GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Ground source cooling — no heat pump ?

Khkiley | Posted in GBA Pro Help on

I’ve been reading a lot of the questions and comments on different variations of ground source heat pumps, I’ve been kicking around an idea for while that may be unconventional because I haven’t seen it mentioned yet.

My requirements may be unconventional as well, as I am interested in comfort, more that cooling, and a solution I am interested in needs to minimize its affect on the original fabric of our 1928 brick Colonial. Which I will explain.

Our 1928 Colonial is a wood framed home, clad in two courses of brick. It retains 99% of its original plaster walls. It retains its original wooden windows, which are fairly well sealed with vintage interlocking weather stripping, meaning there are no noticeable drafts from the windows in the winter. No meaningful insulation at this time.

We cool the house the old fashioned way, we open the windows at night and close them during the day and keep the shades drawn. Cool night breezes keep the home cool and there is a 100 year old oak that shades half of the house.

The house is almost always cooler inside. Normally the house won’t exceed 78 degrees throughout the day during the summer. The house does a fantastic job of absorbing the heat during the day and releasing it during the night.

Until the 95+ day/night heat waves hit. Without the cool nights, after a couple days the home loses its capacity to cool off. We live in NJ in a NYC Suburb.

We are not interested in the cost/intrusion of central air, window units are out because of the negative visual impact to the home, and mini-splits are out as well for the internal visual impact.

We are looking just for improved comfort while we sleep, for the most part we are out and about during the day. Lowering the home to a specific temperature is not important. We just use fans now so looking for an improvement over that, and focused in the bedroom.

My idea is to use a whole house dehumidifier to lower the humidity in the home when needed, this should improve comfort. And to use water cooled air handlers which I can install in the attic above the bedrooms to exchange and cool the air there. I expect I can run the cool water piping in the 2×4 walls and insulate them in the space. This will minimize the impact to the fabric of the home, and can be accomplished during other renovations.

The unconventional part of this is the method of chilling the water. I want to use a 500+ gallon (whatever is required) underground cistern to store and allow the ground to cool the water over time. The cost of deep drilling a heat exchanger puts that aproach out of reach, I don’t think we have enough space in our back yard to lay out loops. The idea is to have a large volume of water that can absorb the heat and return it to the ground during the day when we are not using the system. I borrowed this idea from the electric companies that installed large water tanks in customer homes, heated them during the evening for heating during the day to time shift the electricity usage.

I think this could work. The ground temperature of 57 is higher than normally used, but I think would still be effective. I think the trick is to size the cistern effectively.

And again, the important part is improving comfort, not necessarily lowering the temperature of the home to a specific number. Dehumidifying and cooling the air in the bedroom may be enough to make sleeping during the heat waves possible without waking up in a pool of sweat.

So what could I be missing ? What may trip me up if I try this out ? What haven’t I thought of and should consider ?

Thank you in advance for your thoughts.


GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. GBA Editor
    Martin Holladay | | #1

    There are a lot of possible reasons why this isn't a great idea, but I'll focus on one: Your idea would cost thousands of dollars to implement. You can get a window-mounted air conditioner for your bedroom for $200 or $300.

  2. Jon_R | | #2

    The soil surrounding your tank will start at 57F, but rise quickly as soon as you start pushing heat into it. Consider a ducted mini-split. Or a Chilltrix if you want to use water.

  3. Khkiley | | #3

    Hi Martin,

    Thank you for your reply. Let's assume I have already given window air conditioners some thought and decided to look at alternatives, what would be the other reasons you had in mind that would not make this a good idea ?



  4. Expert Member
    Dana Dorsett | | #4

    To have any reasonable cooling efficiency you need at least a 20F difference between the water temperature an the air temperature in the room you are trying to cool (30F would be much better.) With 57F water the efficiency will be pretty good for a 90F room, but much diminished for an 80F room. Below 77F the power used by the blower motor becomes a larger fraction of the heat you need to remove.

    Note: A smaller sized window-shaker AC unit is about 5000 BTU/hr.

    The painfully simple math:

    500 gallons of water weighs about 4200 lb.

    Putting 1 BTU of heat into 1lbs of water raises it's temperature 1F.

    By the time you've been running a hydronic cooling coil at 4200 BTU/hr (about the capacity of a tiny window unit) for a couple of hours, the 57F water in your 500 gallon buffer will have risen in temperature by 2F to 59F. By the time you've run it 4 hours it would be over 60F. It doesn't take many days of heat wave use (even when used sparingly) to raise the water temperature to a uselessly high number. After 20 hours it would not even very efficient at cooling the room to even 90F.

    Over a five day heat wave, four or five hours per day... you're toast. In a day, three at most, you'd just be heating up the rooms with the power consumed in the blower motors. Maybe if you started out with a 50,000 gallon tank you would have enough cooling capacity to keep it under 80F through a single heat wave, but seriously, heat pumps are the only viable solution.

    In order for hydronic cooling systems to work you need some means of cooling the water in the buffer tank, which is why chillers were invented. With a small air source 2-ton chiller like the Chilltrix & a 50 gallon buffer tank can do a lot, but it has most of the same aesthetic issues of mini-splits.

    You could site a mini-split compressor or chiller a few yards away from the house visually shielded by shrubbery, and trench the lines below grade to a basement buffer tank or ducted mini-split cassette, but that increases the installation costs considerably.

  5. charlie_sullivan | | #5

    Dana's math on the underground tank shows the limitations of that approach.

    I grew up in a NJ house that was cooled by a whole house fan at night, and kept cool by the shade of big trees. It now has central air conditioning, because the humidity starting causing mold and mildew inside.

    So I understand the focus on dehumidification as the primary task. However, the hardware in a dehumidifier is the same as an air conditioner, with the difference being that the dehumidifier heats the air after cooling it to drop the moisture out, whereas the A/C leaves it cold. So it doesn't make much sense to use a conventional dehumidifier unless the temperature in the room is already cooler than you want.

    So that brings me to the same place as Jon R's earlier answer: A ducted mini-split, or perhaps a Chiltrix system. For example, you could mount the ducted mini-split head in the top of a closet, and run ducts a foot or two from it through walls to each of the rooms that adjoins the closet. Unfortunately, I don't know of anyone who makes the equivalent of a ducted mini-split heat for the chiltrix system, even thought that would seem like a versatile component to have in the system.

  6. Khkiley | | #6

    Thanks Dana and Charlie.


    I am guessing from your numbers the ability of the tank to transfer the heat into the ground is going to be overwhelmed by the heat being deposited in it ? If that is the case then you have identified the choke point in my plan, and I would be back to some sort of ground loop system to cool the water that is in the tank. Which may not need to be as large if I use the tank as a buffer and can spread the cooling over time. But still adds cost I was looking to avoid.


    The Chilltrix system is an interesting option. I will do more research in it.

    Aside from a leak from a bathtub, we don't have any discernable mold issues. The NJ humidity is a major comfort problem though. There are many times the house is cool, but the humidity remains a problem.

    Thank you for the advice.


  7. GBA Editor
    Martin Holladay | | #7

    You've been given good advice and analysis by Dana Dorsett and Charlie Sullivan.

    The basic problem with ideas like yours -- using ground loops or buried tanks to temper indoor air temperatures -- is that the value of the heat gathered (or the "coolth" gathered) is too low to justify the high cost of the equipment installation. In many cases, the cost to operate pumps and blowers is also higher than the value of the gathered energy.

    You may be interested in reading my analysis of ground loops used to temper the air flowing through an HRV: Using a Glycol Ground Loop to Condition Ventilation Air. The best data I was able to gather showed that this type of system has a payback period of 4,400 years.

  8. Dana1 | | #8

    "...the ability of the tank to transfer the heat into the ground is going to be overwhelmed by the heat being deposited in it ? "

    Yes. It's all about the RATE of heat transfer between the tank and soil, which isn't nearly high enough.

    The amount of surface area of the tank for heat transfer to the ground is tiny relative to it's volume, the dirt is somewhat insulating. The thermal conductivity of soil varies by soil type & moisture content, but even wet clay isn't nearly conductive enough to move the heat out at a high enough rate. That's why ground source heat pump ground heat exchangers are long skinny pipes, with a high surface area to volume ratio, but they still have to run the fluid temperatures 10s of degrees below the ground temperature to get adequate heat exchange rates. A 500 gallon tub of 60F water in 55F soil will take pretty much forever (months) to cool to 56F, and even a long skinny pipe of 60F water in 55F soil would take well over a day to bring the water to under 56F.

    Unless you're planning to heat with it as well, even a small chiller like the Chilltrix + buffer tank would be an outlandishly expensive way to go, but it would allow you to micro-zone the place for both heating & cooling. A ducted mini-split would likely be half the installed cost (assuming you have somewhere to run an interior duct chase from a basement-mounted mini-duct cassette to the upper floor), and that too could see dual-use for heating, but you won't be able to micro-zone it.

  9. curtkinder | | #9

    Even if you somehow managed to passively create a source of water or air cooled to say, the mid 70s you'd be SOL as to dehumidification. Sufficient moisture removal for comfort depends on lowering the supply air temperature by about 20*F below desired room temperature. We normally design for suppling 55^F air to a 75*F space.

    Using chilled water to accomplish that outcome means the chilled water supply needs to be 45*F or therabouts.

    Since you are willing to go to the expense and aggravation of a central dehumidifier, but you loathe the idea of a classic ducted central air system why not split the difference and go with a low static ducted mini? We deploy the Mitsubishi SUZ / SEZ combo for compact floor plans needing under 1.5 tons heating and cooling.

    The equipment is small, efficient, and whisper quiet...what's not to like?

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |