Model Predictive Control of Energy Storage including Uncertain Forecasts

The intermittency of renewable energy sources, e.g. wind or solar, as well as forecast uncertainties in load, price and renewable infeed profiles call for storage solutions and appropriate control strategies. For the investigations in this paper the energy hub modeling framework is used, which takes into account multiple energy carriers, distributed generation, energy storage systems and renewable energy sources. This paper focuses on the value of energy storage devices when operating them in combination with intermittent infeed from renewable energy sources. Since forecasts never are perfect, a model predictive control (MPC) strategy is used for keeping the consequences of forecast uncertainties at acceptable levels. Instead of compensating forecast uncertainties with fast acting backup generators or balancing energy, the imbalances can be compensated by means of storage devices, placed close to locations where uncertainties arise. Monte-Carlo simulations show that major part of the disturbances can be compensated by means of storage devices without increasing system operation costs significantly. In cases where storage devices are not sufficient, generator settings are changed from day-ahead schedule, representing additional balancing costs.