wrote:

I've been looking at this problem as well. My current best plan is to
fill an insulated, EPDM-lined hole in the ground with drain rock, then
cover that with EPDM and insulation and bury the thing under my patio.

Burying could be bad, maintenance-wise. Indoor heat loss might be nicer.

The point of the drain rock is that it's cheap, available and manageable
in bulk, and takes vertical stresses without any engineered structure.

One might say the same about water :-)

In case it's not clear, this is a water tank, not a rock tank .

That's not clear :-)

The rocks fill about 2/3 of the volume of the store, and store a little
less heat per unit volume than water. So the overall tank has to be
upsized a little.

A little? :-)

density Cm Cv
kg/m^3 kJ/kg-C kJ/m^3-C

granite 2640 0.82 2165
limestone 2500 0.90 2250
marble 2600 0.80 2080
sandstone 2230 0.71 1583
water 998.2 4.186 4178

I'm thinking of using "blueboard" extruded polystyrene insulation,
probably the 30 psi variety...

You might get the hole foamed, or forget insulation, if it's large enough.
And float foamboard under EPDM or build a shallow dome for a cover.

Even including the geotextile felt between the EPDM and the drain rock,
this is about $0.30/gallon, and has no size restriction, shape issue,
or safety problem with overheating.

US Patent No. 3,933,628 "Method and Apparatus for the Anaerobic Digestion
of Decomposable Organic Materials," issued to inventor Frederick T. Varani
of Golden, CO on Jan 20, 1976, assigned to Bio-Gas of Colorado describes
a methane generator for a 100,000-cow feedlot, using 2 EPDM-rubber-lined
trenches, each 700' long x 80' wide x 40' deep. They cost $0.02/gallon,
including self-inflated solar covers and excavation.

I do have some reservations about the idea though:

- Maintenance will be a bitch. The tank I'm planning is large (125
cubic yards).

Wow. Why so big? You could put a hundred tons of water in that 15' cube.

Digging 200 tons of wet drain rock out to get to a leak would be very bad.

Very bad.

I'm wondering what I do to make sure there are never any leaks.

Omitting the rocks might help.

- The top of the (unpressurized) tank to the bottom of the rooftop solar
panels will be 23 feet. I'd like to have a drainback system, but then:
- I am pumping up a 23 foot head, and that'll cost $$ for electricity...

You might put a few coils of PT Industries (800) 44 ENDOT PBJ10041010001
1"x300'100psi NSF-certified pipe on the bottom... $59.99 from True Value
hardware stores. Each coil would have 78.5 ft^2 of surface and might have
78.5x30 = 2356 Btu/h-F of thermal conductance, enough to collect 23,560
Btu/h of heat from about 236 ft^2 of solar collectors with a 10 F temp diff.

Nick