Development and Assessment of a Hierarchical Control Strategy for Electric-Thermal Systems in Household Energy Supply

The future energy infrastructure requires efficient and flexible systems to ensure reliability, security, and economic viability. Model predictive control (MPC) is a control method that enables predictive and optimized behaviour by considering energy production and consumption predictions. MPC can incorporate necessary restrictions and has been successful in industrial processes and thermal power system control. However, optimizing control requires a fast performance or slow global system dynamics to send setpoints to the system’s plants in a timely manner. Deviations between predicted and measured values can occur due to uncertainties in external conditions. These deviations can have different consequences depending on the system’s tolerance. This study focuses on the cost and technical risks of using MPC in an electric-thermal system and proposes a new hierarchical control approach that addresses both thermal and electrical components. The approach is evaluated through result parameter comparison and a laboratory experiment. This work focused on the development of a hierarchical control system for multi-energy systems (MES) that supply energy to households. MES combine electricity and heat components, such as PV systems, heat pumps, cogeneration units, batteries, and thermal storage units, allowing for flexible and efficient energy consumption. The proposed control system enables cost-effective utilization of MES through economic model predictive control (EMPC) and compensates for forecast deviations using an underlying control system. The control system is developed, analyzed, and evaluated within a new software framework, with adjustable parameters to adapt to changing external conditions. The EMPC is validated in a laboratory test environment, specifically for an MES comprising a PV system with battery storage and a combined heat and power plant with heat storage. The results demonstrate the reliable control of MES, with a deviation from the ideal controlled system of approximately 12%. Additionally, a variable variant of the combined control, which utilizes the rule-based control over an extended period, is investigated to save computation time but incurs 33% higher operating costs than the optimum. Changing the cost function to optimize for CO2 reduction or microgrid operation results in 10-12% higher operating costs than cost optimization. The laboratory test confirms the feasibility of using MPC on real components, although deviations are mainly attributed to the IT infrastructure used. To implement an optimized control system, simultaneous recording and prompt transmission of measured values to the optimizer and immediate return of calculated target values to energy components are crucial. Using an underlying faster control, as demonstrated in the combined control, could enhance accuracy and reliability. In summary, the newly developed hierarchical control system effectively addresses uncertainties in multi-energy systems (MES) by providing nearoptimal cost-based control while maintaining the system’s boundary conditions. As Germany shifts away from natural gas towards geothermal and hydrogen-based energy systems, the use of natural gas-based systems in households will diminish. However, the developed control system can be directly applied to larger hydrogen-based CHP plants, such as those used in district heating systems. Moreover, the control system can be adapted for other types of MES, as demonstrated with heat pumps. In the future, additional control approaches can be integrated into the hierarchical control framework.