Posted by Ken Maltby on December 3, 2008, 8:16 pm
Hello Zwerius, all;
One year ago I had a 5 KW system installed (4.2 KW STC). I noted that
the panel specs indicated that there is a .5% per degree C loss. Under
full insolation (Nothern California), the temperature rise on my
panels is 25 to 35 C (roof mounted, 6" space on underside). Spraying
the panels with a garden hose (at mid day) produced the expected ~10%
increase in power output - for about 10 mnutes until the panels warmed
It appeared to me that it might be possible to squeeze another 10% out
of my system without too much trouble. At ~$ per watt installed, this
indicated that an expenditure of up to ~$000 (.1 x 4200 x $) might
Continously flooding or spraying the panels is not an option for me.
Water use would be excessive and my water rates are too high to make
My first inclination was to utilize evaporative cooling - spraying a
fine water mist (fog) into the space behind the panels. My
calculations indicated that the evaporation of approximately 50
gallons of water per day would be adequate to reduce the panel
temperatures by ~15 to 20 C. The objective was to have the applied
water fog evaporate completely and avoid the maintaince complications
of wet surfaces. The costs would be: the water cost, the equipment
(fog nozzles, distribution plumbing, and high pressure pump), power to
run the pump, and installation. The total system would cost
I performed a "proof of concept" test using a group of 6 panels (~ 6
square meters) and quickly found that it DID NOT work as expected.
There was simply not enough air exchange on the backside of the panels
to allow the water mist to completely evaporate. The air quickly
became saturated and the fog simply deposited on the backside of the
panels. The cooling effect was minimal, a few degrees at most.
Once I recognized my oversight, a quick calculation indicated that I
needed an air flow velocity of ~20 miles per hour on the back side of
the panels to provide enough air for the water mist to evaporate. And
this is in California where the relative humidity is relativley low.
The required air flow would be much greater in more humid locations.
Since the fog nozzles and instrumentation were already in place, I
installed a number of fans along the edge of the test array to
provide the necessary air flow. The cooling effect did increase, but
not by as much as desired, a total temperature drop of only 4 to 5
degrees C was achieved. A calculation of the power required for the
fans indicated that the majority, if not all the "extra" power from
the panels would be consumed by the fans. Just the fans alone helped
reduce the panel temperatures - but not enough to satisify their own
Bottom line is that you can increase the power harvested from your
solar panels by active cooling, but it is unlikely to be economical
(solar power itself has questionable economics).
If you have access to large quantities of cool, clean water (and don't
have to pump it very far) and can continously spray the panels, then
you might get a small net increase in power.
All in all, an interesting learning experience but not part of a
solution to global warming
How gulliable do you think we are?? I very much doubt
you did anything like what you described.
And you caculations really suck, 20 MPH wind?? To make
an evaproation scheme work? Why bother with the mist/fog,
in that case, just use a passive heat sink.
The classic method of watercooling silicon chips is to have a
thermal conduction path to a waterblock where the water/cooling
fluid is recirculated, by a pump, through a heat extraction device
[A radiator if the water/cooling fluid temp. is high enough, that
moving ambiant temp air will carry away the amount of heat to
meet the design needs. Or some form of chiller if lower temps
are needed/desired.]; then back through the waterblock.
It is unlikely that your panels were designed to allow for extracting
heat from the backside of the panels. It is even possible that they
would not take kindly to direct exposure to your mist or a spray of
water. (Although it might be easy to "waterproof" them for this
purpose.) In any case I doubt their design included a thermal
conduction path through to the backside of the panel, there could
even be a layer that provides some thermal insulation (intended or
otherwise). This might also be overcome with some ingeniuity, but
it may be something that needs to be addressed in the making of the
A real effort to do as you described, would have been of some
value, your less that creditable post just adds confusion.
Posted by mundt on December 4, 2008, 12:54 am
Hello Zwerius, Morris, Ken, et. al.
It's interesting to see so much interest in this topic - obviously
everyone wants to get the maximum out of their panels. I'll try to
respond to the questions and comments are they were raised.
Morris - Flowing or spraying ambient temperture water (hopefully 40C
or less) over solar panels should work fine. You can do a rough
calculation of how much water would be required by estimating the
amount of thermal energy needs to be removed. I used ~800 W/m^2) in
my calculations. Using the specific heat of water (~ 1 cal/C) and a
reasonable temperature rise for the water (10-20 C). You can figure
the necessary warter flow per m^2. This does not take into account any
heat removed by evaporation so it should be a pretty converative
This will require a fair flow. The water should not be allowed to
evaporate on the panels or the dissolved minerals in the water will
degrade the surface and lower the panel output over time. As you point
out a closed, recirculating system would make the most sense. This
should work provided several issues are addressed.
1. Energy cost - pumping around a fairly large volume of water can
take a fair amount of energy. If the directly powered pump you suggest
can do this, then a major problem is overcome.
2. Water quality - the water will need to be kept pretty clean. At a
minimum I would think that it needs to be deionized (or softened at
least) and the dissolved solids carefully controlled.
3. Distribution and collection system - maybe "soaker hose" and open
gutters is they are already in place?
4. Storage/cooling - a reasonable voulme of water would need to be
stored and provisions made for extracting the adsorbed heat. Maybe a
thermal siphon through a radiator would allow the water to be
precooled at night at little extra energy cost.
The real question is whether a system with this extra level of
complexity can be cost effective (original hardware, maintaince,
lifetime,etc.) My original concept was just the addition of a little
plumbing, a few fog nozzles, and a temperature activated valve. It
seemed like it would be simple (and cheap) enough to be doable. Too
bad it didn't work.
The (semi) closed loop system you propose could work - but I suspect
it would be difficult to do it in a way that makes economic sense. I
would be VERY interested in seeing a "back of the envelope"
calculation of the system elements and costs.
Zwerius - I would highly recommend that you think deeply about what
and how you are trying to do. Use a very sharp pencil to calculate the
costs and benefits. Figure out how to do a reasonably low cost "proof
of concept" prior to trying to scale up. Good luck.
Ken - I'm sorry that you disbelieve my experience, it IS accurate. I
am a retired silicon valley engineer with the time and toys to "play"
with interesting technical things. My solar system was such a thing.
The 20 mph flow velocity came from estimating the amount of heat
energy I needed to remove (see above) and then calculating how much
water woud need to be evaporated to absorb that energy (heat of
vaporization). I then calculated/estimated the quanity of air required
to "hold" this water (at 80% relative humidity I think). Given the
volume of air required (and the cross sectional area of the gap behind
my panels, ~6") I was able to estimate the flow velocity necessary.
If you were to use just the specific heat of the air (simple forced
convection) to remove the necessary energy (at a reasonable
temperature differential) you would need a LOT more air. I don't
remember calculating that case. Maybe you could do it for us? (20 C
delta, 800 W/m^2).
You are correct that my panels (Sharp ND-167U1F) are not specifically
designed for active cooling. The backside of the panels is a white
plastic (Tedlar) sheet encapsulating and sealing the individual cells
against the top glass. There is likey a shorter distance between the
back surface and the cell than between the top surface and the cell.
The relative thermal conductivity of the bottom plastic and the top
glass are likey about the same - obvioiusly neither are anywhere close
The good thing is that the energy density for the solar panel is in
the order of .1 W/cm^2 rather than the 10s of W/cm^2 produced in many
ICs. Thus the temperature differentials are much less. Trying to
couple a liquid cooled "waterblock" to a solar panel would be extreme
overkill - well beyond what I was trying to do, and unnecessary.
If you do a USPTO patent search on my name - Randall S. Mundt, you
will see that I hold a number of patents on the design of
electrostatic chucks - very high tech heat removal devices used in
semiconductor processing equipment. I know what I am talking about and
I did perform the experiments/evaluations I described. Maybe you can
explain in more detail why you find it "less than creditable".
Posted by Morris Dovey on December 5, 2008, 3:59 pm
I'm not an expert on this stuff, but intuitively flow seems much
preferable to spray in order to minimize mineral deposits on the glazing.
I don't see flow volume as a problem, since that will depend on the
capture area of the concentrator and the size/efficiency of the pump
engine. It appears to be a matching exercise. If the installation is
"large", then multiple concentrator/pump subsystems may be appropriate.
When sunlight is most intense (causing greatest panel heating) the pump
efficiency should be maximized - a happy coincidence.
This makes good sense. Perhaps we could entice some of the folks at
Culligan and/or Brita to consider a new potential market.
The open gutters seem pretty reasonable. Instead of a soaker hose I
envision a UV resistant pipe that's perforated in one quadrant or
half-quadrant, fed from a small standpipe.
We might do better to think of some way to extract and use the heat
that's accumulated - perhaps for a coupled DHW (sub)system...
Umm - ITYM hasn't worked /yet/ (The fat lady hasn't even begun to sing.) :)
With a directly solar-powered pump, the only real remaining problem will
be to controllably disable the system when it would produce ice.
Actually, with a suitable cold weather drain-down arrangement, we don't
even care if the pump runs without water to move because there's
esentially no operating cost for the pump.
Back of the envelope:
Trough collector and pump /costs/ should come in under US$00. To that
add the cost of mounting and tracking for the trough, and costs
associated with the "plumbing" (distribution pipe, guttering, filter, etc)
It appears that a reasonable guesstimate of the total cost might be in
the neighborhood of US$000 - and since your guesstimated payback point
was in the US$000 neighborhood, even if I'm off by 100% it's still
within an acceptable range. Mass production could probably significantly
lower the costs and provide an attractive profit margin. :)
DeSoto, Iowa USA
Posted by mundt on December 6, 2008, 5:35 am
A spray might produce more evaparative cooling and less of a
requirement for external cooling of the water. It might also allow any
minerals in the water to build up more rapidly - as your intuition
indicates. Minimizing (not allowing!) mineral deposits is a function
of water quality - not how it is applied.
I looked at the pump materials you had included in your first post. My
first impression is that this pumping method cannot develop enough
lift to be used in this application. I believe that a head of 5' would
be the absolute minimum (for a ground mounted system). My two story
roof mounted system would require ~30' of head.
The issue (as allways) will be cost. What will be the cost per gallon?
In addition to the evaporative loss, I suspect that there will need to
be some amount of water "flushed" to control the solids content.
No problem, just need to make sure that there is a method for insuring
a reasonably uniform distribution over reasonable lengths (30 ft or
so). This usually means large diameter pipes and/or small orifices.
This heat is very low grade - you can't effectively cool with hot
water - i.e. the water exiting the system must still be cooler than
the panels or you are not removing heat from them. You can't have your
cake and eat it too. This is why an "integrated" PV and solar water
heating make little sense. You either produce reasonably hot water
(and don't cool the panels) or you cool the panels with enough flow
and produce tepid water.
If there is danger of producing ice, then you don't need to worry
about keeping your panels cool. Does a pump without water run :-)
I think that you are really underestimating the potential costs and
1. The pump - a tracking trough design, controls, motors, bearings,
2. Water - treatment, monitoring, storage tank, etc.
3. Maintaince and operation - consumables, maintaince (those wet
panels might not last as long, or who knows the reduced thermal
cycling might make them last longer).
I'm not saying that active cooling can't be made to work but...
I would definitely NOT start with the direct solar pump. The surest
way to fail is to try to merge two development projects into one.
Active solar panel cooling is a project with several problems to be
solved, the direct solar pump has its own set of issues to be
Posted by Malcolm \"Mal\" Reynolds on December 6, 2008, 5:55 am
That well may be true, but since ground water is usually cooler than
"tepid", couldn't this tepid water be used to pre-heat water going into
dedicated solar water panels?
And yes, I realize it's a level of complexity.