Case study · 2026
Solar PV Underperformance
How much solar loses, why, and what it costs.
“The difference between segments isn't that smaller installations have more faults. It's that nobody knows how many they have.”
Solar PV Underperformance 2026, section 8.2
30 years in operation: management makes the difference, not technology
A reference 100 kWp plant on the roof of an industrial building. The three scenarios differ only in how the plant is managed.
Underperformance against the investment model, assuming degradation of 0.5% a year. The model excludes the one-off outage during inverter replacement. Scenario C is an extrapolation [ESTIMATE].
What it costs over the plant's lifetime
The value of energy not produced over 30 years per 100 kWp, at €0.11/kWh. The gap between active and neglected management is roughly €54k.
[CALCULATION] Study, section 6.2.
More frequent checks, smaller losses
The strongest data point in the study. The gap of around 4 percentage points is the best available estimate of how much of the loss can be recovered.
Raptor Maps, Global Solar Report 2026. Correlation, not proven causation.
The plant size at which the service pays off
The share of annual output that a €340-a-year service must save to pay for itself. The threshold lies at around 75–100 kWp.
[CALCULATION] Energy price €0.11/kWh, 1,100 kWh/kWp. The recoverable underperformance of ≈3% is derived from the gap between 7% and 3%.
Why nobody sees the faults
The inverter manufacturer's free monitoring reliably catches only a minority of the causes of loss. Most faults don't look like faults there at all.
| Problem | Frequency | Severity | Visible in inverter monitoring? |
|---|---|---|---|
| Inverter outage | High | High | Yes |
| Whole-string outage | High | Medium–high | Partly |
| Connector and DC cabling faults | High | Medium | No |
| Soiling | Very high | Low–medium | Not without a model |
| Gradual shading | Medium | Medium | Not without a model |
| Microcracks | High | Low, rising | No |
| Hotspot | Medium | Medium, safety risk | No |
| PID | Low–medium | High | Partly |
| Faulty bypass diode | Medium, rising | High, safety risk | No |
| Incorrect settings at commissioning | Medium | Medium | Not without a model |
Study synthesis, section 5.5.
A 250 kWp plant in numbers
Underperformance reduces revenue linearly, while fixed costs stay the same. Based on CAPEX of €600/kWp.
| Metric | No losses | 5% underperformance | 10% underperformance | 15% underperformance |
|---|---|---|---|---|
| Annual output | 275 MWh | 261.3 MWh | 247.5 MWh | 233.8 MWh |
| Annual loss | — | €1,513 | €3,025 | €4,538 |
| Simple payback | 5.0 years | 5.2 years | 5.5 years | 5.8 years |
| Loss over 25 years | — | €37.8k | €75.6k | €113.4k |
[CALCULATION] Undiscounted, excluding operating costs. An order-of-magnitude illustration, not a financial model of the project.
Methodology and sources
Main sources: NREL PV Fleet Performance Data Initiative · IEA PVPS Task 13 · Raptor Maps Global Solar Report 2026 · kWh Analytics Solar Risk Assessment 2026 · Kiwa PI Berlin · SolarPower Europe · ERÚ (Czech Energy Regulatory Office). Full list in the PDF, section 23.
How much is your plant losing?
We compare actual output with expected output and show you what is worth fixing.