This thesis explores the extent to which electric cars might reduce oil demand and greenhouse gas emissions in key markets: China, France, Germany, India, Japan and the United States. To meet this objective, a dynamic model capable of simulating the market evolution of nine powertrain technologies between 2000 and 2030 is developed.
The model consists of an econometric sub-model, soft-linked with a system dynamics sub-model. The purpose of the time-series econometric sub-model is to project country-specific total car stock. To this end, six single-equation regressions based on autoregressive integrated moving average or autoregressive distributed-lag techniques are estimated. The purpose of the system dynamics sub-model is to represent feedback processes and facilitate policy analysis. The effects of six policy measures are examined: emission standards, energy taxation, electric car purchase subsidies, investment in recharging stations, investment in hydrogen refuelling infrastructure and desired car occupancy. The dynamic hypothesis of the model captures feedback loops that may stimulate the market development of electric cars. Th ... mehre six countries are interlinked to simulate technological progress concerning the electric vehicle battery. In particular, its cost, price and capacity, together with the resulting electric range of the car, are investigated. Two scenarios are constructed: under the Alternative Scenario, the market uptake of electric cars is faster due to a favourable policy package. This leads to a decline in oil demand and direct greenhouse gas emissions as well as to an increase in electricity demand from cars compared to the Reference Scenario.
The methodological linkage of econometrics and system dynamics, together with the endogenisation of the electric vehicle battery price evolution by explicitly modelling six major car markets, is the main contribution of this study. Its major limitations prompt further research on the representation of supply-side aspects (i.e. battery and vehicle manufacturers) using alternative methods such as agent-based modelling.