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

Natural Creek to Power Ground Source Heat Pumps

TNREM | Posted in General Questions on

Hey folks,

I posted earlier on my current project of sizing/installing mini-split ductless units in our home. In a somewhat related but perhaps off the wall thought, I’m also starting to ponder a bit about a ground source heat pump. To be clear, I wouldn’t ordinarily consider it, because of the nature of our location, however, we do have one ‘resource’ that might make it not only feasible, but the preferred choice. BUT, because experience has taught me that my ‘good ideas’ often aren’t so good, I thought I’d ponder out loud here and see if anyone had any feedback. So, here goes.

We have a little creek that flows by the house over a slate bottom, and then disappears into the stream bed about 50 yards beyond the house where the creek bed turns to gravel. The dry creek bed itself continues on a half mile or so to a much larger creek that basically drains the whole southern part of the county.

The creek flows year-round, with an average minimum flow of about 300 to 400 gal/minute at the driest time of the year (which is Sept/Oct). I know because I set up a weir to measure it and have watched it for several year, with thoughts of perhaps one day putting in micro-hydro. The creek varies from 10 to 20 feet wide and runs from 1 to 3 inches deep. The one-time I measured summer time temp it was 62 degrees F.  It’s spring fed, starting on our property from a multitude of tiny springs, so I honestly expected it to be colder, but I suppose that’s because it’s shallow and the dark slate bottom gets anywhere from full sun to dappled sunlight as it runs through the woods. I’ve never measured the temp in the winter, but I’ve never seen it frozen (though once when it got to around zero F there was ice along the banks. There’s not a great deal of head, but there is about 15 to 20 feet around 200 feet upstream from the house. I’m sure I could bury a pipe without too much difficulty and easily get 10 or 15 gpm to gravity flow to a heat exchanger at the house and then discharge back into the creekbed as an ‘open loop.’  All gravity, no pumping required.

So, it seems like a ‘natural’ for a GSHP, but I also know (or at least have read) that the efficiency of geo-thermal systems is VERY dependent upon design and installation. Also, and as discussed in the other thread on mini-splits, I’ve determined that my cooling load is pretty minimal (realistically about 1 ton, though I’m going to have to install more than that). I can get the mini-splits more or less as ‘canned’ units, where as a GSHP seems more involved and I don’t even know if can get one as small as I need. My thoughts are that even if I have the best, most easily exploited source of ground water in the world, if I still have to run a larger compressor than I need, it might turn out to be a wash (or worse). Also, the installation of the GSHP would (I think) be somewhat more involved (design wise) though I’d probably put the coil the existing ductwork of our propane furnace.

Frankly, I’m still leaning towards the mini-splits because it would be the easiest, but I keep looking at that creek and thinking I should be able to use it somehow. Obviously, given that we’re totally off grid and solar-powered, I would like to minimize power consumption. OTOH, I don’t want to dump extra resource into this for what might be only a marginal improvement.

Any thoughts appreciated.

P.S. We own both sides of the creek in its entirety, from the ‘headwaters’ until well below where it disappears into the earth. We don’t own the ‘mouth’ where the dry streambed connects to the larger creek, but water only flows that far during really torrential rains and then only for a few hours at most.

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. Expert Member
    BILL WICHERS | | #1

    Note that you might run into regulatory issues by discharging heat into that creek.

    I don’t know if you’ll get enough flow by gravity feed, but I’m pretty sure you’ll be able to gain some air conditioning efficiency by using that creek. The cooler you can get the condenser coil while running in air conditioning mode, the more efficient the system will operate. You also don’t have the big expense of a geothermal loop or well system, and you won’t have any issues with ground saturation if you’re using moving water has your heat sink (or source).

    I say go for it and give it a try, just be careful of any potential regulatory or water rights issues.


    1. Expert Member
      NICK KEENAN | | #3

      Second on the regulatory issues. Around here they'd jail you if they caught you doing something like that.

      My experience with creeks is that they are unreliable. They run dry, and they flood. It's hard to design an intake that isn't at risk of getting washed away in a flood. Creeks are full of debris. It's hard to design an intake that's far enough off the bottom that it doesn't pick up debris from the bottom, far enough away from the top that it doesn't pick up floating debris, doesn't let anything in but doesn't clog, and works in all weather and water levels. More common is to dig a well beside the creek and let the earth be your filter.

