Cleaning solar panels with hard water

[JoeS]

I live in Hawaii in an area with exceptionally hard water, in excess of 450 TDS. We also have a significant amount of red dust that collects on everything. I've been using the following approach to clean my panels every 4 months.

1. Start working after 3 PM.
2. Spray one panel with garden hose connected to my 320 PSI 0.8 GPM Cordless Battery Cold Water Power Cleaner.
3. Wash with detergent using a Swiffer and a Costco Ultra Plush Microfiber Towel (I change the towel when they get dirty).
4. Rinse with cordless power washer and garden water.
5. Remove as much water as possible in a squeegee.
6. Rinse with cordless power washer using 2 L bottles of RO filtered water which has a TDS less than 45.
7. Remove as much water as possible in a squeegee.

I get a big placebo effect increase in power after that, lol. I monitor daily with my Tesla app.

Notes:

• I have a 3,500 PSI gas power washer but that seems like too much work.
• I made two test baths out of black granite, making a big square with silicone. I dump unfiltered and filtered water in each. The difference is impressive.
• RO haters (like ChatGPT) keep in mind the waste water goes to my bananas grove. ChatGPT still scolds me for using RO but she approves of my approach for cleaning the panels.
• December of this year I was a net exporter to the grid. Other months I'm a much bigger exporter.

[J.P.M.]

1.) STOP USING THE PRESURE WASHER. The seals around the edges won't take it. Failures will result sooner than later. Check your panel warranty info and spend time interpreting the fine print from the mfg.'s perspective.
2.) After spending several years measuring and documenting my array's performance in more than a little detail, I've found that:
a.) Unless hard water spots can be "felt" and are VERY prevalent, they do not hamper my array's performance in any way I've been able to measure. That is, any performance decrement from hard water spots alone is probably below about 0.2% or so of clean performance.
Also, squeegeing the array made no difference in performance vs. simple drip drying that I could measure.
b.) The admittedly limited literature search I've done on what's common TDS stuff (mostly group IIA chemicals from the periodic table) has mostly convinced me that they are relatively transparent to solar radiation below a silicon cell's cutoff frequency of 1.15 microns or so. That maens that while they may look unsightly, they probably don't impair array performance as much as the human eye may lead us to believe.
c.) I'm not at your site, and while the rate and type of any array fouling is very site specific, I've found my array fouls at a rate such that it causes a production decrease of, on average, about 1% of clean performance per week without rain. That rate seems to mostly agree with published information from the literature for array fouling in SO. CA and the west coast.
d.) A decent rain ( ~ > 0.25" precip. around here) will restore about 3/4 to maybe 90% of my array's clean performance.
e.) Around here, the big, but fortunately no more than occasional bombing raids by the local buzzards and hawks, and the big owl skrocks get localized treatment on an as appearing/needed basis.
f.) Using distilled or deionized water vs. plain old tap water for a final rise after soap cleaning made no difference in clean performance that I was able to measure. Therefore, for my situation, using super clean water seems like a waste of effort and resources to me. Plain, old tap water is fit for purpose.
3.) I now hose my array every 4 weeks if it doesn't rain with simple tap water (spraying from the top of the array), let it air dry and accept an average 2% fouling penalty in my array's performance. That seems to restore most - maybe >90% - of my array's performance that's been lost to dust and dirt. The hard water spots be damned.
I believe all that's a lot easier on the array and it takes maybe 20 minutes.
4.) 1X/yr. about June, I clean my array with a soft brush, dish soap (Dawn) and a double rinse followed by a vinegar douche and then more water. Takes about 90 minutes or so. FWIW, the vinegar removes the spots just fine.

5.) Do what you wish, but you may want to consider less vigorous cleaning methods or at least find a way to get some meaningful measurements of your array's performance and the penalties associated with fouling before you kill your array to save its performance.
At least knock off the power washing.
6.) I've found rinsing/cleaning my array is better if done in the early A.M. before the sun has been on the array for more than say, 1/2 hr. or so as it results in less thermal stresses on the array's glazing and seals. The panels are still pretty hot around here in the afternoon and the possible/likely thermal shock is best avoided .

Most of the above is a summation of posts I've written that describe my array's performance and measurements I've done, as well as my experiments and measurements of array fouling and cleaning methods. Check those posts if interested.

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

[JoeS]

1.) STOP USING THE PRESURE WASHER. The seals around the edges won't take it. Failures will result sooner than later. Check your panel warranty info and spend time interpreting the fine print from the mfg.'s perspective.
2.) After spending several years measuring and documenting my array's performance in more than a little detail, I've found that:
a.) Unless hard water spots can be "felt" and are VERY prevalent, they do not hamper my array's performance in any way I've been able to measure. That is, any performance decrement from hard water spots alone is probably below about 0.2% or so of clean performance.
Also, squeegeing the array made no difference in performance vs. simple drip drying that I could measure.
b.) The admittedly limited literature search I've done on what's common TDS stuff (mostly group IIA chemicals from the periodic table) has mostly convinced me that they are relatively transparent to solar radiation below a silicon cell's cutoff frequency of 1.15 microns or so. That maens that while they may look unsightly, they probably don't impair array performance as much as the human eye may lead us to believe.
c.) I'm not at your site, and while the rate and type of any array fouling is very site specific, I've found my array fouls at a rate such that it causes a production decrease of, on average, about 1% of clean performance per week without rain. That rate seems to mostly agree with published information from the literature for array fouling in SO. CA and the west coast.
d.) A decent rain ( ~ > 0.25" precip. around here) will restore about 3/4 to maybe 90% of my array's clean performance.
e.) Around here, the big, but fortunately no more than occasional bombing raids by the local buzzards and hawks, and the big owl skrocks get localized treatment on an as appearing/needed basis.
f.) Using distilled or deionized water vs. plain old tap water for a final rise after soap cleaning made no difference in clean performance that I was able to measure. Therefore, for my situation, using super clean water seems like a waste of effort and resources to me. Plain, old tap water is fit for purpose.
3.) I now hose my array every 4 weeks if it doesn't rain with simple tap water (spraying from the top of the array), let it air dry and accept an average 2% fouling penalty in my array's performance. That seems to restore most - maybe >90% - of my array's performance that's been lost to dust and dirt. The hard water spots be damned.
I believe all that's a lot easier on the array and it takes maybe 20 minutes.
4.) 1X/yr. about June, I clean my array with a soft brush, dish soap (Dawn) and a double rinse followed by a vinegar douche and then more water. Takes about 90 minutes or so. FWIW, the vinegar removes the spots just fine.

5.) Do what you wish, but you may want to consider less vigorous cleaning methods or at least find a way to get some meaningful measurements of your array's performance and the penalties associated with fouling before you kill your array to save its performance.
At least knock off the power washing.
6.) I've found rinsing/cleaning my array is better if done in the early A.M. before the sun has been on the array for more than say, 1/2 hr. or so as it results in less thermal stresses on the array's glazing and seals. The panels are still pretty hot around here in the afternoon and the possible/likely thermal shock is best avoided .

Most of the above is a summation of posts I've written that describe my array's performance and measurements I've done, as well as my experiments and measurements of array fouling and cleaning methods. Check those posts if interested.

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

Thanks for the detail response. I think you over estimate the PSI from an 18V pressure washer. I can clean my legs with it it's so low pressure, like a very stiff wind.

I think you also underestimate the TDS in Hawaii water. I can scrape 1/6" inch per month from the toilet that only two folks are using.
We get very little rain, sometimes 9 months of zero rain.

Using distilled or deionized water vs. plain old tap water for a final rise after soap cleaning made no difference in clean performance

I'm not using it for cleaning, I'm using it to flush off the hard water. If I was my car and don't dry it off, it has HUGE water spots. It's amazing the amount of minerals in the water here.

Also, squeegeing the array made no difference in performance vs. simple drip drying

Again, that's to prevent mineralization, to pull off as much high mineral water as possible.

Unless hard water spots can be "felt" and are VERY prevalent

I tried to upload a image of my toilet so you can see the heavy mineralization from just a few weeks of not cleaning, but couldn't figure out how to upload. Kept getting server error.

vinegar helps until you get too much mineralization.

I think I'll skip the squeegee and RO rinse on my least productive panel for a couple years and see if I can detect a difference.

Because I'm a huge exporter to the grid and my credits expire 12 months after accruing, the cleaning is more for fun and OCD. We do plan to get a new EV in a year that has triple the kWh tank and do longer trips.

Much Mahalo for your insights.

[Calsun]

We use a window washing scrubber and squeege made by Unger who makes products for pro window cleaners. We only use tap water but if we had residue remaining I would add a little ammonia and liquid detergent to break through the film.

The pressure washer output depends on the nozzle that is used and how wide an angle it sprays. The wider the angle the less pressure taht is hitting the surface being cleaned. There are also two nozzles made specifically for cleaning panels on a roof with a pressure washer - Simpson Cleaning 80183 Nozzles

[JoeS]

Aloha all

We hosted 3 Maui fire displaced ppl and their 3 pets. We found them housing and they are gone now. I imported a little power with the extra load. We put them up in our trailer and the AC was running most of the time.

The 18V Ryobi Cold Water Power Cleaner shouldn't be thought of as a pressure washer. I use it at the beach to clean the sand off my feet. It's a huge help with 0 risk to the panel seals. The 40V Ryobi chain saw is over twice as powerful as my 18V, so I was going to upgrade to the 40 V version of the pressure cleaner, but Ryobi claims they have identical performance. I think that's a lazy copy/paste mistake by marketing so I may get one and test it. If performance is the same, I'll return it.

You've never met hard water like in Hawaii. Toilets here typically get 1/8" reddish brown layer added per year.

> Using distilled or deionized water vs. plain old tap water for a final rise after soap cleaning made no difference in clean performance that I was able to measure.

I stopped doing that. It was too much work. The squeegee removes the bulk of the water and what's remaining evaporates before I can rinse it off.

> A decent rain ( ~ > 0.25" precip. around here) will restore about 3/4 to maybe 90% of my array's clean performance.

I cleaned my panels a couple weeks ago. 2 days later we had a very light sprinkle with 20 MPH winds. The air was filled with dust and my panels looked like they hadn't been cleaned in 6 months. I had to clean them again. Bad timing.

> 6.) I've found rinsing/cleaning my array is better if done in the early A.M. before the sun has been on the array for more than say, 1/2 hr. or so as it results in less thermal stresses on the array's glazing and seals.

Great tip. I've changed from evening to morning cleaning. In addition to your cited advantages, where I live, the wind usually doesn't blow in the morning and almost always blows in the evening. Mineralization is my biggest risk and the wind really accelerated evaporation.


> 4.) 1X/yr. about June, I clean my array with a soft brush, dish soap (Dawn) and a double rinse followed by a vinegar douche and then more water. Takes about 90 minutes or so. FWIW, the vinegar removes the spots just fine.

That's a great tip. I bought a 1.25 gallon bottle from Costco. Costco used to sell 5 gallon buckets of vinegar but I haven't seen them for a couple years. I can spray my panels with my pressure washer and the vinegar in a 2L soda bottle. As you know, the vinegar will dissolve some of the mineralization. I think I'll hose the cleaned panel down with vinegar, leave it for a couple minutes, slight 2nd spray, then use my window scrubber.

[foxxxnick]

Generally speaking, various cleaning technologies commonly used in the market can be divided into three macro areas: manual cleaning, machine cleaning and robot cleaning. In these areas, the cleaning process also varies greatly in the use of detergents, water filters, instruments, and equipment.

The most common way to clean solar energy is to use a resin filter, which desalts the water to avoid residue. Some suppliers also use special cleaning methods. When using a detergent, make sure it is approved by the module manufacturer and is environmentally friendly.

[JoeS]

Generally speaking, various cleaning technologies commonly used in the market can be divided into three macro areas: manual cleaning, machine cleaning and robot cleaning. In these areas, the cleaning process also varies greatly in the use of detergents, water filters, instruments, and equipment.

The most common way to clean solar energy is to use a resin filter, which desalts the water to avoid residue. Some suppliers also use special cleaning methods. When using a detergent, make sure it is approved by the module manufacturer and is environmentally friendly.

This must have been generated by a nonsense bot. You don't clean solar energy. There is no such thing as a resin filter to remove salt. There is no salt in potable water and if there was, desalination is a very expensive process.

[JoeS]

Most of the above is a summation of posts I've written that describe my array's performance and measurements I've done, as well as my experiments and measurements of array fouling and cleaning methods.

array's performance and measurements

I forgot to mention Calsun is right on about performance but I noticed while you might get a few % improvement at noon, in the early morning and evening it's far greater. After cleaning my panels I get 100 watts showing up 45 minutes earlier and later in the evening. The % increase is far bigger when the angle of sunlight hitting the solar panels is more oblique as the light has to pass through more dirt.

[J.P.M.]

Generally speaking, various cleaning technologies commonly used in the market can be divided into three macro areas: manual cleaning, machine cleaning and robot cleaning. In these areas, the cleaning process also varies greatly in the use of detergents, water filters, instruments, and equipment.

The most common way to clean solar energy is to use a resin filter, which desalts the water to avoid residue. Some suppliers also use special cleaning methods. When using a detergent, make sure it is approved by the module manufacturer and is environmentally friendly.

