GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Audio Play Icon Headphones Icon Plus Icon Minus Icon Check Icon Print Icon Picture icon Single Arrow Icon Double Arrow Icon Hamburger Icon TV Icon Close Icon Sorted Hamburger/Search Icon
Q&A Spotlight

Help for an Ailing Water Heater

Would a tankless electric water heater be a good way to boost the output of a heat-pump water heater?

This heat-pump water heater isn't as robust as it once was and now the owner is looking for the best way to increase hot water capacity. One idea he has is to add an electric tankless unit. (Photo: Pat Buerskens)

Pat Beurskens has been happy with his heat-pump water heater, but he fears the time is coming when it won’t be able to keep up with demand.

He foresees the day when his young daughter will be using more hot water, and Beurskens is finishing out the basement of his Seattle home to be used as a rental. That, too, will increase hot water use.

Against this backdrop, Beurskens’s GE Geospring, a model that has been discontinued, is showing signs of fatigue. “Lately,” he writes in a post at the Q&A forum, “the water heater has been maxing out after two showers. I’ll be retrofitting a 1.5 [gallon per minute] adapter, which will help, but I’m still nervous about having a tenant and running out of hot water.”

In order to squeeze more capacity out of his system, Beurskens is considering a solar water heater and an electric tankless water heater.

He’d prefer using the heat-pump water heater most of the time because of its high efficiency. He doesn’t want to replace it with a larger model because of the expense and because he’s not eligible for another rebate.

He asks, “My main question is: Would it work to put the tankless electric heater in a series in front of the heat pump tank so that the tankless kicks on only when the tank starts dropping in temp?”

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

Avoid electric tankless heaters

The best option would be to avoid an electric tankless water heater altogether, argues Dana Dorsett.

“An electric tankless isn’t as sensitive to incoming water temps as fossil burners,” he writes, “but they are really abusive to the grid infrastructure, and one of the least green options there is.”

Dorsett explains it this way: Assuming the incoming water is 52°F,  filling a tub with 110°F water at a a rate of 4 gallons per minute, would take 28,000 watts of electricity — a ginormous load, six times that of a standard electric water heater.

That’s enough power, he adds, “to make the wires jump and the transformer serving your house heat up.”

Even though the load is intermittent, and overall doesn’t add up to a lot of energy use, the grid infrastructure capable of delivering that amount of power has to be built and maintained. So, for those few minutes a week that the water heater might actually be needed, the fixed portion of everyone’s power bill goes up.

Walter Ahlgrim adds that adding an electric tankless water heater as a retrofit in an existing house is “almost never a viable option.” The reason is that any unit that’s large enough to serve the whole house will draw more electricity than the existing panel can provide.

“When you add the tankless water heater to the new electrical panel, with new meter base and heaver wires all the way to the pole, the number is more than most people will go for,” he says.

Norman Bunn writes that he once had a tankless electric that required an additional 120 amps to operate. “Going this route may require you to not just upgrade the panel but get a larger feed from the pole,” he says. “That can be pricey.”

Try a second standard water heater instead

A better plan would be to install a second standard (tank-style) electric water heater in series with the heat-pump model, Dorsett suggests. The system would be designed so that the output of the heat-pump unit would feed the cold water inlet of the standard tank.

Although the system would still suffer the standby losses common to all tank-style heaters, the auxiliary heater won’t be using much electricity until both tanks are nearly depleted.

Consider drain-water heat recovery

An even more attractive option would be install a drainwater heat recovery system — the “tallest and fattest” that will fit in the available space. These passive heat exchangers take the place of a section of standard drain line on a shower. Hot water going down the drain heats up incoming cold water, saving energy in the drainwater that is normally wasted.

A 4-inch-diameter by 48-inch-long heat exchanger can save more than 50% of the heat that’s exiting the drain, Dorsett says. If the drainwater is 105°F, the 52°F incoming water is instantly heated to between 75°F and 80°F, depending on how quickly the water is moving.

