A reflection on using product diffusion models in forecasting the electric vehicle market
Simon Shepherd
Institute for Transport Studies
University of Leeds
UK
Abstract
In this paper I reflect on the use of product diffusion models using a system dynamics approach in the forecasting of transition pathways around the uptake of electric vehicles. I look back at how realistic my previous work was for the UK given we now have 8 years of sales data. I find that the previous forecast was overly optimistic and after re-validating the model the forecast is for a weal uptake at best. I consider the different elements of uncertainty in the used models. There are uncertainties in terms of input parameters based on validation with limited sales data, uncertainty from assumed behavioural parameters within the product diffusion process and the choice model. Structural uncertainty in terms of assumed availability and finally structural uncertainty in terms of delay structure assumed in the typical models used to model fleet turn-over is considered. This last structure has important implications on the most optimistic transitions often in the minds of key stakeholders and policy makers. Finally I look at the use of such a model in the electric scooter market for Bandung.
Keywords: system dynamics, transition pathways, modelling, policy
Topic: EV Socio-Economical Impact