The site and the test
Ecomundo owns and maintains a 48 kWp solar carport in Chile. Cement dust from a nearby precast-concrete products factory had built up in heavy layers and cured onto the modules. Manual cleaning with water alone had already been tried, without success. Scrapers and high-pressure washing were ruled out: they would damage the modules and void the warranty, and the modules in the middle of the carport were hard to reach.
- Modules: 90, split into two halves of 45, each half on its own inverter.
- Treated half: Chemitek Inorganic Removal Agent (CRA), diluted, sprayed onto completely dry modules, brushed thoroughly and rinsed. Ecomundo then coated the same 45 modules with Antistatic Solar Armor (ASA), Chemitek’s antistatic coating.
- Control half: 45 modules cleaned with water, the traditional way. A significant amount of cement stayed on them.
- Dates: both halves were cleaned on 29 September 2022, and Chemitek’s daily table runs from 1 October to 14 November. The report describes the test as 90 days; the dates span 46, so we quote the dates.
- Measure: daily energy from each inverter, with the treated inverter’s extra output expressed as a percentage of the control inverter’s.
Results: the gap kept growing
- 1 Oct4.56%
- 8 Oct2.97%
- 15 Oct6.85%
- 22 Oct8.47%
- 29 Oct9.67%
- 5 Nov11.34%
- 8 Nov11.88%
CRA and ASA inverter against the water-cleaned inverter, 2022 (Chemitek case).
| Date (2022) | CRA + ASA inverter (kWh) | Water-cleaned inverter (kWh) | Gap |
|---|---|---|---|
| 1 Oct | 103.97 | 99.44 | +4.56% |
| 8 Oct | 79.10 | 76.82 | +2.97% |
| 15 Oct | 109.64 | 102.61 | +6.85% |
| 22 Oct | 110.91 | 102.25 | +8.47% |
| 29 Oct | 108.13 | 98.60 | +9.67% |
| 5 Nov | 102.82 | 92.35 | +11.34% |
| 8 Nov | 91.28 | 81.59 | +11.88% |
Every seventh day from 1 October to 5 November in Chemitek’s daily table, plus 8 November, the last day Chemitek counts.
Chemitek’s summary: the gap started at 4.56% and was close to 11.9% by 8 November, which it treats as the last day of available generation. It flagged days when a grid overvoltage restarted the inverters as atypical. Chemitek then projected that the gap could reach up to 15%. That is a projection, not a result: the measured figure is about 11.9%.
A month after cleaning, the coated modules were visibly cleaner than the uncoated ones, and Chemitek attributes the growing gap to the antistatic coating slowing re-soiling. Ecomundo judged the test a success and began working out how often to clean and recoat for its contamination rate.
How to read it. Two cautions. The treated modules got both CRA and ASA, so the test can’t split the gain between removal and protection. And the report gives no comparison of the two inverters before cleaning. Our pilots record treated and control strings under the same sun before any treatment for exactly this reason.
What this means for a South African plant
- Contaminant. Cured cement forms wherever cement, limestone or concrete dust meets moisture: beside cement works, precast yards, quarries and building sites. Moisture dissolves part of the dust, which recrystallises into a bond that dry cleaning can’t break (IEA PVPS, 2022). Our North West pilot on a ±15.6 MWp plant beside a cement works found the same problem.
- Water. More water wasn’t the answer here; chemistry was. That matters in a country that gets about 490 mm of rain a year, about half the world average (WWF-SA).
- Method. Apply CRA to dry glass, keep it wet, rinse before it dries, and protect straight after restoring. Chemitek’s current coating is Antistatic Solar Armor 2.0, launched in 2025; this 2022 test used the earlier ASA. See how we treat cement dust.
- Fit. C&I carports and roofs beside industrial dust sources are the closest match to this site. Utility plants near cement works or quarries face the same chemistry at a larger scale.
