Peer Review of the Optimization Model for Reducing Emissions of Greenhouse Gases from Automobiles (OMEGA) and EPA's Response to Comments


   Peer Review of the Optimization Model
   for Reducing Emissions of Greenhouse
   Gases from Automobiles (OMEGA)
   and EPA's Response to Comments
United States
Environmental Protection
Agency

-------
                Peer Review of the Optimization Model
                for Reducing Emissions of Greenhouse
               Gases from Automobiles (OMEGA) and
                      EPA's Response to Comments
                               Assessment and Standards Division
                              Office of Transportation and Air Quality
                              U.S. Environmental Protection Agency
                                    Prepared for EPA by
                                 Southwest Research Institute
                                 EPA Contract No. EP-C-05-018
                                  Work Assignment No. 4-01
v>EPA
                 NOTICE

                 This technical report does not necessarily represent final EPA decisions or
                 positions. It is intended to present technical analysis of issues using data
                 that are currently available. The purpose in the release of such reports is to
                 facilitate the exchange of technical information and to inform the public of
                 technical developments.
United States                                       EPA-420-R-09-016
Environmental Protection                                 _  ^  ,  „„_
Agency                                          September 2009

-------
In May, 2009, EPA contracted with the Southwest Research Institute (SwRI) to conduct a
peer review of a new model which evaluates the technology and cost associated with reducing
greenhouse gas (GHG) emissions from light-duty motor vehicles. This model is currently
referred to as the Optimization Model for reducing Emissions of Greenhouse gases from
Automobiles (OMEGA). When peer review was initiated, the model was referred to as the
Vehicle Greenhouse Gas Emissions Cost and Compliance Model (VECTOR).

The three peer reviewers selected by SwRI were John German, Dr. Paul Leiby and Dr.
Rubin. EPA would like to extend its appreciation to all three reviewers for their efforts in
evaluating this model. The three reviewers brought useful and distinctive views to all aspects of
the model's use. This included setting up the input files appropriately, running the model
efficiently, and understanding the model's outputs. There are two major sections to this report.
The first section contains the final SwRI report summarizing the peer review of OMEGA,
including the detailed comments of each peer reviewer and an overview of the most significant
comments compiled by SwRI. The SwRI report also contains the peer review charge letter. The
second major section contains our responses to the peer reviewers' comments. In this section,
we repeat the detailed comments from each commenter and, after each section of comments,
provide our response. We have retained the organization reflected in each reviewer's comments
to aid the reader in moving from the SwRI report to our responses.
Table of Contents

1. Peer Review of the EPA Optimization Model for reducing Emissions of Greenhouse
gases from Automobiles (OMEGA) conducted by Southwest Research Institute
I. Introduction
II. Technical Discussion
III. Summary
IV. Appendicies
A. Resumes of Peer Review Panel Members
B. Charge Letter to Reviewers
Attachment 1 - EPA Vehicle GHG Emission Cost and Compliance
Model Description
Attachment 2 - Appendicies to GHG Model Description
Attachment 3 - Model Reference Guide
Attachment 4 - Benefits Calculations Instructions
C. John German' s Revi ew D ocument
D. Paul Leiby's Review Document
E. Jonathan Rubin's Review Document

2. EPA's Response to Peer Review Comments
A. Comments by John German
B. Comments by Paul Leiby
C. Comments by Jonathan Rubin

-------
SOUTHWEST   RESEARCH   INSTITUTE®

6220 CULEBRA ROAD • POST OFFICE DRAWER 28510 • SAN ANTONIO, TEXAS, USA 78228-0510 • (210)684-5111 • www.swri.org
ENGINE, EMISSIONS, AND VEHICLE RESEARCH DIVISION                           HTTP://ENGINEANDVEHICLE .SWRI.ORG

FAX: (210) 522-3950                                                                 ISO 9001 Certified
                                                                              ISO 14001 Certified
               Peer Review of the EPA Optimization Model for
         reducing Emissions of Greenhouse gases from Automobiles
                                   (OMEGA)
                            Office of Transportation Air Quality
                           U.S. Environmental Protection Agency
                                     September, 2009
                                    SAN  ANTONIO,  TEXAS


                         HOUSTON, TEXAS • WASHINGTON, DC  • ANN ARBOR, Ml

-------
SOUTHWEST RESEARCH INSTITUTE®
6220 CULEBRA ROAD • POST OFFICE DRAWER 28510 • SAN ANTONIO, TEXAS, USA 78228-0510 • (210)684-5111 • www.swri.org
ENGINE, EMISSIONS, AND VEHICLE RESEARCH DIVISION HTTP://ENGINEANDVEHICLE .SWRI.ORG
FAX: (210) 522-3950 ISO 9001 Certified
ISO 14001 Certified
In May, 2009, EPA contracted with the Southwest Research Institute (SwRI) to conduct a
peer review of a new model which evaluates the technology and cost associated with reducing
greenhouse gas (GHG) emissions from light-duty motor vehicles. This model is currently
referred to as the Optimization Model for reducing Emissions of Greenhouse gases from
Automobiles (OMEGA). When peer review was initiated, the model was referred to as the
Vehicle Greenhouse Gas Emissions Cost and Compliance Model (VECTOR).

HOUSTON, TEXAS • WASHINGTON, DC • ANN ARBOR, Ml

-------
SOUTHWEST   RESEARCH   INSTITUTE
6220 CULEBRA ROAD • POST OFFICE DRAWER 28510 •
ENGINE, EMISSIONS, AND VEHICLE RESEARCH DIVISION
FAX: (210)522-3950
                               SAN ANTONIO, TEXAS, USA 78228-0510 •  (210)684-5111  • www.swri.org
                                                        HTTP://ENGINEANDVEHICLE.SWRI.ORG
                                                                       ISO 9001 Certified
                                                                      ISO 14001 Certified
                                                June 9, 2009
To:           Environmental Protection Agency
 Contracts             Management Division
              26 West Martin Luther King Drive
 Cincinnati,             OH 45268

Attention:    Ms. Almethyist A. Chambers
 Contracting             Officer

From:        Patrick M. Merritt
 Senior             Research Scientist
              Emissions Research and Development Department
 Southwest             Research Institute
 P.O.             Drawer 28510
              San Antonio, Texas  78228-0510

              Subject: Final Report for Work Assignment 4-1, "Facilitation of an Independent
              Peer Review Process for EPAs VGHG Model"

              Contract No. EP-C-05-018, under SwRI Project 03.14658.01
              Contract Title: "Testing and Analytical Support for Regulation of Motor Vehicles,
              Engines, Fuels, and Fuel Additives"
I.
INTRODUCTION
       On-road vehicles are the predom inant source of greenhouse gas (GHG) em issions in the
transportation sector (principally, CO2 and hydrocarbon emissions from vehicle air conditioners).
Of all on-road vehicles, light- duty passenger cars and trucks pr oduce the majority of these GHG
emissions. As EPA's Office of Transportation and Air Quality explores the regulation of CO2 and
other GHG emission control measures in on-road and non-road vehicles and equipment, there is
a need to evaluate the costs and benefits of any such regulations. As such, EPA has developed its
Vehicle Greenhouse Gas Emissions Cost and Compliance Model, or VGHG model, to facilitate
its analysis of the costs and benefits of th    e control of GHG e  missions from cars and trucks.
Broadly speaking, the primary cost of GHG emission control is the cost of adding technology to
the vehicles, while the p rimary benefit is the value of reduced fuel cons umption in those sam e
vehicles.
                                   SAN  ANTONIO,  TEXAS

                         HOUSTON, TEXAS  • WASHINGTON, DC • ANN ARBOR, Ml

-------
Ms. Almethyist A. Chambers
Environmental Protection Agency
June 9, 2009
Page 4 of 8
The EPA VGHG model is used to apply various technologies to a defined set of vehicles
in order to meet a specified GHG emission target, and to then calculate the costs and benefits of
doing so. The GHG target can be a flat standard app licable to all vehicles within a vehicle class
(e.g., cars, trucks or both cars and trucks) or the 'target' can be in the for m of a c urve which
varies the target as a function of a defined vehicle ' fleet.' GHG emission targets are specified in
terms of CO 2-equivalent emissions. They can simply be CO 2 emissions from the tailpipe or can
be a combination of tailpipe CO 2 and refrigerant emissions. Fleet wide average GHG e missions
can also be used to estim ate a wide array of societal cos ts and benefits associated with the
reduction of GHG emissions.

To assure the highest quality science in its p redictive as sessments, EPA has engaged
SwRI to facilitate an independent peer review of its model for determining the costs and benefits
of changes to vehicle technology for the reduc tion of GHG em issions from passenger cars and
trucks. EPA needs assurance that the proposed structure (and development process) of its VGHG
model will result in a model tha t is viable, accurate, and well-suited for the diversity of uses to
which it may be applied. This report documents the process followed towards that end.
II. TECHNICAL DISCUSSION

EPA's peer review gu idelines specify that all h ighly significant scientific and technical
work products undergo independent peer review per specific agency protocols. The first task was
to select panel members (three) who were qualified, interested, and had time available to devote
to an in-depth review.

Selection of candidates was made from a list of persons familiar with the issues involved
in GHG emissions from ecological, socio-economic, regulatory, and manufacturing perspectives.
The panel m embers who were selected h ave impressive standings in their respective fields and
comprise a balanced an d diverse point of view. The peer review panel consists of the following
individuals:

• John German, The International Council on Clean Transportation
• Paul Leiby, Oakridge National Laboratory
• Jonathan Rubin, University of Maine, School of Economics

Their resum es are presented in Appendix A. Fo llowing selection and determ ination of their
availability, consultancy agreements had to be put in place with each individual.
SwRI Project 03.14658.01 FR

-------
Ms. Almethyist A. Chambers
Environmental Protection Agency
June 9, 2009
Page 5 of 8
As soon as it was practical, the docum ents and software were distributed to the review
panel members. The distribution was accomplished by preparing a charge letter d escribing what
was requested in detail, and attaching the supporting documents. Please see Appendix B.

