The site and the test
TNBES operates an 850 kWp solar hybrid station (solar, batteries and diesel) for TNB at RPS Kemar in Gerik, Perak, Malaysia. Forest surrounds the plant. The soiling is fine dirt and dust, seasonal pollen, and occasionally beeswax and bird droppings. The plant was normally cleaned with manual brushes about every two months in the dry season, and not at all in the rainy season.
- Strings: three, of 285 panels each.
- SCM4: washed with water only, as the normal-maintenance baseline.
- SCM5: washed with Chemitek Solar Wash Protect (SWP), applied by hand on 17 and 18 August 2020: diluted, sprayed, brushed and rinsed.
- SCM6: not washed, as the no-maintenance baseline.
- Water: both washed strings used the same well water.
- Dates: 18 August to 6 September 2020 (20 days).
- Measure: each string’s energy at the inverter, compared with the average of the three. All three saw the same irradiation and ambient temperature.
Results
- Well water only−4.9%
- Solar Wash Protect, same water+3.84%
- Not washed+1.06%
18 August to 6 September 2020 (Chemitek case).
| String | Treatment | Energy, 18 Aug – 6 Sep 2020 | Against the three-string average |
|---|---|---|---|
| SCM4 | Well water only | 3,250.10 kWh | −4.90% |
| SCM5 | Solar Wash Protect, same well water | 3,548.80 kWh | +3.84% |
| SCM6 | Not washed | 3,453.70 kWh | +1.06% |
The string washed with well water produced less than the string nobody washed. Chemitek’s explanation: the well water was so high in total dissolved solids (TDS), minerals and organic matter that it left the panels dirtier than before the clean.
The SWP string, washed with the same water, came out on top; Chemitek credits softeners in the SWP formula for reducing the effect of the TDS. Treating the unwashed string’s +1.06% as the effect of rising irradiation, Chemitek deducted it to give a net 2.78% from SWP over the 20 days. The O&M team also reported faster, easier cleaning and much less water, since water was only needed for rinsing, and moved all its cleaning to SWP.
How to read it. All three figures are relative to the average of the three strings, so the clearest reading is the ranking: SWP first, no wash second, well water last. The report gives no TDS reading for the well water, so it can’t tell you how bad water must be before washing does harm. Only a test of your own water can.
What this means for a South African plant
- Test the water first. Module makers set limits for cleaning water, on different bases. LONGi’s manual allows TDS up to 1,000 mg/L and hardness of 0–450 mg/L without stating the basis; First Solar prefers hardness under 75 mg/L and prohibits 75–180 mg/L water on anti-reflective-coated modules. Our cleaning-water testing checks yours against your modules’ limits.
- Treat the water to match the problem. Chemitek Water Softening Agent captures calcium and magnesium in the wash water, so they can’t deposit as scale when a rinse dries; set the dose from a hardness test with test strips, because a TDS meter reads the agent itself. Softening leaves other dissolved solids in the water, so for water high in TDS, reverse osmosis may be the better fix. See water quality for solar panel cleaning.
- Never let a rinse dry on hot glass. Whatever the water, what is dissolved in it stays behind when it dries. Clean when the glass is cool and rinse section by section. See hard-water scale.
- Borehole and well water. The lesson matters most where a borehole or well is the only water on site, as a well was here. Treat or soften it, and prove the method on a few strings with a pilot first.
