Rolling-horizon scheduling strategies for off-grid systems: On the optimal redispatching frequency and the effects of forecasting errors

Achieving universal electricity access is a serious challenge that developing countries are facing with. Decentralized systems are widely recognized as part of the solution, especially for remote areas where traditional approaches would be too expensive. In particular, mini-grids, which are local systems conceived to produce and distribute electricity, can provide good quality electricity enabling the development of basic industrial and commercial activities. Unfortunately, the significant business risks in that areas are hampering private companies in developing initiative in many areas of the world; hedging risks and reducing costs by achieving the optimal design and operation of mini-grids is then crucial. Advanced rolling-horizon operating strategies can lead to interesting savings in operating costs, which can improve the profitability of the project. In fact, predictive strategies advance the system's requirements and, by re-dispatching the system infra-daily, can better cope with load and RES uncertainties. In the present study, we investigate the effect of both the redispatching frequency and the forecasting errors into the operating costs, in order to help practitioners and researchers in identifying the best parameters to use in design and operation of mini-grids. A numerical case study is proposed for a mini-grid in Soroti, Uganda.