Low Temp Geothermal Snow Melt System (no heat pump)

New_Guy

Hi,

New the community and first time posting.

I'm planning on building a new house in WI in 2025 (hopefull). A friend put a wild idea in my head a few months ago: A driveway snow melt system that is as simple as a circulating pump moving a glycol mixture between the driveway and a geothermal ground loop system WITHOUT A HEAT PUMP or any other kind of heat source or heat exchanger.

The premise is the ground temps 6-8' deep should be able to maintain around 50+ degrees F year-round. The driveway would have PEX loops under the concrete and on top of 2" XPS to keep the heat in the concrete. It would use ground-temp fluid to keep the driveway above freezing all winter so any snow that falls on it would (eventually) melt without the need for me to go clear it with a shovel or snow blower.

I found one reference to a similar system used in New Jersey in 1969. They used ground loops lengths that were 2x the length of the concrete loops and spaced the ground loops 2'. I have the land to run as many ground loops in any length necessary so that's not a problem.

If the premise is solid, should I run the pump 24/7 365? This would essentially pull heat from the driveway in the summer and dump it in the ground, which would then be pulled back to the driveway in the winter. I've also read that I might want to only run the circulating pump starting a day or so before expected snowfall then turn it off after the driveway is clear. The idea there is to try to preserve as much ground temp as possible for deep winter melting needs. I'm skeptical that the intermittent approach would heat the driveway effectively or fast enough.

Curious if anyone has seen/done a similar setup and can provide any guidance.

Please ask any additional questions necessary. I'm happy to have any private discussions with the intent to share useful information back in this forum for others to benefit from.

Thanks!

New_Guy

EBEBRATT-Ed

Not a snow melt expert but with water in the 50s it will take a lot of pex loops to get enough heat transfer to do the job. I would wire the pump on and off to and outdoor air temp switch although they do have snow melt sensors, but they would likely not work fast enough. I see no need to run the pump in the summer.

New_Guy

I haven't used a tool like LoopCAD yet, but I expect to run 5/8" PEX 9" OC with loops of around 250'.

~New_Guy

pecmsg

6 - 8’ in jersey will be cooler then 50*. You need almost twice that depth.
I ASSUME your pumping glycol.

hot_rod

Download the free Uponor SIM snow, ice melt design guide. It takes you through the steps of a design. Supply temperature is generally over 100°F. Uponor suggests staying below 150°

I doubt you will get much sq ft btu with 50° supply.

An example of a SIM design here.

Screenshot 2024-12-07 at 2.26.35 PM.png

https://www.uponor.com/getmedia/50a684b6-091d-4044-b8f7-c06f7705668c/snow%20and%20ice%20melting%20design%20and%20installation%20manual.pdf?sitename=USA

Jamie Hall

As you note in your initial post, @new guy , you won't reach 45 to 50 degree year round ground temperatures until you get at least 6 feet into the native ground. Are you quite prepared to do that much trenching? That's a pretty fair amount of digging, even if you own your own backhoe…

leonz

You need heat to do this to keep it ice free. Your weather will not let you do this very well.

If they use salt in your area you can forget keeping the approach free of ice and slush.

Spend your money on a kubota 1870 with a Pronovost snow blower and snow chains or john deere series 1 sub compact tractor and Pronovost snow blower and snow chains.

A properly installed sub base for an 8,000 pound sand mix cement driveway poured 8 inches thick with properly spaced felt expansion joints 10 foot apart and small grid reinforcing mesh will last for decades and snow blower skids will not ruin it and it will last for 50 plus years.

New_Guy

Thanks for the quick responses. Below are additional responses and information.

@pecmsg - Yes, glycol mix. Likely 40% glycol

@hot_rod - I downloaded and read the Uponor SIM manual earlier this year. It's a great tool, but it doesn't help in this application because their data table doesn't include the values applicable to this application.

Here's my worksheet based on what I am able to determine right now:

@Jamie Hall - I expect to go down to 8 feet. I've been talking with an excation contractor who has done geothermal digs before. There is NOAA data for soil temps down to 5' about 60 miles from me (and Wisconsin isn't very geographically diverse ;)). The NOAA data has soil temps ranging from 42 to 68 degrees F in across all of 2023. Most of the winter temps were between 45 and 55.

What I haven't been able to find yet is how to calculate the BTU/h for the ground at a given temperature and depth. I know that pumping the colder fluid from the driveway will contribute to cooling the ground temp, but I don't know by how much and how fast. My understanding is that a long racetrack loop is better than a shorter slinky loop for each loop in the ground.

~New_Guy

Jamie Hall

There really isn't that much good experimental data on the interaction between buried pipes for geothermal and the ground they are buried in. However, there are some general guidelines or principles which can be applied — though they are probably conservative (or at least one hopes so…).

First and perhaps most useful is that the rate of heat loss or heat gain from the interior of a pipe to the surrounding material is dependent only on the delta T between the interior fluid and the surrounding material, assuming that the surrounding material is similar. In your situation, the surrounding material in the field is compacted earth and concrete in the driveway area. It is probably safe to assume that the heat flow rate in the compacted earth will not be as great as that in the conrete for a given temperature difference — but perhaps for the back of an envelope calculation one can assume no worse than two thirds.

Soo.

In the driveway we are trying to hold a temperature of no less than 32, and we have incoming water at about 45 — for a rough calculation that works out to perhaps a delta T of 13 or so. Interestingly, if we get the flow rate, the return fluid temperature to the field will also be around 32, and the field 45 — so about the same delta T, but only two thirds as much heat transfer per foot of pipe. Scribbling furiously, it can be seen that you need at least half again as many feet of pipe in the field as you have in the concrete — or an absolute minimum length of trench and field pipe of 7,500 feet at the beginning of the season.

The other aspect is that the temperature of the soil will drop over the season as you run the cold water through it. The heat capacity of soil is around a third of that of water, or around 0.3 BTU per degree Fahrenheit. At this point the calculation gets decidedly messy — well beyond the back of an envelope — but the bottom line is that for a conservative estimate for your field you would want to keep the field temperature up to at least 40 to have a hope of melting any snow. I note that you have not determined the BTUh demand of your driveway, but if I were to assume a reasonable value of 10 BTUh per square foot of drive, running for 150 days, I would be looking at 36,000 BTU per square foot of driveway. With a delta T of 5, I could extract that much heat from around 7,000 pounds of soil, or very roughly 50 cubic feet. So around 200,000 cubic feet of soil. By some more scribbling — not so much on the back of an envelope this time — you will need a pipe to pipe spacing in your field of around 6 feet on centres, as a minimum.

Now. I answer questions, but that does not mean I do not have opinions, and in this instance I can't see that this proposal, while clearly it can be done, will be either cost effective or particularly sensible.

hot_rod

how large of an area are you looking at melting?

The slab surface needs to get above 42 f to melt you have edge and down losses

You want around 100 btu / sq ft for residential sims

I just don’t see you having enough delta to warm this slab

The example above uses 132 supply to get 120 btu output

Also the exchange will cool down the loop field

Maybe someone with a geo simulation program could give you some ideas

Maybe the IGSHPA site has some info

Jamie Hall

Thank you for those numbers, @hot_rod . Much better than my guesses. But not so good for the final bottom line, eh? I agree that to get anything like 100 BTUh per square foot you are going to need much higher temperature fluid circulating. So that kind of makes it difficult…

More to the point, instead of a 6 foot spacing, you are going to need more like 90 feet centre to centre, and much longer — perhaps 10 miles instead of 2 — of piping to avoid freezing the ground.

Teemok

I think chilling the loop field would be the biggest problem. Heat transfer from earth to earth and then earth to tube is slow. Many soils has an R valve of .25 to 1. The 12" of soil around the field pipe might be R12 or R3. Once you pull the energy out of the near 12" around the pipe heat flow slows. The delta between the pipe and where the heat is, isn't that big and you've got a growing R valve as you chill soil. The heat has to travel through more and more soil to get into the pipe. Not long before you have R24 between the heated ground and your field pipe. You don't want to wait 10's of hours for snow or ice to melt. Phase changes require more energy than the average person expects. If you have a limited rate of adding energy you will have a long wait.

Jamie Hall

There is another rough way to estimate the land area required… which gives slightly less dramatic requirements (only four miles of trench). It still gives figures which make the project dubious…

leonz

https://forum.heatinghelp.com/discussion/199124/low-temp-geothermal-snow-melt-system-no-heat-pump

I guess if you want to do this, let us do it the right way the first time.

The only way you are going to know what your ground temperature is the year round is to use a high quality soil thermometer and understand that this is the REAL DATA you must use to determine if passive deicing is even possible. I can tell you from personal experience in a self serve car wash that it is not.

Here is a link to a high quality commercial soil temperature gauge with a 36 inch probe-no you do not want to stab it in the ground it to depth, you want to soak the ground deeply with water to depth UNLESS you have sandy soil.

www.gemplers.com/products/361-dial-probe-soil-and compost-thermometer

Gemplers Item Number #109645

This is the only way you will have an answer to your questions which will prevent you making a huge mistake.

If you are still intent on doing this a small coal stoker boiler like the keystoker kaa-2 set with summer temperatures using a mechanical triple aquastat exactly like the L8124L and an Honeywell L6006L single aquastat for a dump zone can be used to make hot water with a cryotek antifreeze/water mix will be the most efficient way to do this using a small circulator, steel compression tank, airtrol valve, and Internal Air Separator to provide the point of no pressure change.

We ran our floor and approach apron heating loops at 140 degrees high limit using black polyethylene pipe from October 1st to May 1st every year.

DCContrarian

A google search tells me snow melting systems need at least 100 BTU/hr/sf. With 3700 square feet you'd need 370,000 BTU/hr. With water entering at 45F and leaving at 32F that's a 13F delta, you'd need a flow of 56 gallons per minute to move 2.7 million BTU/hr. I think when you look at what kind of pump and piping you need to move 56 gallons per minute you'll find it's infeasible.

Also: with water entering at 45F and leaving at 32F the average water temperature is 38.5F. That's 6.5F above the surface temperature. To get 100 BTU/sf/hr with a delta of 6.5F, the insulation between the pipe and the surface has to be no more than R-0.065. Concrete is a pretty good conductor but it's about R-0.1 per inch. Let's assume with 9" spacing every spot on the surface is an average of 5" from the pipe, that's R-0.5, the most you're going to be able to deliver is 13 BTU/sf/hr with that water temperature and surface temperature.

If the air above the snow is colder, there's going to be heat loss into the air, the snow acts as the insulation to prevent that. If the snow layer is thin and the air temperature is cold you could be losing more than 13 BTU/sf/hr to the air and while the surface would be warmer than it would be otherwise, no melting would occur.

Put this all together and for a practical snow-melting system you need an average water temperature of at least 100F; to get a practical flow you want a high temperature drop so you probably want something along the lines of a 140F supply temperature and a 60F return temperature.

GroundUp

There is a fella local to me (MN) who tried exactly what you're suggesting. When it's 30 degrees outside, it'll melt a decent amount of snow. When it's 0 outside, it can run 24/7 and barely makes the snow sticky. This is similar to the folks who try to get free air conditioning by pumping groundwater through a W2A coil and simply put, it doesn't work.