The Use of Machine Learning In Assessing Future Sustainability of Newly Developed Solar Thermal Systems

Abstract

An artificial neural network (ANN) model developed in MATLAB was used to predict the environmental performance of an innovative solar thermal system, ASTEP (Application of Solar Energy to Industrial Processes) over a 30-year period. The system was applied to the industrial processes of two end-users, Mandrekas (MAND) and Arcelor Mittal (AMTP). The ASTEP system was designed to supply thermal energy up to 400°C and consist of three main components: a novel rotary Fresnel Sundial, thermal energy storage (TES) and a control system. The actual GHG emissions of the ASTEP system and a solar thermal plant as presented in the literature were used to evaluate the ability of the ANN model to predict GHG emissions. The actual and predicted emissions were compared to assess the accuracy of the model. Validation results showed a difference of 2.13 kgCO2eq/kWh for AMTP’s ASTEP system, 2.43 kgCO2eq/kWh for MAND’s ASTEP system and 0.32 kgCO2eq/kWh for a third solar thermal plant. These findings indicate that the ANN model could be considered as an effective tool in predicting GHG emissions for solar thermal plants allowing the industry to evaluate their environmental performance and adopt measures to reduce their impact.