Combined Degradation and Soiling With Validation Against Independent Soiling Station Measurements

Operation and maintenance (O&M) is needed to limit losses caused by soiling and degradation in PV power plants, but accurate estimates of both mechanisms are necessary to inform decisions made by the O&M team. In the present work, we propose a novel performance metric, ${\boldsymbol{CP}}{{\boldsymbol{R}}}_{{\boldsymbol{soiling}}}$, that can be used as input to algorithms that estimate soiling and degradation, e.g., combined degradation and soiling (CODS), stochastic rate and recovery (SRR), and the year-on-year (YOY) method, from performance time series data. We show that the proposed metric is robust even for locations with much cloud-cover where a clear-sky filter would remove too much of the data for the CODS algorithm to be effective. ${\boldsymbol{CP}}{{\boldsymbol{R}}}_{{\boldsymbol{soiling}}}$ values have been calculated with daily resolution using 3-years of data acquired from a 150+ MWp utility-scale solar power plant located in a semiarid landscape with a tropical wet and dry climate (“As” in the Köppen–Geiger climate classification system). Estimates of the daily soiling ratio obtained from the SRR and CODS algorithms have been compared with independent measurements carried out with 12 soiling stations and the results are in very good agreement. The average soiling rate for the soiling stations is $ - 0.07$%/day, and the corresponding estimate is $ - 0.09$ and $ - 0.11$%/day with the SRR and CODS algorithms, respectively. With CODS, the distribution of degradation rates is centered at $ - 0.12$%/year, is less dispersed, and has fewer values outside the range reported in the literature for the “As” climate in comparison with the other state-of-the-art algorithms.