Time-based economic hierarchical model predictive control of all-electric energy systems in non-residential buildings

Battery energy storage systems (BESSs) in buildings provide flexibility and can lead to reduced operating costs. In this context, different sources of revenue exist. These require the consideration of different timescales making hierarchical model predictive control (HMPC) promising. Here, typical prediction horizons of the upper layer are 1 month, although peak charges are usually based on annual loads. Additionally, the benefit of HMPC compared to 1-layer MPC and rule-based control (RBC) approaches is seldom quantified. We therefore present an HMPC approach with consideration of a BESS's annual performance and benchmark it to a monthly HMPC, a 1-layer MPC and a RBC approach. The results indicate that the annual HMPC outperforms the others when considering operating costs, computational effort and battery residual lifetime simultaneously.