Photovoltaic Shading Testbed for Module-Level Power Electronics: 2016 Performance Data Update

Deline, Chris¹  Meydbray, Jenya²  Donovan, Matt² 

[1] National Renewable Energy Lab. (NREL), Golden, CO (United States);PV Evolution Labs (PVEL), Davis, CA (United States)

关键词: photovoltaic systems;    PV;    partial shading;    distributed electronics;    DMPPT;    MLPE;   

DOI  :  10.2172/1324394
RP-ID  :  NREL/TP--5J00-62471
PID  :  OSTI ID: 1324394

学科分类：再生能源与代替技术

美国|英语

来源: SciTech Connect

【 摘 要 】

The 2012 NREL report 'Photovoltaic Shading Testbed for Module-Level Power Electronics' provides a standard methodology for estimating the performance benefit of distributed power electronics under partial shading conditions. Since the release of the report, experiments have been conducted for a number of products and for different system configurations. Drawing from these experiences, updates to the test and analysis methods are recommended. Proposed changes in data processing have the benefit of reducing the sensitivity to measurement errors and weather variability, as well as bringing the updated performance score in line with measured and simulated values of the shade recovery benefit of distributed PV power electronics. Also, due to the emergence of new technologies including sub-module embedded power electronics, the shading method has been extended to include power electronics that operate at a finer granularity than the module level. An update to the method is proposed to account for these emerging technologies that respond to shading differently than module-level devices. The partial shading test remains a repeatable test procedure that attempts to simulate shading situations as would be experienced by typical residential or commercial rooftop photovoltaic (PV) systems. Performance data for multiple products tested using this method are discussed, based on equipment from Enphase, Solar Edge, Maxim Integrated and SMA. In general, the annual recovery of shading losses from the module-level electronics evaluated is 25-35%, with the major difference between different trials being related to the number of parallel strings in the test installation rather than differences between the equipment tested. Appendix D data has been added in this update.