Periodicity analysis of charging behavior of electric car drivers: latent class hazard models

The importance of public charging infrastructure has been recognised recently as it represents a key factor in the promotion of plug-in electric vehicles (PEV). Given that a large initial investment from the public sector is essential for its widespread adoption, many studies have focused on the issue how the charging infrastructure should be planned and located. Although it is of critical importance to understand PEV users' charging behavior for improving public charging infrastructure, relatively few number of PEVs makes it difficult to analyze charging behavior of large sample of PEV users. This study uses a Dutch four-year comprehensive charging transaction dataset. Given that electric vehicle users likely exhibit heterogeneous charging behavior in terms of both charging frequency (time interval between successive charging events) and charging regularity (regular charging versus ad-hoc charging), this study applies a heterogeneous hazard-based duration model to examine how often PEV users charge their car at public charging stations and to distinguish the difference characteristics of random (erratic) users and regular (routine) users. The results show that 67 per cent of PEV users charge their car at irregular intervals, while 33 per cent charge regularly. It is also shown that regular users charge their PEV more often and are more likely to have a battery electric vehicle (BEV) compared to random users. Moreover, clear differences exist between the two groups in terms of vehicle and charging characteristics.