Rain, Snow, Dust

[Charlie W]

I start with praise for this site. I consider it reliable and not captured by hype and evangelism. Call me Joe Friday: "Just the facts, ma'am."

In that vein, I am interested in facts relating to cleaning the panels. How much degradation comes from the dirt left by rain and snow, and how about dust? Be scientific. I have asked people, and I have gotten widely differing answers. Links really help. I am here because I want straight answers. Just the facts, ma'am. Thanks much.

[J.P.M.]

Welcome to the forum of few(er) illusions.

Short answer from me is: It mostly depends on the array location, the air quality that comes with the area and several other things of lesser importance, one such factor being array orientation and that orientation relative to prevailing winds. If you're interested in gory details, read on.

I've got a lot of data measured over 8 1/2 + years at my site, but it's only valid for my site. IMO only, the data was gathered in a careful and scientific way, or at least with the same engineering methods I used as an engineer where one of my specialties was studying and measuring heat exchanger fouling, its causes and methods of control and prevention. The data on my array does seem to generally agree with other data on array fouling as can be found in the journals.

Precis:

My array is 16 ea., Sunpower 327's arranged in a 4 X4 configuration with the panels in landscape, 18.75 Deg. tilt, 195 deg. azimuth. Shading late in the day produces an approx. 3.5% performance penalty on an annual basis. Running 365 day average specific output after the first year's burn-in has been 1,722 kWh/365 days per installed STC kW.

Results of measurements:

1.) In a semi rural location in zip 92026, when measured as (net electrical energy output)/(net irradiance input), on average, my array's performance decreases by ~ 0.75%/week when it doesn't rain.

2.) There seems to be little seasonal variation in that rate that I can decern from the data.

3.) That fouling rate at my site seems to be somewhat asymptotic in character and becomes mostly constant or at least greatly reduced rate at an ~ 8% - 10% performance penalty after ~ 10-12 weeks with the rate slowing down and tending to stay constant after that. That number does seem to have some seasonal dependence with winter fouling without rain peaking closer to the 8% number.

4.) When it occurs, rain totals of ~ 0.25" to 0.50" over a few hours will usually restore about 1/2 to 2/3 of the performance lost to fouling that existed prior to the rain event. A "good" rain of ~ 1"- 2" will restore about, maybe 3/4 to 90% of the performance lost to fouling. The clean rate of 0.75% fouling per week then continues as described above.

5.) Somewhat subjectively, a little bit of moisture can be more problematic than no rain. Reason: Dust accumulation and a heavy dew or a few sprinkles will often tend to make a mud soup of sorts. The increased viscosity of that soup makes it less able to flow down a tilted panel. Then, when the sun comes out and heats the mud, it dries. If the mud flowed at all, it got to the lower part of the panel and formed the bathtub ring that is commonly seen on panels. The idea that some parts of a panel get less irradiance than others can have a greater effect on panel performance than if equal fouling over an entire panel is assumed.

6.) Hosing the array from above with plain old tap water at a rate of ~ 1 gal./panel or a bit less has about the same effectiveness at restoring performance lost to fouling as the 0.25" -0.50" of rain, that is, it will restore ~ 1/2 to 2/3 of the performance lost to fouling. The hosings take ~ 10-15 minutes or so for my array, including time to get to/from the roof.

7.) Contrary to what a lot of folks may think or assume, using D.I. or distilled H2O as opposed to plain old tap water, or squeegeeing the panels to remove water spots has no benefit over using plain old hosing and a drip dry as described above that I have been able to measure.

8.) Washing the array with soap and a soft brush will restore just about all the performance lost to fouling. The procedure I use to wash, as opposed to a simple rinse, is to rinse, brush with soft cloth and get rid of the guano, wash with plain old dish detergent (not dishwasher soap) and use the soft cloth, followed by two rinses with tap water, no squeegeeing and a simple drip/air dry. When I'm not measuring fouling, I hose my array ~ every 4 weeks and wash it 2X/year. It takes ~ 90 minutes to wash the array but I take my time doing it.

[peakbagger]

Snow is a PITA and solar salesman typically understate the impact of snow on panels. I live in Northern NH and have 4 arrays of various types and angles and they all can get covered with snow and stay covered for days with zero output if the conditions are right. Snow melts at 32 degrees and snow reflects light rather than absorbing it so the only way snow is coming off a panel is for the air space under the panel to get over 32F and then the snow looses its grip with the panel surface. If the panel angle is steep enough the snow slides down and then part of the panel is exposed. The panel is dark and then its starts to absorb sunlight and its surface temp is raised to the point where the rest of the snow starts melting. Modern homes in snow zones normally are built with vented attic spaces to keep temp cold so its takes a while for roof mounted panels to warm up enough to melt.

My pole mount is adjusted for seasonal angle and even at 30 degrees tilt (from vertical) I can get snow buildup for several days if the outdoor temps stay cold. If I expose some of the panels, even on below freezing days, they will clear off eventually. The snow does not mysteriously disappear, much of it slides down to the base of the panel and piles up to the point where the bottom of the panel is covered in lump of snow until shoveled away. Those with microinverters will get some production for the panels that are not covered but those with string inverters will lose the string. Note many string inverters have multiple MPPT circuits so careful string layout may keep some of the array functional.

The only way around not cleaning snow off panels, such as at a camp or remote location is mount them vertical on a wall under an overhang. In winter, snow on the ground reflects a lot of light so even though the panel angle is not optimal, the output is higher than PV Watts will predict as it does not take into account increased insolation. I just use a snow rake on the panels I can get to and live with the one array on my second floor roof that is far harder to get to having stretches of no output. I can accelerate it coming back on line a bit by exposing the lower edge with my roof rake. I run my snow blower along the lower edge of my pole mount (next one will be higher).

Next house will be all ground mount probably with adjustable angle with plenty of room for the snow to slide.

[bcroe]

An approach to keeping snow clear, go to
Solar Panels for home

Solar Panels for your Home, Grid Tied Solar PV

Stickey: Snow Tolerant PV Mounting

You will find my approach, very successful in recent years, and
some others. The very last post states their independant success
using these methods. Bruce Roe

[Charlie W]

I considered panels after we built our house 5 years ago, but the numbers didn't work given that our electricity costs 9.62 cents/kWh. This past winter, we had 12 feet of snow and were trapped for a while on our rural land. I remarked that even ground mount (the only way to go, IMO) wouldn't have been very good. The snow was waist deep for quite a while, and cleaning them would've been hard. Roof mounted would have been a real disaster; we had three feet of ice and snow up there, with no way to remove it.

Snow is kinda-sorta obvious, so I'm more curious about rain and dust. Starting with rain, we all learned very early that every raindrop forms around a speck of flying dust (or salt in some cases over the oceans), which is why a vehicle left outside in the rain will be dirty. I wonder if there's any quantification of how much of a production drop an uncleaned panel suffers as a consequence of rain. Same thing for dust. These questions are pertinent especially for utility solar arrays, which probably don't get the level of cleaning that at least some residential panels do. I think dust would be a real issue in the arid intermountain West and in the arid Great Plains, but also in urban areas near busy thoroughfares and under flight paths. Also, regardless of location, pollen could be an issue too.

Has anyone quantified this stuff? I hear varying things about the impact on production from rain, dust, and pollen. Another issue that comes to mind is whether the glass in panels is durable. I'd expect that panels would be vulnerable to hail damage. I also wonder if, over time, panel glass gets cloudy or warped. How about condensation? Are these issues, and if so, are there numbers? I'm a numbers guy to the Nth degree.

[J.P.M.]

I considered panels after we built our house 5 years ago, but the numbers didn't work given that our electricity costs 9.62 cents/kWh. This past winter, we had 12 feet of snow and were trapped for a while on our rural land. I remarked that even ground mount (the only way to go, IMO) wouldn't have been very good. The snow was waist deep for quite a while, and cleaning them would've been hard. Roof mounted would have been a real disaster; we had three feet of ice and snow up there, with no way to remove it.

Snow is kinda-sorta obvious, so I'm more curious about rain and dust. Starting with rain, we all learned very early that every raindrop forms around a speck of flying dust (or salt in some cases over the oceans), which is why a vehicle left outside in the rain will be dirty. I wonder if there's any quantification of how much of a production drop an uncleaned panel suffers as a consequence of rain. Same thing for dust. These questions are pertinent especially for utility solar arrays, which probably don't get the level of cleaning that at least some residential panels do. I think dust would be a real issue in the arid intermountain West and in the arid Great Plains, but also in urban areas near busy thoroughfares and under flight paths. Also, regardless of location, pollen could be an issue too.

