Understanding degradation rates is important for the PV industry, along the entire value chain. According to the 2018 PV Innovation Roadmap, a DOE-conducted survey of 89 companies and researchers, better understanding of the degradation rate was cited as the single biggest need under “Reliability and Durability,” which itself was the most popular issue among all categories. According to NREL’s 2016 Compendium of Photovoltaic Degradation Rates, the uncertainty in degradation has a total potential impact of $17/MWh, “even exceeding the impact of the initial cost” of the solar power plant itself.
Today’s state-of-the-art study of degradation is constrained by the lack of real-world performance data, which is why solar risk management company kWh Analytics has won a $1.25 million award from the U.S. Department of Energy Solar Energy Technologies Office (SETO) to use its real-world data to quantify degradation rates. With this award, kWh Analytics will build a machine learning model on its industry-wide data repository to statistically quantify degradation rates on an ongoing basis. The project is titled, “Deciphering Degradation: Machine Learning on Real-World Performance Data,” and will be led by kWh Analytics Data Scientists Ben Browne, Victor Garcia, and Adam Shinn (Principal Investigator).
Degradation rates have far-reaching implications, stretching from BOM decisions made at the module factories to financial assumptions made at the banks. Experts from manufacturing and finance attest to the importance of kWh Analytics’ work to quantify degradation rates:
“kWh Analytics’ approach of using its historical data to deeply understand and quantify degradation can have a significant positive impact on the industry,” said Howard Wenger, clean power investor and former President and CEO of SunPower Systems. “The lack of long-term solar panel performance data has been a real challenge because buyers of inferior panels have been unable to readily know the difference in quality between the various products on the market. With data on one-in-five American solar assets and growing, kWh Analytics is uniquely positioned to address this industry-wide challenge and enable the solar industry to accurately price quality and inform buyers on how to value their solar power assets before and after purchase.”
“kWh Analytics is recognized as an industry expert on PV plant performance, and they have been instrumental in the effort to make solar affordable and brought to scale,” said Jon Previtali, Director of Technology and Technical Services for Wells Fargo’s Environmental Finance team. “kWh Analytics’ quantification of degradation rates using 10+ years of historical performance data will help investors better inform our energy production estimates, project proformas and investment theses.”
Further, kWh Analytics also intends to equip insurers with insights that enable them to identify reliable modules and charge less when insuring them. The introduction of that price signal—explicitly linking module reliability to levelized cost of energy (LCOE) through intelligent insurance premium pricing—is an innovative approach that solves a fundamental incentive misalignment for the industry.
With SETO’s support, kWh Analytics’ work to quantify degradation rates will create systemic, scalable impact by:
• Delivering real-world perspective on the causes of degradation, enabling the industry to improve reliability.
• Creating a price signal that puts a tangible dollar value on high reliability, via insurance premium pricing.
• Contributing to the open source community by sharing the proposal’s resulting machine learning model.
— Solar Builder magazine