PSA: Cleaning panels yielded an almost 25% increase in peak power

[fresnoboy]

Folks, we have a 22 KW array , located on home and garage as well as a poolhouse, and we cleaned the panels for the first time since they were installed about 9 months ago. We just crossed the peak output of the day, and I am seeing a 24.7% boost in peak output power compared with the last 5 days (which here in the SF Bay area have been all clear and sunny and not too warm this week. I was quite surprised about this, and wanted to give folks here a heads up that it may be good thing to do.

Now, we had not cleaned them since they went up, and we had had several weeks or really bad smoke from fires in the area, so it may be a lot worse in terms of grime and dust compared to others, but it's a significant uptick in power, especially in the wintertime.

thx
mike

[bob-n]

Thank you for the data. I'm sure that you are right that the power is increased.

It is very possible that you cleaned off a significant amount of grime to increase the power that much. But it is really hard to have a controlled experiment like this. Could the cloud cover have also changed? The haze factor changed?

Do you have any data to share to suggest that sunlight before and after cleaning were comparable?

[fresnoboy]

Thank you for the data. I'm sure that you are right that the power is increased.

It is very possible that you cleaned off a significant amount of grime to increase the power that much. But it is really hard to have a controlled experiment like this. Could the cloud cover have also changed? The haze factor changed?

Do you have any data to share to suggest that sunlight before and after cleaning were comparable?

Fair point. I compared it to the 5 previous days, and only in one of those days was there a sharp drop because of the clouds. Otherwise, it's been sunny all the time. Temps are warmer today, but that should decrease production, not increase it I think. It's been pretty mild do far here in the Bay Area.

Again, it's been about a year, and we don't get a lot of rain, and there was a lot of dirt on them that required some active scrubbing to remove. So not something most people have to deal with. But folks who have had a lot of smoke for weeks might want to consider doing a thorough washing.

[J.P.M.]

Fresnoboy:

I've no doubt you have had a significant increase in power and energy output after cleaning an array that's not been mechanically cleaned for 9 months. It probably hasn't rained much in the bay area either.

As I've described in prior posts and threads - including some in response to some of your prior posts, and without going into a lot of gory details that Bob-n alludes to as what would be required for a set of controlled experiments, and some of what I believe I've found after 6+ yrs. of data gathering/crunching:

1.) My array fouls at a very variable but average rate such that output falls approx. 0.75 - 0.90 %/week if it doesn't rain and/or I don't mechanically clean it.
That rate increases during Santa Ana season for example.
2.) A decent rain restores very approx. 2/3 to 3/4 of the lost performance. So, for example, if at some point in time, my array was performing at, say, 97% of possible output after adjusting for irradiance/insolation, weather variables, array age, and a few other things, a 1/4" of rain or so might improve that to, say, 99% of max. output for those same operating conditions. Note: that's NOT 99% of STC conditions, only 99% of the clean operating output under those same conditions (say, maybe 80 +% or so of STC output).
3.) Array fouling rates and type of dirt/etc. are quite local, subject to many hard to quantify variables and with probably a few surprises/unknown factors at work for many arrays/locations. Fouling materials for an array next to a freeway will be different than those found next to a fallow field, which will be different from dirt accumulated under a glidepath to O'Hare in Chicago for example.
4.) I believe that fouling rate of ~ 0.75% - 0.90%/week I quote, while variable as f(time, weather )is somewhat linear for maybe the first 8-10 weeks or so of no rain or mechanical cleaning. After that, I believe the fouling rate may decrease to a lower degradation rate/week. I have no data to back that up as I wash the array daily for 2 approx. 70 day periods in the spring/summer each year and measure/calc. array instantaneous efficiency at the minute of min. solar incidence angle for any day when the cos (angle of incidence) of beam irradiance on the array is > 0.99.
Anyway, the fouling rate may become asymptotic at some point as skylights and also vehicle windshields in junkyards never become opaque from the dirt they accumulate - even in the desert.
4.) Heavy dew can be good or bad for array cleanliness. A heavy dew on my array will produce ~ 0.01" - 0.02" precip. in the rain gage at the array. If the array is "dirty" when there's a dew, that much precip. tends to mud up the dust/dirt. Then, when the array heats up, the mud dries up, cakes up, and stays put. The process repeats until it rains. If a heavy dew hits a"clean" array, it tends to clean the array a bit further, but if that happens a lot, the panels can and do get the familiar "bathtub ring" at the bottom 10"-12" of each panel.
5.) I've found hosing my array from above the array, top to bottom using tap water and no scrubbing removes about as much of the fouling as a decent rainstorm. That is, it tends to restore ~ 2/3 - 3/4 of the performance lost to the fouling. I use about 3/4 gal. hose water per panel. I don't scrub (except for big guano strafings or owl skrocks) and I do not squeegee the array. It air dries.
6.) Right after I got the array, I experimented with various methods of cleaning the array to gauge the efficacy of various methods and liquids.
One observation: air dried tap water leaves hard water spots. Squeeged. distilled water doesn't. However, after 30 before/after measurements of instantaneous array output after each type of liquid/method, I was unable to detect a difference, statistical or otherwise in array performance. Conclusions: a.) Hard water spots transmit most solar radiation in the wavelengths that PV panels use, so distilled water/squeegeing is a waste of my time/effort b.) With respect to array cleanliness and associated performance, the human eye is a poor judge of estimating expected performance.
7.) If I ever stop farting around w/measurements on my array, for periods when it doesn't rain, I'll hose the array w/tap H2O about 1X/month @ the rate of maybe 3/4 gal./panel, let it drip dry and call it done. I'll also probably rinse/soap/brush/rinse using a soft cloth around Aug. 01 or so 1X/yr. Doing that may limit my array fouling with the hope of holding the average performance degradation from fouling to something like ~ 2% or so.

Take what you want of the above. Scrap the rest.

Your performance and results will vary and most likely be different than mine. The above is a serving suggestion only.

[foo1bar]

Do you have any data to share to suggest that sunlight before and after cleaning were comparable?

I'm also in the bay area, and I cleaned some of the panels in my array.
Since I have a solaredge system, I'm able to mostly account for the differences in sunlight, temperature, etc by using one of the panels as a control.
I used one of the panels as the scaling factor to account for the environmental issues.
With that, I saw for the ~20 unwashed panels a +/- 5% difference (per panel) from one week to the next. And 1 panel having a 10% drop (haven't looked yet - but I suspect when I look I'll find a bird crapped on it)
For the two that I only hosed down without brushing it was 16% and 12% increase.
For the 5 that I used a plastic bristle brush (floor push broom) on, it was average of 30% gain (low: 25.4% gain, high of 35.8%)

And when I was washing the panels, the water coming off the panels was visibly dirty, so it definitely had gotten quite dirty. I suspect that the ash fall from the fires around here was a significant part of the dirt on those panels.

[scrambler]

I'm also in the bay area, and I cleaned some of the panels in my array.
Since I have a solaredge system, I'm able to mostly account for the differences in sunlight, temperature, etc by using one of the panels as a control.
I used one of the panels as the scaling factor to account for the environmental issues.
With that, I saw for the ~20 unwashed panels a +/- 5% difference (per panel) from one week to the next. And 1 panel having a 10% drop (haven't looked yet - but I suspect when I look I'll find a bird crapped on it)
For the two that I only hosed down without brushing it was 16% and 12% increase.
For the 5 that I used a plastic bristle brush (floor push broom) on, it was average of 30% gain (low: 25.4% gain, high of 35.8%)

And when I was washing the panels, the water coming off the panels was visibly dirty, so it definitely had gotten quite dirty. I suspect that the ash fall from the fires around here was a significant part of the dirt on those panels.

That is great and reliable way to measure it, thank you for that.
I have been putting off doing a mechanical cleaning on mine for a while and they are clearly coated with dirt dust, so I am going to make an effort an get to it this week end

[scrambler]

Sprayed, brushed and rinsed yesterday what was a significant accumulation of dirt dust, and comparing the last few clear sky days at various time during the day, I estimate the dirt accumulation had reduced production by 15/18%. So cleaning them raised production by 18/20%.

Definitely a bigger impact than I had estimated, and worth keeping an eye on in a place where we have no rain (CA)

Note: the amount of dirt dust accumulation is probably higher than usual where I am, due do some land excavation work happening in our area...