Geothermal heat exchanger energy prediction based on time series and monitoring sensors optimization.

Abstract In recent years, the use of renewable energies has been promoted in most of developed countries due to the climate change threat. In this scenario, the importance of geothermal installations has increased. This paper focuses on a heat exchanger present on a geothermal installation. The main aim is to achieve an accurate prediction system using the previous readings of some of the sensors located along the heat exchanger. Different time series modeling techniques were applied obtaining satisfactory results in the prediction of the heat exchanger state during one year. This prediction is made 1 h, 3 h and 6 h in advance. Also, a strong correlation between the sensor readings is concluded, offering the possibility to dispense some of them. Highlights • The performance prediction of a geothermal heat exchanger can improve its efficiency. • From real measurements, the system is modeled with intelligent techniques. • The models predict the state of the installation up to 6 h in advance. • A strong correlation between different sensors is concluded. [ABSTRACT FROM AUTHOR]