Potential assessment of a parabolic trough solar thermal power plant considering hourly analysis: ANN-based approach

Abstract

One of the major challenges of developing and growing of parabolic trough solar thermal power plants (PTSTPPs) is enhancing the techno-economic performance. The goal of this study is to develop a unique artificial neural network (ANN) model that gives the best approach to predict the levelized cost of electricity (LCOE) of two different PTSTPPs integrated with thermal energy storage and fuel backup system; the first one is using thermic oil as primary heat transfer fluid in the solar field, while the other one is based on molten salt. By this way, the optimum designs of the two plants were determined in the LCOE analysis by using the obtained weights and biases of the best ANN topology. The techno-economic potentials of using molten salt in comparison to thermic oil of the two optimized plants were investigated considering both hourly and annual performances.The results show that it is possible to get minimum values of LCOE of 8.3 and 7.0 cent$/kWh for oil and salt configurations, respectively. Moreover, because of the difference in dispatch of thermal energy system and solar field aperture area of the plants, the annual power generation and capacity factor of salt plant are much higher than those of oil plant with a difference of around 26%.