Statistical Analysis and Verification of Performance for CSP Subsystems

Abstract The long term success of the concentrating solar power (CSP) industry depends on having accurate and reliable measurements and predictions of performance. Agreed-upon standards for performance characterization are still lacking. This contrasts with more mature technologies such as gas turbine combined cycle power plants, where performance and testing specifications are well established in spite of system complexities. Lack of standardization inevitably results in extra cost and risk to market participants and the spread of misinformation through hidden “apples-to-oranges” comparisons. To cite just one example, there are multiple distinct metrics for characterizing heliostat tracking accuracy, and the differences can easily go unnoticed. Here we propose rigorous and accurate statistical approaches to two key performance analysis problems for tower CSP. We highlight their advantages over previous approaches, and analyze detailed operational data. Further, we demonstrate why they are sufficiently straightforward and general enough for adoption as industry standards.