PV Reliability and Resilience in Challenging Climates

Research output: NRELPresentation

Abstract

Challenging climates for Photovoltaics are usually based on climate classification. However, extreme weather events such as high wind, flooding, large hail, extreme snow etc. have become more ubiquitous globally. To study the impact of extraordinary weather events on PV reliability we used two of the largest databases in the USA. First, the National Oceanic and Atmospheric Administration (NOAA) database on extreme weather and secondly, the PV Fleet Data Initiative where we have collected high-resolution PV performance data of more than 8 gigawatts or about 6-7% of all commercial and utility systems in the USA. We analyzed almost 200 systems between 2008-20022 that were immediately impacted by these weather events. The immediate impact (outages) was determined to be about 1% of or a median of approximately 3 days of annual lost production. However, the risk these events pose is exemplified by a long tail where 0.4 % of all systems lost more than 2 weeks annual production. We also found a threshold for high wind (90 km/hr) and hail (25mm), above which we observed significantly higher degradation implying long-term damage to the systems. In addition, we are using satellite imagery to quantify visible damage to PV plants. Finally, we share module, design and installation lessons from some observed case studies to improve extreme weather resilience for PV power systems.
Original languageAmerican English
Number of pages37
StatePublished - 2024

Publication series

NamePresented at the KAUST Workshop: Photovoltaic Innovation in Saudi Arabia: Achieving the Goals of the Kingdom's Vision 2030, 22-24 April 2024, Thuwal, Saudi Arabia

NREL Publication Number

  • NREL/PR-5K00-89693

Keywords

  • degradation
  • durability
  • extreme weather
  • performance
  • photovoltaics
  • reliability

Fingerprint

Dive into the research topics of 'PV Reliability and Resilience in Challenging Climates'. Together they form a unique fingerprint.

Cite this