Smart Meter Data to Optimize Combined Roof-Top Solar and Battery Systems Using a Stochastic Mixed Integer Programming Model

Abstract

This paper presents the design and results of a model that uses household smart meter data, electric vehicle (EV) travel load and charging options, and multiple solar resource profiles, to make decisions on optimal combinations of photovoltaics (PV), battery energy storage systems (BESS) and EV charging strategies. The least-cost planning model is formulated as a stochastic mixed integer programming (MIP) problem that makes first stage decisions on PV/BESS investments, and recourse decisions on purchase/sell from/to the grid to minimize expected household electricity costs. The model undertakes a customer-centric optimization taking into consideration net metering policy, time-of-use grid pricing, and uncertainties around inter-annual variability of solar irradiance. The model adds to the existing literature in terms of stochastic representation of inter-annual variability of solar irradiance, together with BESS capacity optimization, and EV charging mode selection. Three case studies are presented: two for a residential house with and without EV load, and a third for a larger community facility. Results from the model for the first residential house case study are compared with commercially available software to show the impacts of an accurate load profile and different policy parameters. The stochastic feature of the model proves useful in understanding the impact of variability in solar resource profiles on PV sizing. Finally, simulations of alternative EV travel patterns and tariff policies that discourage charging during the evening peak show the efficacy of `super off-peak' pricing being introduced in the state of Maryland.

Keywords

• Optimization model
• battery storage
• solar panel sizing
• electric vehicle charging
• smart meter data
• TOU grid pricing