Optimization of residential off-grid PV-battery systems

Abstract

Solar irradiance is abundant, eco-friendly, sustainable, and one of the most promising energy sources to address the shortage of energy in the future. In isolated areas where access to the grid is limited or restricted, a stand-alone photovoltaic (PV) system is particularly effective. In such a system, the available electrical energy mainly depends on the solar irradiance and the capacities of the PV array and battery. In this paper, we develop an optimization method for designing a residential off-grid PV system. The method determines the number of PV panels and battery modules for cost-effectively operating the system. We use a mixed-integer programming model to pre-schedule the daily usage of appliances according to a forecasted solar irradiance. The schedule is then executed on a Monte Carlo simulation that considers the uncertainty of the solar irradiance. An integer Nelder-Mead (N-M) algorithm determines the size of the PV-Battery system. The performance of the method is measured using plane of irradiance data at two locations in the USA. We perform a sensitivity analysis by changing the cost of the battery and the penalty cost of non-served energy. In addition, we investigate the effects of scheduling, forecast solar irradiance variability, and battery degradation.