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

Is a Ground-Source Heat Pump My Best Bet?

With an energy retrofit on the horizon, a homeowner ponders the cost and suitability of a ground-source heat pump

Ground-source heat pumps use the earth as a source of heat in winter, and as a place to dump excess heat in summer. A GBA reader wonders whether this is the right option for her.

Cathleen Dalmeida is budgeting for a heating and cooling system as part of an energy retrofit and is wondering whether a heat pump is part of her future. An obvious question: How much do they cost?

“Is there a general rule of thumb for pricing of a ground-source heat pump and air-to-water heat pump for a medium sized installation?” she asks in a Q&A post at Green Building Advisor.

She wants to know whether anyone can offer an estimate in dollars per ton of heating and cooling, and whether there would be any drawbacks to using a ground-source heat pump in a climate that doesn’t require much cooling.

That’s the topic for this Q&A Spotlight.

Ground-source heat pumps are not the cheapest option

Dalmeida has provided no details about the house, or the climate zone where it’s located, and as a result, it would be impossible to provide any specifics on a heat pump. But, GBA senior editor Martin Holladay points out, “In general, ground-source heat pumps are quite expensive, and aren’t the most cost-effective way to heat or cool a home.”

In an energy retrofit, work typically starts with air-sealing, followed by an insulation upgrade and possibly improvements to existing windows, Holladay says, and heating and cooling loads should be determined with a Manual J calculation.

“In most cases,” he says, “you won’t be using a ground-source heat pump for this type or project.”

Cathleen then asks a follow-up question: “If someone already has radiant floor heating, or wants to install radiant floor heating, what type heating technology would you use?”

Holladay answers that these systems usually include a boiler, which can run on one of several types of fuel. It’s also possible for a radiant-floor system to be built…

GBA Prime

This article is only available to GBA Prime Members

Sign up for a free trial and get instant access to this article as well as GBA’s complete library of premium articles and construction details.

Start Free Trial


  1. Expert Member
    Dana Dorsett | | #1

    Seriously? Coupeville!?!!
    Coupeville is on Whidbey Island, about 5 miles north of one of my relatives' house in Green Bank, where they are heating a similar sized early 1990s vintage house with one 1.5 ton ductless mini-split, installed at a cost of about $4500. They have been laughing out loud every time they pass the propane truck on the road (though they're still using propane for cooking.)

    This is a location with a 99% outside design temp of about +20F, and heat load ratios of ~10BTU/hr-ft^2 or less even for code-min houses. Newer-better mini-splits will deliver a seasonal COP of about 4, in that location (compared to about 3.5 for the one installed at the relative's house in Green Bank.) A somewhat tightened up 1500' house in Coupeville would have a heat load load well within the output of any cold-climate 1-ton mini-split or the better 1-ton mini-duct cassette type mini-splits (or even the 3/4 ton Fujitsu -9RLS3 ), but upsizing to 1.25 ton or 1.5 tons would not necessarily be insane, depending on how the heat load calculations work out.

    This is the also land of cheap 99% carbon-free electricity. If you can't give up the cushy bare-foot comfort of radiant heat, install a ~2kw electric boiler slaved to a floor thermostat to 71-72F for foot comfort and install a mini-split solution to control the room temperature, set to a comparable temp. It's really the best of both worlds! (It may take 4-5kw of electric boiler if you want that to be 100% backup for the mini-split.) The electric boiler + mini-split should still come in under $7K, maybe even under $6K. The electric boiler reduces the average efficiency of the heat pump from a seasonal average COP of 3.5-4.0 down to 3.0-3.5, but so what?

    Spend the difference in upfront cost of the 1-1.5 ton GSHP solution and electric boiler + mini-split solution on photovoltaic solar. The net power use will likely be LESS than what you would with a best-practices GSHP, without the inherent system design risk that comes with GSHP.

    It's possible to estimate the absolute whole house heat load if you have mid to late winter fill-up bills on the propane, with the EXACT dates of the fill-ups, and the nameplate efficiency of your existing boiler. The average thermostat setting during that period is also useful if you need higher accuracy.

  2. user-1111594 | | #2

    analysis of 100,000+ hours of gshp system performance
    Thanks for sharing this.

    There's a lack of GSHP system verification data pulled from real time data. If anyone is curious, there's a case study here that looks at a number of system in New England:

  3. Expert Member
    Dana Dorsett | | #3

    Thanks for that reference!
    Downloading the document:

    It's easy to see how the case for ground source heat pump (GSHP) is getting ever harder if you look at Table 1, p.4. With the best-case variable speed compressors the monitored systems are achieving an seasonal average COP of about 4.2.

    In a climate such as Coupeville WA a ductless Fujitsu ASU/AOU 9RLS3 (costing less than $4K, installed price) will pretty much hit that efficiency. Even the better-class 1.5 ton mini-ducted versions will hit the mid-3s in that climate, which is where the next-best GSHP systems being monitored in that New England study were performing (at comparable deep-well & deep-subsoil temps as western WA.)

    While there may have been a financial rationale for complex & expensive (and design-risky) GSHP systems when air source heat pump systems were operating in the low 2s (or sub-2) for seasonal COP efficiency that rationale has evaporated in most US climates, and the financially-rational line is creeping northward with every new release of cold-climate mini-splits.

    There's NO case for GSHP for small to mid-sized house on Whidbey Island. Even the mini-splits of 5 years ago were averaging comparable seasonal efficiencies to the middle-of-the-road GSHP, according to NEEA's in-situ monitoring. See the Willamette and Puget Sound cluster averages performance in Table 1, p10:

  4. user-2890856 | | #4

    would one partner solar thermal with GSHP unless they were using a storage strategy ? Unless they were planning on using the thermal excess after DHW production to charge the earth you should choose one or the other unless you want to hedge against many situations .
    This individual already has radiant floors which she seems to enjoy and would like to keep or so it seems . Solar thermal with properly sized storage below ground or in a basement is much less costly than Vertical bore holes or almost always done wrong horizontal slinky loops . Probably able to do without any heat pump if someone capable of doing the math performs the calculations . This need not be over complicated with controls that boggle the mind and create problems .

    The other thing that has me wondering about many comments and advice given to folks is this . Most who are researching installing a particular technology that may be more costly than some other technology are more often than not aware that it costs more . Why is it that many often try to change their minds and steer them another way . When was it that our job became to be the guardian of one's checkbook as opposed to giving them sound advice based on fact and not myth pertaining to the technology which they came seeking advice on ?

    I am well aware of the problems and unfortunate happenings with earlier radiant installs , designs , high cost . We are working on eliminating those practices as an industry . What one may choose for his home is not what another wants . I practice what I preach and if I cannot offer good advice because my knowledge of a particular subject is not what I consider expert I keep quiet . If the subject of something I have heard bad things about comes up , I listen and see if there is something I missed in what I heard . When will papers such as ALL ABOUT RADIANT FLOORS , SOLAR THERMAL IS DEAD (and really , really dead) written by folks who don't know all about radiant floors or solar thermal or at a minimum have not had the pleasure to see it done properly stop being referenced as if they are the HOLY GRAIL and end all , be all papers on the subjects .

    How can we even begin to make any recommendations on size and type of equipment with the little information that we have ?

  5. GBA Editor
    Martin Holladay | | #5

    Response to Richard McGrath
    You wrote, "Solar thermal with properly sized storage below ground or in a basement is much less costly than vertical bore holes or almost always done wrong horizontal slinky loops. Probably able to do without any heat pump if someone capable of doing the math performs the calculations."

    The problem is not feasibility. The problem is simply cost. Sure, you can install 8 or 10 solar thermal collectors, each 4x8, and a buried, insulated 1,000 gallon tank. Lots of people have done it. I'm simply asking, "How much does the equipment cost? Does the investment make any sense?" The answers are clear: "a lot" and "no."

    You assume that I've never heard of these systems. But I have. Hoyt Hottel built a house like the one you describe at MIT in 1939; I learned about Hottel's research in 1973, at a Yale engineering class taught by Arvid Herzenberg. You can draw a straight line from Hottel's 1939 house to Peter Amerongen's Riverdale project a decade ago -- and there were dozens or hundreds of similar homes in between.

    Your ideas are not new or untested. They are just extremely expensive compared to much simpler ways of keeping a house warm.

  6. user-2890856 | | #6

    Martin ,
    In comparison to GSHP with VBHs the technology makes absolute economical sense .
    This person has radiant and apparently wishes to utilize it . Do you have any advice that is useful to the radiant subject ? Do you believe and are you advising that GSHP is more cost effective and has a better LCCA than solar thermal ?
    Contrary to popular belief or appearances , I do not come here to debate with you . Lately , however , it seems that way to me . Now i come here to help your readers make INFORMED choices and often find it difficult to help them for debating you . At the end of the day though maybe this is good because they will research further .
    In closing . Is solar thermal a better more economical choice than GSHP partnered with a wwhp?
    But know this . Considering where she is located and average temps , an air / water heat pump may be a nice fit for this home using it's radiant floor heating .

