In-situ heat losses measurements of parabolic trough receiver tubes based on infrared camera and artificial intelligence

Abstract

Thermal problems occurring in glass cover, getters, vacuum, expansion bellows, and HTF lead to significant thermal losses due to convection-conduction, which accelerates the absorber degradation and implies replacing immediately these tubes. The replacement of the absorbers tubes is very expensive and interrupts the normal production of concentrated solar power plants. Therefore, the maintenance of the absorber tubes is one of the most critical challenges. In this paper, the Heat Loss Out System (Heat LOS) is proposed for regular evaluation and monitoring of absorbers tubes to maintain their efficiency and reduce operation and maintenance costs through immediate intervention. The proposed system is based on the intelligent processing of photos and videos in real time taken by a robust infrared camera with exact GPS coordinates and implemented on a vehicle. The Heat LOS evaluates in real-time, precisely, and rapidly, the temperature distribution of the absorbers tubes without contacting the surface or interrupting the normal production. First, an analytical study is performed to determine the appropriate speed of the vehicle, which can reach up to and the distance between the camera and the absorber tube that meet a high accuracy of measurement and do not disturb it which should not exceed 5 m. In addition, several absorbers tubes installed in the Green Energy Park (GEP, Morocco) research center are tested and evaluated using the Heat Loss Out system. The results obtained show a promising performance of the proposed system up to an accuracy of 93% based on deep learning with the CNN method, which will be valid when compared to manual analysis. This solution was assembled on a vehicle to rapidly evaluate several tubes and goes through many loops.