In-situ reflectivity monitoring of heliostats using calibration cameras

Abstract

It is well-known that the efficiency of Concentrating Solar Power (CSP) plants depends strongly on maintaining a high level of heliostat reflectivity. To inform cleaning policies, an online assessment of heliostat reflectivity is highly desirable to provide real-time updates of current mirror conditions and optimise the cleaning policy, as necessary. There is currently, however, a lack of an economically feasible technique for performing a detailed online heliostat reflectivity assessment. In this paper, a low-cost heliostat reflectivity monitoring method based on already-installed targeting calibration cameras (e.g. CCD cameras) is proposed and tested. The key idea is to compare calibration camera images to detect the differences of image power caused by losses in heliostat reflectivity. Unlike previous beam characterization systems, the accuracy and sensitivity of reflectivity estimation for individual mirrors under increasingly soiled conditions is tested, and does not require additional sensors (e.g. radiometers or flux gauges). Using data obtained from field experiments conducted at the CSIRO Energy Centre in Newcastle, Australia, a methodology for estimating the mirror reflectivity from filtered calibration camera images was developed, calibrated, and validated. The model predictions showed a maximum absolute reflectivity error of 1% based on 4 consecutive soilings, which is comparable to the variability of (as accurate as) the reflectometer measurements themselves. The results of this study demonstrate the feasibility of using simple calibration camera images for economical reflectivity monitoring by removing the need for extensive manual measurements or expensive additional equipment. Such a methodology would allow assessment of cleaning benefits in the real-time operational context (e.g. weather, electricity market prices) and enable dynamic, data-driven cleaning decisions.