Optimal Operation of a Building with Electricity-Heat Networks and Seasonal Storage
Eléa Prat, Pierre Pinson, Richard M. Lusby, Riwal Plougonven, Jordi Badosa, Philippe Drobinski
TL;DR
This work tackles the problem of optimally operating a building with interconnected electricity and heat networks and seasonal thermal storage using model predictive control (MPC) across two time scales. It develops a mixed-integer linear programming framework to coordinate short-term electricity storage, long-term heat storage, solar generation, and a heat pump, and evaluates both a full-horizon benchmark and a hybrid rolling-horizon approach with end-of-horizon targets based on historical data. Using a year of hourly data from a university campus, the study identifies a minimum prediction horizon of $H_{ m min}=36$ days to reproduce full-horizon results and shows that a 6-day horizon with a heat-storage end-target achieves a suboptimality of $4.31\%$ relative to the full-horizon solution, while a 42-day horizon with a fixed end-state yields $0.92\%$. The proposed hybrid approach offers a practical balance between computational tractability and near-optimal performance, improving feasibility and runtime compared to long-horizon fixed-end strategies. The results highlight the value of end-of-horizon targets for long-term storage in MPC and point to future work incorporating temperature dynamics and storage degradation for even more realistic planning.
Abstract
As seasonal thermal energy storage emerges as an efficient solution to reduce CO2 emissions of buildings, challenges appear related to its optimal operation. In a system including short-term electricity storage, long-term heat storage, and where electricity and heat networks are connected through a heat pump, it becomes crucial to operate the system on two time scales. Based on real data from a university building, we simulate the operation of such a system over a year, comparing different strategies based on model predictive control (MPC). The first objective of this paper is to determine the minimum prediction horizon to retrieve the results of the full-horizon operation problem with cost minimization. The second objective is to evaluate a method that combines MPC with setting targets on the heat storage level at the end of the prediction horizon, based on historical data. For a prediction horizon of 6 days, the suboptimality gap with the full-horizon results is 4.31%, compared to 11.42% when using a prediction horizon of 42 days and fixing the final level to be equal to the initial level, which is a common approach.