      Does it freeze where you are? If so you're going to have to bury the pipe to keep it from freezing.

  2. Expert Member
    AKOS TOTH | | #2

    If you have an ERV/HRV, probably the best use of the water is a brine loop pre-heat for it. This would still mean a pump that runs 24/7, so it can add up off grid, but if you size the system correctly and keep pressure drop and flow rate to a minimum it can be made to work. You can also run it off a solar pump but than it will only work during the day.

    This way you get the benefit in all seasons without too much extra invested in heat pumps.

    I would also be careful with any water base systems in off grid. Frozen pipes or heat exchangers are very expensive to fix.

    1. TNREM | | #14

      Akos, I have neither and ERV or HRV and in fact, had to look those terms up. LOL

  3. joshdurston | | #4

    I know of someone in here Ontario, Canada who ran about 8 tons of gshp off a very shallow creek that flowed year round. They put a closed loop in the creek. it was a couple hundred feet of 1.25 hdpe pipe. It ran very well at good efficiencies and was installed in an afternoon. This worked well for the 10years they used the property.

    I wouldn't consider an open loop, too much hassle with debris and air locks and such.

    Lake and pond loops are pretty common and negate a lot of the loop cost of GSHPs (especially compared to vertical loops). A creek with moving water should perform even better assuming the temperatures and flow holds for the winter.

    In Ontario you typically use Ethanol for freeze protection. The equipment has a freeze protection safety that should trip long before you freeze a heat exchanger to the point of damage.

    1. TNREM | | #13

      Josh, Good to know and it sounds workable. My only hesitation is running either a loop or other heat exchanger charged with anti-freeze (or anything actually) submerged in a running creek. If the system develops a leak, I'm not real keen on pumping that stuff in the creek. Since here in the Southeast my main load is AC and I have a propane furnace back up for heating, I would likely just charge it with clean water and drain it and go on the gas furnace when temps dropped below freezing.. Or, as an option, I suppose I could go with brine. Not optimal either, but if I did get a leak I'm sure the amount of salt released would be quickly diluted with negligible environmental impact.

      1. joshdurston | | #16

        Ethanol (grain alcohol) is what we use here Ontario, it breaks down pretty quick if leaked, it's already diluted in the loop. Other parts of Canada use methanol.
        Alcohol pumps better than glycol at low temps.
        If it leaks into a lake or pond it's not an issue. For a very small creak perhaps it would to high of concentration. Food grade glycol might be another option that's eco friendly.

        I would say if your only using it for AC you might as well go with a high efficiency (inverter) air source unit. They are cheaper, as efficient at cooling, and virtually silent most of the time if you have inverter compressors and outdoor unit fans.

        1. TNREM | | #22

          Ah yes! Ethanol. Sold during my college days as 'Everclear' and a great punchbowl mixer. How could I forget? Likely the only downside would be tadpoles with severe hangovers. LOL

  4. walta100 | | #5

    The other problem with “open systems” is water quality and how the mineral in your water will react with the metals used to build your system. With a closed system the same water is circulated over and over again with its small quantity impurity. An open system is exposed to unlimited quantities over time.

    Over time open systems have been known to get coated with stuff and get clogged or corrode and leak.

    Water rules are very different from place to place too much for internet advice to be very helpful.

    In most states you do not own the water in the stream just the rocks under the water.

    If you could get the required permits small scale hydroelectric generation has always been interesting to me.

    I do not think any type of electric heat make sense when off grid.

    I have yet to see a BEopt model where a geo system came out with a lower cost to install and operate. You could if someone would install it without a flow pump and if you could avoid the costly loop/ wells. But understand you are unlikely to get any warranty that will cover an open heat exchanger.


    1. TNREM | | #12

      Thanks, Walta

  5. TNREM | | #6

    Hi folks, and thanks to all for the input. First, the regulatory climate is fairly reasonable here. I haven't checked out this particular aspect (that is, cooling), but I have checked out micro-hydro and as long as you stick to 'run of river' and don't do a big water impoundment you can square that with the regulatory folks fairly easily. Since the 'cooling' aspect would be a fraction the water diverted for micro-hydro I don't anticipate this would be an issue. Also, there aren't really any 'downstream' issues as the creek surfaces and subsides 100% on our property and given that I'd be diverting less than 1% of it's volume at it's lowest flow and then 'returning' it just before it went underground, there wouldn't be any 'loss' to anyone downstream (wherever it might 'pop up'') and I really seriously doubt there would be even a measurable increase in creek water temperature.

