Advanced Control Algorithms and New Challenges in Parabolic Trough Solar Plants: A Review

Abstract

The majority of solar thermal plants which produce electricity and deliver it to the grid use parabolic trough technology. As examples of large scale solar trough plants we can mention the SOLANA power station (280 MW of electricity power with 6 hour of thermal storage) or the 280 MW Mojave solar complex. As stated by the National American Academy and the European Commission, one of the main challenges is to improve the overall efficiency of the solar energy plants. To address this challenge, advanced control techniques play a decisive role. This chapter presents a review of the main control techniques applied to parabolic solar trough plants as well as modeling approaches used to describe the behavior of these kind of plants. Since modeling a large scale solar plant very precisely would involve taking into account many elements which compose the plant, the resulting model can be very demanding from the computational time point of view. In this chapter, the general equations to model the solar plant are presented as well as a simple approach used for the commercial solar trough plants of Mojave. This approach provided a very good trade off between precision and complexity. As far as control algorithms are concerned, the chapter mainly focuses on the Model predictive control algorithm and optimization. The control objective of this kind of plants is explained and a description of several control strategies applied to solar trough plants is provided. A example of a model predictive control strategy applied to the old ACUREX solar collector field at the Plataforma Solar de Almería is given. However, the majority of the control strategies existing in the literature are applied to small scale solar trough plants. When dealing with large scale solar trough plants, new problems and challenges appear which have to be addressed. Current commercial solar trough plants have a considerable size and cover a vast extension. They pose additional challenges from the control and optimization point of view. New control algorithms have to be designed to address this issue. Due to the size of these plants, there can be groups of loops affected by different radiation levels due to passing clouds. Furthermore, the efficiency of the loops can be different because a group of them is cleaned and another is not. Some preliminary results published in the literature are reviewed and possible future directions for research are given. Additionally one illustrative example is provided to show the advantage of implementing advanced control strategies using a model of a large scale solar trough plant.