From low output to an actionable diagnosis.

Why it matters

A performance gap can be real while the first suspected cause is wrong. If availability, curtailment, sensor quality, reference conditions and environmental effects are mixed together, a team can spend time correcting an issue that does not explain the observed loss.

Separate the main loss pathways

• Availability: planned and unplanned downtime, communications gaps and equipment states.
• Grid and operating limits: curtailment, export constraints and control setpoints.
• Equipment behaviour: inverter, string, connector, protection and communication patterns.
• Environmental effects: soiling, shading, temperature, irradiance variability and module mismatch.
• Data and baseline: sensor quality, timestamps, missing data and whether the expected-output model fits the question.

Practical diagnostic sequence

1. State the performance question and time window.
2. Check whether the apparent loss is larger than the uncertainty in the data and baseline.
3. Quantify known operating exclusions such as downtime and curtailment.
4. Compare performance at plant, inverter, block and string level where the data supports it.
5. Use site evidence or targeted testing to distinguish remaining explanations.
6. Record the confidence, limitation and next action for each finding.

Assumptions and limitations

Monitoring data alone may not distinguish causes that produce similar patterns. Conversely, a field observation may not quantify energy impact without a suitable baseline. The investigation should state data gaps and avoid treating correlation as proof of root cause.

References

• IEC 61724-1:2021 — terminology, equipment and methods for PV performance monitoring and analysis.
• IEA PVPS Task 13: Analytical Monitoring of PV Systems — methods and models for grid-connected PV performance analysis.

Related SolarOZ content

Discuss an underperformance issue