A multi-year two-stage stochastic programming model for optimal design and operation of residential photovoltaic-battery systems

Abstract

Abstract The residential PV-battery system is becoming economically viable due to technological advances. Selecting optimum investment portfolios and operation strategies are still the challenges faced by homeowners. This study proposed an integrated stochastic sizing-operating framework for optimal design and operation of grid-connected residential photovoltaic (PV)-battery systems. The two-stage stochastic programming method is adopted to address the uncertainties from load and PV productions, where 1) the first stage optimizes the investment portfolio of PV and battery, and 2) the second stage optimizes the energy system operation cost. A multi-year financial model is used to calculate the cash flows during the analysis period. The discounted net present values of expected net benefit (NB) are used as the maximization objective function to make optimal decisions for PV-battery investment and operation. Annual battery degradation factors estimate the aging battery effect. To separately evaluate the PV/battery profitability, we are the first to define the added value/cost metrics of PV/battery. Simulations based on realistic load/PV output profiles and estimated technical-economic parameters validated the effectiveness of the proposed framework and the added value/cost metrics of PV/battery components. The impacts of critical factors, including feed-in tariffs (FiT), tariff profiles and levels, and unit costs, were investigated to provide key findings to stakeholders. This framework can guide the homeowners to cost-effectively invest in and operate the distributed residential PV-battery system to facilitate the transition to a reliable, affordable, and clean building energy system.