How to set a cleaning interval instead of guessing at one
Why a fixed monthly or quarterly cleaning schedule leaves performance on the table, and how to derive an interval from measured soiling rate instead.
6 min readThe calendar is not a measurement
Most rooftop plants get cleaned on a fixed calendar: once a month, once a quarter, sometimes just "before the monsoon and after it." The interval is picked once, usually during commissioning, and rarely revisited. It has nothing to do with how fast the panels are actually soiling.
Soiling rate depends on dust load, wind direction, nearby construction, panel tilt, and season, all of which vary by site and by month. A fixed interval is either too aggressive, in which case you're paying for cleans that barely move the performance ratio, or too conservative, in which case the plant sits at a soiling loss for weeks longer than it should.
What a soiling-loss curve actually looks like
If you plot performance ratio against days since the last clean, most sites show a fairly predictable curve: a sharp drop in the first week or two after rain or a dusty spell, then a slower decline that flattens out. The point where the curve flattens is roughly where a clean stops paying for itself as quickly.
Building this curve doesn't require new hardware if you already have generation data and know your clean dates. A few cycles of that data is usually enough to see the shape. If there's no usable history, a short instrumented monitoring window will get you there faster than waiting out a full season.
Setting the trigger, not just the interval
Once you have a soiling-loss curve, the practical move is to set a performance-ratio threshold as the cleaning trigger, not a date. When the estimated soiling loss crosses that threshold, a clean gets scheduled. This is what we mean by a Soiling Loss Index: a running estimate of how much output is currently being lost to soiling, checked against the threshold that makes a clean worth doing.
In practice this usually means fewer cleans overall, but each one recovers more performance ratio than a clean done on a fixed schedule partway through the soiling curve. The maths favors measurement over habit almost every time.
What changes seasonally
The threshold itself should stay roughly constant, but the interval it produces will move around with the seasons. Expect shorter intervals during dry, dusty months and longer ones after sustained rain. If your cleaning schedule doesn't visibly compress and stretch across the year, it's probably still running on a calendar rather than on data.
Want this run on your own plant?
Send us your generation data and we'll tell you where you actually stand, not just what the theory predicts.