Optimal operations management of hybrid energy systems through short-term atmospheric and demand forecasts

The complexity of systems for energy production through renewable energy sources (RESs) is constantly increasing considering the integration of multiple modules, e.g. different RES sources, energy storage in batteries and connection to national grid for energy trade purpose. This significant complexity could represent a threat but also an opportunity if adequately managed. Aim of this paper is to propose two different approaches to manage the hourly electricity flows between the different components of a hybrid energy system (HES) fueled by PV modules and a wind turbine, equipped with a battery storage system (BES) to satisfy the demand of a user load with the opportunity to sell and purchase the electricity to/from the national grid. The first approach is a heuristic algorithm (HA) which defines robust but constant dispatching criteria of the energy flows between the HES components considering just the current value of energy production and demand with the aim of minimizing the electricity purchased by the grid. On the contrary, the second approach is a mixed integer linear programing (MILP) model which defines the optimal value of the energy flows to maximize the net profit of the HES operations determined by the electricity sales revenues minus the energy purchase costs. The developed MILP leverages the short-term forecast of the atmospheric conditions and user demand as well it considers variable energy sale and purchase pricing in the different daily hours. Both these approaches have been tested and validated through a case study of a residential building in which multiple households live located in the suburban area of Munich (Germany). The obtained results highlight how the MILP outperforms HA considering the net profit achievable weekly due to electricity trade with the grid. In particular, the MILP improve the HA economic performance of the HES operation management of 18% on average over the different months of the year. © 2019 The Authors. Published by Elsevier Ltd.