Targeting Optimal Design and Operation of Solar Heated Industrial Processes: a MILP formulation
The intermittence of solar irradiation displays a great challenge for industrial applications, which often require a constant heat supply. Transient simulation software, such as TRNSYS and EnergyPlus, enable modeling of such systems and the respective dynamic responses. Design and operation in these softwares is usually based on heuristic strategies. In this work, optimal design and operation of a solar heated industrial process with constant heating requirements is investigated based on Mixed Integer Linear Programming (MILP). The MILP model is constrained by surrogate functions that capture the main inefficiencies, but do not represent the transient behavior. In order to investigate the dynamic behavior and feasibility of the results from MILP, a TRNSYS model is created. In this way, the MILP model is verified and the quality of the transient model can be evaluated with respect to the potential non-dynamic optimum. The industrial heating requirements between 60 and 80°C are satisfied with the help of a novel High Concentration Photovoltaic Thermal System (HCPVT), a stratified thermal storage tank, and a back-up burner. The system is located due to high DNI availability in Sede Boqer, Israel. It is demonstrated, that the back-up burner heating requirements of the TRNSYS model even outperform the MILP (-8% fuel consumption) for the same storage size. This stems from the conservatively taken assumptions. It is concluded that the MILP formulation supplies a satisfying approximation of the system performance and, hence, may supply satisfying estimates of the system design and operation. It may also be concluded, that heuristic design and operation strategies can generate very good solutions with respect to the theoretical optimum. In this work, the system is optimized with respect to the thermal behavior and efficiency, and not the economic aspects.