Posted by Andrew Burgess on March 4, 2005, 3:25 pm
How about a spiral loop from the top to the bottom of the tank? That would
preserve stratification. Maybe insulate the vertical rise part. Make sense?
The how water enters the hot top of the tank and loses more and more heat as it
spirals to the cooler tank bottom. The now cool water then returns to the
surface in an insulated part of the tubing so it doesnt pick up heat from the
hotter top layers. Closed cell foam pipe insulation would suffice.
Also, I would consider copper for the exchanger
More reasons to lose the rock :-)
Posted by iain-3 on March 2, 2005, 6:07 am
Lawrence> You can solve a whole bunch of your potential problems
Lawrence> by going to a heat exchanger system.
The panels produce a peak of about 250k BTUs/hour. That's a hell of a
heat exchanger. I would greatly appreciate any pointers to practical,
reasonably inexpensive devices which can move this kind of heat with
low enough temperature deltas.
Here's why I think it's such a problem: suppose the delta-T is 5
degrees F. Suppose the surface area is 50 square feet. The R value
would have to be about 0.001 to make that work. That's really low.
I wasn't able to find R values for liquid-liquid interface on tubes.
Nick has posted many times that fin-tube is 5 BTU/hr-ft-F, but I
suppose that's for water inside, air outside, and I'm sure it would be
better with water outside. But at $/foot, $000 buys just 250 feet,
and the stuff would have to work 40 times better when immersed in water
for it to be sufficient. Does it?
Posted by nicksanspam on March 2, 2005, 11:31 am
...1" plastic pipe. Reread my last post.
I measured 30 Btu/h-F-ft^2 with a paper cup in a pot in my kitchen...
Posted by iain-3 on March 2, 2005, 7:45 pm
Nick> ...1" plastic pipe. Reread my last post.
Nick> I measured 30 Btu/h-F-ft^2 with a paper cup in a pot in my
You measured 30 BTU/hr-F-ft^2 through a paper cup in a pot, and you
want to suppose that 300psi plastic tubing has the same thermal
transfer? Okay, I'll take that jump with you. Maybe the transfer is
dominated by the films at the surface and not by the obviously thicker
and less permeable plastic wall.
Let's look at the 500K BTU/hr pool HX. It will see 60 F
input-to-input temperature delta. Let's suppose the pipe actually
sees half that since there will be some counterflow and general
badness. 1100 feet of 1" plastic tubing will have 287 ft^2 of
transfer area. With your transfer coefficient, that will move 259K
BTU/hr, at a cost of $40 from True Value. Very reasonable.
I don't want a pump on the hot tank side (I have way too many pumps in
this design as it is). So I need to get some convection going.
Suppose I put this 1100 feet of tubing into a 2 foot diameter 16 foot
long plastic culvert pipe (read: chimney) stood on it's end, so that
the bottom has a bit of clearance to the rock (and a diffuser, read
below), and the top sticks out of the top of the tank to the surface
and has a manhole cover. Near the top of the tank I have a bunch of
holes cut in the culvert pipe which lead to some perforated drain pipe
to gather water.
Density of water at 60 C = 140F: 0.983 g/cc
Density of water at 50 C = 140F: 0.988 g/cc
( http://www.csgnetwork.com/h2odenscalc.html )
A 12 foot column of this cooler water will generate a differential
pressure of 0.025 psi. Something like 20 to 40 GPM will need to flow
on the hot tank side. Across the clear 2 foot pipe 40 GPM is 0.34
inches/sec. It'll be higher as the water squeezes past the HX tubing.
To keep the water from having to accelerate into the rock, if I assume
the rock has 2/3 occupancy in the tank, I'll need about 6 square feet
of orifice from the diffuser at top and bottom to the rock matrix.
Does anyone have any idea if 0.025psi will drive water through rock
and pipe at 0.34 inches/sec? 1/40 psi doesn't sound like much. Maybe
I'll put in four chimneys, one for the pool, one for the DHW, and two
more just in case.
Note, though, that this is the easy HX, and the DHW is similar in that
the temperature deltas are large. From the panels to tank is an
entirely different story since the tolerable temperature delta is 1/10
as large. With half the heat flow, it would require 5000 feet of
tubing, which sounds like an accident waiting to happen, not to mention
the $250++ price tag.
Posted by nicksanspam on March 2, 2005, 8:58 pm
...100 psi. Keep up. The plastic pipe wall is about 1 Btu/h-F-ft^2 per inch,
but it's only about 1/16" thick. Add R1/60 for each slow-moving water film.
Where did that come from? You mentioned 250K Btu/h. I said one 300' piece
of 1" pipe would do that with a 10 F temp diff.