skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Comparison of Vehicle Choice Models

Abstract

Five consumer vehicle choice models that give projections of future sales shares of light-duty vehicles were compared by running each model using the same inputs, where possible, for two scenarios. The five models compared — LVCFlex, MA3T, LAVE-Trans, ParaChoice, and ADOPT — have been used in support of the Energy Efficiency and Renewable Energy (EERE) Vehicle Technologies Office in analyses of future light-duty vehicle markets under different assumptions about future vehicle technologies and market conditions. The models give projections of sales shares by powertrain technology. Projections made using common, but not identical, inputs showed qualitative agreement, with the exception of ADOPT. ADOPT estimated somewhat lower advanced vehicle shares, mostly composed of hybrid electric vehicles. Other models projected large shares of multiple advanced vehicle powertrains. Projections of models differed in significant ways, including how different technologies penetrated cars and light trucks. Since the models are constructed differently and take different inputs, not all inputs were identical, but were the same or very similar where possible. Projections by all models were in close agreement only in the first few years. Although the projections from LVCFlex, MA3T, LAVE-Trans, and ParaChoice were in qualitative agreement, there were significant differences in sales shares given bymore » the different models for individual powertrain types, particularly in later years (2030 and later). For example, projected sales shares of conventional spark-ignition vehicles in 2030 for a given scenario ranged from 35% to 74%. Reasons for such differences are discussed, recognizing that these models were not developed to give quantitatively accurate predictions of future sales shares, but to represent vehicles markets realistically and capture the connections between sales and important influences. Model features were also compared at a high level, and suggestions for further comparison of models are given to enable better understanding of how different features and algorithms used in these models may give different projections.« less

Authors:
 [1];  [2];  [3];  [4];  [4];  [5];  [3]
  1. Argonne National Lab. (ANL), Argonne, IL (United States)
  2. Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
  3. National Renewable Energy Lab. (NREL), Golden, CO (United States)
  4. Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
  5. Energetics Incorporated, Columbia, MD (United States)
Publication Date:
Research Org.:
Argonne National Lab. (ANL), Argonne, IL (United States); Sandia National Lab. (SNL-NM), Albuquerque, NM (United States); National Renewable Energy Lab. (NREL), Golden, CO (United States); Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
Sponsoring Org.:
USDOE Office of Energy Efficiency and Renewable Energy (EERE), Vehicle Technologies Office (EE-3V); USDOE National Nuclear Security Administration (NNSA)
OSTI Identifier:
1411851
Report Number(s):
SAND2017-13044R; ANL/ESD-17/19
659226
DOE Contract Number:
NA0003525; AC02-06CH11357; AC36-08GO28308; AC05-00OR22725
Resource Type:
Technical Report
Country of Publication:
United States
Language:
English
Subject:
42 ENGINEERING; 33 ADVANCED PROPULSION SYSTEMS

Citation Formats

Stephens, Thomas S., Levinson, Rebecca S., Brooker, Aaron, Liu, Changzheng, Lin, Zhenhong, Birky, Alicia, and Kontou, Eleftheria. Comparison of Vehicle Choice Models. United States: N. p., 2017. Web. doi:10.2172/1411851.
Stephens, Thomas S., Levinson, Rebecca S., Brooker, Aaron, Liu, Changzheng, Lin, Zhenhong, Birky, Alicia, & Kontou, Eleftheria. Comparison of Vehicle Choice Models. United States. doi:10.2172/1411851.
Stephens, Thomas S., Levinson, Rebecca S., Brooker, Aaron, Liu, Changzheng, Lin, Zhenhong, Birky, Alicia, and Kontou, Eleftheria. 2017. "Comparison of Vehicle Choice Models". United States. doi:10.2172/1411851. https://www.osti.gov/servlets/purl/1411851.
@article{osti_1411851,
title = {Comparison of Vehicle Choice Models},
author = {Stephens, Thomas S. and Levinson, Rebecca S. and Brooker, Aaron and Liu, Changzheng and Lin, Zhenhong and Birky, Alicia and Kontou, Eleftheria},
abstractNote = {Five consumer vehicle choice models that give projections of future sales shares of light-duty vehicles were compared by running each model using the same inputs, where possible, for two scenarios. The five models compared — LVCFlex, MA3T, LAVE-Trans, ParaChoice, and ADOPT — have been used in support of the Energy Efficiency and Renewable Energy (EERE) Vehicle Technologies Office in analyses of future light-duty vehicle markets under different assumptions about future vehicle technologies and market conditions. The models give projections of sales shares by powertrain technology. Projections made using common, but not identical, inputs showed qualitative agreement, with the exception of ADOPT. ADOPT estimated somewhat lower advanced vehicle shares, mostly composed of hybrid electric vehicles. Other models projected large shares of multiple advanced vehicle powertrains. Projections of models differed in significant ways, including how different technologies penetrated cars and light trucks. Since the models are constructed differently and take different inputs, not all inputs were identical, but were the same or very similar where possible. Projections by all models were in close agreement only in the first few years. Although the projections from LVCFlex, MA3T, LAVE-Trans, and ParaChoice were in qualitative agreement, there were significant differences in sales shares given by the different models for individual powertrain types, particularly in later years (2030 and later). For example, projected sales shares of conventional spark-ignition vehicles in 2030 for a given scenario ranged from 35% to 74%. Reasons for such differences are discussed, recognizing that these models were not developed to give quantitatively accurate predictions of future sales shares, but to represent vehicles markets realistically and capture the connections between sales and important influences. Model features were also compared at a high level, and suggestions for further comparison of models are given to enable better understanding of how different features and algorithms used in these models may give different projections.},
doi = {10.2172/1411851},
journal = {},
number = ,
volume = ,
place = {United States},
year = 2017,
month =
}

Technical Report:

Save / Share:
  • Five consumer vehicle choice models that give projections of future sales shares of light-duty vehicles were compared by running each model using the same inputs, where possible, for two scenarios. The five models compared — LVCFlex, MA3T, LAVE-Trans, ParaChoice, and ADOPT — have been used in support of the Energy Efficiency and Renewable Energy (EERE) Vehicle Technologies Office in analyses of future light-duty vehicle markets under different assumptions about future vehicle technologies and market conditions. The models give projections of sales shares by powertrain technology. Projections made using common, but not identical, inputs showed qualitative agreement, with the exception ofmore » ADOPT. ADOPT estimated somewhat lower advanced vehicle shares, mostly composed of hybrid electric vehicles. Other models projected large shares of multiple advanced vehicle powertrains. Projections of models differed in significant ways, including how different technologies penetrated cars and light trucks. Since the models are constructed differently and take different inputs, not all inputs were identical, but were the same or very similar where possible.« less
  • The estimation of emissions from motor vehicles is both complex and challenging. The U.S. Environmental Protection Agency (EPA) and the California Air Resources Board (ARB) have developed computer models to assist in these estimations. In May of this year the Air Pollution Research Advisory Committee (APRAC) of the Coordinating Research Council (CRC) retained Systems Applications International (SAI) to conduct an assessment of the methodologies used in MOBILE4 and EMFAC7 to estimate hydrocarbon (HC), carbon monoxide (CO), and oxides of nitrogen (NOx) emission factors. The assessment, as defined by the CRC, consisted of two parts: a thorough study of the structuremore » and operation of the MOBILE4 and EMFAC7 computer models, and analysis of the sensitivity of MOBILE4 and EMFAC7 to determine which model data components most strongly influence the calculated emission factors. The volume of the report describes in detail the work completed in the structural evaluation of MOBILE4 and EMFAC7E. As a starting point, a brief description of MOBILE4 and EMFAC7 is presented to acquaint the reader with the basic structure of these models. Following this, an in-depth review of the major components of both MOBILE4 and EMFAC7 is presented.« less
  • This analysis represents the biogas-to-electricity pathway under the Renewable Fuel Standard (RFS) as a point of purchase incentive and tests the impact of this incentive on EV deployment using a vehicle consumer choice model. The credit value generated under this policy was calculated in a number of scenarios based on electricity use of each power train choice on a yearly basis over the 15 year vehicle lifetime, accounting for the average electric vehicle miles travelled and vehicle efficiency, competition for biogas-derived electricity among electric vehicles (EVs), the RIN equivalence value and the time value of money. The credit value calculationmore » in each of these scenarios is offered upfront as a point of purchase incentive for EVs using the Market Acceptance of Advanced Automotive Technologies (MA3T) vehicle choice model, which tracks sales, fleet size and energy use over time. The majority of the scenarios use a proposed RIN equivalence value, which increases the credit value as a way to explore the analysis space. Additional model runs show the relative impact of the equivalence value on EV deployment. The MA3T model output shows that a consumer incentive accelerates the deployment of EVs for all scenarios relative to the baseline (no policy) case. In the scenario modeled to represent the current biogas-to-electricity generation capacity (15 TWh/year) with a 5.24kWh/RIN equivalence value, the policy leads to an additional 1.4 million plug-in hybrid electric vehicles (PHEVs) and 3.5 million battery electric vehicles (BEVs) in 2025 beyond the no-policy case of 1.3 million PHEVs and 2.1 million BEVs when the full value of the credit is passed on to the consumer. In 2030, this increases to 2.4 million PHEVs and 7.3 million BEVs beyond the baseline. This larger impact on BEVs relative to PHEVs is due in part to the larger credit that BEVs receive in the model based on the greater percentage of electric vehicle miles traveled by BEVs relative to PHEVs. In this scenario 2025 also represents the last year in which biogas-derived electricity is able to fully supply the transportation electricity demand in the model. After 2025, the credit value declines on a per vehicle basis. At the same time a larger fraction of the credit may shift towards biogas producers in order to incent additional biogas production. The expanded 41 TWh/year biogas availability scenarios represent an increase beyond today s generation capacity and allow greater eRIN generation. With a 5.24kWh/RIN equivalence value, when all of the credit is directed towards reducing vehicle purchase prices, the 41 TWh/year biogas scenario results in 4.1 million additional PHEVs and 12.2 million additional BEVs on the road in 2030 beyond the baseline of 2.5 million PHEVs and 6.1 million BEVs. Under this expanded biogas capacity, biogas-derived electricity generation is able to fully supply electricity for a fleet of over 21 million EVs (15.6 million BEVs and 5.8 million PHEVs) on a yearly basis. In addition to assessing the full value credit scenarios described above, multiple scenarios were analyzed to determine the impact if only a fraction of the credit value was passed on to the consumer. In all of these cases, the EV deployment was scaled back as the fraction of the credit that was passed on to the consumer was reduced. These scenarios can be used to estimate the impact if the credit value is reduced in other ways as well, as demonstrated by the scenarios where the current (22.6 kWh/RIN) equivalence value was used. The EV deployment that results from an equivalence value of 22.6 kWh/RIN equivalence value is roughly equivalent to the EV deployment observed in the 25% case using the 5.24 kWh/RIN equivalence value. A higher equivalence value means that a smaller number of credits, and therefore value, is created for each kWh, and therefore the impact on EV deployment is reduced. This analysis shows several of the drivers that will impact eRIN generation and credit value, and tests the impact of an eRIN point of purchase incentive on EV deployment. This additional incentive can accelerate the deployment of EVs when it is used to reduce vehicle purchase prices. However, the ultimate impact of this policy, as modeled here, will be determined by future RIN prices, the extent to which eRIN credit value can be passed on to the consumer as a point of purchase incentive and the equivalence value.« less
  • In 2014, the EPA approved a biogas-to-electricity pathway under the Renewable Fuel Standard (RFS). However, no specific applications for this pathway have been approved to date. This analysis helps understand the impact of the pathway by representing the biogas-to-electricity pathway as a point of purchase incentive and tests the impact of this incentive on EV deployment using a vehicle consumer choice model.
  • The report assesses consumer behavior towards fuel-efficient vehicles designed to meet recently mandated federal fuel economy standards. The study involves a comprehensive evaluation of existing nationwide survey data as well as the development of a major new econometric forecasting model of household vehicle type choice. As a result, the report describes both an assessment of consumers' current reported sentiments toward fuel-efficient vehicles and insights into expected future changes in household vehicle purchase behavior in response to changes in vehicle designs and prices, demographics, and the energy environment.