Has anyone quantified this stuff? I hear varying things about the impact on production from rain, dust, and pollen. Another issue that comes to mind is whether the glass in panels is durable. I'd expect that panels would be vulnerable to hail damage. I also wonder if, over time, panel glass gets cloudy or warped. How about condensation? Are these issues, and if so, are there numbers? I'm a numbers guy to the Nth degree.

Charlie:

Asking questions one can't find answers to on their own is always a good thing, but it reads to me like you don't plan on a PV array any time soon. If so, and meant respectfully, why all the questions ? Are you curious for it's own sake ? Still investigating the viability of an array for your situation ? An academic looking for data to crunch for the publish or perish world ? What ?

On the rate of panel fouling, I provided information I feel is pertinent to a lot of your questions in my post. Like a lot of array fouling data mine is site specific. I also note now, that my data and findings are probably, IMO only, typical of So. CA rural or semi rural areas. I provided them as a guide or maybe a jump start for you in your own search.

I also noted that array fouling rates and types are very site and application dependent. There are no magic bullet formulas. Array fouling is a highly empirical subject that uses a lot of existing engineering and research techniques in the investigations.

With the cooperation of about 10 or 12 neighbors in my HOA, I informally monitor their array outputs as f(time,weather including precipitation). My no more than anecdotal results for those efforts point to some commonality in fouling rates as f(time, weather including precipitation), but the data has a lot more variation than I might have at first thought - maybe +/- 20% or so. My thinking is array orientation and probably very local wind characteristics may well have a lot to do with the variability of fouling rates over an area of a few square miles as this is a pretty hilly area. That said, my data for those neighbors is less than necessary for any statistical validity and nowhere near as extensive as the data I have for my array.

There are a lot of vetted and peer reviewed data in the open literature, including NREL and the journals as well as any number of textbooks that deal with the subject of array fouling and its mechanisms. I respectfully suggest that, from one numbers guy to another, a literature search can provide a lot of what you're looking for - more than any numbers guy could possibly hope for.

Good luck.

[azdave]

Has anyone quantified this stuff? I hear varying things about the impact on production from rain, dust, and pollen. Another issue that comes to mind is whether the glass in panels is durable. I'd expect that panels would be vulnerable to hail damage. I also wonder if, over time, panel glass gets cloudy or warped. How about condensation? Are these issues, and if so, are there numbers? I'm a numbers guy to the Nth degree.

Well, I'm not your numbers guy but, unless you are in an extreme situation that is ripe for severe fouling, worries about keeping the panels really clean are not worth it. If you think you will lose sleep over such small losses, then install an extra panel to two now (or buy higher output panels) and try to forget about it. I don't clean mine unless I must go up on the roof for some other important reason first. The small benefit I would get from making an extra trip up top a few times a year is not worth the risk I take each time I climb or descend that ladder. How much is a trip to Urgent care these days? Hiring someone else to clean my panels will never pay back either. Mother nature does the bulk of my cleaning chores for me, thanks.

As for the glass, the panels use tempered glass which is extremely resistant to hail and golf balls but still can shatter sometimes. It also has a special coated to help capture the most energy possible. The glass does not warp or get cloudy as it ages but panels can have moisture issues if a seals fail. All rare events in my opinion. You really should worry about all the other electronics involved way before you worry about the panels.

[bcroe]

I live across from a corn field, dust and pollen can be a problem for me
mowing the lawn. But in 9 years there is no accumulation on the 100+
PV panels, the dew and rain take care of that. The only thing ever removed
was dripping tree sap, took a sharp blade, but the 3 panels were restored
to compete recovery. Weather from -27F to 100F, constantly changing, has
not bothered them. Rain, snow, ice, and hail have yet to damage them.
Flight path, we are on the approach to a very high ranking freight transport
field. UPS freighters fly by all day and all night, but I see no impact here.

If poor manufacture or rough handling damage panel sealing, sure moisture
can get inside, failure is certain.

Snow may be somewhat obvious in retrospect, but it is hardly considered
in PV array construction here in the northern ILL snow belt. Arrays continue
to be built the same old, SW desert cost optimized way. Every KWH
harvested here will cost more than there, the question is how best to spend
that cost.

In my experience and opinion, snow impact is never considered before
construction, becoming more obvious later. One approach is write off all
production from times when snow remains. That is almost always the case
of a roof mount. My neighbor has an array held close to the ground by
concrete blocks. Since there is no room to walk between rows, it has no
better snow performance than a roof.

