Soiling Wash Decision at a Tracker Site
At a 4.9 MWDC single-axis tracker site in a semi-arid region, monthly performance reports show PR falling from 82.1% in March to 77.9% in June. The site has a soiling station, but its glass coupon was installed 18 months ago and is cleaned quarterly on the same schedule as the arrays. Water for washing costs $0.004/L delivered; a full wash consumes about 1.1 million L and costs $9,500 in labor and equipment. The energy price is $0.052/kWh, and the site generates about 1,050 MWh/month. The O&M budget has $40,000 remaining for the year. String-level data show the loss is spread evenly, with the tracker rows at the west edge showing slightly higher losses. There was also a five-day inverter outage on one of eight blocks in May. The asset owner asks whether to wash now, wait until fall, or investigate something else first.
1. Quantify the loss and decide whether a full wash is justified on these numbers.
Read the worked answer
Convert the PR gap into money. A 4.2-point drop on 1,050 MWh is about 44 MWh, or roughly $2,290 at $0.052/kWh.
The five-day outage on one of eight blocks accounts for about 2.1% of monthly energy - roughly 22 MWh, or $1,150 - so the outage alone explains a large part of the gap and is not a soiling story. That leaves about 2.1 points, roughly $1,100, as the upper bound for soiling plus other losses.
A full wash costs $9,500. The break-even soiling loss is therefore on the order of 17% of monthly output, far above what the data show. On these numbers, do not authorize a full wash now; the spend would not be recovered at this energy price.
2. How reliable is the soiling station, and what else could explain the west-edge rows?
Read the worked answer
The soiling station is suspect. An 18-month-old coupon cleaned on the same quarterly schedule as the arrays may not represent current soiling, and its glass may have aged differently from the module front glass. Check the cleaning log, compare its output with string currents, and spot-check by cleaning a test row and comparing power or I-V curves before and after.
The west-edge rows reading slightly higher losses could reflect tracker stow position, wind-driven dust deposition, row-to-row shading at low sun, or a tracker pointing error - not uniform soiling. Review tracker availability and pointing data before attributing the loss to dirt.
3. What recommendation should you give the owner?
Read the worked answer
Recommend investigating before washing: verify tracker availability and pointing, confirm the Block 3 outage is correctly excluded from the PR calculation, and clean and validate the irradiance sensors. Defer the full wash until measured soiling approaches break-even or until a higher-value season, and preserve budget for targeted spot cleaning and sensor maintenance.
If a wash is required for contractual or warranty reasons, document the decision basis and consider washing only the highest-loss rows rather than the whole site.