Deville_SAND2023-03207C-1

Sandia National Laboratories is a multimission laboratory managed and operated by National Technology 31(2):149‐160. doi:10.1002/pip.3615 Nameplate deviation of SLTE systems located at Sandia, NREL, and FSEC 6 Data Preparation Raw data from weather station 100% Initial data availability 98% Recorded snow 96% Solar Elevation 15°x90° 36% Effective Irradiance 50x1200 W/m2 35% Ambient Temperature -10x40 °C 34% Current vs Irradiance 30% around slope 32% Average amount of data remaining for any given system: 32% - 2% - 2% - 1% - 1% - 2% - 60% 7 • All models came from pvlib-python • Perez has 11 variations of coefficients • Compared against measured POA from a pyranometer POA Transposition Models ➢Isotropic ➢Perez ( using a location specific Perez model did improve the model’s accuracy - Temperature models’ RMSE improved up to 2.2 ℃ with transient assumptions on cloudy days; these models are even more important for locations with dynamic conditions - The performance models exhibited NMBE within ±2.3% but differences can be seen at varying levels of irradiance, times of day, and times of year - Model complexity does not guarantee any greater accuracy - Input module data could be more significant than the model - This analysis allowed for an apples-to-apples comparison, whereas in our blind modeling comparison efforts the outcomes were dominated by modeler skill and derate assumptions - Future work will include the creation of validation test protocols to reduce modeling errors and create a pathway towards standardized model validation Thank you! Lelia Deville lmdevil@sandia.gov Solar Energy Technologies Office Award Number 38267 Please join the PVPMC at https://pvpmc.sandia.gov/ Contribute, and help increase confidence in PV system performance