Smoothing of renewable energy generation using Gaussian-based method with power constraints

Integration of renewable energy resources to a power system can cause power fluctuations due to their intermittent nature. One way to reduce these effects is to smooth power production using energy storage systems (ESS). A typical approach to tackle the intermittency problem is to use ESS with traditional moving average method. Although it is easy to implement, the moving average method is affected by peaks and troughs during power generations which results in bigger battery sizes. In this paper, we propose a Gaussian-based smoothing algorithm that solves the pitfalls of the moving average methods. Besides smoothing, in big solar plants and wind farms such as in La Réunion island, the grid operator asks energy providers to provide power with a minimum difference from one time to another. We add this constraint on top of the smoothing problem. Then, we determine a minimum possible size of ESS so that the smoothed output power is maintained during the day-ahead forecast period. To test our approach, we use a day-ahead forecast and real data of an industrial site located in France. Bench-marking the moving average methods, we compare performances of the proposed algorithm using two metrics. Based on our simulation results, we obtained at least 34% improvement in smoothness measure and at least 19% reduced ESS size by using the proposed algorithm.