Triangle Park, North Carolina, December 1988.

Quality Review Guidelines For 1990 Base Year Emission Inventories,
EPA450/4-91-022, U. S. Environmental Protection Agency, Office of
Air Quality Planning and Standards, Research Triangle Park, North
Carolina, September 1991.





SIP Air Emission Inventory Management System (SAMS) Version 4.1 and
SAMS User's Guide, U. S. Environmental Protection Agency, Office of
Air Quality Planning and Standards, Research Triangle Park, North
Carolina, September 1991.

User's Guide to MOBILE4.1 (Mobile Source Emission Factor Model),
EPA-AA-TEB-91-01, U. S. Environmental Protection Agency, Office of
Mobile Sources, Ann Arbor, Michigan, July 1991.

Procedures for Estimating and Applying Rule Effectiveness in Post-
1987 Base Year Emission Inventories for Ozone and Carbon Monoxide
State Implementation Plans, U. S. Environmental Protection Agency,
Office of Air Quality Planning and Standards, Research Triangle
Park, North Carolina, June 1989.

Surface Impoundment Modeling System (SIMS) Version 2.0 User's
Manual, EPA-450/4-90-019a, U. S. Environmental Protection Agency,
Research Triangle Park, North Carolina, September 1990.

Background Document for Surface Impoundment Modeling System (SIMS)
Version 2.0, EPA-450/4-90-019b, U. S. Environmental Protection
Agency, Research Triangle Park, North Carolina, September 1990.








                         TABLE OF CONTENTS

                                                                   Page

1.0 INTRODUCTION                                                    1

2.0 OVERVIEW OF THE MOBILE SOURCE CATEGORY                          2
     2.1  INDIVIDUAL MOBILE SOURCE CATEGORIES                       2
          2.1.1 Highway Vehicles                                    2
          2.1.2 Nonroad Sources                                     3
          2.1.3 Aircraft                                            3
          2.1.4 Locomotives                                         4

3.0 EMISSIONS FROM HIGHWAY VEHICLES                                 5
     3.1  GUIDANCE ON THE USE OF MOBILE4.1 VS MOBILE5 FOR THE 1990
     BASE YEAR INVENTORY AND OTHER INVENTORIES                      5
     3.2 MOBILE SOURCE EMISSION ESTIMATION PROCESS                  6
          3.2.1     Overview of Factors Influencing Motor Vehicle  
                    Emission Inventories                            6
               3.2.1.1  Vehicle Fleet Activity                      7
               3.2.1.2  Emission Factors                            7
               3.2.1.3  Fleet Characteristics                       7
               3.2.1.4  Fuel Characteristics                        7
               3.2.1.5  Correction Factors                          8
               3.2.1.6  Control Programs                            8
          3.2.2 Overview of MOBILE4.1 Input Requirements            8
               3.2.2.1 Fleet Characteristics                        9
                    3.2.2.1.1 VMT Mix                               9
                    3.2.2.1.2 Annual Mileage Accumulation Rates     9
                    3.2.2.1.3 Registration Distributions           10
               3.2.2.2 Fuel Specifications                         10
                    3.2.2.2.1 RVP                                  10
               3.2.2.3 Correction Factors                          10
                    3.2.2.3.1 Speed                                10
                    3.2.2.3.2 Temperature                          10
                    3.2.2.3.3 Operating Modes                      11
                    3.2.2.3.4 Minor Correction Factors             11
               3.2.2.4 Tampering and Misfueling                    12
               3.2.2.5 Control Programs                            12
                    3.2.2.5.1 Refueling Emissions                  12
                    3.2.2.5.2 Inspection and Maintenance Programs  12
                    3.2.2.5.3 Anti-Tampering Programs (ATPs)       12
     3.3  GUIDANCE ON SELECTING MOBILE4.1 INPUTS                   13
          3.3.1     Emission Factors                               13
               3.3.1.1   Region                                    14
               3.3.1.2   Calendar Year                             14





                         TABLE OF CONTENTS

                                                               Page

          3.3.2     Fleet Characteristics                          15
               3.3.2.1 Vehicle Miles Traveled Mix by Vehicle Type  15
               3.3.2.2   Annual Mileage Accumulation Rates and
                         Registration Distributions by Vehicle 
                         Type and Age                              16
               3.3.2.3 Trip Length Distribution                    18
               3.3.2.4 Diesel Sales Fractions                      20
          3.3.3     RVP Determination                              22
               3.3.3.1 EPA-Provided 1990 RVP Estimates             26
               3.3.3.2 "Period 1" RVP and "Period 2" RVP           26
               3.3.3.3 Interpolation                               27
               3.3.3.4 Inputs for Future Year RVP                  27
                    3.3.3.4.1 Future Summer RVP                    27
                    3.3.3.4.2 Future Winter RVP                    28
          3.3.4     Oxygenated Fuels                               29
          3.3.5     Correction Factors                             30
               3.3.5.1 Speed                                       30
               3.3.5.2 Temperature                                 34                          
               3.3.5.3 Operating Modes                             38
               3.3.5.4   Additional Correction Factors for Light-
                         Duty Gasoline-Fueled Vehicle Types        40
          3.3.6 Control  Programs                                  43
               3.3.6.1 Refueling Emissions                         43
               3.3.6.2 Inspection and Maintenance Programs         45
                    3.3.6.2.1  I/M                                 47
                    3.3.6.2.2  Start Year                          47
                    3.3.6.2.3  Stringency                          47
                    3.3.6.2.4  First Model Year                    47
                    3.3.6.2.5  Last Model Year                     48
                    3.3.6.2.6  Waiver Rates                        48
                    3.3.6.2.7  Compliance Rate                     49
                    3.3.6.2.8  Inspection Frequency                50
                    3.3.6.2.9  Vehicle Classes                     51
                    3.3.6.2.10   I/M Test Types                    51
                    3.3.6.2.11   Alternate I/M Credits             53
                    3.3.6.2.12   Centralized Programs              53
                    3.3.6.2.13   Decentralized Programs (Manual)   53
                    3.3.6.2.14   Computerized Inspection           54
                    3.3.6.2.15   Tech I-II and Tech IV+            55
               3.3.6.3 Anti-Tampering Programs                     55
                    3.3.6.3.1 ATP                                  57
                    3.3.6.3.2 Tampering and Misfueling             57
                    3.3.6.3.3  Air Pump Inspection                 57





                         TABLE OF CONTENTS
                                                               Page

                    3.3.6.3.4 Catalyst Inspection                  57
                    3.3.6.3.5 Fuel Inlet Restrictor Inspection     58
                    3.3.6.3.6 Tailpipe Lead Detection Test         58
                    3.3.6.3.7 EGR Inspection                       59
                    3.3.6.3.8 Evaporative Control System           59
                    3.3.6.3.9 PCV Inspection                       60
                    3.3.6.3.10     Gas Cap Inspection              60
                    3.3.6.3.11     Tampering Rates                 60
     3.4  VEHICLE MILES TRAVELED                                   62
          3.4.1     Highway Performance Monitoring System          62
               3.4.1.1   Role of the HPMS in SIP Development       62
               3.4.1.2   Overview of HPMS                          63
               3.4.1.3   Consistency Between HPMS and SIP VMT      65
                    3.4.1.3.1 Expansion Factors                    65
                    3.4.1.3.1.1    Non-Attainment Area the Same As
                                   the Federal Aid Urbanized Area  65
                    3.4.1.3.1.2    Non-Attainment Area Inside of
                                   the Federal Aid Urbanized Area  66
                    3.4.1.3.1.3    Non-Attainment Area Outside of
                                   the Federal Aid Urbanized Area  66
                    3.4.1.3.1.4    Non-Attainment Area and Federal
                                   Aid Urbanized Area Crossover    67
                    3.4.1.3.2 Local Functional System              67
                    3.4.1.3.3 Seasonal Adjustment                  68
                    3.4.1.3.4 Daily Adjustment                     68
               3.4.1.4   Allocating VMT to Time of Day             68
               3.4.1.5   Allocating VMT to Functional Systems      69
               3.4.1.6   Estimating VMT in Rural and Small Urban
                         Areas                                     69
                    3.4.1.6.1 Apportionment of Statewide VMT-
                              Recommended Method                   72
                    3.4.1.6.2 Apportionment of Statewide VMT-
                              Alternative Methods                  74
                         3.4.1.6.2.1    Motor Vehicle
                                        Registrations              74
                         3.4.1.6.2.2    Population                 74
                         3.4.1.6.2.3    Fuel Sales                 75
          3.4.2     Travel Demand Network Models                   78
               3.4.2.1   Role of Transportation Models in SIP
                         Development                               78
               3.4.2.2   Background.                               78
               3.4.2.3   Overview of Network Models                79
                    3.4.2.3.1 Level of Service                     81
                    3.4.2.3.2 Physical Attributes                  83
                    3.4.2.3.3 Locational Link Attributes           83
                    3.4.2.3.4 Trip Generation                      86





                         TABLE OF CONTENTS
                                                               Page

                    3.4.2.3.5 Trip Distribution                    86
                    3.4.2.3.6 Modal Split                          86
                    3.4.2.3.7 Traffic Assignment                   86
                    3.4.2.3.8 Feedback                             87

               3.4.2.4   Consistency Between Transportation Model
                         VMT and HPMS                              87
                    3.4.2.4.1 Non-Attainment Area the Same As the
                              Network Model Area                   88
                    3.4.2.4.2 Non-Attainment Area Inside of the
                              Network Model Area                   89
                    3.4.2.4.3 Non-Attainment Area Outside of the
                              Network Model Area                   89
                    3.4.2.4.4 Non-Attainment Area and Network Model
                              Area Crossover                       90
               3.4.2.5   Local Functional System                   90
               3.4.2.6   Seasonal Adjustment                       91
               3.4.2.7   Daily Adjustment                          91
               3.4.2.8   Allocating VMT to Time of Day             91
               3.4.2.9   Allocating VMT to Functional Systems      91
          3.4.3     Exception to the Use of HPMS VMT               92
     Appendix 3-A                                                  94

4.0  EMISSIONS FROM NONROAD, SOURCES                               98
     4.1  Introduction                                             98
     4.2  Inventory Options Under This Guidance                    99
          4.2.1     Options for Areas With EPA Provided
                    Inventories                                    99
          4.2.2     Options For Areas With EPA Provided
                    Inventories                                   101
          4.2.3     Options For Areas Without EPA Provided
                    Inventories                                   102
     4.3  Explanation of EPA Provided Inventory                   102
          4.3.1     Derivation of AMS Inputs                      103
          4.3.2     AMS Inputs                                    105
     4.4  General Methodology Used In Deriving Emission Inventories
          For 33 Areas                                            107
          4.4.1     Explanation of Methodologies to Distribute
                    Equipment Within Each Category Type at the
                    County Level                                  108
          4.4.2     Explanation of Methodologies For Distributing
                    Equipment Within Each Category at the Sub-
                    County Level                                  112
          4.4.3     Seasonal Adjustment Methodology               113
     4.5  New York Non-Attainment Area Example                    115
     Appendix 4-A                                                 117
     Appendix 4-B                                                 124
     Appendix 4-C                                                 132
     Appendix 4-D                                                 135





     TABLE OF CONTENTS
                                                               Page

5.0  EMISSIONS FROM AIRCRAFT                                    137
     5.1  OVERVIEW OF THE INVENTORY METHODOLOGY                 137
          5.1.1     Factors Affecting Emissions                 138
               5.1.1.1   Aircraft Categorization                138
               5.1.1.2   Pollutant Emissions                    139
               5.1.1.3   Aircraft Engines                       140
               5.1.1.4   Operating Modes                        140
     5.2  INVENTORY METHODOLOGY                                 144
          5.2.1     Airport Selection                           144
          5.2.2     Mixing Height Determination                 145
          5.2.3     Activity and Emissions for Commercial 
                     Aircraft                                   149
          5.2.4     Activity and Emissions for General Aviation  
                    Aviation and Air Taxi Aircraft              173
               5.2.4.1   Aircraft-Specific Procedure            173
               5.2.4.2   Alternative, Fleet-Average Procedure      
                         176
          5.2.5     Activity and Emissions for Military Aircraft   
                    178
     5.3  VARIATIONS TO THE INVENTORY CALCULATION PROCEDURE 190
          5.3.1     Variability of Activity - Daily and Seasonal   
                    190
          5.3.2     Operational Activity that Affects Aircraft
                    Emissions                                   191
               5.3.2.1   Reduced Engine Taxiing                 191
               5.3.2.2   Derated Take-off                       192
          5.3.3     Particulate Emissions                       192
     5.4  OTHER EMISSION SOURCES                                192
          5.4.1     Auxiliary Power Units                       192
          5.4.2     Evaporative Emissions                       197
     5.5  EFFECT OF FUTURE CHANGES TO THE FLEET                 197
     5.6  CONVERTING FROM TOTAL HYDROCARBONS (THC) TO VOLATILE   
          ORGANIC COMPOUNDS (VOC)                               198
          5.6.1     Commercial and Military Conversions         198
          5.6.2     General Aviation and Air Taxi Conversions      
                    199
6.0  EMISSIONS FROM LOCOMOTIVES                                 200
     6.1  OVERVIEW OF RECOMMENDED INVENTORY METHODOLOGY         202
     6.2  RECOMMENDED METHODS                                   202
          6.2.1     Class I Line Haul Locomotives               202
               6.2.1.1   Fuel Consumption                       202
               6.2.1.2   Emission Factors:                      204
          6.2.2     Class II and III Line Haul Locomotives      205
               6.2.2.1   Fuel Consumption                       205
               6.2.2.2   Emission Factors                       206
          6.2.3     Yard Operations     206
               6.2.3.1   Number of Yard Locomotives             206
               6.2.3.2   Emissions Per Yard Locomotive          206





     TABLE OF CONTENTS
                                                               Page

     6.3  TAILORING METHODS                                     208
          6.3.1     Locomotive Roster Tailoring Method          208
               6.3.1.1   Identify the locomotives in the area      
                         208
               6.3.1.2   Determine the engine type              209
               6.3.1.3   Sum the total of the conversions       209
               6.3.1.4   Calculate the new fleet average emission
                         factors                                210
               6.3.1.5   Multiply the new emission factors by fuel
                         consumption                            210
          6.3.2     Duty Cycle Tailoring Method                 210
          6.3.3     SO2 Tailoring Method                       212
     6.4  ALTERNATIVE METHOD                                    213
     6.5  RE-ENGINED LOCOMOTIVES                                213
     6.6  CONVERTING FROM TOTAL HYDROCARBONS (THC) TO VOLATILE
          ORGANIC COMPOUNDS (VOC)                               213
     Appendix 6-1                                               215
     Appendix 6-2                                               216
     Appendix 6-3                                               217
     Appendix 6-4                                               219
     Appendix 6-5                                               222
     Appendix 6-6                                               223
     Appendix 6-7                                               225
     Appendix 6-8                                               227





                         1.0 INTRODUCTION

     A fundamental requirement in the effort to control pollution
in any form is to quantify the emissions being released.  This is
necessary to understand the relationships between emissions and the
ambient concentrations that result, and to develop appropriate
policies and methods to ensure that ambient pollutant
concentrations remain within acceptable limits.

     Specific air pollution requirements are set forth in Title 40,
Code Of Federal Regulations, Part 51.321 (40 CFR 51), and in the
Clean Air Act as amended, for the development and maintenance of
ongoing programs to inventory specific pollutant emissions.  States
are required by 40 CFR 51 to prepare and submit annual reports to
the U.S. Environmental Protection Agency (EPA) regarding the
emissions of particulate matter, sulfur oxides, carbon monoxide,
nitrogen oxides, and volatile organics from point sources within
their boundaries.  The amendments to the Clean Air Act require the
development of "...comprehensive, accurate, and current..."
inventories from all sources of each pollutant for every non-
attainment area, in conjunction with the preparation of revised
State Implementation Plans (SIPs).  EPA recognizes that a
significant effort will continue to be needed to develop and
maintain emission inventories to meet the requirements for both
technical analysis and administrative reporting.

     To assist the states in meeting the requirements for emission
inventory development, a five volume series has been prepared that
describes in detail many of the technical aspects of the inventory
process.  This document is the fourth volume in the series, and it
focuses on mobile sources.  Specifically, this document presents
specific methods that can be used to identify sources, estimate
emissions, and establish and maintain a useful, current mobile
source emissions inventory.  Special attention has been given to
preparing the 1990 SIP inventories.

     Following this introductory chapter, Chapter 2 gives an
overview of the mobile source category.  Chapters 3 through 6
present specific methods that should be used to derive emission
estimates for each of the primary mobile source subcategories.

     Chapter 1 - Introduction
     Chapter 2 - Overview
     Chapter 3 - Highway Vehicles
     Chapter 4 - Nonroad Sources
     Chapter 5 - Aircraft
     Chapter 6 - Locomotives





            2.0 OVERVIEW OF THE MOBILE SOURCE CATEGORY

     An inventory of pollutant emission sources should classify
sources into two major categories - point sources and area sources. 
The point source category is described in detail in Volume II of
this series.  The area source category is described in detail in
Volume III of this series.  Mobile sources are a subcategory within
the area source category of pollutant emission sources.  However,
the procedures for preparing and maintaining an inventory of
emissions from mobile sources are presented herein, as a separate
document in this series, because the inventorying procedures are
different from those for other area source subcategories and
because the mobile source emissions inventory represents a major
portion of the total emissions of volatile organics (VOC), nitrogen
oxides (NOx ), and carbon monoxide (CO).

     The mobile sources for which inventory and emission
calculation procedures are presented in this document are highway
vehicles, nonroad mobile sources, aircraft, and locomotives. 
Recreational marine equipment and commercial marine vessels are
discussed in the nonroad mobile source section.  The procedures
describe how to calculate tailpipe emissions and emissions from the
fuel carried on the vehicle (evaporative VOC emissions) for these
four mobile source categories.  The emissions that result from tire
wear and travel over roads or other surfaces should be calculated
from the procedures in Volume III of this series and are
specifically excluded from consideration in this document.

2.1  INDIVIDUAL MOBILE SOURCE CATEGORIES

2.1.1     Highway Vehicles

     Highway vehicles include all vehicles registered to use the
public roadways.  The predominant emissions source in this category
is the automobile, although trucks and buses are also significant
sources of emissions.

     The total highway vehicle population can be characterized by
eight individual vehicle type categories:

-    Light duty gasoline powered vehicles (LDGV);
-    Light duty gasoline powered trucks, from 0 to 6000 lb.
     gross vehicle weight (LDGT1);
-    Light duty gasoline powered trucks, from 6001 to 8500 lb.
     gross vehicle weight (LDGT2);
-    Heavy duty gasoline powered vehicles (HDGV);
-    Light duty diesel powered vehicles, from 0 to 6000 lb. gross
     vehicle weight (LDDV);
-    Light duty diesel powered trucks (LDDT);
-    Heavy duty diesel powered vehicles (HDDV);
-    Motorcycles (MC).


                                 2





     Numerous characteristics for each vehicle type are necessary
before emissions can be calculated.  These characteristics include,
among others, model year, the age distribution of vehicles within
the class, annual mileage by vehicle age, and average speed. 
Chapter 3 of this document presents detailed procedures for
identifying and using these and other key characteristics.

2.1.2 Nonroad Sources

     This mobile source category includes a diverse set of source
types.  The movement of sources in this category occurs on surfaces
other than the public highways.  Nonroad vehicles can be classified
into ten categories:

     -    Lawn and Garden Equipment,
     -    Industrial Equipment,
     -    Airport Service Equipment,
     -    Construction Equipment,
     -    Recreational Equipment,
     -    Agricultural Equipment,
     -    Recreational Marine Equipment,
     -    Logging Equipment,
     -    Light Commercial Equipment,
     -    Commercial Marine Vessels.

     These categories are difficult to inventory, since few data
are available to determine either their activity levels or
operating characteristics.  Chapter 4 of this document provides
procedures for inventorying and estimating emissions from these
categories.

2.1.3 Aircraft

     Aircraft include all types of aircraft, whether civilian,
commercial, or military. Emissions from idling, taxiing, and during
landings and takeoffs are included.  Landing and takeoff cycle
(LTO) emissions are those that occur between ground level and an
attitude of about 3000 feet.  Aircraft emissions above 3000 feet
need not be included in either the base year emission inventory or
in the modeling inventory.

     The larger civil and commercial airports with continuously
manned control towers maintain records of LTO cycles by type of
aircraft as part of their standard operating procedure.  Smaller
airports also maintain these records to the extent that their
control towers are manned or landing fees are recorded.  Difficulty
may be encountered in obtaining data on military aircraft
operations at military airports.

     EPA has compiled a complete set of emission factors for
different types of aircraft operating in the different modes (idle,
taxi, LTO).  Chapter 5 of this document provides instruction on how
to calculate emissions from this mobile source category.

                                 3





     2.1.4     Locomotives

     Locomotives include all fossil fuel fired locomotive engines
operated on railways.
The quantity of fuel used by locomotives and the size, in
horsepower, of the locomotives are
necessary to calculate emissions from this source.  This
information is discussed in Chapter 6.

 0

     3.0  EMISSIONS FROM HIGHWAY VEHICLES

     In most urban areas,     highway vehicles represent the
largest single source of carbon
     monoxide (CO) emissions and   contribute significantly to the
     area's production of volatile
organic compounds (VOC), sulfur oxides (SO2) and oxides of
nitrogen (NOx ).

     Emission estimates for highway vehicles are usually based on
the combination of two fundamental measures of activity: travel and
the average rate of pollutants emitted in the course of travel. 
Both measures reflect complex patterns of behavior.

     The Environmental Protection Agency and the Department of
Transportation Federal Highway Administration (FHWA) have developed
a series of tools/models to estimate the rate of emissions produced
by vehicles per mile of travel and the amount of travel itself. 
The knowledge base and disciplines required to understand and
operate these models are distinct, as are their audiences.  This
distinction generally ensures that environmental analysts have
little appreciation for the accuracy of the travel estimates
produced by transportation analysts and vice versa.

     The purpose of this chapter is to provide guidance for
preparing the highway vehicle portion of mobile source emission
inventories, particularly those associated with the development of
State Implementation Plans (SIPs) for ozone (03) and CO.  The
accuracy of the inventory will be no better than the accuracy of
the estimates of either the emission rates or vehicle miles
traveled (VMT).

     This chapter responds to concerns that little effort has been
devoted to the
development of accurate projections of travel within non-attainment
areas, that projections of attainment dates have been based on
dated information and that highway vehicles are responsible for a
greater portion of the emissions inventory than recent estimates
have suggested.  Because of these concerns, the earlier guidance on
the use of available travel estimates has been carefully reviewed
and updated.

     3.1  GUIDANCE ON THE USE OF MOBILE4.1 VS MOBILE5 FOR THE 1990
BASE YEAR INVENTORY AND OTHER INVENTORIES

     At the time of this writing, MOBILE4.1 is EPA's current
emission factor model.  MOBILE5 will be available within months of
the publication of this document.  EPA will accept 1990 base year
emission inventories prepared with either MOBILE4.1 or MOBILE5
emission factors.  The November 15, 1992 submittal date for
inventories will apply no matter which version of the model is
used.

     Since MOBILE5 will incorporate the new vehicle standards for
VOC and NOx  mandated by the Clean Air Act (CAA), estimates of
those emissions for years after 1990 will be significantly
different than those estimated by MOBILE4.1. Therefore, ozone non-
attainment areas should submit projections using MOBILE5.  However,
the 1990 highway vehicle emissions inventories should be
recalculated as soon as possible after November 15, 1992 using
MOBILE5 so that all required inventories are consistent.  CO

                                 5





non-attainment areas may use MOBILE4.1 for the November 15, 1992
projection submittal, and if they do, recalculation of the 1990
inventory is not necessary.

     The SIP Attainment/Reasonable Further Progress Demonstration,
including projection year inventories, should use MOBILE5 for ozone
non-attainment areas.' In addition, if the base year inventory was
originally developed using MOBILE4.1, it should be recalculated
using MOBILE5 and resubmitted.'

     For CO non-attainment areas, the base year and projection year
inventories may be developed using either MOBILE4.1 or MOBILE5. 
Submissions after the November 15, 1992 submission should use
MOBILE5.  Such submissions may be voluntary or due to bump up or
other provisions of the Clean Air Act.

     The release of MOBILE5 will be accompanied by a supplement to
this document explaining the differences between MOBILE4.1 and
MOBILE5 and the additional inputs contained in MOBILE5.

     3.2  MOBILE SOURCE EMISSION ESTIMATION PROCESS

     3.2.1     Overview of Factors Influencing Motor Vehicle
Emission Inventories

     Many complex processes govern the formation of pollutants in
motor vehicles.  The EPA and the California Air Resources Board
(CARB) maintain large data collection program to quantify the rate
at which pollutants are emitted by individual categories of motor
vehicles.  Both organizations have used this information to develop
models that help analysts in estimating motor vehicle contributions
to the local emissions inventory.  These models, commonly known as
emission factor models, are designed to account for the effect of
numerous vehicle parameters on the volume of pollutants emitted. 
The current EPA model is called MOBILE4. 1.

     The primary components of an emission factor model include the
base emission factors, characterization of the vehicle fleet, fuel
characteristics, vehicle operating conditions and the effect of
local ambient conditions, the effect of alternative I/M programs
and the effect of tampering and misfueling.  None of these factors
is static: technology is continually evolving, leading to changing
in-use emission performance.  Changes in fuel prices and economic
conditions lead to changes in vehicle sales and travel patterns.  A
substantial effort is required to accurately quantify these factors
and to stay current with the influence of all of these factors on
vehicular emission levels.
___________________________

     1  The SIP Attainment/Reasonable Further Progress and
projection year inventories are due on either November 15, 1993 or
November 15, 1994, depending upon the non-attainment
classification.

     2  EPA may set a date prior to November, 1993 for submission
of draft projection and recalculated base year inventories, similar
to the current requirement to submit the draft base year inventory
no later than May, 1992.

                                 6





     3.2. 1.1 Vehicle Fleet Activity

     It is standard practice in preparing highway vehicle emission
inventories to express vehicle activity in terms of vehicle miles
traveled, and the emission factors in units of grams per mile of
travel.  Actually, vehicles also emit hydrocarbons while
stationary.  Estimates of emission-producing activities that do not
involve travel are built into MOBILE4.1. These non-moving emissions
are spread over estimated miles of travel by vehicles of a
particular age and output as an equivalent per mile emission
factor.  Therefore, EPA will accept VMT as the measure of local
vehicle activity for all inventories required under the Clean Air
Act.3

     VMT can be estimated in several possible ways.  Direct
observation via traffic counts (usually at a sample of roadway
points with statistical expansion to represent the universe of all
roadways in the area) and highway/transit network models are the
more preferred approaches.  EPA does not recommend reliance on fuel
sales data, owner reports, or periodic odometer surveys as
substitutes.  The two recommended methods are discussed in Section
3.4.

3.2.1.2 Emission Factors

     Emission rates are computed from test measurements of in-use
vehicles at various odometer readings designed to capture two
fundamental processes: the baseline emission rate and the
deterioration that takes place as the vehicle ages.  Linear
regressions are performed on the data to quantify the level of
pollutants emitted by each model year's vehicles.  The results are
commonly referred to as the intercept, or zero-mile (ZM), emission
rate and the slope, or deterioration rate (DR), that occurs over
each 10,000 mile interval.

3.2.1.3 Fleet Characteristics

     The emission factors quantify the performance of individual
model year vehicle fleets by vehicle type.  The age distribution,
the rate of mileage accumulation and the mix of travel experienced
by each vehicular category can significantly alter the fleet
average emission rate.  While the emission factor models employ
national average distributions for each of the factors, local input
is allowed, often encouraged, and, for some inputs, required. 
Differences between local and national average distributions can
alter the emissions contributions of the individual vehicle
categories.

     3.2.1.4   Fuel Characteristics

     Emission test measurements are conducted on a standardized
test fuel known as Indolene.  The characteristics of this fuel are
well defined and ensure that test results are repeatable.  Since
consumers cannot purchase Indolene at their local service stations
and
___________________________

     3  VMT must usually be disaggregated such that each subset of
it can be reasonably represented by a single emission factor
determined by one set of inputs of the types described below.  EPA
also accepts the trip-based activity methods described in this
document.

                                 7





differences between the volatility of local fuels and Indolene can
influence the level of both evaporative and tailpipe pollutants,
MOBILE4.1 requires local input of fuel volatility.

3.2.1.5 Correction Factors

     To ensure the repeatability of measurements, standardized test
conditions have been specified for each vehicle category.  They
include driving cycle, temperature, humidity, vehicle load, and the
distribution of starting conditions.  Since not all vehicle trips
match these test conditions, a series of correction factors has
been developed to allow the emission factor model to account for
differences.

3.2.1.6 Control Programs

     Emission factors are based on the performance of vehicles
independent of any local control programs such as I/M, anti-
tampering and Stage II refueling.  Each of these programs is
designed to reduce the level of pollutants emitted by vehicles
operating under in-use conditions.  Further, differences in program
designs can have a significant impact on their effectiveness in
reducing emissions.  Therefore, it is important to specify
correctly program parameters in order to estimate correctly their
effect on vehicular emissions.

     3.2.2     Overview of MOBILE4.1 Input Requirements

     MOBILE4.1, EPA's emission factor model, computes separate
emission estimates for eight vehicle categories:

     -    Light-duty gasoline-powered vehicles (LDGV), i.e.,
          passenger cars;
     -    Light-duty diesel-powered vehicles (LDDV), i.e., diesel-
          powered passenger cars;
     -    Light-duty gasoline-powered trucks, type 1 (LDGT1), i.e.,
          pickup trucks and vans that have a gross vehicle weight
          (GVW) of 0 - 6000 pounds;
     -    Light-duty gasoline-powered trucks, type 2, (LDGT2),
          i.e., pickup trucks, vans, and other, small trucks that
          have a GVW of 6001 - 8500 pounds;
     -    Light-duty diesel-powered trucks, types 1 & 2 (LDDT);
     -    Heavy-duty gasoline-powered trucks (HDGV), i.e., all
          vehicles with a GVW greater than 8,500 pounds, powered by
          gasoline engines;
     -    Heavy-duty diesel-powered vehicles (HDDV), i.e., all
          diesel powered trucks with a GVW greater than 8,500
          pounds; and
     -    Motorcycles (MC).

     There are large differences in the emission characteristics of
the vehicles represented by these categories; therefore, it is
important that estimates of local or regional emission rates
incorporate the distribution of VMT by vehicle type.

     The emission factors produced by MOBILE4.1 are derived from
measurements conducted under standardized test conditions.  For
light-duty vehicles, the standard set of test conditions is
referred to as the Federal Test Procedure (FTP).  It involves the
simulated

                                 8





operation of a vehicle over a specific driving cycle, the Urban
Driving Cycle, under controlled operating and environmental
conditions, during which emissions are measured in three sequences. 
The Urban Driving Cycle represents an average trip over an urban
network that includes travel on local and arterial streets, major
arterials, and expressways.  The basic test conditions include:

     -    Ambient temperature range of 68'F to 86'F;
     -    Absolute humidity adjusted to 75 grains of water per
          pound of dry air;
     -    Average speed of 19.6 mph with 18 percent idle operation;
     -    Average percent of VMT` in cold start operation of 20.6
          percent;
     -    Average percent of VMT in hot start operation of 27.3
          percent;
     -    Average percent of VMT` in stabilized operation of 52.1
          percent; and
     -    Average trip length of 7.5 miles.

     In order to understand fully the derivation of emission
factors and the influence of these conditions on emission levels,
refer to Chapter 2 of the MOBILE4.1 User's Guide.  A condensation
of that material is included in Section 3.3 of this report.

     MOBILE4.1 inputs can be altered to reflect city-specific
conditions.  A brief review of each of the primary options is
presented below.  They are not organized as they are in the
MOBILE4.1 User's Guide, but rather in the order in which they will
be discussed in more detail later in this chapter.

3.2.2.1   Fleet Characteristics

3.2.2.1.1 VMT Mix

     The distribution of travel across the eight vehicle categories
determines how the individual emission factors are weighted to
produce a composite emission factor for the entire highway vehicle
fleet.  The LDGVs generally comprise over 50 percent of the travel
recorded in any area of the country and, therefore, tend to be the
dominant source of highway emissions. (HDDVs are an important
source of NOx emissions.) MOBILE4.1 will calculate the VMT mix
based on national data characterizing registration distributions,
annual mileage accumulation rates by age, diesel sales fractions,
and vehicle counts.  These values may not, however, be
representative of certain areas, such as western states where
pickup trucks form a larger share of the vehicle population or
rural areas where a broader distribution of vehicles exists.

3.2.2.1.2 Annual Mileage Accumulation Rates

     The primary effect of the rate of mileage accumulation by age
(in combination with registration data) is to determine the
relative weighting of each model year's contribution to the average
emission factor computed for each vehicle category.  MOBILE4.1
provides the option of using a national average value or inputting
data characterizing local conditions.  The rate of mileage
accumulation may be different from national average conditions in
both rural and urban areas at either end of the economic spectrum.

                                 9





3.2.2.1.3 Registration Distributions

     These are used in concert with mileage accumulation rates to
determine the relative weighting of each model year's contribution
to the average emission factor for each vehicle category. 
MOBILE4.1 provides the option of using national average values or
inputting data characterizing local registrations.  The areas most
likely to be distinct from national average values are rural areas,
areas in which cars do not rust out and urban areas at either end
of the economic spectrum.

3.2.2.2   Fuel Specifications

3.2.2.2.1 RVP

     Evaporative and, to a lesser extent, exhaust emissions vary
with fuel volatility.  EPA's new vehicle certification program and
much of its in-use vehicle testing program use gasoline with a fuel
volatility (RVP) of 9.0 psi.  In recent years much of the country
has been supplied with gasoline of higher volatility.  MOBILE4.1
adjusts estimated emission factors to account for the effects of
volatility.  No national average value for this variable is
available in MOBILE4.1; one must supply this input.

3.2.2.3   Correction Factors

3.2.2.3.1 Speed

     Emission factors are very sensitive to the average speed that
is assumed.  In general, emissions tend to increase as average
speeds decrease from the 19.6 mph average FTP speed.  MOBILE4.1
does not assume an average speed; rather it requires that an
estimate of the speed experienced by vehicles operating in the area
and roadway segment or collection of interest be specified. 
MOBILE4.1 adjusts the emission factors for speeds other than 19.6
mph through the use of speed correction factors.  These
multiplicative adjustments to the base emission factors tend to
follow a non-linear relationship that increases the emission levels
as speeds decline from 19.6 mph and increase beyond 48 mph.4

3.2.2.3.2 Temperature

     Emissions from mobile sources are significantly influenced by
the ambient temperatures under which they are operating. 
Temperature has an effect on both the exhaust and the evaporative
emission levels.  MOBILE4.1 deals with these effects separately. 
In general, exhaust emissions are at a minimum at the temperature
specified for the FTP (75'F), with emissions increasing as
temperature either increases or decreases from that value.  No
ambient temperature is assumed by MOBILE4.1. One must be provided
as an input to the model.

___________________________

     4  The speed correction factors in MOBILE5 may be
significantly revised at speeds above 48 mph.

                                10





3.2.2.3.3 Operating Modes

     Emission factors based on FTP measurements are collected for
three separate segments, usually referred to as bags because the
vehicle exhaust is collected in three separate teflon bags, each
with differing emissions performance.  The three bags correspond to
the following modes of operation: cold start, hot stabilized, and
hot start.  Bag 1, the cold start mode, reflects conditions
experienced at the beginning of a trip when the engine and the
emission control system begin operation at ambient temperature and
are not performing at optimum levels (i.e., the catalyst is cold
and has not reached the "light off" temperature needed to
efficiently control emissions coming from the engine) until part
way through the trip.5 The hot start mode, Bag 3, reflects the
condition of an engine that has been restarted after being turned
off for 10 minutes and, therefore, has not cooled to ambient
conditions.  Under this circumstance the engine and catalyst are
warm and, although not at peak operating efficiency when started,
still have significantly improved emissions performance relative to
the cold start mode.  Bag 2, the hot stabilized mode, reflects the
condition of the engine when the vehicle has been in continuous
operation long enough for all systems to have attained stable
operating temperatures.  The proportion of VMT accumulated in cold
and hot start modes must be specified based on the conditions in
the area to be modeled.  Specifications must be made for catalyst
and non-catalyst vehicles separately.

3.2.2.3.4 Minor Correction Factors

     This category has been added to cover the effects of four
special correction factors that are available:

     -    Air conditioning;
     -    Extra vehicle loading;
     -    Trailer towing;
     -    NOx humidity.

     These factors are designed to account for the effect of
unusual vehicle operating conditions relative to those experienced
in the FTP.  Generally, it is difficult to quantify the extent of
these vehicle operating parameters, and their effect on emission
factors tends to be small.  Therefore, EPA recommends that few
resources be expended to develop the inputs needed.  The effect of
the NOx humidity correction factor is also slight, and, unless
NOx is of particular concern, little effort should be devoted to
its use.


     5  "Bag 1" is usually a mix of cold and warmed operation,
since, except under very cold ambient conditions, the 505 seconds
of driving represented by this bag constitutes a longer period than
is needed for the engine and catalyst to get warm.

                                11





3.2-2.4 Tampering and Misfueling

     The basic emission factors in MOBILE4.1 receive an adjustment
to account for estimates of vehicle tampering rates as a function
of accumulated mileage for each gasoline-fueled vehicle category
and eight categories of tampering (e.g., air pump disablement,
misfueling, etc.). These rates are combined with offsets (the
increase in emissions that results from the given type of
tampering) and added to the non-tampered emission factors.  Options
are available to input local tampering rates.  The use of local
information must be supported by an approved survey.  If locally
developed information is not available, a national average rate
will be used by MOBILE4.1.

3.2.2.5   Control Programs

3.2.2.5.1 Refueling Emissions

     The refueling of gasoline-fueled vehicles results in the
displacement of fuel vapor from the vehicle fuel tank to the
atmosphere.

     There are two basic approaches to the control of vehicle
refueling emissions, generally referred to as "Stage III" (at the
pump) and "onboard" (in the vehicle) vapor recovery systems. 
MOBILE4.1 can model refueling emissions with no controls as well as
with either or both of the control options.

3.2.2.5.2 Inspection and Maintenance Programs

     Many areas of the country have implemented I/M programs as a
means of further reducing mobile source air pollution.  MOBILE4.1
can model the impact of an operating I/M program on the calculated
emission factors.  There is no average national I/M program; local
inputs must be supplied.  Details are given in Section 3.3.6.2 of
this document and in the MOBILE4.1 User's Guide.

3.2.2.5.3 Anti-Tampering Programs (ATPs)

     Some areas of the country have implemented these programs to
reduce the frequency and related emission impacts of emission
control system tampering.  MOBILE4.1 allows the effects of such a
program on the calculated emission factors to be estimated.  Due to
the wide variation in the characteristics of ATPs and the lack of a
national program, there is no national average estimate of ATP
parameters.  Details of the required inputs are given in Section
3.3.6.3 of this document and in the MOBILE4.1 User's Guide.

                                12





3.3  GUIDANCE ON SELECTING MOBILE4.1 INPUTS6

     MOBILE4.1 may be used to develop highway vehicle emission
factors and emission inventories for use in the State
Implementation Plan process.7  The proper version of MOBILE4.1 to
use in preparing SIP inventories is the one dated November 4, 1991. 
Older versions should be discarded or erased.8

     This section contains EPA's recommendations and suggestions
with regard to determining appropriate MOBILE4.1 inputs.  However,
for many inputs there is no single correct answer or recommendation
that is best for every local area.  For those using MOBILE4.1 for
SIP-related modeling purposes, it is important that the appropriate
EPA Regional Office personnel be kept involved in decisions
concerning questionable or controversial assumptions in the
MOBILE4.1 modeling and inventory development process.

     3.3.1     Emission Factors

     Description

     The basic emission rates (BERs) used in MOBILE4.1 are
expressed as linear equations and consist of a zero-mile level and
one or two deterioration rates.9  There are different BER
equations in MOBILE4.1 for each vehicle type/pollutant/model year
group, with the model year groups defined on the basis of
applicable emission standards and emission control technologies
used.

     Although MOBILE4.1 provides the capability to change the BER
equations, the BERs in MOBILE4.1 accurately reflect all promulgated
emission standards as of late 1990, and no locality-specific
changes to these equations are warranted for use in developing
emission factors or inventories for calendar years through 1992. 
Specifically, no need exists for modification of the BERs in
MOBILE4.1 in order to develop emission factors for the development
of base year 1990 emission inventories by the states in response to
the requirements of the Clean Air Act.
___________________________

     6  This section is in part a condensation of material that
appears in the User's Guide to MOBILE4.1, Chapter 2. It is not a
substitute for the User's Guide.  You are advised to obtain and
thoroughly read the User's Guide before running the model.  It is
available from the National Technical Information Service (NTIS),
5285 Port Royal Road, Springfield, VA 22161 (703/487-4650).  The
NTIS accession number is PB91-228759.

     7  Highway vehicle emission factors and emission inventories
for non-attainment areas in California may be developed using the
FISAFAC model.

     8  While future year inventories are discussed in this
document, MOBILE4.1 should not be used for projecting VOC or NOx
emissions beyond January 1, 1994, since it does not reflect new
standards that begin to have an effect after that date.  MOBILE4.1
may be used for CO inventory projections.  MOBILE5 will be released
in final form in August, 1992 and will allow VOC and NOx
projections.

     9  A deterioration rate is the gram per mile increase in
emissions per 10,000 miles accumulated mileage.

                                13





     Guidance

     No need exists for modifying the BERs in MOBILE4.1 in order to
develop VOC or NOx emission factors for any calendar year through
1992 inclusive,10 or to develop CO emission factors for any
calendar year through 2020.

3.3. 1.1 Region

     Description

     MOBILE4.1 provides two options for region: low-altitude and
high-altitude.  Low-altitude emission factors are based on
conditions representative of approximately 500 feet above mean sea
level (+500 ft MSL), and high-altitude factors are based on
conditions representative of approximately +5500 ft MSL. 
MOBILE4.1, like MOBILE4, does not calculate California emission
factors.  There have been no revisions to this variable or how it
is input to the model since the release of MOBILE4.

     Guidance

     For the majority of MOBILE4.1 applications, low-altitude is
the appropriate choice.  For those areas designated as high-
altitude by EPA for mobile source regulatory purposes, generally
those counties that lie "substantially" above +4000 ft MSL, high-
altitude should be selected.11

3.3.1.2   Calendar Year

     Description

     The value used for calendar year in MOBILE4.1 defines the year
(as of January 1) for which emission factors are to be calculated. 
It is frequently referred to as the calendar year of evaluation. 
MOBILE4.1 has the ability to model emission factors for the years
1960 through 2020 inclusive.  There have been no revisions to this
variable or how it is input to the model since the release of
MOBILE4.
___________________________

     10  EPA expects to update the model to version 5.0 to
incorporate all of the requirements of the November 1990 CAA in
time for states to project mobile source HC and NOx emissions and
demonstrate attainment of the National Ambient Air Quality Standard
for ozone.

     11  A list of those counties EPA has designated as high-
altitude appears in 86.088-30, paragraphs (a)(5)(ii) and (iv),
Code of Federal Regulations.

                                14






     Guidance

     The 1990 base year SIP inventories represent emissions during
a typical day in the pollutant season, most commonly summer for
ozone and winter for CO.  Thus, base year VOC inventories should be
based on interpolation of the calendar year 1990 and 1991 MOBILE4.1
emission factors.12, 13, 14

     CO SIP inventories should be based on emission factors from
January 1990 regardless of the three-month period for which CO is
being modeled.

     Similar instructions apply to the development of Reasonable
Further Progress (RFP) inventories.  For modeling of specific
episode days, the best results will be obtained by interpolating
exactly to the day being modeled.  In attainment demonstrations, it
is acceptable to account for fleet turnover through November 15th
of the year being modeled.

3.3.2     Fleet Characteristics

3.3.2.1   Vehicle Miles Traveled Mix by Vehicle Type

     Description

     The vehicle miles traveled mix specifies the fraction of total
highway VMT that is accumulated by each of the eight regulated
vehicle types.  The VMT mix is used in MOBILE4.1 only to calculate
the composite (all vehicle) emission factor for a given scenario on
the basis of the model's eight vehicle class-specific emission
factors.
___________________________

     12  For example, if most exceedances of the ozone National
Ambient Air Quality Ozone Standard occur during the months of June,
July, and August, then the appropriate base year emission factor is
the average of the January 1, 1990 and January 1, 1991 emission
factors.

     13  Since the accuracy gained by interpolating for typical
summer days other than July 1st is minimal and since the AIRSAMS
mainframe version of MOBILE4.1 for VOC and NOx inventories
automatically generates July I emission factors, EPA will accept
1990 VOC and NOx emissions estimates based on July 1st emission
factors.  Areas preparing draft 1990 inventories may select an
input of January 1, 1990 for their ozone season inventory and note
this prominently in the documentation.  However, the inventory must
be switched to a July 1, 1990 basis for the final submission to
EPA.

     14  January 1st and July 1st evaluations only differ in that
the July 1st vehicle fleet is composed of more of the latest model
year vehicles and fewer of the 25th and older model year vehicles,
as a result of new car sales and scrappage of older vehicles
between January I and June 30.  The January I vs.  July I choice is
independent of all temperature and other vehicle operating
conditions, which should represent the appropriate pollutant
season.

                                15





     MOBILE4.1 calculates a typical urban area VMT mix based on
national data characterizing model-year-specific registration
distributions and annual mileage accumulation rates by age for each
vehicle and fuel type,15 the fraction of travel by each vehicle
type that occurs in typical urban areas, and the total number of
vehicles of each vehicle type.

     For SIP-related highway vehicle emission inventory development
in moderate and above non-attainment areas, EPA expects states to
develop and use their own specific estimates of VMT by vehicle type
and highway functional system.16  VMT fractions based on local
estimates of VMT by vehicle type should be used as input to
MOBILE4.1.17

     Each VMT mix supplied as input must consist of a set of eight
fractional values, representing the fraction of total mobile source
VMT accumulated by each of the eight vehicle types.  All values
must be between zero and one, and the eight values must sum to 1.0.
There have been no revisions to how alternate VMT mixes are
supplied to the program as input data since the release of MOBILE4.

