Build it Solar Collector- General Discussion
amdx
I've been looking over the information on Gary's webpages. Thanks very much
for the side by side comparison
of the different type collectors.
http://www.builditsolar.com/Experimental/PEXCollector/SmallPanelTests.htm#Results
Also thanks for the detailed construction information.
http://www.builditsolar.com/Experimental/CopperAlumCollector/CopperAlumCol.htm
http://www.builditsolar.com/Experimental/PEXCollector/PEXCollector.htm
I'm throwing out a bunch of questions in an effort to learn and get the
collective knowledge to see if there is room to improve
the cost and efficiency of your collector design. I'm sure Gary has pondered
many of my questions!
I like the 33% reduction in cost for the PEX unit, but I'm in Florida and
think the heat could be a problem, so I'll stick with
a copper unit.
The bottom line is economics, cost versus performance. You have hit on a
model that does 96% of your all copper unit.
I'm impressed!
Are there any numbers to compare the theoretical maximum solar energy
available to what the best commercial model can extract?
Or to rephrase, What is the best efficiency obtainable versus your
Copper-Aluminum model?
Is there any room for efficiency improvement?
If not then cost would be the place to look for improvement.
I'm curious about how thick does the aluminum need to be to carry the heat
to the copper tubing?
How many copper tubes (of what diameter) is required to take the heat away.
You hit on 0.018" thick aluminum and 7 runs of 1/2" copper pipe.
I wonder, would 0.010" aluminum do as well?
Would six runs (or five) of copper pipe do as well?
Would you need the thicker aluminum if you used six runs (or five) ?
Would a better thermal compound between the copper and aluminum be worth the
cost?
Looking for everyones 2 cents.
Mike
Getting excited about solar energy!
Jeff
fdo...@gmail.com
mounted out of the tank?
Cosmopolite
> I'm curious why a submersible pump is used. Why not use a pump
> mounted out of the tank?
Good suction ? No seals to worry about? cost?
My opinon only.
ga...@builditsolar.com
> I'm curious why a submersible pump is used. Why not use a pump
> mounted out of the tank?
Hi,
I think either works fine.
The submersible pump is actually a little easier to install, and
requires less plumbing.
With the EPDM lined tank, its better if possible to not make
penetrations through the EPDM, and the pumps have to be mounted below
the tank waterline to retain prime after a drainback. So, if you
mount the pump outside the tank, you have to use an inverted U-tube
arrangement. Where the U-tube starts inside the tank near the bottom,
then goes up over the edge, then back down the outside of the tank to
a place below the tank waterline where the pump is mounted. Not a big
deal, but somewhat more work.
I'm going to be trying a new pump that is intended for water cooling
of PC CPU's. Its a Swiftech made by Laing. Supposed to be 50000 hr
MTBF with 140F working temperature capability -- runs on 8 watts with
13 ft startup head, and about 2 gpm flow rate for $55.
Gary
ga...@builditsolar.com
I'd also like to hear any comments/suggestions on the two collectors,
but since there has not been much, I'll put in my 2 cents ...
...
> The bottom line is economics, cost versus performance. You have hit on a
> model that does 96% of your all copper unit.
> I'm impressed!
> Are there any numbers to compare the theoretical maximum solar energy
> available to what the best commercial model can extract?
The efficiency of all of these collectors depend on the solar
radiation level, and on the temperature difference between the fluid
going through the collector (storage temperature) and ambient
temperature -- so there is an efficiency curve, not a single point.
The SRCC website has certification data for many dozens of collectors
they have tested that you can download, and this includes the
efficiency curve parameters for each collector.
Most commercial flat plate collectors are similar in efficiency. For
practical purposes, they have an efficiency under normal collection
conditions in the 50 to 60% area. As it turns out, evac tube
collectors are roughly the same.
I suppose there are always more things that can be done to improve
this, and there is no fundamental Physics reason you can't approach
100%. But, very little has changed in the efficiency levels of these
collectors since the 80's, so I am thinking that further improvements
are not currently cost effective.
The one feature that is included in many of the commercial collectors
is a selective coating on the fins. This might improve the efficiency
by about 5 percentage points compared to a non-selective coated fin.
There are some DIY selective coatings, so you can get some of that
back if you want to, but they are not as effective as the commercially
applied coatings.
> Or to rephrase, What is the best efficiency obtainable versus your
> Copper-Aluminum model?
> Is there any room for efficiency improvement?
> If not then cost would be the place to look for improvement.
I think so.
>
> I'm curious about how thick does the aluminum need to be to carry the heat
> to the copper tubing?
This depends on the width of the fin.
Some of the solar reference books define a "fin efficiency" that is a
measure of how well a particular width, thickness, and material fin
does compared to a perfect fin.
I'll see if I can dig this up for the 0.018 fins I use, but if I'm
recalling correctly the fin efficiency for 6 inch wide by 0.018 thick
fins is good.
> How many copper tubes (of what diameter) is required to take the heat away.
> You hit on 0.018" thick aluminum and 7 runs of 1/2" copper pipe.
The diameter is not much of a practical factor -- a lot of commercial
collectors use 3/8 inch risers, and they work fine.
Its the fin width and thickness that limit performance.
> I wonder, would 0.010" aluminum do as well?
Well, it won't do as well, but not sure how much of a hit you take.
> Would six runs (or five) of copper pipe do as well?
No, the 6 inch spacing is already on the high side.
> Would you need the thicker aluminum if you used six runs (or five) ?
yes -- very thick.
This is also bad from the point of view that as you make the fin
wider, you are trying to push more heat into each inch of the copper
tube through a contact area that is not increasing as you increase the
fin width. Since this thermal connection of the fin to the copper is
very important in the performance, this is a bad thing to do.
> Would a better thermal compound between the copper and aluminum be worth the
> cost?
If the tests are right, you are really looking to get that 4% worse
performance for this homemade version improved -- it seems like there
is not a lot of room for improvement?
The other thing to consider is that silicone is a very good material
to use inside collectors -- it has a high temperature capability, its
glues the parts together in a way that will still accommodate some
differential thermal expansion, and it has a good track record for
providing a long life in this tough environment. Not that the other
compounds might not do just as well or better, but it seems to me
there is some risk. There are some silicone caulks that are filled
for higher thermal conductivity, and this might be a good thing to try
-- I've not found a good source for these yet.
Gary
amdx
<ga...@builditsolar.com> wrote in message
news:a4a80816-b642-4080...@m73g2000hsh.googlegroups.com...
Hi Mike,
I'd also like to hear any comments/suggestions on the two collectors,
but since there has not been much, I'll put in my 2 cents ...
...
> The bottom line is economics, cost versus performance. You have hit on a
> model that does 96% of your all copper unit.
> I'm impressed!
> Are there any numbers to compare the theoretical maximum solar energy
> available to what the best commercial model can extract?
>>The efficiency of all of these collectors depend on the solar
>>radiation level, and on the temperature difference between the fluid
>>going through the collector (storage temperature) and ambient
>>temperature -- so there is an efficiency curve, not a single point.
> >so there is an efficiency curve, not a single point.
Do you think there would be any advantage to a variable speed pump that
would adust its flow rate to keep some (unknown to me) minimum temperature
differential between the bottom and top of the collector?
>The SRCC website has certification data for many dozens of collectors
>they have tested that you can download, and this includes the
>efficiency curve parameters for each collector.
I'll look for the SRCC website.
>> I'm curious about how thick does the aluminum need to be to carry the
>> heat to the copper tubing?
>This depends on the width of the fin.
>Some of the solar reference books define a "fin efficiency" that is a
>measure of how well a particular width, thickness, and material fin
>does compared to a perfect fin.
>I'll see if I can dig this up for the 0.018 fins I use, but if I'm
>recalling correctly the fin efficiency for 6 inch wide by 0.018 thick
>fins is good.
>> How many copper tubes (of what diameter) is required to take the heat
>> away.
>> You hit on 0.018" thick aluminum and 7 runs of 1/2" copper pipe.
>The diameter is not much of a practical factor -- a lot of commercial
>collectors use 3/8 inch risers, and they work fine.
Regarding 3/8" vs 1/2" copper pipe. I have to wonder about about the pipe
diameter.
1/2" pipe has 25% more surface area to absorb the collector heat and 25%
more boundary area to get that heat into the water.
>Its the fin width and thickness that limit performance.
So, do you think the aluminum to copper to water heat path is not a
bottleneck or limiting area?
Sorry to ask so many questions, at one time I worked with a physisist and I
was suprised how often
we worked near the edge of one parameter or another.
>> I wonder, would 0.010" aluminum do as well?
>Well, it won't do as well, but not sure how much of a hit you take.
Probably now a moot point for me, I was at my local metal scrap yard
yesterday
and found hundreds of 0.022" x approx. 2' x 3' aluminum sheets. I think they
might be used on
some type of rolling printing press.
>> Would six runs (or five) of copper pipe do as well?
>No, the 6 inch spacing is already on the high side.
>> Would a better thermal compound between the copper and aluminum be worth
>> the
>> cost?
>If the tests are right, you are really looking to get that 4% worse
>performance for this homemade version improved -- it seems like there
>is not a lot of room for improvement?
>The other thing to consider is that silicone is a very good material
>to use inside collectors -- it has a high temperature capability, its
>glues the parts together in a way that will still accommodate some
>differential thermal expansion, and it has a good track record for
>providing a long life in this tough environment. Not that the other
>compounds might not do just as well or better, but it seems to me
>there is some risk. There are some silicone caulks that are filled
>for higher thermal conductivity, and this might be a good thing to try
>-- I've not found a good source for these yet.
>Gary
Thanks Gary,
Looks like you have put a lot of effort into this collector, I would bet
there are a whole bunch of us that appreciate your time.
Mike
Robert Scott
>Do you think there would be any advantage to a variable speed pump that
>would adust its flow rate to keep some (unknown to me) minimum temperature
>differential between the bottom and top of the collector?
Well, it certainly is beneficial to take into account the temperature
differential when deciding when to turn the pump completely off. It is less
clear when partial speed is preferable to either extreme. If full speed is not
economical in some certain scenario, then I guess the the heat energy you would
have collected at full speed is less than the cost of the electricity to run the
pump (assuming pump wear and tear is not an issue). So what is likely to happen
if the speed dropped to half? Because the effect of viscosity is non-linear,
the energy required to push the liquid around at half speed is less than half
the energy required to push it around at full speed. And if dropping to half
speed allowed the temperature differential to rise to the magic threshold, then
you would be getting exactly half the heat energy as you would at full speed
(with the same temperture differential). So here you are spending less than
half the electrial energy to get exactly half the heat energy. So yes, lowering
the speed would be cost-effective in theory.
Having said that, there are other practical considerations. A variable speed
control is a one-time expense that needs to be amortized. In calculating the
pay-back period of a variable speed control, you have to ask yourself how often
it would be called upon to work in variable speed mode. And when it is working
in that mode, the amount of heat energy it is collecting is only slightly more
than the electrical energy used to run the pump. So while there is a benefit,
it is quite small.
Robert Scott
Ypsilanti, Michigan
Jeff
Well, the fins are collecting the heat. As far as the contact area,
this is of much less importance for copper/aluminum as you can get
pretty good bonds (I use heatsink compound), at least compared to the
thermal resistance of the fin itself. For plastic piping, this is a
different story.
Also note that you have much less flow resistance with 1/2" over
3/8", that will be more important in thermosyphon systems. But you will
find that the flow through the system, to get the temperature as low as
possible, has a fair size impact on efficiency.
Jeff