A heuristic-iterative technique for optimal battery size selection towards dispatchable wind generation

Abstract

This paper aims to develop a simple yet effective technique for estimating the size of a Battery Energy Storage System (BESS) in order to make a Wind Energy System (WES) work as a dispatchable unit in unit commitment problem. The technique proposes an approximate method that can estimate the initial kWh of the battery and then upgrade the kWh size of the battery iteratively on the basis of heuristic rules that can mitigate the probabilistic forecasted error of wind power generation. An approximate method for initial size of the BESS has been proposed based on the longest continuous discharging cycle of the BESS to obtain near-optimum solution. After determining the initial BESS size, two heuristic rules are used to update the initial value in each iteration when the constraints are violated. A series of probabilistic forecasted wind power generation errors generated by the Autoregressive Dynamic Adaptive (ARDA) technique and then mapped on the Normal Distribution Curve (NDC) for each time block of a day to generate a more appropriate error or load cycle of the BESS to deal with wind speed variability. To demonstrate the justification of the proposed technique and the cost of the BESS, different load cycles have been generated using mean and 1σ of the NDC values. The proposed iterative technique was validated by comparing it to the Genetic Algorithm (GA) when applied to the BESS sizing optimization problem.