Integrating the ASME Performance Test Code 19.1 Approach of Calculating Measurement Uncertainty With the Regression Based Test Methods for Reporting Photovoltaic System Performance

Abstract

Solar Photovoltaic (PV) power plants have high performance test measurement uncertainty due to instrument precision limitations and spatial variations associated with irradiance and soiling measurement. Accurate prediction of the measurement uncertainty is critical for both the Owner and the EPC contractor to appropriately manage their risk. While there are several methods for testing the performance of PV plants, regression analysis based methods, like the PVUSA Method and the PPI rating method, are widely used. However, there is limited guidance on uncertainty analysis when using these methods. Most utilities and power producers have familiarity with the ASME PTC 19.1 code for measurement uncertainty analysis and often require the guidelines of PTC 19.1 be followed for evaluating the measurement uncertainty for the performance testing of PV plants. However there is lack of published literature on using the ASME PTC 19.1 approach with regression based PV performance test methods. This paper expands on the limited guidance provided by ASME PTC 19.1 Section 8-6 for regression based analysis and presents a detailed approach of calculating measurement uncertainty for PV power plants when using regression based testing methods. The paper also presents the importance of obtaining a good regression fit to the measurement uncertainty and elaborates on methods to reduce the measurement uncertainty. The overall approach discussed in this paper was applied on performance testing of two large utility-scale PV plants.