You’re probably asking “why didn’t he just google it and find some nice scientific paper that told him how much of an issue the dust was?”. Well I have spent a fair bit of time looking and still haven’t found any nice simple scientific papers that provide a simple enough answer for me. The numbers that do get thrown around range from anything between 1% to 50% which doesn’t really help in answering the basic question “how often should I be cleaning my panels”. I have listed a few good ones below but if anyone can send through some new links I’d be most appreciative.
Anyway I felt the best thing to do was to run a good old back yard (well roof top) experiment on my parents Solar PV grid connect system in Perth Western Australia. I installed this system back in June 2007, so coming up to it’s seventh birthday, and it hasn’t missed a beat, caused a house fire, electrocuted anyone or blown up the local network. It’s provided beautiful, problem free renewable energy and even almost paid for itself. Back then Solar PV grid connect system retailed for around $10Wp installed and now we are down to around $1/Wp, nice learning curve. It was nice just to get up on the roof to reminisce about the good old days as an installer and remember how fast the renewable energy industry has moved.
Enough reminiscing – back to the subject!
The system hasn’t been cleaned (other than rain) for over five years and Perth has just had it’s driest summer on record, so if they was ever going to be a “dusty to the max” situation the time was now.
The system is 2.79kWp with 18 x 155Wp BP Solar poly crystalline panels (yes back then 155Wp was considered a pretty big, efficient and fandangled panel) coupled with an SMA SB2500. It’s orientated pretty close to due north at around 30deg pitch on some custom made “Alex Bruce” welded steel frames.
I waited until I had two fairly similar days in solar irradiation and temperature lining up consecutively (10th and 11th of March). I then recorded the total energy output on day one, cleaned them after the sun went down and then recorded the total energy output after day two.
- Day one = 13.68kWhr
- Day two (after cleaning) = 13.59kWhr
A drop of 0.7%!
Now before you decide that you should go through dust on your panels I need to mention a couple of details. Solar irradiation (how much the sun did shine without clouds or stuff in the way) was pretty much spot on but it was hotter on day two (33degC vs 28degC) with a little bit less wind to help cool the panels. For those of you who know that panels don’t like getting too hot you’d probably be spotting a pretty big flaw in my highly scientific experiment.
Under these conditions I would have expected a 2-3% drop in performance. This was verified by going onto PVOutput and seeing how a few systems in the same area performed over the same two days (I had to assume some sneaky bugger hadn’t gone and cleaned them on the night in question). I took an average performance for three systems and it seemed to show a drop of around 3.2% between the 10th and 11th of March. So instead of our system dropping by 3.2% it only dropped by 0.7% suggesting the cleaning saved or improved performance by 2.5% which is a pretty good outcome.
However the devil is in the detail.
When I was cleaning the panels I started with the hose on mist to see how much dirt would come off while simulating rain. Again very unscientific but I would suggest 70-80% of the dirt including the odd bird poo came off with the fake rain. Funny thing is we were forecasted for a storm the next day so I probably could have just let nature do the cleaning. Furthermore, this would suggest that 70-80% of the dirt has only been there since the last rain, so this massive 2.5% improvement only is really only over the last few months. Over the course of a year, I reckon I could improve the situation by maybe 1.5% at most by regular cleaning.
There was a bit of lichen starting at the corners and I had to scrub at these bits to get them off. If these had grown further and ended up over the cells then there would be more substantial drop in output (might have taken another five years). I’ve also seen branches and leaves sitting over panels so it’s definitely worth having a periodic look or monitoring your systems output (saves you getting on the roof) to pick up any substantial decrease in performance.
Now for the money side. Everyone is on different tariffs so to make it easy I’ll base my calculations on $0.25/kWhr. This system is now (after seven years) probably knocking out 10kWhr/day average over a year. So a 1.5% improvement would amount to about $14/year. It took me about 20minutes to get on the roof, give them a clean and get back down (excluding the ten minutes of reminiscing). I’ll excluded travel time and costs as I rode around on my bike for exercise and like visiting my parents anyway. With this all in mind I’d be worth $40/hr cleaning panels. If you did it for a living and included travel, writing out invoices, insurance etc you’d probably be down to $10/hr.
I should also mention I went through about 50l of good drinking water and a squirt of washing liquid in the process. I think the broom could do another 100 or so systems before the bristles fell out…
On the environmental front you could suggest that the cleaning will save another 50kgCO2e/year from avoided fossil fuel burn so maybe that alone is enough motivation to get on the roof and give them a scrub.
I think I’ll wait for the rain to clean them next time but if I ever feel like a good reminisce about how fast things have moved for renewable energy, I’ll take a bucket and a broom and head on up to the roof….
The above story is just one slightly scientific anecdotal experience and:
- If you’re in a dryer dustier place than average suburban Perth Western Australia,
- or if your panels are much flatter (less than say 10deg pitch) and wont self clean as easy,
- or you’re in the path of some large migratory birds topped up on mulberries,
- or you’re directly under very low flying air craft (see link below),
- or you just like things being clean….
- or you just may think differently
Some links with much more scientific rigour behind them:
Cano, J. (2011). Photovoltaic Modules: Effect of Tilt Angle on Soiling. ARIZONA STATE UNIVERSITY. Retrieved from http://repository.asu.edu/attachments/57216/content/Cano_asu_0010N_11001.pdf
Mejia, F., & Kleissl, J. (n.d.). Soiling Losses for Solar Photovoltaic Systems in California. University of California. Retrieved from http://maeresearch.ucsd.edu/kleissl/pubs/MejiaKleisslSE2013_Soiling.pdf
Sulaiman, S. A., Hussain, H. H., Siti, N., Leh, H. N., & Razali, M. S. I. (2011). Effects of Dust on the Performance of PV Panels. World Academy of Science, Engineering and Technology, 58, 588–593. Retrieved from http://www.solarwash.ca/dat/content/impact_of_dust.pdf
Denver, J., Miller, J. T. A., Manager, P., Jackson, J., Engineer, S. D., Gupta, V., … Hoffner, J. (2009). Impact of Soiling and Pollution on PV Generation Performance Performance Loss Due to Pollution (pp. 1–5). Retrieved from http://www4.eere.energy.gov/solar/sunshot/resource_center/sites/default/files/san_jose_pv_module_soiling_impact.pdf