| Description |
The proliferation of electric vehicle (EV) adoption introduces a suite of opportunities and risks to power system operation that have the potential to significantly increase loading, reduce equipment lifespan, impact design of rate structure, and offer energy flexibility. Electrified transportation, and the diversity of use cases therein, remains a nascent industry that lacks a maturity of data and research to make informed decisions regarding EV infrastructure deployment and operation. EV charging applications can be classified into one of four use cases: residential, workplace, transit corridor (or public), and electrified buses. Each unique use case presents distinctive infrastructure needs and utilization patterns. Despite positive growth trends in EV adoption, utilities continue to operate distribution infrastructure according to standard operating procedures and ad-hoc processes developed in response to local changes in EV utilization without data-driven, consensus-based best practices that consider the totality of possible loading scenarios. Probabilistic methods are paramount to understanding and quantifying the uncertainty that electrified transportation may present to the power system. The uncertainty in space, time, power, and energy presented by EV charging demand creates a unique challenge to the development of utility best practices. Developing forward thinking approaches to harnessing the opportunities of electrified transportation is further challenged by the relative sparseness of real-world EV charging data. Thus, we present and propose probabilistic methods to develop a data-driven understanding of EV charger utilization. First, data-driven probabilistic models quantifying EV loading and equipment lossof- life in the residential distribution charging use-case are presented. Then, a high-level review of EV charging use cases, challenges, and screening criteria are presented as a foundation to the development of EV charging infrastructure deployment best practices. Finally, in the proposed work a hierarchical Bayesian model is proposed to learn the hourly arrival rate of EVs to public EV charging stations. |
