I have had my solar panels for just over a year now, and I was honestly not planning on cleaning them… They looked pretty clean from the street view… It didn’t look like there was much visible dust or “dullness” to the panels…they still seemed to “shine” and “glisten” in the daytime…
But after looking at my numbers from last year, my production is down over 10% day-to-day… Last June 2020 I was seeing daily production numbers in the 24kWh range… And this year (same exact calendar days), I am seeing numbers in the 21kWh range…
So this morning I decided to go up and give them a quick cleaning… Took about an hour between the hours of 7am and 8am. I have a total of 10 panels, so I didn’t suspect it would take a whole lot of time…
I’ll report back my results here later… But I thought I would start a thread to see if any of you have done some fine measurement, and whether you have concluded that solar panel cleaning is worth it (or not)…
I did some analysis a while back, looking at solar panel “degradation” over time. I tried looking at daily production over 6 years. Came up with about 2% degradation per year.
It might be a mix of both. I have probably had my panels cleaned 4 times since 2013 when they first went live. Some are easy to clean, some are hard (second floor roof). Here’s a view of monthly production with a prediction for the next 3 years based on Prophet, a Facebook open-source team based prediction algorithm. Black dots represent monthly production, the dark blue line is the fit line based on the Prophet prediction and the light blue is the 90% confidence band. There are a few points that fall outside the band, either due to startup issues or internet outages that reduced the usage sent to SolarCity/Tesla.
And here’s a view of the decomposition / analysis Prophet uses to fit the data and make the predictions. You can see the trend line dropping by about 10kWh / month per year.
The good news is that yesterday’s production numbers compared to the day before were up 15%. Pretty significant. I want to do a 5 day average to smooth out any anomalies, but I am pretty confident that the 15% increase is real… Every single day this month has yielded about 21-22kWhs… Yesterday yielded 25kWhs. Yesterday’s production was slightly better than my all-time daily record of 24.9kWhs.
So the bad news is that it looks like panel cleaning matters in a meaningful way… I was actually hoping that my results would show that my production numbers increased, but not by enough to justify climbing up on my roof and washing them on a regular basis… 15% of my last year’s 5600 kWh production is an extra 840 kWh. And my monthly consumption averages 380 kWh, so that’s over two months of energy! The problem is that I can’t yield that entire 15% benefit throughout the year with a single annual cleaning. So either I need to figure out a regular cleaning cadence throughout the year to optimize my production… Or I need to just clean them once a year and live with the fact that my solar production will under-perform as my next annual cleaning nears…
I’ll report back in a few days to see how the next 5 days or so looks…
panel cleaning matters but not as much as you think, its too subjective to expect the same generation year over year on similar months due to changes in weather and cloud patterns, cleaning does help but not at the % of generation
Yeah… I get that… And additionally, I’ve had my system now for just over a year, and the panels started out clean, right? So I really can’t expect a 15% benefit across the entire year…
I am confident that my 15% is actually 15%… I live in a location that gives me a majority of cloudless, bright sunny days that shine down on my rooftop with absolutely no obstructions… My summers are full of perfect bellcurve production days…
Last June, my most productive day was 24.9 kWhs… Before cleaning my panels, I was peaking at 22.4kWhs… After cleaning, I am back at 25.0 kWhs…
I agree with you that there are a lot of variables that make forecasting unpredictable… But I do feel like my location is as pretty consistent, even though I understand that I am still “in the wild”.
What state is everyone in? I watched a “one year later” video on a guy’s Tesla roof. He’s in California and said the wildfire last year coated his roof in a fairly hefty layer of soot. Most other states won’t have that problem, and states with more rain might be able to naturally wash that off.
Mine are going on 11 years. And had them installed on the garage. I can get to them with a step ladder.
When they are covered in snow I clean them off. A lot of dirt comes with the snow, This can be several times a week if the weather is bad, so the routine is shovel the driveway, side walk, clean the snow off the solar panels. In the summer if they are really dirty and no rain is due I spray them with a hose and clean them off. But at the most once or twice a summer. It doesn’t rain often here (Colorado) but enough to keep them sort of clean.