     Guidance

     Techniques for calculating estimated VMT by vehicle type (and
thus, total VW and the VMT mix fractions) from available data
sources are described in Chapter 6 of the report, "Techniques for
Estimating MOBILE2 Variables.18 Metropolitan Planning
Organizations (MPOs) and state Departments of Transportation (DOTs)
should also be consulted.  Information from these agencies can be
used to determine the proportion of passenger vehicles and light
duty trucks relative to heavy duty trucks by time of day and
facility class.  These two groups of vehicles can then be allocated
into the eight MOBILE4.1 vehicle classes using the national default
mix within each group provided by MOBILE4.1.

3.3.2.2   Annual Mileage Accumulation Rates and Registration
          Distributions by Vehicle Type and Age

     Description

     MOBILE4.1's emission factor calculations rely in part on
travel fractions for vehicles of each given age within a vehicle
type, which in turn are based on estimates of the average annual
mileage accumulation by age (first year to 25th-and-greater years
of operation) for
___________________________

     15  Total HDDV registrations and annual mileage accumulations
are also distributed within the model by truck weight class.

     16  Highway functional systems are commonly designated as the
interstate system, other freeways and expressways, other principal
arterials, minor arterials, and collectors.

     17  Ozone and carbon monoxide non-attainment areas classified
as marginal, submarginal, or transitional may use the MOBILE4.1
default VMT mix if no local estimate is readily available.

     18  Techniques for Estimating MOBILE2 Variables" and
"Additional Techniques for Estimating MOBILE2 Variables," Energy
and Environmental Analysis, Inc. for EPA (EPA Contract Number 68-
03-2888).

                                16





each of the eight vehicle types, and the registration distributions
by age (age 0-1 to age 24 and 25+) for each vehicle type, except
motorcycles, for which annual mileage accumulation rates and
registration distributions are only provided for the first to 12th-
and-later years of operation (ages 0-1 to 11 and 12+).  For all
vehicles except motorcycles, this represents an increase in detail
from the 20 years of operation used in MOBILE4.19

     To use locality-specific annual mileage accumulation rates by
age, a total of 200 input values is required: the estimated annual
mileage accumulated by vehicles of each of the eight types for each
of 25 ages.

     To use locality-specific registration distributions by age, a
total of 200 input values is also required.  For each vehicle type,
a set of 25 values is required to represent the fraction of all
vehicles of the given type that are of a given age.

     If both annual mileage accumulation rates by age and
registration distributions by age are supplied, the annual mileage
accumulation rate corresponding to any vehicle type/age combination
accounting for a non-zero fraction of registrations must be
positive.  That is, if vehicles of a certain type and age are
registered, then they are assumed to be driven.20

     If locality-specific mileage accumulation rates and/or
registration distributions by age are not used, the information in
MOBILE4.1 is used for all calendar years evaluated.

     Guidance

Mileage Accumulation

     EPA recommends, the use of the national annual mileage
accumulation rates by age that are included in MOBILE4. 1. Most
local sources of data on mileage accumulation rates by age are
subject to sampling bias or data entry errors, and the use of such
data should be approached with caution.  States that wish to use
alternate mileage accumulation rates in their development of
highway vehicle emission inventories in response to the
requirements of the new CAA should obtain prior approval from their
EPA Regional Office before using such rates in their emission
factor modeling.20
___________________________

     19  MOBILE4.0 modeled vehicles from age 0 through age 19 with
the 20th year representing all vehicles 20 years and older.

     20  MOBILE4.1 will issue an error warning if vehicles of a
certain type and age are registered but do not accumulate mileage. 
A warning will also be issued if there are no vehicles of a certain
type and age yet the mileage accumulation distribution includes a
positive value for that category.

     21  If local annual mileage accumulation rates are used, they
normally should not change from one evaluation year to the next.

                                17






 Registration Distributions

     EPA recommends and encourages the use of actual locality-
specific 1990 registration distributions by age to develop base
year SIP emission inventories.22, 23  Local registration
distributions are particularly appropriate for those inventory
areas where there are significant differences from the national
average.  The exception to the use of local data would be in those
areas that have relatively few local HDDV registrations, but that
experience significant interstate trucking activity.  Such areas
may want to retain and use the MOBILE4.1 national registration
distributions.

     EPA will issue at a later date additional guidance on how 1990
registration distributions by age can be adjusted to reflect future
years.  This guidance will provide a mathematical routine that
preserves the average age of the fleet in 1990, while retaining the
general shape of the local distribution for 1990 and earlier model
years.24, 25

     Methods for estimating the annual mileage accumulation rates
by age and the registration distributions by vehicle type and age
are presented in Chapters 2 and 3, respectively, of the report
Techniques for Estimating MOBILE2 Variables.26

3.3.2.3   Trip Length Distribution

     Description

     Running loss emissions are a form of evaporative volatile
organic compound (VOC) emissions that occur while the vehicle is
being operated.  Running loss emissions are different from
"traditional" evaporative emissions that occur after the vehicle
has been driven(hot soak evaporative emissions) and while it is
parked during periods of rising ambient temperatures (diurnal
evaporative emissions).  MOBILE4 was the first version of the
emission factor model to account for these emissions.  In MOBILE4.1
estimates of running loss emissions have been extensively revised.
___________________________

     22  Marginal, sub-marginal, and transitional non-attainment
areas may use the MOBILE4.1 distributions for all vehicle types, if
local distributions are not available.

     23  Registration distributions by age may be developed from
data available through state motor vehicle registration records,
either directly or commercially through R.L Polk Company.

     24  Effectively, the routine will apply a scrappage curve to
the existing 1990 registration distribution.  The result will be
that the pattern of high and low vehicle sales will propagate down
the Registration distribution as vehicles age with successive
evaluation years.

     25  EPA will not accept a locally developed registration
distribution that implies that the average age of the vehicle fleet
is becoming younger in the future than is reflected in the
registration distribution used for the base year unless the state
provides adequate justification for the new distribution .

     26  "Techniques for Estimating MOBILE2 Variables" and
"Additional Techniques for Estimating MOBILE2 Variables," Energy
and Environmental Analysis, Inc. for EPA (EPA Contract Number 68-
03-2888).

                                18





     EPA has determined through its running loss emission test
programs that the level of running loss emissions depends on
several variables: average vehicle speed, ambient temperature, fuel
volatility, and the length of the trip.27  Test data show that for
any given set of conditions (average speed, ambient temperature,
and fuel volatility), running loss emissions are zero to negligible
at first, but increase significantly as trip duration lengthens.

     In MOBILE4, running loss emissions were modeled as direct
functions of the input temperature and volatility; average speed
and trip duration were held constant within the model to values
representative of typical urban area traffic patterns.  In
MOBILE4.1 running loss emissions are modeled as a direct function
of the input temperature, fuel volatility, average speed and trip
duration.

     The input data record of the VMT-weighted trip duration
distribution must list the fraction of all travel (VMT) being
accumulated over the time period that the emission factors apply:

     -    Under 10 minutes
     -    11 to 20 minutes
     -    21 to 30 minutes
     -    31 to 40 minutes
     -    41 to 50 minutes
     -    51 minutes and longer

     Note that the first value should be the fraction of VMT that
occurs in trips that end within 10 minutes of their start, not the
fraction of VMT that occurs within 10 minutes of trip start for
longer trips.  The other values are defined similarly.  Note also
that the running loss emission factor that is calculated by
MOBILE4.1 is a fleet and area-wide average that applies to all of
the VMT in all of the trips for each vehicle type.  Any geographic
disaggregation by VMT density will be only approximate.  Situations
more heavily affected by emission rates at the end of trips, such
as a central business district in the morning rush hour, are more
complex to model.  EPA staff should be consulted in such cases.

     If this option of specifying trip length distributions is not
selected, then MOBILE4.1 will calculate the running loss emission
factors on the basis of the typical trip duration included in the
model.

     Guidance

     Since reliable local data on the distribution of trip
durations is often unavailable, EPA will accept the use of the
model's typical distributions for the estimation of running loss
VOC emission factors for the 1990 emission inventory.  Where the
transportation modeling process

___________________________

     27  "Length of trip" as used here refers to the duration of
the trip (how long, in minutes, the vehicle has been traveling),
not on the distance traveled in the trip (how far the vehicle has
been driven).

                                19





can produce reliable inputs for trip duration, use of such inputs
will produce a more accurate estimate of the benefits attributable
to SIP measures which shorten average trip lengths without
eliminating entire trips.28

     Most SIP inventories will be constructed by adding together
emission estimates for several functional classifications of
roadway.  EPA recommends that one area-wide trip length
distribution be used for all roadway classifications, due to the
complexity of trying to develop separate distributions.

3.3.2.4   Diesel Sales Fractions

     Description

     Sales of diesel powered light-duty vehicles and trucks
underwent a surge in the late 1970's and early 1980's, peaking at
5.9% of LDV sales in the 1981 model year, and at 9.3% of LDT sales
in the 1982 model year.  Since then diesel sales have fallen
precipitously, to virtually zero for LDVs29 and to about 0.2% of
LDTs since the 1988 model year.  While MOBILE4 contained forecasts
of increasing diesel sales for both LDVs and LDTs through the early
1990's, MOBILE4.1 assumes a more limited and slower increase from
the current, very low diesel sales rates.  MOBILE4.1 assumes that
future LDV diesel sales never exceed 0.3% and that future LDT sales
never exceed 2.15%.

     MOBILE4.1, like earlier versions of the model, uses a single
set of registration distributions by age and annual mileage
accumulation rates to describe all LDVs, and another set to
describe all LDTs.  This is due in part to the fact that it is
nearly impossible to develop such information for gas and diesel
LDVs and LDTs separately, and in part since there is little
evidence to suggest that typical use patterns and mileage
accumulation rates are different for gas and diesel LDVs and for
gas and diesel LDTs.

     Diesel sales fractions represent the share of all sales30 in
a given model year that are diesel-fueled vehicles.  The use of
model-year-specific diesel sales fractions allows MOBILE4.1 to
internally split the LDVs and LDTs into gas and diesel sub-
categories, which have distinctly different emission rates.

     If vehicle registration data that distinguish between gas and
diesel LDVs and gas and diesel LDTs exist, it is possible to input
local diesel sales fractions by model year.  These data must be
supplied for every calendar year of evaluation; since they apply to
vehicles of
___________________________

     28  The use of trip length distributions other than those
included in MOBILE4.1 should be adequately documented in the SIP.

     29  LDV diesel sales accounted for less than 0.05% of total
LDV sales in the 1988-1990 model years.

     30  Diesel sale fractions apply only to LDVs and LDTs.  Heavy
duty gasoline and diesel vehicles are treated separately within the
MOBILE models.

                                20





Ages 1, 2, 3, ..., to 25-and-older, different sets of fractions are
required for each calendar year.  Including this information will
create a more accurate highway mobile source emission inventory
estimate.

     For each scenario, the fractions of LDV and of LDT sales that
were diesel for each model year from the calendar year of
evaluation back to 25 model years ago must be entered as a model
input.  For example, if the calendar year of evaluation is 1990,
then diesel sales fractions for model year 1990, 1989, 1988, ...,
1967, and 1966-and-older LDVs and LDTs must be provided.  If two
different scenarios are being run, both for calendar year 1990 but
with other differences, then the same set of diesel fractions would
have to be supplied again as part of the second scenario.  If a
scenario with calendar year 1995 was also being run, then the
diesel sales fractions would represent model year 1995, 1994,...,
1972, and 1971-and-older vehicles.  The same values would be used
for the model years in common to the two sets of sales fractions,
but the five oldest model year values would not be used in the
second sequence to make room for the five most recent model years
sales fractions.

     The 50 diesel sales fractions, 25 each for LDVs and LDTs, must
be specified as fractions.  For example, if in a given area the
1990 model year had diesel sales of 1.1% of LDVs and 1.8% of LDTs,
the diesel sales fractions are 0.011 and 0.018 respectively.  The
values are supplied in pairs: The first two values on the first
record are the diesel sales fractions for one year old LDVs and
LDTs;31  the second two values are the sales fractions for two
year old LDVs and LDTs, and so on, with the last two values on the
third record being the sales fractions for LDVs and LDTs 25 years
and older.

     Guidance

     This option has been provided in MOBILE4.1 for two reasons. 
First, some users performing highway vehicle emission factor
modeling may have access to vehicle registration data, or data from
other sources, enabling them to characterize diesel sales of LDVs
and LDTs in the area being modeled.  Particularly if these sales
fractions differ significantly from those included in MOBILE4.1, it
will enhance the accuracy of the emission factors and inventory to
use those sales fractions as model input.  Second, as can be seen
by the sharp rise and equally sharp fall of diesel sales in the
late 1970's and early 1980's, it is extremely difficult to forecast
diesel sales fractions for future model years.  This provision will
allow modelers to account for future increases in diesel sales, if
such increases occur.32
___________________________

     31  A vehicle is assumed to be one year old if the model year
of that vehicle is the same as the evaluation year.  Thus, a 1990
model year vehicle is assumed to be one year old in 1990. 
Similarly, a 1989 model year vehicle is assumed to be two years old
in 1990.

     32  EPA does not envision any circumstances in which a state
or locality should substitute its own projection of future diesel
sales for that built into MOBILE4.1.

                                21





3.3.3 RVP Determination

Description

     The basic emission rates that underlie the emission factor
calculations are developed from vehicles tested at FTP conditions,
including a fuel volatility of 9 psi Reid Vapor Pressure (RVP). 
For other fuel volatility levels, MOBILE4.1 adjusts the emission
factors for exhaust and evaporative emissions as well as for
running loss, resting loss, and refueling loss emissions.

     Vehicle emission rates increase as the volatility of the fuel
increases, for temperatures between 45'F and 75'F and for RVP
values between 9.0 and 11.7 psi.  This effect is most pronounced at
higher RVP levels and at higher ambient temperatures.  Since there
is a significant interaction effect between RVP and temperature, it
is important that RVP and temperature inputs to MOBILE4.1 be
consistent.  That is, RVP and temperature should be chosen in such
a way that they represent the same time period.33  In general, use
July 1990 RVP levels to estimate VOC and CO emissions during the
ozone non-attainment season . Use January 1990 RVP levels to
estimate CO emissions during the CO non-attainment season.

Guidance

     Gasoline survey data should be used to determine historical
RVP, if quality-assured survey data are available.  The survey
samples should be drawn at the pump, not "upstream" of the pump at
a refinery or fuel distribution terminal.34, 35, 36
___________________________

     33  High RVP fuel is used in the winter months to facilitate
vehicle starting.  If the same high RVP fuel were used in the
summer, a vehicle could experience vapor lock and stall.

     34  EPA will also accept the use of RVP determined from
either of two regularly published gasoline volatility surveys, one
performed by oil companies and compiled by the National Institute
for Petroleum and Energy Research (NEPER) and the other sponsored
by the Motor Vehicle Manufacturers' Association (MVMA) and
conducted by the Southwest Research Institute (SwRI).  Since the
NIPER survey is not city-specific, the MVMA survey is the preferred
choice.

     35  A third survey is sponsored by a consortium of oil
companies, the American Petroleum Institute (API), and is also
conducted by Southwest Research Institute.  This survey includes
more cities and sampling months, but the data from it axe
proprietary.

     36  A final possible source of RVP data is the sampling done
by some states to enforce their state RVP limits.  However, before
using this approach, it should be discussed with EPA to determine
if RVP data collected for enforcement purposes are suitable for
determining average RVP for inventory purposes.

                                22





Procedure for Determining RVP Using the MVMA Survey

     Obtain the appropriate edition of the MVMA National Gasoline
Survey, published semi-annually37  by the Motor Vehicle
Manufacturers' Association.  Ordering and price information is
available from:

             Motor Vehicle Manufacturers' Association
                      300 New Center Building
                         Detroit, NE 48202
                       Phone (313) 872-4311

     Use the summer MVMA survey to estimate VOC and CO emissions
during the ozone non-attainment season.  Use the winter MVMA survey
to estimate CO emissions during the CO non-attainment season.

     If the average RVP for a specific city is desired and that
city is included in the MVMA survey, use the RVP for that city.38

     Find the average RVP value(s) for the city or cities selected
from the summary table that appears near the end of the MVMA
survey.  The average RVP for regular unleaded gasoline is provided
for all cities; the average RVP for premium and/or mid-grade
unleaded and/or regular leaded gasoline is also provided for many
cities.  Ignore RVP values for any ethanol blends that may also be
listed.39

     Calculate the overall average RVP from the averages supplied
for different grades of gasoline as follows:
___________________________

     37  The data reflected in MVMA National Gasoline Survey are
generally collected as of January 15th and
July 15th.

     38  If no city from the inventory area is included in the
MVMA survey, use the RVP for a city that is both geographically
close to the city with the largest population within the inventory
area and that was subject to the same EPA, state, or ASTM
volatility limit at the time.  If fuel distribution patterns are
known, give preference to a survey city with the same distribution
system.

     If the RVP for outlying areas of a state is desired (for
example, to complete the inventory for the fringes of an Airshed
modeling domain) and a city within that state is included in the
MVMA survey, use the RVP for that city.  If two or more cities in
that state are included in the MVMA survey, average the RVPs from
those cities.  If no city within the state is included in the MVMA
survey, use the RVP for a city that is both geographically close to
the state and that was subject to the same EPA, state, or ASTM
volatility limit at the time.  If fuel distribution patterns are
known, give preference to a survey city with the same distribution
system.

     39  As the use (and market share) of regular leaded fuel
continues to decline, survey values for mid-grade unleaded are
replacing those for regular leaded.

                                23





-    If only the average RVP for regular unleaded gasoline is
     provided, use that value;

-    If the average RVP for regular unleaded and one of the other
     fuel grades (premium unleaded, mid-grade unleaded, or regular
     leaded) is provided, weight the values using the local sales
     mix, if known, or at 75 percent regular unleaded and 25
     percent of the other grade for which the RVP is provided,
     according to equation 3-1.

Average RVP =  0.75 - (average RVP of regular unleaded) +
               0.25 - (average RVP of premium unleaded or mid-grade
               unleaded or regular leaded)
                                                              (3-1)

-    If the average RVP is provided for three fuel grades, weight
     the values, using the local sales mix, if known, or at 50
     percent regular unleaded, 25 percent premium unleaded, and 25
     percent mid-grade unleaded or regular leaded, according to
     equation 3-2.

Average RVP =  0.50 - (average RVP of regular unleaded) +
               0.25 - (average RVP of premium unleaded) +
               0.25 - (average RVP of mid-grade unleaded or regular
               leaded)
                                                              (3-2)

     If the average RVP is provided for all four fuel grades,
weight the values using the local sales mix, if known, or at 50
percent regular unleaded, 20 percent premium unleaded, 20 percent
mid-grade unleaded, and 10 percent regular leaded, according to
equation 3-3.

Average RVP =  0.50 - (average RVP of regular unleaded) +
               0.20 - (average RVP of premium unleaded) +
               0.20 - (average RVP of mid-grade unleaded) +
               0.10 - (average RVP of regular leaded)
                                                              (3-3)

The RVP thus calculated is used as the value of historical RVP in
MOBILE4.1.

Procedure for Determination of RVP Using the NIPER Survey

     Obtain the appropriate edition of the report Motor Gasolines,
published semi-annually by NIPER.  Samples for the summer survey
are taken in June, July, and August.  Samples for the winter survey
are taken in December, January, and February.

                                24






     The cost per report is $60, and it is available from:

               Cheryl L. Dickson
               National Institute for Petroleum and Energy Research
               P. 0. Box 2128
               Bartlesville, OK 74005
               Phone (918) 336-2400

     Use the summer NIPER survey to estimate VOC and CO emissions
during the ozone non-attainment season.  Use the winter NIPER
survey to estimate CO emissions during the CO non-attainment
season.

     The NIPER survey divides the country into seventeen districts,
which are described in a table and illustrated on a map of the U.S.
Use the district in which the inventory area is located.  If the
RVP for an entire state is desired and that state lies entirely
within one district, use that district.  If the state lies within
two or more districts, average the RVPs from the districts within
which the state lies.

     Table 4 of the NIPER survey presents the average RVP of three
grades of gasoline for each district: regular unleaded, regular
leaded, and premium unleaded.40

     Determine the overall average RVP of gasoline in a district by
weighting these three values by the local sales mix, or, in the
absence of local data, by an assumed sales mix of 50 percent
regular unleaded, 25 percent premium unleaded and 25 percent
regular leaded according to equation 3-4.

     Average RVP =  0.50 - (average RVP of regular unleaded) +
                    0.25 - (average RVP of premium unleaded) +
                    0.25 - (average RVP of regular leaded)
                                                              (3-4)

     The calculated RVP (or the average of the calculated RVPs, if
the area for which the RVP is being determined resides within two
or more districts) is then used as the value of historical RVP in
MOBILE4.1.

Procedure for Determining RVP from Applicable RVP Limit

     For an area without its own survey data and for which it is
not possible to use a city or district for which survey data exist,
RVP can be determined from the applicable RVP limit41 adjusted by
either a non-compliance margin or a compliance safety margin. 
Where ASTM
___________________________

     40  Ignore RVP values for any ethanol blends that may be
listed.

     41  Historically, fuel volatility was subject to voluntary
limits according to ASTM Standard D439, "Standard Specification for
Automotive Gasoline." More recently, fuel volatility has been
subject to federal and/or state regulatory requirements.

                                25





limits were the applicable limit, average RVP often exceeded the
ASTM limit (noncompliance margin) by an amount that varied with the
year and the ASTM limit itself.  On the other hand, gasoline
regulated by EPA or state RVP limits usually has had an average RVP
below the EPA or state ceiling.

     To estimate RVP using the "limit" approach, apply the
historical margin to the applicable limit according to equation 3-
5.

          RVP = Applicable Limit + Margin
                                                              (3-5)

where

     Applicable Limit =  Federal or state regulatory limit, or, if
                         none applies, ASTM standard for state and
                         month for which an inventory is being
                         estimated;

     Margin =  Non-Compliance Margin, if average RVP is greater
               than the applicable limit, or Compliance Safety
               Margin, if average RVP is less than the applicable
               limit.42


3.3.3.1 EPA-Provided 1990 RVP Estimates

     For states that do not wish to apply one of the above methods
themselves, EPA will post the 1990 RVPs recommended for use in
modeling mobile source HC, CO, and NOx emissions on a typical
summer day and the RVPs recommended for use in modeling mobile
source CO emissions on a typical winter day on the Chief Bulletin
Board System maintained by the Office of Air Quality Planning and
Standards.

3.3.3.2 "Period 1" RVP and "Period 2" RVP43

     MOBILE4.1 requires two RVP inputs, one for "period I" and one
for "period 2".  The purpose of having two RVP inputs is to allow a
step change in fuel volatility as of a specific calendar year.44
___________________________

     42  The non-compliance margin is always positive; the
compliance safety margin is always negative.

     43  "Period 1" RVP was called base or pre-control RVP in
MOBILE4; "period 2" RVP was called in-use RVP in MOBILE4.

     44  MOBILE4.1 assumes this change to occur as of January 1 of
the specified calendar year.


                                26






     The value to be used for the "period 1" RVP is the average in-
use RVP of gasoline in either the time before a volatility control
program took effect or the years preceding a change in the
controlled RVP level such as will take effect for most areas in
1992 when EPA's Phase I volatility control limits are superseded by
the Phase II volatility control limit.45  Period 1 RVP can be
between 7.0 psi and 15.2 psi inclusive.  "Period 2" RVF can be
between 6.5 and 15.2 psi inclusive.  The earliest allowed "period
2" start year is 1989.

     There have been no revisions in the input of these two
variables since the release of MOBILE4; only the names of these
variables have been changed.

3.3.3.3   Interpolation

     If emission factors are being calculated on a month-by-month
basis, or if the consecutive three-month period with the highest
frequency of NAAQS exceedance days occurring in the inventory area
is some period other than June, July, and August for ozone modeling
or November, December, and January for carbon monoxide modeling,
the RVP appropriate to each of the specific months being modeled
should be used.  The July RVP value may be used for the entire
period of the EPA RVP control program (May through September).  For
periods other than the period of EPA's control program, it is not
correct to average RVP values from different months or seasons
together, and it may be incorrect to use RVP from a time period
other than that used to determine the temperatures input to
MOBILE4.1.46, 47, 48

     3.3.3.4   Inputs for Future Year RVP

3.3.3.4.1 Future Summer RVP

     All parts of the United States will be subject to more
stringent EPA summer RVP limits beginning in 1992, with the highest
limit being 9 psi.  The 1992 summer survey data may not be
available in time to prepare draft or final inventories for 1996
and beyond.  Also,
___________________________

     45  To model the effects of the Federal volatility control
program promulgated by EPA, in which volatility is limited in the
summer months (May - September), see the relevant Federal Register
notices (54 FR 11868, March 22, 1989; 55 FR 23658, June 11, 1990),
or contact an EPA Regional Office to determine the applicable RVP
limits for a specific State and month.  The interim (Phase I)
controls were in effect during 1989, 1990, and 1991, and the final
(Phase H) controls took effect during 1992.

     46  The 1990 base year SIP inventories represent emissions
during a typical day in the pollutant season, most commonly summer
for ozone and winter for CO.  The procedure for choosing typical
day temperatures is described in section 3.3.5.2.

     47  For attainment demonstrations, states should use
temperature and RVP values that reflect the conditions of the
specific episodes being modeled.

     48  MOBILE4.1 does not model effects of RVP on emissions at
temperatures of less than 45' F (T Q, nor does it model effects of
RVP greater than an in-tank (weathered) level of 11.7 psi.  Under
summer temperature conditions an in-tank RVP of 11.7 psi
corresponds to a dispensed fuel RVP of approximately 12.5 psi.

                                27





since 1992 is the first year of the "Phase II" volatility
regulation, it may not be representative of long-term RVP levels. 
ASTM Class A and B areas will have a limit of 7.8 psi.  To forecast
future RVP, each area should determine its compliance safety margin
in 1990 and/or 1991 relative to the 1990 EPA or state limit
(ranging from 9.0 to 10.5). This margin should be subtracted from
the future EPA or state limit, if that limit is 9.0. Areas without
1990 or 1991 survey data should subtract a default compliance
safety margin of 0.3 psi.  Areas with a 1990/91 limit of 10.5 psi
and a safety margin significantly greater than 0.3 (which may have
been the result of distribution of fuel intended to comply with a
9.0 limit in nearby areas, or which may be the result of other
unique circumstances) should consult with EPA.  It may be
appropriate to use the 0.3 psi default, rather than the 1990/91
compliance safety margin for future years.  Areas with a 9.0 limit
in 1990/91 which also observed a safety margin significantly in
excess of 0.3 psi should also consult with EPA as to the
representativeness of the surveys involved.

     Areas with a future summer RVP limit of 7.8 should, in
general, not assume a safety margin, since RVP reductions below 7.8
psi are more costly than those below 9.0 psi, and loss of RVP
between refinery and pump will be lower.  However, if an area
subject to the 7.8 psi limit in 1992 determines that a safety
margin does exist in 1992 based on quality assured survey data, it
may request that EPA review 1992 summer MVMA and/or API survey data
from several cities to support its claim.

3.3.3.4.2 Future Winter RVP

     There are no plans for EPA to establish winter RVP limits.  If
there are no state standards for future winter RVP or if they are
the same as the state limit in 1990/91, the 1990/91 winter RVP
input should be used for future years.

     If a state is tightening its winter RVP limit from a limit
actively enforced in 1990/91, the 1990/91 RVP compliance safety
margin should be applied to the future limit.  If no margin was
applied in the 1990/91 inventory, none should be applied for the
future year.

     If a state is establishing a winter RVP limit where no limit
or only an ASTM limit (not backed by state law including active
enforcement) applied in 1990/91, the safety margin relative to the
new limit should be the limit calculated from the 1990/91 (or more
recent) summer survey.49
___________________________

     49  Alternatively, the default margin of 0.3 psi may be
assumed.

                                28





3.3.4 Oxygenated Fuels

Description

     MOBILE4.1 can model the effects of two types of oxygenated
fuels, gasoline/alcohol blends and gasoline/ether blends, on
exhaust carbon monoxide (CO) emissions50 provided that the
following information is input:

     -    Ether blend market share (as a fraction);
     -    Average oxygen content of ether blend fuels (percent
          weight, expressed as a fraction);
     -    Alcohol blend market share (as a fraction);
     -    Average oxygen content of alcohol blend fuels (percent
          weight, expressed as a fraction);
     -    RVP waiver. (If oxygenated fuels must meet the same RVP
          limits as gasoline, this indicator is set to 1; if such
          fuels have been granted a 1.0 psi waiver, this indicator
          is set to 2.51)

Guidance

     Areas that are known to have significant market
penetration52, 53 of ether blends and/or alcohol blends should
characterize the relative market shares and oxygen content of these
fuel blends and account for them in their mobile source emission
inventory.

     EPA should be contacted for assistance in modeling the effects
of oxygenated fuels if any of the following situations apply:
___________________________

     50  Reductions in exhaust CO emissions are estimated for
gasoline-fueled vehicle types (LDGV, LDGT1, LDGT2, HDGV, and MC). 
No effects on exhaust VOC or NOx emission factors or on any of the
evaporative components of VOC emissions are currently modeled with
the exception that, if the oxygenated fuels have a higher
volatility than base gasoline in an area, exhaust and evaporative
emissions will be increased to reflect the increased volatility of
the oxygenated fuels.  MOBILE5 will contain adjustments for exhaust
HC.

     51  Gasoline/ether blends axe assumed to have the same RVP as
gasoline, indicated by the regular RVP value input.

     52  If, together, ethanol blends account for less than 2% of
total gasoline sales within an inventory area, and if there is no
mandatory or locally endorsed voluntary program for ether blends,
oxygenated fuels need not be explicitly modeled for the 1990 base
year inventory.  Market shares for ethanol blends are readily
available by state.

     53  There have been no recent significant sales of other
oxygenated blend types (e.g., gasoline/methanol).

                                29





`    -    the fuels available in an area include blends containing
          both ether(s) and alcohol(s) in the same fuel ;
     -    an RVP waiver greater than 1.0 psi is applicable to
          oxygenated fuels in an area;
     -    no RVP waiver is in effect, but the volatility of base
          gasoline is currently below the regulated limit (in this
          situation, the practical effect may be same as if a
          waiver were in effect);
     -    if two or more types of alcohol blends are marketed under
          different RVP waiver treatment (for example,
          gasoline/methanol blends might not be given the same
          waiver as gasoline/ethanol blends).

3.3.5     Correction Factors

3.3.5.1 Speed

     Description

     There is considerable variation in vehicle emission factors as
average vehicle speed changes.54  In general, however, exhaust
emissions are at a minimum at about 48 mph.55  All emission rates
(VOC, CO, NOx ) display very high emissions at very low speeds,
with emissions decreasing (sharply at first and then more slowly)
as average speed increases, until minimum emissions are reached at
around 48 mph.  Above 48 mph, further increases in speed result in
increased emissions.

     MOBILE4.1 will calculate emission factors for average speeds
of 2.5 to 65.0 mph, in increments of 0.1 mph.56  One average speed
may be used for all vehicles, or a different average speed may be
used for each vehicle type.

     Guidance

     Selection of vehicle speeds is a difficult and complex
process.  Although it is appropriate for some purposes to use an
average speed for all vehicle trips and vehicle types within urban
areas as a whole, such an approach is not suitable for SEP
inventory preparation.  Instead, VMT should be left disaggregated
into subsets that have roughly equal speed, with separate VOC, CO,
and NOx emission factors for each subset.57  At a minimum, speeds
should be estimated separately by roadway functional class.
___________________________

     54  The speed correction factors in MOBILE4.1 are
substantially revised from those in MOBILE4.
     55  The average speed of the Highway Fuel Economy (HFE) test
cycle is approximately 48 mph.
     56  The maximum average vehicle speed allowed in MOBILE4.0
was 55 mph.
     57  Since emissions are a non-linear function of speed, with
significant curvature at low and high speeds, total daily area-wide
emissions are to some degree incorrectly estimated if VMT "events"
occurring at significantly different speeds are averaged together.

                                30






 Travel Demand Network Approach

     The recommended approach to estimating speeds is to post-
process the output from a local travel demand network model.58, 59 
Two documents that provide guidance on speed estimation for areas
using network models are: Highway Vehicle Speed Estimation
Procedures for Use in Emissions Inventories and A Study of Highway
Vehicle Emission Inventory Procedures in Selected Urban Areas.60

     The primary purpose of speed within a transportation planning
model is to allocate travel across the network.  It is used
primarily as a measure of impedance to travel rather than as a
prediction of accurate travel times.

     The report, Highway Vehicle Speed Estimation Procedures for
Use in Emissions Inventories, focuses on speed estimation methods
that are extensions of traffic assignment procedures.  The basic
method presented takes the link-specific traffic estimates provided
as an output by a UTPS-type highway assignment model and calculates
speeds based on the estimated highway volume-to-capacity ratios and
a set of speed formulas that are more specific to different road
types than the formula built into most assignment models.61  As
such, this method is not as simple as using the direct traffic
assignment output without modifications.62

     A second approach to estimating speeds using travel demand
network models is to use the output from traffic assignment
directly.63  If the network model assigns traffic to links on the
basis
___________________________

     58  Travel demand models that do not meet the performance and
validation requirements for use in forecasting VMT growth may
nevertheless be suitable for deriving speed estimates.  However,
reasonable efforts and success in validating the model are still
required.

     59  Once link-by-link speeds are determined from speed
formulas, the results may be aggregated into functional classes.

     60  Both documents were prepared for EPA by Cambridge
Systematics, Inc.  Much of the information contained in the
remainder of this section was taken directly from these documents.

     61  In most metropolitan areas, transportation planners
calibrate their highway assignment models to replicate observed
volume levels, treating highway speeds only as tools to obtain good
volume estimates rather than as critical outputs in their own
right.  In many cities, assignment-predicted speeds are too high to
match actual conditions; in some cities, they are too low.


     62  For those urban areas that can demonstrate that their
assignment-predicted link speeds closely match observed speed data
and/or speeds estimated using the Federal Highway Administration's
Highway Capacity Manual, the assignment-predicted link speeds may
be used directly in vehicle emissions inventories.

     63  If the year for which the inventory is being calculated
is not the same as one of the years for which the network model as
been run, speeds may be interpolated between chronologically
adjacent network model runs.

                                31





of a capacity restraint algorithm, then the associated link speeds
are likely to be more accurate than if another type of assignment
methodology is used.64, 65   However, the unique manner in which
the traffic assignment algorithm manipulates speed for a particular
link does not necessarily provide an accurate estimate of speed for
that link but rather provides a value that optimizes the traffic
assignment over the entire congested network.

Highway Performance Monitoring System (HPMS) Roadway Classification
Approach

     Post-processing with better speed formulas is often combined
with a direct link to the emission factor model, and link speeds
are either used directly as MOBILE inputs66 or grouped into ranges
based on the speed at which VMT occurs on each link.67

     One way to further reduce the number of MOBILE4.1 runs is to
use FHWA's Highway Performance Monitoring System (HPMS) roadway
classification scheme to group portions of VMT by the functional
classification of the roadways on which they occur.  This results
in 12 subsets of VMT.68  Within each subset, speed is weighted by
VMT to calculate an average speed and emission factor.

     This disaggregation of VMT by functional system avoids most of
the undesirable VMT averaging that might otherwise cause errors in
the emission inventory.  Further accuracy improvements can be
obtained by dividing the day into separate time periods so that
congested VMT and free-flowing VMT are not mixed.  While two
periods are the minimum split to get more homogeneity in vehicle
speeds, more than two periods are possible.  Each functional system
can, for example, be characterized by four average speeds during
distinct periods of the day: a morning peak period, a mid-day non-
peak period, an afternoon peak period, and a late evening/night
non-peak period.69, 70  Under this approach separate MOBILE4.1
emission factors are calculated for each
___________________________

     64  The capacity restraint method is a common type of traffic
assignment algorithm.  It is based on the inverse relationship
between speed and congestion.  It attempts to model congested
speeds during peak conditions for all facility types.  As
congestion increases, vehicle operating speeds decrease.  The
capacity restraint methodology is used as the default formula in
many urban areas' traffic assignment models.

     65  A potential problem with the use of any single function
is that it may not account well for the variations in traffic
operating conditions across all types of links, especially on very
congested links.  A single formula may be unable to accurately
estimate speed for facility types having very different operating
characteristics.  A more appropriate procedure would include
separate equations for estimating speed for each facility class for
each condition, i.e., peak versus off-peak travel.

     66  In this case, emission factors are developed for each
highway link and multiplied by the VMT on each link to calculate
link-specific vehicle emissions.

     67  Typically, a range would be one or two mph.  If this
approach is used, emission factors are calculated for the midpoint
of each speed range and multiplied by the associated VMT.

     68  FHWA designates roadway segments separately within urban
and rural areas into six functional classes each.

     69  AIRS/AMS is set up according to this approach, with up to
12 roadway classifications and one to four time periods within a
day.

                                32





period based on the speeds and temperatures prevailing during the
period.71  This approach has most of the advantages of link-
specific hour-by-hour modeling, but requires fewer MOBILE4.1
runs.72

     Estimating emissions for separate time periods within the day
also requires that particular attention be paid to the treatment of
the temperature inputs to MOBILE4.1. The sum of the emissions
within the four periods should be logically consistent (except for
the effect of the speeds) with that which would result from using
the 24-hour approach.  In order to achieve this consistency, the
24-hour minimum and maximum temperatures should be used to
determine diurnal evaporative emissions for each of the time
period-specific MOBILE4.1 runs.

     Ambient temperature, on the other hand, should be set to the
VMT-weighted average temperature of the period in question.73  For
example, if the night period extends from 7 pm to 6 am, the
temperature for each of the hours occurring during the night should
be weighted by the percent of night period VMT in each hour.74, 75

Highway Performance System National Estimates76

     If no network model is available, and in marginal and sub-
marginal non-attainment areas, the national speed estimates listed
in Table 3-1 may be used.  Individual areas may be able to obtain
comparable but locally specific speed estimates through their state
DOT or FHWA division office.  These speeds are calculated from HPMS
traffic counts and site-specific speed formulas and are not actual
speed observations.
___________________________

     70  The start and end times of periods should be locally
determined to reasonably separate higher from lower speed traffic
periods.

     71  Inputs other than Speed and temperature might also differ
by functional system and time of day.  The hot/cold mix of vehicle
operation is one example of such an input.

     72  This approach is most worthwhile when a significant
portion of the highway network gets much more congested during part
of the day, with considerable VMT in both the congested and non-
congested periods.

     73  The TEMFLG control flag should be set to accomplish this. 
See the User's Guide to MOBILE4. 1, section 2.1.14.

     74  The recommended method of apportioning daily VMT to
specific hours is to use the state's continuous monitors available
within the FAUA.  If no such monitors exist within the inventory
area, then the state may rely on other continuous monitors located
in areas similar in geographic, land use and demographic
characteristics, or on those areas' final Airshed Emission
Preprocessor profiles.

     75  Hour-by-hour temperatures should be determined from the
10-worst-days method used to determine the minimum and maximum
temperatures for inventory purposes.  See section 3.3.5.2.

     76  SOURCE: Federal Highway Administration, Highway
Performance Monitoring System, Impact Analysis for 1989 Base Year.

                                33





                             Table 3-1


Geographic            Roadway           Autos, Vans        All
  Area       Functional Classification    Pickups         Trucks

Rural     Interstates                        57.3           43.6
          Other principal arterials          45.4           36.0
          Minor arterials                    39.9           33.3
          Major collectors                   35.1           29.8
          Minor collectors                   30.5           24.4

Urban     Interstates                        46.3           39.0
          Other freeways and expressways     43.3           36.5
          Other principal arterials          18.9           16.0
          Minor arterials                    19.6           19.6
          Collectors                         19.6           16.4


Consistency Over Time

     Speed estimates for years other than 1990 must be logically
related to the 1990 methodology and estimates, with no arbitrary or
unsupported assumptions of speed changes.

3.3.5.2 Temperature

Description

     The basic emission rates that underlie the emission factor
calculations are developed from emission data from vehicles tested
at FTP conditions, including an ambient temperature of 75'F (24'C). 
MOBILE4.1 uses temperature correction factors to correct the
emission factors for other temperatures.

     MOBILE4.1 provides temperature correction factors for
temperatures in the range of 0'F (-18'C) to 110'F (43'C).  If a
temperature below 0'F is entered, a warning message is issued, and
0'F is used in the calculations.  Similarly, if a temperature above
110'F is entered, a warning message is also issued, and 110'F is
used in the calculations.

     The temperature used to adjust the exhaust emission factors
for all three pollutants, the hot soak component of evaporative
emissions, refueling emissions, and resting loss and running loss
emissions can be calculated on the basis of the input minimum and
maximum daily temperatures.77  Alternatively, the model can use a
single temperature that represents ambient conditions at a
particular time.
___________________________

     77  The maximum temperature must not be less than the minimum
temperature.

                                34





     However, even if a single temperature is used as the basis of
the temperature correction factors for all exhaust emissions, hot
soak evaporative emissions, refueling emissions, and resting loss
and running loss emissions,78 minimum and maximum daily
temperatures will still be used to calculate the diurnal component
of evaporative emissions.

     Minimum and Maximum Daily Temperatures

     Minimum and maximum daily temperatures are used directly in
MOBILE4.1 to calculate the diurnal portion of evaporative VOC
emissions79 and to estimate the temperature of dispensed fuel for
use in the calculating refueling emissions.  Unless overridden80
the temperatures used in calculating temperature corrections for
exhaust VOC, CO, and NOx emissions, the hot soak portion of
evaporative emissions, and resting loss and running loss VOC
emissions are also calculated by MOBILE4.1 based on the minimum and
maximum temperatures entered as input to the model.

     Since the basic exhaust emission rates for VOC, CO, and NOx
are based on the standard test temperature of 75'F (24'C),
MOBILE4.1 also adjusts these rates for other temperatures.  Using
the minimum and maximum daily temperatures and a representative
profile of temperature versus time of day, MOBILE4.1 first
calculates i temperature for each pollutant representing average
emissions over the course of the day and then adjusts the exhaust
emission factors for temperature effects accordingly.81

     Hot soak emissions at FTP conditions are based on a
temperature of 82'F (28'C).  Again using the minimum and maximum
temperatures, MOBILE4.1 calculates a temperature by which to adjust
hot soak emissions.
___________________________

     78  Logically, the single temperature used to represent a
typical day must be between a typical day's minimum and maximum
temperatures.

     79 Diurnal emissions are most frequently measured for a
temperature range of 68-86'F (20-30'C).  However, MOBILE4.1 adjusts
diurnal emission rates for the minimum and maximum temperatures
provided as input based on special EPA testing over additional
temperature ranges.

     80  A single ambient temperature can also be used to
determine the temperature corrections for exhaust VOC, CO, and NOx
emissions, hot soak evaporative emissions, dispensed fuel
temperature in the refueling emissions calculations, and resting
loss and running loss emissions, through the choice of a value for
the control flag TEMFLG (see the "User's Guide to MOBILE4.1,"
section 2.1.14). This approach is not recommended unless modeling a
short time period, such as an hour.  Refueling emissions should
always be modeled using the "full day" approach; hourly
temperatures should not be used.  Diurnal emissions can only be
modeled directly in MOBILE4.1 using the "full day" approach, since
the algorithm used is inaccurate over the very small temperature
rises (1' to 5'F) typical of a single hour.

     81  The algorithm used in MOBILE4.1 to determine temperatures
for correcting emissions on the basis of the input minimum and
maximum temperatures takes into account both the typical 24-hour
diurnal temperature profile for a day having the specified minimum
and maximum, and the typical distribution of travel over the course
of 24 hours.  Thus, the emission factors calculated in this way are
appropriately weighted for trips, vehicle miles traveled, and
emissions at different temperatures and result in factors that can
be multiplied by total daily VMT when total daily emissions are the
desired result.

                                35





     Resting loss and running loss VOC emissions are also dependent
on temperature.  As in the cases of exhaust and hot soak emissions,
MOBILE4.1 calculates appropriate average temperatures for
estimating resting loss and running loss emissions, weighted to
account for differing emission levels at different temperatures in
the range of the minimum and maximum daily temperatures and
differing travel fractions over the course of a day.  Restrictions
on these temperatures are: the maximum temperature must be greater
than or equal to the minimum temperature, and the ambient
temperature should be between the minimum and maximum (minimum 
ambient  maximum).

     There have been no revisions to this variable's use or input
data format requirements since the release of MOBILE4.

     Guidance

     EPA recommends that the minimum and maximum daily temperatures
be used to determine the temperatures for corrections to the
emission factors, if daily average, rather than hour-by-hour,
emissions are to be estimated.82, 83

     Minimum and maximum temperatures are normally calculated from
the most recent three-year period for which validated ozone and/or
CO monitoring data exists at the time the emission inventory is
due.  For 1990 inventories, the period to be used for temperature
determination should be 1988-1990.84
___________________________

     82  If hourly diurnal emissions are required for
photochemical or other models, an acceptable approach is to first
calculate the daily diurnal emissions, then allocate to specific
hours in proportion to the temperature rise per hour.  For example,
if during the modeling day the temperature increases from 60-84'F
within the 7 a.m. to 5 p.m. ten-hour period, and if during the 1-2
p.m hour temperature increases from 76-80'F then, assuming the
total diurnal emission factor is 3 g/vehicle, the emission factor
for the 1-2 p.m. hour is 0.5 g/vehicle.  The diurnal emission
factor for hours other than 5 am. to 3 p.m. is zero.  Other
reasonable methods may also be acceptable.  States wishing to use
another method should consult with EPA staff.

     Hourly Emission Factor = Hour-Specific Temperature
Increase/(Maximum Daily Temperature - Minimum Daily Temperature) -
Total Diurnal Emission Factor
                                                              (3-6)

     83  For CO modeling inventories, the recommended temperature
is the average of the 8-hour high concentration period temperatures
rather than the minimum and maximum temperatures used to calculate
a typical winter day inventory:



Click HERE for graphic.


     84  The temperatures used in the 1990 inventory must also be
used for all projection inventories.

                                36





Procedure

     Determine the consecutive three-month period with the highest
frequency of NAAQS exceedance days occurring in the inventory
area.85 The same consecutive three-month period applies for each
year, with a total of nine months used to determine temperature. 
If the months containing the highest frequency of exceedances are
not consecutive, or if two or more sets of consecutive months have
the same frequency, use the months of June, July, and August for
ozone modeling and the months of November, December, and January
for carbon monoxide modeling.

     Next, list the 10 highest concentrations86 that occurred in
the inventory area during those nine months and the dates of those
concentrations.87
     The ten highest ozone concentrations for each site in a county
and the dates on which they occurred are contained in the
Aerometric Information Retrieval System (AIRS) AMP440/Maximum
Values Report.  Eight-hour average CO concentrations and the dates
on which they occurred can be found in the AIRS AMP350 raw data
report.  The AMP355/Standard Report contains the CO values that
exceed the NAAQS.  These reports are available from EPA's National
Air Data Branch.  Be sure to specify the year(s) and counties of
interest and indicate that the request is for preparation of a SIP
emission inventory to avoid being charged the normal processing
fee.  To obtain copies of these reports, contact Tom Link, U.S. EPA
Office of Air Quality Planning and Standards, at (919) 541-5456.

     Determine the maximum and minimum temperatures for each of the
10 days for the area being inventoried.  This information is
contained in the Local Climatological Data Monthly Summary for the
inventory area and is available from:

               National Climatic Data Center
               Federal Building
               Asheville, NC 28801-2696
               Telephone: (704) 259-0682

     Maximum and minimum daily temperatures are located in columns
2 and 3, respectively, on page 1 of the Summary.
___________________________

     85  Consider the three-year period as a whole when making
this determination.
     86  The 10 highest concentrations need not all be
exceedences.
     87  There are four exceptions to selecting the 10 highest
concentrations: 1) More than 10 concentrations may be needed to
identify 10 unique dates. 2) If the 10th ranked concentration level
occurs on more than one day, all of those days should be included
in the temperature calculation. 3) If only two years of validated
monitoring data exist for the entire MSA monitoring network, use
the seven highest concentrations. 4) If only one year of data
exists for the entire MSA monitoring network, use the four highest
values.

                                37





     If there is more than one meteorological station within the
non-attainment area, use the station that best represents the
conditions of most mobile source emissions within the inventory
area.

     The temperatures recommended for use in modeling mobile source
HC, CO, and NOx emissions on a typical summer day and the
temperatures recommended for use in modeling mobile source CO
emissions on a typical winter day will be posted on the Chief
Bulletin Board System maintained by the Office of Air Quality
Planning and Standards.

3.3.5.3   Operating Modes

     Description

     One important determininant of emissions performance is the
mode of vehicle operation.  EPA's emission factors are based on
testing over the FIT cycle, which is divided into three driving
segments (referred to as "bags"), each with differing associated
emissions performance:

     Bag                  Operating Mode

     1         Cold start - first 505 seconds of a cold-start trip
               (or less, if the trip ends before 505 seconds);

     2         Stabilized - all operation beyond 505 seconds of a
               trip;

     3         Hot start - first 505 seconds of a hot-start trip
               (or less, if the trip ends before 505 seconds).

     Emission data from each of these modes reflect the fact that
emissions generally are highest when a vehicle is first started,
i.e., is operating in the cold-start mode.  At that time, the
vehicle, engine, and emission control equipment (particularly the
catalytic converter and oxygen sensor) are all at ambient
temperature and, therefore, are not performing at optimum levels.

     The hot start mode represents the case of a vehicle that was
operating, was turned off, and then was restarted.  In this case
the vehicle was not turned off for sufficient time to have cooled
completely to ambient temperatures.  Since the vehicle is partially
warmed up, emissions during a hot start are generally lower than
during a cold start.  However, since the vehicle is not yet
completely warmed up, emissions are generally higher than when the
vehicle is completely warmed up and operating in what is known as
stabilized mode.  During stabilized mode the vehicle has been in
continuous operation long enough for all emission control systems
to have attained relatively stable, fully "warmed-up" operating
temperatures.

                                38





     MOBILE4.1 uses three inputs to indicate vehicle operating
mode.88  These inputs represent VMT. that is, the percentage of
VMT` (not the percentage of vehicles) accumulated by non-catalyst
vehicles in cold-start mode (PCCN), by catalyst-equipped vehicles
in hot-start mode (PCHQ, and by catalyst-equipped vehicles in cold-
start mode (PCCC).  The three specified values must all be
expressed as percentages (not as fractions).  Each value must lie
between 0.0% and 100.0%, and the sum of PCHC + PCCC must not exceed
100%.  There have been no revisions in the definitions or in the
use or format requirements of these variables since the release of
MOBILE4.

     Guidance

     Historically EPA has defined cold starts to be any start that
occurs at least four hours after the end of the preceding trip for
non-catalyst vehicles and at least one hour after the end of the
preceding trip for catalyst-equipped vehicles.  Hot starts are
those starts that occur less than four hours after the end of the
preceding trip for non-catalyst vehicles and less than one hour
after the end of the preceding trip for catalyst-equipped vehicles. 
The shorter time interval associated with the cold/hot start
definition for catalyst-equipped vehicles reflects the fact that
catalytic converters do not operate at intended efficiency until
they are fully warmed up.89

     In the absence of supporting data for values other than those
listed above, EPA believes that the values reflecting FTP
conditions are appropriate.  This is particularly true when the
emission factors being modeled are individually or collectively
intended to represent a broad geographic area (Metropolitan
Statistical Area, entire state) and/or a wide time period (days,
months).  When the modeling is intended to represent highly
localized conditions (specific highway links) or very limited
periods of time (single hours), it may be possible to develop more
representative values for these variables.90  Areas known to have
average
___________________________

     88  The values corresponding to the FTP cycle are:
                         PCCN 20.6 %
                         PCHC 27.3 %
                         PCCC 20.6 %

     89  Catalysts begin to operate at full efficiency once they
reach about 600'F (316'C).  Since non-catalyst vehicles do not
depend on attainment of such high temperatures for stabilization of
emissions performance, they can remain partially warmed up for up
to four hours.

     90  Some transportation emissions modeling approaches are
based on the concepts of trip-start emissions and running
emissions, rather than the method described above.  In this
alternative approach, trip-start emissions are assumed to be
instantaneous and are calculated as the difference between
MOBILE4.1 total "start" emissions and total "stabilized" emissions. 
Total start emissions per trip are the product of the 100% cold-
(or hot-) start emission factor in grams per mile and the 3.59-mile
distance attributed to the first 505 seconds of the FTP driving
cycle.  Total stabilized emissions are the product of the 100%
stabilized emission factor in grams per mile and the same 3.59-mile
distance and same speed as that used to estimate start emissions. 
Start emissions are calculated as the grains per trip event times
trip productions, and are typically located at the centroids.

     Total stabilized emissions are the product of the sum of the
MOBILE4.1 stabilized exhaust and running loss emission factors in
grams per mile and the total distance traveled during the course of
the trip.

                                39





lengths significantly shorter or longer than 7.5 miles may also
merit the use of alternate values.91

     Thus for SIP-related modeling, EPA will accept the use of the
FTP operating mode values except for small-scale scenarios where
their use would clearly be inappropriate.  EPA will not accept SIP-
related modeling that includes different operating mode fractions
for the base and projection years without a clear demonstration
that such a shift is warranted.

3.3.5.4   Additional Correction Factors for Light-Duty Gasoline-
          Fueled Vehicle Types Description

     MOBILE4.1 can provide four additional corrections to the
exhaust emission factors for LDGVs, LDGT1s, and LDGT2s.  These
corrections are used to represent unique conditions not typically
assumed in MOBILE4.1 runs, and include the emissions effect of air
conditioning (A/C) usage, extra loading, and trailer towing.  There
is also a humidity correction factor, which applies only to exhaust
NOx emissions.92

     If these corrections are to be applied, either six or ten
inputs will be required.  If six values are required, they are an
A/C usage fraction (for all LDGVs and LDGTs), three extra load
usage fractions (for LDGVs, LDGT1s, LDGT2s), a trader towing
fraction (for all LDGVs and LDGTs), and a humidity level (for all
LDGVs and LDGTs plus motorcycles).  If ten values are required,
they are an A/C usage fraction (for all LDGVs and LDGTs), three
___________________________

     Trip end emissions are simply the "hot soak" emissions
expressed in grams per trip event times trip attractions.  Trip end
emissions are also typically located at centroids.

     Daily diurnal emissions are calculated as the emissions in
grams per vehicle times vehicles present and must also be assigned
to specific hours and locations.  Methods for this assignment vary.

     if the transportation model produces reliable estimates for
the relevant trip parameters, this alternative method is believed
to be a more accurate way to locate mobile source emissions within
the inventory area than is the use of average FTP hot/cold
percentages.

     Recently it has been reported that one of the commercially
available transportation models was modified to distribute start
VMT along individual links based on travel time.  Since there is
considerable variation among vehicles in the time interval required
for catalyst light-off, EPA recommends that those using this
approach for their 1990 SIP submittal evenly distribute "start"
emissions on the basis of the first 505 seconds of the FTP rather
than try to estimate the exact amount of time within that period
required for catalyst light-off, As more information becomes
available, it may be possible at a later time to meld the
instantaneous and distributed approaches to locating "start"
emissions.

     91  The driving cycle used in FTP testing is 7.5 miles long. 
If shorter trips are preponderant within a given area, it is
possible that the percentage of VMT occurring in one of the "start"
modes is greater than the national average.  Similarly, if longer
trips predominate, the percentage of VMT occurring in the
"stabilized" mode may be greater.

     92  The humidity correction factor is also applied to NOx
emissions from motorcycles.

                                40





extra load usage fractions (for LDGVs, LDGT1s, LDGT2s), three
trailer towing fractions (for LDGVs, LDGT1s, LDGT2s), a humidity
level (for all LDGVs and LDGTs plus motorcycles), and dry bulb and
wet bulb temperatures (used to calculate an A/C usage fraction for
LDGVs and LDGTs).

A/C Usage Fraction

     In the six-input option, a correction factor for A/C usage
will not be applied, regardless of the value that is entered. 
Enter 0.0 in this case.  If you wish to include the effect on the
exhaust emission factors of A/C usage, enter a non-zero fractional
value for this variable and appropriate dry and wet bulb
temperatures, as explained below.

     In the ten-input option, this variable acts as a flag, and the
A/C usage fraction is calculated on the basis of the dry bulb and
wet bulb temperatures (see below).  If 0.0 is entered for A/C, no
correction factor will be applied, although values of dry and wet
bulb temperature must still be provided if the ten-input option has
been chosen.

Extra Load Usage Fractions

     These values are used to model the exhaust emissions effect of
vehicles carrying an extra 500 lb (227 kg) load.  To include this
effect, three fractional values are entered (one each for LDGVs,
LDGT1s, and LDGT2s), representing the fraction of all vehicles of
the given type carrying such an extra load.  These inputs are
restricted to the range of zero to one.  If the value entered is
zero, no correction for the effects of extra load is applied.

Trailer Towing Usage Fraction

     These inputs are used to modify exhaust emissions of vehicles
towing trailers.  Enter one or three values that represent the
fraction of vehicles of a given type that are to be assumed to be
towing trailers.  These inputs are also restricted to the range of
zero to one.  If the value entered is zero, no correction for the
effect of trailer towing is applied.  In the six-input option, one
value is entered and is applied to LDGVs, LDGT1s, and LDGT2s.  In
the ten-input option, three values are entered, and one each is
applied to LDGVs, LDGT1s, and LDGT2s.

NOx Humidity Correction

     This input is used to correct exhaust NOx emission factors
for absolute humidity.  The value entered is the absolute
(specific) humidity, expressed as grains of water per pound of dry
air.  Absolute humidity is restricted to the range 20 to 140.93
___________________________


     93  A value of 75 corresponds to the absolute humidity
condition of the FTP.  If 75 is entered as the input, then no
correction will be applied.

                                41





Dry and Wet Bulb Temperatures

     MOBILE4.1 will estimate the fraction of AX-equipped vehicles
with the air conditioning operating on the basis of a "discomfort
index".94 The discomfort index is calculated from the dry bulb and
wet bulb temperatures, which are restricted to the range 0'F (-
18'C) to 110'F (43'C).  In addition, the wet bulb temperature must
be less than or equal to the dry bulb temperatures95

     There have been no revisions to any of the variables discussed
in this section, or to how they are supplied to the model as input
data, since the release of MOBILE4.

     Guidance

     In most cases, ozone pollution episodes occur during summer
months and very warm to hot temperatures.  It is reasonable to
assume that vehicle air conditioning usage is high under such
conditions.  The air conditioning correction factors that are
calculated in MOBILE4.1 will increase vehicle emissions, and areas
that believe their motor vehicle inventory has been underestimated
in the past may choose to use them.  However, EPA will accept SIP
inventories that do not attempt to explicitly account for vehicle
air conditioning use.96

     The same approach that is taken in developing the base year
inventory must also be used for projection inventories.

     The humidity correction for NOx emissions accounts for the
fact that when "excess" water vapor is present, some of the heat of
combustion heats water vapor rather than enhancing NOx formation. 
As with the air conditioning correction, EPA will accept SIP
inventories that do not attempt to account for local humidity.  If
the humidity correction is applied in the base year, it must also
be used in any projection inventories.  While the humidity
correction factors were developed in the late 1970's, limited
testing on current technology vehicles indicates that they are
still adequate.
___________________________

     94 These values (in 'F) will be used to calculate the A/C
usage fraction on the basis of the discomfort index only if the ten
input option is selected and a non-zero value is entered for the
variable AC.  If used, this calculated value overrides the value
read in for AC, which serves as a flag indicating that this
correction is desired (see above).

     95  If any of these three conditions is not met, MOBILE4.1
will print an error message.

     96  There is some uncertainty surrounding the air
conditioning correction factors.  The emissions effect of operating
the air conditioner for late model year vehicles is not well
quantified.  Also, the fraction of vehicles equipped with air
conditioning (built into MOBILE4.1) is substantially higher for the
vehicle fleet of the late 1980's than it was for the fleet of the
late 1970's, which magnifies the consequence of a possible error. 
Thus, the use of the air conditioning corrections to emissions is
acceptable but not required in the development of SIP inventories.

                                42





3.3.6     Control Programs

     In general, VMT should be disaggregated so that vehicles
generating the VMT are subject to a common control program.97 
This issue applies especially for I/M versus no I/M and AT`P versus
no ATP areas, specifically interstate areas but also where the
inventory area is partially designated attainment and partially
designated non-attainment.  EPA will accept the use of a single
correct pair of VMT` fractions (with and without I/M) for the
entire inventory area, even if these fractions may be incorrect
within one county or a state portion of a non-attainment area.98

3.3.6.1   Refueling Emissions

     Description

     The refueling of gasoline-fueled vehicles results in the
displacement of fuel vapor from the vehicle fuel tank to the
atmosphere.  There are two basic approaches to the control of
vehicle refueling emissions, generally referred to as "Stage H" (at
the pump) and "onboard" (on the vehicle) vapor recovery systems
(VRS).  MOBILE4.1 has the ability to model uncontrolled levels of
refueling emissions (i.e., assuming no requirements for Stage If or
onboard VRS systems) as well as the effects of the implementation
of either or both of the major types of vapor recovery systems.

     There are five approaches available in MOBILE4.1 for modeling
vehicle refueling emissions:

     -    Model uncontrolled refueling emissions for all gasoline-
          fueled vehicle types;
     -    Model refueling emissions assuming a Stage II VRS
          requirement;
     -    Model refueling emissions assuming an onboard VRS
          requirement;
     -    Model refueling emissions assuming both Stage II and
          onboard VRS requirements;
     -    Account for refueling emissions by some means other than
          MOBILE4.1.

     No additional inputs are required for either the first or the
last approach.  Additional information is needed, however, to model
the effects of either or both VRS requirements on refueling
emissions.
___________________________

     97  VMT should also be disaggregated so that the set of
operating conditions under which the vehicles are generating the
VMT is fairly homogeneous.

     98  For example, if it is difficult to tell which of two
states' vehicles produce the VMT on each side of the border, both
states could assume that the VMT on their side comes from a random
mix of their collective vehicle population

                                43





     Four inputs must be provided to model the effect of a Stage II
VRS requirement: the start year (calendar year in which the
requirement takes effect), the phase-in period (number of years for
Stage III VRS installation to be completed), and the system
efficiency (in percent) at controlling refueling emissions from
light-duty vehicles and trucks, and from heavy-duty vehicles.99

     Only two inputs are required to model the effect of an onboard
VRS requirement: the starting model year and which of the four
possible vehicle types (LDGV, LDGT1, LDGT2, HDGV) are subject to
the requirement.100

     Both sets of inputs must be supplied to model both VRS
requirements concurrently.

Guidance

     EPA recommends that states and others use MOBILE4.1 to model
refueling emissions for highway vehicle emission inventories.  The
refueling emission factors can be calculated in grams per gallon of
dispensed fuel (g/gal) or in grams per mile (g/mi).  The preferred
approach is to calculate g/gal refueling emission factors that
reflect the applicable Stage II requirements, then multiply the
g/gal emission factor by total gasoline sales.  This is the most
accurate method of estimating the contribution of refueling
emissions to the inventory, particularly for areas with reliable
data on gasoline sales.101  This method also accounts for
refueling emissions generated when gasoline is purchased in an area
but consumed largely outside of the area, and does not include
refueling emissions for through traffic that does not refuel in the
area.102 When good data on gasoline sales is not available, the
use of the g/mi refueling emission factor is more convenient and,
while also more approximate, acceptable for SIP inventory
development.

Stage II

     The overall effectiveness of a Stage II vapor recovery system
at controlling refueling emissions depends on a number of factors,
including the baseline efficiency of the system used, the portion
of total area gasoline consumption handled by service stations
exempt from Stage II requirements, and the frequency and stringency
of enforcement programs.  In general, the effectiveness of a Stage
II VRS at controlling refueling emissions will be greater for
light-duty vehicles and trucks than for heavy-duty vehicles, since
HDGVs are more likely to
___________________________

     99  There are no national average or default values for Stage
II efficiency.

     100  EPA has no plans to promulgate regulations for onboard
vapor recovery.

     101  State tax revenue receipts on county gasoline sales are
often used for this purpose.

     102  One alternative to using MOBILE4.1 to calculate
refueling emissions is to use the applicable AP-42 emission
factors.  However, the effects of an onboard vapor recovery system
requirement cannot be modeled accurately using this approach. 
MOBILE4.1 makes use of improved predictive equations to calculate
refueling emission factors, and these have not yet been
incorporated into AP-42.

                                44





be refueled at service stations (or other fuel dispensing
locations, such as private refueling depots) that will be exempted
from Stage II requirements.

     EPA has estimated the in-use efficiency of Stage II programs
based on the dispensed volume of the stations exempted from the
requirement and the frequency of inspections at stations subject to
it.  The 1990 Clean Air Act exempts from the Stage II requirement
stations that sell less than 10,000 gallons of gasoline per month
(50,000 gallons per month for independent small marketers, as
defined in the Act).  This exemption level, along with a semi-
annual inspection frequency, results in 83% in-use efficiency for a
Stage II program.  If inspections occur annually, efficiency is
estimated to be 77%.  Minimal inspections reduce the efficiency to
56%.  EPA will accepts these efficiency levels as MOBILE4.1 inputs
when modeling Stage II controls as part of the 1990 SIP submittal.

Spillage

     Emissions from fuel spillage also can be modeled using
MOBILE4.1. The "baseline" spillage factor (assuming no controls) is
0.31 g/gal. of dispensed fuel.  If no controls are assumed, this
factor is added to MOBILE4.1's calculated displacement loss, which,
in turn, is based on ambient temperature and RVP.  If Stage 11 is
modeled, then the in-use efficiency percentages input determine the
reduction in both displacement and spillage.

     Displacement loss can be separated from fuel spillage, using
the grams per gallon emission factors from the expanded evaporative
output.  If no controls are used, fuel displacement emissions
should be calculated by subtracting the 0.31 g/gal spillage factor
from the emission factor.  In this case, spillage would be 0.31
g/gal.

     To model the effects of Stage II on displacement and spillage
losses, the percent emission reduction for LDGVs, LDGTs, and HDGVs
must be input.  These percentages are then multiplied by the 0.31
g/gal spillage factor to arrive at a value for spillage. 
Displacement loss is then calculated as the gram/gallon emission
factor minus the calculated spillage loss.

3.3.6.2   Inspection and Maintenance Programs

     Description

     Many areas of the country have implemented inspection and
maintenance (I/M) programs as a means of reducing mobile source air
pollution.  MOBILE4.1 can model the effect of an operating I/M
program, based on the specification of certain parameters that
describe the program.  Standard low-altitude area emission
reduction credits are contained within the MOBILE4.1 code itself,
while standard high-altitude area emission credits are included as
a separate file on the MOBILE4.1 diskettes and tapes.103
___________________________

     103  The model can also accept alternate credit matrices as
input.  These must be developed by EPA.  Areas for which the
standard emission reduction credit matrices are inappropriate
should contact the Office of Mobile Sources (Air Quality Analysis
Branch, 313/668-4325) to obtain the required matrices.

                                45





     The following program parameters must be specified in order to
model an I/M program:

     -    Program start year (calendar year that program begins);
          Stringency level (percent);
     -    First (earliest) and last (latest) model years of
          vehicles subject to the requirements of the program;
     -    Waiver rates (percent of failed vehicles; one rate
          applicable to pre-model year 1981 vehicles and one rate
          applicable to 1981 and later model year vehicles);
     -    Compliance rate (percent);
     -    Program type (centralized; decentralized and
          computerized; or decentralized and manual);
     -    Frequency of inspection (annual or biennial);
     -    Vehicle types covered by the program;
     -    Test type (idle, 2500/idle, loaded/idle);
     -    Whether alternate I/M credits are to be supplied;
     -    IM240 transient test first model year;
     -    Purge system test first model year;
     -    Pressure system test first model year.

     The last three parameters in the list above are optional. 
They refer to the earliest model year of vehicles subject to:

     -    transient testing of HC and CO emissions (where the
          vehicle is tested on a chassis dynamometer over a
          transient driving cycle, and mass emissions are measured
          using a constant volume sampling (CVS) system);
     -    functional purge testing of the evaporative emission
          control system;
     -    functional pressure testing of the evaporative emission
          control system.

     While these three types of testing may be incorporated into
many future I/M programs, there were no areas using any of these
tests as part of their I/M programs in 1990.  Thus, these three
parameters104 should not be included in estimating the 1990 base
year emissions inventory.

     The following sections discuss the terminology used to
describe I/M programs for purposes of modeling the emission
benefits of such programs using MOBILE4.1. In general, MOBILE4.1
assumes that the I/M program is mandatory, periodic, and covers a
well-defined group of vehicles.105
___________________________

     104  MOBILE4.1 will run correctly if an I/M program is
specified and only the first ten parameters from the above list are
provided.  Additional information on the IM240 and purge and
pressure testing can be found in the "User's Guide to MOBILE4.1"
(sections 2.2.5.4,2A.1.16, 2A.1.17, and 2A.1.18).


     105  There are many details, such as instrument
specifications, that are beyond the scope of this document to treat
I/M program planners should consult with EPA (Emission Control
Strategies Branch, 313/668-4476) if they have further questions
regarding program requirements.

                                46





3.3.6.2.1 I/M

     I/M refers to "inspection and maintenance" programs, which are
inspections of vehicles using a measurement of tailpipe emissions
and which require that all vehicles with tailpipe emissions higher
than the program cutpoints be repaired to pass a tailpipe emission
retest.  Inspections that are aimed at verifying the presence and
proper connection of emission control devices are called anti-
tampering programs.106

3.3.6.2.2 Start Year

     The year in which the periodic inspection program begins to
require both inspection and repairs is called the start year. 
MOBILE4.1 only provides for a January 1st start date.  Other start
dates will require interpolation between two MOBILE4.1 runs to give
accurate estimates of benefits.  Separate start dates may be
entered for the tailpipe emissions check and anti-tampering
portions of an I/M program.

3.3.6.2.3 Stringency

     Stringency is the tailpipe emission test failure rate expected
in an I/M program among pre-1981 model year passenger cars or pre-
1984 light-duty trucks, based on the short test107 emission
cutpoints.108 The expected failure rate can be determined by
applying the program cutpoints to a representative sample of
vehicles tested in a survey.  Actual failure rates reported by a
program can also be used to determine stringency, but only when
there is no possibility of significant testing or data reporting
errors.  MOBILE4.1 assumes that the failure rate remains fixed at
the stringency level for each evaluation year.  MOBILE4.1 will not
allow a stringency level less than 10% or greater than 50%.

3.3.6.2.4 First Model Year

     The first model year refers to the oldest model year vehicle
that is always included in the inspection program.  MOBILE4.1
assumes that all vehicle classes have the same model year coverage
and does not allow for a separate coverage period for each vehicle
class.  Some programs do not fix the model years covered by the
program, and instead use a coverage it window" to define those
vehicles that must be inspected.  For example, such a program may
___________________________

     106  Such tailpipe I/M and anti-tampering programs are
sometimes referred to collectively simply as 'W programs" in other
EPA documents.

     107  A tailpipe emission short test is any one of several
emission testing procedures authorized by EPA for use in emission
testing programs.  They include a simple idle test, where emissions
are sampled as the vehicle idles in neutral gear, to more elaborate
tests, such as the IM240, where the vehicle's emissions are sampled
during a simulated drive using a chassis dynamometer.

     108  Emission cutpoints are the emission level measurements
used to determine whether a vehicle passes or fails the short test. 
If the vehicle's tailpipe emission level exceeds the cutpoints set
for any of the measured pollutants, the vehicle fails.

                                47





cover only vehicles 15 years old or younger.  Since model year
cohorts that leave the I/M program in this way retain a partial I/M
influence for a period of some uncertain number of years beyond
their last inspection, such programs cannot be modeled exactly
using MOBILE4.1. However, as a practical matter, EPA is willing to
assume that a vehicle that has left the I/M program as of a certain
evaluation year was never inspected.109

3.3.6.2.5 Last Model Year

     The last model year refers to the youngest (newest) model year
vehicle that is subject to the inspection program.  The combination
of first and last model year inputs makes it possible to model a
program that covers only particular model years.  Many programs
routinely include new model year vehicles in the program as they
reach their one year anniversary.  In such cases the year 2020
should be designated as the last model year.110 If inspection is
delayed until vehicles are two or three years old, then the input
value for the last model year will be different for each evaluation
year.111

     3.3.6.2.6 Waiver Rates

     Many I/M programs waive the requirement to pass a retest ff
certain defined criteria are met.  Typically, waivers are granted
for vehicles whose owners have spent more than a certain dollar
amount repairing the vehicle in an attempt to pass the test.

     The waiver rate inputs to MOBILE4.1 reduce the estimated
benefit of the program.  The waiver rates are always calculated as
a percent of non-duplicate initial test failures.  Waiver rates
must be provided for pre-1981 and for 1981 and later model year
light duty Vehicles.112
___________________________

     109  An alternate approach is to add one extra model year to
the coverage window to represent all model years still experiencing
some residual but declining IN benefit.  EPA will accept this
approach for purposes of estimating 1990 emissions.

     110  MOBILE4.1 assumes, in the calculation of I/M credits,
that vehicles less than one year old are exempt from inspection.

     111  Programs may include delayed inspections because a state
considers such a program to be more cost effective, since emissions
of new vehicles are generally very low.

     112  MOBILE4.1 assumes that tampered or misfueled vehicles
cannot receive waivers, and so does not reduce the ATP benefit
based on the waiver rate.

                                48





     Guidance

     For an historical inventory, the actual waiver rate must be
determined and used as the input to MOBILE4.1. For future year
inventories, the historical waiver rate should be used unless a
change will be made to the criteria for granting waivers.113

3.3.6.2.7 Compliance Rate

     Compliance rate refers to the level of compliance with the
inspection program.114  For example, assume a program required
that all passenger cars be inspected each year, and that 100,000
passenger cars were registered in the area covered by the program. 
If in a given year only 95,000 passenger cars completed the
inspection process to the point of receiving a final certificate of
compliance or a waiver, the remaining 5,000 vehicles may have
avoided the inspection requirement.  If those vehicles did, in
fact, avoid the inspection requirement, the compliance rate for the
program would then be 95%.115

     MOBILE4.1 uses a single compliance rate to reduce both the I/M
and ATP portions of the program benefits.  The reduction in benefit
is not linear.  The benefit loss per vehicle assumes that the
failure rate among non-complying vehicles will be larger than the
expected failure rate in the fleet.  As the rate of non-compliance
increases, the non-complying failure rate will approach and finally
equal the expected failure rate.

  Table 3-2 shows the loss of benefit assumed for the enforcement
fraction:
___________________________

     113  If tighter criteria are planned, a lower waiver rate may
be assumed.  For areas subject to the requirements for Enhanced
I/M, including the $450 expenditure requirement, a future waiver
rate as low as one percent may be assumed, but planners should
realize that an underestimation of the future waiver rate may cause
later problems in demonstrating reasonable further progress
milestones.  EPA may also consider a finding of SIP non-
implementation if an actual waiver rate substantially exceeds the
rate assumed in inventory forecasts.

     114  The compliance rate input is also used to account for
vehicles that are waived from compliance without any testing.  For
example, vehicles with special testing problems or vehicles owned
by certain groups of individuals may be automatically waived.

     115  Other possible reasons for the 5000 vehicle discrepancy
include vehicles registered but scrapped or transferred out-of-
state prior to the inspection due date.

                                49





                             Table 3-2

             Benefit Assumed for Enforcement Fraction


               Non-   Non-Complier      Fraction       Fraction
Compliance Compliance Failure Rate       Benefit        Benefit
   Rate        Rate    Adjustment         Loss         Remaining

   100%         0%       2.0            .000           1.000
    99%         1%       2.0            .020            .980
    98%         2%       2.0            .040            .960
    97%         3%       2.0            .060            .940
    96%         4%       2.0            .080            .920
    95%         5%       1.5            .095            .905
    90%        10%       1.4            .169            .831
    85%        15%       1.3            .238            .762
    80%        20%       1.2            .302            .698
    75%        25%       1.1            .361            .639
    70%        30%       1.0            .415            .585
    50%        50%       1.0            .615            .385


Guidance

     Historical compliance should be determined by sticker surveys,
license plate surveys, or a comparison of the number of final tests
to the number of vehicles subject to the I/M requirement.116
Planners should not assume a compliance rate of 100%.  An area with
a registration denial system using automatically generated
compliance documents that uniquely identify the complying vehicle
and that are serially numbered and accounted for, that rely on
centralized processing by government clerks with management
oversight, may assume a 98% rate unless there is evidence to
indicate otherwise.117

3.3.6.2.8 Inspection Frequency

     MOBILE4.1 allows for two inspection frequencies.  "Annual"
means that all covered vehicles must be inspected once each year. 
"Biennial" means that each vehicle is inspected once every two
years, such that either half of the vehicles of each model year are
inspected
___________________________

     116  The number of initial inspections should not be used to
calculate the compliance rate, since some cars may drop out after
failing one or more tests.

     117  An overestimation of the future compliance rate may
cause problems later on in demonstrating that Reasonable Further
Progress milestones have been met.  Also, EPA may issue a finding
of SIP non-implementation, if the actual compliance rate is
substantially less than the rate assumed in inventory forecasts.

                                50





each calendar year, or vehicles of one model year are inspected in
alternate calendar years.  Any other inspection frequency would
require alternate I/M credits provided by EPA.

3.3.6.2.9 Vehicle Classes

     MOBILE4.1 program benefits are calculated separately for each
gasoline-fueled vehicle class.  No emission benefits are estimated
for diesel vehicles or motorcycles.  The vehicle class designations
are based on the same definitions under which vehicles are
certified: 118

     -    LDGV - light-duty gasoline-fueled vehicles (passenger
          cars);
     -    LDGT1 - light-duty gasoline-fueled trucks less than 6000
          lbs gross vehicle weight (lighter pick-up trucks and
          vans);
     -    LDGT2 - light-duty gasoline-fueled trucks greater than
          6000 lbs but less than 8,500 IN GVW (heavier pick-up
          trucks and vans and many commercial trucks);
     -    HDGV - heavy-duty gasoline-fueled vehicles greater than
          8500 lbs GVW (heavier commercial trucks, including
          highway hauling trucks).

3.3.6.2.10 I/M Test Types

     There are four I/M tailpipe test types allowed in MOBILE4.1.
These test types only apply to the inspection of 1981 and newer
model year passenger cars and 1984 and newer light-duty trucks.119
The chosen test type is assumed to be applied to all 1981 and newer
passenger cars and 1984 and newer light-duty trucks both at the
initial inspection and at the retest.120

Idle Test

     The idle test is a measurement of VOC and CO emission
concentrations of a fully warmed vehicle as it idles in neutral or
park.

2500/Idle Test

     The 2500/idle test is a measurement of VOC and CO emission
concentrations of a fully warmed vehicle operating at 2500 rpm,
first, in neutral or park and second, at idle.  The vehicle must
pass both 2500 rpm and idle tests.
___________________________

     118  Those areas that do not use the same vehicle class
designations in their vehicle registration data as are used in
MOBILE4.1 must take care not to claim coverage for too many
vehicles.

     119  The concept of stringency already takes into account the
effect of test type on the benefits that accrue to older vehicles.

     120  In all cases MOBILE4.1 assumes that the cutpoints used
for the inspections are 1.2% CO and 220 ppm VOC.

                                51





Loaded/Idle Test

     The loaded/idle test is a measurement of VOC and CO emission
concentrations of a fully warmed vehicle operated first, on a
chassis dynamometer at a constant cruise speed and dynamometer
load, and second, at idle in neutral or park.  The vehicle must
pass both the cruise and idle tests.

IM240 Transient Test

     The IM240 transient test is a measurement of VOC and CO
emission concentrations of a vehicle operated over a range of
speeds.  The IM240 was patterned after the Urban Dynamometer
Driving Schedule used to conduct the Federal Test Procedure (FTP). 
In order for MOBILE4.1 to estimate the effects of using the IM240
test in an I/M program, a set of alternative I/M credits must be
entered as input.121  MOBILE4.1 assumes that transient testing is
applied to all model years and to all vehicle types covered by the
I/M program.

Purge Test

     The purge test is a functional test of the purge capabilities
of the evaporative emission control system.  The flow rate of
canister purge is measured during transient operation of a vehicle
on a chassis dynamometer using a flow measurement device, and
cutpoints for minimum flow rate are used to determine if the
vehicle passes or fails.  Vehicles failing the purge test are
required to have repairs performed to enable the vehicle to pass
the test.  MOBILE4.1 assumes that all model years and vehicle types
subject to the I/M program are also subject to functional purge
testing, if a first model year for such testing is specified as an
input.

Pressure Test

     The pressure test is a functional test of the evaporative
emission control system for leaks.  The fuel tank and related hoses
and pipes are pressurized, and the pressure loss is monitored over
time.  Cutpoints defining the maximum allowable loss of pressure
are set and used to determine if the vehicle passes or fails. 
Vehicles failing the pressure test are required to have repairs
performed to enable the vehicle to pass the test.  MOBILE4.1
assumes that all model years and vehicle types subject to the I/M
program are also subject to functional pressure testing if, a first
model year for such testing is specified as an input.
___________________________

     121  These are provided on diskettes and tapes distributed
with the model.

                                52





3.3.6.2.11 Alternate I/M Credits

     In special cases where the design of an I/M program does not
fit into any of the categories defined in MOBILE4.1,122 an
alternative set of credits may be input.  These credits are then
used by the model to determine the benefits of the I/M program.123

     One set of alternate I/M credits applies to purge and pressure
system checks of the evaporative control system.  To estimate these
credits MOBILE4.1 requires that the oldest model year covered by
the purge system check and the oldest model year covered by the
pressure check be entered as inputs.  MOBILE4.1 assumes that all
vehicle types covered by the I/M program are also covered by the
purge and pressure system checks.  It also assumes that the purge
system check is done as part of a transient exhaust emission check.

3.3.6.2.12 Centralized Programs

     Centralized inspection programs refer to those programs that
completely separate vehicle inspection from vehicle repair. 
Usually, high-volume inspection stations, run either by the local
agency itself or by a contractor, perform all initial tests and
retests after repair.  Garages and other repair facilities are not
allowed to perform official tests.  Independent centralized
programs are the standard used to determine the emission benefits
for I/M and ATP program designs.124

3.3.6.2.13 Decentralized Programs (Manual)

     Decentralized inspection programs refer to those programs
where the local program agency licenses service stations and
garages to perform official inspections, and re-inspections.  These
licensed inspection stations are allowed to perform repairs on the
vehicles they inspect.  The number of licensed inspection stations
in decentralized programs is larger, and the volume per station is
smaller than for centralized programs.
___________________________

     122  Examples of such programs are those with a semi-annual
or tri-annual inspection frequency.

     123  Normally these actors will be supplied by EPA at the
request of the program manager or air quality planner.

     124  Test-only programs are considered centralized even if
they are decentralized in the sense of multiple businesses.

                                53





     Decentralized program have been found to be less effective in
reducing emissions than are centralized programs.  As a result,
MOBILE4.1 reduces the emission benefits of a decentralized program
to 50% of that attributed to a centralized design for both the
tailpipe test portion and the ATP portion of the program.125 
Modelers who can demonstrate a higher level of effectiveness for a
decentralized I/M program should contact EPA.126

3.3.6.2.14 Computerized Inspection

     Some decentralized I/M programs require the use of
computerized emission analyzers.  These analyzers contain small
computers which keep track of all official inspection activity,
automatically calibrate the instrumentation, and prompt the
inspector during the inspection procedure.  The computer also
prepares a machine-readable record of all official inspections and
calibrations, and will not allow inspections whenever it determines
that the instrumentation is out of calibration.

     MOBILE4.1 assumes that the W portion of a decentralized
computerized inspection program will be 50% as effective as a
centralized program of similar stringency (i.e., the benefits of
the program are discounted by 50%).  This benefit discount is the
same as for manual decentralized programs.127  As noted above,
this benefit reduction includes the effect of waivers, if any, and
is not applied on top of a waiver-related loss of potential
benefits.128

     Decentralized computerized inspection programs will also have
some of the benefits from the ATP portion of the program reduced by
50%.129  Modelers who can demonstrate a higher level of
effectiveness for a decentralized I/M program should contact
EPA.130
___________________________

     125  The 50% reduction in benefits from the tailpipe portion
of the test includes the loss due to waivers, if any.  For
decentralized I/M programs, the waiver rate input of the model is
disabled so that user input of waiver rates has no effect on I/M
benefits.

     126  Since the degree to which manual decentralized programs
are less effective than centralized programs is not a MOBILE4.1
input, it is necessary for EPA to prepare a special version of the
model to account for the effects of a decentralized I/M program
that has been shown to have an effectiveness level different from
that of a centralized program by more or less than 50%.  The
demonstration of increased effectiveness should rely on historical
data and not on arguments for an anticipated increase in
effectiveness.

     127  Although the computerized analyzers make it easier for
inspections to be performed correctly, EPA audits have shown that
the overall performance of inspectors in computerized decentralized
I/M programs is no better than in manual decentralized I/M
programs.

     128  For decentralized I/M programs, the waiver rate input of
the model is disabled so that even if waiver rates are input to
MOBILE4.1, they will have no effect on I/M benefit calculations.

     129  Most of the deterrence effect of the program, which
deters tampering that has not yet occurred, is unaffected by the
decentralized discount.

     130 Since the degree to which computerized decentralized
programs are less effective than centralized programs is also not a
MOBILE4.1 input, it is necessary for EPA to prepare a special
version of the model to account for the effects of a decentralized
VIA program that has been shown to have an effectiveness level
different from that of a centralized program by something other
than 50%.  The demonstration of increased

                                54






3.3.6.2.15 Tech I-II and Tech IV+

     The calculation of I/M benefits for MOBILE4.1 was done by
technology group, which can roughly be determined by model year for
each vehicle type.  These technology groups have come to be
referred to by numbers.  The table below summarizes the technology
groupings used in MOBILE4.1 and their respective application to
gasoline-fueled passenger cars and light trucks.  Within the Tech
IV group, there are separate I/M credits for each model year of
LDGVs, and a mapping of LDGT model years to similar technology LDGV
model years.131

                        Model Years Covered

          Technology
          Grouping       LDGV      LDGT1     LDGT2

               I         Pre-1975  Pre-1975  Pre-1979

               II        1975-80   1975-83   1979-83

               IV+       1981+     1984+     1984+


     3.3.6.3   Anti-Tampering Programs

Description

     Some areas of the country have implemented anti-tampering
programs (ATPs) to reduce the frequency and resulting emission
effect of emission control tampering (e.g., misfueling, removal or
disablement of catalytic converters).  MOBILE4.1 can estimate the
emission factor effects of such programs.  The following inputs are
required to model an anti-tampering program:
___________________________

effectiveness should rely on historical data and not on arguments
for an anticipated increase in effectiveness.

     131  Sets of alternate I/M credits may contain both Tech I
and II credits, only Tech IV+ credits, or Tech I and II and Tech
IV+ credits together.  This is usually indicated in the header
block of the alternate I/M credit deck.

                                55






     -    Start year (calendar year in which the program begins);
     -    First (earliest) and last (most recent) model years of
          vehicles subject to the program;
     -    Vehicle types covered by the program;
     -    Program type (centralized or decentralized);
     -    Frequency of inspection (annual or biennial);
     -    Compliance rate (percent);
     -    Inspections performed (air system, catalyst, fuel inlet
          restrictor, tailpipe lead deposit test, EGR system,
          evaporative system, PCV, gas cap).132

     There have been no revisions to the information required to
model ATP effects since the release of MOBILE4.

     The following sections discuss the terminology used to
describe anti-tampering program inspections for purposes of
modeling the emission benefits of such programs using MOBILE4.1. In
general, it is assumed that the program is mandatory, periodic, and
covers a well-defined group of vehicles.

     It is also assumed that the inspections are primarily visual
rather than functional and involve no disassembly or disconnection
to gain access to hidden components (other than removal of the gas
cap to view the fuel inlet restrictor).  However, program
regulation writers are encouraged to define failure in broad enough
terms of visual damage and proper operating condition so that any
emission control component determined by the inspector to be non-
functional can be properly failed and repaired.

     A program that inspects for tampering only when a vehicle has
failed its tailpipe I/M inspection, or only when a vehicle owner
requests a test waiver, is not considered as an anti-tampering
program, gainers no emissions benefits, and should not be modeled
in MOBILE4.1.

     The following sections discuss the terminology used to
describe anti-tampering programs for purposes of modeling the
emission benefits of such programs using MOBILE4.1. In general,
MOBILE4.1 assumes that the anti-tampering program is mandatory,
periodic, and covers a well-defined group of vehicles.133
___________________________

     132  MOBILE4.1 will only model an ATP with an evaporative
system inspection and provide appropriate emission credits if a gas
cap inspection is also included.  If an evaporative system
inspection is indicated, but a gas cap inspection is not indicated,
MOBILE4.1 will issue a warning message, and no emission credit will
be given for the evaporative system inspection.  However, the
converse is not true.  A gas cap inspection may be indicated
without an indication of an evaporative system inspection.


     133  There are many details (such as replacement catalyst
specifications) that are beyond the scope of this discussion. 
Program planners should consult with EPA's Office of Mobile Sources
(Emission Control Strategies Branch, 313/668-4476) if there are
questions regarding the requirements of ATP inspections.

                                56





3.3.6.3.1 ATP

     Anti-tampering programs are periodic inspections of vehicles
to detect damage to, disablement of, or removal of emission control
components.  Owners are required to restore the vehicle's emission
control system and have the vehicle reinspected.  Programs that
inspect for tampering only those vehicles failing an I/M tailpipe
test are not considered to have an anti-tampering program and
should not be included in estimating a mobile source emissions
inventory.

3.3.6.3.2 Tampering and Misfueling

     Any physical damage to, or disablement or removal of, an
emission control component is considered tampering in MOBILE4.1.
This does not limit tampering only to deliberate disablements or
only to those disablements of which the vehicle owner is aware. 
Tampering, therefore, can often be a result of poor maintenance
rather than some deliberate action by the vehicle owner or service
mechanic.

     Misfueling is the use of leaded fuel in any vehicle that is
equipped with a catalytic converter.  This includes inadvertent use
of leaded fuel without the knowledge of the vehicle owner.

3.3.6.3.3 Air Pump Inspection

     Air pump systems supply fresh air needed by the catalytic
converter to reduce engine emissions before they leave the
tailpipe.  Inspectors should check for missing belts and hoses and
proper connection at the exhaust manifold.  Sometimes the entire
pump and its plumbing are removed.  A valve is sometimes used to
route air away from the exhaust stream during certain operating
modes.  This valve should be checked for proper hose and wire
connections.  Often the air is injected directly into the catalytic
converter underneath the vehicle.  If so, this connection should be
checked.  Any missing, damaged, or altered components of the air
pump system should be replaced.

3.3.6.3.4 Catalyst Inspection

     The catalytic converter, sometimes referred to simply as the
catalyst, oxidizes excess volatile organic compounds and carbon
monoxide from the engine exhaust into water and carbon dioxide. 
Newer catalysts also reduce oxides of nitrogen in the exhaust.  The
metals that accomplish this task are most commonly coated on a
ceramic honeycomb inside the stainless steel shell of the catalyst. 
The catalyst resembles a muffler in some ways, but would not be
confused with a muffler because it is farther forward on the
vehicle, and its stainless steel shell will not rust.

                                57





     Some cars will have more than one catalyst, so the number of
catalysts expected should be determined before the inspection
begins.  Some catalysts are located very near the exhaust manifold,
so the inspector should be sure to check the entire length of the
exhaust piping from the exhaust manifold to the muffler before
determining that the catalyst is not present.

     Emission credit should not be claimed using MOBILE4.1 unless
regulations provide a mechanism to ensure that failed cars are
correctly repaired with original equipment manufacturer (OEM) or
approved aftermarket replacements.  Program planners should consult
with EPA to avoid incorrectly claiming credit.

3.3.6.3.5 Fuel Inlet Restrictor Inspection

     Vehicles requiring the use of only unleaded gasoline have been
equipped with fuel inlets that only allow use of narrow fuel
nozzles.  Leaded fuel is required to be dispensed only from pumps
using wider nozzles.  Any vehicle found to have a fuel inlet which
allows a leaded fuel nozzle to be inserted, such as having the
nozzle size restriction. removed, is assumed to have used leaded
fuel.  Leaded fuel permanently reduces the ability of the catalytic
converter to reduce emissions.  Therefore, vehicles found with a
fuel inlet that allows insertion of a leaded fuel nozzle should be
required to replace the catalytic converter.  In addition, the
vehicle's fuel inlet should be repaired to allow only the insertion
of unleaded fuel nozzles.