  7. GBA Editor
    Martin Holladay | | #7

    Response to Richard McGrath
    Q. "Do you have any advice that is useful to the radiant subject?"

    A. If someone has an existing radiant-floor heating system, there is no reason to change it if they are happy with it. Here in New England, these systems are common; most New England systems are heated by oil-fired boilers. I don't think that Cathleen Dalmeida said that she has any problems with her radiant-floor system. If it works, she should keep it.

    Q. "Do you believe and are you advising that GSHP is more cost-effective and has a better LCCA than solar thermal?"

    A. For most homes, the investment in either a ground-source heat pump (GSHP) or a solar-thermal space heating system is hard to justify compared to cheaper alternatives, but there are exceptions. The owners of large single-family homes located in areas where GSHP contractors quote low prices may find that these GSHP systems make sense, especially if they can claim a 30% federal tax credit. I've never seen an active solar-thermal space-heating system that was cost-effective compared to other ways to heat a house.

  8. Expert Member

    You wrote:
    "Why is it that many often try to change their minds and steer them another way . When was it that our job became to be the guardian of one's checkbook as opposed to giving them sound advice"

    Wouldn't it be fair to say that you consistently champion radiant heating against any of the other alternatives? Surely that's the same behaviour you are complaining about?

  9. GBA Editor
    Martin Holladay | | #9

    More on "the guardian of one's checkbook"
    Amory Lovins used to say that people don't really want a refrigerator -- all they really want is cold beer. It's easy to forget what we are aiming for.

    Homeowners want to be comfortable. If they can heat or cool their house comfortably for much less money using System A than System B, I'm not sure why anyone would advise them to spend more on System B.

  10. user-672886 | | #10

    air-to-water heat pump
    Since the radiant distribution system is already in place, why not consider an air-to-water heat pump? Solar-PV-friendly and less complicated than electric boiler + mini-split or solar-thermal, no?

    Here is a new North American model designed for hydronic heating:

    and there are others discussed in this thread:

  11. user-917153 | | #11

    Thanks for references
    Thanks Dana and Chris for the links to the monitoring study. COPs roughly in the 3 to 3.5 range were seen in this study, "Residential Ground-Source Heat Pumps: In-Field System Performance and Energy Modeling" by Puttagunta, et. al.:

    Plug those COPs into a heating cost calculator such as:

    Or, to compare alternatives, or to run scenarios using different assumptions.

    I like Bruce Harley's advice in his book, _Insulate and Weatherize_: "They are also quite expensive to install: my preference is to invest in a more efficient building enclosure"

  12. user-2890856 | | #12

    Radiant and refrigerators
    Malcolm ,

    I speak about radiant when that is the topic of discussion . I also design systems that use a water heater to feed hot water coils in ducted systems where people have the capability of utilizing outdoor reset to the coil , vary air temp , and condition all the air in the house . I sell comfort Malcolm and comfort is not a number on a thermostat , nor is it hearing your equipment . My theory is simply that heating and cooling systems should not be designed for homes but for the occupants . Far be it for me to steer their decision unless it is a terrible one arrived at by garbage opinions from the internet based on poorly designed , if at all and installed systems .

    Martin ,
    Amory also spoke of the negawatt . The systems I design , specify and build use less energy than the things you often recommend . I don't use theories from the first half of the 20th century nor would I ever reference them since everything is so much more capable today . I can buy a refrigerator for 450.00 and keep my beer nice and cold if I drank . But you would have me buy a 1,000.00+ refrigerator because it costs 22.00 less a year to operate . Could you give me the LCCA on that and have it make sense ?

    Your comments are always so general and blase . Where do you come up with a thousand gallon tank and thsi and that without knowing what would be required . Hell , Thorsten's house in Fairbanks only uses a 5,000 gallon tank and IT WORKS . My guess is that Cathleen's home could quite possibly do well with a tank under 500 gallons and 1/2 the panels Thorsten has , and a bit of math by someone knowledgeable may be just the ticket . I guarantee that it is less expensive to install than the GSHP option , and that was her question . Don't forget the part where I said in her climate and area an air / water heat pump would probably be the best all around decision she could make . This also gives her the option of radiant cooling if she will be using a DOAS system also .

  13. GBA Editor
    Martin Holladay | | #13

    Response to Richard McGrath
    You have told me, "Your comments are always so general and blasé."

    With your help, Richard, I'll try my hardest to be less general and increasingly specific. As for becoming less blasé, that's somewhat of a head-scratcher. I suppose that one way to do that would be to become enthusiastically naïve. That's a challenge, for sure -- but I'll do my best.

  14. Expert Member
    Dana Dorsett | | #14

    Thanks Jonathan
    I'd seen the Puttagunta et al study of three systems previously. GSHP pros I've had look at it have commented that the designs were not well-optimized for minimizing pumping power, so they're not exactly best practices. But there's nothing I've seen that would make me think they don't represent the typical.

    Even the smallest of the houses in that study is nearly 2x the size of the Coupeville WA home. The Virginia home that was in a zone 4 climate (like the Coupeville home), but was more than 4x the size, and in a location with a somewhat lower outside design temp. But they were all higher performance houses (HERS 21, 33, & 42) unlike what I'd expect for the Coupeville home, so the loads and system size requirements may not be very different. The cheapest of the three systems cost $28K. Typical seasonal COP The Coupeville house could get the same net operating efficiency using an electric boiler to keep the floor slightly warm, with a mini-split to keep the house warm (carrying the bulk of the heat load) for under $10K.

    The cheapest air-water heat pump systems sufficient to cover the likely loads would all be well over $10K, and would only be marginally more efficient. There's an outside chance that the heat load of that house would be within the output of a Sanden EcoCute air to water heat pump which would also be under $10K, but that's not too likely, and if it needs an electric boiler as backup, it would be more expensive than the mini-split + electric boiler solution. And the ~$20K cost difference buys enough rooftop PV to more than offset any efficiency deltas, even at the wretched wintertime insolation levels they receive in that location.

    The 4 ton SpacePak Harry Voorhees pointed to costs $8-9K just for the heat pump, and is more than 2x oversized in capacity @ the +20F outside design temp for the likely load of this ~1500' house, and runs at a COP in the low to mid-3s, lower than a better-class mini-split would in that climate. Their 3-ton version has a more appropriate +20F output, but only delivers a COP of 2.2-2.5 at the average wintertime temp in that location, and costs about $7K just for the heat pump.

    In the land of 10 cent electricity & low wintertime insolation, solar thermal enhancements or air-to-water solutions don't have a very good financial rationale either.

    Cost WAS the original question: "Is there a general rule of thumb for pricing of a ground-source heat pump and air-to-water heat pump for a medium sized installation?"

    Best guesstimates:

    $25-35K for a GSHP solution

    $15-20K for a pretty-good air to water solution such as a Daikin Altherma, delivering comparable or better efficiency than the GSHP.

    $7-10K for an electric boiler + mini-split solution of somewhat lower efficiency. The as-used efficiency is highly dependent upon how high one cranks the floor thermostat relative to the mini-split's setpoint- it's directly under user control. If taking that route it's worth buying the appropriate wall thermostat to achieve more precise control of the mini-split to avoid any unpleasant surprises on power bill. But it won't be tough to beat the propane boiler by quite a bit on heating costs.

  15. user-948834 | | #15

    Take the money for a gshp and
    Take the money for a gshp and put it into insulation, airsealing and better windows, (R-5 or better) and get an air to air heat pump from Mitsubishi and call it best.

  16. Expert Member
    Dana Dorsett | | #16

    But Tim...
    R5 / U0.20 or better replacement windows in an existing home are ridiculously non-cost-effective, even against $0.10/kwh resistance electricity in a zone 4 climate with a +20F 99% outside design temp. It's even less cost effective than solar thermal here, and the daylighting hit in that foggy-rainy location would be noticeable. The load is almost certainly already within the +20F output capacity of a 1.25-2 tons of cold-climate mini-split, even without envelope upgrades (TBD). If the windows need upgrading for either comfort or energy use reasons, the retrofit treatment that might make sense would be hard-coat low-E storm windows over whatever they currently have, which would be a tiny fraction of the cost of replacing them with U0.20 windows.

    And they DO seem to want the cush-factor of warm floors. The cost delta between a Daikin Altherma and a mini-split + electric boiler solution may or may not pay for sufficient rooftop PV to cover the power use difference, which depends on just how warm they run the floor relative to the mini-split setpoint. But on Whidbey Island even a 100% electric boiler solution would be both cheaper to run and greener than the pre-existing propane boiler at recent years' propane pricing in island locations.

Log in or become a member to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |