Control-Oriented Modeling, Simulation, and Predictive Control of District Heating Networks

District heating networks (DHNs) are recognized as pivotal infrastructures to reduce the carbon footprint of the heating sector. However, their large-scale dimension and the involved nonlinear dynamics lead to significant issues regarding their modeling and optimal control. Tackling these challenges, two dynamic DHN modeling frameworks are presented: a detailed simulation model and a control-oriented one. The simulation model allows the development of lightweight dynamic DHN simulators, accurately represents the main system dynamics in reduced computational time, and is well-suited for testing online control algorithms. This model is leveraged to develop a simulation library for DHN systems in the Modelica environment, named DHN4Control, freely distributed with this article in Nigro (2023). Then, a methodology for deriving a control-oriented DHN model is presented, suitable for designing a nonlinear model predictive control (NMPC) regulator, ensuring computationally efficient and cost-effective regulation while respecting operative constraints. The developed modeling and control methods are tested in simulation considering as benchmark a Test Facility district heating plant located at Ricerca sul Sistema Energetico -RSE S.p.A. in Milan (Italy), showing promising results in terms of modelling accuracy, computational efficiency and energy savings.