    The other concerns (trash, silt/sediment, water chemistry, etc.) are all valid, and a 'closed loop' submerged in the creek might be a better idea. Maybe I can turn the pump with a small paddle wheel and save a bit of energy there. LOL The 'creek rising' IS a very big concern however, because the watershed is mostly vertical on both sides of the creek. All things to consider.

    However, in focusing on the creek and its possible usage, I think I've buried my original question (entirely my fault). As some of you may have seen in another thread, I was inquiring about sizing a mini-split (and that's the way I'm headed at the moment). So this question here is really not so much about the creek, in that whether I use an open system with 'gravity feed' creek water or a closed loop submerged in the creek, I'm sure the resource of the creek can be incorporated into any GSHP solution. But the REAL question is how the GSHP might compare with the mini-splits in terms of efficiency and most especially power consumption. I like the mini-splits for a number of reasons, so if it's a close call in terms of energy usage, I'd go mini-split. What I'm really looking for is feedback or opinions on whether a 'creek cooled' GSHP would be significantly more 'energy thrifty' than a mini-split of the same cooling capacity.

    So again, sorry for 'disguising' my real question in descriptions of the creek.

  6. Expert Member
    NICK KEENAN | | #7

    While GS has the theoretical edge, the AS manufacturers have been much more aggressive about adopting technological innovation. So in the real world right now GS has no efficiency advantage, and AS has a significant price and complexity advantage.

    I recently was reading an AS manufacturer website that claimed their heat pump exceeded the performance that a GSHP would need to be Energy Star rated.

    1. TNREM | | #10

      Thanks DCContrarian, That's just the sort of feedback I was looking for.

  7. Expert Member
    BILL WICHERS | | #8

    >”Since the 'cooling' aspect would be a fraction the water diverted for micro-hydro I don't anticipate this would be an issue.”

    The issue is how much you would warm or cool the water with your heat pump system, not just how much you’d effect volume of water flow. I’ve never had any expieirncd with a system as small as you’re proposing (I had some peripheral involvement with a powerboat plant that used a river for cooling, but that was around a 1,000 MW plant so a little bigger than your system :-). Depending on how much difference your system makes, you can actually alter the ecosystem in the creek. The power plants can raise the water temperature of a river in the area of their outlet enough to change the habitat. Some critters find this change beneficial, others do not. This is the “other” regulatory concern you have to think about besides just water volume and flow.

    The efficiency improvement you get has to do with how much your geothermal loop shrink the delta T your heat pump sees. It takes more energy to air condition a space to 74F when it’s 100F outside than it its 80F outside, for example. If your creek acts to keep the condenser cooler than the outdoor ambient air temperature, then you’ll see an efficiency increase in the operation of the system. The same goes in reverse for heating. How much of an efficiency improvement you’ll actually see is complicated to predict and needs a lot of additional information, and you need to deduct the energy the circulating pump uses on the “creek loop” from whatever potential compressor efficiency gains you’d see.


    1. TNREM | | #9



      "The issue is how much you would warm or cool the water with your heat pump system, not just how much you’d effect volume of water flow. "

      LOL, I think maybe you missed my last sentence, which read:

      "I really seriously doubt there would be even a measurable increase in creek water temperature."

      To elaborate, the water would be 'returned' to the stream just as it filters across the gravel creek bed and sinks underground. I obviously can't know what subterranean water dwelling critters might exist, but I'm fairly confident the tadpoles in the surface creek wouldn't be impacted. Also, this is pretty wet 'karst country' and we have two caves (that I know of, likely more) on our property alone and both of them have flowing water. I suspect (but can't know for sure, that my little creek is one of many that flow into these underground water courses. I totally understand your concern, but I honestly can't imagine that raising 1% of a flowing volume of water by a delta T of even 30 or 40 degrees and returning it to the stream (just as it goes 'subterranean' into a massive constant 50 degree 'heat sink') would have a measurable environmental impact. Otherwise, I wouldn't do it.

      Also, on the regulatory side and with regard to riparian rights, etc., I've been looking at and researching micro-hydro for over a decade, with a view towards putting in a system. So rest assured that anything I do with the creek will be in accordance with applicable regulatory guidelines and rules. I totally understand that you may think I'm ignoring that aspect (and perhaps I contributed to that impression with my previous responses), but trust me when I say I'm not. I'm totally dialed in to the requirements and my responsibilities on that end of things.

      I also understand the working principles of the heat pump and realize efficiency gains are relative to shrinking the delta-T at the heat pump, and that just 'how much' that improves efficiency can only be guessed at without some real world numbers (or at least reasonably close estimates) to run the calculations. Obviously, no one can give me those at this point.

      I guess what I was really looking for was informed opinions as to whether or not there might be a sufficient difference in efficiency/energy savings between the GSHP (creek cooled) and an AS (mini-split) system to justify going with the GSHP. And for me, that would have to be significant difference because I've lived with mini-splits while living overseas and really like some of the features (specifically 'zone control')

      So, I guess my question at this point is are you in agreement with DCContrarian's opinion above that the GSHP wouldn't make enough of a (real world) difference to make it worth the effort? I understand that can only be a 'qualified' opinion but at this point, that's all I'm asking for.

      Also, I figure if I go with a mini-split, I can always temporarily bolt a new car radiator (cost about $150) in front of the air inlet side as a 'pre-cooler' and pump cold creek water through it via a garden hose as an experiment to see if it improves things. Maybe I can snag the 'best of both worlds. LOL

      Anyway, I appreciate the feedback. Opinions on the above welcome.


      1. Expert Member
        BILL WICHERS | | #17

        I was just pointing out there is a potential regulatory issue with the temperatures. I didn’t mean to try to scare you off or anything like that :-) just be aware that there MIGHT be an issue there that some govt busybody might try to nail you on so be sure to check.

        Regarding a guess as to the efficiency improvement, I think it would take you a while to recover the expense of putting in the system. I suppose with you’re being off grid that “expense” is a little harder to quantify too since you can’t just say “I’ll save $x on my electric bill”.

        I think you’d find a micro hydro setup to be more useful. If you can combine a micro hydro setup with the geothermal loop, you can maybe double dip and share some of the costs, but otherwise I probably wouldn’t bother doing the geothermal system alone.

        If you want to read up on something similar for some real-world experiences, there are people that use their air conditioner to heat their swimming pools. There are commercial heat exchangers made for this purpose. This is similar enough to what you’re thinking about that you might get some ideas from those people that would help you decide if it’s worth trying on your own system or not.


        1. TNREM | | #23

          Thanks for the tip (on the swimming pool thing). I'd actually seen some sights selling those heat exchangers, but I think they'd be way overkill for my purposes. However, I'll research it a bit more for future reference. However, I think for the time being the best route all around is the mini-splits.

      2. Expert Member
        NICK KEENAN | | #20

        Wetlands are federally regulated.

        I'm not a lawyer but you need to talk to someone who is one before sinking money into this. Water rights is one of the most complicated areas of the law, it's definitely not an area where knowing a few general principles and some common sense gets you by. There are lots of stories of people who have lost enormous sums of money doing things they thought were insignificant on land they owned.

        1. TNREM | | #21

          Good to know. Thanks for your concern.

  8. TNREM | | #11

    Thanks, Walta

  9. Jon_R | | #15

    +1 on needing a lot of information to calculate it. If pressed to guess, maybe a 33% drop in cooling energy use with ground source.

    > we’re totally off grid and solar-powered

    This changes things considerably. You have three very different costs of electricity - when the sun is shining, when running on batteries and when running on generator. Time-of-use matters, perhaps more than COP.

    > mini-split .. really like ... 'zone control'

    Zoning (ie, independent temp control) is a downside to mini-splits and an advantage of hydronic systems. Same for using a large water tank for time-shifting the heating/cooling load to match the availability of cost effective PV solar. But hydronic output is also available with air source (for example, Chiltrix).

    1. TNREM | | #24

      Hey Jon, I'd actually never heard of the Chilltrix system until a couple of days ago when I saw it mentioned on this site. However, I also saw a post where someone was looking for 'real' reviews and many of the comments indicated that perhaps Chilltrix wasn't quite ready for prime time. I love the idea, but I'm pretty sure I don't want to be on the 'bleeding edge' of that.

      Regarding costs, my main concern is not having the genny kick on when we run the AC on warm summer nights. It's not just the cost of the propane but nightly running really puts unnecessary hours on the genny, means more maintenance (oil/filter changes, etc.) and is generally more of a hassle. On the other hand, we ran it this summer to support the little window unit, because it takes a braver man than I am to tell my wife she has to sweat all night because I don't want to burn 5 bucks worth of propane. LOL

      1. Jon_R | | #26

        There are other options - for example, Spacepak Solstice. Would be interesting to see your economic analysis if you do one.

        1. TNREM | | #28

          Hmm! Spacepak Solstice? Haven't heard of that either. I'll check it out. Thanks!

  10. joshdurston | | #18

    If you're totally off grid, modulating "inverter" style compressors are the way to go to eliminate the inrush spike, it takes a pretty big AC inverter to start a conventional compressor with an "across the line" contactor start.
    Inverter is pretty much standard for mini splits, and still a rarity for GSHPs, and even then you might have a tough time finding one small enough to suite your needs. Plus the cost of GSHPs isn't going to be competitive with ASHPs.
    You may want to go with a air to water style heat pump with a fan coil (or infloor heating) and a big storage tank. Store lots of hot or cold water while the sun is shining, and use a low power ecm pump and ecm fan coil to trickle it out as necessary.

    1. TNREM | | #25

      Josh, you're correct that inrush current is a concern for refrigeration compressors and that the inverter technology in mini-splits have the edge there. That said, there are after market 'soft start' kits you can buy as add-ons for AC compressors not so equipped. However, for all the reasons I've given elsewhere, I think the mini-splits are the way to go. Regarding in-floor heating, big storage tanks, etc., I'm afraid that ship has pretty much already sailed. The house is built (or 'installed' I guess, since it's a manufactured home) and most things are not easily changed at this point. Right now, I'm just looking for a way to 'make it through the night' on my existing battery bank, and it looks like the zoned mini-splits the most viable solution.

  11. walta100 | | #19

    Being off grid what is the plan to power the system?

    General off grid solar struggle to have enough battery storage to get thru the night and 3 or more cloudy day will get a generator running on systems that are not providing heating.
    A battery large enough to run a mini for 3 days is likely larger than the cabin and costs 5X what the land would sell for.


  12. TNREM | | #27

    Hey Walter,

    The solar is already in place and has been working without problem for a couple of years. System specs as follows:

    6.7 KW of panels
    8800 Watts of inverter (2 – 4400 units slaved together)
    24.5 KWh LiFePo battery bank (Simpliphi)
    12 KW Kohler generator on auto-start for back up

    FYI, we currently have a propane furnace for heat, cooking, and hot water, and for the last two years, we've only had to crank up the genny to 'exercise' it, even with long periods (5 or 6 days) of overcast days. That all changed when we spent our first summer here 'full time' this year, and decided we needed some AC. As a temporary 'stop gap' I spent $400 at Lowes on the biggest capacity (14K btu/h) 110 VAC window unit I could find. We hung it in the window of one bedroom and circulated cool air through the house with 4 box fans. (VERY inefficient, but the house is tight and well-insulated and it kept us cool, even on the hottest days.) On a sunny day, we power the AC and everything else just fine and still generate enough power so that the battery bank is full and floating by anywhere between 1 and 3 PM

    Obviously, all bets are off when the sun goes down. Even with a full battery bank, the AC will suck the battery bank down and the genny will kick on somewhere between 3 AM and 6 AM to charge the batteries. That is, we can 'almost' make it to sunshine but not quite. Therefore, I'm quite sure that with a more efficient AC and 'spot cooling' (i.e. only cooling the occupied bedrooms) that we'll be fine.

    As far as heating goes, we'll use the heat pumps if the sun is shining or fall back on the existing propane furnace if it's not. My focus is the AC during our hot, muggy, southern summers. The heat from the heat pump is just (to my thinking) icing on the cake, so to speak.

    Obviously, I know we can't run a mini (or any AC) 3 days (or even one day) on batteries alone. As I said, the genny is set up on autostart tied to battery bank 'state of charge.' My goal is to stay cool during summer nights without having the genny kick on.



Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |