Assessment of a thermocline control-oriented model in closed-loopoperation for concentrated solar thermal generation

Industrial process heat is among the most energy-intensive and polluting activities in the industry. Concentrated solar energy can be harnessed to supply heat for these processes, thereby contributing significantly to sector decarbonisation. However, the specific heat demand of industrial processes is difficult to meet with solar energy alone because of its intermittency. A practical solution is to combine real‑time control of concentrated solar plants with energy storage systems that can buffer the variability. Model Predictive Control (MPC) has proven to be an effective strategy in this context. Accordingly, this paper proposes an MPC scheme whose primary objective is to satisfy a prescribed heat demand by modulating the mass‑flow rate of the heat transfer fluid (HTF) inside a thermocline storage tank. Existing dynamic models for concentrated solar thermal (CST) plants - and in particular for thermocline tanks - are typically too computationally intensive for real‑time control. To overcome this limitation, we develop a control‑oriented CST plant model that explicitly captures the HTF temperature dynamics in the solar collectors, the thermocline tank, and the heat exchangers. In order to ensure that the associated MPC algorithm remains computationally tractable, we apply a systematic reduction of an exisiting thermocline tank model. The reduction aims to minimise execution time while retaining acceptable predictive accuracy. After tuning both the reduced model and the MPC parameters, the resulting control‑oriented model is fast enough for real‑time implementation. The MPC algorithm successfully meets a constant heat demand of 30 kW over a six‑hour horizon, with only minor deviations: 10.59 kWh for the tuned model, versus 11.18 kWh for the original model.