     Repair of the fuel inlet restrictor only is not considered a
repair that will reduce the emissions of the vehicle.  Since the
damage to the emission control of the vehicle occurs in the
catalyst, it is the catalyst that must be replaced to result in any
substantial emission reduction.  The inlet restrictor must be
replaced simply as protection for the new catalyst.  If the program
regulators do not require catalyst replacement, the MOBILE4.1
inputs should indicate that an inlet check is not performed.134

3.3.6.3.6 Tailpipe Lead Detection Test

     Leaded fuel permanently reduces the ability of the catalytic
converter to reduce engine emissions before they leave the
tailpipe.  Therefore, vehicles found to have used leaded fuel
should be required to replace the catalytic converter.  EPA has
allowed for the use of a lead detection test in the vehicle
tailpipe as a method to detect the use of leaded fuel.  Since this
is a chemical test, care must be taken to ensure that the test is
properly conducted and that the results are properly interpreted.

     Vehicles with evidence of lead deposits in the tailpipe have
used leaded fuel.  Since the damage to the emission control of the
vehicle occurs in the catalyst, it is the catalyst that

___________________________

     134  MOBILE4.1 assumes that inspectors are not allowed to
skip this inspection because the fuel inlet is concealed by a
locked door.

                                58





must be replaced to result in any substantial emission reduction. 
If the program regulators do not require catalyst replacement, then
the input to MOBILE4.1 should indicate that a tailpipe lead test
was not preformed.135  Anti-tampering programs that require
failure of both the fuel inlet restrictor inspection and the
tailpipe lead detection test before requiring replacement of the
catalyst get credit for neither.  The input to MOBILE4.1 should not
indicate either inspection was performed.

3.3.6.3.7 EGR Inspection

     The exhaust gas recirculation (EGR) system reduces oxides of
nitrogen by routing some of the exhaust back into the intake
manifold.  Although the primary component of the system is the
valve that controls the flow between the exhaust and intake
manifolds, most systems are quite complex, with various sensors and
valves which together control the operation of the entire system. 
Any system observed with missing or damaged components or misrouted
or plugged hoses and wires should be failed and repaired.136 
While MOBILE4.1 has an input flag for EGR inspections, there are no
emission reductions associated with them.  This reflects EPA's
assessment that these difficult visual inspections are virtually
never performed correctly and result in virtually no repairs. 
Programs with evidence to the contrary should consult EPA.

3.3.6.3.8 Evaporative Control System

     The evaporative control system collects gasoline vapors from
the gas tank and carburetor bowl and stores them in a charcoal
canister.  During certain engine operations, the canister purges,
releasing the vapors, which are then routed to the engine to be
burned.  In addition to the evaporative canister itself, the system
includes varying numbers of hoses, wires, and control valves.137  
Any system observed with missing or damaged components or misrouted
or visually obvious plugged hoses and wires should be failed and
repaired.  This inspection flag should not be used to indicate any
functional pressure or purge testing of the evaporative emission
control system.  The benefits of such tests are calculated
separately.138  Purge and pressure testing is, however, assumed to
detect all evaporative control system tampering.
___________________________

     135  The tailpipe as well as the catalyst should also be
replaced to avoid failing a subsequent inspection test.

     136  Hoses may be plugged, either deliberately or by neglect.

     137  Hoses may be plugged, either deliberately or by neglect

     138  Modeling the benefits of functionally testing purge
and/or pressure evaporative systems requires that the initial    
model year of vehicles subject to such tests be input.  See section
3.3.6.2.10.

                                59





3.3.6.3.9 PCV Inspection

     The positive crankcase ventilation (PCV) system routes the
vapors from the crankcase to the intake manifold where they can be
burned by the engine.  The PCV system has two major loops.  The
most critical connects the crankcase to the throttle or the intake
manifold via a hose and usually contains a valve.  Another hose
connects the crankcase to the air cleaner to provide the crankcase
with filtered fresh air.  Any system observed with damaged or
missing components or with hoses misrouted or plugged should be
failed and repaired.

3.3.6.3.10 Gas Cap Inspection

     Gas caps are actually part of the evaporative control system. 
Without a properly operating gas cap, fuel vapors from the gas tank
would escape.  On some vehicles, a missing gas cap will also cause
the evaporative system canister to purge incorrectly.  Inspectors
should examine the fuel inlet area of each vehicle to determine
that the gas cap is present.  If not, the vehicle should be failed
and the gas cap replaced.139  Pressure testing is assumed to
detect all missing gas caps.

3.3.6.3.11 Tampering Rates

     Description

     MOBILE4.1 calculates tampering rates as a piece wise linear
function of accumulated mileage for each gasoline-fueled vehicle
type140 and for eight types of tampering.141  These rates am
combined with the corresponding fractions of vehicles equipped with
the given control technology and emissions rates to obtain the
tampering offsets.142  These offsets are later added to the non-
tampered emission factors.

     MOBILE4.1 uses tampering rates based on EPA Office of Mobile
Sources (OMS) analysis of multi-city tampering survey results.  EPA
recommends that the tampering rates included within MOBILE4.1 be
used.143
___________________________

     139  MOBILE4.4 assumes that inspectors are not allowed to
skip this inspection even if the fuel inlet is concealed by a
locked door.

     140  The four vehicle types are LDGV, LDGT1, LDGT2, and HDGV.

     141  The eight tampering types are air pump disablement,
catalyst removal, overall misfueling, fuel inlet restrictor
disablement, exhaust gas recirculation system disablement,
evaporative control system disablement, positive crankcase
ventilation system disablement, and missing gas caps.

     142  Tampering offsets are the increases in emissions that
result from a given type of tampering.

     143  If EPA or local authorities have performed a
statistically valid anti-tampering survey in a particular area, the
Office of Mobile Sources will consider whether it is possible to
develop locality-specific tampering rates from that survey for use
in modeling for the area.  A sample size larger than that collected
by EPA in the typical one week of its survey is essential.

                                60





     EPA has determined through its tampering surveys that
tampering rates are lower in areas with operating I/M programs than
in areas without such programs.  Two complete sets of tampering
rates representing I/M and non-I/M cases therefore come with the
model.

     If no I/M program exists, the set of non-I/M program tampering
rates must be input.  If an I/M program does exist, however, then
two sets of tampering rates must be input; the one representing the
non-1/M case accounts for tampering that occurred before the start
of the I/M program, and the one representing the I/M case accounts
for tampering that occurred after the start of the I/M program.

     MOBILE4.1 uses three rate equations for each type of
tampering, one for each vehicle type subject to tampering144 in
non-I/M areas and three more rate equations for each type of
tampering in the I/M areas.145

     The only change to anti-tampering program inputs since MOBILE4
is that when alternate tampering rates are used, additional
equations are required.  This is due both to the increase in model
year groups (from two in MOBILE4 to three in MOBILE4.1), and the
use of a second deterioration rate to describe the increase in
tampering rates as a function of mileage for mileages over
50,000.146, 147

     Guidance

     The tampering rates built into MOBILE4.1 are the rates that
should be used in all Clean Air Act (CAA) mandated development of
mobile source emission inventories.  Use of any other tampering
rates in CAA-related work must be based on actual in-use tampering
surveys, and must be approved in advance by EPA.  A local tampering
survey would have to be quite large to justify reliance on it in
preference to the MOBILE4.1 rates.  For guidance regarding EPA
approval of local tampering surveys and the development of
tampering rates based on such surveys, contact the Office of Mobile
Sources' Field Operations and Support Division, 202/382-2633.  For
guidance on the analysis of data collected in a local tampering
survey or for further guidance on developing the information
required to model the emissions effect of an anti-tampering
program, contact OMS's Emission Planning and Strategies Division,
313/668-4367.
___________________________

     144  The three equations axe used to model pre-1981 model
year vehicles, 1981-83 model year vehicles, and 1981 and later
model year vehicles.

     145  These rate equations are based on OMS analysis of
national tampering survey data.


     146  A second, higher deterioration rate is applied only to
LDGVs in MOBILE4.

     147  MOBILE4.1 assumes that the maximum rate of tampering for
vehicles of any given model year is the rate at 130,000 mi (ZML +
5*DR1 + 8*DR2).

                                61





3.4  VEHICLE MILES TRAVELED

     This section describes two methods of estimating highway
vehicle miles traveled.  Both are related to the Department of
Transportation's (DOTs) Highway Performance Monitoring System
(HPMS).

3.4.1     Highway Performance Monitoring System

3.4.1.1   Role of the HPMS in SIP Development

     EPA and DOT have both endorsed the Highway Performance
Monitoring System (HPMS) as the appropriate source of VMT
estimates.  All states, except for the states of California,
Connecticut, Florida, Hawaii, Maine, Michigan, Missouri, North
Carolina, New York, Ohio, Oregon, South Carolina and Washington,
should base their 1990 estimates of actual annual VMT on unique
sample panels for each Federal Aid Urbanized Area (FAUA) within the
state, since sampling has already occurred at that level of
geographic detail.148  The effect of this agreement is that the
VMT used to construct mobile source emission inventories should be
consistent with that reported through the HPMS.  However, since the
Federal Aid Urbanized Area geographic boundaries of HPMS are not
generally coincident with EPA's non-attainment area boundaries, the
two estimates of VMT will not necessarily be identical.

     Ideally, the VMT in that portion of the non-attainment area
that is encompassed by the FAUA should be the same as the VMT
reported for that FAUA to the U.S. Department of Transportation. 
If a non-HPMS method used by the state does not achieve this goal,
then an adjustment factor should be applied to the state-estimated
VMT to make it match the HPMS report.
___________________________

     148  In 1990, the states of California, Florida, Hawaii,
Maine, Michigan, Missouri, North Carolina, Oregon, South Carolina,
and Washington based HPMS VMT estimates on one or more collective
sample panels while the states of Connecticut, New York, and Ohio
based HPMS VMT estimates on a combination of individual and
collective sample panels.  In some of these states, such as
California, the state reports area-specific VMT estimates to HPMS
as if they were derived in the same way as in other states.  EPA
recognizes that such estimates may not be as reliable as estimates
based on individual sample panels and therefore may allow the state
to submit a 1990 SIP inventory based on VMT estimates that are not
fully consistent with the data submitted to HPMS.  EPA Regional
Offices are advised not to agree that the HPMS data are less
reliable than VMT estimates made by another state-requested method
without consulting with divisional or regional FHWA officials who
have direct knowledge of the HPMS data associated with the non-
attainment area.


                                62






     Consistency between HPMS VMT149 and SIP VMT150 Means, in
general, that the same factors used within the FAUA to seasonally
adjust and expand the HPMS 24- and 48-hour traffic count samples by
functional system and volume group to annual average daily traffic
(AADT) VMT should also be applied to all segments within the non-
attainment area.  Similarly, whatever method is used to estimate
VMT on local facilities within the state should also be applied to
local facilities within the non-attainment area.  Finally,
consistency means that the SIP functional systems are identical to
the HPMS functional systems.151, 152

     Since the VOC, NOx , and summer CO emission inventories are
typical summer day inventories, VMT for ozone non-attainment areas
should be adjusted to the summer season using the inverse of the
factors used to adjust summer 24- and 48-hour counts to AADT. 
Similarly, VMT for winter CO emission inventories should be
adjusted using the same technique.  Modeling inventories for
particular days should also be adjusted for average day-of-week
variations in VMT.

3.4.1.2 Overview of HPMS

     Urban areas with populations over 50,000 are required to
maintain formal transportation planning programs in order to meet
Federal requirements for securing transportation funds.  These
programs are intended to establish the analytical basis for
assessing current and future transportation needs and for
evaluating projects that will satisfy those needs.  Although not a
requirement of this process, the Highway Performance Monitoring
System can be very useful to it.

     The Highway Performance Monitoring System was developed by the
U.S. Department of Transportation Federal Highway Administration in
the mid-1970's to collect and report information on the nation's
highways.  Traffic data reporting for the system is documented in
the HPMS Field Manual153 and the Traffic Monitoring Guide.154
___________________________

     149  HPMS VMT refers to the vehicle miles traveled reported
to the Federal Highway Administration's Highway Performance
Monitoring System.  Under this program, traffic counts taken on
samples of an area's roadway network are adjusted for day-of-week
and season and expanded to include the area's entire roadway
network.

     150  SIP VMT refers to the vehicle miles traveled that is
used to construct the mobile source emissions inventory.

     151  SIP functional systems and the VMT reported thereon may
Dot be identical to those reported to the U.S. DOT, if the state
demonstrates that HPMS data are sufficiently uncertain and the
competing alternative proposed by the state is more accurate. 
Consult with divisional or regional FHWA officials who have direct
knowledge of the HPMS data associated with the non-attainment area.

     152  Exceptions to these general guidelines are described in
sections 3.4.1.3 and 3.4.2.4.

     153
     U.S. D.O.T.Code     Title
     M5600.1A            Highway Performance Monitoring System
                         (HPMS) Field Manual

                                63





     The HPMS universe consists of all public highways or roads
within a state.  The reporting strata for the HPMS include type of
area (rural, small urban, and individual or collective urbanized
areas) and functional class.  In rural areas the functional classes
are Interstate, Other Principal Arterial, Minor Arterial, Major
Collector, and Minor Collector.  In urban areas they are
Interstate, Other Freeway or Expressway, Other Principal Arterial,
Minor Arterial, and Collector.  A third level of stratification
based on 13 volume groups was added as a statistical device to
reduce sample size, insure the inclusion of higher volume sections
in the sample, and increase the precision of VMT at a lower sample
rate.

     The HPMS sample design is a stratified simple random sample. 
The sample size estimation process was tied to annual average daily
traffic, although nearly 75 data items are collected.  The decision
for using AADT was based partly on the fact that AADT is one of the
most variable data items in the HPMS.  As such, the reliability of
the other data is expected to equal or exceed that of AADT.

     The HPMS sampling element was defined on the basis of road
segments or links that include both directions of travel and all
travel lanes within the section.  AADT variability was estimated
based on data from the 1976 National Highway Inventory and
Performance Study (NHIPS).  Sample size was determined and the
sample selected as a simple random sample within strata according
to predetermined levels of precision.  For the higher volume strata
the sample size is estimated on the basis of providing a 90%
confidence that the sample strata mean was within  5% of the
universe strata mean.

     Typically 24- or 48-hour counts are taken on each sample
segment once every three years.  These short counts are adjusted,
based on day-of-week and season, to annual averages using a small
number of continuous traffic recorders.  The HPMS expansion factors
are computed as the ratio of universe mileage to sample mileage
within each stratum.155  This procedure expands the HPMS sample to
represent the universe of all roadways in the area by multiplying
each segment's AADT, segment length, and expansion factor and
summing the product for each sample section.  Axle correction
factors are also incorporated into the process.156  Growth factors
are used to convert counts on sections unmonitored during a year to
current year annual estimates, i.e., segments not counted during
the year of record are estimated by multiplying the earlier counts
by growth factors.
___________________________

     M 5600.1A, Chg. 1   Highway Performance Monitoring System
                         (HPMS) Field Manual Updates
     M 5600.1A, Chg. 2   Highway Performance Monitoring System
                         (HPMS) Field Manual
     (Deleted by Chg. 3)
     M 5600.1A, Chg. 3   Highway Performance Monitoring System
                         (HPMS) Field Manual

     154  Traffic Monitoring Guide, June, 1985, U.S. Department of
Transportation, Federal Highway Administration, Office of Highway
Planning.

     155  Expansion factors for different FAUAs may be different,
even within the same state.

     156  Vehicle counts taken by axle counting equipment require
adjustment by axle correction factors.  An adjustment factor is the
ratio of vehicles to axles as determined by a vehicle
classification count.

                                64






     While a statistically valid sample design could be developed
independently of the HPMS, the availability of the clearly defined
and implemented HPMS sample design results in a significant
reduction of effort.  The HPMS sample has now been implemented in
every State, the District of Columbia, and Puerto Rico.  It
provides a statistically valid, reliable, and consistent data base
for analysis within States, between States, and for any aggregation
of States up to the national level.

3.4.1.3   Consistency Between HPMS and SEP VMT

     VMT used to construct mobile source emission inventories
should be consistent with that reported through the HPMS.  However,
since the Federal Aid Urbanized Area geographic boundaries of HPMS
are not generally coincident with EPA's non-attainment area
boundaries, the two estimates of VMT will not necessarily be
identical.

3.4.1.3.1 Expansion Factors

     Consistency between HPMS VMT and SIP VMT means, in general,
that the same factors used to seasonally adjust and expand the HPMS
24- and 48-hour samples by functional system and volume group to
annual average daily traffic (AADT) VMT within the FAUA should also
be applied to all segments within the non-attainment area.157 
This in effect assumes that all roads of a given functional system
and volume group are used with equal intensity, allowing geographic
allocation of total HPMS VMT within the non-attainment area based
only on knowledge of roadway miles.

3.4.1.3.1.1    Non-Attainment Area the Same As the Federal Aid
               Urbanized Area

     If the boundaries of the non-attainment area are coincident
with the boundaries of the FAUA, then the SIP VMT and HPMS VMT
should be identical for each functional system.  This may require
that state-estimated non-attainment area VMT be adjusted to make it
equal the VMT reported to HPMS.
___________________________

     157  There are three exceptions to this generality.  In the
first case, separate expansion factors may be available for
functional systems in the rural portion of the non-attainment area
outside of the FAUA, and they should be used for that area.  In the
second case, if a state uses a network model to spatially and/or
temporally allocate HPMS VMT (adjusted to match the HPMS total)
within a non-attainment area, then the model's assignment to
different functional systems may prevail.  In this second case, the
model may show that intensity of use by functional system may be
quite different from one traffic analysis zone to another.  See
sections 3.4.2.4.1 and 3.4.2.4.2. Third, this rule does not apply
to any areas with weak HPMS data that have been set aside in favor
of a competing alternative.

                                65





3.4.1.3.1.2    Non-Attainment Area Inside of the Federal Aid
               Urbanized Area

     If the boundaries of the non-attainment area are entirely
within the boundaries of the FAUA, then the FAUA VMT by HPMS
functional system should be allocated to the non-attainment area
(by HPMS functional system) on the basis of the number of roadway
miles according to equation (3-7).158


VMT(SIP,f,v) =

(Roadway Miles(SIP,f,v) / Roadway Miles(FAUA,f,v) *
VMT(FAUA,f,v)

                                                              (3-7)

where     FAUA =    Federal Aid Urbanized Area

          SIP  =    SIP Non-Attainment Area

          f    =    Interstate System, Other Freeways and
                    Expressways, Other Principal Arterials, Minor
                    Arterials, Collectors159

          V    =    Volume Group160

          VMT(SIP,f,v) =     VMT(SIP,f,v)
                            all V

3.4.1.3.1.3    Non-Attainment Area Outside of the Federal Aid
               Urbanized Area

     If the non-attainment area is entirely outside the boundaries
of all FAUAs within the state, states may use any reasonable method
to estimate 1990 VMT on the separate functional systems within the
non-attainment area.
___________________________

     158  EPA will accept an adjustment based on total FAUA VMT
and not require that separate adjustment factors be applied to make
VMT identical for each functional system and volume group, if a
state demonstrates that the more refined approach is infeasible. 
If a state elects this option, it should use equation (3-7a).


VMT(SIP,f,v) =
          (Roadway Miles(SIP) / Roadway Miles(FAUA) * VMT(FAUA)

                                                             (3-7a)

     159  The functional systems listed are for an HPMS Urbanized
Area.  Rural areas classify highway facilities as interstates,
other principal arterials, minor arterials, major collectors and
minor collectors.

     160  Volume groups vary by geographic area and functional
system.  Appendix F from the HPMS Operations manual, included at
the end of this report, is a listing of functional systems and
volume groups within each HPMS geographic area.

                                66






     However, the recommended method of estimating VMT in a
completely rural non-attainment area is on the basis of the
expansion factors used for the universe of rural samples within the
state.161 The recommended method of allocating VMT to this type of
non-attainment area is described by equation (3-8).162, 163

VMT(SIP,f,v)  =

(Roadway Miles(SIP,f,v) / Roadway Miles(HPMS,f,v))  * 
VMT(HPMS,f,v)
                                                              (3-8)

     where HPMS     =    HPMS Statewide Rural Area

     SIP            =    SIP Non-Attainment Area

     f              =    Functional System

     v              =    Volume Group

     VMT(SIP,f)    =        VMT(SIP,f,v)
                         all v

3.4.1.3.1.4    Non-Attainment Area and Federal Aid Urbanized Area
               Crossover

     If the boundaries of the non-attainment area are such that
some of the non-attainment area is inside an FAUA and some of it
lies outside of any FAUA, then VMT in that portion of the non-
attainment area within the FAUA should be estimated by allocating
HPMS FAUA VMT to the non-attainment area on the basis of equation
(3-7),164 (assuming that only some of the FAUA is in the non-
attainment area; if all of it is, allocation is not necessary),
while VMT in that portion outside of the FAUA should be estimated
according to equation (3-8) and Section 3.4.1.3.1.3.

3.4.1.3.2 Local Functional System

     Section 3.4.1.3.1.3 above applies only to VMT on functional
systems designated collector and above.  While HPMS includes state-
provided estimates of VMT on the local functional system, these
estimates are not now generally based on current ground counts at
statistically representative sites.  Instead, the estimates are
based on a method chosen by the state in light of its own
circumstances.  States may continue to use the same method as they
___________________________

     161  Within HPMS, all rural areas within the state are
grouped into one sampling universe.

     162  Alternatively, a state may estimate VMT on the basis of
another rural area similar in terms of land use, etc., based on
equation 3-8.

     163  Other alternative methods of estimating VMT in this type
of non-attainment area are discussed in section 3.4.1.6.

     164  If all of the FAUA is inside of the non-attainment area,
equation 3-7 does not apply, since all FAUA VMT is included as part
of the total non-attainment area VMT estimate.

                                67





have used in the past to estimate 1990 VMT on the local functional
system within the FAUA, and they may also apply that method to the
portion of the non-attainment area outside of the FAUA.165  For
example, local road VMT within the non-attainment area might be
estimated as a simple percentage of all collector and higher VMT.

     Proper estimation of actual travel on the local functional
system is most important for areas subject to the highest ambient
ozone and CO concentrations.  However, increasing the accuracy of
these estimates will require some lead time.

3.4.1.3.3 Seasonal Adjustment

     HPMS-based annual average daily VMT should also be adjusted
for seasonal effects.  Since the VOC, NOx , and summer CO emission
inventories are typical summer weekday inventories, VW for ozone
non-attainment areas should be adjusted to the summer season using
the inverse of the factors used to adjust the 24- and 48-hour
counts to AADT.  Similarly, VMT for winter CO emission inventories
should be adjusted using the same technique.166

3.4.1.3.4 Daily Adjustment

     Since base year emission inventories must also be calculated
for a typical day, a similar adjustment using the inverse of the
factors used to adjust the 24- and 48-hour counts to AADT should be
made to convert typical summer day VMT to typical summer weekday
VMT and to convert typical winter day VMT to typical winter weekday
VMT.167

3.4.1.4   Allocating VMT to Time of Day

     It may also be necessary to allocate daily VMT to each hour of
the day.  This is commonly done for purposes of preparing emissions
estimates for the photochemical grid models used in forecasting
ozone concentrations.168  The recommended method of
___________________________

     165  A state may substitute, for the 1990 estimate of actual
VMT accumulated on the local functional system, a methodology
superior to that used for HPMS reporting in the past, provided that
the substitution is reflected both in the required emission
inventories and the attainment demonstrations.  Similarly, if,
after a state submits a required emission inventory or an
attainment demonstration, it wishes to change the methodology it
used to estimate VMT on the local functional system, it should re-
submit its emission inventory and attainment demonstration using
the same alternative methodology.

     166  Weather conditions may be too severe to take 24- or 48
hour traffic counts during the winter.  In that case, it may be
possible to use continuous monitors within the inventory area to
adjust summer counts to winter levels.  National VMT monthly
profiles could also be used to make the adjustment.

     167  Modeling inventories for particular days can also be
adjusted for average day-of-week variations in VMT.

     168  The Urban Airshed model and its associate emissions
preprocessor is an example of such a model.  EPA also maintains the
Regional Oxidant Model, a mesoscale photochemical model providing
coverage of the Northeastern United States.

                                68





apportioning daily VMT to specific hours is to use the state's
continuous monitors available within the FAUA.  If no such monitors
exist within the non-attainment area being modeled, then the state
may rely on other continuous monitors located in areas similar in
geographic, land use, and demographic characteristics, or on those
areas' final Airshed Emission Preprocessor profiles.169

3.4.1.5 Allocating VMT to Functional Systems

     To be consistent with HPMS, the SIP functional systems should,
with few exceptions, be identical to the HPMS functional systems
(Interstate System, Other Freeways and Expressways, Other Principal
Arterials, Minor Arterials, and Collectors).170

     The recommended method for estimating VMT on each HPMS
functional system within the non-attainment area is to follow the
procedures outlined in Section 3.4.1.3; that is, to allocate VMT on
the basis of roadway miles and functional class.  The underlying
assumption in this methodology is that VMT is generally a function
of total roadway miles and that this relationship becomes stronger
as individual types of highway facilities within specific areas are
considered.

3.4.1.6   Estimating VMT in Rural and Small Urban Areas171

     The general process of estimating VMT in rural and small urban
areas involves the apportionment of statewide VMT to the county or
other area for which mobile source emissions estimates are
required.  Statewide VMT data are tabulated by all state
transportation agencies and reported to the Federal Highway
Administration, which, in turn, annually publishes these and other
similar data in Highway Statistics.  Highway Statistics is based
upon and consistent with HPMS.172

     Central to the overall method of estimating VMT in rural and
small urban areas is the development of apportioning factors,
which, when applied to the statewide total, yield areawide VMT. 
Several options exist for apportioning VMT, such as roadway miles,
motor
___________________________

     169  The Airshed Emission Preprocessor System has a default
profile that can be used.

     170  One exception is that the Interstate System and Other
Freeways and Expressways may be combined into a single functional
system.  A second exception is that the Collector, Local, and
Frontage Roads may be combined into a single functional system.

     171  If a state has the capability of reporting rural or
small urban area VMT directly, the indirect methods described here
for apportioning state-level VMT data to the applicable area are
not appropriate.

     172  HPMS data submitted to FHWA on or before September 15th
of the year following the year of record are included in Highway
Statistics.  FHWA therefore considers the data in Highway
Statistics preliminary and subject to revision within the following
twelve months.

                                69





vehicle registrations, population, and fuel sales.173  The option
selected will depend on the availability of the required data.  A
key component of this process is the assistance of state and county
transportation engineers and planners who should be able to provide
the best assessment of both data availability and quality.

     Once the apportioning method has been developed, the resulting
factors are applied to statewide VMT to produce estimates of
areawide travel.  Other data are applied to yield VMT by roadway
classification and vehicle type.

     Statewide VMT estimates are available directly from state
transportation and highway agencies and from Table VM-2 in Highway
Statistics (included here as Table 3-3).  To provide the most
accurate estimates of travel that occurred in 1990, the data
contained in the 1991 edition of Highway Statistics should be used.

     Areawide VMT is estimated by applying the allocation factor,
Ff, to statewide VMT statistics obtained from the state
transportation agency, or from Table 3-3, from Highway Statistics. 
Expressed as an equation, the estimated annual VMT occurring in the
area is:

          VMT(SIP, f) =- Ff * VMT(State, f)                (3-9)

where:

     VMT(SIP, f)   =    estimated annual areawide VMT;

          Ff       =    the apportioning factor to be applied to
                         statewide VMT to estimate areawide VMT for
                         functional system f, derived from one of
                         equations 3-10 through 3-13;

     VMT(State, f) =    total statewide VMT for functional system
                         f, obtained from the state transportation
                         agency or from Table 3-3.


     In all instances except where the apportioning factor is based
on roadway miles (equation 3-10), the value assigned to VMT(State
,f) is found in the TOTAL column for ALL HIGHWAY CLASSES in Table
3-3.  If equation 3-10 is used, the individual values of VMT for
each roadway type in urban and rural areas in Table 3-3 are used
for VMT(State,f).  Four methods for apportioning statewide VMT to
county or other subareas are presented below.
___________________________

     173  The recommended method for apportioning VMT in a
completely rural non-attainment area is by roadway miles.

                                70





apportioning daily VMT to specific hours is to use the state's
continuous monitors available within the FAUA.  If no such monitors
exist within the non-attainment area being modeled, then the state
may rely on other continuous monitors located in areas similar in
geographic, land use, and demographic characteristics, or on those
areas' final Airshed Emission Preprocessor profiles.169

3.4.1.5   Allocating VMT to Functional Systems

     To be consistent with HPMS, the SID? functional systems
should, with few exceptions, be identical to the HPMS functional
systems (Interstate System, Other Freeways and Expressways, Other
Principal Arterials, Minor Arterials, and Collectors).170

     The recommended method for estimating VMT on each HPMS
functional system within the non-attainment area is to follow the
procedures outlined in Section 3.4.1.3; that is, to allocate VMT on
the basis of roadway miles and functional class.  The underlying
assumption in this methodology is that VMT is generally a function
of total roadway miles and that this relationship becomes stronger
as individual types of highway facilities within specific areas are
considered.

3.4.1.6   Estimating VMT in Rural and Small Urban Areas171

     The general process of estimating VMT in rural and small urban
areas involves the apportionment of statewide VMT to the county or
other area for which mobile source emissions estimates are
required.  Statewide VMT data are tabulated by all state
transportation agencies and reported to the Federal Highway
Administration, which, in turn, annually publishes these and other
similar data in Highway Statistics.  Highway Statistics is based
upon and consistent with HPMS.172

     Central to the overall method of estimating VMT in rural and
small urban areas is the development of apportioning factors,
which, when applied to the statewide total, yield areawide VMT. 
Several options exist for apportioning VMT, such as roadway miles,
motor
___________________________

     169  The Airshed Emission Preprocessor System has a default
profile that can be used.

     170  One exception is that the Interstate System and Other
Freeways and Expressways may be combined into a single functional
system.  A second exception is that the Collector, Local, and
Frontage Roads may be combined into a single functional system.

     171  If a state has the capability of reporting rural or
small urban area VMT directly, the indirect methods described here
for apportioning state-level VMT data to the applicable area are
not appropriate.

     172  HPMS data submitted to FHWA on or before September 15th
of the year following the year of record are included in Highway
Statistics.  FHWA therefore considers the data in Highway
Statistics preliminary and subject to revision within the following
twelve months.

                                69








                                71





3.4.1.6.1 Apportionment of Statewide VMT-Recommended Method

     The recommended method of allocating statewide VMT to rural or
small urban areas is on the basis of roadway miles by functional
class.  The underlying assumption in this method is that VMT is
generally a function of total roadway miles and that this
relationship becomes stronger as individual types of highway
facilities within specific areas are considered.

     State transportation and highway agencies report total roadway
miles by functional class (i.e., interstate, other freeways and
expressways, other principal arterials, minor arterials, and
collectors) in urbanized, small urban, and rural areas, for both
the Federal-Aid and non-Federal-Aid portions of the highway system. 
A summary of these data is published by FHWA in Highway Statistics
as Table HM-20, which is shown here as Table 3-4.  As Tables 3-3
and 3-4 show, roadway miles and VMT are reported in exactly the
same format, i.e., by functional class on Federal-Aid and non-
Federal-Aid highways in both urban an( rural portions of each
state.  The one additional set of data required is the areawide
roadway miles disaggregated by the same categories as those for the
statewide data.  This information should be available directly from
the state transportation agency, or it can be derived based on
input from both state and county transportation agencies.

     The VMT allocation factor is then derived as the ratio of
areawide to state roadway miles for each category of highway in the
applicable rural or small urban area:

     Ff = Roadway Miles(SIP, f)  / Roadway Miles(State, f)(3-10)

where Ff =    the apportioning factor to be applied to statewide
               VMT to estimate areawide VMT for functional system
               f;

Roadway Miles(SIP, f)   =    roadway miles of functional system f
                              in the area;

Roadway Miles(State, f) =    roadway miles of functional system f
                              in the state.

     The result will be as many as 13 individual allocation factors
in each area, each of which is applied to the corresponding VMT
figure in Table 3-3 from Highway Statistics.

                                72





Click HERE for graphic.


                                73





3.4.1.6.2 Apportionment of Statewide VMT-Alternative Methods

     One of the remaining three apportioning methods may be used as
an alternative to apportioning by roadway miles in rural counties
if the state can demonstrate that the recommended method is either
not feasible or less accurate than one of the alternative methods. 
Such might be the case in rural counties entirely outside of the
non-attainment area but within a photochemical modeling domain.

3.4.1.6.2.1    Motor Vehicle Registrations

     The first of these alternative methods utilizes motor vehicle
registration data to derive an apportioning factor.  The premise
here is that the amount of travel occurring in an area is a
function of the area's vehicle population.  This method assumes
that travel by vehicles registered outside the area is balanced by
travel outside the area by vehicles registered in the area and that
individual areawide travel patterns are not significantly different
from the statewide average for the area in question.

     The allocation factor is the ratio of areawide-registered
vehicles to state-registered vehicles.  Registration statistics can
be obtained from the state Department of Motor Vehicles (DMV) for
both the state and county, although there may be a requirement for
special processing of state registration files to produce areawide-
specific data.  The factor is defined by:

               F = V(Area) / V(State)                      (3-11)

     where F   =    the allocation factor to be applied to
                    statewide VMT to derive areawide VMT;

     V(Area)  =    total number of motor vehicles of all types
                    registered in the area;

     V(State) =    total number of motor vehicles of all types
                    registered in the state.


3.4.1.6.2.2    Population

     A second alternative method for allocating statewide VMT
apportions statewide VMT to the area based on the relative area and
state population.  This method has the advantage of utilizing data
that are routinely available from several sources, such as the
Bureau of Census and state and county agencies, and therefore VMT
estimates can be developed with minimal effort.  It is important to
use 1990 data to allocate the travel in order to have a
representative estimate of the travel that occurred in the base
year of the emissions inventory.  The apportioning factor is:


                     F = P(Area) / P(State)                (3-12)


                                74





where F   =    the allocation factor to be applied to statewide VMT
               to derive areawide VMT;

P(Area)  =    total population in the area;

P(State) =    total population in the state.


3.4.1.6.2.3    Fuel Sales

     Finally, statewide VMT can be apportioned to the area based on
fuel sales data.  An assumption inherent in this method is that VMT
and fuel sales are directly related.  In order for this method to
be practical, areawide fuel sales data must be available.  If such
data are not available, one of the other methods will need to be
used.

     All states collect taxes on motor fuel sold within their
boundaries, and formal records are maintained regarding both the
fuel throughput and the revenue derived therefrom.  Since these
statistics are usually aggregated as state totals, special
processing may be required to identify fuel sales by area.  The
data requirements should be discussed with appropriate officials in
the state taxation or revenue agency.

     As a minimum, total annual sales (gallons) of gasoline and
diesel fuel in both the area and state are required.  The monthly
distribution of sales is also required.  Note that each state
reports monthly fuel sales data to the Federal Highway
Administration, which, in turn, reports various fuel use statistics
in Highway Statistics and in a monthly publication by FHWA entitled
Motor Fuel Consumption Reports (MF26G).

     The apportioning factor is the ratio of areawide motor fuel
use to state fuel use.  Tables MF-25 and MF-26, respectively, in
Highway Statistics tabulate the monthly highway use of special
fuels and gasoline for each state in the U.S. These tables are
included here as Tables 3-5 and 3-6.  Special fuels are essentially
diesel fuel and some liquefied petroleum gases.  Table MF-26
indicates the actual use of gasoline and gasohol by highway
vehicles; the data shown in that table have been adjusted to
account for handling losses and exclude gasoline used for non-
highway purposes.  The VMT apportioning factor is:

                     F = G(Area) / G(State)                (3-13)

where F   =    the allocation factor to be applied to statewide VMT
               to derive areawide VMT;

G(Area)  =    total quantity (gallons) of gasoline and diesel fuel
               sold in the area, obtained from the state revenue
               agency;

G(State) =    total quantity (gallons) of gasoline and diesel fuel
               sold in the state, obtained from the state revenue
               agency or from Tables MF-25 and MF-26 in Highway
               Statistics.

                                75





Click HERE for graphic.


                                76





Click HERE for graphic.


                                77





3.4.2     Travel Demand Network Models

3.4.2.1   Role of Transportation Models in SIP Development

     A common and highly useful information source for the
preparation of the baseline inventory is a transportation planning
model run configured to the same base year (1990).  With this tool
one is able to combine the appropriate travel estimates with
emission factors developed by MOBILE4.1 to prepare an estimate of
the emissions inventory.  In addition, network model results can be
used to develop certain critical inputs to the emission factor
program.  These include speeds, vehicles' operating conditions,
trip starts, trip ends, number of trips per day per vehicle, and
vehicle mix.  In addition, network models can be used to spatially
and temporally allocate VMT, and therefore emissions, within the
non-attainment area.  However, since EPA and DOT have both endorsed
the Highway Performance Monitoring System as the most appropriate
means by which to measure VMT, the VMT estimates produced by the
transportation planning process should be made consistent with
HPMS.  The mechanism for making this adjustment is discussed in
Section 3.4.2.4.

     The initial development of new transportation planning model
runs for both a base and a future year can take a substantial
effort, especially for a large urban area.  EPA recognizes that,
generally, local air quality agencies do not operate the
transportation planning models themselves and that they must work
with the agencies responsible for the operation of such models.

3.4.2.2   Background

     An important element of the transportation planning process is
an assessment of the regional highway network.  An extensive effort
is required to collect and integrate the information needed to
assess how the network is currently used and where growth will
occur in the future.  The mathematical models used to assess the
effects of growth on the highway network allow the transportation
analyst to evaluate the improvements that are needed and when they
should be constructed.  From the perspective of air quality,
important products of these models are estimates of VMT and speed
on each of the links coded into the highway network.

     Basic requirements of the transportation planning process are
an understanding of where travel occurs, what factors stimulate it,
and how demand is satisfied.  The Federal Highway Administration
and the Federal Transit Administration (FTA)174 developed a series
of models to help communities satisfy these requirements. 
Historically, the most frequently used model has been the Urban
Transportation Planning System (UTPS).175  In recent years
___________________________

     174  The Federal Transit Administration was formerly the
Urban Mass Transit Administration (UMTA).

     175  Supplement 1 to "Methodology to Calculate Emission
Factors for On-Road Motor Vehicles", California Air Resources
Board, Technical Support Division, January 1988.

                                78





many variations of the UTPS have been developed in the private
sector as Federal funding for the model decreased and sufficiently
powerful microcomputers became available.176

     Responsibility for implementing and operating transportation
planning models generally falls into two categories: metropolitan
planning organizations and state Departments of Transportation.  In
some cases, transportation or planning agencies within communities
operate the models.  Responsibility for operating the models
generally falls outside of the jurisdiction of the air quality
planning agency.  Thus, the development of an emissions inventory
for an urban area requires a cooperative effort between the air
quality agency and the relevant transportation planning agency (the
one operating the UTPS-type model).  This should not come as a
surprise to most communities, since a cooperative effort between
the two types of agencies was required to develop the SIP
inventories in 1979 and 1982.

     Despite the fact that air quality planners are not responsible
for operating the highway network models, it is important that they
have an understanding of the principles guiding the operation of
the models and the information that they generate.  Section 3.4.2.3
provides an overview of UTPS-type operation and outputs.  It is not
designed to supplant the need to have a transportation planner
actively involved in both generating and assessing the travel
information used to prepare a highway emissions inventory estimate
but does provide an introduction to the transportation planning
process.

3.4.2.3   Overview of Network Models

     UTPS-type models consist of manual and computerized planning
procedures that provide decision-makers with information on long-
range transit and roadway travel patterns.  UTPS-type computer-
based packages allow planners to simulate the operation of a
transportation system to determine what would happen if population
and economic activity increased and/or if changes were made in
either the roadway or transit networks.  When population, economic
activity, and roadway and transit network inputs are matched to
historical conditions, an estimate of VMT for that year is
produced.  The estimate, however, is based on many assumptions and
estimates, not on direct observation of travel in that year.  UTPS-
type computer packages consist of a number of programs that
parallel steps in the transportation planning process.  In general,
this process involves the following major steps:

     -    Representation of the roadway or transit system;
     -    Estimation of the number of current and future drivers
          and transit riders, the numbers of trips of various types
          they will choose to take in a typical day, and their trip
          origins and destinations;
     -    Assignment of trips to appropriate roads and transit
          routes; and
     -    Preparation of maps, tables and graphs to display results
          and compare different transportation alternatives.
___________________________

     176  UTPS as a computer system is no longer supported by
either FHWA or FTA.  However, although some of the conventions used
by UTPS may not be identical to those used by the other UTPS-type
programs, they are similar, since, generally, UTPS served as the
prototype for these other programs.

                                79





     The capabilities of a UTPS-type package include estimation of
the impacts of long-range land development, transportation system
costs, travel demand, and major facility and corridor travel
volumes.  The package has been characterized as "data hungry".  For
most applications, planners must prepare a description of the
roadway and/or transit networks as well as detailed demographic and
economic forecasts.  In addition, policy makers must agree on
transportation alternatives to be tested and identify the impacts
about which they are interested.  Depending on the complexity of
problems to be addressed, the availability of raw data, and the
experience of the analytical staff, preparing initial inputs for a
UTPS-type model can take from two months to over two years.  The
following discussion provides a more detailed overview of steps
involved in configuring a UTPS-type package to a community.

     The development of a realistic abstraction of the existing
highway and/or transit network is the most time-consuming step
required to implement a UTPS-type package.  A network177 describes
the characteristics of roads or transit lines to the computer in
the same way a map describes roads to a driver.  The first step in
network coding is the development of the zone system.

     Zones are geographic areas dividing the study area into
relatively homogeneous areas of land use, land activity, and
aggregate travel demand.  Zones represent the origins and
destinations of travel activity within the study area.  Since it is
not computationally feasible to represent every household, place of
employment, shopping center, and other activity as a separate
origin and destination, these entities are first aggregated into
zones and then further compressed into a single node.  A centroid
is a point that represents all travel origins and destinations in a
zone.  Typically in the highway network, these centroids are
connected to the highway system at several points to represent the
many paths over which each of the discrete origins and destinations
within a zone access the balance of the highway system.

     Once the zone system is developed and mapped, zonal
socioeconomic data can be assembled for the transportation planning
process.  Zone centroids can be located in the center of activity
of the zone, using land use maps, aerial photographs, and local
knowledge.  The center of activity is not necessarily the
geographic center, but it is the midpoint of activity.  The maximum
number of allowable zones in UTPS is 2,500.  UTPS-type
microcomputer packages may have different limits.

     Selection of links is the second major step in developing a
network, since links represent those facilities (highways, roads,
streets, etc.) that actually comprise the highway system.  The two
nodes that mark a link's end points define the link in the
transportation network.  Nodes can be defined as those locations in
the highway system where vehicles are able to change direction of
travel (e.g., intersections, interchanges, etc.) or where the level
of service of a highway facility alters significantly (i.e., where
a road narrows from four lanes to two lanes).  It should also be
noted that some nodes represent origins and destinations of
___________________________

 177  This description was abstracted from "UTPS Highway Network
Development Guide", prepared by COMSIS Corporation for the Federal
Highway Administration, U.S. Department of Transportation, January
1983.

                                80






travel within the study area.  These nodes are actually centroids
and, as such, represent geographic units of travel demand at a
single point.  Figure 3-1 provides an example of network
components.

     In most UTPS-type applications travel time must be assigned to
the link, since links define the actual paths along which traffic
flows through the study area.  UTPS conventions also call for
separate links to designate opposite travel directions.  One link
is used to represent the east-west direction of flow, for example,
and another to represent the west-east direction.

     UTPS allows for a variety of highway attributes to be
associated with each link, including:

     -    X-Y coordinates to locate the nodes for plotting
          purposes;
     -    zones, which generally follow census data boundaries;
     -    geographic locations that generally define a corridor or
          larger (than zone) indication of location.
          zone) indication of location;
     -    area types to describe the kind of business or
          residential development that may be occurring around the
          node or link; and
     -    turn penalties and prohibitions to provide an indication
          of additional time required to make a particular movement
          through a node.

     Link attributes summarize basic network information about
highway facilities and are generally grouped into three broad
categories.

3.4.2.3.1 Level of Service

     Level of service is a qualitative measure describing
operational conditions within a traffic stream and their perception
by motorists and/or passengers.  A level-of-service definition
generally describes these conditions in terms of such factors as
speed and travel time, freedom to maneuver, traffic interruptions,
comfort and convenience, and safety.

     Level of service is used to determine path choice and vehicle
assignments in the network.  Six levels of service are defined for
each type of facility for which analysis procedures are available. 
They are given letter designations, from A to E with level-of-
service A representing the best operating conditions, and level-of-
service E the worst.178
___________________________

     178  Highway Capacity Manual, Special Report 209,
Transportation Research Board, National Research Council,
Washington, D.C., 1985.

                                81





Click HERE for graphic.


                                82





3.4.2.3.2 Physical Attributes

     Physical attributes describe the type of area where a link is
located, the number of lanes, and its observed volume.  The
functional system of a link serves as an index to the link
speed/capacity tables.  These tables assign free flow speed and
per-roadway capacity to the links using a single digit integer. 
The integer corresponds to default speed/capacity values:

     Functional
     System Code    Functional System

          1         Freeway
          2         Expressway
          3         Two-way arterial with curb parking
          4         One-way arterial without parking
          6         Two-way arterial without parking

     All other coded values (5 or greater than 6) are treated as
centroid connector links in the default condition.  The
interpretation of functional system codes can be changed by
modifying the speed/capacity table values.  Functional system is
also used to aggregate and summarize traffic assignment results in
output reports.

     Example speed/capacity tables are provided as Figures 3-2 and
3-3.

     Area type describes the part of the area in which the facility
is located.  Like functional system, area type is indexed to the
speed/capacity table so that free flow speed and capacity can be
determined.  Single digit integers used to represent link area
types are:

     Area Type
        Code        Area Type

          1         Central business district (CBD)
          2         CBD Fringe
          3         Residential
          4         Outlying

     All other values (greater than 4) are interpreted in the
default table as rural values.  The interpretation of area type
codes can be changed by modifying the default values in the
speed/capacity table.  Information on traffic assignment results
can also be aggregated by area type in output reports.

3.4.2.3.3 Locational Link Attributes

     Locational link attributes define each link's place in the
study area in relation to other links.  These attributes include
the area type in which the link lies, the analysis zone in which it
is found, its geographic location, and the group of links with
which it belongs.  With the exception of geographic location, the
other link attributes have already been defined.

                                83





Click HERE for graphic.


                                84





Click HERE for graphic.


                                85





     It is important that air quality analysts understand the basic
structure of the networks employed in a UTPS-type package so that
they can interpret and manipulate the various output reports that
are available from the individual programs.  The remaining
steps179 in the transportation planning process are as follows.

3.4.2.3.4 Trip Generation

     The relationships between trip making and the social-economic
characteristics of the residents of an urban area, as well as the
relationships between trip making and land use, are obtained from
travel data and land use inventories.  Trip generation procedures,
which relate these characteristics, are then developed.  There are
several trip generation procedures available, such as regression
analysis, cross classification analysis, and rate analysis.

     Forecasted social, economic, and land use data are substituted
for base year data in the trip generation procedure, and forecasts
of trips are obtained for each analysis unit.

3.4.2.3.5 Trip Distribution

     The origin-destination data collected in dwelling unit, truck,
taxi, and external surveys provide an estimate of the existing
travel taking place within, into, out of, and through the urban
area on an average day.  The trip distribution model is developed
to simulate the manner in which trips are made between small
analysis units within the study area.  The Gravity Model is used in
various urban areas throughout the United States.

3.4.2.3.6 Modal Split

     This term is used to define the division of total person trips
in an urban area between public (buses, trains, etc.) and private
(automobiles, trucks, etc.) transportation, or the process of
separating person trips by mode of travel.  Thus, a modal split
model is one that is used to forecast the amount of person trip
travel that will use mass transit facilities.  The calibration of
this model is dependent upon the relationships that have been found
from the travel data collected for the study year.  Modal split
analysis varies in the degree of complexity, ranging from simple
estimates in smaller urban areas to complex mathematical relations
in larger areas.  It generally involves characteristics of the trip
maker, the trip, and the transportation system.

3.4.2.3.7 Traffic Assignment

     The trip data collected in travel surveys describe the total
number of trips occurring between small analysis areas; however,
these surveys do not collect data that describe the specific routes
used to travel from an origin to a destination.  Consequently, a
traffic assignment technique is developed.  This technique can be
used to estimate the routes of
___________________________

     179  The following descriptions of trip generation, trip
distribution, modal split, and traffic assignment were obtained
from Urban Origin-Destination Surveys, U.S. Department of
Transportation, Federal Highway Administration, Washington, D.C.

                                86





travel that are used for trips occurring within the urban area. 
Traffic assignment may be described as the process of allocating
trip interchanges to a specific transportation network.  It is a
tool that allows the transportation planner to assign either
present trips (those trip interchanges between zones obtained from
the travel surveys) or future trips (those forecasted for some
future year) to various alternative transportation systems.

3.4.2.3.8 Feedback

     Feedback refers generally to the relationships within and
among the various steps of the transportation process.  For
example, in trip assignment, the constrained equilibrium approach
first assigns trips to pathways that result in the quickest journey
under an initial set of link-speed assumptions.  However, under the
constrained equilibrium algorithm, the initial set of speeds is
altered as trips are assigned.  As more and more trips are assigned
to links, vehicle speed is reduced, often to the point where
alternate routes from point A to point B are quicker.

     Feedback among different planning steps is also logically
important but often not implemented.  For example, travel
projections from transportation models are often not fed back to
the original regional growth and land use projections that formed
the basis for the transportation model.  Ideally, this type of
feedback should also proceed to equilibrium.  However, there is now
only limited information in the literature that describes exactly
how to incorporate feedback effects, particularly feedback all the
way back to regional growth and land use assumptions.  Further,
estimates of the degree to which incorporating such effects changes
predictions of travel behavior are also limited.180, 181

3.4.2.4   Consistency Between Transportation Model VMT and HPMS

     EPA encourages the use of transportation models in the
development of SIP emission inventories.  A transportation planning
model ran configured to the same base year (1990) as the required
emission inventory can be an excellent source of geographic and
temporal detail for transportation activity levels.  However, since
EPA and DOT both endorse the Highway Performance Monitoring System
as the appropriate means by which to measure VMT, the detailed VMT
estimates produced by the transportation planning process should be
made consistent in the aggregate with HPMS.

The process of making the network model VMT estimates consistent
with HPMS VMT estimates begins with the procedures discussed in
Section 3.4.1. Recall that since the Federal Aid Urbanized Area
boundaries of HPMS are not generally coincident with EPA's
___________________________

     180 The Federal District Court of Northern California ruled
that where a model had the capability to incorporate feedback
effects, the planning agency was obligated to project travel with
those effects included.

     181  EPA considers that the feedback effect between trip
assignment and the trip origin/destination distribution is the most
important at this time, given the current state of modeling
practice and the potential for model improvement that incorporating
such effects may have.  The link travel times used for trip
distribution should be consistent with the result of the trip
assignment step.

                                87





non-attainment area boundaries, the HPMS and non-attainment area
VMT estimates will not necessarily be identical.  Therefore, the
first step in making network model VMT consistent with HPMS VMT is
to make non-attainment area VMT consistent with HPMS VMT.  Once
that is accomplished, that consistent non-attainment area VMT
becomes the benchmark VMT to which network model VMT should be
adjusted.

     The actual adjustment of network model VMT is straightforward
and parallels the discussion in Section 3.4.1.3.1182

3.4.2.4.1 Non-Attainment Area the Same As the Network Model Area

     Even if the boundaries of the non-attainment area are
coincident with the geographic domain of the network model, there
are several reasons why the estimates of VMT produced by the HPMS
and the network model may be different.

     -    Not all higher level functional system links183 may be
          coded into the network;
     -    Links on local functional systems may not be coded at
          all;
     -    Network VMT may be estimated for different time periods;
          e.g., annual rather than seasonal, average day rather
          than average weekday, separate peak and off-peak travel
          estimates rather than an average weighted by hour-of-the-
          day.

     If the boundaries of the non-attainment area are coincident
with the boundaries of the travel demand network model, then
network model VMT and non-attainment area VMT184 should be made
identical.  The adjustment required to achieve this identity is
described by equation 3-14.185, 186
___________________________1

     182  Using HPMS rural area expansion factors is equivalent to
allocating state rural totals by roadway miles.

     183  Interstate system, other freeways and expressways, other
principal arterials, minor arterials, and collectors.

     184  Non-attainment area VMT is the HPMS-based VMT estimated
by the procedures described in Section 3.4.1.

     185  States are encouraged to make separate adjustments for
each HPMS functional system.  However, a state may demonstrate that
the more refined approach is infeasible due to inconsistent
classification schemes, data problems, or constraints on time and
staff.  If it is not be possible to identify network links with
HPMS functional systems, a state may adjust VMT and other network
factors according to equation 3-14a.


VMT(Adj. Network, i)    =
     (VMT(SIP) / VMT(Network)) * VMT(Network, i)         (3-14a)

     186  Other network model parameters, such as the number of
trips, number of cold starts, etc., (if used) should be increased
or decreased by the same factor(s) so that, for example,
evaporative emissions are correctly converted to grams per mile
emission rates and so that cold start emissions are correctly
weighted into an overall emission factor.

                                88





VMT(Adj. Network, i)
     (VMT(SIP, f) / VMT(Network, f)) * VMT(Network, i)    (3-14)

where:

 VMT(Adj. Network, i)   =    Adj. Network Model VMT

(VMT(SIP, f)            =    SIP Non-Attainment Area VMT187, 188

VMT(Network, f)         =    Original Network Model VMT

f                        =    Functional System

i                        =    Individual network link or group of
                              links (e.g., links of one functional
                              system in one area type)

3.4.2.4.2 Non-Attainment Area Inside of the Network Model Area

     If the boundaries of the non-attainment area are entirely
within the geographic domain of the network model, then the
procedures described in Section 3.4.1.3.1.2 should be used to
calculate an adjusted network VMT.  This requires that network
links or portions thereof are identified as to whether or not they
are in the non-attainment area.

3.4.2.4.3 Non-Attainment Area Outside of the Network Model Area

     If the boundaries of the. non-attainment area are entirely
outside of the geographic domain of the network model, then the
procedures described in Section 3A.1.3.1.3 should be used to
calculate VMT.

     The recommended method of estimating VMT in a completely rural
non-attainment area is on the basis of the HPMS expansion factors
used for an area within the state that is comparable in terms of
land use, transportation use, and demographic characteristics.189
___________________________

     187  Defined by equations 3-7, 3-7a and/or 3-9.

     188  Under some circumstances the adjustment factor could be
based on an area other than the non-attainment area.  For example,
a designated CO non-attainment area may be considerably smaller
than either the geographic domain of the network or the FAUA. in
this case, if the network domain exceeds the FAUA boundaries, then
the VMT contained within the FAUA, including VMT on local and other
functional systems, could be the point of departure for equation 3-
14.  Or, if the network domain is smaller than the FAUA, network VW
could be supplemented with reasonable estimates of VMT outside of
the network but within the FAUA, and that total VMT could be the
point of departure for equation 3-14.  A state may elect this
option, provided that it demonstrates that it produces a more
accurate estimate of VMT than equations 3-7, 3-7a and/or 3-8. 
Further, the state must continue to utilize the same approach for
all of its CAAA reporting requirements.  EPA Regional Offices are
advised that use of equations 3-7, 3-7a and/or 3-8 is presumed to
be the better approach for ozone non-attainment areas.

     189  Within HPMS all rural areas within the state are also
grouped into one sampling universe.  One alternative to estimating
VMT on the basis of another rural area similar in terms of land
use, etc., is to base the estimate on all other fully rural areas
combined.

                                89





The recommended method of allocating VMT to this type of non-
attainment area is described by equation (3-15).190  This equation
is similar to equation (3-8), which describes how to allocate VMT
to a non-attainment area completely outside of the FAUA.

                (Roadway Miles(SIP, f, v)
VMT(SIP, f, v) = __________________________ * VMT(HPMS, f, v)

                 Roadway Miles(HPMS, f, v)                  (3-15)

where:
     HPMS =    Applicable HPMS Area
     SIP  =    SIP Non-Attainment Area
     f    =    Functional System
     v    =    Volume Group

3.4.2.4.4  Non-Attainment Area and Network Model Area Crossover

     If the boundaries of the non-attainment area are neither
entirely outside nor entirely inside the geographic domain of the
network model, then a combination of the approaches described in
Sections 3.4.2.4.2 and 3.4.2.4.3 will need to be applied.  VMT in
that portion of the non-attainment area that is within the network
model's geographic domain should be based on equation 3-14; VMT in
that portion of the non-attainment area that is outside of the
network should be based on the procedures described in Section
3.4.1.3.1.3.

3.4.2.5 Local Functional System

Regional planning analyses frequently only represent the higher
functional systems.191  While in smaller urban areas the level of
detail on the lower functional systems may be higher, conversations
with FHWA officials have indicated that the local roads normally
excluded from the network can represent up to 15 percent of the VMT
that occurs in the planning area.

     Further, the low speeds generally recorded on these streets
can magnify their emissions contribution to the inventory.  This
omission can lead to a significant underestimate of the highway
contribution to the total emissions inventory.

     Therefore, if not accounted for by the highway network model
explicitly, a separate estimate of the travel occurring within the
transportation planning area on these lower functional systems must
be prepared.  The estimate of VMT may be based on MPO or DOT
___________________________

     190  Other alternative methods of estimating VMT in this type
of non-attainment area are discussed in section 3.4.1.6.

     191  Interstate system, other freeways and expressways, other
principal arterials, minor arterials, and collectors.

                                90





judgment and/or FHWA studies of areas of comparable size.192  The
estimate of off-network VMT should be added to network VMT prior.to
use of equation 3-14.

3.4.2.6   Seasonal Adjustment

     HPMS Annual Average Daily VMT should also be adjusted for
seasonal effects.  Since the VOC, NOx , and summer CO emission
inventories are typical summer weekday inventories, VMT for ozone
non-attainment areas should be adjusted to the summer season using
the inverse of the factors used to adjust the 24- and 48-hour
counts to AADT.  Similarly, VMT for winter CO emission inventories
should be adjusted using the same technique.

3.4.2.7   Daily Adjustment

     Since base year emission inventories must also be calculated
for a typical day, a similar adjustment using the inverse of the
factors used to adjust the 24- and 48-hour counts to AADT should be
made to convert typical summer day VMT to typical summer weekday
VMT and to convert typical winter day VMT to typical winter weekday
VMT.193

3.4.2.8   Allocating VMT to Time of Day

     It may also be necessary to allocate daily VMT to each hour of
the day.  This is commonly done for purposes of preparing emissions
estimates for the photochemical grid models used in forecasting
ozone concentrations.  The recommended method of apportioning daily
VMT to specific hours is to use the HPMS continuous monitors
available within the FAUA.  If no such monitors exist within the
non-attainment area being modeled, then the state may rely on other
continuous monitors located in areas similar in geographic, land
use, and demographic characteristics.

3.4.2.9   Allocating VMT to Functional Systems

     To be consistent with HPMS, the SIP functional systems should,
with few exceptions, be identical to the HPMS functional systems
(Interstate System, Other Freeways and Expressways, Other Principal
Arterials, Minor Arterials, and Collectors).194
___________________________

     192  See section 3.4.1.3.2.

     193  Modeling inventories for particular days should also be
adjusted for average day-of-week variations in VMT.

     194  One exception is that the Interstate System and Other
Freeways and Expressways may be combined into a single functional
system.  A second exception is that the Collector, Local, and
Frontage Roads may be combined into a single functional system. 
Finally, if it is not possible to map the classification system
built into the network model into the HPMS functional systems, then
the state may develop its own typing scheme as long as it is
reasonable and represents the entire roadway network.

                                91





     The recommended method for estimating VMT on each HPMS
functional system within the non-attainment area is to follow the
procedures outlined in Section 3.4.1.3; that is, to allocate VMT on
the basis of roadway miles and functional class.  The underlying
assumption in this methodology is that VMT is generally a function
of total roadway miles and that this relationship becomes more
direct as individual types of highway facilities within specific
areas are considered.

3.4.3     Exception to the Use of HPMS VMT

     Since 1990 ground counts submitted to HPMS may not be as
comprehensive and of as high a quality as FHWA intends all states
to obtain for 1993 and later and since it may be possible for
network-based travel demand models to be validated for 1990,195
this guidance allows for the use of travel demand models to
estimate 1990 VMT under certain circumstances.  However, this
method is not considered to be viable for most areas due to the
general disrepair of a large number of network models.  Areas
should use this method only if their network model is particularly
strong and their 1990 HPMS data are particularly weak, and only
after consulting with EPA.196

     An affected area with a strong network-based travel demand
model that is based on reasonably recent demographic trip-making
data may use its model to estimate 1990 VMT after consultation with
EPA197 and under the following conditions:

     -    Urban areas within the state were not sampled separately
          under HPMS in 1990;

     -    The state Department of Transportation did not adequately
          follow HPMS guidance in 1990, resulting in poor quality
          traffic counts;

     -    The state has made substantial progress in preparing its
          1990 inventory using a network-based model and does not
          have the time and resources to switch approaches and
          still meet the Clean Air Act expectations for the
          November 1992, 1993, and 1994 SIP submittals.
___________________________

     195  If 1990 demographic and economic input data are used, a
model validated for 1987 or later could provide an acceptable VMT
estimate for 1990.

     196  EPA Regional Offices are advised not to agree that the
HPMS data are weak without consulting with divisional or regional
FHWA officials who have direct knowledge of the HPMS data
associated with the non-attainment area.  Desirable features of a
network model that is a candidate for SEP VMT estimation are that
a) the model is validated with 1990 ground counts and that b) the
model uses demographic inputs properly updated to 1990. (At a
minimum, population by zone should be updated to 1990; preferably
all socioeconomic variables would also be updated to 1990.)

     197  EPA Regional Offices are encouraged to consult with
divisional or regional FHWA officials who have direct knowledge of
the network-based model under consideration.

                                92





     An area using this method should make sure that all VMT in the
entire non-attainment area is included in the estimate.  Most
network-based models normally do not assign intra-zonal trips, VMT
on some local roads, or trips on functional classes outside of the
modeling area.  States may use any reasonable method to estimate
VMT on those functional systems that are within the non-attainment
area but that are not included in the model.

     States adopting this approach should realize that, beginning
in 1993, all urbanized areas with a population above 200,000 will
be required to conduct HPMS sample panels for individual FAUAs and,
therefore, will be required to estimate mobile source emissions in
such a way that the VMT estimates used by the state are consistent
with HPMS.  This means that there should be no reason for an ozone
area with poor HPMS data in 1990 not to base the "periodic
inventory" for 1993 and the Reasonable Further Progress tracking
inventory for 1996 on HPMS estimates, even if EPA accepted another
method for the 1990 inventory.  This switch in basis for the
inventory may reveal that the method used for 1990 was not
accurate, and it may be disadvantageous in terms of demonstrating
progress in emissions reduction.  A possible solution is for the
area to project 1990 VMT backwards from the higher quality 1993
HPMS figures and submit a revision of its 1990 inventory using this
revised estimate of 1990 VMT.  Where the disparity between 1990 and
1993 estimates by two different methods appears to EPA to
constitute an erroneous estimate of actual VMT changes, EPA may
require such backward projections from 1993 data.


                                93





                           Appendix 3-A
           Prescribed Volume Groups and Precision Levels

                                94





Click HERE for graphic.


                                95





Click HERE for graphic.


                                96





Click HERE for graphic.


     1/   Precision levels for individual urbanized areas.
     2/   Precision levels for collective urbanized areas.
     3/   For individual urbanized areas, use (70-15) precision
          level for States -with 3 or more individual urbanized
          areas.  Use (80-10) precision level for States with less
          than 3 individual urbanized areas.

                                97





                4.0  EMISSIONS FROM NONROAD SOURCES

4.1  Introduction

     Nonroad sources include motorized vehicles and equipment which
are normally not operated on public roadways to provide
transportation.  The study and regulation of nonroad emission
sources is mandated by the 1990 Clean Air Act.  Section 213(a) of
the 1990 Clean Air Act directs EPA to conduct a study of emissions
from nonroad engines and vehicles in order to determine if such
emissions cause, or significantly contribute to, air pollution
which may be reasonably anticipated to endanger public health or
welfare.  This study was completed in November 1991.198  The Clean
Air Act also requires EPA to regulate emissions from nonroad
engines and vehicles within 12 months after completion of the study
if EPA makes a determination that these sources are significant
contributors to concentrations of ozone or carbon monoxide (CO) in
more than one area which has failed to attain the National Ambient
Air Quality Standards (NAAQS) for these pollutants.

     This chapter summarizes and discusses information from several
studies, which EPA used to definite emission factors and
inventories for nonroad equipment.  It also discusses work updating
and expanding nonroad equipment inventories for use by state and
local agencies.  The nonroad inventories for the 24 non-attainment
areas in the November 1991 EPA nonroad report are presently being
updated to include areas that are within the recently redefined
nonattainment boundaries.  Inventories will also be compiled for
nine additional areas.  A sample inventory of nonroad equipment
populations and emissions in the New York-New Jersey area is
provided in this chapter (Appendix 4-A).  States should use this
chapter to guide them in the preparation of nonroad emission
inventories for use in determining needed VOC reductions and
developing implementation plans.  The ten nonroad equipment
categories are listed below.

     -    Lawn and Garden Equipment
     -    Agricultural Equipment
     -    Logging Equipment
     -    Light Commercial Equipment
     -    Industrial Equipment
     -    Construction Equipment
     -    Airport Service Equipment
     -    Recreational Equipment
     -    Recreational Marine Equipment
     -    Commercial Marine Vessels

___________________________

     198  U.S. Environmental Protection Agency.  Nonroad Engine
and Vehicle Emission Study, Report and Appendices, EPA-21A-2001,
Washington, D.C, Office of Air and Radiation, November 1991.

                                98





     A description of these categories, including the 79 equipment
types within them, can be found in Appendix B of this chapter.  In
general, each equipment type comes in three possible engine types:
diesel, 4-stroke, and 2-stroke.  However, for some types of
equipment, there are no two-stroke engines, but units fueled by
propane or CNG are in service.  For simplicity, propane and CNG
equipment is included in the 2-stroke engine type tables, but with
their correct emission factors when operated on gaseous fuel.

4.2  Inventory Options Under This Guidance

     In its nonroad report, entitled "Nonroad Engine and Vehicle
Emission Study," EPA developed two emission inventories for the
first nine categories listed above and a single inventory for the
tenth.199

     The first nine categories consist of nonroad engines and
vehicles for 24 areas.  Inventories for the last equipment category
(commercial marine vessels) are only approximate and available for
only six areas.

     For the first inventory, designated as Inventory A, EPA used
commercially and publicly available data so that the method could
be repeated by state agencies and other groups.  EPA used
confidential industry-supplied sales and other data, which are not
publicly available, for the second inventory, Inventory B. The
second inventory provided EPA with a cross check for the first
inventory results.  Both inventories agreed reasonably well.

4.2.1     Options for Areas With EPA Provided Inventories

     EPA has contracted with Energy and Environmental Analysis,
Inc. to update the nonroad equipment inventories based on new non-
attainment boundaries for ozone and carbon monoxide for the 24
areas in the nonroad report.  These areas, which were selected to
be geographically representative of areas with significant air
pollution problems, are defined as metropolitan statistical areas
(MSAs), consolidated metropolitan statistical areas (CMSAs), north
east county metropolitan areas (NECMAs), or air basins.  The exact
definitions of these terms can be found in the State and County
Metropolitan Area Data Book, U.S. Bureau of the Census, 1986. 
These areas are listed below, as presented in Table 1-03 of the EPA
nonroad report.
___________________________

     U.S. Environmental Protection Agency, Nonroad Engine and
Vehicle Emission Study,  Report and Appendices, EPA-21A-2001,
Washington, D.C., Office of Air and Radiation, November 1991.

                                99





     -    Atlanta, Georgia MSA
     -    Baltimore, Maryland MSA
     -    Baton Rouge, Louisiana MSA
     -    Boston-Lawrence-Salem-Lowell-Brockton, Massachusetts
          NECMA
     -    Chicago-Gary-Lake County, Illinois, Indiana, Wisconsin
          CMSA
     -    Cleveland-Akron-Lorain, Ohio CMSA
     -    Denver-Boulder, Colorado CMSA
     -    El Paso, Texas MSA
     -    Hartford-New Britain-Middletown-Bristol, Connecticut
          NECMA
     -    Houston-Galveston-Brazoria, Texas CMSA
     -    Miami-Fort Lauderdale, Florida CMSA
     -    Milwaukee-Racine, Wisconsin CMSA
     -    Minneapolis-St.  Paul, Minnesota-Wisconsin MSA
     -    New York-Northern New Jersey-Long Island, New York-New
          Jersey-Connecticut CMSA/NECMA
     -    Philadelphia-Wilmington-Trenton, Pennsylvania-New Jersey-
          Delaware-Maryland CMSA
     -    Provo-Orem, Utah MSA
     -    St. Louis, Missouri-Illinois MSA
     -    San Diego, California Air Basin
     -    San Joaquin, California Air Basin
     -    Seattle-Tacoma, Washington CMSA
     -    South Coast, California, Air Basin
     -    Spokane, Washington MSA
     -    Springfield, Massachusetts NECMA
     -    Washington, D.C.-Maryland-Virginia MSA

     Air quality non-attainment boundaries, established by the
Office of Air Quality Planning and Standards, for these and all
other areas are defined in the November 6, 1991, Federal Register
Notice, "Designation of Areas for Air Quality Planning
Purposes".200  The new boundaries sometimes divide up counties and
MSAs/CMSAs.  A county or MSA/CMSA was divided in this manner if a
state and/or local government could show that sources in a part of
the county or MSA/CMSA did not contribute significantly to
violations of the ambient standard.

     In addition, EPA is developing inventories for nine additional
areas.  With these additional areas, EPA will have determined
nonroad inventories for all moderate-2 or worse non-attainment
areas for CO (greater than 12.7 ppm), all the areas in serious or
worse nonattainment for ozone (16.0 ppm or greater), and several
miscellaneous areas.  These additional areas are listed below.
___________________________

     200  U.S. Environmental Protection Agency.  Designation of
Areas for Air Quality Planning Purposes, 40 CFR Part 81, Final
Rule, Washington, D.C., Office of Air and Radiation, November 6,
1991.

                                100





     -    Anchorage, Alaska
     -    Tucson, Arizona
               Pima County
     -    Las Vegas, Nevada
     -    Phoenix, Arizona, Maricopa County
     -    Muskegon, Michigan
     -    Portsmouth-Dover-Rochester Area, New Hampshire
               Rockingham County
               Stratford County
     -    Providence, Rhode Island
     -    Beaumont-Port Arthur, Texas
     -    Sheboygan, Wisconsin

4.2.2     Options For Areas With EPA Provided Inventories

     For the 33 areas listed above, states have several options as
listed below.  For the 24 areas originally included in the EPA
nonroad report, there are three EPA inventories that can be used.

     -    (A+B)/2 Inventory
     -    Inventory A
     -    Inventory B

     State and local agencies wanting to examine these inventories
should send in the request sheet at the end of this chapter for a
disk with the inventories in Lotus compatible files and a report
discussing them.  In general, the average of inventories A and B is
preferred since it utilizes all available input, both from the EPA
study and that provided by industry.  However, either the A or B
inventory may be used by either category or equipment type if local
agencies have data or knowledge to support its use over the average
of the two inventories.  With sufficient technical justification,
states may also move equipment types and emissions inventories
among counties within the non-attainment area in a way which they
believe to be more realistic.  AR such changes should be noted in
the documentation of the inventory.  Also, any changes that are
made should be kept within the given non-attainment area, unless
the state does a comprehensive analysis of the equipment type
across the entire CMSA/MSA.

     Limited data for commercial marine vessels can be found in the
EPA nonroad report for the six areas listed below.

     -    Baltimore           -    New York-New Jersey
     -    Baton Rouge         -    Philadelphia
     -    Houston-Galveston   -    Seattle/Tacoma

                                101





     The three types of commercial vessels include harbor, ocean-
going, and fishing vessels.  States may use the commercial marine
inventories found in the Booz, Allen, and Hamilton report for these
six areas, making and documenting whatever changes are needed to
include only partial county areas where appropriate.201  States
may also use the methodology in this report to create local
inventories.  Upon request, EPA will supply the states with the
Booz, Allen, & Hamilton report.  States may also use the
methodology in the previous edition of Volume IV, being certain
that commercial and recreational marine are handled as separate
categories.

4.2.3     Options For Areas Without EPA Provided Inventories

     For areas not among the 33 areas listed above, there are
several options.  The states may choose one of the 33 areas, which
is similar in terms of climate and economic activity, so that
emission inventories can be produced by applying the ratio of the
populations of the two areas.  States may opt to produce
inventories by doing a ratio of the populations at the county
level, choosing counties from several of the 33 areas if
appropriate.  Please use the form at the end of this chapter to
request one or more of the 33 non-attainment area inventories. 
States may also use the EPA methodology, using local data to
develop inventories for one or all of the nine categories.  States
may also apply the Energy and Environmental Analysis, Inc.
methodology themselves or obtain consultant services to do so.  If
for some reason one of the two methods discussed above cannot be
used for areas not among the 33 areas already inventoried, states
may use the 1989 Volume IV guidance.  However, this should only be
considered as an option of last resort.  Other approaches may also
be used if they have technical support at least equal to or better
than the EPA methodologies.

4.3  Explanation of EPA Provided Inventory

     Nonroad emission inventories have been calculated for 33
areas, including all areas that are serious and above for ozone and
moderate-2 and above for carbon monoxide.  The inventories can be
used in turn to develop the Area and Mobile Source (AMS) inputs for
the Aerometric Information and Retrieval System (AIRS), which is
discussed in Section 4.4. The AMS inputs can in turn be used to
recreate nonroad inventories.  Since the AMS methodology is the
official way to calculate and document the inventory, the use of
AMS inputs is needed.
___________________________

     201  Booz Allen & Hamilton, Inc.  Commercial Marine Vessel
Contributions to Emission Inventories, Final Report to
Environmental Protection Agency.  Los Angeles, California, October
7, 1991.


                                102





     Section 4.3.1 explains how the AMS inputs are derived from the
EPA inventories.  Section 4.3.2 explains the AMS inputs themselves
and how they can be used to calculate emission inventories.  The
general methodology used in developing the EPA inventories will be
explained later in Section 4.4.

4.3.1     Derivation of AMS Inputs

     The following general information is included in the overall
EPA nonroad inventory tables.

     -    Equipment population (for ozone and CO non-attainment
          areas)
     -    Emission factors (g/hp-hr, g/hr, g/gal, g/day)
     -    Hours/year (gallons/year in some cases)
     -    Average horsepower
     -    Average load
     -    Evaporative emission season (229 days)
     -    l/SAF202  (tons per year/tons per summer day)
     -    I/SAF (tons per year/tons per winter day)
     -    VOC tons per summer day (ozone non-attainment area)
     -    NOx tons per summer day (ozone non-attainment area)
     -    CO tons per summer day (ozone non-attainment area)
     -    CO tons per winter day (CO non-attainment area)
     -    Particulates tons per year
     -    Population (human) and tons/person

     Each non-attainment area is broken down into individual
counties and, where appropriate, portions of counties that are
included within the non-attainment boundaries (i.e., townships and
boroughs).  The methodology for developing the county and sub-
county data is summarized in Section 4.4. The above information is
given as needed for each county or sub-county, as well as the non-
attainment area as a whole.  The sub-county boundaries are defined
in the November 6, 1991, 40 CFR Part 81.203

     Emission factors are given for 79 equipment types and include
VOC, CO, NOx , and particulates.  VOC emissions are divided into
the following subcategories.
___________________________

     202  SAF stands for seasonal adjustment factor.  The columns
l/SAF are taken directly from the spread sheets used for the EPA
nonroad report.  Although, SAF itself is in units of tons per
seasonal day/tons per year and is used to multiply the annual
inventory to obtain seasonal tons/day.

     203  U.S. Environmental Protection Agency.  Designation of
Areas for Air Quality Planning Purposes, 40 CFR Part 81, Final
Rule, Washington, D.C., Office of Air and Radiation, November 6,
1991.


                                103





     -    exhaust
     -    evaporative
     -    crankcase
     -    refueling

     These emission factors are derived from EPA and other test
data with appropriate extrapolations to the different equipment
types.  Evaporative emissions presently include only diurnal
emissions assumed to occur for 229 days of the year.  Note that
particulate inventories are being calculated simply because the
information is available.  These calculations are only being done
for ozone and CO non-attainment boundaries, rather than the
specific boundaries for PM10.  The reason for this is that the
emphasis of the present work is on ozone and CO non-attainment.  A
sample inventory calculation using the given information is shown
in Appendix 4-C.

     It is important to note that the definition of VOC excludes
methane and ethane but includes formaldehyde and acetaldehyde. 
Instead of VOC, the 1991 EPA nonroad report contains total
hydrocarbons (THC, as measured by a flame ionization detector),
which includes methane and ethane but excludes formaldehyde and
part of acetaldehyde.  In the process of revising the emission
inventories for the original 24 areas and developing inventories
for the additional areas, THC has been converted into VOC for 4-
stroke and diesel equipment using a correction factor.  This is
being done in order for the data to be compatible with AMS, as
discussed in Section 4.3.2. For 2-stroke engines it is being
assumed that THC is equivalent to VOC (that is, a T`HC to VOC
correction factor of 1.00). One reason for this assumption is that
2-stroke emissions contain significant amounts of unburned fuel ,
for which THC and VOC are the same, compared to 4-stroke engines. 
Another reason is that no quantitative data exists on methane,
ethane, and aldehyde emissions from 2-stroke engines.

     The seasonal adjustment factor is provided to calculate winter
and summertime inventories for VOC, CO, NOx , and particulates. 
The annual inventories are adjusted so that the output is in tons
per summer day (tpsd) for the ozone non-attainment areas.  Emission
inventories for CO are also adjusted so that the output is in tons
per winter day (tpwd) for the CO non-attainment areas.

     For each CMSA, tables are available in hard copy (paper) or on
disk in a text or Lotus format.  States may use the order form in
the back of this chapter to request the data for the non-attainment
area of interest.

     Again, separate information is given for 2-stroke, 4-stroke,
and diesel engines.  All 79 equipment types spanning 9 nonroad
engine categories are included.  However, as mentioned earlier, the
commercial marine vessel category is not included.


                                104





     Another important point to make on the eventual application of
these data is that diesel, 4-stroke, and 2-stroke engine emissions
all have different exhaust HC specifications.  The 4-stroke
gasoline and diesel engines can use the standard EPA motor vehicle
specifications.  For the present time, we recommend that the
standard EPA specification for non-catalyst gasoline engines also
be used for 2-stroke and propane engines.  In addition, evaporative
emission specification differs from exhaust, consisting of more
volatile fuel components with no combustion products.  However, in
order to keep inventory specification simpler, the two categories
are being combined based on the relative amounts of exhaust and
evaporative emissions for the 2-stroke and 4-stroke engines.  EPA
will provide later guidance concerning the typical fraction for
combining the exhaust and evaporative specifications, or the user
can determine it from the detailed inventory tables.

4.3.2     AMS Inputs

     The AMS system has the following five mandatory inputs.

     -    Annual activity level (hp-hr, gal, hrs, day)
     -    Emission factor (g/hp-hr, g/gal, g/hr, g/day)
     -    Period throughput (% annual activity based on emission
          mass during 3 month period)
     -    Adjustment factors204 (weekday, Saturday, Sunday,
          operating fractions)
     -    Category operating parameters

     The first two inputs have to be compatible so that one can
calculate overall emissions.  The EPA nonroad report generally
gives exhaust emissions in g/hp-hr units but sometimes they are
given in g/gal and g/hr units.  Evaporative emissions are in g/day
but can be artificially converted into one of the above units by
knowing how many horsepower-hours, gallons, or hours of operation
occur each day.

     Item #1, the annual activity level for an equipment type, has
to be the sum of all activity within an area (that is, equipment
type population times activity per unit piece of equipment) rather
than activity per unit equipment type, as given in the EPA nonroad
report.  Therefore, these levels will be greatly different from one
area to another.  It is important to note that the activity levels
will be different for each county or portion of a county within a
non-attainment area due to the changes in equipment population from
one county to another.

     Item #2, emission factors, are directly available from the
November 1991 EPA nonroad study for CO and NOx .  Approximate
factors are available for particulates.  These factors are in g/hp-
hr, g/gal, and g/hr.  The underlying emission factors are the same
for all non-attainment areas.  However, as with motor vehicles,
there are separate emission factors for exhaust, crankcase,
evaporative (diurnal), and refueling VOC.  In addition, there will
___________________________

     204  Adjustment factors are not mandatory for calculations to
proceed in AMS, but should be provided if modeling will be done.

                                105





eventually be information on evaporative (hot soak), running loss,
and resting loss HC.  There is no nonroad emission factor model,
such as MOBILE 4.1, to combine exhaust and evaporative emissions
that are in different units.  The nonroad report multiplies each
emission factor by the appropriate activity level to compile the
composite VOC inventory, as illustrated in Appendix D of this
chapter.

     For development of the base year inventory and for AMS data
entry, a single composite emission factor for VOC, combining the
different sources of VOC (i.e., evaporative versus exhaust), is
being provided.  This composite factor is being created by dividing
the total VOC in annual tons by the activity levels.  The total
evaporative inventory, which is dependent only on equipment
population and days per year, is divided by activity levels (hp-hr,
hours, or gallons of fuel ) as one part of the emission factor. 
Since this particular fraction (evaporative inventory/activity
level) will vary from area to area, the overall Item #2 VOC
emission factors will vary slightly from one area to another.

     Item #3, period throughput, is the percent of annual activity
occurring in the 3 month summer and winter periods.  This is being
calculated from the EEA information for summer and winter.  The
tons per winter day and tons per summer day are being compared to
tons per year in deriving these factors.  The throughput is based
on emission tons rather than activity level.  Therefore, it is
compatible with the annual emission factors in calculating summer
and winter inventories.

     Item #4, adjustment factors, account for activity variation in
different periods of the week (weekdays, Saturdays, Sundays).  EPA
has no quantitative information here.  EPA recommends the same
treatment that is implicit in the November 1991 nonroad report,
assuming all nonroad activity is distributed equally over every day
of the week.  For this case, the adjustment factor is 1.0. This is
an approximate assumption and, in the aggregate, may be justified
because some categories (e.g., construction equipment) operate
mostly during the week while others (e.g., recreational equipment
and recreational marine equipment) operates largely on weekends. 
The states may use their own judgement or data to assign other
factors.

     Item #5, category operating parameters, are days/week and
weeks@year for equipment operation.  The states should assume that
the equipment operates seven days/week and 52 weeks/year.  This is
an artificial treatment which will make the calculations work out
properly, since the period throughput already accounts for
weeks/year because it is emission weighted.

     Not all of the detail needed to revise nonroad inventories, if
emission factors are revised at a later date, will be provided in
the AMS format.  Instead, inventory revisions would be made (if
decided by the states) by going back into the EEA input, which is
available separately, and altering specific items (such as
equipment populations, hours of

                                106





usage, and emission factors) and then recalculating the factors
used for AMS.  For this purpose, the specific EEA data discs for
the items given in Section 4.3. 1, which were used to calculate the
AMS inputs for the non-attainment areas that were examined, must be
used.

4.4  General Methodology Used In Deriving Emission Inventories For
     33 Areas

     This section explains the methodology developed and used by
EEA and EPA to update emission inventories for a given non-
attainment area from the 1991 EPA nonroad report.  Much of the
information in this section has been taken directly from an EEA
draft report entitled, "Methodology to Estimate Nonroad Engine and
Vehicle Emission Inventories at the County and Sub-County Level,"
which gives information primarily for the New York area and
describes the methodology planned for use for the other areas.205 
This methodology is frequently generic in nature with
extrapolations being made to smaller areas from equipment
populations in larger areas.  This might lead to some minor cases
of some equipment types being listed in areas where they would not
be expected to be found.  EPA and EEA will be reviewing the
inventories to try to screen out these anomalies.

     The following inputs are used to develop nonroad emission
inventories.

     -    The equipment populations in a given area;
     -    The annual hours of use of each type of equipment
          adjusted for geographic region and for the season of
          interest for each pollutant studied;
     -    The average rated horsepower of each type of equipment;
     -    The typical load factor for each type of equipment;
     -    An emission factor (EF).

     The emission factor is defined as the average emissions of
each pollutant per unit of use (gram/horsepower-hour) for each
category of equipment.  In order to calculate emission factors for
nonroad equipment, EPA compiled and used data from past tests and
studies, as well as new data supplied to EPA from engine
manufacturers.  Using these data, EPA developed emission factors
for tailpipe exhaust, refueling, evaporative, and crankcase
emissions with appropriate adjustments to account for in-use
emissions.  These emission factors can be found in Appendix I of
the EPA nonroad report.206
___________________________

     205  Energy and Environmental Analysis, Inc.  Methodology To
Estimate Nonroad Engine and Vehicle Emission Inventories At the
County and Sub-County Level, Draft Report to the Environmental
Protection Agency.  Arlington, Virginia, February 11, 1992.


     206  U.S. Environmental Protection Agency.  Nonroad Engine
and Vehicle Emission Study, Report and Appendices, EPA-21A-2001,
Washington, D.C., Office of Air and Radiation, November 1991.

                                107





     For the EPA nonroad report, EEA estimated equipment
populations by CMSA or NECMA using regression analysis of state
level populations and activity indicator statistics.207  EEA
relied on Power Systems Research (PSR) as a major source of data
for national and state equipment populations, annual hours of use,
load factors, horsepower, 2-cycle/4-cycle distributions, and
LPG/CNG penetrations.  Linear relationships were derived between
economic activity indicators and equipment category populations. 
County-level indicators were summed across each CMSA or NECMA and
these sums were plugged into the corresponding fitted regression to
arrive at equipment population estimates for each area in the
study.  However, the methodology from the EEA final report,
"Methodology to Estimate Nonroad Equipment Populations by
Nonattainment Areas," did not address developing inventories for
counties and sub-counties which EPA later calculated and included
within the boundaries of the 24 CO and ozone non-attainment areas
in the EPA nonroad report.

4.4.1     Explanation of Methodologies to Distribute Equipment
          Within Each Category Type at the County Level

     Adjustments to the emissions inventories contained in the EPA
nonroad report must be made through changes in the county
populations of nonroad equipment, since load factors, annual hours
of use, emission factors, average horsepower, 2 cycle/4 cycle
distribution, LPG/CNG penetration, and seasonality factors remain
essentially constant throughout an area and, in some cases, on a
state or national level.  In order to do this, activity indicators
used in the previous EEA analysis are used at the county level to
determine equipment populations by county and equipment type.  In
general, these indicators are derived from economic data presented
in various census publications.  By using regression analysis, it
can be determined whether there is a strong relationship between
specific activity indicators and an equipment category's state
population.  These regressions were done for the areas in the EPA
nonroad report208  If an activity indicator proved to have a
strong relationship with the equipment category of interest, it was
used to distribute the equipment types within a category at the
county level, using data supplied by Power Systems Research.  Shown
below is the general formula (equation 4-1) used to distribute
national equipment populations to the local county level unless
otherwise specified.

               County Activity Indicator x National Pop
County Pop =   _________________________________________      (4-1)

                    National Activity Indicator

___________________________

     207  Energy and Environmental Analysis, Inc.  Methodology To
Estimate Nonroad Equipment Populations By Nonattainment Areas,
prepared for U.S. Environmental Protection Agency, September 1991.

     208  U.S. Environmental Protection Agency.  Nonroad Engine
and Vehicle Emission Study, Report and Appendices, EPA-21A-2001,
Washington, D.C, Office of Air and Radiation, November 1991.

                                108





     Using logging activity (employees) as an indicator, logging
equipment populations were determined in the EPA nonroad report at
the county level by taking the ratio of county logging employment
over national logging employment and multiplying it by national
logging equipment populations.

     For the lawn and garden equipment category, the number of
single family housing units and the number of landscape and
horticultural service employees in the county of interest were used
to distribute equipment populations.  However, for chainsaws, which
are included in the lawn and garden category,EPA used a methodology
suggested by the Portable Power Equipment Manufacturers Association
(PPEMA) using a combination of inventories A and B.209  This
method was based on a combination of activity indicators, including
rural area population, urban population outside of major urbanized
areas, and landscaping/horticultural employment.  This methodology
is shown below in equation 4-2.


                    NB local
     NA local =    ____________  x  NA national             (4-2)

                    NB national

     In this formula, N refers to the number of chain saws (all
sizes), and A and B refer to inventories A and B.210

     Motorcycle dealers were used as an activity indicator in order
to distribute recreation equipment (e.g., nonroad motorcycles,
minibikes, golf carts, snowmobiles, and specialty vehicle carts)
from the state to the CMSA level.  However, this indicator does not
accurately reflect recreation equipment populations at the county
level.  Another problem is that there exists an inverse
relationship between population and recreational equipment (i.e.,
recreation equipment is not usually used in densely populated
areas).  Therefore, EEA developed another methodology to use at the
county level.  First, the mean population density (people per
square mile) is calculated for the MSA, CMSA, NECMA, or air basin
and all the counties within it.  Second, the standard deviation
(sigma) from the mean. is calculated using the individual county
data.  Third, those counties with a population density greater than
a certain amount (e.g., the mean minus sigma) are assigned an
equipment population of zero.  Fourth, the equipment populations
are then distributed among the remaining (lower population)
counties according to the following formula (equation 4-3).

     Cntypop = Areapop x [Area/Sum(Area)]                     (4-3)

___________________________

     209  U.S. Environmental Protection Agency.  Nonroad Engine
and Vehicle Emission Study, Report and Appendices, EPA-21A-2001,
Washington, D.C., Office of Air and Radiation, November 1991.

     210  Ibid.

                                109





     The variable Cntypop is the population of a particular
recreation equipment type in a county which has a human population
density less than that of the CMSA.  The variable Areapop is the
population of a recreation equipment type in the CMSA, of which the
county (Cnty) is a part.  The ratio within the brackets consists of
a particular county's area divided by the sum of the area for all
counties in the CMSA which have a population density less than that
of the CMSA.

     Air carrier operations were used as an activity indicator to
distribute airport service equipment populations in counties in
which an airport was located.  If more than one airport was located
in a county at which air carrier operations took place, the
relevant activity levels at these airports were summed
together.211

     For agricultural equipment, EPA decided to use the following
methodology, shown below in equation 4-4, for both the A and B
inventories .

                                          PSR national 
     County pop = Census county pop x   ____________________  (4-4)

                                        Census national pop

     Both census data and PSR data were used because census data
provided the best indicator of county distribution of agricultural
equipment (i.e., total number of units in the county), while the
PSR data gave a better estimate of equipment actually used
regularly in agricultural activity versus idle equipment.212

     In order to distribute light commercial equipment (<50 hp) to
the county level, the total amount of wholesale activity (number of
establishments) was used as an activity indicator.  EPA used the
number of employees in manufacturing at the state and county level
and regressed these data on PSR's state population for industrial
equipment.  For construction equipment, total construction activity
(number of employees) was used to determine county level
populations.  For industrial equipment, the number of people
employed in manufacturing was used as an activity indicator and was
used to distribute state level populations to the county level. 
For construction equipment, the number of employees involved in
total construction activity, including road construction, was
chosen to distribute state level construction equipment populations
to the county level.

     For recreational marine vessels, the nonroad report used local
boat registration data to ascertain the number of each marine
equipment type in each non-attainment area.  However, an adjustment
was made in the marine vessel data to derive the number of marine
engines.
___________________________


     211  U.S. Environmental Protection Agency.  Nonroad Engine
and Vehicle Emission Study, Report and Appendices, EPA-21A-2001,
Washington, D.C., Office of Air and Radiation, November 1991.

     212  Ibid.

                                110





This was done in order to match the vessel data to the horsepower
and hours of use estimates, which were calculated per engine.  The
second adjustment was done to determine how many engines were used
in the non-attainment areas.  Survey results provided by NMMA were
used in this adjustment for 8 of the 24 areas, which are listed
below.

     -    Baltimore      -    Boston
     -    Chicago        -    Denver
     -    Hartford       -    Houston
     -    Milwaukee      -    Seattle

     Data from the survey was used in the following formula
(equation 4-5).

                 Fuel Usedna
     EU = ER x __________________ + % Use Offshore            (4-5)

               Fuel Usedboats na

     The variable EU stands for the number of engines used in the
non-attainment area.  The variable ER is equal to the number of
engines registered in the non-attainment area Fuel Usedna is equal
to the sum of the reported amount of fuel consumed inside the non-
attainment area by boats registered inside the non-attainment area,
plus the fuel consumed within the non-attainment area by boats
registered outside the non-attainment area.  Fuel Usedboats na is
equal to the total reported amount of fuel consumed by boats
registered inside the non-attainment area without regard to where
the fuel was consumed.  The variable % Use Offshore is equal to the
percentage of boats used in coastal waters or the Great Lakes 0-1
mile from the shore.

     For the 16 non-attainment areas that NM]MA did not include in
their survey, the average ratios derived from the eight surveyed
areas were applied.  However, in some cases, the average ratios
were too large for non-attainment areas with lesser amounts of
navigable water area.  In these cases, a calculation was made of
the maximum number of boats that could be operated normally on the
available water area, which was supplied by EEA.  The area required
per boat was supplied by NMMA.  This calculation is shown in
equation (4-6).

                    Water Surface Areana
     Max.  Boats = ______________________                     (4-6)

                    Area Requiredboat

     Max.  Boats is equal to the theoretical maximum amount of
boats that can operate on the available water surface area.  Water
Surface Areana is equal to the total water surface area in the non-
attainment area.  Area Requiredboat is equal to the average water
surface area required per boat.

     The result of the above calculation multiplied by the
available hours of prime boating use (assumed nationwide), which is
384 hours/year (12 weeks/year x 4 days/week x 8 hours/day),
provided the theoretical maximum number of summer boat hours inside
the non-

                                111





attainment area, which was compared to the amount of summer boat
hours calculated from the survey results and the local boat
registrations.  In the cases where the summer boat hours calculated
from registrations and survey results exceeded the theoretical
maximum calculated using 384 hours/year, the calculated number of
engines used in the non-attainment area was reduced by the ratio of
the theoretical maximum summer boat hours to the calculated summer
boat hours.  Because this correction ratio does not include
offshore boat use, the average offshore use was subtracted out
prior to applying the correction ratio.  For areas on the ocean or
on a Great Lake, the average of the offshore usage proportion for
all the areas with offshore use was added back after applying the
correction ratio.  Note that recreational marine equipment was not
allocated to the county level in the November 1991 nonroad report.

4.4.2     Explanation of Methodologies For Distributing Equipment
          Within Each Category at the Sub-County Level

     When the non-attainment boundaries include only a portion of a
county, the nonroad equipment population at the county level must
be adjusted.  This adjustment can usually be made based on the
following fraction (equation 4-7).


                Human population at subcounty level
           ____________________________________________       (4-7)

                 Human population at county level


     This adjustment factor works well for nonroad equipment
populations that can be expected to track human populations
reasonable well in going from the county to subcounty level.  This
adjustment is used for the following nonroad equipment categories.

     -    lawn and garden
     -    light commercial
     -    industrial
     -    construction

     For other equipment categories, different methodologies were
used.

     The following equipment categories are adjusted from a county
to subcounty level using the same methodology, found in Section
4.4.1, for distributing recreation equipment from the CMSA, MSA,
NECMA, or Air Basin to the county level.

     -    recreational
     -    logging
     -    agricultural


                                112





     This methodology uses a mathematical relationship based on
population density to distribute the equipment population to
sparsely populated areas of a sub-county.  Equation (4-8) shows
this calculation below.

Subcnty = Cntyeqip pop. x [(Area)subcnty / (Area)cnty)]    (4-8)

     The variable Subcnty is a township or borough in the partial
county of interest with a population density that is less than that
of the CMSA.  Cntyequip pop. is the population of a particular
recreational equipment type within the county of which the partial
county is a part.  The ratio of Areasubcnty over Areacnty is the
ratio of the area of the partial county of interest to the area of
the total county.  In effect, this moves this equipment completely
from highly populated areas of the partial county, where this
equipment is not used much if at all, to the lower population areas
where available land exists.

     For recreational marine equipment, a similar method is used
for county to subcounty allocations of equipment population, except
that it does not limit the allocation of equipment to only partial
counties that have a population density less than that of the CMSA. 
This is because in many cases, recreational marine equipment is
used in bodies of water that are close to populated areas (e.g.,
the Potomac River in Washington, D.C.). Although this could be done
by inspecting maps and considering other factors, such as usage
patterns for the available water, it is quite difficult to
determine the water area available for recreational marine
equipment from a county to subcounty level.

     For airport service equipment, subcounty populations are
obtained simply by inspecting a map to see if the airport is in the
subcounty portion being considered.

4.4.3     Seasonal Adjustment Methodology

     EEA and EPA also applied seasonal adjustments to the emission
inventories based on annual hours of use data in eight regions of
the United States at different times of the year.213, 214  These
adjustments were also based on the fact that ozone violations
usually occur in the summer and CO violations usually occur in the
winter.  In Inventory A, EEA and EPA used eight seasonality
adjustments based on classifying each of the 24 original areas into
eight regions, which are listed below.
___________________________

     213  U.S. Environmental Protection Agency.  Nonroad Engine
and Vehicle Emission Study, Report and Appendices, EPA-21A-2001,
Washington, D.C., Office of Air and Radiation, November 1991.

     214  Energy and Environmental Analysis, Inc.  Methodology To
Estimate Nonroad Equipment Populations By Nonattainment Areas,
prepared for U.S. Environmental Protection Agency, September 1991.

                                113






     - Northeast              -    Southeast
     - Mid-Atlantic Coast     -    Great Lakes
     - Southwest              -    Rocky Mountains
     - Northwest              -    West Coast


     These regions were also classified into cold, medium and warm
regions, which were geographically similar to regions found in a
1973 report by Hare and Springer.215  The United States was
divided up into these three temperature regions based on average
January temperatures.  A warm region was classified as being above
45'F a medium region was classified as being between 35'F and 44'F
and a cold region was classified as being below 35'F.  For more
detailed information, please refer to Appendix L of the EPA nonroad
report.

     The formulas listed below (equations 4-9 and 4-10) were
applied using this information to determine summertime total HC and
NOx emissions and wintertime CO emissions.

     tpsd tpy  SAFsummer                                     (4-9)
     tpwd tpy  SAFwinter                                    (4-10)

     TPSD and TPWD refer to tons per summer and winter day,
respectively.  TPY stands for tons per year.  SAF is the seasonal
adjustment factor, which can be derived by the following formulas
(equations 4-11 and 4-12).

SAFsummer = 4  *  (% activity during summer/365 days)       (4-11)

SAEwinter = 4  *  (% activity during winter/365 days)       (4-12)

     The seasonal adjustment factors were calculated based on data
from the Hare and Springer report mentioned above, the CARB
Technical Support Document concerning lawn and garden equipment,
1987 SIP emission inventories, the Motorcycle Industry Council
(MC), and the National Marine Manufacturers Association (NMMA). 
Some seasonal activity percentages are listed in the table below,
which is the same as Table L-02, excluding
___________________________

     215  Hare, C.T., and K.J. Springer.  Exhaust Emissions from
Uncontrolled Vehicles and Related Equipment Using Combustion
Engines, Part 5, No. APRD-1494.  San Antonio, TX, Southwest
Research Institute, October 1973.

                                114





commercial marine equipment, in the EPA nonroad report. 
Recreational marine and recreation equipment seasonal activity
percentages can be found in tables L-03, and L-04 in Appendix L of
the EPA nonroad report.216


Equipment      Cold/North       Medium/Central         Warm,/South
Category     Summer   Winter    Summer   Winter      Summer  Winter

Agricultural   50%        6%       40%        6%       34%       6%

Construction   43%       10%       38%       15%       33%      20%

Industrial     30%       20%       25%       25%       25%      25%

Lawn & Garden  50%        6%       40%        6%       34%       6%
(excluding chainsaws)

Snowblower/     0%      100%        0%      100%       0%      100%
Snowmobiles


Airport        25%       25%       25%       25%       25%      25%
Service

Logging        25%       25%       25%       25%       25%      25%
(including chainsaws)

Light          25%       25%       25%       25%       25%      25%
Commercial


4.5  New York Non-attainment Area Example

     EEA has prepared an updated nonroad inventory for the New York
CMSA, including partial counties included in the non-attainment
boundaries.  Some example tables from the final report containing
the updated New York CMSA inventory for diesel, 4-stroke, 2-stroke,
and propane/CNG engine types can be found in Appendix 4-A at the
end of this chapter.  These tables include information for some of
the following items, most of which are in the spreadsheets for the
November 1991 nonroad report.  AR of these items are being listed
in the final EEA report for New York and the other areas.
___________________________

     216  U.S. Environmental Protection Agency.  Nonroad Engine
and Vehicle Emission Study, Report and Appendices, EPA-21A-2001,
Washington, D.C., Office of Air and Radiation, November 1991.

                                115






     -    Equipment population (for ozone and CO non-attainment
          areas)
     -    Emission factors (g/hp-hr, g/hr, g/gal, g/day)
     -    Hours/year
     -    Average horsepower
     -    Average load
     -    Evap season (days)
     -    1/SAF217 (tons per year/tons per summer day)
     -    1/SAF (tons per year/tons per winter day)
     -    VOC - tons per summer day
     -    NOx - tons per summer day
     -    CO - tons per summer day (ozone area)
     -    CO tons per winter day (CO area)
     -    Particulates tons per year
     -    Population (human) and tons/person

     Table 4-1 lists the 26 counties and partial counties included
in the New York CMSA, as well as the air quality classification of
each county or partial county for CO and ozone.  Tables 4-2 through
4-4 present nonroad inventory data for the New York CMSA using the
(A+B)/2 inventories.  Table 4-2 presents emission inventories for
diesel equipment.  Table 4-3 presents emissions per person for
diesel equipment.  Table 4-4 shows the AMS input parameters for
diesel equipment.
___________________________

     217  Seasonal Adjustment Factor.

                                116





                           Appendix 4-A
            Emission Inventory Tables For New York CMSA

                                117





                             Table 4-1

              Counties and Sub-County Areas Included
          In the New York CMSA Example

                               Ozone                       CO      
     County               Classification             Classification

Bergen, NJ                 Nonattainment              Nonattainment

Essex, NJ                  Nonattainment              Nonattainment

Hudson, NJ                 Nonattainment              Nonattainment

Hunterdon, NJ              Nonattainment              Nonattainment

Middlesex, NJ                                         Nonattainment
     City of
     Perth Amboy           Nonattainment

Monmouth, NJ                                          Nonattainment
     Borough of
     Freehold              Nonattainment

Morris, NJ                                            Nonattainment
     City of
     Morristown            Nonattainment

Ocean, NJ                                             Nonattainment
     City of Toms
     River                 Nonattainment

Passaic, NJ                                           Nonattainment
     Clifton City          Nonattainment
     Patterson City        Nonattainment
     Passaic City          Nonattainment

Somerset, NJ                                          Nonattainment
     Borough of
     Sommerville           Nonattainment

Sussex, NJ                 Nonattainment              Nonattainment

                                118





                        Table 4-1 continued

                               Ozone                       CO      
     County               Classification             Classification

     Union, NJ             Nonattainment              Nonattainment

     Bronx, NY             Nonattainment              Nonattainment

     Kings, NY             Nonattainment              Nonattainment

     Nassau, NY            Nonattainment              Nonattainment

     New York, NY          Nonattainment              Nonattainment

     Orange, NY            Nonattainment                 Attainment

     Putnam, NY            Nonattainment                 Attainment

     Queens, NY            Nonattainment              Nonattainment

     Richmond, NY          Nonattainment              Nonattainment

     Rockland, NY          Nonattainment                 Attainment

     Suffolk, NY           Nonattainment                 Attainment

     Westchester, NY       Nonattainment              Nonattainment

     Fairfield, CT         Nonattainment              Nonattainment

     Litchfield, CT        Nonattainment
          Bridgewater Town                            Nonattainment
          New Milford Town                            Nonattainment
          Bethlehem Town                              Nonattainment
          Thomaston Town                              Nonattainment
          Watertown                                   Nonattainment
          Woodbury Town                               Nonattainment

     New Haven, CT         Nonattainment              Nonattainment

                                119





Click HERE for graphic.


                                120





Click HERE for graphic.


                                121





Click HERE for graphic.


                                122





Click HERE for graphic.


                                123





                           Appendix 4-B
                       Equipment Categories

                                124





     Category 1, lawn and garden equipment, includes 14 types of
equipment.218, 219  These are mostly powered by small gasoline
engines having less than 25 horsepower.  Examples of these are
listed below.

     -    lawn mowers
     -    trimmers
     -    edgers
     -    brush cutters
     -    chainsaws

     Some larger lawn and garden equipment types, such as those
listed below, have diesel engines.

     -    Chippers/grinders
     -    rear engine riding mowers
     -    wood splitters
     -    commercial turf equipment

     The main source of data used by EPA in its nonroad report to
derive the emission factors for gasoline engines in this category
was the California Air Resources Board (CARB) Technical Support
Document (TSD).220  CARB relied on testing done by manufacturers,
Southwest Research Institute, and Heiden Associates for the
Portable Power Equipment Manufacturers Association (PPEMA).  Since
there were no emissions data available for the small percentage of
lawn and garden equipment that have diesel engines (rear engine
riding mowers, lawn and garden tractors, wood splitters/chippers,
stump grinders, and commercial turf equipment), the emission
factors for diesel light commercial equipment (less than 50
horsepower) were considered to be the substitutes for use in the
nonroad equipment study.  Please refer to tables I-3 and I-4 in the
appendix of the nonroad report for the actual emission factors. 
The activity indicators used by Energy and Environmental Analysis,
Inc. (EEA) for developing lawn and garden equipment populations
included the number of single family housing units in a given area
and SIC 078 -Landscape and Horticultural Services (Employees).
___________________________

     218  U.S. Environmental Protection Agency.  Nonroad Engine
and Vehicle Emission Study, Report and Appendices, EPA-21A-2001,
Washington, D.C., Office of Air and Radiation, November 1991.

     219  Energy and Environmental Analysis, Inc.  Methodology To
Estimate Nonroad Equipment Populations By Nonattainment Areas,
prepared for U.S. Environmental Protection Agency, September 1991.

     220  California Air Resources Board.  Technical Report
Documents for California Exhaust Emission Standards and Test
Procedures for 1994 and Subsequent Model Year Utility and Lawn and
Garden Equipment Engines, attachment C to CARB Mailout #90-64, El
Monte, California: State of California, December 1991.

                                125





     The agricultural equipment category is comprised of 11 types
of equipment.  Examples of these are listed below.

     -    tractors                           -    balers
     -    combines                           -    tillers
     -    swathers                           -    sprayers
     -    fertilizer spreaders               -    harvesters
     -    agricultural mowers (> 5 h.p.)     -    Strippers
     -    cotton pickers

     Some specialized equipment, such as cotton pickers and
strippers have relatively small populations and can only be found
in certain areas of the country.  For agricultural equipment using
gasoline, the emission factors used to calculate emission
inventories were from the Fourth Edition of AP-42.221  The factors
found in AP-42 were selected because no other sources had specific
emission factors by equipment type for gasoline nonroad equipment. 
For particulate emission factors for gasoline equipment, a value of
1.64 lb./1000 gallons was used.  For diesel agricultural equipment,
emission factors were taken from a report by Environmental Research
and Technology, Inc. (EAT) entitled, "Feasibility, Cost, and Air
Quality Impact of Potential Emission Control Requirements on Farm,
Construction, and Industrial Equipment in California".222 
Emission factors were given by equipment category.  Emission
factors for particulate matter were taken from AP-42, Fourth
Edition.  Table I-07 in the nonroad report presents the chosen
emission factors from EAT in grams/horsepower-hour, and table I-08
gives the emission factors converted to pounds/1000 gallons of
fuel.  EEA used data from the 1987 Census of Agriculture to derive
activity, indicators for agricultural equipment.  SIC 07 -
Agricultural Services (Employees), modified to exclude SIC 78 (used
for the lawn and garden category), was found to be the best
activity indicator for this category.
___________________________

     221  U.S. Environmental Protection Agency.  Compilation of
Air Pollutant Emission Factors, Volume II, EPA Report No. AP-42,
Fourth Edition, Research Triangle Park, North Carolina, Office of
Air Quality Planning and Standards, September 1985.

     222 Environmental Research and Technology, Inc.  Feasibility,
Cost, and Air Quality Impact of Potential Emission Control
Requirements on Farm, Construction, and Industrial Equipment in
California, Document PA841, sponsored by the Farm and Industrial
Equipment Institute, Engine Manufacturers Association, and
Construction Industry Manufacturers Association, May 1982.

                                126





 The logging equipment category includes the following equipments
types.

     -    chainsaws (> 5 h.p.)
     -    shredders (> 5 h.p.)
     -    skidders
     -    delimbers
     -    tellers
     -    bunchers
     -    other miscellaneous equipment

     Significant amounts of logging equipment can only be found in
parts of the United States where large-scale logging operations
take place, such as the Pacific Northwest.  Emission factors for
this category were taken from the CARB Technical Support Document
and data submitted to EPA by the Engine Manufacturers Association
(EMA).223   SIC 241 (Employees) was chosen by EEA as the best
activity indicator to distribute logging equipment populations.

     Light commercial equipment is categorized as having engines
under 50 horsepower.  This equipment is used in various
wholesaling, retailing, and manufacturing capacities.  Examples of
this category include various types of the following equipment.

     -    electrical generators
     -    pumps
     -    compressors

     Emission factors recommended by SwRI224 and contained in a
report produced by Radian Corporation225 were used for equipment
fueled with diesel in this category.  For equipment using gasoline,
emission factors for utility and lawn and garden equipment from the
CARB Technical Support Document were used.

     Somewhat related to the light commercial equipment category,
the industrial equipment category includes equipment used in
manufacturing and warehousing operations.  This category includes
the types of equipment listed below.
___________________________

     223  Ingalls, M.N. Nonroad Emission Factors of Air Toxics,
Report No. 08-3426-005.  San Antonio, Texas, Southwest Research
Institute, February 1991.

     224  Weaver, C.S. Feasibility and Cost Effectiveness of
Controlling Emissions from Diesel Engines in Rail, Marine,
Construction, Farm, and Other Mobile Off-Highway Equipment, Final
Report for U.S. Environmental Protection Agency.  Sacramento,
California, Radian Corporation, February 1988.

     225  Hare, C.T., and K.J. Springer.  Exhaust Emissions from
Uncontrolled Vehicles and Related Equipment Using Internal
Combustion Engines, Final Report Part 5, Heavy-Duty Farm,
Construction, and Industrial Engines.  San Antonio, Texas,
Southwest Research Institute, October 1973.

                                127





     -    forklifts           -    vacuums
     -    boom lifts          -    scrapers
     -    scissor lifts       -    Winches
     -    industrial sweepers -    hoists
     -    scrubbers           -    conveyors

 blowers

     For both gasoline and diesel industrial equipment, emission
factors found in Volume I of AP-42 were used.  These factors were
developed by SwRI in 1973.226  EEA use total wholesale activity
(number of establishments) as the indicator for the distribution of
light commercial equipment populations.

     The category of construction equipment includes 27 kinds of
equipment.227, 228  These types are listed below.

     -    paving equipment    -    cranes
     -    roofing equipment   -    cement/mortar mixers
     -    signal boards       -    crushing/processing equipment
     -    cable layers        -    dozers
     -    drilling rigs       -    backhoes
     -    excavators          -    loaders
     -    industrial saws     -    tractors

     The emission factors used for diesel construction equipment
were derived from data from EMA.  For some types of diesel
equipment, EMA emission factors were unavailable.  In these cases,
EPA used factors from AP-42, Fourth Edition, which were originally
derive from the EAT report mentioned earlier.229  For the actual
emission factors, please refer to Table I-09, which compares the
AP-42 and the EMA emission factors, in the appendix of the EPA
nonroad report.  For gasoline construction equipment, the emission
factors which EPA selected came from the fourth edition of AP-42. 
For particulate matter and aldehyde
___________________________

     226  California Air Resources Board, Mailout #90-58.  El
Monte, California, State of California, September 7, 1990.

     227  U.S. Environmental Protection Agency.  Nonroad Engine
and Vehicle Emission Study, Report and Appendices, EPA-21A-2001,
Washington, D.C., Office of Air and Radiation, November 1991.

     228  Energy and Environmental Analysis, Inc.  Methodology To
Estimate Nonroad Equipment Populations By Nonattaimment Areas,
prepared for U.S. Environmental Protection Agency, September 1991.

     229  Environmental Research and Technology, Inc. 
Feasibility, Cost, and Air Quality impact of Potential Emission
Control Requirements on Farm, Construction. and Industrial
Equipment in California, Document PA841, sponsored by the Farm and
Industrial Equipment Institute, Engine Manufacturers Association,
and Construction Industry Manufacturers Association, May 1982.

                                128





construction equipment emission factors, EPA used the gasoline and
diesel agricultural equipment emission factors.  Local construction
activity was used by EEA to determine local construction equipment
populations.

     For airport equipment, only equipment owned by each airline
was included (EEA).  Airport equipment owned and operated by the
airport authority was inventoried within the other categories. 
This was done to prevent double counting.  This category also does
not include aircraft engines, which are addressed in Chapter 5.
Examples of this equipment are listed below.

-    load lifters                  -    starting units
-    de-icing equipment            -    baggage conveyors
-    heating units                 -    towing/pushback tractors
-    utility service equipment     -    misc. service vehicles

     The emission factors for industrial equipment were also
applied to airport service equipment.  Air carrier operations,
including certified carriers, air taxis, supplemental air carriers,
commercial operators of large aircraft, and air travel clubs, were
used by EEA as the activity indicator for determining local airport
equipment populations.

     The recreational category includes a varied array of
equipment.  Examples of these are listed below.

     -    all terrain vehicles (ATVs)
     -    minibikes
     -    snow and ice maintenance equipment,
     -    go-carts
     -    gasoline powered golf carts
     -    snowmobiles
     -    industrial personnel carriers/ATVs

     Some equipment in this category is limited to certain specific
areas or regions.  Golf carts are mostly found in resorts and golf
courses.  Snowmobiles and snow/ice maintenance equipment are mostly
found in areas that have a significant amount of snowfall. 
Emission factors developed by CARB for off-road motorcycles with
both 2- and 4-stroke engines were used by EPA.  EPA also applied
these emission factors to all-terrain vehicles, minibikes, golf
carts, and specialty vehicle carts.230  For snowmobiles, very
little data exist on emission rates.  EPA is considering using
emission factors found in AP-42, which were derived from
___________________________

    230 HAM, C.T., and K.J. Springer.  Exhaust Emissions from
Uncontrolled Vehicles and Related Equipment Using Internal
Combustion Engines, Final Report, Part 7, Snowmobiles, San Antonio,
Texas, Southwest Research Institute, April 1974.


                                129





testing done by SwRI in 1974.231  SIC 557 - Motorcycle Dealers
(Establishments) was the primary indicator used by EEA to determine
local recreational equipment populations.  For some areas SIC 557
had no data.  In such cases, SIC 55 - Automotive Dealers and
Service Stations (Employees) - was used as a substitute.

     Five types of recreational marine vessels are addressed in the
nonroad report.  These are listed below.

     -    vessels with inboard engines
     -    vessels with outboard engines
     -    vessels with stern drive engines
     -    sailboats with auxiliary outboard engines
     -    sailboats with auxiliary inboard engines

     Emission factors for outboard engines were derived from test
data supplied to EPA by the National Marine Manufacturers
Association, which tested 25 two-stroke and three four-stroke
outboard engines.  Please refer to Tables I-IIa and I-IIb in the
November 1991 EPA "Nonroad Engine and Vehicle Emission Study" for
further information.  For four-stroke outboards, emission factors
recommended by SwRI were used for particulate matter emissions.232
Since no data were available for 2-stroke outboard engine
particulate matter emissions, EPA used emission factors from the
CARB Technical Support Document for utility and lawn/garden
equipment as approximations.232  For inboard/stern drive gasoline
engines, EPA derived emission factors on the basis of test data on
three 4-stroke gasoline marine inboard/stern drive engines supplied
by NMMA (See table I-IIc in the appendix of the EPA nonroad
report).  The particulate emission factor used was 1.64 lb./1000
gal (0.74 grams/gallon).  Please refer to Section 1.2.2 of Appendix
I of the EPA November 1991 nonroad study for more information.  EPA
used test data on a small diesel sailboat inboard and three large
diesel inboard engines, which NMMA supplied, as the basis for
calculating emission factors for inboard diesel engines.  Please
refer to table I-I Id in appendix of the EPA nonroad study for more
information.  The activity indicator that EEA used to distribute
marine engine populations at the county level consisted of taking a
ratio of the water surface area of the given county to the total
water surface area of the state in which the county is located. 
This includes miles of public beach with an assumed operating
distance from shore of one mile, as well as inland waterways.  Data
on miles of public beach were found in the
___________________________

     231  Booz Allen & Hamilton, Inc.  Commercial Marine Vessel
Contributions to Emission Inventories, Final Report to
Environmental Protection Agency.  Los Angeles, California, October
7, 1991.

     232  U.S. Environmental Protection Agency.  Designation of
Areas for Air Quality Planning Purposes, 40 CFR Part 81, Final
Rule, Washington, D.C., Office of Air and Radiation, November 6,
1991.

     233  Energy and Environmental Analysis, Inc.  Methodology To
Estimate Non-road Engine and Vehicle Emission Inventories At the
County and Sub-County Level, Draft Report to the Environmental
Protection Agency.  Arlington, Virginia, February 11, 1992.

                                130





National Oceanic and Atmospheric Administration's (NOAA) National
Estuarine Inventory: Data Atlas, 1988.  Data on inland water
covered surface area were derived from a census report entitled,
Area Measurements Reports, GE-20.  No. 1, 1970.

     Commercial marine vessels can be subdivided into three
categories, including oceangoing, harbor, and fishing vessels. 
These vessels have similar characteristics of size, speed, engine
design, and distance traveled.  Booz Allen & Hamilton developed the
commercial marine vessel inventories and emission factors under
contract to EPA for the EPA nonroad study.  These are contained in
The Booz Allen & Hamilton final report.234  In addition, the
emission factors are contained in tables I-12a and I-12b in
Appendix I of the EPA nonroad report.  In developing commercial
marine populations for six ports (Baltimore, Baton Rouge, Houston-
Galveston, New York-New Jersey, Philadelphia, and Seattle-Tacoma),
Booz Allen & Hamilton requested data from several sources,
including port authorities, Lloyds exchange, local marine
exchanges, bar pilots associations, maritime trade organizations,
state regulatory and licensing boards, and the U.S. Army Corps of
Engineers.  For the ports of Houston, Galveston and Baton-Rouge,
inventories were based on data published in Waterborne Commerce of
the United States, Calendar Year 1988.235  For fishing vessel
populations, data supplied by the National Marine Fisheries Service
was used.236  For ports in non-attainment areas not addressed in
the Booz Allen & Hamilton report, EPA relied on data from SIP
inventories and the 1985 National Emissions Report.237
___________________________

     234  Booz Allen & Hamilton, Inc.  Commercial Marine Vessel
Contributions to Emission Inventories, Final Report to
Environmental Protection Agency.  Los Angeles, California, October
7, 1991.

     235  U.S. Army Corps. of Engineers.  Waterborne Commerce of
the United States, Calendar Year 1988, Water Resources Support
Center.

     236  U.S. Department of Commerce, National Oceanic and
Atmospheric Administration, National Marine Fisheries Service. 
Fisheries of the United States, 1990, Washington D.C., U.S.
Government Printing Office, May 1991.

     237  U.S. Environmental Protection Agency. 1985 National
Emissions Report, Research Triangle Park, NC, Office of Air Quality
Planning and Standards, March 1991.

                                131





                           Appendix 4-C
                   Sample Inventory Calculation

                                132





         Sample OMS and EEA Inventory Calculation for VOC


Exhaust

     (g/hp-hr) x (rated hp) x (percent of rated hp typically used) 
x (hrs/yr per engine) x (equipment population) x (ton/g)
     = (tons/yr)

Evaporative

     (g/day) x (229 days/yr)* x (equipment population) x (ton/g)
     = (tons/yr)

     Evaporative diurnal VOC emissions are assumed to occur 229
days/year, which includes each day of the ozone season, no winter
days, and most other days.  Future OMS work will determine emission
factors for hot soak, running loss, and resting loss emissions.

Refueling

     Refueling emissions are the sum of the spillage and vapor
displacement emissions.  They are calculated on a g/gal basis and
then, as appropriate, converted to a g/hp-hr or g/hr basis to be in
the same unit as the exhaust emission factors.  Please see Appendix
I (pages 26-3 1) of the EPA nonroad study for details on the
calculations.  The example below is shown for a g/gal emission
factor.

     (g/gal) x (gal fuel used) x (equipment population) x (ton/g)
     = (tons/yr)

     The total of these three emission categories are combined to
give total annual VOC emissions as shown below.

Total annual VOC emissions =
     Exhaust VOC + Evaporative VOC + Refueling VOC

     Annual emissions for CO and particulates are calculated using
just the exhaust portion of this example.

                                133





     The total annual emissions (tons/year) can be apportioned to
tons/summer day or tons/winter day by use of one of the conversion
factors developed by EEA.

     SAF (tons per summer day/tons per year)
     SAF (tons per winter day/tons per year)

     These factors consider different fractions of summer and
winter operations for the nine equipment categories in three
different geographical areas of the country (northern, central, and
southern).  Thus, the factors vary from area to area.

     By use of these conversion factors, the following inventories
are obtained.

     VOC tons per summer day
     NOx tons per summer day
     CO tons per summer day (O3 area)
     CO tons per winter day (CO area)
     Particulate tons per summer day
     Particulate tons per winter day

     The conversion factors can also be applied to particulates if
one is considering the 24-hr. particulate National Ambient Air
Quality Standard instead of the annual average.

                                134





                           Appendix 4-D
                      Inventory Request Form

                                135





                    5.0 EMISSIONS FROM AIRCRAFT

     This chapter describes the procedure for calculating emissions
from civilian and military aircraft within an inventory area.  The
basic methodology determines aircraft fleet make-up mid level of
activity and then calculates air pollutant emissions on an annual
basis.  Variations to the methodology, which account for seasonal
changes or specific operational considerations, are discussed. 
Changes expected in the fleet in the future and the effect on
emissions are also briefly described.  Finally, a method for
converting total hydrocarbon (THC) emissions to volatile organic
compound (VOC) emissions is presented at the end of the chapter.

     The inventory methodology and emission factors have been
updated since the last edition of this report.  This chapter also
updates the emission factor information that appears in Compilation
of Air Pollutant Emission Factors, Fourth Edition and Supplements. 
AP-42.238  Subsequent to the publication of this document, AP-42
will be formally updated and may include some additional data,
primarily on general aviation and military aircraft, which was
unavailable when this report was prepared.

5.1  OVERVIEW OF THE INVENTORY METHODOLOGY

     Preparing an emissions inventory for aircraft focuses on the
emission characteristics of this source relative to the vertical
column of air that ultimately affects ground level pollutant
concentrations.  This portion of the atmosphere, which begins at
the earth's surface and is simulated in air quality models, is
often referred to as the mixing zone.  The aircraft operations of
interest within this layer are defined as the landing and takeoff
(LTO) cycle.  The cycle begins when the aircraft approaches the
airport on its descent from cruising altitude, lands, and taxis to
the gate.  It continues as the aircraft taxis back out to the
runway for subsequent takeoff and climbout as it heads back up to
cruising altitude.  Thus, the five specific operating modes in an
LTO are:

     -    Approach
     -    Taxi/idle-in
     -    Taxi/idle-out
     -    Takeoff
     -    Climbout

___________________________

     238  Compilation of Air Pollutant Emission Factors, Volume
II: Mobile Sources, AP-42, U.S. Environmental Protection Agency,
Ann Arbor, Michigan, September, 1985. (Aircraft data from February
1980.)

                                137






Most aircraft go through a similar sequence during a complete
operating cycle.  Helicopters may combine certain modes such as
takeoff and climbout.

5.1.1 Factors Affecting Emissions

     The LTO cycle provides a basis for calculating aircraft
emissions.  During each mode of operation, the aircraft engines
operate at a fairly standard power setting for a given aircraft
category.  Emissions for one complete cycle for a given aircraft
can be calculated by knowing emission factors for specific aircraft
engines at those power settings.  Then, if the activity of all
aircraft in the modeling zone can be determined for the inventory
period, the total emissions can be calculated.  Each of the
dominant factors that affect the emissions from this source is
discussed below.

5.1.1.1   Aircraft Categorization

     For a single LTO cycle, aircraft emissions vary considerably
depending on the category of aircraft and the resulting typical
flight profile.  Aircraft can be categorized by use.  Commercial
aircraft include those used for scheduled service transporting
passengers, freight, or both.  Air taxis also fly scheduled service
carrying passengers and/or freight but usually are smaller aircraft
and operate on a more limited basis than the commercial carriers. 
Business aircraft support business travel, usually on an
unscheduled basis, and general aviation includes most other non-
military air-craft used for recreational flying, personal
transportation, and various other activities.

     For the purpose of creating an emissions inventory, business
aircraft are combined with general aviation aircraft because of
their similar size, use frequency, and operating profiles.  In this
inventory methodology they are referred to simply as general
aviation.  Similarly, air taxis are treated much like the general
aviation category because they are typically the same types of
aircraft.  Military aircraft cover a wide range of sizes, uses, and
operating missions.  While they often are similar to civil
aircraft, they are handled separately because they typically
operate exclusively out of military air bases and frequently have
distinctive flight profiles.  Helicopters, or rotary wing aircraft,
can be found in each of the categories.  Their operation is
distinct because they do not always operate from an airport but may
land and takeoff from a heliport at a hospital, police station, or
similarly dispersed location.  Military rotorcraft are included in
the military category and non-military rotorcraft are included in
the general aviation category since information on size and number
are usually found in common sources.  However, they are combined
into a single group for calculating emissions since their flight
profiles are similar.


                                138





     Commercial aircraft typically are the largest source of
aircraft emissions.  Although they make up less than half of all
aircraft in operation around a metropolitan area their emissions
usually represent a large fraction of the total because of their
size and operating frequency.  This may not hold true, of course,
for a city with a disproportionate amount of military activity or a
city with no major civil airports.

5.1.1.2   Pollutant Emissions

     Aircraft pollutants of significance are hydrocarbon (HQ,
carbon monoxide (CO), oxides of nitrogen (NOx), sulfur dioxide
(SO2), and particulates (PM10).  The factors that determine the
quantity of pollutant emitted are the emission index for each
operating mode (pounds of pollutant per 1000 pounds of fuel
consumed), the fuel consumption rate, and the duration of each
operating mode.  HC and CO emission indexes are very high during
the taxi/idle phases when aircraft engines are at low power and
operate at less than optimum efficiency.  The emission indexes fall
as the aircraft moves into the higher power operating modes of the
LTO cycle.  Thus, operation in the taxi/idle mode, when aircraft
are on the ground at low power, is a significant factor in
calculating total HC and CO emissions.  For areas which are most
concerned about the contribution of aircraft to the inventory of HC
and CO, special attention should be paid to the time the aircraft
operate in the taxi/idle modes.

     NOx emissions, on the other hand, are low when engine power
and combustion temperature are low but increase as the power level
is increased and combustion temperature rises.  Therefore the
takeoff and climbout modes have the highest NO. emission rates.  If
NO. is a primary concern for the inventory area, special effort
should focus on determining an accurate height of the mixing layer,
which affects the operating duration of climbout.

     Sulfur emissions typically are not measured when aircraft
engines are tested.  In evaluating sulfur emissions, it is assumed
that all sulfur in the fuel combines with oxygen during combustion
to form sulfur dioxide.  Thus, sulfur dioxide emission rates are
highest during takeoff and climbout when fuel consumption rates are
high.  Nationally the sulfur content of fuel remains fairly
constant from year to year at about 0.05% wt. for commercial jet
fuel , 0.025% wt. for military fuel , and 0.006% wt for aviation
gasoline.  This is the basis for the sulfur dioxide emission
indexes in the tables included in this methodology.  If the sulfur
content of fuel varies significantly on a local basis, the emission
index can be adjusted according to a ratio of the local value to
the national value.

     Particulates form as a result of incomplete combustion. 
Particulate emission rates are somewhat higher at low power rates
than at high power rates since combustion efficiency improves at
higher engine power.  However, particulate emissions are highest
during takeoff and climbout because the fuel flow rate also is
high.  It is particularly difficult to estimate the emissions of
this pollutant.  Direct measurement of particulate emissions from
aircraft engines

                                139





typically are not available, although emission of visible smoke is
reported as part of the engine certification procedure. 
Particulate emission factors for only a few aircraft engines are
included in this chapter.

5.1.1.3   Aircraft Engines

     The aircraft powerplant is the source of emissions of the key
pollutants that result from fuel combustion.  Emission rates vary
depending on the fuel consumption rate and engine specific design
factors.  In 1984, EPA established standards for HC emissions.  In
developing the emission limits, EPA defined an operating regimen to
standardize the engine certification testing procedure and method
for determining engine HC emissions.  The standard applies to jet
engines over 6,000 lbs-thrust and emissions are calculated based on
a specific LTO cycle.  EPA considered in-use engine deterioration
when the standards were developed but concluded that, because of
the high levels of maintenance of aircraft engines for reasons of
safety and fuel economy, emission performance would not deteriorate
significantly.  The operating parameters used in the standard for
the LTO cycle can be used as default values in calculating
emissions when more specific information is not known.  These
default values are defined in later sections of this methodology.

     When the standards went into effect, some engines in
production could already meet them due to design changes made
previously for improved fuel efficiency.  Other engines had to be
redesigned to reduce their HC emissions so that they could remain
in production.  In-service engines were not required to be
retrofitted in the normal course of periodic servicing and
rebuilding.  These older engines, many of which remain in service,
have HC emissions that exceed the standard.  New engine designs,
produced since the standards went into effect, have HC emissions
much lower than the standards.  As a result of design changes made
to the engines that meet the HC standard, emissions of CO also
generally went down while NOx emissions tended to increased. 
However, the change in these pollutants was much less dramatic than
the decrease in hydrocarbons.  The smoke number for the newer
engines also is lower due to specific design changes intended to
reduce smoke production, which is regulated by EPA.

5.1.1.4   Operating Modes

     During the LTO cycle, aircraft operate for different periods
of time in various modes depending on their particular category,
the local meteorological conditions, and operational considerations
at a given airport.  The "Time-In-Mode," or TIM, as used in this
methodology, takes these factors into consideration.  Table 5-1
shows representative LTO cycle times for several aircraft
categories.

                                140





Click HERE for graphic.


                                141





                  TABLE 5-1: DEFAULT TIME-IN-MODE
           FOR VARIOUS AIRCRAFT CATEGORIEs1(Concluded)


1   SOURCE:  AP-42, Compilation of Air Pollutant Emission Factors,
     Volume II: Mobile Sources, U.S. Environmental Protection
     Agency, Ann Arbor, Michigan, September, 1985. (Aircraft data
     from February 1980).

2   Civil aircraft data is for large congested metropolitan
     airports.

3   USAF - U.S. Air Force, USN - U.S. Navy.

4   Fighters and attack aircraft only.

5   Time-in mode is highly variable.  Taxi/idle out and in times
     as high as 25 and 17 minutes, respectively, have been noted. 
     Use local data base if possible.

6   Includes all turbine aircraft not specified elsewhere (i.e.,
     transport, cargo, observation, patrol, antisubmarine, early
     warning, and utility).

                                142





     Duration in approach and climbout depends largely on the local
meteorology.  Since the period of interest is during operation of
the aircraft within the air modeling zone, the inversion layer
thickness determines how long the aircraft is in this zone.  The
inversion layer thickness is also known as the mixing height or
mixing zone since the air in this layer is completely mixed and
pollutants emitted anywhere within the layer will be carried down
to ground level.  When the aircraft is above the mixing layer,
whether on descent or when climbing to cruising altitude, the
emissions tend to disperse, rather than being trapped by the
inversion, and have no ground level effect.

     Taxi/idle time, whether from the runway to the gate
(taxi/idle-in) or from the gate to the runway (taxi/idle-out),
depends on the size and layout of the airport, the amount of
traffic or congestion on the ground, and airport-specific
operational procedures.  Taxi/idle time is the most variable of the
LTO modes.  Taxi/idle time can vary significantly for each airport
throughout the day, as aircraft activity changes, and seasonally,
as general travel activity increases and decreases.

     The takeoff period, characterized primarily by full-throttle
operation, typically lasts until the aircraft reaches between 500
and 1000 feet above ground level when the engine power is reduced
and the climbout mode begins.  This transition height is fairly
standard and does not vary much from location to location or among
aircraft categories.

     This methodology describes techniques and data sources for
determining the critical variables in the inventory calculations. 
When an inventory is being created for a particular area, the fleet
make-up, aircraft activity, and times-in-mode will be specific to
that area.  Engine emission indexes, on the other hand, depend on
the engine design and are provided in reference tables.

     Where specific information may be difficult to obtain,
simplifying assumptions are discussed.  An automated (computerized)
calculation procedure, which can simplify data management, has been
developed by the Federal Aviation Administration (FAA) i support
from EPA and can be obtained from the FAA Technology Division,
Office of and Energy, 800 Independence Avenue, SW, Washington, DC
20591, (202) 267-8933.  The FAA Aircraft Engine Emission Database
(FAEED) includes information on the engines mounted on specific
aircraft with emission factors for each of the engines, in addition
to a menu-driven procedure for calculating an aircraft emissions
inventory.

                                143





5.2  INVENTORY METHODOLOGY

     The steps in the methodology are basically the same for each
aircraft classification and each location, even though several
factors used in creating an inventory are site specific.

     (1)  Identify all airports to be included in the inventory

     (2)  Determine the mixing height to be applied to the LTO
          cycle

     (3)  Define the fleet make-up for aircraft category using each
          airport

     (4)  Determine airport activity as the number of LTOs for each
          aircraft category

     (5)  Select emission indexes for each category

     (6)  Estimate a time-in-mode for each aircraft category at
          each airport

     (7)  Calculate an inventory based on the airport activity,
          TIM, and aircraft emission factors.

     For a specific region where an emissions inventory is being
created, steps one and two, the airports to be included and the
mixing height, will be determined largely by the assumptions used
in defining the scope of the modeling area.  Steps three through
six are repeated for commercial aircraft, general aviation,
military aircraft, and helicopters.  The primary difference in
creating an inventory for each type of aircraft is the references
used to determine the fleet make-up and activity.  The following
sections discuss each of these steps.  Steps one and two are
discussed in terms of the specific modeling area while steps three
through six are addressed together for each aircraft category.

5.2.1     Airport Selection

     Maps and regional information directories are good sources for
identifying civil airports and military air fields.  Sectional
aeronautical charts, published by the Aeronautical Charts
Distribution Division (C44), National Ocean Survey, NOAA,
Riverdale, MD 20840, (301) 436-6990 ($5.25 per map), particularly
show the location of large and small airports.  Specific airports
to be included will be limited by the geographic boundaries of the
modeling area.  A secondary reference is AOPA's Aviation USA239
which lists publicly and privately owned civil airports, including
heliports and seaplane bases, and locates them with directions
relative to specific cities, as well as providing latitude and
longitude coordinates.  Much like the sectional aeronautical
charts, this reference provides general information on all but a
few small landing strips.  These small air fields are unlikely to
be considered for most analyses because they have low activity,
typically can accommodate only small general aviation aircraft, and
therefore, contribute insignificantly to the emissions inventory.
(Many private
___________________________

     239  AOPA's Aviation USA, Aircraft Owners and Pilots
Association, 1990.

                                144





use landing sites are listed in AOPA's Aviation USA by city and
site name but a telephone number is the only information given). 
FAA Air Traffic Activity240 lists all airports with air traffic
control towers operated by the FAA.  While this is a subset of the
airports listed in these other references, all of the airports in
urban areas with significant air traffic are included.

5.2.2     Mixing Height Determination

     The height of the mixing zone influences only the time-in-mode
for approach and climbout.  This factor is significant primarily
when calculating NOx emissions rather than HC or CO.  If NOx
emissions are an important component of the inventory, specific
data must be gathered on mixing heights.  If NOx emissions are
unimportant, mixing height will have little effect on the results
and the default value of 3000 feet can be used for more generalized
results.

     Mixing height should be determined in conjunction with those
responsible for the air quality modeling of the region to insure
that assumptions used for creating different sections of the
overall inventory are consistent.  If the inventory is being
created independently of any air quality modeling, the mixing
height can be determined by contacting the National Meteorological
Center at (301) 763-8298 or alternatively the National Climatic
Data Center (NCDC) at (704) 259-0682.  Another source of mixing
height data is the EPA Office of Air Quality Planning and
Standards' SCRAM (Support Center for Regulatory Air Models)
Bulletin Board.241  This electronic date base contains data used
by various air quality models.  Mixing height data, which appears
under the Meteorological Data Main Menu, comes from the NCDC.  As a
third alternative, typical mixing heights can be found on Figures
5-1, 5-2, and 5-3 which come from Mixing Heights, Wind Speeds, and
Potential for Urban Air Pollution Throughout the Contiguous United
States.242  These figures, which show mixing height for a mean
annual morning, a mean summer morning, and a mean winter morning,
illustrate the seasonal variation in the mixing height.  The
morning data corresponds to the few hours centered near the morning
commuter rush hours, which roughly coincide with the diurnal
maximum concentration of slow-reacting pollutants in many urban
areas.  Figure 5-1, showing annual mixing heights, may be used for
creating an annual inventory.  If
___________________________

     240  FAA Air Traffic Activity, U.S. Department of
Transportation, Office of Management Systems, Federal Aviation
Administration, Fiscal Year 1989.  NTIS Report Number ADA 226063.

     241  U.S. Environmental Protection Agency, (SCRAM BBS).  To
access SCRAM BBS with a modem: (919) 541-5742 (XModem, 8 Bit
System, NO Parity, 1 Stop Bit).  Research Triangle Park, North
Carolina.

     242  Mixing Heights, Wind Speeds, and Potential for Urban Air
Pollution Throughout the Contiguous United U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina, January
1972.  NTIS Report Number PB 207103.

                                145





Click HERE for graphic.


                                146





Click HERE for graphic.


                                147





Click HERE for graphic.


                                148





a seasonal inventory is being used for evaluating emissions during
a peak ozone period, the summer morning data from Figure 5-2 may be
preferred.  Episodes lasting two to five days occur most frequently
during the winter for much of the U.S. If these episode periods are
of primary interest, the data from Figure 5-3 should be used. 
Mixing Heights, Wind Speeds, and Potential for Urban Air Pollution
Throughout the Contiguous United States should be consulted for
additional information on the use of these figures.  As a final
alternative for mixing height, a default of 3000 feet may be used. 
This value, which is used as the default value for the EPA standard
LTO, is incorporated into the calculations used for determining
time-in-mode.

5.2.3     Activity and Emissions for Commercial Aircraft

     The next four steps relate specifically to creating an
emissions inventory for commercial aircraft.  The procedures for
other aircraft categories are discussed subsequently.  Definition
of the mix of commercial aircraft that uses each airport (step
three) can be found in Airport Activity Statistics of Certified
Route Air Carriers,243 published annually by FAA.  Figure 5-4, a
copy of a page from Table 7 of that report, shows the information
that is included by airport.  All of the commercial aircraft that
used the airport for the given year are listed, along with the
number of departures during the year.  This is the fleet that
should be used for the inventory.

     In step four the number of LTOs is determined by aircraft
type.  Since Airport Activity Statistics of Certificated Route Air
Carriers lists departures, which are equivalent to LTOs, it is
again the preferred source.  From Table 7, the total departures
performed for all service (both scheduled and non scheduled) should
be used as the number of LTOs for each aircraft type.

     The engines used on each aircraft type must be determined to
select the emission factors for step five.  Table 5-2 lists
aircraft and the corresponding engines used to power them.  Many
aircraft use only a single engine model, while others have been
certified to use engines from two or three different manufacturers. 
When a single engine is listed for an aircraft model, emissions
data for that engine should be used.  For aircraft with engines
from more than one manufacturer, defining the specific engine mix
used on the fleet of aircraft operating at a specific airport may
be extremely difficult.  Individual airlines probably are the only
source of detailed fleet data on specific engine models and they
likely do not have it readily available.  To develop a
representative engine mix for aircraft with more than one engine
model, the percentage of each model likely to be found on those
aircraft in the U.S. fleet is shown adjacent to the engine model
number in Table 5-2.  The recommended procedure for compensating
for the lack of detailed engine data is using the percentages

___________________________

     243  Airport Activity Statistics of Certificated Route Air
Carriers, U.S. Department of Transportation, Research and Special
Programs Administration, Federal Aviation Administration, Calendar
Year 1989.  NTIS Report Number ADA 229303.

                                149





Click HERE for graphic.


                                150





Click HERE for graphic.


                                151





Click HERE for graphic.


                                152





Click HERE for graphic.


                                153





Click HERE for graphic.


                                154





shown in the table as weighing factors.  For example, Boeing 757-
200 cargo aircraft have been sold to U. S. airlines with Pratt &
Whitney PW2040 engines as well as Rolls Royce RB.211-535E4 engines. 
The number of aircraft with each engine model is 15 and 43,
respectively, to give the percentages shown in Table 5-2 of 26 and
74.  These percentages can be used to divide the total LTOs for B
757-200 cargo aircraft into two groups representing the two engine
types.  This makes the inventory more representative than assigning
a single engine for all cargo versions of B 757-200's, since the
emission factors are different for each engine.

     After identifying the engines included in the fleet, engine
emission factors are used to calculate mass of emissions.  For some
of the engines shown in Table 5-2, emission factors have never been
determined.  For these engines it is necessary to use emission
factors from an alternative engine.  Table 5-3 lists alternative
engines recommended by the engine manufacturers.  For most of these
engines, emission factors are available for a very similar engine,
usually one of the same model and a related series.  For a small
number of engines there is no emissions data available and there
are no suggested alternatives.  In these instances there are three
approaches available.  First, the needed data may appear in the
latest update of the FAEED data base.  The FAA should be contacted
for the latest version of the data base as mentioned earlier. 
Second, for an aircraft with several potential engine types, where
no emissions data is available for one engine, the recommended
procedure is to reallocate the market share among the engines for
which data is available.  Third, if emission rate information (fuel
consumption and emission index) for an engine model still cannot be
located the engine manufacturer should be contacted directly.

     After the engine types have been identified, fuel flow rates
and emission indexes can be found in Table 5-4.  The data in this
table has been updated since the last edition of this reference and
of AP-42, to include new engine models and to reflect new data on
models already in AP-42.  The next version of AP-42 may have some
additional new data for engines that have not been updated here.
(Updates primarily will be for general aviation aircraft engines.)
The fuel flow rates and emission indexes that appear in Table 5-4
for commercial aircraft are based on information engine
manufacturers provide to FAA and the International Civil Aviation
Organization.  These data are representative of production engines. 
Emission indexes are given for specific fuel flow rates which are
representative of the power settings used during the different
operating modes.  The emission index multiplied by the fuel flow
rate gives an emission rate.

     Step 6 is to specify a time-in-mode for each aircraft type. 
Take-off time is fairly standard for commercial aircraft and
represents the time for initial climb from ground level to about
500 feet.  The default take-off time for calculating emissions is
0.7 minutes (42 seconds) and, unless more specific data is
available, should be used in this methodology.  The time in the
approach and climbout modes depends on mixing height.  As mentioned
earlier, a

                                155





          TABLE 5-3: ALTERNATIVE SOURCE OF EMISSION DATA
                    FOR SOME AIRCRAFT ENGINES1

                                              Source for           
     Manufacturer          Engine Model    Emissions Data2        
     GE                        CF6-6              CF6-6D
                              CF6-50            CF6-50E/C1/E1/C2/E2
                              CT7-5A              CT7-5
                              CT7-5A2             CT7-5
                              CT7-7E              CT7-5

     GE (SCNECMA)             CFM56-2             CFM56-2B
                              CFM56-2-C1          CFM56-2B
                              CFM56-5A            CFM56-5A1

     P&W                      MD series      Contact manufacturer3
                              JT8D-7D             JT8D-7f7Af7B
                              JT8D-15B            JT8D-15
                              JT9D-3A          Contact manufacturer
                              JT9D-7A-SP          JT9D-7Ff7A
                              JT9D-7AH            JT9D-7F[7A
                              MD-20               JT9D-7Ff7A
                              JT9D-70A            JT9D-70/59flQ
                              PW4060              PW4460

     RR                       RB211-535E5    Contact manufacturer4
                              RB211-535F5      Contact manufacturer
                              TRENT 600 series Contact manufacturer
                              TRENT 700 series Contact manufacturer
                              SPEY M006        Contact manufacturer
                              SPEY NIK555-15      SPEY NM55
                              SPEY NM55-15P       SPEY NIK555
                              SPEY NIK555-15H     SPEY NIK555
                              SPEY NIK512      Contact manufacturer
                              TAY MK651        Contact manufacturer

                              Dart 514-7          Dart RDa7
                              Dart 528-7E         Dart RDa7
                              Dart 532-7          Dart RDa7
                              Dart 532-7N         Dart RDa7
                              Dart 532-7P         Dart RDa7
                              Dart 532-7R         Dart RDa7
                              Dart 535-7R         Dart RDa7
                              Dart 536-7E         Dart RDa7
                              Dart 542-4          Dart RDa10
                              Dart 542-10J        Dart RDa10
                              Dart 542-10K        Dart RDa10
                              Dart 552-7R         Dart RDa7

                                156




          TABLE 5-3: ALTERNATIVE SOURCE OF EMISSION DATA
                    FOR SOME AIRCRAFT ENGINES1
                            (Concluded)


1   FAA Aircraft Engine Emission Database does not identify these
     alternative emission factors.  A manual adjustment to the
     database output may be required.

2   As recommended by engine manufacturers.

3   For information, contact the Office of Certification &
     Airworthiness, Commercial Engine Business, United Technologies
     Pratt & Whitney, 400 Main Street, East Hartford, Connecticut
     06108, 203/565-2269.

4   For information, contact Manager Project Combustion, Rolls
     Royce plc.  P.O. Box 31, Derby DE2 88J England.  Telephone -
     0332 242424.

                                157





Click HERE for graphic.


                                158





Click HERE for graphic.


                                159





Click HERE for graphic.


                                160





Click HERE for graphic.


                                161





Click HERE for graphic.


                                162





Click HERE for graphic.


                                163





Click HERE for graphic.


                                164





Click HERE for graphic.


                                165





Click HERE for graphic.


                                166





Click HERE for graphic.


                                167





Click HERE for graphic.


                                168





Click HERE for graphic.


                                169





    TABLE 5-4: MODAL EMISSION RATES - CIVIL AIRCRAFT ENGINES1
                            (Concluded)


1   SOURCE: ICAO Engine Exhaust Emissions Databank (ICAO Committee
on Aviation Environmental Protection, Working Group 3 Meeting,
Mariehamn, Aland., October 1989), unless otherwise noted.

2   MANUFACTURERS: All. - Allison, Con - Teledyne/Continental,
GE - General Electric, Grt - Garrett AiResearch, Lyc -
Avco/Lycoming, P&W - Pratt & Whitney, P&WC - Pratt & Whitney
Canada, RR - Rolls-Royce

3   SO2 emissions based on national average sulfur content of
aviation fuels from Aviation Turbine Fuels, 1989, Dickson, Cheryl
L. and Paul W. Woodward, March, 1990, NIPER Report Number NIPER-164
PPS, National Institute for Petroleum and Energy Research, UT
Research Institute, Bartlesville, Oklahoma.

4   Source of data is AP-42, Compilation of Air Pollutant Emission
Factors, Volume II: Mobile Sources, U.S. Environmental Protection
Agency, Ann Arbor, Michigan, September, 1985. (Aircraft data from
February 1980).

5   Source of engine data is General Electric Office of Combustion
Technology, GE Aircraft Engines, One Neumann Way MD A309,
Cincinnati, Ohio 45215-6301, 513/774-4438.

6   Source of data is AP-42.  Source of Particulate data is AP-42
Reference 4 (M.  Platt, et al., The Potential Impact of Aircraft
Emissions upon Air Quality, APTD-1085, U.S. Environmental
Protection Agency, Research Triangle Park, NC, December 1971).  The
indicated reference does not specify series number for this model
engine.

7   Source of engine data is ICAO, ICAO Engine Exhaust Emissions
Databank.  Data are sales weighted averages of two versions of this
engine.  The basis is 93% high emission combustors and 7% low
emission combustors.

8   Source of engine data is ICAO, ICAO Engine Exhaust Emissions
Databank.  Data are sales weighted averages of two versions of this
engine.  The basis is 77% high emission combustors and 23% low
emission combustors.

9   Source of engine data is Rolls Royce Combustion Research
Department, Rolls Royce plc.  P.O. Box 31, Derby DE2 88J England. 
Telephone - 0332 242424.

                                170






default mixing height of 3000 feet was assumed for calculating an
approach time of 4 minutes and a climbout time of 2.2 minutes,
which can be used if specific information on mixing height is
unavailable.  The procedure for adjusting these times to correspond
to a different mixing height is shown below.

     The mode most likely to vary by time for each specific airport
is taxi/idle time.  Total taxi/idle time for a very congested
airport can be as much as three or four times longer than for an
uncongested airport.  Taxi/idle-in time typically is shorter than
taxi/idle-out time because there are usually fewer delays for
aircraft coming into a gate than for aircraft lining up to takeoff. 
For a large congested airport the taxi/idle-out time can be three
times longer than taxi/idle-in time.  Taxi/idle time also may vary
by aircraft type.  For example, wide-body jets may all use special
gates at the terminal that place them further from the runway than
narrow-body jets or small regional commuter aircraft so their
taxi/idle-in and taxi/idle-out times are longer.  Because of the
variation in taxi/idle time, it is important to get data specific
to the airports of interest in the inventory.  Commercial airlines
must keep track of their taxi/idle time at each airport for
different aircraft types so that their flight schedules reflect
anticipated daily and seasonal variations.  These data are
important to the airlines since they report schedule delays to the
Department of Transportation as a measure of their operating
performance.  Therefore, the airlines' Flight Operations
departments at their headquarters locations are the best source of
data for taxi/idle time by aircraft type at a particular airport. 
Since all airlines using a particular airport will experience
similar taxi/idle times it is only necessary to get information
from a single source.  If taxi/idle times are not available for a
particular airport, Table 5-1 lists default values of taxi/idle
periods, as well as other modes, for different aircraft
classifications.  For commercial aircraft this information is based
on data collected prior to 1971 at large airports during periods of
congestion.  Idle times that reflect more recent experience will be
incorporated in the next version of AP-42.  For the inventory
calculations, taxi/idle-in and taxi/idle-out time are added
together to get a total time for the taxi/idle mode.

     The final step in the procedure is to calculate total
emissions for each aircraft type and to sum them for a total
commercial aircraft emission rate.  The following series of
equations illustrates the calculation:

Adjust Approach and Climbout TIM to Represent Local Conditions

     These equations adjust the times-in-mode, which are based on a
default mixing height of 3000 feet, to an airport specific value
based on the local mixing height.  Equation 5-2 assumes the
climbout mode begins with the transition from takeoff to climbout
at 500 feet and continues until the aircraft exits the mixing
layer.

                                171





     TIMapp-C =    4 X (H/3000)                              (5-1)

     TIMclm-C =    2.2 X [(H-500)/2500]                      (5-2)

     TIMapp-C -    time in the approach mode for commercial
                    aircraft, in minutes

     TIMclm-C -    time in the climbout mode for commercial
                    aircraft, in minutes

     H         -    mixing height used in air quality modeling for
                    time and region of interest


Calculate Emissions for Each Aircraft Type

     Eij =     (TIMjk) X (FFjk/1000) X (EIijk) X (NEj)  (5-3)

     Eij =    total emissions of pollutant i, in pounds, produced
               by aircraft type j for one LTO cycle

     TIMjk =  time in mode for mode k, in minutes, for aircraft
               type j

     FFjk =   fuel flow for mode k, in pounds per minute, for each
               engine used on aircraft type j (from Table 5-4)

     EIijk =  emission index for pollutant i, in pounds of
               pollutant per one thousand pounds of fuel, in mode k
               for aircraft type j (from Table 5-4)

     NEj =    number of engines used on aircraft type j (from
               Table 5-2)

                                172





     Calculate Total Emissions for All Commercial Aircraft

     ETi(C)   =    (Eij)  X  (LTOj)                       (5-4)

     ETi(C)   -    total emissions of pollutant i, in pounds,
                    produced by all conunercial aircraft operating
                    in the region of interest (where j covers the
                    range of commercial aircraft operating in the
                    area)

     LTOj          -    total number of LTO cycles for aircraft
                         type j, during the inventory period
                         (annual data available from Airport
                         Activity Statistics of Certificated Route
                         Air Carriers, Table 7)


     After completing this series of equations, the inventory of
emissions is complete for commercial aircraft.  The next series of
calculations is a repeat of steps three through six for general
aviation aircraft.

5.2.4     Activity and Emissions for General Aviation and Air Taxi
          Aircraft

5.2.4.1   Aircraft-Specific Procedure

     The overall methodology for general aviation and air taxi
aircraft is similar to that for commercial aircraft. 
Unfortunately, defining the fleet mix and associated activity level
of general aviation and air taxi aircraft is more difficult than
for commercial aircraft.  FAA does not track operations by aircraft
model for the general aviation category and no other sources of
these data cover all states.  For some states, this information is
available for some airports from the State Airport Authority or
from the operations officials at individual airports.  Detailed
model information for aircraft operating in the inventory area is
difficult to locate, except perhaps for air taxis, and may add only
relatively small improvement in accuracy to the emissions inventory
compared to treating general aviation and air taxis as though they
were made up of a representative mix of aircraft.  For some smaller
airports, air taxi activity may predominate and it may be possible
to locate aircraft specific information on the operations there.

     Where information on specific aircraft is available, it can be
combined with enginespecific emission factors and the time in each
operational mode to estimate total engine emissions from the
general aviation and air taxi categories.  Table 5-4 shows emission
factors for the various engines.  Table 5-5 shows some examples of
the aircraft and engine combinations.  Information on these
categories may be expanded in the next update of AP-42 to include
more aircraft and engine combinations as well as emission indices
for additional engines.

                                173





Click HERE for graphic.


1   Source of aircraft, corresponding engines, and number of
     engines is FAA Aircraft Engine Emission Database (FAEED), U.S.
     Department of Transportation, Federal Aviation Administration,
     Office of Environment and Energy, 1991.  Source of number of
     seats, aircraft type, and number of aircraft is Census of U.S.
     Civil Aircraft, U.S. Department of Transportation, Federal
     Aviation Administration, Office of Management Systems,
     Calendar Year 1989.

2   No. of Aircraft refers to Total U.S. Registered Aircraft as of
     December 31, 1989.

3   ENGINE MANUFACTURERS ABBREVIATIONS: Con -
     Teledyne/Continental, GE - General Electric, Grt - Garrett
     AiResearch, Lyc - Avco/Lycoming, P&W - Pratt & Whitney, PWC -
     Pratt & Whitney Canada, RR - Rolls-Royce

4   Engine refers to a PA-18-150 Super aircraft.

5   Engine refers to a PC6/B2H2 aircraft.

6   Engine refers to a PA-42 Cheyenne aircraft.

                                174





     If the detailed estimation procedure is being followed based
on specific aircraft and engines, airport specific estimates on
time-in-mode might be used if available from airport officials. 
These data likely vary quite widely because of the many different
types of services provided by this aircraft category.  Otherwise,
the estimation procedure is based on the default times-in-mode from
Table 5-1.  The rest of the detailed estimation procedure uses the
same set of equations used for commercial aircraft.  Emissions
should be calculated separately for the general aviation and air
taxi categories.

  Adjust Approach and Climbout TIM to Represent Local Conditions

     TIMapp-G =    6 X (H/3000)                              (5-5)

     TIMclm-G =    5 X [(H-500)/2500]                        (5-6)

     TIMapp-G -    time in the approach mode, in minutes

     TIMclm-G -    time in the climbout mode, in minutes (assumes
                    transition from takeoff to climbout occurs at
                    500 feet)

          H    -    mixing height used in air quality modeling for
                    time and region of interest

     Calculate Emissions for Each Aircraft Type

     The emission factors that appear in Table 5-4 for general
aviation and air taxi aircraft have not been updated since the last
version of AP-42.  The next edition of AP-42 should include updates
to much of the date that appears in the table.

     Eij =    (TIMjk) X (FFjk/1000) X (EIijk) X (NEj)   (5-7)

     Eij      =    total emissions of pollutant i, in pounds,
                    produced by aircraft type j for one LTO cycle.

     TIMjk    =    time in mode for mode k, in minutes, for
                    aircraft type j

     FFjk          =    fuel flow for mode k, in pounds per
                         minutes for each engine used on aircraft
                         type j (from Table 5-4)

     EIijk    =    emission index for pollutant i, in pounds of
                    pollutant per one thousand pounds of fuel , in
                    mode k for aircraft type j (from Table 5-4)

     NEj      =    number of engines used on aircraft type j (from
                    Table 5-5)


                                175





        Calculate Total Emissions for All General Aviation
                       or Air Taxi Aircraft

          ETi(G)   =     Eij                               (5-8)

          ETi(G)   =    total emissions of pollutant i, in pounds,
                         produced by all general aviation or air
                         taxi aircraft operating in the region of
                         interest (where j covers the range of
                         aircraft operating in the area)

          LTOj          =    total number of LTO cycles for
                              aircraft type j, during the inventory
                              period

5.2.4.2   Alternative, Fleet-Average Procedure

     Where detailed information on specific aircraft mix and
activity is unavailable, a rough estimate of emissions for each
aircraft category can be made using emission indices based on a
representative fleet mix.244  The following indices were
calculated based on 1988 fleet data for general aviation aircraft.

     HC         0.394 pounds per LTO
     CO        12.014 pounds per LTO
     NOx       0.065 pounds per LTO
     SO2       0.010 pounds per LTO

     Since air taxis have fewer of the smallest engines in their
fleet and more turboprop and turbojet engines, their emission
factors are somewhat different.

     HC         1.234 pounds per LTO
     CO        28.130 pounds per LTO
     NOx       0.158 pounds per LTO
     SO2       0.015 pounds per LTO

     Airport activity for general aviation aircraft and air taxis
can be found in FAA Air Traffic Activity.  Figure 5-5 is a copy of
a page from Table 4 which reports airport operations at airports
with FAA-operated traffic control towers.  Table 22 from the same
report lists operations at airports with FAA contractor-operated
traffic control towers.  In this report, an operation could be
either a takeoff or landing, so the number of operations should be
divided by two to get LTOs.  In addition to these airports, general
aviation and air taxi activity is common at smaller airports and
landing strips not included in FAA's reporting system.  These
airports must be contacted directly to determine if information is
available on
___________________________

     244 Memorandum from S. Webb to R. Wilcox, "General Aviation
Generalized Emission Indexes," June 10, 1991.

                                176





Click HERE for graphic.


                                177





general aviation activity.  Air taxi operators located at the
airports, may be a source for information on air taxi activity. 
These steps may have little impact on the inventory and should be
considered discretionary.

     The annual emissions are then calculated as the product of
airport activity in LTOs from FAA Air Traffic Activity and the
emission index in pounds per LTO listed above.  Total emissions
should be computed separately for general aviation and air taxis.

5.2.5     Activity and Emissions for Military Aircraft

     FAA Air Traffic Activity contains information on the number of
military operations at airports with FAA-operated traffic control
towers.  This information can be used in much the same way as for
general aviation aircraft, however, military air bases are not
included in this reference.  The information only addresses
military operations at civil airports.  Military air bases included
in the modeling area should be apparent from maps of the area.  For
these bases, it likely will be difficult to get good information on
fleet make up and activity.  In some cases, information may be
available from the Office of the Base Commander on fleet make-up
and possibly some measure or estimate of activity such as LTOs for
one day or one month.  Where specific information is available for
aircraft type and LTOs, Table 5-6 lists military aircraft and their
engines, and Table 5-7 lists the modal emission rates for these
engines.  Much of the data in Table 5-7 has been updated since the
last version of AP-42.

     Where data on military aircraft operations and fleet make-up
cannot be obtained from the base commander, a centralized support
office may be able to provide the required information.  The
Navy245 and Air Force246 both have environmental support offices
responsible for information on emissions from military aircraft
including complete inventories for many bases.  If inventory
information is unavailable after contacting the Navy or Air Force
environmental support office, a letter requesting an inventory
should be sent to the base commander through the EPA regional
office with copies to the appropriate environmental support office.

     If data on fleet make up and activity are obtained from the
base commander or the environmental support offices, the procedure
for calculating an inventory for military aircraft is the same as
that used for both commercial and general aviation.  The
calculations for each subsequent step follow.
___________________________

     245  Aircraft Environmental Support Office (AESO). 
Commanding Officer, Attn: AESO, Code 04210, Naval Aviation Depot,
North island, San Diego, California 92135-5112, (619) 545-2901.

     246  Air Force Engineering and Service Center, RDUS, Tyndall
AFB, Florida 32403.

                                178





Click HERE for graphic.


                                179





Click HERE for graphic.


                                180





      TABLE 5-6: MILITARY AIRCRAFT TYPES AND ENGINE MODELS1
                            (Concluded)


1   SOURCE: FAA Aircraft Engine Emission Database (FAEED), (U.S.
     Department of Transportation, Federal Aviation Administration,
     Office of Environment and Energy, 1991) unless otherwise
     noted.

2   Source of Type information is "Aviation Week & Space
     Technology," McGraw-Hill Publication, March 18, 1991.  TYPES:P
     - Piston, TF - Turbofan, TJ - Turbojet, TP - Turboprop, TS -
     Turboshaft

3   Source of Operator information is Encyclopedia of Modem
     Military Aircraft, Taylor, Michael, 1987.  OPERATORS: Army, NG
     - National Guard, USAF - U.S. Air Force, USCG - U.S. Coast
     Guard, USMC - U.S. Marine Corps, USN - U.S. Navy, US - USAF
     USCG, USMC, & USN.

4   ENGINE MANUFACTURERS: All. - Allison, GE - General Electric,
     Grt - Garrett AiResearch, Lyc - Avco/Lycoming, PW - Pratt &
     Whitney, W - Curtis Wright

5   Source of aircraft and corresponding engine information is
     Example of an Air Base Emissions Inventory for the County of
     San Diego (1987), Aircraft Environmental Support Office, AESO
     Report No. 2-91, San Diego, California, March 1991.

6   Sources: Engines - Summary Table of Gaseous and Particulate
     Emissions from Aircraft Engines, Aircraft Environmental
     Support Office, AESO Report No. 6-90, San Diego, California,
     June 1990.
          Aircraft, Type, and No. of Engines - "Aviation Week &
     Space Technology," McGraw-Hill Publication, March 18, 1991.
          Classification and Operator - Encyclopedia of Modern
     Military Aircraft, Taylor, Michael, 1987.

7   Source: "Aviation Week & Space Technology," McGraw-Hill
     Publication, March 18, 1991.

                                181





Click HERE for graphic.


                                182





Click HERE for graphic.


                                183





Click HERE for graphic.


                                184





Click HERE for graphic.


                                185





Click HERE for graphic.


                                186





Click HERE for graphic.


                                187





   TABLE 5-7: MODAL EMISSION RATES - MILITARY AIRCRAFT ENGINES1
                            (Concluded)

1   SOURCE: Example of an Air Base Emissions Inventory for the
     County of San Diego (1987) (Aircraft Environmental Support
     Office, AESO Report No. 2-91, San Diego, California, March
     1991), unless otherwise noted.

2   MANUFACTURERS: All. - Allison, Con - Teledyne/Continental,
     CP - United Aircraft of Canada, GE - General Electric, Grt -
     Garrett AiResearch, Lyc -Avco/Lycoming, P&W - Pratt & Whitney,
     RR - Rolls-Royce, W - Curtis Wright

3   SO, emissions based on national average sulfur content of
     aviation fuels from Aviation Turbine Fuels, 1989, Dickson,
     Cheryl L. and Paul W. Woodward, NIPER Report Number NIPER-164
     PPS, National Institute for Petroleum and Energy Research, ]IT
     Research Institute, Bartlesville, Oklahoma, March 1990.

4   Source of data is AP-42, Compilation of Air Pollutant Emission
     Factors.  Volume II: Mobile Sources, U.S. Environmental
     Protection Agency, Ann Arbor, Michigan, September, 1985.
     (Aircraft data from February 1980).  Nitrogen oxides reported
     as NOx.  HC refers to total hydrocarbons (Volatile organics,
     including unburned hydrocarbons and organic pyrolysis
     products).

5   Includes all "condensable particulates," and thus may be much
     higher than solid particulates alone (AP-42).

6   Source of data is FAA Aircraft Engine Emission Database
     (FAEED), U.S. Department of Transportation, Federal Aviation
     Administration, Office of Environment and Energy, 1991.

7   Source of data is Summary Tables of Gaseous and Particulate
     Emissions from Aircraft Engines, Aircraft Environmental
     Support Office, AESO Report No. 6-90, San Diego, California,
     June 1990.

8   Source of data is ICAO Engine Exhaust Emissions Databank, ICAO
     Committee on Aviation Environmental Protection, Working Group
     3 Meeting, Mariehamn, Aland., October 1989,

9   Includes all "condensable particulates," and thus may be much
     higher than solid particulates alone.  Data are interpolated
     values assumed for calculational purposes, in the absence of
     experimental data (AP-42).

10  Particulate data refers to TF34-GE-400A engine.

11  Particulates refer to dry particulates only (AP-42).

12  Source of Particulate data is Table 4. Particulate Mass
     Emissions From the TF-30-P-414 Engine, Summary Tables of
     Gaseous and Particulate Emissions from Aircraft Engines,
     Aircraft Environmental Support Office, AESO Report No. 6-90,
     San Diego, California, June 1990.

                                188





  Adjust Approach and Climbout TIM to Represent Local Conditions

     TIMapp-M =    4 X (H/3000)                              (5-9)

     TIMclm-M =    1.4 X [(H-500)/2500]                     (5-10)

     TIMapp-M -    time in the approach mode for military
                    aircraft, in minutes

     TIMclm-M -    time in the climbout mode for military
                    aircraft, in minutes (assumes transition from
                    takeoff to climbout occurs at 500 feet)

     H         -    mixing height used in air quality modeling for
                    time and region of interest


     Calculate Emissions for Each Aircraft Type

     Eij =    (TIMij) X (FFjk/1000) X (EIijk) X (NEj)  (5-11)

     Eij      -    total emissions of pollutant i, in pounds,
                    produced  by aircraft type j for one LTO cycle


     TIMij    -    time in mode for mode k, in minutes, for
                    aircraft type j

     FFjk          -    fuel flow for mode k, in pounds per
                         minute, for each engine used on aircraft
                         type j (from Table 5-7)

     EIijk    -    emission index for pollutant i, in pounds of
                    pollutant per one thousand pounds of fuel , in
                    mode k for aircraft type j (from Table 5-7)

     NEj      -    number of engines used on aircraft type j (from
                    Table 5-6)


     Calculate Total Emissions for All Military Aircraft

     ETi(M)   =     Eij                                   (5-12)

     ETi(M)   -    total emissions of pollutant i, in pounds,
                    produced by all military aircraft operating in
                    the region of interest (where j covers the
                    range of military aircraft operating in the
                    area)

                                189





     After completing the emissions inventory for military
aircraft, the overall inventory is complete, made up of emissions
from commercial, general aviation, and military aircraft.  The
final three sections of the report address changes to the inventory
due to alternative operating practices, addition of minor emission
sources, and changes to the aircraft fleet in the future.

5.3  VARIATIONS TO THE INVENTORY CALCULATION PROCEDURE

     There are several variations to the basic inventory procedure
that can adjust the period covered by the inventory or address some
operational procedures followed by some pilots or airlines that
affect aircraft emissions.  These adjustments to the inventory are
discussed in this section.

5.3.1     Variability of Activity - Daily and Seasonal

     The calculation procedure described in the methodology does
not address daily or seasonal variations.  If the air quality
modeling period requires emissions data that accounts for these
variations, certain adjustments must be made to the equations.  The
daily or seasonal variations will be exhibited in LTOs, mixing
height, and idle time, primarily idle-out.

     The references for determining LTOs in Section 5.2 give data
on an annual basis and adjustment may be necessary to capture
changes over time.  The frequency of LTOs at most civil airports
are reasonably uniform during daylight hours with lower activity
during the night and uniform during week days with lower activity
on the weekends, although some airports that cater to recreational
flying may show higher activity on weekend days.  For most large
urban airports, LTOs are uniform on a monthly basis with a slight
increase in activity during the summer, which typically is a time
of high travel, although some regions may attract more travelers
during the winter as a result of their climate.  The seasonal
variation in activity at smaller urban airports or airports that
serve smaller cities may be more pronounced because of factors that
affect travel on a local basis such as tourism or seasonal business
activity.  Obtaining specific information on daily and seasonal
variation is difficult.  The best source likely will be the airport
operators, many of who keep some type of records of activity such
as total number of LTOs, number of visitors/passengers, number of
cars using the parking lots, or some similar measure that may be
representative of the daily or seasonal variation in use of the
airport.  Another source of information on the daily and weekly
variation of LTOs is published flight schedules.  These schedules
can be reviewed to evaluate the number of scheduled flights during
daylight hours versus night-time hours or week day versus weekend. 
It would be difficult to use this source to evaluate seasonal
variations.

     Mixing height changes throughout the day and from season to
season depending on meteorological conditions such as wind, cloud
cover, temperature, and humidity.  The adjustments to the time in
approach and climbout mode should be based on a weighted average of
the mixing heights for the time periods of interest, using
variations in LTOs as the weighing factors.  See Section 5.2.2 for
more information about determining the mixing height.

                                190





     Taxi/idle time may vary in proportion to variations in LTOs
because they are partially a function of airport congestion such
that the greater number of LTOs the more likely that airport
congestion will increase the time for aircraft to taxi to the
runway.  The airlines' scheduling departments are the best sources
of taxi/idle-time data and their projections typically show daily
variations estimated for a particular season.  Airport operators
also may have information on taxi/idle time variation during a day
or from one season to another.  Availability of this data will be
highly variable.

5.3.2     Operational Activity that Affects Aircraft Emissions

     There are variations to standard operating procedures which
pilots follow that will affect the aircraft's emissions.  Two
examples, which may be found in commercial operations, are single-
engine taxiing and derated takeoff.  Both of these procedures have
the potential to save fuel as well as reduce emissions.  Where
detailed air quality modeling is being performed, these refinements
may merit consideration.  However, in most cases these procedures
are performed at the discretion of the pilots and their use may not
be consistent or predictable.

5.3.2.1 Reduced Engine Taxiing

     Single-engine taxiing or reduced-engine taxiing is, as the
name implies, taxiing with one or more engines shutdown.  This is
usually practiced during taxi-out.  An aircraft can taxi using a
single engine at idle without significantly increasing the
emissions of that engine since adequate power for taxi generally is
available at idle power setting.  The emissions reductions are
equal to the calculated emissions of the engines that are shutdown. 
The change to the calculation procedure to account for single-
engine taxiing is shown in Equation 5-13.

     Eij       (TIMjk) X (FFjk/1000) X (EIijk) X (NEjk)    
(5-13)

     Eij -    total emissions of pollutant i, in pounds, produced
               by aircraft type j

     TIMjk -  time in mode for mode k, in minutes, for aircraft
               type j

     FFjk -   fuel flow for mode k, in pounds per minute, for each
               engine used on aircraft type j (from Table 5-4)

     EIijk -  emission index for pollutant i, in pounds of
               pollutant per one thousand pounds of fuel , in mode
               k for aircraft type j (from Table 54)

     NEjk -   number of engines used on aircraft type j, for mode
               k (from Table 5-2)

     NE for the taxi/idle-out mode would be the number of engines
actually used rather than the number on engines shown in Table 5-2.

                                191





5.3.2.2 Derated Take-off

     A derated take-off is a procedure where the pilot sets the
throttle for takeoff at less than 100%.  The derated throttle
setting is determined based on worst-case operating conditions,
i.e., performance of the aircraft as though it were at maximum
weight on a hot day.  In some cases this may allow a takeoff
throttle setting of 90% or less.  To adjust the emissions
calculations to account for this change, engine manufacturers
recommend a linear interpolation between the takeoff and climbout
fuel flow rates and emission factors.  Information on the degree
and frequency of derating; for takeoff should be collected directly
from the airlines.

     Other operational factors may affect engine exhaust emissions,
such as the use of full throttle, reverse thrust to decelerate the
aircraft during landing.  These effects may also be significant and
are being evaluated by EPA.  Any additional information on
operational factors will be included in the next update to AP-42.

5.3.3     Particulate Emissions

     As mentioned in Section 5.1.1.2, very few measurements have
been made of particulate emissions from aircraft engines.  However,
for most turbine engines, EPA does limit the amount of smoke that
may be emitted.  This limit is specified as a smoke number. 
Attempts have been made to derive a correlation between smoke and
particulates which could be used to create a particulate emission
index based on smoke number.  Thus far, these efforts do not match
experimental results very closely.  If particulates are of concern
for the inventory area it may be of help to discuss the issue
further with the engine manufacturers or the FAA Office of
Environment and Energy.  EPA will continue to investigate this area
and may provide further information in the next update to AP-42.

5.4  OTHER EMISSION SOURCES

5.4.1     Auxiliary Power Units

     When large aircraft are on the ground with their engines shut
down, they need power and preconditioned air to maintain the
aircraft's operability.  If a ground-based power and air source is
unavailable, an auxiliary power unit (APU), which is part of the
aircraft, is operated.  These units are essentially small jet
engines, which generate electricity and compressed air.  They bum
jet fuel and generate exhaust emissions like larger engines.  In
use, APUs essentially run at full throttle.  Table 5-8 lists
several APUs and the aircraft on which they are installed. 
Emission factors for some of these APUs are provided in Table 5-9. 
Emission factors for APUs under load should be used where
information is available on time of use.


                                192




              TABLE 5-8: APU'S AND AIRCRAFT MODELS1

     Auxiliary
     Power Unit                                              No. of
     (Shaft HP)           Aircraft Model                   Aircraft


COMMERCIAL

Allied-Signal Aerospace Company
Garrett Auxiliary Power Division

     GTP 30 Series       Fairchild F-272

     GTCP 30 Series      Dassault-Bregue Falcon 202
                         Jet Commander2

     GTCP 35-300         Airbus A-3213

     GTCP 36 Series      Airbus A320                            132
     (80 HP)             Aerospatiale ATR-422
                         Brit. Aero. BAe 146                    149
                         Canadair CL600/CL6012
                         Dassault-Bregue Falcon 502
                         Embraer EMB-1202
                         Fokker F-28                            193
                         Fokker F-100                            56
                         NAMC YS-112
                         Saab Fairchild 3402

     GTC 85              Convair CV-5802

     GTCP 85 Series      Boeing B-707                           206
     (200 HP)            Boeing B-727                         1,652
                         Boeing B-737                         1,825
                         Lockheed L-1002
                         McDonnell Douglas DC-8                 300
                         McDonnell Douglas DC-9                 842
                         McDonnell Douglas MD-80                806

     GTCP 331 Series     Airbus A-300-600                      854
     (143 HP)            Airbus A-310                           175
                         Airbus A-3303
                         Airbus A-3403
                         Boeing B-757                           328
                         Boeing B-767                           343
                         Boeing B-7773

     GTCP 660            Boeing B-747                           671
     (300 HP)
     TSCP 700            Airbus A-300-B2                         52
                         Airbus A-300-B4                        184


                                193






              TABLE 5-8: APU'S AND AIRCRAFT MODELS1
                            (Continued)

     Auxiliary
     Power Unit                                              No. of
     (Shaft HP)           Aircraft Model                   Aircraft

     (142 HP)            McDonnell Douglas DC-10                365
                         McDonnell Douglas MD-11                  3

     Hamilton Standard

     ST-6                Lockheed L-1011                        226

     PraTT & Whitney

     PW 901A             Boeing B-747                           103


MILITARY5

     Allied-Signal Aerospace Company
     Garrett Auxiliary Power Division

     GTC 36-200          McDonnell Douglas F-18 Hornet

     GTCP 36 Series      Gulfstream II (VC-11A)
     (80 BP)             Gulfstream III (C-20A/B)
                         Lockheed S-3A Viking

     GTC 85 Series       Gulfstream I (VC-4A)
                         Lockheed C-130 Hercules                221

     GTCP 85 Series      McDonnell Douglas C-9                   23
                         Lockheed C-141 StarLifter              254
                         Boeing T-43                             15

     GTCP 660-4          Boeing E-4 NEACP                         4
     (300 BP)
     JFS 100 Series      Douglas A-4M Skyhawk
                         Vought A-7D Corsair 11               3356

     JFS 190-1           McDonnell Douglas F-15 Eagle           895

                                194





              TABLE 5-8: APU'S AND AIRCRAFT MODELS1
                            (Concluded)


1   SOURCES:
     Civil - Federal Express Fleet Guide (Federal Express Aviation
     Services, Inc., January, 1991), unless otherwise noted.
     Military - Reference Guide - Auxiliary Power Systems, Garrett
     Turbine Engine Company, Phoenix, Arizona.

2   SOURCE: Reference Guide - Auxiliary Power Systems, Garrett
     Turbine Engine Company, Phoenix, Arizona.

3   New aircraft scheduled to enter production.

4   No. of Aircraft refers to Airbus A-300 aircraft.

5   No. of Aircraft refers to the total number of aircraft in the
     Air National Guard, Air Force Reserve, Air Force, and Coast
     Guard inventories.
    SOURCES:  Air National Guard, Air Force Reserve, Air Force -
               "AIR FORCE Magazine," Air Force Association, May
               1991.
               Coast Guard - United States Coast Guard, 2100 Second
               Street, SW, Washington, DC 20593-0001, 202/267-0952.

6   No. of Aircraft refers to 14 A-7 and 321 A-7D aircraft.

                                195





     TABLE 5-9: MODAL EMISSION RATES - AUXILIARY POWER UNITS1


                                   Emission Rates (lb/1000 lb)
Model-Series          Fuel Flow    HC        CO        NOx   SO2
(Shaft HP      Mode   (lb/min)
at Load)

GTC85-72       Load      3.50      0.13      14.83     3.88   0.54
(200)


GTCPIOO-54     Load      6.88      0.16       5.89     5.95   0.54
(400)


GTPC95-2       Load      4.88      0.36       3.20     5.65   0.54
(300)

T-62T-27       Load      1.70      7.79      42.77     3.94   0.54
(100)


WR27-1         Load      2.33      0.21       5.66     4.63   0.54
(85)


1   SOURCE: Summary Tables of Gaseous and Particulate Emissions
from Aircraft Engines (Aircraft Environmental Support Office, AESO
Report No. 6-90, San Diego, California, June 1990), unless
otherwise noted.

                                196





Where emission factors are unavailable for a specific APU, factors
for an alternative unit of the same or similar horsepower should be
used.  It will be necessary to contact the airlines or military
base commander directly to find out whether APUs are used regularly
at a specific airport and, if so, how long an aircraft is expected
to stay at a gate with the APU running.  This information may be
difficult to get.

5.4.2     Evaporative Emissions

     For general aviation aircraft, there are evaporative emissions
that result from refueling and fuel spillage.  Emissions also occur
from preflight checks of the aircraft and diurnal temperature
cycles that cause the fuel tanks to vent.  Refueling emissions are
addressed in Volume I, Section 5.4. 1. EPA is continuing to
evaluate the other emission sources and may provide information in
the next update to AP-42.

5.5  EFFECT OF FUTURE CHANGES TO THE FLEET

     Airlines continually acquire newer aircraft, gradually phasing
out older models.  While commercial aircraft often remain in
service for more than 25 years, over time, this process phases out
the aircraft using engines that do not meet EPA's hydrocarbon
emission standard.  The current world aircraft fleet averages 12.4
years old according to the 1990 World Jet Inventory published by
the Boeing Corporation.247  Significant among the older aircraft
are engines that do not meet the EPA standard such as the Spey
NIK511 and older JT8Ds and CF6-50s.  The JT8Ds and CF6-50s are
prevalent on B-727s, DC-9s, and DC-10s, many nearly 20 years old. 
As new aircraft are added to the fleet the older aircraft are the
most likely to be retired.  The effect is one of replacing older,
dirty engines with newer engines on the new aircraft that are much
cleaner from an emissions standpoint.  Airport noise regulations
also are forcing changes to the commercial aircraft fleet. 
National noise regulations which were recently passed by Congress
are forcing airlines to phase out use of loud aircraft by 2000. 
This can be accomplished by retiring the loud, older aircraft,
replacing their engines with newer, quieter ones, or modifying the
engines to muffle the noise.  The first two alternatives result in
aircraft with reduced emissions.  Because this legislation is so
new, the airlines are yet to formulate specific plans meeting the
requirements.  However, as the equipment is updated, the changes to
the fleet will be reflected in FAA's reports on aircraft activity. 
Since there is a significant engineering and development leadtime
for producing new aircraft engines, most of the commercial aircraft
to be added to the fleet in the next five to seven years will be
powered by engines that are included in Tables 5-2, 5-3 and 54.

     Since specific plans to upgrade their fleets have not been
announced recently by the airlines, it is difficult to project what
future changes will be and how they will effect the inventory of
emissions for all locations.  Some carriers will update their
fleets more quickly
___________________________

     247  World Jet Airplane inventory, Boeing, Commercial
Airplane Group, Year-End 1990.

                                197





than others so there may be changes that can be captured on an area
specific basis.  If it is desirable to project changes to the
inventory for this source category, the predominant airlines for
the airports included in the inventory area should be contacted for
their specific plans.  EPA is continuing to look at better data
sources and methods for projecting changes to aircraft fleet
emissions.

     Another change that will affect future emissions from aircraft
is the growth in travel.  Air travel has experienced strong growth
over the past several years and this growth is expected to continue
for the foreseeable future.  Many existing airports are near
capacity and others will reach their capacity limits in the near
future.  This will have two effects: air traffic at small feeder
airports and regional hubs will grow and the current major hubs
will experience additional congestion.  The net effect these
changes will have on air quality is unclear.  Increased congestion
at some airports will increase taxi/idle times but the expanded use
of smaller airports may relieve congestion at others.

5.6  CONVERTING FROM TOTAL HYDROCARBONS (THC) TO VOLATILE ORGANIC
     COMPOUNDS (VOC)

     EPA recognizes that it may be necessary to determine the level
of volatile organic compounds (VOC) emitted from aircraft engines. 
Since the emission factors for exhaust HC contained in this
document represent total hydrocarbons (THC), this section
illustrates the method that is recommended for converting THC to
VOC emissions.248

5.6.1     Commercial and Military Conversions

     The commercial and military aircraft fleets are dominated by
turbine engines.  Therefore, a single correction factor can be used
to convert THC to VOC emissions for each aircraft category as
follows.

     VOCCOMMERCIAL =    THCCOMMERCIAL X 1.0947             (5-14)

     VOCMILITARY   =     THCMILITARY X 1.1046              (5-15)

___________________________

     248  Memorandum from R. Cook to R. Wilcox, "Exhaust THC to
VOC Correction Factors for Aircraft," July, 1992.

                                198





                 Non-Road Mobile Sources Inventory
                       Guidance Request Form
                       Please Print or Type

Requester: ___________________________________________________

Agency: ______________________________________________________

Street: ______________________________________________________

City, ST, Zip: _______________________________________________

Fed.  Exp.  Acct: _________________ Telephone: _______________

Nonattainment
Areas Under
Your Agency: _________________________________________________

______________________________________________________________

                        Materials Requested

     _____     New York - New Jersey inventory example

     _____     Inventory disks for areas listed above
            (If among 33 areas analyzed by EPA - see back)

     _____     Booz-Allen study on commercial vessels in six areas
               (Oct. 1991)

     _____     Inventories for others among the 33 areas, for
               purposes  of population-ratio estimates (limit 5)

     _______________      _______________           _______________

     _______________      _______________

Return this form to:     Natalie Dobie  Voice: (313) 741-7812
                         U.S. EPA       FAX:   (313) 668-4368
                         2565 Plymouth Rd.
                         Ann Arbor, MI 48105





                          Areas Available


     1.   Anchorage                                              AK
     2.   Atlanta                                                GA
     3.   Baltimore                                              MD
     4.   Baton Rouge                                            LA
     5.   Beaumont-Port Arthur                                   TX
     6.   Boston-Lawrence-Worcester                              MA
     7.   Chicago-Gary-Lake County                         IL-IN-WI
     8.   Cleveland-Akron-Lorain                                 OH
     9.   Denver-Boulder                                         CO
     10.  El Paso                                                TX
     11.  Hartford-New Britain-Middletown-Bristol                CT
     12.  Houston-Galveston-Brazoria                             TX
     13.  Las Vegas                                              NV
     14.  Miami-Fort Lauderdale                                  FL
     15.  Milwaukee-Racine                                       WI
     16.  Minneapolis-St. Paul                                   MN
     17.  Muskegon                                               MI
     18.  New York-Northern New Jersey-Long Island         NY-NJ-CT
     19.  Philadelphia-Wilmington-Trenton               PA-NJ-DE-MD
     20.  Phoenix                                                AZ
     21.  Portsmouth-Dover-Rochester                             NH
     22.  Providence                                             RI
     23.  Provo-Orem                                             UT
     24.  San Diego                                              CA
     25.  San Joaquin Valley Air Basin                           CA
     26.  Seattle-Tacoma                                         WA
     27.  Sheboygan                                              WI
     28.  South Coast Air Basin                                  CA
     29.  Spokane                                                WA
     30.  Springfield                                            MA
     31.  St. Louis                                              MO
     32.  Tucson                                                 AZ
     33.  Washington                                             DC





5.6.2 General Aviation and Air Taxi Conversions

     The general aviation (GA) fleet and, to a much lesser extent,
the air taxi (AT) fleet may have a significant proportion of piston
engines, in addition to turbine engines.  Therefore, separate
correction factors should be used for each engine type within the
respective aircraft categories if the detailed, aircraft-specific
inventory methodology described in Section 5.2.4.1 was used to
estimate THC emissions.  Otherwise, if the alternative, fleet-
average procedure described in Section 5.2.4.2 was used, a single
correction factor can be used for each aircraft category.


Detailed Methodology


     VOCGA PISTON  =    THCGA PISTON  X 0.9649             (5-16)

     VOCGA TURBINE =    THCGA TURBINE  X 1.0631            (5-17)

     VOCAT PISTON  =    THCAT PISTON  X 0.9649             (5-19)

     VOCAT TURBINE =    THCAT TURBINE  X 1.0631            (5-20)


Alternative, Fleet-Average Methodology

     VOCGA FLEET = THCGA FLEET X 0.9708                    (5-18)
     VOCAT FLEET = THCAT FLEET X 0.9914                    (5-21)

                                199





                  6.0 EMISSIONS FROM LOCOMOTIVES

     This chapter illustrates how a state or local agency can
calculate emissions from locomotives within an inventory area.249
Railroad locomotives used in the United States are primarily of two
types: electric and diesel-electric.250  Electric locomotives are
powered by electricity generated at stationary power plants and
distributed by either a third rail or overhead catenary system. 
Emissions are produced only at the electrical generation plant,
which is considered a point source and therefore not of interest
here.  Diesel-electric locomotives, on the other hand, use a diesel
engine and an alternator or generator to produce the electricity
required to power its traction motors.  Emissions produced by these
diesel engines are of interest in emission inventory development. 
Emissions for hydrocarbons (HQ, carbon monoxide (CO), oxides of
nitrogen (NOx), sulfur dioxide (SO2), and particulate matter (PM)
from this source are covered in this chapter.

     This chapter is a complete revision of the corresponding
chapter in the previous edition of this document.  In addition,
this chapter also updates the emission factor information that
appears in Compilation Of Air Pollutant Emission Factors, Fourth
Edition And Supplements, AP-42.  Subsequent to the publication of
this document, AP-42 will be formally updated.

     Other sources of emissions from railroad operations include
the small gasoline and diesel engines used on refrigerated and
heated rail cars.  These engines are thermostatically controlled,
working independently of train motive power, and fall in the
category of off-highway equipment which are addressed elsewhere in
this document. (See Section 3.3 of Volume IV.)

     Railroads can be separated into three classes based on size:
Class I , Class 11, and Class III.  Class I railroads251 represent
the largest railroad systems in the country. (See Appendix 6-1 for
a complete list.) Because of their size, Class I railroads operate
over a large geographic area.  Also, they carry most of the
interstate freight252 and carry most of the passenger service. 
They are required to keep detailed records of their operations and
to report yearly to the Interstate Commerce Commission (ICC).
___________________________

     249  The term inventory area can be quite diverse and may
refer to an area as large as a multi-state CMSA or an area as small
as a county, or part of a county, within a state.

     250  A third type, steam locomotives, is used in very
localized operations, primarily as tourist attractions, and
emissions from these locomotives are insignificant.  In addition,
the particulate emissions from operating steam locomotives is so
large that nearly all of it falls to the surface within 50 meters.


     251  Class I railroads are classified by the Interstate
Commerce Commission as having annual revenues greater than $93.5
million.

     252  Class I railroads carried 93 percent of total freight
revenues in 1989.

                                200





     Class II253 and III254 railroads represent the remainder of
the rail transportation system and generally operate within
smaller, localized areas.255 These smaller railroads are not
subject to the same reporting requirements, and their recordkeeping
may be less extensive.  Also, their fleet of locomotives tends to
be older, with the Class I railroads buying almost all of the new
locomotives.

     Locomotives within each of the Classes can perform two
different types of operations: line haul256 and yard (or switch). 
Line haul locomotives, which perform the line haul operations,
generally travel between distant locations, such as from one city
to another.  Yard locomotives, which perform yard operations, are
primarily responsible for moving railcars within a particular
railway yard.

     This chapter of the guidance document will be divided into six
sections plus eight appendices.  Section 6.1 will be an overview of
the recommended methodology.  Section 6.2 will specifically
describe the recommended methods for calculating the emissions from
various types of rail service based on generic or national
operating characteristics.  Section 6.3 will present procedures for
tailoring the recommended methods to more closely reflect local
operating conditions.  Section 6.4 will introduce alternative
methods which are not fully discussed in this chapter.  Section 6.5
will discuss "remotored" locomotives; locomotives which have had
their original diesel engines replaced by newer, more efficient
power plants.  Section 6.6 will illustrate the conversion factor
method recommended for converting total hydrocarbons (THC) to
volatile organic compound (VOC) emissions.

     All correspondence pertaining to this chapter of the guidance
document should be directed to:
     Emission Planning and Strategies Division
     U.S. EPA
     2565 Plymouth Road
     Ann Arbor, NE 48105
     (313) 668-4200
___________________________

     253  Class II railroads are classified by the Interstate
Commerce Commission as having annual revenues greater than $18.7
nation but less than $93.5 million.

     254  Class III railroads are classified by the Interstate
Commerce Commission as having annual revenues less than
$18.7 million.

     255  A "smaller" area can still be an area as large as a
state.  The term "smaller" is used in contrast with the large
interstate areas which are covered by Class I railroads.


     256  In this chapter, line haul operations include intermodal
freight service, mixed freight service, and passenger service.

                                201





6.1 OVERVIEW OF RECOMMENDED INVENTORY METHODOLOGY

     Three steps are necessary in order to assess locomotive
emissions within an inventory area.  First, railroad operations are
separated into three distinct categories: Class I line haul, Class
II and Class III line haul, and yard.  Second, emissions for each
pollutant are calculated for each of the three categories using
either the recommended methods described in Section 6.2 or, if
circumstances explained later occur, the alternative methods
described in Sections 6.3 or 6.4. Third, the total locomotive
emissions in the inventory area are calculated by summing the
quantities of each pollutant for each of the three categories.

     The methods illustrated in this chapter are based on annual
inventories and annual data.  Developing inventories for shorter
time periods is straight forward because railroad traffic is
relatively constant throughout the year and therefore, less than
annual calculations can be done by simple apportionment.  In
addition, the recommended methods described in Section 6.2 are
based on a national locomotive fleet mix and average fuel
consumption figures.

6.2 RECOMMENDED METHODS

     The recommended methods for each of the three categories, as
follows: Class I line haul, Class III and Class III line haul, and
yard, are discussed separately below.

6.2.1 Class I Line Haul Locomotives

     For Class I line haul locomotives, emissions are calculated by
multiplying the amount of fuel consumed in the inventory area by
the appropriate emission factors.

     Inventory Area Emissions = Fuel Consumption x Emission Factors

6.2.1.1 Fuel Consumption

     If Class I line haul locomotives only traveled within the
inventory area, fuel consumption could be determined directly from
the amount of fuel dispensed into the units.  However, these line
haul locomotives travel predominantly interstate.  Hence, they do
not necessarily bum the fuel in the same location where the fuel
was pumped, making it impossible to determine fuel consumption in
the area of interest in this manner.

     In order to determine inventory area fuel consumption, it is
necessary to allocate the total amount of fuel consumed
"systemwide" for Class I railroads to the inventory area.  This is
done by dividing the traffic density (expressed in Gross Ton Miles
or GTM) for each Class I railroad track segment within the
inventory area by the systemwide fuel consumption index (expressed
in Gross Ton Miles per gallon or GTM/gal) for that railroad.  This
process is repeated for each railroad.

                                202





     In any given area, there will be only a few active Class I
railroads, and the railroad company staff should be able to perform
this step and provide the amount of fuel consumed within the
inventory area on request.  In addition, EPA has included a
detailed explanation of how this step is performed based on
published data and information which is generally available from
each railroad.

     Fuel consumption, for each Class I railroad within an
inventory area, is therefore specifically calculated using the
following formula:

Fuel Consumption =
     Traffic Density (GTM) / Fuel Consumption Index (GTM/gal)

     The inventory area traffic density and the fuel consumption
index are described separately below.

     Traffic Density

     For every track segment within a state, each Class I railroad
maintains information on traffic density (GTM), length (miles),
direction, and geographic location.  Therefore, it is possible to
calculate the traffic density for an inventory area by summing the
traffic densities for each track segment or portion thereof within
the inventory area.

     This information can be obtained, for each area, either
directly from the individual railroads or from the Association of
American Railroads in Washington, D.C. The information should
contain enough detail so that track segments or portions thereof
can be assigned to the inventory area.  However, if the agency is
unable to perform this task, it may become necessary to obtain
assistance from the Class I railroad in order to determine where
the inventory area boundary intersects the track segment.

     The gross ton mile information may be supplied in one of two
ways.  The first way is without the weight of the locomotives
included.  The second way is with the weight of the locomotives
included.  This distinction is important when calculating the fuel
consumption index.

     Fuel Consumption Index

     The fuel consumption index (GTM/gal), for each Class I
railroad within an inventory area, should be calculated by dividing
the systemwide gross ton miles (GTM) by the systemwide fuel
consumption (gal).  See the following formula:

     Fuel Consumption Index (GTM/gal) = System Gross Ton
Miles/System Fuel Consumption


                                203





     Each Class I railroad is required to report these statistics
each year to the ICC in an annual report entitled "R-1." The R-1
report should be used, for each carrier, to obtain information on
annual fuel consumption (Schedule 750: line 1), total gross ton
miles including locomotives (Schedule 755: line 104), and, when
needed, total gross ton miles excluding locomotives (Schedule 755:
line 104 minus line 98).  An example of these schedules is included
in Appendix 6-2.

   The fuel consumption index will vary depending on whether the
weight of the locomotives is included in the calculation.257 
Also, calculating fuel consumption within the inventory area
requires the multiplication of traffic density by fuel consumption
index; therefore, it is important to match the units of each of
these components.  If traffic density is supplied without the
weight of the locomotives included, then the fuel consumption index
should be determined without the weight of the locomotives included
in the calculation.  If traffic density is supplied with the weight
of the locomotives included, then the fuel consumption index should
be determined with the weight of the locomotives included in the
calculation.

     The fuel consumption index, with locomotives, is calculated by
dividing total gross ton miles with locomotives, Schedule 755: line
104, by the total fuel consumed, Schedule 750: line 1. The fuel
consumption index, without locomotives, is calculated by dividing
total gross ton miles without locomotives, Schedule 755: line 104
minus line 98, by the total fuel consumed, Schedule 750: line 1.
Examples of these calculations are shown in Appendix 6-3.

6.2.1.2 Emission Factors

     Now that fuel consumption has been calculated, inventory area
emissions are determined by multiplying that value by the fleet
average emission factors for each pollutant (expressed in pounds
per gallon of fuel burned (lbs/gal).  The recommended default
emission factors for all line haul locomotives are shown in Table
6-1.

       Table 6-1.  Line Haul Locomotive Emission Factors258

                                        Emission Factor
                             Pollutant       (lbs/gal)
                                HC            0.0211
                                CO            0.0626
                               NOx           0.4931
                               SO2*          0.0360
                                PM            0.0116

     *  SO2 calculated based on a fuel sulfur content of 0.25
percent by weight.

___________________________

     257  The fuel consumption rate will be less if the weights of
the locomotives are not included in the calculation.

     258  Locomotive Emission Factors for Inventory Guidance
Document, Office of Mobile Sources, U.S. EPA, June 1991.

                                204





     Appendix 6-3 gives a full example of how to calculate
emissions from the Class I line haul locomotives in an inventory
area, using the Santa Fe railroad in the State of Illinois.

6.2.2     Class II and III Line Haul Locomotives

     Similar to the recommended method for Class I line hauls,
emissions from Class II and III line haul locomotives are
calculated by multiplying the amount of fuel consumed in the
inventory area by the appropriate emission factors.

Inventory Area Emissions = Fuel Consumption x Emission Factors

6.2.2.1 Fuel Consumption

     Since Class II and III railroad companies are not required to
file R-1 reports, annual fuel consumption should be obtained
directly through interviews or letters with each Class II and III
railroad operating within the inventory area.  This approach is
sufficient because, unlike Class I line haul operations, most Class
II and III line haul travel is predominantly within a relatively
small geographic area (see footnote 7).  Therefore, in many
instances, it is unnecessary to apportion system fuel use to an
inventory area, because the fuel is consumed by the locomotives
within the inventory area.

     However, for the small number of Class II and III railroads
operating outside the inventory area, EPA recommends simply
allocating the fuel consumption by track length or track density
(GTM).  Each Class II and III railroad can supply both track length
and track density information.259  So, the percentage of fuel
consumed is based on the percentage of track length or track
density within the inventory area.  If, for example, 30 percent of
the track length, for a particular railroad, runs within the
inventory area, then, in order to apportion the total fuel consumed
in the inventory area, multiply the total fuel consumption for the
railroad by 0.30.
___________________________

259  In addition, tabulations of track mile data may be available
for counties, Metropolitan Statistical Areas (MSA), or urban areas
from some state transportation agencies, or from the county or
metropolitan planning organization within whose jurisdiction the
study area is located.  Alternatively, route miles can be obtained
by direct measurement from an appropriate map, such as the County
Series maps (obtained from the state transportation agency), U.S.
Geologic Survey maps, U.S. Transportation Zone maps, or locally
prepared maps.

                                205





6.2.2.2 Emission Factors

     The emission factors for Class II and III line haul
locomotives are assumed to be the same as Class I locomotives.260 
A complete list of the emission factors is provided in Table 6-1
(See Section 6.2.1.2).

6.2.3     Yard Operations

     The recommended method for yard locomotives is different from
the method used for line haul locomotives.  Yard locomotive
emissions, for each pollutant, are derived by multiplying the
number of yard locomotives operating within the inventory area by
the amount of emissions generated by each unit during the year. 
See the formula below:

Inventory Area Emissions =    Number of Yard Locomotives x Annual
                              Emissions Per Yard Locomotive

6.2.3.1 Number of Yard Locomotives

     Since yard locomotives operate within the boundaries of a
railway yard, it is possible to calculate the number of locomotives
operating within an inventory area through inter-views with the
railway yard managers, who keep accurate records of yard locomotive
operations.  If this first approach proves unproductive, the number
of yard locomotives can be determined by actually counting the
units operating in each railway yard during a day.  This is
sufficient because the number of yard locomotives in operation each
day remains relatively constant throughout the year.  Switch yard
engines are sent to railroad maintenance facilities according to
regular schedules.  When a particular yard locomotive is away
getting maintenance or repair, the yard will replace the unit with
another of approximately the same horsepower.

6.2.3.2 Emissions Per Yard Locomotive

     EPA estimates that the average yard engine emits the following
amount of each pollutant per year:
___________________________

     260  In actuality, Class II and III railroads tend to have an
older fleet mix.  With local information, this could be included in
the calculation of an appropriate emission factor, as shown in
Section 6.3, but the improvement in accuracy is believed to be
small.

                                206





     Table 6-2.  Annual Emissions Per Yard Locomotive

                                        Annual Emissions
                             Pollutant       (lbs/yr)
                                   HC         4,174
                                   CO         7,375
                                   NOx           41,608
                                   SO2*           3,075
                                   PM         1,138

     * SO2 calculated based on a fuel sulfur content of 0.25
percent by weight.

     These emission levels were calculated as follows.  EPA
estimated that, based on a reasonable activity or duty cycle and
typical fuel consumption rates, the average yard engine consumes
228 gallons of fuel per day.261   Since yard locomotives can be
assumed to operate 365 days a year262 (this assumes that when a
yard engine is taken in for repairs it is replaced during that
period), the average yard engine consumes 82,490 (226 X 365)
gallons of fuel per year.

     EPA also determined that the average yard locomotive has the
following emission factors (lbs/gal):

       Table 6-3.  Emission Factors For Yard Locomotives263

                                        Emission Factor
                             Pollutant    (lbs/-gallon)
                                HC           0.0506
                                CO           0.0894
                               NOx          0.5044
                               SO2*         0.0360
                                PM           0.0138

     *  SO2 calculated based on a fuel sulfur content of 0.25
percent by weight.
___________________________

     261  The fuel consumption data used for this calculation were
considered proprietary by the locomotive manufacturers and hence
could not be printed in this document.

     262  Some yards operate partially or not at all on weekends. 
Information on the operation schedule for yards may be easily
obtained from the railroads, and the figure of 365 can be adjusted
accordingly.

     263   Locomotive Emission Factors for Inventory Guidance
Document, Office of Mobile Sources, U.S. EPA, June 1991.

                                207





     Therefore, the annual emissions per yard locomotive, Table 6-
2, were determined by multiplying the fuel consumption estimate
(85,410 gal/year) by each emission factor in Table 6-3.

6.3  TAILORING METHODS

     EPA recognizes that railroad operations may vary significantly
from the national average and that some state and local air quality
agencies may have access to more detailed information regarding the
locomotive activity in their inventory area.  Because of this, EPA
has included two additional methodologies to allow these agencies
to tailor the emissions calculations based on actual locomotive
fleet, or roster, data and local operational characteristics. 
Using these tailoring methods is not required, but they have been
included here should a state or local agency decide that a more
precise calculation is desirable.

     As explained in Section 6.2, the recommended method for
calculating total emissions in an inventory area requires
multiplying a fleet averaged emission factor by fuel consumption. 
An implicit element of the composite emission factor is the
locomotive roster.  If the actual roster for an inventory area is
different from the one used in the recommended method, then the
composite emission factor, calculated using the recommended method
for the inventory area, could be different than if the composite
emission factor were calculated using the actual locomotive roster.

     Another element implicit in both the composite emission factor
and the estimate of fuel consumption is the duty cycle which an
engine goes through during operation.  If the actual duty cycle in
an inventory area is different from that assumed in the recommended
method, then the values used in the recommended method for the
inventory area could be different than if the values were
calculated using the actual duty cycle for that area.

6.3.1     Locomotive Roster Tailoring Method

     The roster tailoring method requires the development of an
area specific roster, and subsequently the calculation of new fleet
average emission factors.  These new emission factors will then be
substituted for the national fleet average emission factors in
Section 6.2, and subsequently, will be multiplied by the fuel
consumption figure to give the emissions for the inventory area. 
The tailoring method for both line haul and yard locomotives will
be the same, but, the new tailored rosters should be calculated
separately.  The roster tailoring method is composed of the
following steps:

6.3.1.1   Identify the locomotives in the area

     The first step in the roster tailoring method is to identify
all of the locomotives within an inventory area.  Identification
should be made based on make and model number (EMD GP9 for
example).

                                208





     EPA recognizes that identifying Class I line haul locomotives
within an inventory area may be difficult because these locomotives
travel predominantly interstate.  EPA recommends that the agency
performing the emissions inventory contact each Class I railroad
within the inventory area and ask for an estimate of the number and
type of locomotives operating within the area.

     If it is not possible to identify the Class I line haul
locomotives within the inventory area using the above procedure,
EPA suggests, as a next best alternative, that the agency
performing the emissions inventory contact each Class I railroad
operating within the inventory area and request a copy of its
systemwide locomotive roster.  The agency may then assume that
these locomotives operate on routes that enter or lie within the
inventory area.

     Identifying Class II and III line haul locomotives should be
easier than identifying Class I line hauls, due to the fact that
the former locomotives travel predominantly within a small area. 
EPA again recommends contacting each Class II and III railroad
which operates within the inventory area and requesting a copy of
its locomotive roster.

     Identifying yard engines within an inventory area has already
been explained in Section 6.2.3.1, and that approach should be
followed here as well.

     Once the locomotives have been identified, the agency
calculating the emissions within the inventory area should create a
list, including number of engines and model type, similar to the
one listed as "Locomotive Model" in Appendix 6-4.  "Locomotive
Model" is a detailed listing of most of the locomotives models
operating today which could exist in any fleet. (For Example:
"GP9", "SD50", "WOC", etc.)

6.3.1.2 Determine the engine type

     The second step is to convert from the locomotive model to a
specific engine type for which EPA has emissions data.  For
example, a "GP9" model number would equate to a "16-567C" diesel
engine type.  Again, Appendix 64 shows how this conversion would
appear.

6.3.1.3 Sum the total of the conversions

     Once all of the conversions have been made, the third step is
to calculate the new inventory area roster by summing the total of
the conversions and then calculating the percent of the total for
each engine type.  Appendix 6-5 illustrates how this procedure was
performed for the national roster used in Section 6.2.

                                209





     It is important to remember that it is possible for some
engines to perform both yard and line haul service (usually at
different times).  If this is the case, the same engine type may be
entered as both a line haul and a yard engine.264

6.3.1.4   Calculate the new fleet average emission factors

     The fleet average emission factors given in the recommended
methods (Section 6.2) were derived using a national roster
(Appendix 6-5).  Since the roster changes with the tailoring
method, the fleet average emission factors will change as well. 
The fourth step, therefore, is to re-calculate the emission factors
using the new roster.  To perform these calculations, simply
multiply the new roster percentage for each engine by the emission
factor (Appendix 6-6) for each pollutant for that engine.  Appendix
6-7 illustrates how this is done.

6.3.1.5   Multiply the new emission factors by fuel consumption

     Now that the new emission factors have been calculated for
both line haul and yard locomotives based on the new roster,
emissions are calculated by multiplying by fuel consumption.  This
procedure is the same as the recommended methodology which is
explained in Section 6.2.

6.3.2     Duty Cycle Tailoring Method

     The Agency believes that the duty cycles which are indicated
in the recommended methods are reasonably representative of
railroad operations across the nation, including nonattainment
areas.  These duty cycles, for line haul and yard locomotives, are
shown in the following table and again in Appendix 6-9.
___________________________

     264  When the national roster was calculated, locomotives
which might perform both functions were not entered as indicated
here; thus Appendix 6-5 does not show any overlap of any
locomotives.

                                210





                Table 6-4.  Locomotive Duty Cycles

                                  Time in Notch (%)
          Throttle Notch         Line Haul   Yard
               8                     1         1
               7                     3         0.5
               6                     4         0.5
               5                     4         1
               4                     5         2
               3                     4         4
               2                     4         7
               1                     4         7
               Idle                 49        77
               Dynamic Brake        12         0
               Total               100       100


     If local conditions are likely to result in duty cycles that
are substantially different from those shown above, State or local
authorities may consider adjusting the emissions inventory
methodology to incorporate this information.  Any modifications,
however, should be done in consultation with local railroad
officials to determine if actual duty cycle measurements are
available for this purpose, or if less specific adjustments are
appropriate, based on past operating experience.

     The most significant effect of altering the duty cycle would
be to change the amount of fuel that is estimated for the inventory
area.265  Generally, the fuel consumption rate of a locomotive
engine is determined by throttle notch position.  The more time a
locomotive spends in higher throttle notches, the more fuel it
uses.  Conversely, the more time spent in lower throttle notch
positions, the less fuel is used.  Therefore, the amount of fuel
consumed is generally proportional to the firne a locomotive spends
in each throttle notch position.

     The preferred approach for incorporating a tailored duty cycle
into the analysis is for State or local authorities to recalculate
the fuel consumption of each locomotive model in the inventory
area, using the new time-in-notch data and the appropriate fuel
consumption rate for each throttle position.  Unfortunately, EPA is
unable to provide notch-specific fuel consumption rates in this
document because the locomotive manufacturers have claimed that
such information is proprietary.  Fortunately, there are two
alternative methods for incorporating alternative duty cycles
without having these data.
___________________________

     265  This effect pertains only to the fuel consumption value
obtained by apportioning systemwide fuel to the inventory area
(Section 6.2.1.1 or 6.2.2.1), or the estimated fuel consumption
value for each yard locomotive (Section 6.2.3.2).

                                211






     The first approach is to use the information obtained in
consultation with railroad officials to develop a "general
correction factor" that can be applied to the fuel consumption
estimates determined with the recommended methods in Section 6.2.
The second approach is to contact EPA at the address previously
provided in this chapter with the alternative duty cycle data. 
After analyzing this information in conjunction with the State or
local agency, EPA could provide a "tailored correction factor" for
the inventory area.

     A secondary effect of altering the duty cycle is to change the
emission factors (i.e., pounds of pollutant per gallon of fuel ). 
This occurs because the combustion characteristics of the engine
vary by throttle notch position.  Also, the significance of this
effect varies for each of the primary pollutants from this source
(i.e., HC, CO, and NOx ).

     In most internal combustion engines, fuel is usually burned
more nearly completely as the throttle position increases. 
Generally, this results in a smaller mass of certain pollutants
being produced for each unit of fuel consumed.  For locomotives,
this effect is most significant for HC emissions, less so for CO
emissions, and essentially non-existent for NOx emissions (i.e.,
the amount of this pollutant is essentially constant per unit of
fuel regardless of notch position).  Therefore, only state or local
agencies that are particularly concerned about HC emissions from
railroad operations are likely to benefit from including this
effect in the inventory.

     Accounting for this phenomenon is similar to that discussed
above for estimating the amount of fuel used in the inventory area,
because the preferred methodology involves the use of proprietary
notch-specific fuel consumption data.  As a consequence, any agency
wishing to include this aspect of duty cycle tailoring in the
emissions inventory must contact EPA with the requisite information
for a customized composite emission factor.

6.3.3     SO2 Tailoring Method

     The emission factors for SO2 are calculated based on the
amount of sulfur contained in the fuel.266  These emission factors
are based on a sulfur content of 0.25 percent sulfur by weight. 
EPA recognizes that the amount of sulfur contained in diesel fuel
may vary significantly from one inventory area to another.  In
order tailor to sulfur emission factors based on the inventory area
fuel sulfur content, a recalculation should be performed by first
calculating an adjustment factor and then multiplying by the sulfur
emission factor.  Tailored Sulfur = Adjustment Factor x Guidance
Document SO2 Emission Factor.
___________________________


     266  It is assumed that all of the elemental sulfur in the
fuel is oxidized into SO2.

                                212





     The adjustment factor is determined by dividing the inventory
area sulfur percentage by the existing sulfur percentage.  If the
inventory area had a fuel sulfur content of 0.20 percent, then the
adjustment factor would be 0.80 (0.20/0.25). In order to
recalculate the emission factor, multiply the sulfur emission
factor, 0.0340 lbs/gal, by the adjustment factor, 0.80, to get an
answer of 0.0272 lbs/gal.

6.4  ALTERNATIVE METHOD

     EPA believes that the line haul locomotive methods presented
in this chapter are practical, feasible, and accurate.  However, if
for some unforeseen reason it is not possible to acquire, or
apportion, fuel consumption as required by these methods, an
alternative approach, such as the one described in EPA's Report to
Congress On Railroad Emissions - A Study Based On Existing Data,
(no date available at time of printing) may be appropriate.  This
particular alternative, if needed, may be most appropriate for
inventorying congested urban areas.  If it proves necessary to
consider any alternatives, please contact EPA at the address
previously provided in this chapter for additional details and
assistance.

6.5  RE-ENGINED LOCOMOTIVES

     EPA recognizes that some older locomotives have been re-
engined with newer, more efficient engines.  Some of these engines
are manufactured by EMD and GE and are most likely represented in
the data base.  However, a small number of older GE and EMD chassis
also have been retrofitted with engines from other manufacturers,
such as Caterpillar, Sulzer (Germany), and Cummins.  EPA
acknowledges the existence of these engines, but because there are
so few engines of this type in the fleet and because EPA has not
been able to obtain emission data on these engines, they are not
included in the data base.

6.6  CONVERTING FROM TOTAL HYDROCARBONS (THC) TO VOLATILE ORGANIC
     COMPOUNDS (VOC)

     EPA recognizes that it may be necessary to determine the level
of volatile organic compounds (VOC) emitted from locomotives. 
Since the emission factors for HC contained in this document
represent total hydrocarbons as measured by a flame ionization
detector (THCFID), this section illustrates the conversion factor
method recommended for converting THCFID to VOC emissions.

     The method listed below was derived from an April 21, 1992,
EPA memorandum from Greg Janssen to Phil Lorang entitled "THC to
VOC Correction Factors for Nonroad Emissions Inventories" (the
memorandum).  Since locomotive emissions are created by large
diesel engines, and since locomotive diesel correction factors do
not exist, EPA has assumed that the correction factors for heavy-
duty diesel vehicles (HDDV) can also be used as locomotive
correction factors.

                                213





     The method for converting THC to VOC was based on the
methodology for nonroad conversion and is as follows:

                                      NMHCFID HDDV      VOC HDDV
VOCLocomotive = THCFID Locomotive X ________________ X
____________

                                       THCFID HDDV    NMHCFID
                                   HDDV
where:

     THCFID Locomotive  represents the total hydrocarbon emissions
                         measured from locomotives by FID,

       NMHCFID HDDV
     __________________

        THCFID HDDV          is a non-methane hydrocarbon
                              correction factor and represents the
                              ratio of non-methane hydrocarbons to
                              total hydrocarbons, as emitted by
                              heavy duty diesel vehicles, and

     VOC HDDV
   _______________

     NMHCFID HDDV       is a VOC correction factor and represents
                         the ratio of VOC to non-methane
                         hydrocarbons, as emitted by heavy duty
                         diesel vehicles.

     Using the same method illustrated in Attachment 3 of the
memorandum, EPA determined that NMHCFID HDDV = 1.07 g/mile and
THCFID HDDV = 1.10 g/mile.267  Furthermore, the VOC correction
factor, as derived for HDDV in Appendix 3 in the memorandum, was
determined to equal 1.0332.

     Thus, EPA has determined that the correction factor to
determine VOC emissions from locomotives is as follows:

                                    1.07
VOCLOCOMOTIVE = THCFID Locomotive X _____  X 1.0332

                                    1.10

                                or

VOCLOCOMOTIVE = THCFID Locomotive X 1.005

___________________________

     267  Springer, 1979 (EPA-460/3-79-007) (PP. 47,98)

                                214





                           Appendix 6-1
     Class I Railroad Systems In The United States as of 4/91*

AMTRAK

Atchison, Topeka and Santa Fe Railway Company (Santa Fe)
     (AZ, CA, CO, IL, IA, KS, LA, MO, NB, NM, OK, TX)

Burlington Northern Railroad Company
     (AL, AR, CA, CO, FL, IA, ID, IL, KS, KY, MN, MO, MT, MS, NM,
     ND, NE, OK, OR, SD, TN, TX, WA, WI, WY)

Chicago and North Western Transportation Company

Consolidated Rail Corporation (Conrail)
     (CT, D.C., DE, EL, IN, KY, MA, MD, MI, NJ, NY, OH, PA, VA, WV)

CSX Corporation
     CSX Transportation, Inc. (Includes Chessie System and Seaboard
     System)
     (AL, DE, D.C., FL, GA, IN, IL, KY, LA, MD, Ng, MS, NC, OH, PA,
     SC, TN, VA, WV)

Denver and Rio Grande Western Railroad

Florida East Coast Railway
     (FL)

Grand Trunk Corporation
     Grand Trunk Western Railroad

Guilford Industries
     Boston and Maine Corporation

Illinois Central Railroad
     (IL, TN, KY, MS, LA)

Kansas City Southern Railway

Norfolk Southern Corporation
     Norfolk and Western Railway
     Southern Railway System

Soo Line Railroad

Southern Pacific Transportation Company
     St. Louis Southwestern Railway

Union Pacific Railroad Corporation
     Missouri Pacific Railroad

     *  Sources:    Association of American Railroads and Federal
                    Railroad Administration

                                215




Click HERE for graphic.


                                216





                           Appendix 6-3

       Sample Calculation of Inventory Area Fuel Consumption
                     for Santa Fe in Illinois

Inventory Area Emissions = Fuel Consumption x Emission Factors

Fuel Consumption = Traffic Density x Fuel Consumption Index


Traffic Density Without Locomotives

If the traffic density data for Illinois are supplied without
locomotive weight included, emissions would be calculated as
followed:

Traffic Density

     Santa Fe Traffic Density in Illinois:
     (Furnished by Santa Fe without
     locomotives) = . . . . . . . . . . . . . . . 7,329,000,000 GTM

Fuel Consumption Index
     Santa Fe System Fuel Consumption:
     Schedule 750: line 1 = . . . . . . . . . . . . 304,370,694 gal

     Santa Fe System
     Gross Ton Miles (w/o locomotives):
     Schedule 755: line 104 - line 98
     (186,661,355,000 - 24,631,118,000) = . . . 162,030,237,000 GTM

     Santa Fe Fuel Consumption
     Index (w/o locomotives)
     (162,030,118,000 / 304,370,694) =. . . . . . . . . 532 GTM/gal

Fuel Consumption
     Fuel Consumption for
     Santa Fe in Illinois
     (7,329,000,000 / 532) =. . . . . . . . . . . . .13,776,316 gal

                                217





Traffic Density With Locomotives

if the traffic density data for Illinois are supplied with
locomotive weight included, emissions would be calculated as
follows:

Traffic Density

     Santa Fe Traffic Density in Illinois:
     (Furnished by Santa Fe with
     locomotives) = . . . . . . . . . . . . . . . 8,445,000,000 GTM

Fuel Consumption Index

     Santa Fe System
     Fuel Consumption:
     Schedule 750: line 1 = . . . . . . . . . . . . 304,370,694 gal

     Santa Fe System
     Gross Ton Miles (with locomotives):
     Schedule 755: line 104 = . . . . . . . . . 186,661,355,000 GTM

     Santa Fe Fuel Consumption
     Index (with locomotives)
     (186,661,355,000 / 304,370,694) =. . . . . . . . . 613 GTM/gal

Fuel Consumption

     Fuel Consumption for
     Santa Fe in Illinois
     (8,445,000,000 / 613) =. . . . . . . . . . . . .13,776,508 gal

Emission Factors

The emission factors for line haul locomotives are located in Table
6-1 above.  Emissions can now be calculated as follows:

Emissions (Tons) =

        Emission Factor (lbs/gal) x Fuel Consumption (gal)
     _________________________________________________________

                               2,000

     Example for HC: (0.0211 X 13,776,508)/2,000 = 150.16 Tons

                                218





                           Appendix 6-4
         Conversions from Locomotive Model to Engine Type

                                EMD

     Locomotive                                              Engine
       Model                    HP                            Type 
______________________     ____________                  __________
     E8A                       2250                        12-567BC

     F40C                      3200                        16-645E3
     F40PH                     3000                       16-645E3B
     F40PH-2                   3200                       16-645E3B
     F45                       3600                        20-645E3
     FP45                      3600                        20-645E3

     GP7                       1500                        12-567BC
     GP9                       1750                         16-567C
     G18U                      1100                          8-645E
     G18W                      1100                          8-645E
     Gl8AlA                    1100                          8-645E
     GP15-1                    1500                         12-645E
     GP15T                     1650                        8-645E3C
     GP18                      1800                         16-567C
     GP20                      2000                         16-567C
     GP28                      1800                         16-567C
     GP30                      2250                        12-645E3
     GP35                      2500                        12-645E3
     GP38                      2000                         16-645E
     GP38-2                    2000                         16-645E
     GP38-2P                   2000                         16-645E
     GP38-AC                   2000                         16-645E
     GP39-2                    2300                       12-645E3B
     GP40                      3000                        16-645E3
     GP40-2                    3000                       16-645E3B
     GP40-P                    3000                       16-645E3B
     GP40P-2                   3000                       16-645E3B
     GP40X                     3500                        16-645F3
     GP49                      2850                       12-645F3B
     GP50                      3500                        16-645F3
     GP59                      3200                   12-71OG3/G3A+
     GP60                      3600                   16-71OG3/G3A*
     GP60M                     3800                       16-71OG3A


     +  For locomotives built after 4/91, use the 12-71OG3A
     *  For locomotives built after 4/91, use the 16-71OG3A

                                219





                                EMD
                     ________________________
     Locomotive                                              Engine
       Model                    HP                            Type 
  _______________          _____________               ____________

     MP15AC                    1500                         12-645E
     MP15                      1500                         12-645E
     MP15T                     1650                        8-645E3C

     SD9                       1750                         16-567C
     SD35                      2500                        12-645E3
     SD38                      2000                         16-645E
     SD38-2                    2000                         16-645E
     SD38AC                    2000                         16-645E
     SD39                      2500                        12-645E3
     SD40                      3000                        16-645E3
     SD40-2                    3000                        16-645E3
     SD40A                     3000                        16-645E3
     SD4OT2                    3000                        16-645E3
     SD40X                     3500                        16-645F3
     SD45                      3600                        20-645E3
     SD45-2                    3600                        20-645E3
     SD50                      3500                       16-645F3B
     SD60                      3800                        16-71OG3
     SD60M                     3800                   16-71OG3/G3A*
     SDF40-2                   3000                        16-645E3
     SDF45                     3600                        20-645E3
     SDP45                     3600                        20-645E3
     SDL39                     2300                        16-645E3
     SDP40                     3000                        16-645E3
     SDP40FM                   3000                        16-645E3
     SW7                       1200                        12-567BC
     SW900                      900                          8-645E
     SW1000                    1000                          8-6453
     SW1001                    1000                          8-645E
     SW1200                    1200                        12-567BC
     SW1500                    1500                         12-645E

     *  For locomotives built after 4/91, use the 16-71OG3A

                                220





                                GE
                    ___________________________

     Locomotive                                              Engine
     Model                      HP                            Type 
  _________________       _______________              ____________

     B23-7                     2250                         12-2500
     B30-7                     3000                         16-3000
     B30-7A                    3000                         16-3000
     B30-7AB                   3000                         16-3000
     B36-7                     3600                         16-3600
     B36-8                     3600                         16-3600
     B39-8                     4100                         16-4100
     B40-8                     4100                         16-4100
     B40-8W                    4100                         16-4100

     C30-7                     3000                         16-3000
     C30-7A                    3000                         16-3000
     C32-8                     3000                         16-3000
     C36-7                     3600                         16-3600
     C39-8                     4100                         16-4100

     DASH 8-32B                2500                         12-2500
     DASH 8-40C                4100                         16-4100

     U23B                      2500                         12-2500
     U23C                      2500                         12-2500
     U25C                      2500                         12-2500
     U25C                      2500                         12-2500
     U28B                      3000                         12-3000
     U28C                      3000                         12-3000
     U30B                      3000                         12-3000
     U30C                      3000                         12-3000
     U33B                      3000                         16-3000
     U33C                      3000                         16-3000
     U34CG                     3600                         16-3600
     U36B                      3600                         16-3600
     U36C                      3600                         16-3600
     U36CG                     3600                         16-3600

                                221




                           Appendix 6-5

                    National Fleet Percentages

Category  Engine Type         Number    Percentage of Total
Line Haul 16-645E3             1562     16.1
          16-645E3B            2693     27.7
          16-645F3              232     2.4
          16-645F3B             400     4.1
          20-645E3              723     7.4
          16-71OG3              537     5.5
          16-71OG3A             250     2.6
          12-2500               843     8.7
          12-3000               145     1.5
          12-3300                0      0.0
          16-3000               801     8.3
          16-3600               451     4.6
          16-4100              1029     10.6
          12-645F3B              6      0.1
          12-71OG3               2      0.0
          12-71OG3A             34      0.4


Yard      12-567BC              131     2.9
          12-645E              1216     26.7
          16-567C              1279     28.1
          16-645E              1763     38.8
          12-645E3B             125     2.7
          12-645E3              32      0.7
          8-645E                 1      0.0
          8-645E3C              42      0.8

                                222





                           Appendix 6-6

                 Emission Factors For Locomotives
                    Emission Factors (lbs/gal)

Eng. Type       HP    Operation    HC        CO      NOx   PM

EMD 8-645E     1100      Yard   0.0525     0.1571   0.4984  0.0162

EMD 8-645E3C   1650      Yard   0.0223     0.0477   0.7232  0.0121

EMD 12-567BC   1200      Yard   0.2209     0.1608   0.3936  0.0165

EMD 12-645E    1500      Yard   0.0321     0.0752   0.5141  0.0138

EMD 12-645F3B  2950   Line Haul 0.0155     0.0490   0.4962  0.0104

EMD 12-645E3   2300   Line Haul 0.0192     0.0939   0.5156  0.0113
                         Yard   0.0277     0.1130   0.5803  0.0135

EMD 12-645E3B  2500   Line Haul 0.0196     0.0878   0.5475  0.0115
                         Yard   0.0266     0.0761   0.5107  0.0133

EMD 12-710G3   3200   Line Haul 0.0155     0.0373   0.4306  0.0103

EMD 12-710G3A  3200   Line Haul 0.0070     0.0447   0.4568  0.0108

EMD 16-567C    1750   Line Haul 0.0284     0.0857   0.5109  0.0114
                         Yard   0.0435     0.0784   0.4452  0.0136

EMD 16-645E    2000   Line Haul 0.0192     0.0604   0.5443  0.0114
                         Yard   0.0308     0.0747   0.5243  0.0136

EMD 16-645E3   3000   Line Haul 0.0198     0.0721   0.4744  0.0114

EMD 16-645E3B  3000   Line Haul 0.0160     0.0460   0.5236  0.0071

ENM 16-645F3   3500   Line Haul 0.0178     0.0529   0.6060  0.0116

EMD 16-645F3B  3600   Line Haul 0.0153     0.0295   0.6687  0.0111

EMD 16-710G3   3600   Line Haul 0.0170     0.0244   0.4986  0.0111

EMD 16-710G3A  3600   Line Haul 0.0091     0.0935   0.4862  0.0110

EMD 20-645E3   3800   Line Haul 0.0200     0.0528   0.4590  0.0113
GE 12-2500     2500   Line Haul 0.0229     0.0970   0.4271  0.0117

                                223





                 Emission Factors For Locomotives

                    Emission Factors (lbs/gal)

Eng. Type       HP  Operation   HC        CO       NOx          
PM
GE 12-3000     3000 Line Haul 0.0229    0.0852    0.4572    0.0115

GE 16-3000     3000 Line Haul 0.0354    0.1025    0.4471    0.0196

GE 12-3300     3300 Line Haul 0.0229    0.0852    0.4572    0.0115

GE 16-3600     3600 Line Haul 0.0311    0.0780    0.4663    0.0173

GE 16-4100     4100 Line Haul 0.0302    0.0659    0.4851    0.0171

                                224





                           Appendix 6-7
        Sample Roster Tailoring For Line Haul Locomotives*

Step 1 - Identify the locomotives in the area

Assume that there are ten line haul locomotives which operate
within the inventory area: 5 -EMD GP59's and 5 - GE C36-7's

Step 2 - Determine the data base equivalent

According to the list in Appendix 6-5, the two locomotives convert
as follows:

     Locomotive     Data Base Equivalent
     5 - GP59       5 - 16-710G3
     5 - C36-7      5 - 16-3600

Step 3 - Calculate the tailored roster

     The new tailored roster for line haul   locomotives looks as
follows:

     Line Haul

     Engine Type    Population     Fraction of Total
   _____________   ____________   ____________________

     16-710G3             5             0.50
     16-3600              5             0.50
                       ______         _________
               Total     10             1.00

Step 4 - Locomotive emissions factors

Select the appropriate emission factors for each engine type from
Appendix 6-6.  The emission factors for each locomotive in the
tailored roster are:

     EMD IL-710G3           GE 16-3600
   _________________     __________________

 Pollutant     lbs/gal   Pollutant      lbs/gal
     HC        0.0161      HC           0.0274
     CO        0.0233      CO           0.0624
     NOx           0.4805      NOx         0.4655
     PM        0.0107      PM           0.0157
     SO2      0.0360      SO2         0.0360


     *  Note:  The same procedures would be followed for yard
               locomotives.


                                225






Step 5 - Calculate new emission factors

The new fleet average emission factors are calculated by
multiplying the engine roster percentage by the engine emission
factor and then summing the weighted emission factors for all
engines.  For this example, the result is as follows:

HC emission factor: [(0.50 x 0.0161) + (0.50 x 0.0274)] = 0.0218


                                226





                           Appendix 6-8
                            Duty Cycles

Line haul duty cycle data were taken from the Average California
Profile as determined in the Booz-Allen & Hamilton, Inc. report
entitled Locomotive Emission Study, 1991.  It represents the most
extensive analysis of time in notch profiles for line haul
locomotives operating in both mountainous and flat terrain and, in
terms of actual testing done, it is without equal.  It is also
reasonably consistent with the existing EMD and GE line haul duty
cycles.  EPA believes that the Average California Profile
adequately represents average line haul locomotive travel
throughout the nation.

Line Haul

     Notch           Percent of Time In Notch  Hrs Per Day In Notch
  ____________       ________________________   ___________________
     8                          11                             2.64
     7                           3                             0.72
     6                           4                             0.96
     5                           4                             0.96
     4                           5                             1.20
     3                           4                             0.96
     2                           4                             0.96
     1                           4                             0.96
     Idle                       49                            11.76
     Dyn Brk                    12                             2.88


Yard

Yard duty cycle data were taken from the Report to Congress.  This
duty cycle was the same as the EMD duty cycle and is consistent
with most other recommended duty cycles.

     Notch           Percent of Time In Notch  Hrs Per Day In Notch
  ___________         _______________________   ___________________
     8                           1                             0.24
     7                          0.5                            0.12
     6                          0.5                            0.12
     5                           1                             0.24
     4                           2                             0.48
     3                           4                             0.96
     2                           7                             1.68
     1                           7                             1.68
     Idle                       77                            18.48

                                227








(AQP.html)
Jump To Top