Search

Parabolic trough power plants are complex hydraulic and thermal systems that use parabolic mirrors to heat a fluid for energy generation. The solar field is divided into several loops through which the fluid flows and is heated. A loop can be more than 600 m long [3] and consists of parabolic mirrors that focus the sun's radiation onto an absorber tube. To achieve the desired outlet temperature, the amount of fluid brought into the solar field can be controlled. The fluid is distributed in the solar field in such a way that approximately the same amount flows through each loop. In case of homogeneously distributed irradiation, e.g. under clear sky conditions, the control with the mass flow works well. Problems may occur in transient conditions such as cloud passing. In this case, shaded loops can receive too much mass flow and the fluid cools down, or loops that are not shaded by a cloud can receive insufficient mass flow and overheat in the process. As an emergency function, the parabolic mirrors are moved out of focus to prevent damage in the event of high temperatures. In order to be able to react better in such situations, a system was developed that provides a better overview of the irradiation in the solar field. A camera system with an all sky imager (ASI) takes pictures of the sky and the software creates spatially resolved irradiation maps. This information can be used in different ways. This paper presents different implementation depths for parabolic trough power plants. The implementation levels range from the ASI system as stand-alone system up to a fully integrated automatic control of the solar field. For each implementation level, the benefits of the provided information are shown and an overview on the implementation for the operator effort is provided. [ABSTRACT FROM AUTHOR]

The solar field of a parabolic trough power plant is a complex thermal and hydraulic system. It is used to bring the heat transfer fluid (HTF) up to the desired temperature. The mass flow for the entire solar field is usually controlled from the plant control room. In the ideal case, the controller will adjust the mass flow so that the desired temperature is reached and all collector in the solar field are in full focus. Under transient conditions like a cloud passing through, different temperatures of the HTF can occur in the solar field. For example, the area not covered by a cloud can become hotter whereas temperatures in the area with cloud cover drop. In case the mass flow is adapted to the average irradiance level in the field, this scenario can lead to significant defocusing of those collector rows seeing high DNI but suffering from reduced mass flow. State of the art solar field controllers can use information on current DNI derived from one or a few pyrheliometers distributed over the solar field. Information on DNI distribution or an appropriate average DNI are thus not available. Recent developments in all sky imager (ASI) systems provide such spatially resolved DNI information and open the way for new control approaches. In this study, an irradiance forecasting system is used based on all sky imagers (ASI) to provide spatially resolved DNI maps of the whole solar field. With the ASI system, the controller can access information about the spatial DNI distribution and thus has knowledge about individual clouds in the solar field. In this paper, a model based approach to improve the feed forward control of the mass flow is investigated The detailed model used by the controller represents the current state of the solar field using most recent sensor data from the real solar field. With the forecasts of the ASI system as input, several simulations are performed to determine a suitable mass flow for the upcoming minutes. Under cloudy conditions, the new system shows advantages compared to a reference controller using only the DNI information from the elapsed time step. The investigation are carried out with ideal forecast and forecasts including uncertainties. For a test set of 27 days, the new control concept is able to reduce the number of emergency defocusing events by 9% under ideal forecasts and 5.7% when considering uncertainties. The average focus state of the field is increased by about 0,25%. [ABSTRACT FROM AUTHOR]

The yield of solar thermal power plants is strongly affected by disturbances in the irradiation condition through passing clouds. This effect is even more prominent when the clouds shade the solar field only partially as the irradiance situation cannot be fully measured by few measurement points in the field. In this paper, the benefit of using nowcasting systems to produce highly resolved irradiance maps on the field during start-up is studied. A simulation tool is used to dynamically model all the loops and piping in a reference commercial solar field. Automatic solar field controllers allow the model to operate during different situations. The performance of the field under a reference state-of-the-art controller is compared against a controller using the irradiance maps from nowcasting systems. The benefit during start-up processes is found to be significantly lower than the benefit the system makes during normal solar field operation. However, there is more potential shown for more advanced controllers, like model-predictive control. [ABSTRACT FROM AUTHOR]