Posted by nicksanspam on August 3, 2008, 8:47 am
With a density 62.46-0.01(T-68) lb/ft^3 at T (F), 40 gallons at 55 F weighs
62.59 pounds per ft^3, ie 62.59x40/7.48 = 334.7 pounds. Heating it to 140 F
raises the volume from 5.348 to 5.421 ft^3, an increase of 0.073 ft^3, or
0.55 gallons. In a large house with a working check valve on a city water
supply, those 2 quarts might expand the pipes elastically with no damage at
say, 60 psi, but that seems unlikely, since copper doesn't stretch much
at that pressure.
"Tiny" as in 2 quarts :-) But we can't push water back through a check
valve, which is often a safety requirement with city water supplies...
This could be a non-problem if the water heater were a more elastic $0
1"x300' 13-gallon black plastic HDPE pipe coil in 140 F solar-heated water
in a 4'x8'x3' deep plywood box tank with a folded 10'x14' EPDM rubber liner.
Posted by nicksanspam on August 3, 2008, 7:17 pm
I assumed the temperature in most of the piping did not change, but the OP
said the T&P valve popped at 150 psi, after a lot of hot water use, so maybe
the hot water pipes are already expanded after the hot water use, and they
shrink while the water in the heater tank expands.
Posted by Solar Flare on August 4, 2008, 2:07 am
Champagne bottles have a concave bottom in order to strengthen the glass so
the bottle cannot flex.
Posted by Solar Mike on August 4, 2008, 11:37 am
Domestic HWC systems have a cold water expansion valve fitted on the
cylinder side of the non-return or check valve to the mains street supply.
Any pressure increase beyond its setting pops it open, draining cold water
to waste, thus equalizing the pressure due to expansion. This is the case in
NZ anyway. The valve is normally slightly lower setting than the
overpressure/temp valve fitted to the top of the HWC, thus cold water is
wasted, not hot due to expansion.
Without these valves the cylinders would explode.
Posted by daestrom on August 4, 2008, 5:56 pm
David Williams wrote:
Well let's see...
Copper has an expansion coefficient of 9.4e-6 / degF. So a rise from 55 to
140 (85F) means that a 0.7500 inch diameter pipe is now 0.7506 inches
diameter. And if the total piping in the house is something on the order of
200 feet, it is now 200.16 feet.
So the original volume was 0.61359 ft^3 and the new volume if *all* the
piping heats up is 0.61506 ft^3. A gain of 1.47e-3 ft^3 or 0.01 gallon.
Nope, that doesn't contain the expansion.
Looking at 3/4 inch copper pipe...
It has a wall thickness of 0.065 inches. Copper has a modulus of elasticity
of around 15e6.
So for every linear foot of such pipe, pipe walls have a stress on them
stress = force-of-water / pipe-wall-area = 12*(0.75)*Pwater / (2*0.065*12) =
9* Pwater / 1.56
When the water pressure is 30 psi, the stress in the pipe wall is 30*9/1.56
= 173 psi
When the water pressure is 150 psi, the stress is 865 psi
So the copper 'stretches/expands' (865-173)/1.5e7 = 4.61e-5 dL/L. That
means it's 'new diameter' after raising the pressure is 0.7500346 inch.
This is less change than the thermal expansion so that isn't enough to
accomodate a 0.5 gal volume expansion.
Pure copper has a yield strength of about 10,000 psi. So that 3/4 inch pipe
would start to yield (bulge) at about 1733 psi water pressure. Of course a
joint or the water tank would probably yield first.
Don't know about everywhere, but here in upstate NY and also in MI, there is
no check valve in the supply from city main water. So the two quarts just
expands out the supply line. If one has a well, then the well pump's
expansion tank deals with it.
If the check valve in the supply line is *perfect*, then heating that much
water could burst a pipe. But even without a TPV, most houses would have
leakage past various faucet seals before the pipe burst. And that check
valve probably *isn't* that tight a seal. It would only prevent gross
'back-feeding', not stop every last drop of water.