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

Community and Q&A

HPWH (in garage) in series with Electric Water Heater (in house)?

truesouth | Posted in Energy Efficiency and Durability on


I’m hoping to get a sanity check before proceeding with my water heater plan for a house I’m building, located in central Oregon (zone 5).

My plan is to install:

1) 50-gal Heat Pump Water Heater (GE 50-gallon model) in unconditioned garage space

> connected in series to downstream >

2) 50-gal Electric Water Heater (50-gal super insulated Marathon unit) in conditioned space in adjacent laundry room.

Some reasons this seems to be a good setup in my particular situation:

– The garage walls are CMU with 3″ polystyrene insulation board and the house is tightly sealed, so I’m not worried about heat loss trade-off due to the cooling effect in the garage.

– The house is designed for a lot of passive solar gain, and the HPWP would sit in front of a large 84″ x 60″ high-SHGC (~0.70) south-facing window in the garage, so it will benefit from solar gain.

My thinking is that this arrangement would mean that the electric water heater will almost never turn on unless I exceed 100 gallons of use, since the 50 it has been keeping warm will be the first used, and then next 50 it gets from the HPWH located two feet away in the garage will already be hot.

So… am I missing anything here?

Many thanks for any guidance!

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. davidmeiland | | #1

    The electric heater will have to run enough to keep the water in it at the temperature you want at your faucets. It cannot simply sit there, never running, because it will cool down, and then the hot water from the HPWH will get diluted to lukewarm. Yes, you will be feeding it hot water from the HPWH, but that is not the whole picture.

    I think you'll want to install a couple of valves so that you can bypass the electric heater when you don't need it, which will probably be most of the time.

    You might want to just go with a slightly larger HPWH, maybe a 65, and skip the second heater. Marathons are not cheap, and if you can save money by not buying one, you should.

  2. iLikeDirt | | #2

    You can buy a 66 gal ATI AirGenerate for $500 right now with free shipping:

    I wonder what's wrong with it that they're practically giving it away like that, but with utility/state rebates lowering the price to zero or below in many places (such as Oregon, see their rebates page), you'd almost be crazy not to buy one. I bought just one today because this price was too insane not to jump on with the addition of rebates.

    Instead of a $1,000 GE GeoSpring, get two of these $500 units! In case the first one breaks, hook up the second instead, or cannibalize it for parts. With all the rebates you can get up there in OR, maybe get four or five if they'll let you.

    What a crazy world we live in.

  3. truesouth | | #3

    David: True, I shouldn't have said the electric heater would never run, just that it will run very little to maintain temp vs. having to heat cold incoming water. And good suggestion re: just going with a larger HPWH. I considered that, but liked the option of having the electric water heater in conditioned space do the heavy lifting during very cold periods. It will be more efficient, especially in our heated room, than the electric backup on the HPWH during those times. So my plan would be to set the HPWH to "heat pump only" mode in this dual heater configuration.

    I have to admit I also like the idea of having 100 gal capacity but also having a redundant/failover capacity of 50-gal in case (or when) one fails, provided I plumb them with a bypass valve. Basically I will have a 100 gallon HPWH with half of its storage in conditioned space and half in a solar heated unconditioned space, but either half can run independently if needed.

    It may be that I'm splurging needlessly on the Marathon unit ($850) and should just go with a standard $350 electric heater though. We plan to live here for the long haul so I'm rationalizing the lifetime warranty on that one I'm sure. I may rethink that and pocket some cash there.

    Nathaniel: I almost bought the same ATI-66 for that price this morning. It's a great deal, but my understanding is that it's because the mfg is no longer warrantying the units. I've read so many reviews citing system failure on both the AirTap and GE GeoSpring units that for once I was afraid to go for the best deal and instead bought the GE unit from Lowes for $699 so I could get a 9-yr extended warranty for $99 more. With the rebates I'll be getting, it's still unbelievably cheap.

    Thanks for the input!

  4. Dana1 | | #4

    Unless you have some monster sized tub to fill the series HW heater is all cost, no benefit. The effective EF of the power used in the resistance electric HW heater would be extremely low, and it won't last any longer than if it had been the primary HW heater, since longevity is primarily a function of how much fresh water has moved through it.

    Most homes (even those with soaker tubs) would do fine with a 65-80 gallon HPWH, such as the Stiebel Eltron Accelera 300 or Ati-66 or Ati-80. (The ATI units have all-stainless tanks, and the tank would likely last 2x as long as the glass lined Stiebel Eltron). Spend the money on a bigger HPWH, not an electric tank with resistance heaters.

    If you want to extend showering time, a gravity-film type drainwater heat recovery unit with better than 50% heat recoveray @ 2.5gpm (about $500-600 if purchased from EFI: ) would extend showering time considerably, and reduce recovery time by half. It won't do a thing for tub-filling, since the heat recovery occurs only during simultaneous drain & potable flows. Oregon has a statewide rebate subsidy for drainwater heat recovery:

  5. truesouth | | #5

    After reviewing the construction budget last night I decided that if I go with the dual tank system I'll just stick with the standard $300 electric tank my plumber recommends rather than the spendy Marathon.

    D: thanks for the feedback. Is the cost you're referring to the operating cost (electricity) of the two series tanks vs one HPWH or is it the upfront purchase cost? If the latter, it's actually the opposite for me. The dual tank system is $999 total before rebates but single larger 80 gal HPWH units are $1900 for the GE and $2300 for the Accelera 300. (The fire sale on the ATI units is out because one of my requirements is mfg warranty and local service and parts capability)

    Otherwise a single 80 HPWH does seem simpler. My wife has specified a soaking tub (ugh) and we have two young kids so 100 is a nice luxury but I'm sure we would survive with 80. But that $1k in cost savings plus the failover benefit with dual 50-gal tank system is tempting me, provided there's not a big hit in the ongoing electric cost.

  6. Expert Member
    Dana Dorsett | | #6

    It's the full lifecycle cost that matters. With 2 tanks you have twice the standby losses, and the difference in upfront cost needs to be balanced against the additional standby loss. A quick sketch:

    A typical electric HW heater uses about 4600kwh/year, about 10% of which is standby loss, so you're looking at 460 kwh /year in additional standby loss with the 2-tank system.

    At Oregon's state average electricity price in January 2015 (the most recent posted by the EIA) that comes to about $50. Assuming a 12-20 year lifespan for a glass lined tank and NO electricity price inflation, that's $600-1000 in additional power use over it's lifespan.

    The $300 hot water heater also adds at least another $100 in installation labor (unless that's the quoted installed price.)

    The warranties for the AO Smith PHPT-80, & GE 80 gallon units are comparable to the Stiebel Eltron unit which makes it hard to rationalize the $400-600 upcharge for it, except for the better reputation for reliability. Any of them should outlast their 10 year limited warranty period. But "should" doesn't always mean "will"- the upcharge might be viewed as insurance (even though it isn't, really) based on reputation alone.

    Of course, for the price of the 80 gallon Geospring you can now buy too ATI-80s (or three for the price of an Accelera 80) and leave one(s) in the crate until the first one fails. :-) The stainless steel tank will last forever- if only they had their act completely together on the rest of it! It's a relatively small startup company that got into it without the deep manufacturing & support experience of a big player like GE or AO Smith. I hope they continue to survive as a company. (They have 3 left at the fire sale price, free shipping included: )

  7. truesouth | | #7

    Dana, thanks for the helpful response. I was just in the middle of trying to calculate that standby loss $ to figure out whether to go with the cheaper 2-tank setup or a pricier single HPWH, so that's very timely.

    I was back to using the Marathon's 0.95 EF just to see what best/lowest standby cost might be and was getting $23.38/year for standby loss (at $0.10/KWH). Not fair if I go with a cheap glass lined unit I realize, since the Marathon has lower standby loss b/c no scaling on its tank material and more insulation. (Side note: I read somewhere today that if in conditioned space, water heater will approach 1.0 in heating season b/c loss is heat energy that is captured -- we have ERV so would get this if it's true. That's another post topic though.)

    Anyway, anything in that low cost range per year has a very long payback period if I save ~$1K now, so for me seems to make it worthwhile to just pocket the $1K in savings. I'll have to make myself invest it so I can use it to buy a new unit or units in 10 - 15 years when there are hopefully even better choices available.

    Thanks again!

  8. charlie_sullivan | | #8

    The idea that standby loss is zero if it's in the conditioned space only works if you are heating the space with resistive electric heat. If you are heating via something less expensive, then heating with the standby loss of the tank is a very inefficient way to do it.

  9. propeller | | #9

    Thanks Nathaniel for sharing the information. I've just placed an order for one.
    The silly part is that I have access to employee pricing on GE Geo Spring but the sale of that product just got discontinued in Canada as we break ground to build our house. But thanks to you, even factoring the exchange rate, duty and taxes, the ATI66 is still cheaper than I would have paid!

  10. iLikeDirt | | #10

    Happy to help! Mine is on the way too. I wonder how long it will last. Regardless, my garage gets nice and hot for about 7 months of the year so I'm hoping the operating costs will be really, really low during that time. And I get to seal up a roof penetration, too!

  11. Expert Member
    Dana Dorsett | | #11

    If any people in your house are shower-bathers the $600 drainwater heat exchanger (less OR rebate subsidy) would have a better lifecycle ROI than a Marathon vs. cheap tank in the series-plumbed configuration, but you have to do the math. A gravity film drainwater heat exchanger is good for at least 40 years (to where it's efficiency has degraded to 75-80% of it's day-1 efficiency), and it has significant residual value at end-of-life as scrap copper.

    Charlie is correct that standby loss into conditioned space is going to be the MOST expensive portion of the space heating fraction, even with 10-cent electricity. (How ARE you heating the place?)

    The additional standby loss of a lower EF tank is also a cooling load, but in a central OR location, but in a ~150 cooling degree day climate (base 65F) that's barely worth calculating. The lifespan of the Marathon vs. a cheaper tank + multiple replacements over the life-cycle of a Marathon is a much bigger cost factor than the cost of the standby losses. The question is really, how many replacements and at what installed cost per replacement?

    It may be worth considering the hedge value of higher efficiency against future electricity cost inflation. My own gut feeling is that electricity price inflation is going to be severely constrained by the continued falling cost of PV. In 2015 the lifecycle cost of small scale rooftop PV is probably slightly higher than 10 cents/kwh in OR (don't know, haven't seen any recent system quotes in OR, nor do I know what level of subsidy there might be), but by 2030 the cost of PV (any scale) is going to have a strong deflationary pressure on retail electricity prices. (Analysis from multiple financial industry sources seem to back up this view.)

    In that case there is nearly-zero energy cost hedge value in going with a Marathon, even if there is still a financial rationale for going with a bigger EF 2+ heat pump water heater without the in-series tank. Run some net-present value calculations on the different approaches, playing some "what if" games with the anticipated lifespan of cheap 0.90EF electric tanks, the standby loss costs & replacement costs.

    BTW: It doesn't much matter where the HPWH is relative to the location of the high-gain window. Direct solar gain has no effect on it's operating efficiency. It's an air source heat pump- it's only the air temperature that matters. It'll use the same amount of power whether it's in the opposite corner of the room, or prominently in front of the window blocking the view.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |