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

Heat-pump water heater or something else?

SteerageFlyer | Posted in General Questions on

We moved into our house last fall and I’m in the process of figuring out what type of hot water to replace our 25 year old standard vent unit with.  We can’t do a direct replacement because I don’t believe it was installed correctly.  To give you a little background, the hot water tank is located in a utility closet (with the boiler) in the garage.  The garage is actually below grade (really common in our area) so it’s kind of like being in an unconditioned basement other than all the waste heat coming off the boiler and associate piping).  There is natural gas to the closet (for both the hot water heater and boiler).  There is no 220v however.  We live in the Seattle are so our electricity actually comes from 90%+ green sources.

So the options I am considering are 1) natural gas direct vent 2) natural gas condensing power vent 3) electric heat pump water heater (HPWH) 4)on demand natural gas.  All have pros and cons and I’m trying to figure out what makes the most sense both from a carbon footprint and cost perspective.

For the HPWH I’d have to run get a 220 line run to the garage and I figure that’d cost me around $1K or so given the location of the panel and long run.  And I’d have to decide if I want to pull external air or vent the cold air outside (probably vent the cold air outside).  But this is probably the greenest of all the options, but not sure about operating costs in the cool/cold seattle winter and not really very warm summers.  May need a condensate pump depending on the outlet height.

The condensing unit would need a condensate pump.

The direct vent would probably be the most simple replacement but by far the lease efficient.

The on demand would be the most efficient natural gas, but I don’t have a large enough gas line over there so I’d have to run a new gas line as well and that won’t be cheap either.  If I went direct vent I could also probably replace the boiler with a higher efficiency combo unit but that’s probably pushing over 10k and the boiler is working mostly fine.

Anyway, I’m just wondering what you all think and if you have any suggestions or recommendations.  Thanks in advance!!

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. user-2310254 | | #1

    Martin Holladay recently published an article that might be helpful:

    FWIW. I decided to replace my natural gas water heater with a HPWH before reading this article. (I live in one of the few states where electricity is cheaper than natural gas for heating water.) For me, removing combustion sources from my home was a bigger objective than lowering my hot water costs.

    1. SteerageFlyer | | #4

      Thanks! I'll take a look at that.

      Can I ask which unit you went with and how you like it? Did you vent or pull air from outside?

      1. user-2310254 | | #8

        I have a Rheem 80 gallon HPWH, which is overkill for a two-person household. But the distributor sold it to my plumber for about the same price as a 60 gallon. It definitely chills the garage during the winter, but I'll probably skip venting the outside. I have been surprised by how infrequently the heat pump runs.

        1. SteerageFlyer | | #13

          Yeah I was thinking about going big potentially and having the spare capacity. My wife and daughter both like really hot showers/baths so having the extra capacity might be good.

          I figured i might vent outside, in part because my water heater shares space with the boiler and there is loads of copper in that room as the boiler water goes to all the different zones in the house. I'm afraid of essentially dumping that cold air into my home heating system.

  2. Jon_R | | #2

    Only partially related, but note that the most cost effective option is usually low flow (say 1 GPM) shower heads.

    1. SteerageFlyer | | #3

      Totally agree, but I gotta ease the family into that. :-)

      1. burninate | | #11

        I don't see why you couldn't do an energy-efficient option for high-flow shower heads, a drainpipe thermal exchanger, and a big thermal storage tank - so the actual water heater is a low-powered heat pump (maybe an extremely low-powered one), but you get 200-500gal of hot water capacity in reserve..

        1. SteerageFlyer | | #14

          Not sure I have enough space for that though.

  3. Expert Member
    Dana Dorsett | | #5

    A HPWH can still live in the closet if it's ducted to the garage, but it needs access to large quantities of air not found in a small utility closet. It will have the additional benefit of drying out the garage air (not that high humidity in semi-conditioned spaces is particularly common in the PNW.)

    There is yet another option not mentioned in your list of four:

    The 5th option is to heat the domestic hot water with the boiler using an indirect-fired tank operated as the priority zone. Almost all hydronic boilers are oversized for the heat loads of normal sized homes in Seattle, and increasing the duty cycle by giving it more load improves the as-used net efficiency as compared to a standalone (even tankless) gas water heater.

    To assess that option...

    What boiler make/model is it?

    What is your 99% design heat load? If you're not sure, run a fuel use based heat load calculation using wintertime-only gas bills following this methodology:

    1. SteerageFlyer | | #6

      Yeah I had thought about the airflow issue some. Figured I'd pull air from the garage and exhaust air out the side of the house. But was still debating that.