SwRI then arranged a teleconference between the peer reviewers and EPA technical staff
as a kick-off m eeting. In that c onference, held on May 1, 2009, th e charge to the reviewers was
discussed, and the presentation of the supporting documents was reviewed. As there were seven
attachments, it was i mportant to have assura nee that everyone understood how the docum ents,
appendices, and attachments are related.

During the kick-off m eeting, an infor mal question dialog allowe d pa nel me mbers t o
interact with EPA' s work assignm ent m anager (WAM) and the technical staff who are m ost
familiar with the GHG model. It was agreed that any further questions would be subm itted in
writing to S wRI, and the question and EPA's r esponse would be distributed to all reviewers.
Only one s uch questio n was sub mitted. Fina lly, it was agreed tha t each would attem pt to
complete their work prior to the next teleconference, which was scheduled for May 28, 2009.

The review documents of two of the three reviewers had been received and forwarded to
EPA before the teleconference held on May 28th. The third followed shortly thereafter. The May
28th teleconference was nonetheless productive, w ith discussion am ong EPA technical staff and
the panel members. Those documents are presented in Appendices C, D, and E.

Because the third docu ment had not been re ceived before the teleco nference, it was
determined that another teleconf erence would be held. The idea wa s that all parties could have
additional time to read the comm ents of each of the p anel members and determine if there are
other topics or issues that need clarification or discussion. That discussion was also productive,
with discussion of whether one equation had an error or if it simply needed more explanation, for
example. In addition, investm ent costs of technology with regards to tooling and plant
conversions, capital budgets, and lead time were also discussed.
III. SUMMARY

A very brief summary of each of the three reviewers' comments is presented below.
SwRI Project 03.14658.01 FR

-------
Ms. Almethyist A. Chambers
Environmental Protection Agency
June 9, 2009
Page 6 of 8
Paul Leiby:

•   Stay focused first on clearly and rigorously   modeling the fuel-economy technology choice
    and cost-effectiveness considerations, for various GHG emission levels.
•   It is essential to be very explicit about whose behavior and objectives are being modeled.
•   Some confusing term s are includ ed in the TA  RF, m ost notably the  non-standard way in
    which VMT is discounted for the purposes of    this TARF (See equation top of page 11,
    line 1).
•   The inclusion of "IR" ("the annu al increase in  the value of CO2") in the discount factor is
    done without explanation or ju  stification.  (Is  it  meant to be the growth rate in GHG
    damages, abatement cost, or a CO2 tax?).
•   CostEff TARF would not seem   to be a cons   ideration f or vehicle manufacturers whose
    objective is to produce a new-car fleet m     eeting consum er needs and a GHG em   ission
    standard at least cost. What objective was intended with this hybrid aspect of the TARF?
•   Model Documentation:
          •   Restructure the presentation, perhaps folio  wing the pattern of a j ournal article
              (e.g., begin with stated purpose and b    ackground. Place this m   odel in the
              constellation of related models and indicate what is different and why.  Describe
              approach, data sources, sample results.)
          •   Bringing description of the "Core P rogram" and what the model does toward the
              front.
          •   Clarify and condense the model description
          •   State model objective (typically stating wh at is m aximized, minimized, or what
              final solution condition is sought)
          •   State model constraints
          •   State and discrim inate between principl e decision variables, exogenous inputs,
              parameters, and internally calculated  results
          •   State the solution algorithm and termination condition
          •   Be rigorous in use of notation.
          •   Use consistent variable names
          •   Clarify subscripts and carefully apply them
          •   Carefully state units.
•   Appropriateness and completeness of the contents of the sample input files:
          •   In all data input files, specify units
          •   The "Data Validation" capability and error report is a very useful feature
•   Fuels input file, Appendix 4
          •   This list does not yet reflect biofuels  or renewable fuels
          •   Some provision m ay be needed for th  e variable energy  and GHG content of
              gasoline
          •   Provision may also be needed for E 85, and the uncertain fraction of E85 used by
              FFVs.
SwRI Project 03.14658.01 FR

-------
Ms. Almethyist A. Chambers
Environmental Protection Agency
June 9, 2009
Page 7 of 8
          •   The net fuel economy and emissions by PHEVs
•  Calculation of compliance to attribute-based standards:
          •   On page 7, equation for the logistic-based   footprint, there appears to b e a s ign
              error in the denominator (should be l+exp((x-C)/D) not l-exp((x-C)/D)). This is
              likely a typo in the documentation alone.
          •   No discussion or provision for m   arket-based (perm it trading) standards is yet
              made. This should at least be acknowledged.
          •   One strategy for doing m  ore flexible st andards would be to sim  ply m erge the
              datasets and technology-sequence stage for all manufacturers and vehicle types in
              a trading group. However, this would not   provide information about potential
              permit prices and burdens across manufacturers.
•  Clarity, co mpleteness, and accuracy of the     model's visualization output, in which the
   technology application is displayed:
          •   It would be very helpf  ul to have som  e graphical summ aries of the input and
              output results.
          •   All output files should embed clear documentation on the inputs used.
          •   The .log file does list names of the 4 input files, which is essential.
          •   The "Visualization Output" file does not   (yet) report the input files (but the
              information could be retrieve from the XML file).

John German:

•  "Accounting m odel" has advantage of si   mplicity, avoids "overm odeling."  However,!   t
   "requires a great deal more sophistication and work by anyone using the model to prepare the
   inputs properly."
•  Modeling by redesign cycles (rather than annually) is a good idea.
•  Leadtime issues modeled far too simplistically
          •   One of the most important issues in  standard setting
          •   Inappropriate to treat all m  anufacturers the sam e, regards potential penetration
              rates and costs for adding technologies (given differing experience)
          •   No such thing as a hard cap on technology penetration rates
•  Recommends handling leadtim  e  c onstraints and tech penetrati  on with an assessm ent of
   capital costs by each manufacturer, with a capital budget  each design cycle
          •   [Q:  How would that budget be set?]
•  Short-term: use manufacturer-specific caps on max penetration rate per year
•  Use m ax pe netration caps (total)  to reflect m arket re strictions (dem and lim its) r ather than
   leadtime constraints
•  Re: rank-ordering technologies for each vehicle type: "Requiring the user to input technology
   in rank order of cost-effectiveness" has some challenges.
          •   Requires a great d  eal of analysis to cr   eate m odel inpu ts: "real analyses and
              modeling are in these input files"


SwRI Project 03.14658.01 FR

-------
Ms. Almethyist A. Chambers
Environmental Protection Agency
June 9, 2009
Page 8 of 8
          •  "It only works if the lear   ning rate is the same for    all technologies and if no
             technology changes effectiveness over time."
          •  "model must be able to handle multiple pathways" of technical progression
          •  Synergies depend on order of introduction  (?) and must be assessed according to
             different pathways
•  Very valuab le, perhap s ultim ately n ecessary for regulatory evaluation, to allow model to
   "maximize net social value"

Jonathan Rubin:

•  Be clear and explicit about "accounting stance"
          •  "costs to whom?"
•  Describe and report costs and benefits to as many as three groups (consumers, manufacturers,
   society)
          •  Account for subsidies and taxes
          •  Show component costs
          •  Allow distinct discount rates and treatment of risk for the three groups
•  Improve notation, consistent use of subscripts
•  Concerns about certain formulations in model equations
          •  1/i in FS equation
          •  Discounting factors used
•  Avoid discounting physical quantities, and mixing physical phenomena with economic costs
   and benefits
•  Need to account for environm ental and fuel econo my implications of al ternative fuels m ore
   carefully
          •  Renewable and biofuels
          •  Electricity
•  Future work:
          •  Significant enhancement: make probabilistic, reflecting uncertainty
          •  Account for hedonic (vehicle attribute)   implications of large changes in GHG
             emissions
          •  Consider implications for gas excise tax revenue
          •  User manual, describing impact and power of key assumptions
          •  Output in SI units
SwRI Project 03.14658.01 FR

-------
Ms. Almethyist A. Chambers
Environmental Protection Agency
June 9, 2009
Page 9 of 8
IV.    CLOSING

       Southwest Research In stitute has p repared this final report to W ork Assignment 4-1 to
describe the process followed in the peer re view of the EPA's Vehicle Greenhouse Gas Model.
Please contact Patrick Merr itt at 210-522-5422 (e-m  ail pmerritt@swri.org) with any questions.
Thank you for this opportunity to be of service.
Prepared by:
 Reviewed by:
Patrick M. Merritt
Senior Research Scientist
Emissions Chemistry
Emissions Research and Development
 E. Robert Fanick
 Manager
 Emissions Chemistry
 Emissions Research and Development
Approved:
Jeff J. White, Director
Emissions Research and Development
Engine, Emissions and Vehicle
   Research Division
/tyd
       Christine Brunner, EPA NVFEL
       Kent Helmer, EPA NVFEL
       Richard Rykowski, EPA NVFEL
       Sherry Twilligear, SwRI Contracts
U:\Davison\wp\PROJECTS\14658 EPA FR WA 4-1 VGHG Model Peer Review PMM.doc
This report shall not be reproduced, except in full, without the written approval of Southwest Research
Institute.  Results and discussion given in this report relate only to the test items described in this report.
SwRI Project 03.14658.01 FR

-------
Ms. Almethyist A. Chambers
Environmental Protection Agency
June 9, 2009
Page 10 of 8
                         ATTACHMENT


Appendices                                      Number of pages

  A    RESUMES OF PEER REVIEW PANEL MEMBERS	17

  B    CHARGE LETTER TO REVIEWERS	2

        ATTACHMENT 1 - EPA VEHICLE GHG EMISSION COST AND
          COMPLIANCE MODEL DESCRIPTION	45
        ATTACHMENT 2 - APPENDICES TO GHG MODEL DESCRIPTION
        ATTACHMENT 3 - MODEL REFERENCE GUIDE
        ATTACHMENT 4 - BENEFITS CALCULATIONS INSTRUCTIONS

  C    JOHN GERMAN'S REVIEW DOCUMENT	11

  D    PAUL LEIBY'S REVIEW DOCUMENT	13

  E    JONATHAN RUBIN'S REVIEW DOCUMENT	8

-------
                APPENDIX A

RESUMES OF PEER REVIEW PANEL MEMBERS

John German, The International Council on Clean Transportation
Paul Leiby, Oakridge National Laboratory
Jonathan Rubin, University of Maine, School of Economics

-------
John M. German
730 Brooks St., Ann Arbor, MI 48103
(734) 213-0537 (H) (734) 222-5962 (W)
PROFESSIONAL EMPLOYMENT
January SENIOR FELLOW, INTERNATIONAL COUNCIL ON CLEAN TRANSPORTATION
2009 to • Primary responsibility for technology innovation and U.S. policy development.
present • Managing project to track technology costs and benefits worldwide.
February MANAGER, ENVIRONMENTAL AND ENERGY ANALYSIS, PRODUCT REGULATORY
1998 to OFFICE, AMERICAN HONDA MOTOR CORPORATION
January • Provide policy and technical analyses on vehicle-related emissions and energy issues.
2009 • Liaison between Honda R&D, both in the U.S. and Japan, and external organizations,
including government agencies, environmental groups, other manufacturers,
academia, and state representatives.
Primary Honda representative on fuel economy and global warming issues, including
testifying before Congress, writing testimony, writing responses to CAFE
rulemaking, and making presentations.
October, SENIOR TECHNICAL ADVISOR, u.s. EPA OFFICE OF MOBILE SOURCES. Supervised up to 8
1986 to employees, managed development of regulations and guidance, and served as
January, technical consultant on a wide variety of issues.
1998 • Technical manager for study on Tier II emission standards for cars and light trucks.
• Designed and managed extensive research project evaluating in-use driving behavior
and its impact on emissions in support of revisions to the Federal Test Procedure.
Created and managed extensive usage of teams across organizational boundaries.
• Managed the development of a nonroad emission inventory and the issuance of the
Nonroad Engine and Vehicle Emission Study.
• Managed rulemaking for Cold Temperature Carbon Monoxide Standards.
• Worked with transporatation planners to help create and develop a computer
simulation model for vehicle emissions.
• EPA senior technical advisor on greenhouse gas and fuel economy issues, including
CAFE alternatives, in-use fuel economy factors, and advanced technology. Active
member of EPA global warming team and an inter-agency modeling team.
• Developed initial concepts for On-Board Diagnostics.
• Created and managed rulemaking assessing LOT CAFE test procedure adjustments.
• Developed policy guidance for 48" roll electric dynamometer, driver-selectable
devices, mileage accumulation fuel requirements, coastdown procedures and
dynamometer power absorption settings, and model year definition and duration.
May, 1985 TEAM LEADER, U.S. EPA OFFICE OF MOBILE SOURCES. Supervised 3 employees and
to Sept., managed manufacturer motor vehicle emissions compliance program.
1986 * Wrote guidance on numerous certification procedure issues.

-------
December,
1981
to May,
1985
ENGINEERING SUPERVISOR, CHRYSLER POWERTRAIN. Supervised 6 engineers,
supported product planning, and developed strategies to optimize vehicle fuel economy
and to ensure compliance with all fuel economy requirements.
• Planned and coordinated activities of staff.
• Chrysler's principal technical advisor on fuel economy and methods to improve CAFE
• Provided technical analyses and written responses to proposed regulations.
• Represented Chrysler on fuel economy matters with EPA and NHTSA.
• Provided CAFE projections and analyzed impacts of future product changes on
    CAFE.
• Team leader on a project with all areas of engineering to implement Shift Indicator
Lights. Independently developed computer algorithms to eliminate cost of a sensor.
November,
1976 to
December,
1981
ENGINEER, CHRYSLER POWERTRAIN.  Designed and implemented, from scratch,
Chrysler's system to comply with extensive EPA fuel economy regulations issued in
1975. Also the corporate expert on fuel economy regulations, coordinated fuel
economy testing, served as liaison with EPA, helped write responses to proposed
regulations, and worked on special projects.
AWARDS and ADVISORY COMMITTEES
2008
2006
2004
2002-2003
2002-2003
2001-2002
1995
1994
1993
1992
1991
National Research Council - COMMITTEE FOR A STUDY OF POTENTIAL ENERGY
SAVINGS AND GREENHOUSE GASS REDUCTIONS FROM TRANSPORTATION
SAE Engineering Meetings Outstanding Oral Presentation Award, FOR "IT'S A HlGH-
MPG VEHICLE ISSUE, NOT A HYBRID ISSUE" AT SAE GOVERNMENT/INDUSTRY MTG.
Barry McNutt Award for Excellence in Automotive Policy Analysis
1ST RECEIPIENT OF ANNUAL AWARD FROM THE SAE
advisory board, ADVANCED POWER TECHNOLOGY ALLIANCE, CENTER FOR
AUTOMOTIVE RESEARCH, ANN ARBOR, MI
sae industrial lectureship program, TO PROMOTE INTERACTION BETWEEN PRACTICING
ENGINEERS AND FACULTY AND STUDENTS VIA CAMPUS VISITS
SILVER MEDAL, U.S. EPA for strategies to reduce air pollution from nonroad engines
EPA SCIENCE ACHIEVEMENT AWARD in Air Quality. Only person in EPA's Office of
Mobile Sources ever to receive this award.
OUTSTANDING TECHNICAL COMMUNICATION in the 1992-93 Society for Technical
Communication of Southeastern Michigan Technical Publications Competition, for
"Nonroad Engine and Vehicle Emission Study"
BRONZE MEDAL, U.S. EPA for the "Nonroad Engine and Vehicle Emission Study"
BRONZE MEDAL, U.S. EPA for the Cold Temperature Carbon Monoxide Rulemaking

-------
LEADERSHIP TRAINING
2000 Honda Leader's Program - Center for Creative Leadership
1997 Modeling and Computer Simulation of Internal Combustion Engine—U. of Mich, course
1996-7 Excellence in Government Fellows Program—Council for Excellence in Government
1995 Diversity Workshops - University of Michigan
1993 Total Quality Management
1992 Looking Glass Workshop: Leadership in Multilevel Organizations - Creative Leadership
1991 Use of Consultative Methods - EPA Institute
1990 Work Group Leadership - Conservation Foundation
1989 Regulation Development in EPA - EPA
1988 Planning Effective Meetings - EPA
1987 Zenger-Miller Supervision program on Behavior Modeling - EPA
1985 Personnel Management for Managers and Supervisors - OPM
1984 Interaction Management - Chrysler Institute
1982 Organizational Leadership and Productivity - Mansare Corp.
1982 Leadership Effectiveness Training - Chrysler Institute
1981 Supervisory Skills Training - Chrysler Institute

PUBLICATIONS
John German, " Leadtime, Customers, and Technology: Technology Opportunities and Limits on the
Rate of Deployment". Reducing Climate Impacts in the Transportation Sector. D. Sperling and J.
Cannon, Springer Press, 2008.
D. Greene, J. German, and M. Delucchi, " Fuel Economy: The Case for Market Failure ". Reducing
Climate Impacts in the Transportation Sector. D. Sperling and J. Cannon, Springer Press, 2008.
J. German, "Reducing Vehicle Emissions Through Cap and Trade Schemes". Driving Climate Change:
Cutting Carbon from Transportation. D. Sperling and J. Cannon, Elseview & Academic Press, 2006.
Hybrid Gaseoline-Electric Vehicle Development edited by John German, SAE PT-117, 2005.
John German, "Hybrid Electric Vehicles", Encyclopedia of Energy, Elsevier & Academic Press, 2004
John German, Hybrid Powered Vehicles, SAE Technology Profile T-l 19, book published by Society
of Automotive Engineers, Warrendate, Pa., 2003.
John German, "Hybrid Vehicles Go to Market", TRNews #213, March-April 2001.
K. Aoki, K. Nakano, J. German, S. Kajiwara, H. Sato, and Y. Yamamoto, "An Integrated Motor
Assist Hybrid System - Development of the Insight, a Personal Hybrid Coupe", SAE 2000-01-2216,
2000.
John German, "VMT and Emission Implications of Growth in Light Truck Sales", Air and Waste
Management Association Emission Inventory Conference proceedings, Oct. 1997.
J. Alson, J. German, K. Gold, R. Larson, and M. Wolcott, "Transportation Energy Demand Models:
Why They Underestimate Greenhouse Gas Emissions", Climate Change Analysis Workshop
Proceedings, June 6-7, 1996.
John German, "Off-Cycle Emission and Fuel Efficiency Considerations", Asilomar conference on

-------
Transportation and Energy, 1995.
John German, "Observations Concerning Current Motor Vehicle Emissions", SAE 950812, Feb. 1995.
J. Koupal and J. German, "Real-Time Simulation of Vehicle Emissions Using VEMISS", CRC On-
Road Vehicle Emissions Workshop, April 1995.
S. Sheppard, J. Fieber, J. Cohen, and J. German, "Cold Start Motor Vehicle Emissions Model", Air
and Waste Management Association, Cincinnati, 1994.
P. Enns, J. German, and J. Markey, "EPA's Survey of In-Use Driving Patterns: Implications for
Mobile Source Emission Inventories", AWMA/CARB Specialty Conference on Emission Inventory,
Pasadena, CA, October, 1993.

EDUCATION
1980-1984   University of Michigan.  Completed 34 hours towards M.B. A.  GPA: 7.9 (A=8.0)

1970-1975   University of Michigan,  B.S., Physics (minor in Math).
            Honors: National Merit Finalist, Honors Program, Dean's List
            Activities: U. of Michigan Marching Band and Concert Band

-------
Insert file




Leiby Paul N. resume_2.0page_2008May02.pdf

-------
Insert file




Rubin%20CV.pdf

-------
       APPENDIX B




CHARGE LETTER TO REVIEWERS

-------
SOUTHWEST   RESEARCH  INSTITUTE
                                                                                        ®
6220 CULEBRA ROAD • POST OFFICE DRAWER 28510 • SAN ANTONIO, TEXAS, USA 78228-0510  • (210)684-5111 • www.swri.org
ENGINE, EMISSIONS, AND VEHICLE RESEARCH DIVISION                           HTTP://ENGINEANDVEHICLE .SWRI.ORG
FAX: (210) 522-3950                                                                 ISO 9001 Certified
                                                                             ISO 14001 Certified
Gentlemen:

       Thank you for agreeing to review EPA' s proposed vehicle emission effects model which
estimates the technology necessary for vehicle manufacturers to meet a specified greenhouse gas
(GHG) standard. The m   odel is contained in    the enclosed com   puter program and the
documentation, Description and Methodologies of the EPA Vehicle Greenhouse Gas Emissions
Cost and Compliance Model  and its appen dices. This report illu  strates the concepts an  d
methodologies behind EPA'  s proposed GHG model.   No independen t data analy sis will be
required for this review. Specifically, EPA staff are seeking your expert opinion on the concepts
and methodologies upon which the m odel relies and whether or not the model will  execute these
algorithms correctly. Toward this end, we as   k that you please review and comm    ent on the
following items:

       1)     The overall approach to the specifi    ed m odeling purpose and the particular
              methodologies chosen to achieve that purpose;
       2)     The appropriateness and completeness of the contents of  the sam pie input files.
              EPA staff are not seeking comment on the particular values of the contents of the
              input files, which are samples  only. Thes e input files are included as Appendices
              to the model description:
          a)  The elem ents of the Market input f ile, as shown in Appendix 1 of the m    odel
              description, which characterize the vehicle fleet;
          b)  The elem ents of the Technology input    file, in Appendix 2, that constrain the
              application of technology;
          c)  The definition of the standard  and econom ic conditions in the Scenario input file,
              as shown in Appendix 3;
          d)  The elements of the Fuels input file, as shown in Appendix 4, which characterize
              the fuel types, properties, and prices; and
          e)  The reference data contained in Appe ndix 5 which are currently  hard-coded into
              the model but, in the very near future, w ill be contained in a user contro lied input
              file.

       NOTE: The types of infor mation which can   be input to the m   odel point to both the
       flexibilities and constraints of the model.
                                SAN ANTONIO, TEXAS

                         HOUSTON, TEXAS • WASHINGTON, DC • ANN ARBOR, Ml

-------
       3)     The accuracy and app  ropriateness of the m  odel's conceptual algorithm  s and
             equations for technology application and calculation of compliance;
       4)     The congruence between the con    ceptual m ethodologies and the program
             execution;

       NOTE: Thi s can be verified by com   paring spreadsheet calculation  s to the outputs
       provided by EPA or by changing the input va  lues and exam ining the results with good
       engineering judgment.

       5)     Clarity, co mpleteness and accu  racy of th  e calcu  lations in the Benefits
             Calculations output file, in which costs and benefits are calculated;
       6)     Clarity, completeness, and accuracy of the model's visualization output, in which
             the technology application is displayed; and
       7)     Recommendations for any functionali  ties beyond what we   have described as
             "future work."

       In making your comments, you should disti nguish between recommendations for clearly
defined improvements that can be readily m ade based on data or literature reasonably available
to EPA and im provements that are more explor  atory or dependent on in  formation not readily
available to EPA.  Comm  ents should be suffici  ently clear and detailed to allow a thorough
understanding by EPA or other parties familiar with the  model. EPA requests that you not release
the peer review m aterials or you r comments until the Agency  makes its model and supporting
documentation public. EPA will  notify the reviewers when this occurs.

       If you have questions about what is required   in order to com plete this review or need
additional background m   aterial, please contact Patric    k Merritt at 210-522-5422 or
pmerritt@swri.org.  If you have any questions about the EPA peer review process itself, please
contact Ms.  Ruth Sche nk in EPA's Quality   Off ice,  National Vehic le and Fuel Em  issions
Laboratory by phone (734-214-4017) or through e-mail  (schenk.ruth@epa.gov).
                                    With best regards,
 Patrick                                    M. Merritt
                                    Senior Research Scientist
                                    Emissions Research and Development Dept.
Attachments (4)

-------
i ATTACHMENT 1
2
3 EPA Vehicle GHG Emission Cost and Compliance Model Description
4
5
6
7 Background and Overview
8
9 On-road vehicles are the predominant source of GHG emissions from the
10 transportation sector. Of all on-road vehicles, light-duty vehicles and light-duty trucks
11 (hereafter referred to as cars and trucks) produce the majority of the GHG emissions.
12 There are many methods for reducing GHG emissions from cars and trucks due to the
13 myriad of technology options available to improve the efficiency of vehicles. A detailed
14 analysis of the costs and benefits of various GHG emissions reduction requires a
15 specialized application that optimizes and accounts for all the promising technologies,
16 going beyond what can be accomplished with simple spreadsheet tools . Therefore,
17 EPA's Office of Transportation and Air Quality (OTAQ) has developed the Vehicle
18 Greenhouse Gas Emissions Cost and Compliance Model (hereafter referred to as the
19 "EPA model") to help facilitate the analysis of the costs and benefits of reducing GHG
20 emissions from cars and trucks..
21
22 Broadly speaking, the EPA model applies technologies with varying degrees of
23 cost and effectiveness to a defined vehicle fleet in order to meet a specified GHG
24 emission target and calculates the costs and benefits of doing so. The technologies are
25 combined into a series of vehicle "packages" over a series of model years, which defines
26 the fleet input file.
27
28 The vehicle fleet can be characterized very simplistically (one vehicle) or more
29 precisely (over a thousand vehicle models). Vehicle sales can vary over time in the
30 model. The vehicle description includes the baseline level of GHG emissions along with
31 any other attribute used in setting the target GHG emission level, such as footprint, which
32 is discussed further below.
33
34 GHG control technology packages can be applied "one at a time" or in groups or
35 bundles. The costs and effectiveness of these technologies are assumed to be the same
36 for all vehicle models falling within a vehicle type category, such as midsize cars with V6
37 engines or minivans. The model considers whether a specific vehicle model already has
38 a specific technology package or whether a technology can or cannot be applied to it.
39 The volume of a specific vehicle model's sales which can receive a technology package
40 can be limited by indicating a fraction of its baseline that already contains some
41 effectiveness and cost of each specific technology package. The volume of a given
42 vehicle type's sales which can receive a specific technology package can also be limited
43 with a market penetration "cap", if desired. The effectiveness and application limits of
44 each technology package can vary over time, if desired.
45
VGHG Model Documents Page 1 of 43

-------
1 Technology is applied to individual vehicles using a ranking process. Within a
2 vehicle type, the order of technology packages is set by the user. Across vehicles,
3 technology is applied to that vehicle with the lowest Technology Application Ranking
4 Factor (hereafter referred to as the TARF). The TARF considers the cost of the
5 technology, the value of any reduced fuel consumption considered by the vehicle
6 purchaser, and the mass of GHG emissions reduced over the life of the vehicle. Fuel
7 costs by calendar year and annual vehicles travelled per vehicle are provided by the user.
8
9 Technology is applied to vehicles until all the technologies have reached their
10 caps or until the sales-weighted GHG emission average of a given manufacturer's
11 vehicles complies with the specified GHG emission target or stringency. The GHG target
12 can be a flat standard applicable to all vehicles within a vehicle class (e.g., cars, trucks or
13 both cars and trucks). Or, the GHG target can be in the form of a linear or logistic
14 function, which varies the target as a function of vehicle footprint (vehicle track width
15 times wheelbase).
16
17 The GHG emission target can vary over time, but not on a model or calendar year
18 basis. One of the fundamental features of the EPA model is that, over a specified vehicle
19 redesign cycle, a manufacturer has the capability to redesign any or all of its vehicles.
20 The EPA model does not attempt to determine exactly which vehicle will be redesigned
21 by each manufacturer in any given model year. Instead, it focuses on the GHG emission
22 goal several model years in the future, reflecting the capability of longer term planning
23 on the part of auto manufacturers. Any need to further restrict the application of
24 technology can be affected through the caps on the application of technology to each
25 vehicle type mentioned above. Approximate costs and benefits of complying with
26 gradually decreasing GHG emission targets within the endpoints of a redesign cycle are
27 produced via linear interpolation, despite that in reality these functions may resemble step
28 functions more closely for any given vehicle.
29
30 GHG emission targets are specified in terms of CO2 equivalent emissions. They
31 can simply be CO2 emissions from the tailpipe or a combination of tailpipe CO2
32 emissions and air conditioner refrigerant emissions. In the case of the latter, the
33 descriptions of vehicles and technologies must include baseline refrigerant emissions and
34 the effectiveness of each technology in reducing these emissions.
35
36 Once technology has been added so that every manufacturer meets the specified
37 targets (or exhausts all of the available technologies), average costs per vehicle by
38 manufacturer and industry fleet are determined. Fleet-wide average GHG emissions are
39 also approximated for each calendar year and are used to estimate a wide array of societal
40 costs and benefits associated with the GHG emission control.
41
42 The model outputs the costs and the benefits of the control scenario. The primary
43 cost of GHG emission control is the cost of the added technology as compared to the
44 baseline. The primary benefit is the value of reduced fuel consumption and can be
45 sensitive to the user assumed price of fuel. However, the value of a number of other
46 costs and benefits are also evaluated, as listed below:
VGHG Model Documents Page 2 of 43

-------
 1
 2       1) The reduction or co-benefits in VOC, CO, NOx, sulfur dioxide (SOx), and PM
 3          emissions associated with reduced fuel production are estimated, as well as their
 4          value to society,
 5       2) Societal benefits associated with reduced crude oil use which are not reflected in
 6          the price of crude oil,
 7       3) The value of reduced time necessary to refuel vehicles, and
 8       4) The value of GHG emission reductions.
 9
10          GHG emission control tends to encourage technologies that improve vehicle fuel
11    efficiency and reduces the cost of driving, which in turn can result in more driving.  This
12    feedback effect is commonly referred to as the "rebound effect", the extent of which can
13    be specified by the user.  Estimates are made of a number of potential costs and benefits
14    associated with driving, as follows:
15
16       1) The increase in vehicular VOC, CO, NOx, SOx, and PM emissions,
17       2) The increased vehicular noise, congestion, and accidents,
18       3) The value of the increased driving,  and
19       4) The cost of fuel required by the increased driving.
20
21          Where the EPA model differs from other similar models is that it adds technology
22    to vehicles on a redesign cycle basis.  Some models try to predict the model year when
23    each vehicle model (of hundreds) will be redesigned and then adds technologies to each
24    vehicle on its redesign year. In reality, the timing of vehicle redesign is difficult to
25    predict, because it is constantly adjusting to changes in the market, corporate needs, or
26    government regulations, which these other models are not able to capture. In addition,
27    since there are hundreds of specific vehicle models being sold in any given model year,
28    the technology application process becomes computationally intensive. This
29    methodology creates a situation where the model attempts to achieve a level of precision
30    greater than its accuracy.
31
32          The EPA model avoids these two issues by taking a mid to long term approach to
33    vehicle redesign, assuming a 5+ year planning horizon.. In this methodology, the user
34    designates a redesign cycle length (currently this is hard-coded as 5 years, but in the
35    future this will be a user input), and the model redesigns the entire fleet in the final model
36    year of the redesign cycle.  This philosophy is based on the assumption that vehicles
37    undergo redesign at a rate consistent throughout the fleet. In addition, the EPA model
38    looks much further into the future than other models, with the capability of adding
39    technology to vehicles over four redesign cycles.  Because it is next to impossible to
40    determine each vehicle model's subconfiguration (there are over 1000) 20+ years in the
41    future, the methodology of adding technologies incrementally to each vehicle model by
42    model year does not add value to the model results. Moreover, the EPA model avoids
43    "overmodeling" by allowing for the simplification of the fleet by dozens (or fewer)
44    representative vehicles, rather than  the hundreds that are currently forecasted by
45    manufacturers.
46
                                VGHG Model Documents Page 3 of 43

-------
 1
 2   Input Files
 O
 4          The EPA model is designed to be flexible in almost every respect.  Very little data
 5   is hard-coded in the model; since the model relies heavily on its input files, the user can
 6   alter them at will to create new vehicle models, types, and technologies, or to change the
 7   model's operating parameters for sensitivity analysis or "what-if' scenarios.  For example
 8   the following can all be modified by the user: vehicle descriptions, the technologies
 9   which are available to be applied to each vehicle, as well as their costs and effectiveness.
10
11   Vehicle Fleet Characterization
12
13          The sample input file, "Market-1", contains a list of the vehicle models that
14   describe the vehicle fleet and five different categories of data that describe them: 1)
15   Vehicle efficiency data, such as baseline fuel economy and GHG emissions (columns N
16   and O); 2) Vehicle attribute data that the model uses in its technology ranking and
17   compliance calculations, such as footprint, weight, and refrigerant type (columns P, Q,
18   and AB respectively); 3) Vehicle attribute data that the model will eventually use when
19   calculating output statistics, such as engine displacement, horsepower, and drive type,
20   although output statistics are not calculated in the current version (columns R, S, and W);
21   4) Vehicle sales data, which is presented in the form of annual sales in the baseline and in
22   the final "redesign" year of each redesign cycle (columns E through M); 5) An indication
23   of whether a technology package exists in the reference case, and if so, to what degree  the
24   technology package effectivenesses and costs are reflected. An example of the Market
25   input file is in Appendix  1.
26
27          Reference case technology is tracked to avoid double counting technology costs
28   and GHG improvement.  Columns AD through AW represent the fraction  of the
29   technology package effectiveness for  the different vehicles that is present in the reference
30   case; for example, a value of 35% for technology package 1 means that 35% of the
31   effectiveness of technology package 1 on that vehicle type is already present in the
32   reference case. Columns AY through BQ represent what fraction of the technology
33   packages' cost is reflected in the reference case.  Likewise, a value of 75% in any of
34   these columns means that 75% of the  technology package's cost on the particular vehicle
35   has been included in the reference case. This is to prevent double counting of
36   technologies that are already in the baselines.
37
38   Technology Characterization
39
40          EPA designed the model to allow the user to add GHG reducing technologies one
41   at a time or in packages or bundles that would reasonably and likely be added by
42   manufacturers within a redesign cycle. In addition, the user can combine similar vehicle
43   models into "vehicle type" groups which are likely to receive the same list of technology
44   packages. For each vehicle type, the user must rank the technology packages in order of
45   how the EPA model  should add them  to that specific vehicle type. This approach puts
46   some onus on the user to develop a reasonable sequence of technologies.  However, the
                                VGHG Model Documents Page 4 of 43

-------
 1    model also produces information which helps the user determine when a particular
 2    technology or bundle of technologies might be "out of order". The approach also
 3    simplifies the model's calculations and enables synergistic effects among technology
 4    packages to be included to the fullest degree possible.
 5
 6          The sample input file entitled "Technology-1" defines the technology packages
 7    that are available for the model to add to the fleet.  The data in this file can be categorized
 8    in three ways: 1) Parameters the model uses to calculate CO2 improvement, such as
 9    effectiveness and market penetration cap (the latter being the cap of sales for each vehicle
10    model of a vehicle type that can receive a technology package) (columns D through T);
11    2) Data the model uses to calculate technology costs, such as the package cost and cost
12    learning coefficients (columns X through AC); 3) Properties of the technology package,
13    such as refrigerant type and fuel type (columns U through W).
14
15          Each worksheet in the file contains a user-defined list of technology packages for
16    a different vehicle type. In order to avoid the complexity of synergistic effects, the user
17    must record this list within each worksheet from  top to bottom in the order of how the
18    model should add them to the particular vehicle.  Since the user defines the technology
19    packages contained in the input file,  technologies are dynamic (user defined) and not
20    "hard-coded" within the model. Values in columns L through S represent the Average
21    Incremental Effectiveness (AIE) of the technology  package over the previous package (or
22    over the baseline in the case of technology package 1) which the user can adjust on a
23    redesign cycle basis.  For example, a value of 7.0% in the AIE column would denote a
24    7.0% tailpipe CO2 improvement beyond any CO2 improvements already realized.
25    ("Tailpipe CO2" refers to the CO2 emitted over a test cycle - that relates it to vehicle
26    efficiency.) This value of 7% would include any synergistic effects that components of
27    the technology package may have with technologies that have already been added to the
28    vehicle type.  Currently, the refrigerant effectiveness noted in column T is based on the
29    fraction reduction in direct refrigerant leakage emissions and is separate from tailpipe
30    CO2.  An example of the Technology input file is in Appendix 2.
31
32          When technology is sufficiently new, or the leadtime available prior to the end of
33    the redesign cycle is such that it is not reasonable to project that it could be applied to all
34    vehicle models that are of the same specific vehicle type, for example, all minivans, the
35    user can limit its application to minivans through the use of a market cap of less than
36    100% in columns D through K. This cap can vary by redesign cycle.  When a technology
37    package is applied to fewer than 100% of the sales  of a vehicle model due to the market
38    cap, the effectiveness of the technology group is  simply reduced proportionately to reflect
39    the total net effectiveness of applying that technology package to that vehicle's sales.
40    The EPA model does not create a new vehicle with the technology package and retain the
41    previous vehicle which did not receive the technology package, splitting sales between
42    the old and new vehicles. If subsequent technology packages can be applied to the
43    vehicle, the user should consider whether in reality the new technology would likely be
44    applied to those  vehicles which received the previous technology or those which did not,
45    or a combination of the two. The effectiveness of adding the subsequent technology may
46    depend on which vehicles are receiving it.
                               VGHG Model Documents Page 5 of 43

-------
 1
 2    Fuel and Energy Prices
 3
 4    The Fuels-1 input file contains data relevant to fuel and electricity, including energy,
 5    mass, and carbon density (columns B through D in the Fuel tab), and annual price
 6    forecasts for up to 20 years (columns E through X).  There is a small subset of fuel
 7    information not included in this file that has been hard-coded into the model, which
 8    reflects the societal cost of importing fuel, which is discussed briefly below and in the
 9    section on benefits calculations. An example of the Fuels-1 file is in Appendix 4.
10
11    Regulatory Scenarios
12
13           The Scenario input file contains all data specifying the number and types of
14    model runs. The Scenarios tab acts as  a directory for different model runs, where the user
15    can create an entry for any number of runs that the model can perform in succession. At
16    present, the model is only capable of performing one scenario at a time, but in the future,
17    it will be capable of batch processing.  In the Scenarios tab, the user must specify the
18    base year, type of compliance target (CO2 or MPG), type of compliance function
19    (universal = 1, linear attribute = 2, or logistic attribute = 3), the number of redesign
20    cycles,  and the names of the other input files that describe the vehicle fleet, technology
21    packages, and fuel properties.  At present, the model is limited to a CO2 standard, but our
22    future plans include the capability to analyze an MPG standard as well. These elements
23    are entered in the Scenario input file in columns C through K, and the user can create a
24    name for the run in column B.
25
26           As stated above, there are three options for compliance targets.  The universal
27    target option is simply a numerical designation which the manufacturers' average fleet
28    CO2 cannot exceed. In contrast, the attribute-based linear target function is described by
29    up to four coefficients and has the following piecewise linear mathematical form:
30
31
32
33
34
35
36
37                                        x:  Vehicle footprint
38
39
y = \
      ' Y < Y
      , A ^ A.mm
       B-A
               _  _  mm       max
     x   — x
      max    min
           Where A: CO2 minimum
           B: CO2 ceiling
< x < x     xm;n: Intersection of lower asymptote with slope
           xmax:  Intersection of upper asymptote with
           slope
                      VGHG Model Documents Page 6 of 43

-------
                        Example of Footprint-Based CO2 Piecewise Linear Target Function
1
2
3
4
5
6
7
 9
10
11
12
13
14
15
     E
     3
     d
     o
                                               Footprint (fr)
    The footprint-based logistic curve (shown below) is described by four coefficients and
    has the mathematical form described below.
           (B-A)
                       x-C
                   1-e
    Where A: CO2 minimum
    B:  CO2 maximum
    C:  Midpoint
    D:  A sort of "inverse slope"
     x:  Vehicle footprint
                             VGHG Model Documents Page 7 of 43

-------
                              Example of Footprint-Based CO2 Logistic Target Function
 2
 O
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
       E
       2
       d
       o
                                            Footprint (ft2)

       The Scenario input file also contains the economic parameters that the model uses
to calculate the Technology Application Ranking Factors (TARFs), which are the
optimization equations that determine the order of technology application (See the Model
Logic section below). Such economic data includes the discount rate, vehicle payback
period, cafe fine cost, the "gap" between on-road CO2 and test cycle CO2, threshold
technology cost (the cost at which manufacturers add technology to only enough vehicles
to meet the standard as opposed to adding technology to all of a model line), and the
increase in price of CO2 over time.  The user enters these economic values into columns
B through G in the Economics tab of the Scenario file. An example of the Scenario input
file is in Appendix 3.
Input data "hard-coded" into the model

       There is a set of data that has been hard-coded into the model in the development
phase, which will be moved to a new "References" input file in the very near future. This
data is illustrated in Appendix 5 and is comprised of vehicle age data, scrappage rate,
vehicle miles travelled, upstream criteria pollutant emissions from fuel/energy
production, storage, and distribution, regression coefficients for downstream criteria
                               VGHG Model Documents Page 8 of 43

-------
1 pollutant emissions, emissions damage cost (cost per ton), economic cost of importing
2 fuel, external costs of driving, and global warming potentials for various GHGs.
3 Model Operation
4
5 Because the model depends so heavily on the input files, the current graphic user
6 interface (GUI) is simple to operate. After installing the program and double clicking on
7 the VGHG icon, click on the file menu and choose "Open," and a window will appear
8 with a set of input files. Click on the sample "Scenario" and it loads the data. The
9 Scenario file indicates which of the other input files are to be used in the run, and a
10 portion of the scenario description is shown in the top box of the GUI. The Market and
11 Technology input files identified in the Scenario file are shown in the two boxes below.
12 When everything is loaded, the car icon turns green. Mouse click on the car button to run
13 the program. In less than a minute, a text file showing the sequence of technology
14 addition by manufacturer and by redesign cycle appears. This file is automatically named
15 Results-"date-time". This file can be saved by the user with a more descriptive name, if
16 desired, by simply using the standard Windows "File", "Save As" actions. There is an
17 additional saved output file entitled "Tarf = "date-time", which includes the results of the
18 TARF calculations for each vehicle type-technology package combination for each
19 redesign cycle. Go back to the file menu in the GUI, and click on "Save," which is
20 necessary to enable the "visualization" function. In the file menu, now click on
21 "visualization" to examine the results of the technology application process for each
22 vehicle model in the market data file. In the near future, the information in the
23 visualization will be generated directly in an output spreadsheet, but at present it is only
24 available in the visualization, and if desired, the user can copy and paste the results
25 manually. Economic costs and benefits can be loaded in a spreadsheet file entitled
26 "Benefits Calculations," located in the "Output" folder. Open the "Benefits
27 Calculations" excel file, and in the "Load" tab, press the "Load" button to get the
28 economic impact results.
29
30
31 Model Logic
32
33 The model's programming is organized into three main sections. At first, a pre-
34 processing section reads in the input files and performs intermediate calculations to get
35 the input data into the desired form for calculations in the second section, the "core"
36 calculations. The core model backs out the existing technology from the baseline, and
37 calculates the Technology Application Ranking Factor (the TARF) for each vehicle type-
38 technology package combination and determines the order to which technology packages
39 should be added to vehicles. The core model then adds the effectivenesses and the costs
40 of the technology addition until each manufacturer has met the standard or until all
41 technology packages have been exhausted. Finally, a post-processing section computes
42 the societal costs and benefits of the GHG emissions reduction scenario, including the
43 tons of CO2 reduced and the gallons of fuel saved and outputs these results in a
44 Microsoft© Excel© file.
45
VGHG Model Documents Page 9 of 43

-------
1 Before the model runs the optimization algorithms, a the user can validate the data
2 in the input files, and the model then reads in the data. Each redesign cycle "starts from
3 scratch" in that every vehicle will include only those technologies present in the base
4 year.
5
6 The model then executes three time loops in the following order: 1) A calendar
7 year loop that interpolates the annual sales in between redesign cycles; 2) A redesign
8 cycle loop adds technology packages to vehicles until the fleet is in compliance (the
9 "core" programming); 3) A final loop calculates the costs and benefits of the CO2 or
10 MPG program and exports the economic impact results to a Microsoft Excel file (a post
1 1 processing step).
12
13 EPA provided the contractor with the equations and conceptual methodologies for
14 the model with a series of spreadsheets, and the contractor developed the program. These
15 equations and methodologies are outlined after the variable and parameter definitions,
16 below.
17
18
19 Pre-processing:
20
21
22 1) The user can validate the data in the spreadsheets, if desired.
23 2) Read in the input files
24 3) Perform calculations to convert input data into useful form for later calculations:
25
26 A) First, the model calculates the annual per-vehicle VMT from the survival
27 fraction and annual miles driven for cars and trucks provided in the Reference
28 file.
29
30 VMT = SurvivalFraction * AnnualMilesDriven
31
32 B) The model then calculates the discounted annual VMT and discounted
33 refrigerant leakage (RCO2) for year 1 through the vehicle's useful life
34 according to the following equations. The subscript, i, represents the specific
35 year of calculation. DR is the discount rate and IR is the annual increase in
36 value of CO2.
37
38 i. Discounted VMT for fuel savings calculations
39
41
42 ii. Discounted VMT for GHG calculations
43
VGHG Model Documents Page 10 of 43

-------
                                                  1+
                                                     DR-IR
 2
 3
 4
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
                               VMT
                                   D C02 ,.
                                    '   '
                                                 (l + DR-IRj
             in.  Discounted refrigerant leakage
                           RefLeakage. = LeakRatei
                                                     1 +
                                                        DR-IR
                                                    (l + DR-IRj
       C) Finally, the model calculates the approximate per-mile refrigerant leakage
          emissions in CO2 equivalents from the discounted leakage rate (calculated
          above), the Global Warming Potential (GWP) of the refrigerant provided in
          the Reference file, and the discounted VMT (calculated above).
Lifetime
(years)
2] RefLeakage, .xGWP
Lifetime
[years) )
i=l
Lifetime
(years)
z
1=1
LeaL
Lifetime
(years)
z
1=1
DR-IR
Po/jP -•'
(\ + DR- IR)'
\ HDR-IR
T/7UT - 2
xGWP
LifetimeLeakage x GWP

(I + DR- IR)'
LifetimeVMT
       D) At this point, the model will linearly interpolate the sales in between redesign
          cycles in order to estimate intermediate annual sales.  The final version will
          incorporate cost learning based on sales volume. Since the current version of
          the model does  not account for cost learning, this calculation is currently
          performed in the start of the "post processing" section of code, which
          calculates the social costs and benefits of the modeled regulation, rather than
          in the core code. Therefore, it will be discussed in the section below on post
          processing.
 Core Program:

       The core program is divided into two main parts: The first part determines the
order in which the model must add technology packages to vehicle types, and the second
part is the technology application until all manufacturers have met the standard provided
in the Scenarios input file (or the penetration caps have been met). The core code
contains the baseline technology accounting, the technology application ranking
optimization, the technology application and cost calculations, and whether the
manufacturers have met the standard.  Subscripts (t-1) and (t) indicate vehicle conditions
before and after technology package "t" addition, respectively.
                               VGHG Model Documents Page 11 of 43

-------
 1    1)  Determine the order of technology application
 2
 3       A) First, the model "backs out" any advanced technology that might have been
 4          present in the baseline. This step ensures that costs and benefits aren't double
 5          counted when the model is optimizing the order of technology package
 6          application.  This is done for both Tailpipe and Refrigerant CO2. The subscript, i,
 7          refers to the current technology package and the subscript, n is the final
 8          technology package for the vehicle type.
 9
10

                                       H C02REP x (l - AIE^ (Cap^ x 7£BM ))
11                    BackedoutCO2  = -i	-,	r	
                                              1 - AIE,.x (Capt.x TEBt)
12
13
14                                        l\C02REF x (1 - RIE^ (Cap^ x 7£BM ))
I5                      BackedoutRCO21=^	-.	r	
16                                               1 - RIEJ x (Capt x ^EB^)
17
18
19             whereas AIE is the average incremental effectiveness of the technology
20             package, RIE is the refrigerant incremental effectiveness of the technology
21             package, CAP is the market cap, and TEB is the percent of the technology
22             package's benefit that has been reflected  in the baseline.  The RIE reflects
23             either a change in refrigerant or a reduction in refrigerant leaks, as opposed to
24             an improvement in A/C efficiency resulting in a fuel efficiency gain (from a
25             smaller compressor, for example) are reflected in the AIE.   The backed-out
26             CO2 is calculated for each technology package on each vehicle.
27
28       B) Intermediate calculations for each vehicle type-technology package combination:
29
30                 i.  Calculate the new tailpipe and refrigerant CO2 after technology is
31                    added to the baseline fleet (any advanced technology in the baseline
32                    has been removed via the "backout" calculations) and thus CO2t-i and
33                    RC(92f_7 are equivalent to the BackedoutCO2t and BackedoutRCO2j
34                    from above.
35
36                              CO2t = CO2t_, x (l -  AIE)

37                        RCO2t =RCO2t ,x(l-RIE)
                                          t-i  v      ;

38
39                  The model uses this "backed out" CO2 in the equations above because it
40                  helps compare the effectiveness of the various vehicle-technology
41                  package combinations on a level playing field. If the model did not back
                              VGHG Model Documents Page 12 of 43

-------
 1                   out the effectiveness of the technology in the baseline, the model would
 2                   not be able to appropriately compare the effectiveness of the vehicle-
 3                   technology package combinations, and the technology application order
 4                   would not be as robust.
 5
 6                 ii. Calculate the fuel consumption (FC) before and after technology
 7                    addition:
                                         CA-ix
10                                FC,=-
                                                  UgC
                                              C02t
                                              _  UgC
11
12                    Whereas CDt_i and CDt represent the carbon density of the liquid fuel.
13                    The model does not use the equation,  since it does not account for a
14                    change in fuel type. The electric consumption of the vehicle before
15                    and after technology addition is a direct input from the Market and
16                    Technology file, respectively.
17
18                iii.  Calculate the fuel  savings (FS). Vehicles can have up to two separate
19                    energy sources, which are distinguished by subscripts  1 and 2.
20                    Currently, the model recognizes the subscript "1" as the liquid fuel and
21                    "2" as electricity, for plug-in hybrid vehicles.
22

23         FS =

24
25                    Since the fuel price (FP) changes from year to year, the fuel savings
26                    will be calculated  based on the average fuel price from the year of
27                    manufacturer to a  "Payback Period" (PP), specified by the user.
28
29       C) The model then calculates the Technology Application Ranking Factors (TARFs),
30          which are used to compare the effectiveness technology packages on the different
31          vehicle types, and optimize the order in which the model should add technology
32          packages to the fleet. At this point, the model calculates one TARF for every
33          technology package-vehicle type combination.
34
35              Currently, there are two TARF equations from which the user can choose:
36              "Effective  Cost" and "Cost Effectiveness - Manufacturer."
37
38                 i.  The "Effective Cost" TARF is defined as the cost of the technology
39                    (variable plus amortized fixed) minus the discounted fuel savings over
                               VGHG Model Documents Page 13 of 43

-------
 1                    a specified payback period of vehicle use minus the implicit reduction
 2                    in CAFE non-compliance fee.  Quantitatively, it is defined as follows:
 3
                                        pp r         i      1            (  1       1^1
 4              EffCost = TechCost - FS x Y \VMTn ™ ,. x -- FEE x
                                                        -
 5
 6                    Where the GAP is the difference between real-world and test cycle
 7                    CO2 and the FEE is a fee for non-compliance, both of which are inputs
 8                    to the model.  Appendix 7 contains the full equation form (without
 9                    intermediate steps).
10
11                 ii.  The Cost Effectiveness-Manufacturer incorporates the effective cost
12                    (illustrated above) but also accounts for the GHG benefit. It is defined
13                    as the cost of the technology minus the discounted fuel savings over a
14                    specified payback period minus the implicit reduction in CAFE non-
15                    compliance fee (which is the "effective cost" definition), all divided by
16                    the amount of lifetime GHG emission reduction in kg of CO2
17                    equivalent. It is represented quantitatively as the following:
18
19
20
21
22

23
ZJ
24
25                    Appendix 7 contains the full equation form (without intermediate
26                    steps).
27
28             Finally, the model determines the order in which technology packages are
29             added to vehicles. The model first compares the TARFs corresponding to
30             technology package 1 on all of the different vehicle types in the fleet and
31             chooses the combination with the lowest TARF. (The lowest effective
32             cost/most cost effective combination is represented by the smallest TARF,
33             whereas  a negative TARF indicates a negative cost.)
pp r i i f i
- - C-^) W
L,osmjjMFR .+35
z
This equation can be rewritten as:
C^n vtT^ff —
Y [iy/?rV)9 RCD7 \ + (CC)7 Cm ^\\yVA/fT 1\
/ j li\ ?-l 1 / \ ?-l ? /J D,CO2,i J
z
1 1
FCt_J
1
(l-Gop)
1
'(l-Gop)
                               VGHG Model Documents Page 14 of 43

-------
 1
 2       D) Application of technology packages and compliance calculations
 3
 4          At this point, the model has determined the optimal order of technology
 5    application and must apply the technologies'  effectivenesses and costs until each
 6    manufacturer has achieved compliance or exhausted all technology package options.
 7
 8          A) Starting with the vehicle type-technology package combination having the
 9             lowest TARF, the model adds the technology package and calculates the new
10             vehicle CO2 emissions according to the following equations:
11
12                    i.  First, the model calculates the tailpipe CO2 (TCO2); i.e., this is the
13                       CO2 the vehicle would emit on a test cycle.
14
15                           TCQ2  _TC02t_lx(l-CAPxAIE)
                                           \-AIExTEB
16
17                       Where TEB is the "Technology Effectiveness Basis," which is the
18                       percent of the technology package's effectiveness that is reflected
19                       in the baseline for the vehicle to which it is being applied.
20
21                    ii.  The model  also calculates the refrigerant CO2  (RCO2), which is
22                       based on improvements in leakage or a change in refrigerant

24                      RC02, = RCO^l-CAP
                                       \-RIExTEB             _
25
26                    iii. The model  must then calculate the cost this technology addition. It
27                       calculates average incremental cost, the average per-vehicle cost
28                       (averaged over a manufacturer's fleet), and the cumulative cost up
29                       to this point for all manufacturers.
30
31                       The average incremental cost is performed for each technology
32                       addition (but is not reported). It is the cost of the technology  added
33                       up to its market cap minus the cost of the technology present in the
34                       baseline.
35
36                       IncrementalCost = TechCost * (CAP - CEB)
37
38                       Where CEB is the "Cost Effectiveness Basis,"  which is the amount
39                       of the technology package's cost that is in the baseline for the
40                       particular vehicle.
41
42                       Next, the average vehicle cost for each manufacturer is calculated
43                       and reported.
                              VGHG Model Documents Page 15 of 43

-------
 1
                                            TechCost * ModelSales
MFR
 2
                                           _    TotalFleetSales
 3
 4                       The model then updates the cumulative costs of the program. In
 5                       future versions of the model this can also be expressed in terms of
 6                       cost per redesign cycle.
 7
                                   (Compliance                \
                                     ^ AvgVehicleCostMpx x TotalFleetSales ^^
                                     r=i                  J
 9                    whereas T is the number of the technology package.
10
1 1                    iv. The next step is to add the tailpipe and refrigerant CO2 equivalent.
12                       The result is the vehicle's total CO2 and is used in the calculation
13                       for compliance.
14
15                                CO2t =TCO2t  +RCO2t
16
17          B)  The model must then recalculate the fleet CO2 for each manufacturer that
18              produces the vehicle type that received the technology package.
19
20                             FleetCO2 =
                                           Models
                                                ^
                                                ^ Sales
                                              MFKModels
21
22              The model must calculate this value for each manufacturer and compare the
23              result with the standard that the user has indicated for that manufacturer in the
24              Scenario file input. If the manufacturer has not met the standard, the model
25              iterates back and compares the remaining TARFS for technology package 1
26              along with the TARF for technology 2 on the same vehicle type to which the
27              previous technology package was just added. The model then repeats the
28              process of choosing the next lowest TARF out of the new group, applying the
29              technology package effectiveness and cost values, calculating the new total
30              CO2, and recalculating the new fleet averages.  The model performs this loop
3 1              until each manufacturer has met the standard or until the technology packages
32              have been exhausted.
33
34    Appendix 6 contains a compilation of variable definitions used in this section.
35
36    Post-Processing - Calculation of Costs and Benefits
37
38          The model estimates discounted costs and benefits on both calendar year and
39    model year bases. We describe these calculations on a calendar year basis first, followed
40    by that for an entire model year's vehicle sales. In both cases, the discount rate is
                               VGHG Model Documents Page 16 of 43

-------
1 specified by the user, as described above. Except for technology costs, all costs and
2 benefits are estimated on a fleetwide basis. Industry fleetwide technology costs are
3 estimated, but they are also estimated on a disaggregated basis by manufacturer.
4
5 Costs and Benefits by Calendar Year
6
7 A list of the types of costs and benefits evaluated by the EPA Model was provided
8 in the Background and Overview section above. The first step in estimating these costs
9 and benefits is to develop estimates of vehicle costs, GHG emissions and fuel
10 consumption by model year. This is done by linearly interpolating vehicle sales, the cost
11 of GHG emission control per vehicle, GHG emissions per mile, and fuel consumption per
12 mile between the baseline case and the first redesign cycle and subsequently between
13 redesign cycles. For example, if the baseline year is 2010, redesign year 1 is 2015 and
14 redesign year 2 is 2020, we linearly interpolate between the 2010 and 2015 sales to
15 estimate annual sales for years 201 1-2016. Likewise, we linearly interpolate between the
16 2015 and 2020 sales to estimate annual sales for years 2016-2019. The same is done for
17 the other factors. The following equation depicts the arithmetic involved.
18
1 „ , ,0 , 0 , redesignyear2
1 9 AnnualSales = Sales , + gy
. .
redesignyear2 - redesignyearl
20
21 With an estimated cost per vehicle and vehicle sales by model year, the cost of
22 added technology is simply the product of these two values. For simplicity, we assume
23 that all the sales of a given model year's vehicles occurs in the calendar year of the same
24 value.
25
26 Most of the other costs and benefits depend on an estimation of the amount of
27 VMT occurring in each calendar year by vehicles of a certain model year vintage. As
28 reduced fuel consumption per mile can lead to increased driving, the amount of VMT in
29 the baseline and control cases can differ and increases the complexity of the calculations.
30 Thus, the next step, conceptually, is to estimate the percentage increase in VMT that
3 1 might result from reduced driving costs.
32
33 The rebound effect is defined as the ratio of the percentage change in VMT to the
34 percentage change in incremental driving cost, which is typically assumed to simply the
35 incremental cost of fuel consumed per mile. As mentioned above, the economic concept
36 is that as driving becomes cheaper, people tend to drive more. Since VMT increases with
37 a reduction in fuel consumption, the sign of the rebound effect is negative. The rebound
38 effect (REB) is an input on the "Economics" worksheet of the "Scenario.xls" file. The
39 percentage increase in VMT for a given change in fuel consumption per mile is
40 calculated as follows:

43
FleetFCold
VGHG Model Documents Page 17 of 43

-------
1 Since fuel consumption changes by model year, each model year's vehicles will reflect a
2 different change in VMT. This change in VMT is assumed to continue throughout the
3 life of the vehicle, since fuel economy is assumed to be constant throughout vehicle life.
4
5 Only new vehicles are affected by the additional technology. Thus, in any given
6 calendar year, some of the VMT will be by vehicles whose GHG emissions and fuel
7 consumption has changed and the remainder will be by vehicles which were unaffected
8 by the addition of technology. The model estimates this split by predicting each model
9 year's vehicles contribution to total VMT in each calendar year.
10
11 The model currently contains estimates of annual VMT per vehicle by vehicle
12 age, as well as the fractions of new vehicles still on the road as a function of age. These
13 estimates are taken from EPA's MOBILE6 model and are contained in cells M1:Q39 of
14 the Shared Inputs worksheet of the Benefits Calculation spreadsheet. The values for
15 these two parameters are currently hard-coded in the model. However, they will soon be
16 made a part of a References spreadsheet and modifiable by the user. For the purposes of
17 benefit calculation, the user can already change these values in the Benefits Calculation
18 spreadsheet and the changes will flow through to the calculation of fuel consumption and
19 VMT-related costs and benefits.
20
21 The total VMT by a specific model year's vehicles in specific calendar year is
22 determined by multiplying 1) new vehicle sales for that model year, 2) the fraction of
23 new vehicles remaining on the road according to the age of those vehicles in that calendar
24 year and 3) the annual VMT for that vehicle class at that age. Historic vehicle sales are
25 currently hard-coded in the model, like annual VMT and survival fractions. However,
26 they will soon be made a part of a References spreadsheet and modifiable by the user.
27 For the purposes of benefit calculation, the user can already change these values in the
28 Benefits Calculation spreadsheet and the changes will flow through to the calculation of
29 fuel consumption and VMT-related costs and benefits. Vehicle sales starting with the
30 baseline year are input to the model through the Market spreadsheet, as described above.
31 Historic sales are shown in cells V2:W39 of the Shared Inputs worksheet of the Benefits
32 Calculation spreadsheet. Vehicle sales starting in the baseline year are shown on the
33 Benefits2 (Sales) worksheet of the Benefits Calculation spreadsheet.
34
35 These VMT values for each combination of model year and calendar year are then
36 multiplied by fuel consumption per mile and emission factors and summed by calendar
37 year to determine total fuel consumption and emission levels by calendar year both
38 before and after GHG emission control. Fuel consumption per mile and CO2-equivalent
39 emissions before and after control are direct outputs of the compliance model. Pre-
40 control levels are specified in the Market spreadsheet, while post-control values are a
41 direct function of the GHG standard specified. The resultant fuel consumption and
42 emission levels are presented on the Emissions_Fuel_Consv worksheet of the Benefits
43 Calculation spreadsheet. Intermediate calculations are performed on the VMT_Lookup
44 and VMT_Rebound_Effect worksheets of the Benefits Calculation spreadsheet.
45
VGHG Model Documents Page 18 of 43

-------
1 The model also estimates the additional emissions of CO, VOC, NOx, PM and
2 SOx which result from the rebound effect. The emission factors for these pollutants are
3 currently hard-coded in the model, like annual VMT and survival fractions. However,
4 they will soon be made a part of a References spreadsheet and modifiable by the user.
5 For the purposes of benefit calculation, the user can already change these values in the
6 Benefits Calculation spreadsheet and the changes will flow through to the calculation of
7 social costs and benefits. These emissions are currently specified as a function of age
8 using either linear or quadratic equations. The coefficients for these emission factors are
9 taken from EPA's MOBILE6 emission model and are shown in cells HI :R11 of the
10 Exclusive Inputs worksheet of the Benefits Calculation spreadsheet. The resultant
11 emission levels are presented on the Emissions_Fuel_Consv worksheet of the Benefits
12 Calculation spreadsheet.
13
14 The value of the change in fuel consumption is determined by simply multiplying
15 the change in fuel consumption by calendar year by the price of fuel less taxes. Fuel
16 prices are input to the model via the Fuels spreadsheet, as described above. They are
17 shown in cells Dl :E44 of the Shared Inputs worksheet of the Benefits Calculation
18 spreadsheet. Fuel taxes are not currently input to the model, but are simply part of the
19 Benefits Calculation spreadsheet. We plan to add these taxes by fuel type to the Fuels
20 spreadsheet. Fuel taxes are assumed to be constant over time and are shown in cells
21 AE2:AF2 of the VMT Adj Costs worksheet of the Benefits Calculation spreadsheet. The
22 value of fuel savings are shown on column E of the Externaties worksheet of the Benefits
23 Calculation spreadsheet, and also carried forward to the Non-Tech Costs and All Costs
24 worksheets.
25
26 The model also estimates the value of externalities related to crude oil use. These
27 externalities include, for example, monopsony effects within the crude oil market, the
28 economic impact of periodic price shocks, and military costs related to protecting oil
29 production and supply overseas. These values are all in terms of $ per gallon. They are
30 currently hard-coded in the model, like annual VMT and survival fractions. However,
31 they will soon be made a part of a References spreadsheet and modifiable by the user.
32 For the purposes of benefit calculation, the user can already change these values in the
33 Benefits Calculation spreadsheet and the changes will flow through to the calculation of
34 costs and benefits. The value of crude oil related externalities are shown on column F of
35 the Externaties worksheet of the Benefits Calculation spreadsheet, and also carried
36 forward to the Non-Tech Costs and All Costs worksheets.
37
38 The value of the changes in CO2 and other pollutant emissions are based on
39 estimates of the value of these pollutants per ton of emission. The value of CO2 emission
40 reductions is input to the model through the Scenario spreadsheet. This value in real
41 terms can vary over time (e.g., increase at 2.4% per year in real dollars). The values for
42 the other pollutants are currently hard-coded in the model, like annual VMT and survival
43 fractions. However, they will soon be made a part of a References spreadsheet and
44 modifiable by the user. They are shown in cells A15:B23 of the Shared Inputs worksheet
45 of the Benefits Calculation spreadsheet. For the purposes of benefit calculation, the user
46 can already change these values in the Benefits Calculation spreadsheet and the changes
VGHG Model Documents Page 19 of 43

-------
1 will flow through to the calculation of costs and benefits. The value of changes in
2 vehicle emissions are shown on the DownstreamCosts ($) worksheet of the Benefits
3 Calculation spreadsheet, and also carried forward to the Non-Tech Costs and All Costs
4 worksheets.
5
6 The model also estimates the amount of emission reduction related to reduced fuel
7 production and distribution. The estimates of the emissions of each pollutant associated
8 with the production and distribution of each fuel are currently hard-coded in the model,
9 like annual VMT and survival fractions. However, they will soon be made a part of a
10 References spreadsheet and modifiable by the user. They are shown in Columns E and F
11 of the Shared Inputs worksheet of the Benefits Calculation spreadsheet. For the purposes
12 of benefit calculation, the user can already change these values in the Benefits
13 Calculation spreadsheet and the changes will flow through to the calculation of costs and
14 benefits. The value of changes in upstream emissions are shown on the UpstreamCosts
15 ($) worksheet of the Benefits Calculation spreadsheet, and also carried forward to the
16 Non-Tech Costs and All Costs worksheets.
17
18 The model estimates five additional vehicle related costs and benefits. Three are
19 related to the additional VMT resulting from reduced fuel consumption: noise, congestion
20 and accidents. The values for these factors are expressed in terms of $ per mile. Thus,
21 their total values are simply the product of these per mile values and the additional VMT
22 resulting from the reduced fuel consumption. The values of these three vehicle related
23 impacts are currently hard-coded in the model, like annual VMT and survival fractions.
24 However, they will soon be made a part of a References spreadsheet and modifiable by
25 the user. They are shown in cells Al 1 :B14 of the Exclusive Inputs worksheet of the
26 Benefits Calculation spreadsheet. For the purposes of benefit calculation, the user can
27 already change these values in the Benefits Calculation spreadsheet and the changes will
28 flow through to the calculation of costs and benefits. The value of changes in these
29 vehicle impacts on society are shown on the External VMTCosts ($) worksheet of the
30 Benefits Calculation spreadsheet, and also carried forward to the Non-Tech Costs and All
31 Costs worksheets.
32
33 The fourth vehicle related impact is related to the time required to refuel vehicles.
34 As fuel consumption per mile decreases, if fuel tank size doesn't decrease, or at least
35 does not decrease proportional to fuel consumption, the number of vehicle refuelings will
36 decrease. The parameters involved in estimating the reduction in refuelings are not
37 currently input to the model, but are simply part of the Benefits Calculation spreadsheet.
38 We plan to add these parameters to the References spreadsheet. The value of the
39 reduction in the number of refueling is estimated as the product of:
40
41 1) The reduction in fleetwide fuel consumption for a specific calendar year,
42 2) Average fuel tank size,
43 3) Average refueling volume, as a percentage of fuel tank capacity,
44 4) Ratio of the change in fuel tank size to the change in fuel consumtion,
45 5) Average time required to refuel a vehicle,
46 6) Value of time to the driver and other occupants,
VGHG Model Documents Page 20 of 43

-------
1 7) The number of total occupants in the vehicle, including the driver.
2
3 The input values and the annual benefits are in columns H and I of the
4 ExternalVMTCosts ($) worksheet of the Benefits Calculation spreadsheet, and also
5 carried forward to the Non-Tech Costs and All Costs worksheets.
6
7 The final vehicle related benefit category is the value of the increased driving
8 associated with the rebound effect. People decide to drive more because the value of
9 travelling somewhere exceeds the cost. This value consists of two components. The first
10 is the sum of the direct costs of the additional driving. This is assumed to be cost of fuel
11 consumed plus that of the additional congestion caused. The additional noise and
12 accidents are not considered to be direct costs. Noise is primarily experienced by non-
13 drivers and therefore not considered by drivers in their decision to take an additional trip.
14 While accidents tend to be proportional to mileage, especially for a given driver, we
15 assume that most of the cost of accidents is borne by insurance, where there is only a
16 weak association with mileage driven. Congestion, on the other hand, is totally
17 experienced by drivers and experienced fully each trip.
18
19 The second component of the value of the additional driving is the change in the
20 consumer surplus of the demand for VMT versus the cost of driving. We estimate this
21 change in surplus as one half the change in VMT times the change in the cost of driving,
22 which here is the reduction in fuel cost per mile. All of these terms have already been
23 estimated for other purposes within the Benefits Calculation spreadsheet. The value of
24 the additional driving is calculated and shown in column I of the ExternalVMTCosts ($)
25 worksheet of the Benefits Calculation spreadsheet, and also carried forward to the Non-
26 Tech Costs and All Costs worksheets.
27
28 The first step in loading the Benefits Calculation model is to "save" the results of
29 a model run (click on "File", then "Save" in the menu bar of the model window). This
30 saves the model results in an ".html" file, whose name is the current date/time stamp.
31 The next step is to open the Benefits Calculation spreadsheet (if it is not already opened),
32 go to the Load worksheet, and click on the Load button. From the list of files available to
33 load, choose the last one on the list. This loads the latest model results into the
34 spreadsheet and updates all the cost and benefit estimates.
VGHG Model Documents Page 21 of 43