Very simple process I used for cleaning… First, it’s important to note that I have a one-story ranch style home with a pitched roof… Stuff can roll off the roof, but for the most part, I am able to balance myself, and balance my stuff on the roof pretty well…
Equipment
Ladder
Six-Foot Telescopic Pole
Window Washing “Scrubber”
Window Squeegee
Professional Easy Glide Glass Cleaner
Two Buckets
I leverage the same equipment and supplies that I use to wash the exterior windows of my home, but I put a twist on my window washing process by using a “two-bucket” washing technique used by car detailers and enthusiasts when washing/detailing cars.
I fill one bucket with warm wash water and the prescribed 4 capfuls of glass cleaner per gallon of water. The other bucket is for clear rinse water. I really like the window cleaner fluid that I use because it doesn’t create a lot of heavy suds, and it is designed specifically for cleaning glass windows.
I bring all of the equipment up to the roof, in addition to my water hose and perform the following process:
Rinse the section of panels that I am working on with the garden hose. First with a light raindrop setting, and then with a more forceful spray (but not too forceful that the panels will be damaged). The goal here is to get as much of the surface dirt off the panels as possible.
Clean panels methodically working in columns first (up/down) and I work my way from left to right.
For each panel, I use a “two-bucket” process. I first dip my wash scrubber in the warm soapy water and wash the panel up/down and left-to-right, flipping the scrubber midway through. After washing each panel, I dunk & agitate the scrubber into the rinse bucket to remove as much of the major filth, dirt and debris as possible and wring as much excess water as possible before dunking the scrubber back in the soapy bucket… And as the old saying goes… “Lather, Rinse, Repeat”…
The rinse bucket will start to get filthy…but don’t worry about it… The main reason for the rinse bucket is to reduce the amount of filth that gets into the soapy bucket so you have a nice effective soapy bucket for as long as possible.
Rinse panels
Squeegee each panel dry. Again, working from top/bottom and left-to-right. If you have hard water like I do, you’ll want to get as much of the moisture off your panels as quickly as possible before the water air dries. I washed my panels in the early morning, so the panels stay wet long enough for me to squeegee them dry…
Here is a photo of the equipment/supplies that I use…
When you clean them, the evaporating water also cools them (for a while) so you will see increased numbers but not from cleaning alone.
I have panels from 2002. I also happen to have 1 new panel (spare) that has been stored inside in the dark, no light on it what so ever.
When I compare the output of the “new” (aged but kept in the dark) with the ones that were out in the sun for 19 years, I see about a 6% difference in performance.
Yeah… I wonder if my “day 1” performance after cleaning was impacted at all by this…
I did the cleaning at 7am… And I squeegee’d all the water off of the panels… And conductive temps probably were lower as a result… But by 8am, when I completed the cleaning, I would have thought that the cleaning impact on temperatures would be an absolute non-factor…
A lot of people take reading before and after cleaning and say “i gained xx % by cleaning”
Your method of comparing a day does not include real temperate changes because of cleaning.
I am in socal and even though we have pure blue sky most of the time, there is a lot of things you can’t see that make the daily generation fluctuate. Last 3 days where “sunny & warm”
Yet there was a 6% change in generated power.
Comparing day before/after cleaning is not really reliable.
In my experience (I have a lot of flat panels that collect a lot of dirt) when I just hose them off when they are very dirty, I gain about 5% over the next week.
@dannyterhaar Agreed. I think it’s too difficult to discern cleaning improvement on the following day. Like you, here in SoCal, two, back to back, sunny, clear skies will produce different production amounts.
Additionally, as it gets hotter I’m sure it takes it’s toll on the Enphase Microinverters performance on the roof here in the desert. And as you know living in SoCal, some days are smoggier than others with the particulate matter floating around.
I’m sure cleaning works, I just think it’s too difficult to compare it day to day.
I think the panels loose more power under high temperatures than the microinverters.
This is one of the differences between poly & mono crystalline solar panels.
Most mono crystalline solar cells have a temperature coefficient of around -0.3% / C to -0.5% / C
So when the temperature rises 1 degree Celsius, the mono crystalline solar cell will temporarily lose 0.3% to 0.5% of its efficiency. Poly crystalline PV cells have a higher temperature coefficient
eg my 19 year old poly cells:
