Evey year, in wide areas of West Africa, from December to February or March, the harmattan season descends, characterised by cooler temperatures, a dry dusty wind and hazy skies. The dust, purportedly blown from the Sahara, is pervasive and quickly becomes apparent to the naked eye on all exposed surfaces. Solar panels, located outdoors, usually on difficult-to-reach rooftops, are very susceptible to this. The effect is insidious by the very nature of solar energy. Completely silent and effective through an accumulation of energy over many hours rather than a sudden release over minutes, solar panel output is difficult to monitor by the casual observer, and a drop in output will often go unnoticed.

How much does an accumulation of Harmattan dust affect solar output? I set out to investigate on a late February day, cloudless but moderately hazy. 100 Watt solar panels on my rooftop had not been cleaned since the start of the Harmattan season, indeed since the last downpour of the rainy season, possibly in November. Located in Sierra Leone, 7.5 deg North of the equator, my panels are mounted flat on my rooftop, on sloping, corrugated roofing sheets, at azimuth angles of approximately 15 degrees. This is technically not ideal, but for the purposes of ease of mounting least troublesome.

During the investigation solar panel output was monitored over a period of time before, during and after the panels were cleaned with a soft cloth and moderately soapy water. Output readings driven by computer software were taken every 12 seconds and saved directly to an Excel spreadsheet. The cleaning exercise lasted less than ten minutes, so change in output due to the movement of the sun rather than to the cleaning would have been small. Skies were cloudless, so  any variation due to cloud cover would have been small. However the effect being measured, the effect of the dust that had already settled on the panels, is separate from the effect on solar radiation of the general Harmattan haze, which of course was still present.

  A graph of solar output versus time is shown. Readings were very consistent before and after the cleaning and  a significant rise in output is observed. Solar output monitored for several days before the cleaning and after the cleaning confirm this pattern.

CONCLUSION: Based on these results I conclude that the effect of settled Harmattan dust on output of uncleaned solar panels at the end of the Harmattan season is of the order of a thirty percent reduction.