Unlike traditional underperformance detection methods based on irradiance data, PV Health Scan detects underperformance by means of automated statistical analysis of the power output data of all the plant’s individual inverters or any other equipment (called ‘subsystem’) for which monitoring is available. The algorithm uses the data to calculate a virtual reference subsystem, subsequently offsetting each subsystem’s performance against it. This results in a heatmap, clearly highlighting underperforming subsystems.
The analysis also produces a health score spider chart, which indicates the plant’s relative performance to the fleet with respect to major and minor underperformance events, time between failures, health recovery time, downtime, and production deviation. This allows PV fleet owners to quantify production losses, assess plant market value, prioritize maintenance, and provide valuable return on experience for developing new assets.
Benchmarking PV plants to address underperformance
In March 2022, the PV Health Scan tool was integrated within the ENGIE Darwin platform, which collects data from every ENGIE renewable plant in real time. This integration means that a PV Health Scan analysis, including health score and KPI calculation, can be launched at any time in every plant connected to Darwin.
Fleet performance was further investigated later in 2022 with a first proof of concept and in 2023 with a pre-industrialization of the fleet performance analysis tool based on PV Health Scan on solar plants in ENGIE Green’s fleet in France. It allowed to identify and solve some issues related to data extraction and to highlight data quality issues (30% of PV parks having more than 10% missing power data at inverter level). This fleet performance analysis tool offers a different approach to rank the sites less dependent on external parameters (such as irradiance measurement and unavailability calculation by the operator) than classical approaches based on performance ratio.
Low performers are quickly identified; KPIs allow to understand the main reason behind this classification and a comparison with values observed in the fleet supporting ENGIE Green in developing corrective action plans to reduce underperformance and downtime.
Extending automated analysis
In a next step, the fleet performance analysis tool will be used on all ENGIE PV assets to look at asset-specific conditions such as the local climate, grid characteristics, and plant design. By extending automated analysis to the entire PV fleet, ENGIE will be able to detect problems early on, prioritize maintenance actions, carry out urgent corrective actions, and eventually avoid revenue loss. Statistics over the fleet will also provide return of experience for future investments (e.g. performance related to inverter’s make).
Meanwhile, the PV Health Scan programme was awarded an ENGIE Innovation Trophy in the Renewables category.