A probabilistic capacity planning methodology for plug-in electric vehicle charging lots with on-site energy storage systems

Bayram, I. Safak and Galloway, Stuart and Burt, Graeme (2020) A probabilistic capacity planning methodology for plug-in electric vehicle charging lots with on-site energy storage systems. Journal of Energy Storage, 32. 101730. ISSN 2352-152X

[img] Text (Safak-Bayram-etal-JES-2020-A-probabilistic-capacity-planning-methodology-for-plug-in-electric-vehicle)
Safak_Bayram_etal_JES_2020_A_probabilistic_capacity_planning_methodology_for_plug_in_electric_vehicle.pdf
Accepted Author Manuscript
Restricted to Repository staff only until 12 August 2021.
License: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 logo

Download (2MB) | Request a copy from the Strathclyde author

    Abstract

    Plug-in electric vehicles (PEV) have gained popularity to support environmental sustainability and reach net-zero emission goals. However, accommodating large numbers of PEVs is a complex problem as concurrent PEV demand significantly increases peak demand and stresses supporting network elements. In this paper, we present a large-scale PEV charging lot equipped with an on-site storage. Power drawn from the grid is utilized to meet customer demand and charge the storage unit which, in return, is employed to lower peak load and demand charges. By considering the probabilistic nature of the customer demand, the proposed architecture is modelled by a Markov-modulated Poisson Process and a matrix-geometric based algorithm is developed to solve the associated capacity planning problem. Station outage probability (defined as the probability of not serving PEV demand) is used as the main metric to size station resources. Case studies show that by accounting for the statistical variations in customer demand, the power required for the station is significantly less than the sum of chargers' rated power. In addition, on-site storage can considerably reduce the stress on the supporting grid components and lower stations' running cost.