Advanced diagnostics for PV systems

Australia's rapid shift toward renewable energy has placed rooftop solar photovoltaic (PV) systems at the centre of the clean energy transition.

Yet for commercial and industrial operators, the full potential of these systems is frequently undermined by faults, degradation, and grid disturbances that in-built monitoring tools are too coarse to detect or diagnose with precision.

Without granular, actionable insights, operators face reactive maintenance cycles, unnecessary downtime, and reduced energy yield, ultimately weakening confidence in solar investment.

"The size and scale of rooftop and distributed solar continues to expand across the National Electricity Market as households and businesses increasingly invest in distributed generation for both economic and environmental reasons," said ISF Senior Research Consultant Jonathan Rispler.

"With this growth comes benefits for asset owners and the entire electricity system, but for installers and asset managers this comes with the challenge of managing and maintaining growing fleets for their customers. Having smart, timely and insightful monitoring systems becomes essential at scale to ensure that system capacity across a fleet portfolio is generating as expected."

The UTS Institute for Sustainable Futures (ISF) is a co-lead on this RACE for 2030 project, working alongside academic lead UNSW and industry partner Diagno to develop next-generation diagnostic capabilities for PV systems. The research applies data analytics and machine learning to two interconnected diagnostic challenges.

The first focuses on transforming raw inverter error codes into meaningful, traceable fault diagnoses. Rather than simply flagging that something has gone wrong, the research enriches automated fault detection with contextual, device-level information, enabling operators to pinpoint the nature and location of a fault and take targeted corrective action.

This shift from reactive troubleshooting to evidence-based maintenance has direct implications for system uptime, maintenance costs, and the longevity of solar assets.

The second challenge involves accurately distinguishing between faults originating within the PV system itself versus those caused by the broader electricity network. Grid-side disturbances are frequently misattributed, leading to unnecessary site visits and wasted resources.