Understanding processes and travel behaviour changes connected to electric vehicle adoption

Abstract: The use of electric vehicles (EVs) has the potential to increase the sustainability of the transport system, especially in case of unchanged or decreased car use, eco-driving and charging during hours of low electricity demand in case of an electricity generation mix with a large share of renewable energy sources. However, EVs still use much energy and EV-use does not solve problems regarding accessibility, social equity, traffic safety and has only a limited beneficial effect on liveability (e.g. through decreased noise). The focus of this study is on the process of the transition from conventional vehicle use to electric vehicle use. Economic and socio-psychological theories have been used in order to get more insight into the motivations for people to start using EVs, the process of EV-adoption and travel behaviour, as well as the interaction between car users, the electric vehicle and policy measures. The aim is to better understand the ongoing transition process to EV-use and the potential behavioural implications of this transition. This study is largely based on a two-wave survey that has been deployed among active car drivers in the metropolitan area of Stockholm, Sweden. In total, 294 respondents have participated in the first wave of the study, while 269 respondents have completed all waves. Besides this survey, one paper of this study is based on another two-wave survey that has been deployed among people renting a car on the island of Gotland, Sweden. In total, 158 respondents have participated in the first wave of this study, while 69 respondents have completed all waves. Electric vehicle adoption implies a considerable initial investment and a behavioural change because of range limitations. Therefore, the change towards electric vehicle use could be considered as a process rather than an event. Using the Transtheoretical Model of Change (Paper 1), it has been iv found that certain socio-economical, behavioural and socio-psychological determinants are correlated with being in a more advanced stage-of-change. Knowledge levels and self-efficacy for electric vehicle use are increasing from stage to stage. The level of response efficacy increases from stage to stage for the non-EV users, but is slightly lower for the group of EV-users that might have a more realistic view of the range and energy use of EVs. There is no direct effect between environmental awareness and stage-of-change, but there is an indirect effect through goal intention to decrease one’s CO2-emissions. In this study (Paper 2), it has also been found that the respondents using EVs make more trips on average and that they also use the car for a larger share of their total distance travelled. The differences in the number of trips and modal share of the car are statistically significant even after controlling for socio-economic variables, which might imply a rebound effect. The risk for a rebound effect is also explainable because of the fact that the marginal cost of EV-use is considerably lower than the marginal cost of conventional car use. Another potential reason for increased car use is the extremely good image that the EV has. It has been found that the EV is perceived to be more environmentally friendly than conventional cars, which was expected, but also more environmentally friendly than public transport modes. Policy makers in many countries provide incentives to make EVs more attractive. Using a stated choice experiment (Paper 3), the effects of several potential policy incentives on EV-adoption has been investigated in this study. Both purchase-based benefits and use-based benefits have a significant positive effect on EV-adoption. Purchase-based benefits intervene with the high investment costs of EVs, while use-based benefits intervene with the already low marginal costs of EV-use. Use-benefits, incentives that are given to EV-users in some countries, such as free parking or access to bus lanes, further decreases the marginal cost of EV-use, increasing the risk for rebound effects. On the other hand, the study shows v that use-based benefits do have a large effect on EV-adoption. Including the stage-of-change of the respondents, EV-adoption rates increase in the stated choice experiment for people in more advanced stages-of-change. However, the price-sensitivity decreases for people in more advanced stages-of-change. Also people with a high self-efficacy and response efficacy are more likely to adopt EVs. Seen from a policy perspective, it might be more efficient to provide use-benefits rather than purchase based benefits. However, there is a risk for a rebound effect because of the decrease of the already low marginal costs of EV-use. Because current electric vehicle users are a small part of the population, future EV-use has also been investigated using stated adaptation methods. Two stated adaptation experiments have been carried out: one concentrating on travel patterns and one on charging patterns. The first stated adaptation experiment (Paper 4/5) was carried out among all respondents, taking the initial travel patterns registered in a one-day travel diary as a starting point. The respondents got scenarios with a kilometre budget that was based on the travel distances during the one-day travel diary day. In case of shortage of range or perceived range limitations, different behavioural alterations have been selected, among trip cancellation, destination change and change of travel mode towards alternative travel modes were most frequently selected. Non-mandatory activities were more likely to be cancelled, as well as trips for which the public transport alternative is rather unattractive in terms of travel time and number of transfers. In case of abundant range and an electricity cost that is five times lower than the fuel cost per kilometre, a non-negligible number of additional trips has been reported, predominantly leisure trips and shopping trips. Besides, for a number of trips, a modal shift “towards the car” has been registered for a non-negligible number of trips, the majority of them being trips to work or school which are often carried out during rush vi hour. So, the existence of a rebound effect under the condition of abundant range has been confirmed. Charging behaviour has a significant effect on the sustainability of EV-use. The timing of charging events can increase peaks in electricity demand or fill the valleys of electricity demand. In this study (Paper 6), it has been investigated when people prefer to start a four-hour charging event and how temporal price differentiation influences these preferences. Based on this study, it has been found that the afternoon rush hour is by far the most preferred charging time if the price for charging events is fixed throughout the day. However, temporal price differentiation significantly affects preferred charging time. Both the existence and degree of temporal price differentiation matters: different behavioural responses were observed using two different price differentiation schemes: a high level of price differentiation causes the majority of charging events to move to night time. In the final paper of this thesis (Paper 7), it has been investigated whether electric vehicle rental affects the process of electric vehicle adoption as described in Paper 1. Using a before-after study, the long-term effects of renting an EV on the Swedish island of Gotland has been investigated. The results of this study show that EV-rental does not seem to significantly affect the stage-of-change towards EV-adoption. However, there seems to be a selection effect: the EV is more likely to be selected as a rental car if the rental guest is in a more advanced stage-of-change. Besides, the driving patterns of EV rental cars do not differ much from those of ICEV rental cars, which is an indicator of EVs being adequate for EV-rental in Gotland.