Water cooled solar panels for significant output boost

[russ]

Fair enough, but water used in this way will never pull the temp down by 20 deg C. Here is an example of a misting system using a whole lot more water that achieved 2 deg of cooling. Any water that recirculates will rise in temperature without an additional heat exchanger to cool it back down, further defeating the cooling power of the system.

Not nice to confuse them with facts - conjecture is so much more fun!

[FishGun]

Sensij:
Yeah and to be honest I didn't pore over his numbers...didn't watch the vid...just took a quick swag at the delta and your point about the .4-.5% loss. It fits. I probably shouldn't have commented.

But beyond the exact details of his experiment, I think there are reasonable gains waiting to be squeezed-out by a good idea.

[Ian S]

Fair enough, but water used in this way will never pull the temp down by 20 deg C. Here is an example of a misting system using a whole lot more water that achieved 2 deg of cooling.

Here in the Arizona desert, evaporative coolers in June's dry heat can reduce the air temperature by 20C so it's not beyond the realm of possibility.

[russ]

I think there are reasonable gains waiting to be squeezed-out by a good idea.

That is the catch - the "good idea" part - it isn't here. This is old hat looked at by many, many bright people. It is a .

[Sunking]

Here in the Arizona desert, evaporative coolers in June's dry heat can reduce the air temperature by 20C so it's not beyond the realm of possibility.

You are correct when in the shade and bone dry air. Will not have any great effect on surface temps in direct sun light. This road has been traveled already by many bright minds. To date no one has come up with anything that has an energy net gain. One of the reasons Solar and Hydrogen are not fuels.

[bouncintigger]

I believe the O.P. is talking ambient temperature of 35C not panel temperature. Panel temperature can easily run 20 C deg higher under full mid day sun and calm wind conditions.

On the 5th of January, here in Perth when ambient was 44°C, I wasn't running the system, but I checked the thermocouple and it was sitting on 78°C.

[bouncintigger]

Fair enough, but water used in this way will never pull the temp down by 20 deg C. Here is an example of a misting system using a whole lot more water that achieved 2 deg of cooling. Any water that recirculates will rise in temperature without an additional heat exchanger to cool it back down, further defeating the cooling power of the system.

Evaporation is the cooling method. At steady state, the recirculated water reaches a peak temperature, the panels are at 35-38°C and the slight evaporation keeps the temperature in balance.

[bouncintigger]

That is the catch - the "good idea" part - it isn't here. This is old hat looked at by many, many bright people. It is a .

Well, I suppose I'm yet another bright person to look at it, and as far as I can tell, it appears to be working.

The parasitic power is < 5% of the average gain. The water consumption is 12c per day and it's clear as mud from the datalogging over several weeks of turning the system on and off, that the gain is ~ 3-4 kWh per day. Granted, the water filter cost was $300, but no single item in the rest of the setup cost more than $25.

You're really good at telling people how useless they are, but you're not particularly good at providing facts to back up your negativity or poking holes in the science. Care to give it a try?

[sensij]

Well, I suppose I'm yet another bright person to look at it, and as far as I can tell, it appears to be working.

The parasitic power is < 5% of the average gain. The water consumption is 12c per day and it's clear as mud from the datalogging over several weeks of turning the system on and off, that the gain is ~ 3-4 kWh per day. Granted, the water filter cost was $300, but no single item in the rest of the setup cost more than $25.

You're really good at telling people how useless they are, but you're not particularly good at providing facts to back up your negativity or poking holes in the science. Care to give it a try?

The most compelling evidence I think you could provide would be a pair of panels, side by side, one with the cooling system and one without. Collect data for a few days or weeks to establish a correlation between the panels, then see what happens to the correlation when the cooling is applied to one but not the other. If you don't have individual panel monitoring, making two strings of panels and comparing them in a similar way might work too.

[pleppik]

The most compelling evidence I think you could provide would be a pair of panels, side by side, one with the cooling system and one without. Collect data for a few days or weeks to establish a correlation between the panels, then see what happens to the correlation when the cooling is applied to one but not the other. If you don't have individual panel monitoring, making two strings of panels and comparing them in a similar way might work too.

What he said. Given the variability of solar power over time, a convincing test requires a side-by-side test rather than comparing production on the same panel at different times. Do you have microinverters on your array? If so, then you can monitor output at the panel level, and do a side-by-side test by turning off half your sprayers.

Bouncintigger, props for experimenting and learning, and double-props for posting your ideas in a public forum and letting other people poke holes in your ideas.

But I'm skeptical that this is working as well as you think it is, for a bunch of reasons:

1. There's embedded cost and power usage you're not taking into account. For example, 30 liters per hour is a LOT of water usage. A quick back-of-the-envelope calculation says that if I were to try something like this with my utility rates, the cost of the tap water would be about half the value of the extra electricity. So there goes 50% of the value right off the top. You lose another 8% to 10% from the circulation pump. That's a lot of the gain to give back before we even start considering the costs of things like replacing the filter media, broken water pumps, etc.
2. In many places, fresh water is a limited and valuable resource. You can look at this system as a way to convert fresh water into electricity. In some places that's a bad tradeoff. I'm guessing that Australia, like California, is one of those places.
3. It's not clear what the long-term effect of this will be on your panels. You already had problems with scale buildup when you used unsoftened water. Will you need to clean the panels regularly even with the treated water? What effect will this have on the life of the PV panels and energy output?
4. As others have pointed out, the effect of cooling on solar output is well-known, and a lot of smart people have explored ways to cost-effectively cool solar panels. Usually they've found that the energy gained isn't worth the extra cost and complexity of the cooling system. I'm not seeing how this scheme is all that different from other approaches which have been tried.
5. I'm not convinced that your system would increase output enough to make it with the cost of installation, water, and all the extra maintenance that comes with pumps and plumbing. For a system expected to pay back in 5-10 years and last 25+ years, even a small amount of additional maintenance can destroy the ROI. Spray-cooling does not strike me as a low-maintenance approach.

But whether it turns out to make economic sense or not, I'm glad you posted your data here. Worst case scenario, you turn it off and we all learned something.

[bouncintigger]

The most compelling evidence I think you could provide would be a pair of panels, side by side, one with the cooling system and one without. Collect data for a few days or weeks to establish a correlation between the panels, then see what happens to the correlation when the cooling is applied to one but not the other. If you don't have individual panel monitoring, making two strings of panels and comparing them in a similar way might work too.

Totally agree, but unfortunately I don't have individual panel monitoring. I have separate monitoring on both sides of my house, however they're not at the same orientation. Instead, what I plan to do is to benchmark against other available data for systems within < 10km of my house.

All that said, it's clear as mud that when the system is hot and I switch on the cooling, the peak goes through the roof. If I keep the system running, the peak follows an elevated bell curve that peaks around 4.7 kW without fail. When I switch it off, the bell curve drops back to the typical "hot day" bell curve that peaks at ~ 4-4.2 kW.

Here's another graph showing two cooling system "events" on an otherwise sunny, hot day... I have plenty more.... I can understand the level of scepticism, but at some point the evidence speaks for itself.

[sensij]

You know that "clear as mud" means "not clear at all" right?

The attachment is coming up invalid. Maybe try again?

[bouncintigger]

You know that "clear as mud" means "not clear at all" right?

The attachment is coming up invalid. Maybe try again?

Heh "clear as day" I believe I meant

Strange - the attachment links works for me when I click on it. Perhaps it's finding a local copy... Let's try again:

Peaks 2.jpg

[bouncintigger]

What he said. Given the variability of solar power over time, a convincing test requires a side-by-side test rather than comparing production on the same panel at different times. Do you have microinverters on your array? If so, then you can monitor output at the panel level, and do a side-by-side test by turning off half your sprayers.

Bouncintigger, props for experimenting and learning, and double-props for posting your ideas in a public forum and letting other people poke holes in your ideas.

But I'm skeptical that this is working as well as you think it is, for a bunch of reasons:

1. There's embedded cost and power usage you're not taking into account. For example, 30 liters per hour is a LOT of water usage. A quick back-of-the-envelope calculation says that if I were to try something like this with my utility rates, the cost of the tap water would be about half the value of the extra electricity. So there goes 50% of the value right off the top. You lose another 8% to 10% from the circulation pump. That's a lot of the gain to give back before we even start considering the costs of things like replacing the filter media, broken water pumps, etc.
2. In many places, fresh water is a limited and valuable resource. You can look at this system as a way to convert fresh water into electricity. In some places that's a bad tradeoff. I'm guessing that Australia, like California, is one of those places.
3. It's not clear what the long-term effect of this will be on your panels. You already had problems with scale buildup when you used unsoftened water. Will you need to clean the panels regularly even with the treated water? What effect will this have on the life of the PV panels and energy output?
4. As others have pointed out, the effect of cooling on solar output is well-known, and a lot of smart people have explored ways to cost-effectively cool solar panels. Usually they've found that the energy gained isn't worth the extra cost and complexity of the cooling system. I'm not seeing how this scheme is all that different from other approaches which have been tried.
5. I'm not convinced that your system would increase output enough to make it with the cost of installation, water, and all the extra maintenance that comes with pumps and plumbing. For a system expected to pay back in 5-10 years and last 25+ years, even a small amount of additional maintenance can destroy the ROI. Spray-cooling does not strike me as a low-maintenance approach.

But whether it turns out to make economic sense or not, I'm glad you posted your data here. Worst case scenario, you turn it off and we all learned something.

Thanks for this reply and thanks for the compliments and constructive criticism.

I agree with you that there's a risk the long term impact on the panels will outweigh the gains and I agree with you that the water consumption is high, however this is only when the system is working hard and I expect to be able to improve on this figure considerably with more tinkering to avoid losses onto the tiles. The long term impact is as you say easy to avoid, as if it becomes apparent that the scale can't be managed economically, I can switch it off and cut my losses. That said, I haven't regenerated my resin in acid yet, and given the composition of our water, I anticipate that a simple acid backwash will vastly improve the minimal scale I currently have.

The ~ 200 ppm Cl in the water can be tasted (as NaCl) in areas near the bottom of the panels where channelling has caused streams of water to evaporate to dryness. I plan on using Rain-X on the panels after a good clean, to minimise channelling and hopefully avoid localised regions of evaporation to dryness. Provided the water keeps flowing everywhere, the salts should stay well below their solubility limits and shouldn't be able to crystallise out. The resin acid regeneration option may do more harm than good: the ~ 200 ppm Cl will become HCl, although this is pH ~ 2.3 so isn't much different from vinegar in acidity, although the chlorides aren't pleasant! I'll make sure the pH is acceptable before putting anything crazy on my roof. If that fails, the next step could be mixed cation / anion exchange with a mixed bed resin, but regeneration would be more problematic and the cost of resin replacement would likely be too high.

I won't do anything stupid... I'm keeping an eye on the system to make sure it's going smoothly and there isn't any sign of corrosion. I'm slowly getting a handle on the best way to minimise scaling and am confident it will work out.

I must admit that the main point of the exercise is the challenge. Long term I'm hopeful it will work, though I may well lose out but I'll have fun along the way.

[russ]

1) Rain-x does not suggest to use their product on solar panels

2) Have fun on your fools game - People that know and understand what they are doing and understand how to test have tried it and learned it is useless.