Measuring & Mitigating Electric Vehicle Adoption Barriers

Loading...
Thumbnail Image

Date

2015-01-19

Authors

Tommy, Carpenter

Advisor

Journal Title

Journal ISSN

Volume Title

Publisher

University of Waterloo

Abstract

Transitioning our cars to run on renewable sources of energy is crucial to addressing concerns over energy security and climate change. Electric vehicles (EVs), vehicles that are fully or partially powered by batteries charged from the electrical grid, allow for such a transition. Specifically, if hydro, solar, and wind generation continues to be integrated into the global power system, we can power an EV-based transportation network cleanly and sustainably. To this end, major car manufacturers are now producing and marketing EVs. Unfortunately, at the time of this writing, drivers are slow to adopt EVs due to a number of concerns. The two greatest concerns are range anxiety—the fear of being stranded without power and the fear that necessary charging infrastructure does not exist—and the unknown return on investment of EVs over their lifetime. This thesis presents computational approaches for measuring and mitigating EV adoption barriers. Towards measuring the barriers to adoption, we build a sentiment analysis system for programmatically mining detailed perceptions towards EVs from ownership forums. In addition, we design the most comprehensive electric bike trial to date, which allows us to study several aspects of electric vehicles, including range anxiety, at a much lower cost. Towards mitigation, we develop algorithms for managing a network of gasoline vehicles to be used by EV owners when a planned trip exceeds the range of their EV. Further, we design a model for taxi companies to compute whether it is profitable to transition a fraction of their fleet to EVs. To summarize our findings, we find that sentiments towards EVs are very positive, especially regarding performance and maintenance, but there are concerns over range anxiety and the higher initial price of EVs. There is a delicate balance between these two adoption barriers. Larger batteries cost more, so alleviating range anxiety with larger batteries leads to pricier vehicles. Conversely, EVs with low range capabilities can also induce costs, because drivers and fleets that own EVs may have to often acquire (or own as an additional vehicle) a gasoline vehicle to fully meet their mobility demands. As a result, EVs are best suited for drivers and fleets that are able to make long-term return on investment calculations, and whose mobility patterns do not include many very long trips. Fleets can greatly reduce their operating costs by adopting EVs because they have the capital to make upfront investments that are profitable long-term. We show that even under conservative assumptions about revenue loss due to battery depletion, EVs are already profitable (the company saves more than enough money to recoup all initial investments) for a large taxi company in San Francisco. Similarly, EVs can be profitable for two-car families (those who already have a gasoline car) and for those who can easily acquire a gasoline vehicle when needed, hence our work on sizing networks of gasoline-vehicle pools for EV owners. Finally, we find that not only are electric bikes and EVs operationally similar, the sentiments towards the two technologies are as well. Advancements made in the battery sector, especially those that reduce costs or weight, are likely to accelerate sales in both markets. The results presented in this thesis, as well as in prior work, suggest that EVs are suitable for many drivers and will hence serve a role in our eventual transition away from fossil fuels.

Description

Keywords

electric vehicles, queueing theory, sentiment analysis, return on investment

LC Keywords

Citation