Soiling Loss & Cleaning Economics
The measured soiling rate — not an assumed one — sets the economic cleaning cycle.
The challenge
Cleaning on a fixed calendar over-cleans in some months and under-cleans in others. Both are expensive: one wastes wash cost and water, the other leaves generation on the table through the dustiest weeks of the year.
The reason it persists is that most sites have no measurement of their actual soiling rate — only a schedule inherited from the EPC.
The approach
Soiling produces a distinctive sawtooth: PR decays steadily between washes and steps back up after each one. The rate of decay is specific to the site and the season, and it can be measured directly from the plant's own performance data.
Once the rate is known, the economic cleaning trigger falls out of it — the point at which accumulated loss value exceeds the cost of a wash. A box plot across inverters then separates permanent degradation from cleaning quality.
- Measure the soiling decay rate between wash events
- Separate seasonal variation from permanent degradation
- Set the cleaning trigger where loss value exceeds wash cost
- Use inverter-level spread to grade cleaning quality
- Re-baseline after each wash to keep the rate current
The outcome
A cleaning cycle set by measured economics rather than by calendar, and a clear split between what cleaning can recover and what it cannot.
- 0.8–1.5% recoverable PR identified
- ≈1.2 MU/yr on a 50 MW plant
- Cleaning cadence set by measured rate, not assumption
- Degradation separated from cleaning quality per inverter
Curious what this would find in your portfolio?
A diagnostic runs across a representative slice and establishes the baseline before any build commitment.