| Work Detail |
Australian researchers have developed multi-stage algorithms to remotely detect and accurately diagnose underperforming solar panels in residential and commercial photovoltaic systems.
Researchers at the University of New South Wales (UNSW) and the University of Technology Sydney have developed algorithms that they say can automatically detect a number of common poor performance problems with solar panels, such as wiring faults, degradation and shading.
Fiacre Rougieux, senior lecturer at UNSWs School of Photovoltaic and Renewable Energy Engineering, says the technology can also detect outages, disconnections and export limits, and can revolutionize fault diagnosis in photovoltaic systems.
“This is a game changer for Australian residential and commercial system operators,” he says. “By analyzing inverter and maximum power point data every five minutes, this algorithm can accurately diagnose poor performance issues, enabling early intervention and maximizing energy production.”
Rougieux explained that the researchers, collaborating on a New South Wales Smart Sensor Network project, used sensors and different types of analytical methods to develop a two-tier method for diagnosing poor performance of solar panels, which is It is estimated to cost A$7 billion (US$4.6 billion) in avoidable losses worldwide.
“We have created a high-level diagnostic based solely on AC power data, which can detect broad categories of problems, such as zero generation and disconnections,” he explains. “The advantage of this approach is that the diagnostics are completely technology independent and can work with any inverter and brand of maximum power point tracker.”
Since many inverter brands provide abundant AC and DC information, Rougieux explains that the team has also developed a more detailed algorithm that uses both AC and DC data, which can provide more actionable information to asset owners by detecting and classify more specific bugs, such as shading and string problems.
“This type of diagnosis requires both rule-based statistical methods and machine learning approaches for cases that cannot be detected with conventional rule-based methods,” he explains.
The technology has already been fully integrated into a commercial production platform that uses Global Sustainable Energy Solutions, the projects industrial partner, to monitor more than 100 MW of solar energy.
Ibrahim Ibrahim, head of the UTS team, explained that the technology, which can be applied to more than 1,200 photovoltaic systems, has enabled proactive measures to be taken that maximize energy production and improve system reliability.
“By significantly reducing avoidable losses, valued in billions worldwide, these technologies guarantee substantial cost savings to PV system owners,” he said.
Rougieux explained that the software could replace the need for expensive contractors to go on site to find out why a solar system is not performing enough.
“We had a city council that had an underperforming system for five months in a row,” he explains. “The contractor had an operation and maintenance contract, but the problem was not detected for months. Our algorithms detected it almost instantly. The big surprise for us was the staggering number of systems where an operations and maintenance contractor completely overlooked the poor performance we detected.”
The research team is now working on improving the algorithm so that it can diagnose a wider range of problems, such as shadows, dirt, and detailed network faults. |