“That has two effects,” Dorsett says. “Less hot water is needed to mix with ~78°F water to get 105°F at the shower head, so less hot water is being drawn, and the water entering the water heater only has to be raised 27 F° rather than 53 F° by the water heater, which means a slower depletion time and a shorter recovery time.”

That has the effect of increasing the effective capacity of the heat-pump water heater from 50 gallons to somewhere between 65 gallons and 75 gallons when someone is showering. (It won’t help when someone is drawing a tub.)

Keep the new system simple

Beurskens has heard enough to drop plans for an electric tankless and opt instead for a second tank-style heater, possibly with a recirculation pump.

You don’t need a pump, GBA editor Martin Holladay tells him. Just install the new water heater downstream from the existing heat-pump water heater (HPWH) and set the new water heater’s aquastat 10 F° lower than the aquastat on the main water heater.

A recirculation system would only increase the standby loss for a very small improvement in capacity, Dorsett adds.

“Insulate all of the hot water distribution plumbing with R-3 foamy pipe insulation, including the 10 feet of cold feed nearest the HPWH, and the nearest 10 feet of temperature and pressure overflow plumbing,” Dorsett says.

What about a new mixing valve?

Beurskens suggests one other option. “Speaking of simplicity, and assuming the new hot water needs can be limited with flow restriction,” he says, “what are your thoughts on a Tank Booster mixing valve? Basically turning up the thermostat on the HPWH to increase load capacity. This would negate the need for any new tank.”

A mixing valve could be used for those times when extra hot water will be needed, Bunn replies, but it wouldn’t be a good idea when the heat-pump water heater was capable of meeting the demand because that would mean more power consumption.

Thermostatic mixing or tempering valves between water heaters and sinks and tubs are required by code in most locations, Dorsett says, although some pre-plumbed units can reduce flow significantly.

“Turning up the storage temperature to 150°F or higher, and then mixing it down to 115°F with a thermostatic mixing valve can provide greater apparent capacity,” he says. “The down side to that is that a substantially higher storage temperature increases standby losses (not by more than adding a second tank, though), and lowers the heat pump’s raw efficiency.”

The bottom line, Dorsett says, is that turning up the thermostat on the existing HPWH and using a mixing valve will be cheaper in terms of total energy consumption than adding a second tank. That’s because standby losses will be lower, and because the HPWH is more efficient than a standard electric tank even when set a higher temperature.

Our expert’s opinion

Here’s what GBA Technical Director Peter Yost has to add:

Pat Beurskens has clearly done a lot of work and research to get the best domestic hot water he can configure for his situation. Here are some additional thoughts, particularly an extended one on drain water heat recovery.

When a heat-pump water heater is in the basement, particularly in low-load homes, keep an eye on the temperature and relative humidity in your basement during the winter. The HPWH is removing BTUs from your basement air to heat water.

Beurskens has a clever setup for his recirculating pump such that he decides when it runs. He has has set it up to manage energy and water efficiency for his most problematic draw, the most-used bath sink that shares a line with the shower.

I sometimes worry that we are repeating the low-flow toilet debacle with showerheads. That is, reducing flow without making sure that performance is maintained. EPA WaterSense has added performance tests to its shower head spec (see EPA WaterSense Specifications Appendix A and B). But I still hear from plenty of consumers who feel that WaterSense-approved showerheads are not delivering enough water. In my own home, I resorted to purchasing a series of low-flow showerheads until my teenage daughter finally signed off. Long, thick hair seems to be the issue.

It baffles me that given the difference in temperature between the tank water temperature and the air surrounding any tank water heater (let’s assume a 120°F tank temperature and 65°F air temperature in the basement — meaning a delta-T of 55 F°), that we don’t better insulate these tanks. See this ACEEE report indicating that an external insulation blanket can save approximately 28%.

More on drain water heat recovery

Mechanical engineer Dan Cautley, my close friend and colleague at SeventhWave in Madison, Wisconsin, who was my mentor in my first year at the NAHB Research Center in 1993, has done at least two research projects over the years on drainwater heat recovery, and he gave me this extended perspective on the topic:

“A vertical pipe drain water heat recovery unit is a remarkably simple way to capture waste heat, and the ‘effectiveness’ (effectiveness is the fraction of available energy that’s captured through a heat exchanger) can be surprisingly high (50% and higher). But this doesn’t mean it’s necessarily cost-effective.

“Horizontal pipe (sloped to drain) drain water heat recovery units may work fine, but are outside my experience.

“Overall system heat recovery effectiveness will always be lower than rated heat exchanger effectiveness, mostly due to heat loss from the drain upstream of the heat exchanger. In my detailed study of three units in commercial buildings, measured overall system heat recovery effectiveness was generally 5 to 10 percentage points lower than the nominal effectiveness rating of the heat exchanger. Effectiveness also depends on flow rate – an issue that is probably beyond the scope of a short blog.

“Dynamics (thermal mass) play a role. In general, thermal mass will reduce performance — think of taking a shower using 12 gallons of water, where it takes 3 gallons of flow to heat up the drain line and heat exchanger before reaching steady-state performance. (I found that it took about 6 gallons of flow to reach steady state in one system, and that wasn’t an extreme case.) This effect generally works against you – you don’t get full heat recovery instantly when the shower starts, plus you leave warm water in the heat exchanger when the shower ends, which is likely to cool back to room temperature. (And any cold water flowing down the drain will take this heat with it.) Thus the best application of a drain water heat recovery unit is where there are long episodes of hot water usage with simultaneous drain flow – a bank of showers in a fitness center would be ideal, while a residential shower is less so.

“But that’s not the whole story. Dynamics can work to help heat recovery where there are short flows of warm drain water and cold supply water offset in time. We found surprisingly good performance in a restaurant in which a dishwasher filled, ran, and then drained, with little simultaneous flow of supply water and drain water. The thermal mass of the heat exchanger absorbed enough heat when the dishwasher drained to significantly heat the supply water passing through it a few minutes later. The water use in each dishwasher cycle was low — a gallon or so — which was an important factor.

“Based on a 36-home study I did at the NAHB Research Center in the 1990s, my estimate of typical overall water heating energy savings in a residence is around 20%, but this will of course vary greatly with the specifics.

“All this said, a drain water heat recovery unit will certainly increase the effective capacity of any storage tank water heater, i.e. allow longer showers, for the reasons that Dorsett mentions. The Dorsett claim does seem somewhat optimistic in terms of the size of this effect.

“The bottom line: I love this simple, elegant technology, but it’s going to be really attractive only in limited applications. Here is a link to my report from the three-site study of commercial applications that I did 5 years ago. Take a look at page 34 of the report (page 38 of the PDF) for a list of criteria for selecting good applications. It’s written for commercial use, but the principles apply to residential use.”


  1. User avater
    Jon R | | #1

    If you limit flow to 1.5 gal/min (I use < 1 gal/min), then a secondary heater can be less than 35A (and usually 0A), with a low correlation to neighbor's use. Save the standby losses and space and use an small electric tankless.

  2. User avater
    Dana Dorsett | | #2

    Dan Cautley's study while interesting doesn't appear to have studied a good analog for a single family or 2-3 family home with, much shorter water distribution & drain runs than the 26 unit apartment building or a 6 suite student housing, where the distribution losses are much higher. I'd be curious to see what his 36 home NAHB Research Center study had for particulars back in the 1990s, but I'm basing my performance WAG on multiple Canadian studies from 2005 or later (one of which was in Vancouver, not too far from Pat Beursken's location.)

    My temperature numbers are based on in-situ meaurements at my house (though not measured to NIST standards of accuracy, and no data logging over time :-) ). The heat exchangers themselves and code requirements for pipe insulation have improved quite a bit since the 1990s. The new gravity film units from EcoDrain even leave Renewability's PowePipe series in the dust, and those PowerPipe units beat anything available in the 1990s by comfortable margins, inch for inch.

    As Dan stated:

    >"In general, thermal mass will reduce performance — think of taking a shower using 12 gallons of water, where it takes 3 gallons of flow to heat up the drain line and heat exchanger before reaching steady-state performance."

    No kidding!?! To get all the way up three gallons of water takes 150' of 3/4" pipe between the water heater and shower! That's certainly not my house, and probably not yours (or Pat's) either. But it's a good point- volumes/distances matter.

    At my house there is about 15' of (insulated) copper plumbing from the water heater to the shower mixer, most of which is half inch, and well under one gallon total volume. YMMV- but it's easy enough to estimate your ACTUAL distribution volumes (and losses) rather than try to infer it from a much larger multi-unit residential situation, or unspecified 36 houses from the 1990s study.

    Cast iron vs. plastic drain pipe would make a real net performance difference too, as does vertical vs. horizontal drain. With vertical drain pipe there is a lot greater loss due to the greater surface area contact from the gravity film, whereas with horizontal drain it's at most 20% of the surface area in contact with the water, even with 2" drain, whereas with vertical drain it's nearly 100% surface contact (which is the very feature gravity film heat exchangers are exploiting) Most newer small scale residential houses are using plastic, which has both lower thermal conductivity and less thermal mass than cast iron. In my house there is about 10' of 4" cast iron drain between the shower and the drainwater heat recovery unit, 8' of which is horizontal.

    Again, YMMV, but Pat Beursken's case is more likely to be more similar to my house than either the student housing or a 26 unit apartment building.

    >"But this doesn’t mean it’s necessarily cost-effective."

    From a net present value of the installed cost MERELY against lifecycle energy costs it varies with local rates and the efficiency of the water heater. An essential part of the cost-effectiveness in Pat's case is the avoided/traded cost of installing a second water heater (tankless or tank) or an oversized heat pump water heater or some other solution. The avoided/traded cost is a large fraction of the cost of installing drainwater heat recovery unit right up front. The present value of the cost DIFFERENCE varies with both the electricity rates and the estimated cost & efficiency of the alternative solution.

  3. Jayson Berger | | #3

    Would it make sense to look at Water Heater Booster at the top of the tank to increase the apparent capacity?

    It seems like the homeowner is mainly concerned about edge use cases, not daily use, so adding a bit of capacity for just those events seems appropriate. The devices are only a few hundred dollars and I'd assume he has the proper electrical in place since it's heatpump unit today.
    The only time the booster would kick on is when the pump could not keep up with demand so no standing losses to worry about.

    1. Pat Beurskens | | #4

      That's an interesting product, not sure why I didn't see those in the past. Since it runs only on a 30 amp circuit, I wonder if the affect on the grid that Dana mentioned is still a concern.

      The current solution of a drain water heat recovery pipe and flow restriction seems to be working for now. I may consider a water heater booster if/when our demand increases.

  4. JBB123 | | #5

    Just curious - have you put you're water heater into hybrid mode?

  5. Steve Black | | #6

    One other thought - if the existing heat pump water heater’s output is falling off there is a good chance the efficiency isn’t what you think it is either. Might want to get it serviced, may be low on Freon

  6. User avater
    Rich Cowen | | #7

    How do you test the efficiency of a GE heat pump water heater? I know I can connect mine via wired Internet. If I do that, does the unit maintain any statistics that can be viewed if I point a web browser to the IP address of the water heater?

  7. JBB123 | | #8

    The eithernet port on the water heater is for the wireless adapter which is no longer supported since GE sold out the heat pump water heater. Be careful plugging any thing into the port as I blew my control panel out trying to get the wireless to work. ( just plugged it in)

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |