Developing Decision Support for Electric Vehicle (EV) Adoption & Charging Infrastructure Planning

Abstract

Majorly reliant on petroleum products, the transportation sector accounts for 25% of the global energy demand and 23% of carbon dioxide emissions. Among all the modes of transportation, road vehicles are the primary sources of greenhouse gas emissions and other pollutants. As urban pollution worsens and energy scarcity becomes more pronounced, nations and the automotive industry is turning to innovative solutions. In India, the number of motor vehicles have increased from 141 million in 2011 to 300 million in 2019. These numbers are expected to further increase with the country's growing gross domestic product (GDP). This growth exacerbates oil consumption and carbon dioxide emissions. Under the Paris Agreement (2015), India aims to reduce emissions by 33-35% by 2030 compared to the levels of 2005. To achieve this level of decarbonisation, the government sees transportation as a major sector and as such is focusing on transitioning to electric mobility through various policies and initiatives. However, despite all the monetary and non-monetary incentives offered to manufacturers and buyers, the acceptance and penetration rates of electric vehicles are still significantly low, especially in developing economies. In fact, the adoption patterns for most developing economies (except for China) are pretty low, which points towards a massive window of opportunity for electric vehicles to revolutionise mobility behaviour. Hence, this study aims to develop a modelling framework that can be used as a decision-support tool for the widespread adoption of electric vehicles by framing policies and planning infrastructural facilities in India and potentially other developing economies. Along those lines, this research, with a particular emphasis on the setting of a developing economy, aims to (1) study the factors affecting the electric vehicle adoption intention among young and educated Indian adults, (2) study the factors affecting the electric two-wheeler and four-wheeler adoption behaviour among potential two-wheeler and four-wheeler buyers, (3) study the preferences for public electric vehicle charging stations locations among the potential vehicle buyers, and (4) provide a modelling framework to plan an optimal network of charging infrastructure for an Indian city. Towards this end, this study first uses structural equation modelling to analyse the effects of latent factors such as environmental enthusiasm, technological enthusiasm, anxiety (or perceived risk), social image, social influence, perceived benefits, performance expectancy and facilitating conditions on the intention to adopt electric vehicles among young and educated Indian adults. The study further uses these latent variables to cluster and profile this population cohort into different consumer segments. Segmentation technique, based on k-means clustering, revealed three clusters in the respondents, which are found to show positive, neutral (slightly positive), and negative perceptions towards all clustering variables and are subsequently labelled as “innovation adoption leads”, “innovation adoption indifferents”, and “innovation adoption idlers” respectively. Further, this study expands the research domain to a wider and more diverse population termed as “potential electric two-wheeler and four-wheeler buyers”. This study first analyses the impact of latent factors such as environmental enthusiasm, social values, technological enthusiasm, perceived monetary benefits, perceived environmental benefits, lack of infrastructural readiness, perceived fee, and perceived risks affecting the adoption behaviour of potential electric two-wheelers and four-wheelers using structural equation models. Second, discrete choice models are developed to study the impact of vehicle attributes, such as purchase price, operating costs, driving range, emissions, and service-related attributes, like charging time and charging infrastructure, on electric vehicle adoption behaviour. In addition, the impact of some supporting schemes, such as reduction in registration fees, exemption from toll charges, and free parking for electric vehicles, is also studied. Third, this study estimates discrete choice models with the workplace, leisure place and highway as the location choice alternatives to investigate the electric vehicle public charging location preferences of the potential electric vehicle buyers. Mixed multinomial logit (MMNL) models and integrated choice and latent variable (ICLV) models are developed based on the attributes of the charging stations, viz. charging time, waiting time, charging cost, distance to the nearest charging station, and emissions, and the characteristics of the individuals such as age, gender, income, and daily travel distance. Fourth, this study proposes a mixed integer programming (MIP) model for the optimal placement of charging stations in the city of Bangalore by minimising the weighted distance travelled along with penalising the shortfall in demand served. The optimisation framework considers the ward-level produced (electric vehicles owned) and attracted (electric vehicles travelling into the ward) demand for deciding the optimal locations of the charging stations. The empirical models are estimated using primary data collected in two phases. In the first phase, data was collected from 660 students at an Indian technological university. In the second phase, data was collected from 1301 and 1375 potential four-wheeler and potential two-wheeler buyers, respectively. The results reveal environmental enthusiasm, social image, technological enthusiasm, monetary benefits, and environmental benefits have a significant positive impact on electric vehicle adoption intention. At the same time, perceived fees, perceived risks, and the lack of infrastructural readiness hinder electric vehicle adoption. The findings also indicate a significant influence of vehicle attributes such as purchase price, operating cost, and driving range and service attributes such as charging time and charging infrastructure density on electric vehicle adoption behaviour. Moreover, socio-demographic variables are also found to be significant determinants of electric vehicle adoption behaviour. The results also reveal the disparity in terms of the impact of the above-mentioned variables on two-wheeler and four-wheeler adoption behaviour. In terms of the charging station location station preferences, this study finds negative utility associated with higher values of charging times, waiting times, charging costs, distance to the nearest charging station, and emissions when choosing the preferred location for charging. The results also indicate that the marginal disutility related to waiting time is higher than that of charging time. The findings of this study contribute to a deeper and improved awareness of the reasons that shape consumer preferences and intentions towards electric four-wheelers and two-wheelers in India and provide some significant implications for policy and decision-makers that can help in the widespread adoption of electric vehicles.