Smart Battery Control for Production Facilities

Electricity costs for industrial consumer in Germany have increased significantly over the last years. At the same time, the costs for battery energy storage systems (BESS) and possibilities for behind‑the‑meter generation, in particular photovoltaics, have decreased. BESSs are therefore increasingly discussed as an option to reduce electricity costs for industrial consumers through peak shaving and increased self-consumption. Due to limited capacity, battery degradation effects as well as fluctuating load and generation the optimal control of BESS for behind-the-meter applications still represents a challenge. In summer 2019, the research project Smart Battery Control for Production Facilities (SmartB4P) was launched. The consortium of research and industry partners focusses on three main contributions in order to improve existing behind-the-meter control strategies for BESSs:

  • Detailed modelling of the BESS, including degradation effects
  • Investigation of artificial intelligence, specifically reinforcement learning, as opposed to linear programming based control approaches for operating the BESS
  • Integration of existing thermal energy storage and loads with demand side management into the control strategy under consideration of all production relevant boundary conditions

In this context, a software environment is developed and appropriate models are designed in order to train a reinforcement learning agent under consideration of the abovementioned factors. The derived control strategy will be translated into a prototype and validated in field tests. Next to the design and validation of the derived control strategy, the focus of the Institute for Electrical Energy Storage Technology lies in the development of the models for the BESS and thermal energy storage.


This project is funded by the Bavarian Research Foundation (grant number: AZ-1376-19). The full list of project partners can be found in the below press release:

The responsibility for the content of this publication lies with the author.


Project members
Collath, Nils; M.Sc. +49 (89) 289 - 26925 room 3019 kein Bild