PVMPC-Clean-Power-Research-Data-Sources-and-Validation-24062020

Copyright © 2020 Clean Power Research, L.L.C. SATELLITE IRRADIANCE MODEL ACCURACY IMPROVEMENTS: ACCESS TO LATEST INPUTS AND 20-YEAR VALIDATION 2020 PV Systems Symposium Webinar June 24, 2020 Data Reduce risk on your solar project Get the most accurate, bankable solar resource data. SystemCheck® Validate PV system performance Automatically monitor and assess performance of PV systems and fleets. FleetView® Effectively integrate solar into your grid Plan for solar adoption on your distribution system with site-to- feeder-specific PV production. Forecast Forecast solar power Reliably predict production from utility- scale PV with the most accurate, solar-specific forecast. Today’s presentation ▪ Motivation ▪ Input and validation data ▪ V3.4 model performance ▪ Key results Need for consistent and real-time solar data is increasing ❖ Benchmarking performance ❖ Solar resource tuning ❖ Weather trends Why temporal consistency matters Short Validation 1998 2020 Annu al Inso lation Resource Data Dataset B Dataset A On-site data Correlation On-site data Dataset B Dataset A Correlation Tuned Resource Data Why temporal consistency matters Short Validation Complete Validation 1998 2020 Annu al Inso lation On-site data On-site data Dataset B Dataset A SolarAnywhere® Correlation Correlation Resource Data Quality and volume of data enable more accurate models ❖ New satellites ❖ Numerical weather models ❖ 20+ years of ground measurements ❖ Leveraging software techniques New satellites offer better performance Comparison of GOES-13 and GOES-16 However, maintaining consistency is critical Half Hourly RMSE of GHI for Western Validation Stations Ground measurements provide an excellent long-term reference, but different biases must be considered Detecting Calibration Drift at Ground Truth Stations A Demonstration of Satellite Irradiance Models’ Accuracy Richard Perez1, James Schlemmer1, Adam Kankiewicz2, John Dise2, Alemu Tadese2 & Thomas Hoff2 1 Atmospheric Sciences Research Center, SUNY, Albany, New York, 12203, USA 2 Clean Power Research, Napa, California, 94558, USA Today’s presentationDirectional response present in pyranometer data Today’s presentationIndirectly measured GHI shows better alignment of clear sky irradiance Today’s presentation ▪ Motivation ▪ Input and validation data ▪ V3.4 model performance ▪ Key results Long-term bias errors provide a quick view of accuracy https://www.solaranywhere.com/validation/leadership-bankability/data-validation/ Annual statistics are more important for many use cases https://www.solaranywhere.com/validation/leadership-bankability/data-validation/ SolarAnywhere v3.4 shows 18% reduction in distribution of annual errors in North America… https://www.solaranywhere.com/validation/leadership-bankability/data-validation/ … and excellent consistency https://www.solaranywhere.com/validation/leadership-bankability/data-validation/ Today’s presentation ▪ Motivation ▪ Input and validation data ▪ V3.4 model performance ▪ Key results Histogram 20-year Change in GHICLR (%) Mean: 0.2% SD: 0.5% ΔGHICLR (20-yr) AoD AoD Effect of Shutting Coal Assets M. Perez et al., Observed recent trends in the solar resource across North America: changing cloud cover, AOD, and the implications for PV yield, IEEE PVSC 2020. Histogram 20-year Change in GHI (%) Mean: 0.6% SD: 1.5% ΔGHI20-yr M. Perez et al., Observed recent trends in the solar resource across North America: changing cloud cover, AOD, and the implications for PV yield, IEEE PVSC 2020. For more information, contact: Patrick Keelin Lead Product Manager pkeelin@cleanpower.com Mark Grammatico Senior Technical Account Executive markg@cleanpower.com www.cleanpower.com Thank you