All those special cleaning methods and potions are IMO only, are ripoff B.S. (but I'd respectfully submit that's an opinion that's more informed from an academic and practical standpoint than most other's).
From what I've done and what I've learned from it as well as some prior experience in an unrelated field, solar array fouling and array fouling rates are quite local and varied.

While some situations/applications are less favorable, and so a tougher clean, my bet is most residential applications are pretty standard, mostly dust and some guano.
For those applications that are tougher service - such as near sappy trees, seashores or arrays with low slopes or other situations, some more aggressive methods and/or more frequent cleaning may be appropriate.

However, and regardless of the cleaning method, - aggressive or not - an array will still get fouled and usually at a rate such that array performance rolls off at something like maybe - and regardless of the cleaning method - between 0.5% and maybe 1+%/week or so. At some point, the performance deterioration will probably level off and become somewhat constant in an asymptotic way until it rains or until the array gets cleaned. What that deterioration will be is again very site dependent.
FWIW and as an example only, I believe my array's fouling levels off at something like maybe a 10% or so performance loss. I suspect other regions/areas are worse (deserts with near horizontal arrays for example), and some are better (like places that get a lot of rain for example).

For the common residential application, Hosing an array down as I've described with plain old tap water at a rate of something like 3l/panel will restore about 2/3 to maybe 90% of the lost performance. Doing so on a regular 1x/month basis or so if it doesn't rain will take about 20 minutes and is a lot less onerous that all the B.S. methods that don't get that much more dirt (and which will still come back at the same rate regardless of how involved and expensive the cleaning method is).

I've cleaned my array a lot. It's a real PITA. Along the way I believe I've learned a lot in the way of about what works and what is a ripoff and what is achievable in a practical sense that is a balance between pristine, and practical and workable.
I hose my array with tap water 1X/month if it doesn't rain and accept an approximate 3% performance degradation due to array fouling.

[J.P.M.]

array's performance and measurements

I forgot to mention Calsun is right on about performance but I noticed while you might get a few % improvement at noon, in the early morning and evening it's far greater. After cleaning my panels I get 100 watts showing up 45 minutes earlier and later in the evening. The % increase is far bigger when the angle of sunlight hitting the solar panels is more oblique as the light has to pass through more dirt.

It's probably close to a certainty that the fouling penalty will be a greater percentage of clean performance at low solar incidence angles than higher solar incidence angles as is common in early morning and late(r) afternoon, but depending on the nature and characteristics of the fouling layer itself - including, as you suggest, to some extent on the thickness of the fouling layer, it's most likely due to a combination of the thickness of the fouling layer and perhaps even more a function of the reflectivity of the fouling layer as f(incidence angle) of the irradiance.

To maybe keep things in perspective a bit, the performance of a panel is a reasonably strong function of the angle of the irradiance for 2 big reasons. One is that the Plane Of array (P.O.A.) irradiance is f(cos (incidence angle)) of the Beam Normal Irradiance.
The other is that the reflectivity of the panel surface is a somewhat strong function of not only the panel glazing reflectivity but also as f(the angular dependence of the fouling layer's reflectivity) to wavelengths in the solar spectrum that are of use to a silicon PV cell.

So at low solar incidence angles like those common in the hours when mostly equator facing arrays not only see less P.O.A irradiance available due to the angle of incidence, but more of that is also lost due to a higher percentage of that lower irradiance being reflected away from the surface.

Overall, the amount of reflectivity from a surface will increase at a greater rate (as f(incidence angle)) than the P.O.A. irradiance will decrease with both of those rates of change being nonlinear functions of solar incidence angle and also different from one another perhaps (in the case of a fouled PV panel) also due to the variable reflectivity of the fouling layer as a function of its composition and physical characteristics.

One other thing that may be contributing to a later in the day differential of clean to dirty for panels' output that I hadn't investigated, considered (until now) or even read about in my solar odessey might be partly due to the nature of the diffuse portion of the irradiance and the idea that it tends to stay more constant over the day (and so, be a larger portion of the total P.O.A. irradiance farther away from solar noon - even while the P.O.A. is lower), and also as well as maybe due to the nature of the diffuse irradiance - it's reflectivity is a generally considered to have be more/less constant as f(incidence angle) as well as the diffuse portion of the irradiance being a combination of shorter wavelengths (Google "Rayleigh Scattering" for an explanation of why if you're interested) and so more readily absorbed by silicon PV cells.