      Excellent point. I had thought about indirect as well and should have included it on the list. The contractors I had give me an estimate for that wanted like 5k to do the work. One didn't want to touch it. I'm not sure I'd be comfortable doing that work on my own like i would with the other options.

      I should do the calcs and get the boiler info. Ill try and do that today.

      1. Expert Member
        Akos | | #9

        Heating hotwater using an indirect is pretty inefficient. Even if you have a modcon, you are looking at around 80% efficiency as the return water temperature is too high to get much condensation. This is about the same efficiency as a power vented tank, which is way cheaper to buy and install.

        If you want semi DIY friendly install, one option is to go with a larger electric tank with 4kW elements. You can connect this to your existing 120V, it will just mean you get 1/4 the power. A typical hotwater tank uses 13kWh per day, so the tank at 120V will just run about half a day to recover. As long as the tank is big, recovery time won't matter, it should not run out.

        You can also connect a sidearm heat exchanger to the tank. Put sidearm in series with outlet of your boiler. This way during the heating season most of your hot water will come from the boiler.

        1. Expert Member
          Dana Dorsett | | #17

          >"Heating hotwater using an indirect is pretty inefficient. Even if you have a modcon, you are looking at around 80% efficiency as the return water temperature is too high to get much condensation."

          In real world residential hot water draw profiles 80% is about where a condensing tankless falls, due to the flue purges and ignition losses on low volume draws. One might still characterize it at "...pretty inefficient...", but has to be considered in the context of "Inefficient compared to what?"

          With a lower storage temp (125-130F instead of 140F) it's possible to get steady-state efficiency in the min-90s with an indirect and a mod-con, and an average efficiency in the mid-80s, comparable to or better than a condensing tankless.

          >"This is about the same efficiency as a power vented tank, which is way cheaper to buy and install."

          Gotta call BS on that one. The STEADY STATE efficiency of non-condensing power vented tank is about 80%, but the average efficiency (after standby and other losses) is lower.

          It's hard to find a non-condensing power vented water heater with a UEF higher than the mid-0.6s. Even most condensing tank water heaters don't come in much higher than 0.8, though a few hit the high 0.8s and a few in the low 0.9s.

          Also, any decent indirect will last more than 20 years, unlike a standalone power vented water heater. An all stainless condensing tank heater can last that long though.

          The 50 gallon all stainless "Phoenix Light Duty" comes in at a UEF of 0.90, but would be more expensive to install than an indirect (just the water heater is over $2K f.o.b. the distributor's warehouse):

          While it's possible to use the Light Duty as a combi unit to also heat the house, that is really only something to consider at boiler replacement time.

          1. Expert Member
            Akos | | #21

            Maybe its a Canadian thing, but there are plenty of power vented tank units with mid 70's UEF efficiency. Those run in mid 80's for steady state efficiency.

            For the cost delta, you would have to use a lot of hot water to justify the additional cost of an indirect.

            I've also had issues with scale buildup with indirect coils over time and loosing capacity. The scale buildup also makes the RWT even higher, a double whammy for capacity and efficiency.

            Overall, your idea of replacing both boiler and DHWT with a condensing tank is probably the lowest cost/high efficiency solution.

      2. Expert Member
        Dana Dorsett | | #10

        >"I had thought about indirect as well and should have included it on the list. The contractors I had give me an estimate for that wanted like 5k to do the work."

        That would be a serious gouge. An indirect isn't cheap, but it's not a huge project. Most contractors doing that work regularly could be in and out in a day.

        A possible fly in the ointment for your option #4 (the gas tankless on-demand solution) would be the capacity of the gas meter. In Seattle a 2-bathroom house really calls for a 180-199KBTU/hr burner. Depending on the boiler and other gas appliances it could require a service upgrade. In most cases there would need to be a dedicated home-run gas line of sufficient size routed between the tankless and the meter/regulator too. Between the additional gas plumbing and venting on top of the raw cost of the unit, of installing a tankless is usually more than the cost of installing an indirect onto an existing hydronic heating system.

        1. SteerageFlyer | | #15

          >"That would be a serious gouge. An indirect isn't cheap, but it's not a huge project. Most contractors doing that work regularly could be in and out in a day."
          Yeah I thought $5k was crazy high too. Got that quote from two different installers. I'm always amazed at how expensive all installs/repairs of anything is in Seattle. It's kinda mind boggling.

          >"Heating hotwater using an indirect is pretty inefficient. Even if you have a modcon, you are looking at around 80% efficiency as the return water temperature is too high to get much condensation. This is about the same efficiency as a power vented tank, which is way cheaper to buy and install."

          The bright side is endless hot water for not as much cost, theoretically. I'm not sure o the efficiency. Won't my boiler just have to fire less often as the hot water returns back to the boiler. I did remember seeing something about the turndown ratio though and that it could possibly stress the boiler, particularly in the summer when the hot water is the only demand.

    2. SteerageFlyer | | #16

      >"What boiler make/model is it?"
      So the boiler is 80kbtu in and 67kbtu out it seems. Is that as simple then as knowing it is 83% efficient (67/80?). I still need to do the heat calcs though.

      1. Expert Member
        Dana Dorsett | | #19

        Yes, the steady state efficiency is 83%, but with standby and distribution losses it's lower, due to the low duty cycle, further degraded due to being installed outside of conditioned space, where standby losses are truly lost. A typical 2 x 4 framed house with 2000' of conditioned space and clear-glass double-panes would come in under 25,000 BTU/hr, and smaller houses even lower. The DOE output of that thing is 67,000 BTU/hr, so even on design day the duty cycle is less than 25/67= 37%.

        The I=B=R net rating is 58K, which is a fudge factor invented to account for the standby losses being truly lost when installed outside of conditioned space. So if it were right-sized per the AFUE standard (1.7x the load at the 99% outside design temp) the as-used efficiency would be 58/80= ~73%, (sorta-kinda) but your oversize factor is probably closer to 3x, and the lower duty cycle delivers a even lower as-used AFUE.

        If the thing is 30 years old (looks like it could be approaching that) the steady state efficiency may have slipped a bit too, but at such a low duty cycle maybe not. If you would be considering replacing it with a more appropriately sized condensing boiler (or turning a condensing tank water heater into a combi), run the napkin math zone-by-zone on the radiation too, in addition to the heat load calc:

        With the heat load number and the radiation sizing it's possible to estimate the water temperature requirements to see if it would be in the condensing mode the majority of the time. That is very likely to be the case, if the radiation is as oversized for the load as the boiler.

        1. SteerageFlyer | | #28

          Sounds like I have some math to do. I don't really mind the lost heat all that much as it does keep the garage a little warmer in the winter months.

          1. Expert Member
            Dana Dorsett | | #29

            If you're going to heat the garage *even "passively" with standby losses), insulate it.

          2. SteerageFlyer | | #30

            Totally agree. We live in a 1940s house in Seattle. Insulation was an after thought. We have several ecterior walls in our house that have no installation in them at all. It's on the list of things to do. Attic insulation is in good shape.

  4. Jon_R | | #7

    Probably too expensive, but a Chiltrix (or similar) heat pump can eliminate (or greatly reduce) your burning of fossil fuel for space and water heating.

    1. SteerageFlyer | | #18

      So I hadn't heard of those, but it looks like they would also provide A/C for the house wherever I elected to put in an appropriate units. That looks to be a pretty spiffy (i.e. expensive) system.

      1. Expert Member
        Dana Dorsett | | #20

        You'd have to run both the heat load numbers and the radiation numbers to know if it's possible to even run the existing system with a Chiltrix, let alone run it as efficiently as a mini-split. Having local support (or a willingness to wait if it ever craps out) would be important for going that route.

        Part of your problem finding contractors willing to do an indirect at a reasonable price is the fact that hydronic heating is far less common in your area than ducted hot air, so there is an inexperience factor. Some of those bids may have come from contractors who had zero or very little experience installing an indirect water heater and needed to cover the potential down side.

        In my area most houses are heated with hydronic or steam boilers. About a decade ago I had an entire hydronic heating system installed that included an indirect, an 80KBTU 85% efficiency boiler and a bunch of baseboard for about $8K in one of my wife's family's rental properties. That also included legal disposal of the asbestos encrusted 100 year old steam boiler an asbestos insulated plumbing. Even at east lake upcharge pricing $5K for just an indirect is crazy.

        Look at Table 3 in this document and assume 2-3x oversizing, compare the steady state efficiencies to the net average efficiency at 3x oversizing, ignoring the fuel type (it's comparable between cast-iron gas vs. cast iron oil.)

        Your steady state efficiency is about 83%, which makes it most comparable to System 1, even though you'd be using an indirect, not a tankless coil. Rounding down, that means you might expect 70-74% efficiency out of that boiler + indirect as an annual average. If you're keeping the Vaillant for more than a couple of years, adding an retrofit heat purging controller such as an Intellicon HW+ or Fuel Smart HydroStat 3250-Plus would reduce that efficiency hit substantially bringing the net efficiency north of 80%.

        Take a look at how little the oversizing hit is with system #3, the only system in that lab test with heat purge control. That's going to be "worth it", if you're keeping the boiler for more than one more heating season. (That Brookhaven Nat'l Labs program is why almost all new cast iron boilers now come fitted with heat purging controls.) It's DIY-able for under $200 for those with electrician skills.

  5. FrontierEnergy | | #12

    Besides which, air-to-water heat pumps have an outdoor condenser/evaporator, so no chilling the garage, and they deliver space heating and cooling as well. Chiltrix is fairly affordable, but tariffs are driving up the cost.

  6. tommay | | #22

    I would install a direct vent water heater. Cheaper, easier, no moving parts or electricity needed. That way there if power ever goes out, you still have hot water. They also make small, external heat ex-changers that work off a boiler zone. You could pipe/circulate water from the hot water tank (out the relief valve tapping then back into the drain tapping) into this ex-changer along with a pump and aquastat to use during the winter when the boiler is running, hopefully keeping the gas off in the tank to save some money. Then you have the tank itself to use in the summer when hot water demands are less and the boiler is off.

  7. Expert Member
    Dana Dorsett | | #23

    >"Maybe its a Canadian thing, but there are plenty of power vented tank units with mid 70's UEF efficiency. Those run in mid 80's for steady state efficiency."

    Show me!

    The Energy Star listings for gas water heaters stop at 0.73 for non-condensing units, making a quantum leap to 0.80 for the crummiest condensing versions (maybe this ridiculous link will work... or not):

    FWIW: Seattle's water is soft enough that I don't have the slightest concerns about scaling on a indirect's heat exchanger (or tankless, for that matter.) Even those on wells in greater King County tend to have very little mineral content to the water (a benefit of high local rainfall and snowmelt replenished artesian aquifers.) In parts of Washington state east of the Cascades it can start to be an issue.

    1. Expert Member
      Akos | | #24

      Check out:

      0.72 EUF with 0.86% recovery efficiency. It is a pretty good unit for budget combi DHW/space heat for low load apartment.

      My issue with indirect is the efficiency is dependent on install and setup. Theoretically you can hit high 80% efficiency, I doubt that happens often in the real world. For what its worth, from my experience, an indirect uses about 40% more natural gas in the summer than a condensing tankless (1.3 m^3/day vs 2.3m^3/day).

      1. Expert Member
        Dana Dorsett | | #25

        A condensing Phoenix Light Duty is only ~USD$400-$500 more than the GPDT 50, has an all stainless tank, a quieter modulating 76K burner and a 3:1 turn down ratio, and UEF of 0.90 in hot-water only mode (higher as a combi-heater), making the upcharge "worth it" when kludging together a combi system for a 2 bathroom house.

        For a 1 bath apartment a condensing stainless RGH20-76F with a UEF of 0.89 and a 2:1 turn down modulating burner is priced comparably to the GPDT 50 (often cheaper), and at less than 4o" tall- small enough to tuck away in some pretty tight places:

        The burner is big enough to support a continuous 2gpm shower with enough left over to cover the heat load of a low load apartment without needing more heat buffering mass than the 20 gallons. It'll fill a standard sized tub faster than a 199K tankless, but it won't fill a big soaker tub in a reasonable amount of time. There's a 100K burner version too, if the combined load makes a 76K condensing burner marginal, and could even serve some 2-bath apartments if one of the baths was shower-only, no tub.

        Despite moderate wintertime temps, Seattle's heating season is 10-10.5 months out of the year, so the summertime water heating only efficiency hit of an indirect isn't as relevant as it is in some other locations. The higher net efficiency delivered by a higher duty cycle on the oversized boiler would more than make up the difference of lousier summertime performance during the other 40 weeks out of the year.

        1. Expert Member
          Akos | | #26

          The HTP units are great, they are just pretty hard to get up here. I also think they would fall into the same category as a tankless, when something goes wrong the plumber will just want to swap out the unit. With a power vent, at least you can get parts like a blower replaced. Also no need to deal with condensate.

          Summer or winter, heating the DWHT with an indirect will still have the same issue. With a priority zoned indirect the RWT (135 to 140) I've seen a modcon will run at best at mid 80's efficiency, maybe with larger draws the efficiency will be better, but the average is still not that good. The RWT temperature gets even worse when re-heating to make up for standby losses.

          Properly set up, it should be better than a power vent, but that is a lot of cost for a bit less gas use, the payback is in range of decades. I can't comment on how much fuel reducing the cycling will save, so maybe on that front there might be a reasonable payback.

          Bottom line, if you want to heat water with dinosaur farts for cheap, stick to a direct vent/power vent.

          If you want to do it cheap and efficient, can't beat a tankless unit.

  8. Expert Member
    Dana Dorsett | | #27

    >"The HTP units are great, they are just pretty hard to get up here."

    In SteerageFlyer's Seattle the Westinghouse branded versions of the HTPs are available through the big orange box store. (Same unit, different warranty & distribution chains, but the Westinghouse tech line rings at HTP headquarters.)

    The 100KBTU version of the apartment friendly 20 gallon water heater under the Westinghouse label is model WGRGH20NG100F, the 75K version is WGRGH20NG75F, and less than $1600 USD, full-retail at the box store. (Often under $1500 when on-sale, like it is right now).

    The Westinghouse 50 gallon all stainless that is probably good enough to run 2 bathrooms AND the space heating load, is the WGR050NG076 (currently on sale for less than USD$2K at the box store).

    >"Properly set up, it should be better than a power vent, but that is a lot of cost for a bit less gas use, the payback is in range of decades. I can't comment on how much fuel reducing the cycling will save, so maybe on that front there might be a reasonable payback."

    In my area the installed cost of adding an indirect to an existing boiler is sometimes even less than the internet list price of the GPDT 50 power vent. That's instant "payback", even if the net efficiency improvement on the boiler turns out to be negligible.

    But if it costs the aforementioned $5K quote payback isn't decades- it's "never", making the heat pump water heater a much better deal.

    I suspect retiring an aging cast iron Vaillant boiler early is "worth it" on it's own terms, but what replaces it would affect what makes the most sense for domestic hot water. If a mod-con boiler is the replsement, assigning the full cost of a mod-con boiler replacement to the domestic hot water portion isn't legit- only the additional cost of the indirect and associated plumbing & controls.

  9. SteerageFlyer | | #31

    So after all of this discussion I think we're leaning towards the HPWH. My wife really likes the idea of not needing to burn NG for hot water.

    I think we're probably going to keep the boiler as is for now. It works ok (other than not firing sometimes but I think that's a sticky relay that needs to be replaced). When that does, we may think about going to heat pump for the whole house and running the ductwork as much of a pain as that might be.

    I'd be interested in what everyone thinks though.

    1. Expert Member
      Dana Dorsett | | #32

      A heat pump water heater is a great way to go, given the low cost (and low carbon content) of your local electricity.

      If the boiler is working fine, keep it. It's rarely economic to scrap an existing fully functional boiler, even if substantially oversized for the load.

      Due to the combination of large oversize factor and location outside of conditioned space, a retrofit heat purging economizer is still going to be "worth it". There are several out there that are inexpensive and DIY-able, eg:

      As a DIY a heat purging boiler control can come in under $200, sometimes under $150, and would pay for itself in one heating season or less. If you're going to be replacing a sticky relay or aquastat anyway, maybe that's the time to install the economizer control.

      Most newer cast iron boilers already come fitted with similar (or identical) controls now, since they have been reliably shown to take back most of the efficiency hit from oversizing. While it may be difficult to separate out your water heating gas use from your space heating use when analyzing the gas bills, in your case.

      The issue of standby loss as a fraction of the total fuel use is pretty easily understood in this graphic:

      If your boiler is 3x oversized for the design heat load, even on one of the coldest days of the year you'd be at the 33% capacity tick on that curve, not quite making 70% as-used efficiency, and most of the time you'd be to the left of that curve. A heat purging economizer reduces the average standby temperature, which extends the knee of the curve to the left considerably. The comparable curves for the boilers tested in that Brookhaven study can be in found in the corresponding appendix for each.

      Compare Figure A3-2 (the boiler with heat purging controls) to any of the other systems tested. The better the jacket insulation, the better the curve, but a heat purging boiler control moves the knee where efficiency slides off a cliff quite a bit to the left. At 3x oversizing for design load you're probably averaging something closer between the 10-20% capacity marks on the curve, which in Figure A3-2 is still pretty good.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |