Archived

This content is available here for research, reference, and/or recordkeeping.

Abstract

Future smart grid virtual power plants (VPPs) are considered for development based on industry communication standards for electric vehicle (EV) chargers such as Open Charge Point Protocol (OCPP), IEC 15118, and IEC 61851. To support research and development of computationally intelligent controls for distributed EV batteries, a python-based API OpenDSS VPP framework is utilized with thousands of experimental smart meter profiles, the IEEE 123 node test feeder, and hundreds of national survey-based EV modules for conventional and optimal charging and vehicle-to-grid (V2G) control development to mitigate any voltage violations and reduce peak load. A methodology is proposed for model-predictive control (MPC) optimization considering power flow calculations for physical network impacts as part of the objective function for each candidate design. It is simulated in a first approach to minimize utility cost and obtain a near-constant substation load. Within the combined experimental and synthetic neighborhood, natural disaster shark-curve type charging events and typical patterns are evaluated through the newly proposed terms of “EV hourly and daily homogeneity” to find adverse effects starting at 20% and 50%, respectively with a sensitivity study showing up to 1.45MWh increase of system losses during maximum overlap. Reductions up to 26% in cost during the VPP time window are found given assumptions for CA, USA retail rate and EV owner compensation. In the second formulation, a multi-objective problem to minimize both system losses and the substation load to a near constant with 5% tolerance and 45% reduction in system losses, i.e. 135 and 40kW respectively, shows the power saving benefits of VPP operation beyond cost reductions.

Document Type

Article

Publication Date

2026

Notes/Citation Information

© 2026 The Authors. This work is licensed under a Creative Commons Attribution 4.0 License.

Digital Object Identifier (DOI)

https://doi.org/10.1109/ACCESS.2026.3660975

Funding Information

This work was supported by the National Science Foundation (NSF)under Award No. #1943035 and the NSF Graduate Research Fellowship under Grant No. #2239063. The support of the Leverhulme Trust under their visiting professorship scheme and University of Kentucky through the L. Stanley Pigman Chair in Power Endowment and the Lighthouse Beacon Foundation is also gratefully acknowledged. Any findings and conclusions expressed herein are those of the authors and do not necessarily reflect the views of the sponsors.

Download

Share

COinS