Modelling an automatic controller for parabolic trough solar fields under realistic weather conditions

Controlling parabolic trough solar thermal power plants during transient events, such as passing clouds, is challenging for the power plant operators. Solar field controllers try to operate the field with the suitable fluid mass flow to reduce any waste of solar irradiation under very different weather and operational conditions. However, finding the optimal operation trajectory is not possible or guaranteed due to the lack of information about many influencing parameters in the solar field including, for example, effect of spatial variation of irradiance and flow maldistribution in the field. In this paper, the in-house transient simulation tool, the Virtual Solar Field (VSF) is used to test full-sized solar power plants under realistic weather conditions. Automatic field controllers have been developed with the help of our industrial partners to act as reference for further developments in field control. Thus, the VSF can estimate the potential benefit of advanced controllers using, for instance, nowcasting systems provided by cloud cameras, as well as loop control valves to manipulate the flow control in individual loops. It is shown how the VSF can act as a platform to test different control strategies and provide assessments of the detailed performance of the controller with different control components. The implemented automatic field controllers proved to be robust for different irradiance types and, hence, can be used as basis for further analysis. In addition, the controller can be instructed to dump some of the incident solar energy through defocusing in cases when the available solar energy is higher than the energy consumption by the power block and storage. In summary, VSF allows the developers of solar field controllers to test their implementations under different weather conditions and provide accurate assessment without disturbing power plant operation and before investing in new technologies, for instance nowcasting systems.