Geothermal driveway heating

[Dean Carpenter]

A friend is in construction of his house right now, and plans to install a home-grown geothermal heating system in the driveway to keep snow off. I'm not convinced it will work, but since we're planning to build next year, if it does work, it sounds very cool, and I'm going to want to do it too :)

Basically, he's planning on putting down pex (I think ?) tubing before the asphalt is laid down, so it's embedded in the asphalt. That will lead to several loops dug down below the frost line for heat transfer.

Hrm - not sure if he's going vertical or wide horizontal loops. It's a closed loop system, with a small pump to move the glycol solution through. He hasn't done any calculations at all, just "it should work" :) It sounds like it will, but I don't know how to calculate if it will or not.

I thought of using a manifold system, with two main pipes along the sides of the driveway and ladder-like rungs of pipe between them. Each run would have several loops, to ensure sufficient time for heat transfer.

He's against that - worried about the T-joints leaking. He wants to run one or two looong loops. I'm thinking that at the end of the loop, all the heat will have already been removed from the fluid, and it won't be effective. So you would get one end nice and clear of snow, the other end as if the system was off.

So, questions are ... Is this even feasible ? Would there be enough heat transferred to keep the driveway above freezing point ? I would think so, sort of, since below-ground is a fairly static 50degF or so, isn't it ? And what would the best layout be ? Long loop, or manifolds ?

Anything else ?

D.

[nicksanspam]

Sounds like this could work in principle, but it might require a LOT of water pumping and tubing. You can find some clues here:

They describe a Japanese system in which 60 F water circulates through a heat exchanger under a sidewalk, melts off snow, cools to 45 F, then gets sprinkled onto the road next to the sidewalk.

And a $3 million Swiss system that melts snow off a 14K ft^2 bridge by storing about 20% of the summer's heat (512K million Btu) from tubing below the asphalt surface in about 2 million ft^3 of sandstone via 91

200' boreholes with heat exchangers :-) The store loses 35% of the heat by wintertime, when the rest is recovered to melt snow off the bridge.

It looks like most snow melting systems aim at 100 Btu/h-ft^2 min, enough to melt about 1" of snow per hour. You might sprinkle the driveway and collect water it in a trench along each side. A few leaks might help, since the main mechanism for upwards heatflow in soil is evaporation from lower soil layers, upward vapor migration through pores, and condensation in layers above.

Nick

[daestrom]

Darn!!! So when it snows over 6"/hour here (a near weekly occurance in Jan/Feb), I'm out of luck :-/

Oh, well. Time to PM the snowblower anyway as it's already October.

daestrom

[nicksanspam]

...describes a Japanese system in which 60 F water circulates through a heat exchanger under a sidewalk, melts off snow, cools to 45 F, then gets sprinkled onto the road next to the sidewalk.

The batch simulation below seems to indicate that a groundwater system with a 2'x2' trench on one side of a driveway might keep up with 10"/hour of snow, not counting the useful energy in the original trenchful of water. It seems to do surprisingly well, with a 3.3 hour layer time constant and 0.1 hour timesteps. The 10" of snow over a 10' wide x 1' long strip of driveway is like 53 pounds of ice, which requires 7680 Btu to melt, with no heat loss to the outdoors (in this first-order model.) The first 4" layer of soil below the trench has a heat capacity of 1361 Btu/F, so it (alone) can supply the snow melting energy with a 7680/1361 = 5.6 F temperature drop, which is close to the final temp drop in the simulation.

We might keep the soil under the trench damp in wintertime by measuring its lengthwise conductance and automatically adding water as needed with a solenoid valve when the soil conductance becomes too low.

Nick

20 SNOWDEPTH=10'(inches) 30 SNOWDENSITY=6.4'(lb/ft^3) 40 DRIVEWIDTH=10'feet 50 MELTLOAD=144*SNOWDEPTH/12*SNOWDENSITY*DRIVEWIDTH'(Btu) 60 TG=55'deep ground temp (F) 70 GC=20'damp soil conductivity (Btu-in/h-F-ft^2) 80 CG=50'damp soil heat capacity (Btu/F-ft^3) 90 RI=24'trench radius (inches) 100 THICKNESS=4'layer thickness (inches) 110 FOR LAYER = 0 TO 10'(10 is deepest) 120 RLAYER=RI+LAYER*THICKNESS/2'mean layer radius (inches) 130 SLAYER=3.14159*RLAYER'mean layer surface (ft^2) 140 R(LAYER)=THICKNESS/GC/SLAYER'layer resistance (h-F/Btu) 150 VLAYER=SLAYER*THICKNESS/12'layer volume (ft^3) 160 C(LAYER)=VLAYER*CG'layer capacitance (Btu/F) 170 TEMP(LAYER)=TG'initialize layer temps (F) 180 NEXT LAYER 190 DT=.1'timestep (h) 200 TEMP(0)=32'trench temp (F) 210 FOR LAYER = 0 TO 9 220 IF LAYER=0 THEN Q=0:GOTO 240 230 Q=-(TEMP(LAYER)-TEMP(LAYER-1))/R(LAYER-1)*DT'heatflow out of layer (Btu) 240 Q(LAYER)=Q+(TEMP(LAYER+1)-TEMP(LAYER))/R(LAYER)*DT'flow into layer (Btu) 250 TEMP(LAYER)=TEMP(LAYER)+Q(LAYER)/C(LAYER)'new layer temp (F) 260 NEXT LAYER 270 T=T+DT'elapsed time (h) 280 ICEMELT=ICEMELT+Q(0)'total ice melting energy (Btu) 290 IF ICEMELT