Lowest panel should be high enough that snow sliding off will not pile high
enough to block the panel. Leaving a half foot snow gap between rows of
panels will greatly reduce the height of each snow pile, and greatly reduce
the weight and distance cleared snow must be pushed. With landscape
mount it greatly encourages self cleaning. Of course a high angle for winter
helps a lot, if adjustable. good luck, Bruce Roe

[slinthicum]

I bought specific equipment to clean off the panels about 4 years ago, shortly after I bought the system. I did a single cleaning shortly after the equipment arrived. Shortly after that my wife ordered me to stay off the ladder (understandable since I'm in my '70s) so I gave it away. The panels have not been cleaned since that first cleaning. I'm noticing no downturn in production, except for those days here in SoCal where the early morning clouds roll in from the Pacific Ocean. I should have read this article before that equipment purchase.

[bcroe]

my wife ordered me to stay off the ladder

On a ladder? Listen to your wife. Bruce Roe

[J.P.M.]

I bought specific equipment to clean off the panels about 4 years ago, shortly after I bought the system. I did a single cleaning shortly after the equipment arrived. Shortly after that my wife ordered me to stay off the ladder (understandable since I'm in my '70s) so I gave it away. The panels have not been cleaned since that first cleaning. I'm noticing no downturn in production, except for those days here in SoCal where the early morning clouds roll in from the Pacific Ocean. I should have read this article before that equipment purchase.

I belonged to the parent society of the Journal of Solar Energy for ~ 20 years until 1995. I only saw that paper after my original research was done and results were on paper. The Patringenaru paper was published long after I'd given up my membership and access to the journal.

It was interesting (to me anyway) to note that the 7.4% figure the paper noted for an average performance loss to fouling after 21 weeks seems roughly in line with the 8 - 10 % figure for an asymptotic value I found after about 10-12 weeks.

I tried to call the author a couple of times to talk shop, but got no response.

As for paying to have panels cleaned, old timers around here may remember that was one of the wild hairs in my butt around here for several years, partly because of the asymptotic nature of the fouling, and the observation my me and others around here that a typical (at least in CA) array will most likely get dirty in a way that the loss of array revenue from dirt, even without rain, will most likely tend to a limit and, in effect, not get much worse, and so leveling off after a relatively short period for many, but I'd note certainly not all arrays. That makes cleaning an array not cost effective.

The analogy I used is: Does a skylight continue to get dirtier after a few months in service ?

I along with other here and elsewhere have noted the $$ spent of cleaning an array would often be better spent by oversizing the array by that 8 - 10% if there is room for the excess surface.

FWIW, my situation didn't allow for any excess area so I hose the array every month or so if it doesn't rain and live with about a 2 - 3 % output penalty due to fouling.

As for getting on a roof, I designed the array for easy access and servicing - including hosing from above - on a roof that's got < 20% slope as measured.

[slinthicum]

Better not let the water police (or neighbors) catch you here in SoCal.

[Ampster]

Better not let the water police (or neighbors) catch you here in ...

I have also heard thar dihydrogen monoxide is an essential ingredient of rain

[J.P.M.]

Better not let the water police (or neighbors) catch you here in SoCal.

I suppose.
I don't read/record my water meter readings daily like my electric meter, but my daily household potable water use has been about 25 gal./day , or about 100 ft.^3/month for as long as I've owned property. If the H2O police see 16 - 18 gal. of water or less from my property going down the gutter ~1X/month as array runoff, so be it. Hell, I probably produce more urine than that in a month. Compared to others, my gross water use is a pittance, but if restrictions really do hit in a blanket % reduction fashion, I'll get hit just like everyone else. As for the water police, I usually hose the array between 6 and 7 A.M. I've never seen the water cops out that early. As for my neighbors, they're mostly a bunch of retired and entitled old white folks, about as environmentally unaware as it's possible to get and belligerent as hell when asked (or especially hostile when required) to do something that may even slightly inconvenience them and their lifestyles. The mobile carwashes and car detailers thrive in my HOA. Rosebushes abound.

[nerdralph]

There are a few tall (about 20m) pines in a neighbor's yard, and I never noticed just how much they shed needles until I installed PV panels on my garage. With the amount of wind and rain we get here, it doesn't last long. I've only ever cleared snow off the garage panels. It's a nominal 4.4kW array, and on a sunny day covered with snow I was getting only 230-240W. After clearing off the snow, production was 2500W.