musingsheader image
5 Helpful?

Are Tankless Water Heaters a Waste of Money?

The energy savings you’ll get from a tankless water heater are usually too low to justify the high purchase price

Posted on Apr 6 2012 by Martin Holladay

Although tankless water heaters are, on average, more efficient than traditional tank-style water heaters, they’re also more expensive — so expensive, in fact, that many potential customers wonder whether their high cost can ever be justified by likely energy savings.

Before you can decide whether to buy a tankless water heater, you’ll need to know how much energy you’ll save. Can you trust the information provided by tankless water heater manufacturers — for example, the estimate from Rinnai’s online calculator that you’ll save $178 per year?

Before I get around to answering that question in detail, suffice it to say: probably not.

Real-world answers from a monitoring study

To figure out the payback period for the incremental cost of a tankless water heater, it would be useful to know:

  • The installed cost of a tankless water heater;
  • The number of gallons of hot water used per day by the average American family;
  • The in-use efficiency of a typical tank-type water heater and the in-use efficiency of a typical tankless water heater;
  • The annual natural gas savings and the annual dollar savings attributable to switching from a tank-type to a tankless water heater.

To find the answers to all of these questions, a group of researchers in Minnesota undertook a monitoring study to measure the performance of tank-type and tankless water heaters in actual homes. The researchers concluded that most tankless water heaters will fall apart from old age before they save enough energy to justify their high cost.

The researchers — Dave Bohac, Ben Schoenbauer, and Martha Hewett of the Center for Energy and Environment in Minneapolis, along with Tom Butcher of Brookhaven National Laboratory and Mary Sue Lobenstein of Lobenstein Consulting — monitored water heaters in ten homes for over a year. Their data have been published in a report, “Actual Savings and Performance of Natural Gas Tankless Water Heaters”.

The ten families enrolled in the study were chosen based on household size. The number of people in these families matched the household size distribution shown in the census data for the Minnesota in 2000: two homes had 1 resident each, three homes had 2 residents, two homes had 3 residents, two homes had 4 residents, and one home had 5 residents.

A total of twenty-four water heaters were installed in the ten homes; each home got at least two water heaters. Eight homes got a tank-type water heater (an A.O. Smith GCV40 40-gallon natural gas water heater with atmospheric venting). In addition, each home got at least one natural gas tankless water heater. Ten tankless water heaters were tested; a variety of models were chosen from among those sold by five manufacturers (Bosch, Noritz, Rheem, Rinnai, and Takagi). The researchers did not test any electric tankless models.

Monitoring equipment for each test set-up included a dedicated gas meter, a dedicated water meter, and temperature sensors that measured the temperature of the incoming water as well as the temperature of the hot water. (The incoming water temperatures varied seasonally by about 30°F to 35°F.)

At each house, researchers adjusted gas and water valves to alternate between the tank-type water heater and the tankless water heater at monthly intervals. Only one water heater was used at a time. The changeover schedule was adjusted at each site so that every tested heater operated over the full seasonal spectrum of incoming water temperatures and outdoor air temperatures. An average of 363 days of useful data were collected from each home.

Installation costs are high

The study found that tankless water heaters cost more to install, and save less energy, than many energy experts claim. “From interviews with eight local contractors, installed costs for whole-house gas tankless water heaters as a retrofit were estimated from $2,000 to $5,000, with typical price range of $2,500 to $3,400. These costs are considerably higher than estimated by others. … For comparison, these same contractors estimated the installed cost of a conventional standard water heater to range from $900 to $1,300, with an average cost of about $1,100.”

Installing a tankless water heater in an existing home is significantly more complicated that swapping an existing tank-type heater for a new tank-type heater. In most cases, the tankless water heater will require a different method of venting — usually double-walled vent pipe through a side wall — as well as a larger gas line, reconfigured water piping, and in some cases the installation of a new 120-volt electrical receptacle.

EF ratings are less than measured efficiency

Among the data gathered by the researchers for each water heater were the following:

  • The temperature of the incoming cold water;
  • The temperature of the hot water leaving the heater;
  • The number of gallons water flowing through the water heater;
  • The amount of natural gas burned by the water heater.

Using these data, it was a simple matter to calculate the actual efficiency of each water heater. The researchers’ efficiency calculations included standby losses but not distribution system losses. The researchers reported, “Across all sites and water heaters, the measured annual efficiency averaged 16% less than the DOEUnited States Department of Energy. EF” — that is, 16% less than the Energy Factor calculated in a laboratory according to a standard established by the U.S. Department of Energy.

The tank-type water heater used in the study (A.O. Smith GCV-40) has an EF rating of 0.60, but its monitored efficiency was only 49% (11 percentage points lower than the EF).

The non-condensing tankless water heaters had EF ratings that ranged from 0.82 to 0.83, but the monitored efficiency of these units ranged from 69% to 78% (about 9 percentage points lower than the EF ratings).

The condensing tankless water heaters had EF ratings that ranged from 0.89 to 0.95, but the monitored efficiency of these units ranged from 62% to 89% (about 17 percentage points lower than the EF ratings).

In their report, the researchers noted two major factors responsible for these low efficiencies. The first factor is that the hot-water draw schedule used in the EF laboratory testing procedure is unrealistic. The test procedure requires six draws of over 10 gallons each at one hour intervals followed by a 19-hour period of inactivity. In the real world, however, only 3% of actual hot-water draws are greater than 10 gallons. In a real home, there are a great many low-volume draws rather than a few high-volume draws.

The second factor is that the EF laboratory testing procedure requires a daily draw of 64.3 gallons, which is higher than the amount of hot water used by average families. In this study, the average daily hot water use per house was 41 gallons.

Tankless water heaters use a significant amount of electricity

Unlike old-fashioned atmospherically vented tank-type gas water heaters, tankless water heaters use a fair amount of electricity. The measured standby electricity use averaged 5 watts; during operation, these tankless gas water heaters consumed between 50 and 80 watts of electricity. If there is any need for the units’ freeze-protection heater to turn on, electricity consumption rises to between 120 watts and 182 watts.

According to the researchers report, “Electricity consumption … accounts for about 5% to 18% of operating costs for tankless water heaters.”

Calculating simple payback

The researchers calculated the simple payback period for the incremental costs of these tankless water heaters. To make the calculation, they assumed that the incremental cost of installing a non-condensing tankless water heater ranged from $1,500 to $2,500. Using these figures, the simple payback period for non-condensing tankless water heaters ranged from 21 to 35 years.

The researchers assumed that the incremental cost of installing a condensing water heater ranged from $2,500 to $3,500. The payback period for condensing tankless heaters (not including the Navien CR-240A, an outlier with dismal performance) ranged from 27 to 38 years.

The Navien CR-240A was in a category by itself. Equipped with a very poorly insulated 0.5-gallon buffer tank, this water heater proved to be an energy hog. The Navien CR-240A (a condensing unit) had energy savings of only $49 per year; assuming that it had an incremental cost of $2,500 to $3,500, the payback period for this unit ranged from 51 to 71 years.

In general, the condensing tankless units weren’t worth the upcharge in cost compared to the non-condensing tankless units. The researchers noted, “On average, switching from a non-condensing tankless water heater to a condensing tankless water heater only increased savings by 3%, which is small compared to the 9% average difference in Energy Factor ratings.”

Should I buy one?

Tankless water heaters have at least two drawbacks: they are expensive and mechanically complicated. They also have at least two virtues: they are compact and can provide “endless” quantities of hot water. For some applications, these benefits are desirable enough to tip the balance in their favor.

As long as you realize that you'll never save enough energy to justify the high purchase price, you may want to buy a tankless water heater for its performance specifications and compact size.

Last week’s blog: “Energy Modeling Isn’t Very Accurate.”

Tags: , , , , , ,

Image Credits:

  1. John Eisenschenk

Apr 6, 2012 7:42 AM ET

by shane claflin

That's why I bought an indirect system with a modulating boiler. My bills are still high. Buderus s120 with a Utica USC5.

Apr 6, 2012 8:31 AM ET

by James Morgan

It wouldn't hurt to mention that this study is based on a best-case scenario using natural gas fired heaters. Propane tankless heaters are becoming depressingly popular in these parts. It would be interesting to see how they compare to a well-insulated electric resistance heater like a Marathon. I suspect the payback period would be ... never.

Apr 6, 2012 8:46 AM ET

Response to Shane Clafin
by Martin Holladay

In a previous blog on water heaters, the posted comments included a discussion of Marc Rosenbaum's monitoring data for his indirect water heater connected to a Buderus boiler. Like you, he had high bills, and his monitoring data showed that he was burning a lot of oil to make hot water, especially in the summer. Low-use households will find these indirect boilers to be especially inefficient.

Apr 6, 2012 8:47 AM ET

Response to James Morgan
by Martin Holladay

I agree with your comments. In most areas of the country, it's more expensive to operate a propane water heater than an electric-resistance water heater. Propane is an extremely expensive fuel.

Apr 6, 2012 9:13 AM ET

My Success
by Ryan Polasek

I installed a propane tankless heater a few years ago for about $1800 (part and labor) which includes the 30% federal rebate given at that time. Once thing the article does not mention is the life space of a tank vs tankless water heater, tankless being almost double the life span making saving even greater. In my case, I am saving a lot more money yearly because I went from an electric tank heater in a basement that could reach 35 degrees in the winter. My propane has gone up 10 cents in the last 5 years but my electric has gone up 60% so my cost are more consistent. Be sure to do all your research for your area before deciding because it was a great benefit for me!

Apr 6, 2012 11:36 AM ET

new construction
by Hunter Dendy

This seems to be focused on replacing an existing tank heater. What about new construction where you have to buy either and start either installation from scratch? Seems the "up charge" would be much less and then is it worth it from a payback perspective? Since the tankless units are efficient, you would need to consider it against a more expensive tank, not a bottom line.

Apr 6, 2012 11:52 AM ET

Edited Apr 6, 2012 11:52 AM ET.

Response to Hunter Dendy
by Martin Holladay

The math is fairly straightforward, so it should be easy for you to do your own calculations if your incremental cost to install a tankless water heater differs from the figures in the cited study.

On average, the tankless water heaters used 37% less energy than the tank-type heaters. The annual savings ranged from $42 per year for families that used small volumes of hot water to $121 per year for a high-use family.

Apr 6, 2012 6:29 PM ET

Gas draw very high
by David Goldman

As a homeowner who follows a number of heating sites, I've noticed that many of the pros point out that since the btu hot water load exceeds the heating load in many houses, the tankless will require much larger gas draw and supply piping. As it is with so many switching to gas, the aging street line infrastructure may become an issue sooner than later. I'd much prefer an indirect to use storage to give me hot water i need.

Apr 6, 2012 7:47 PM ET

That's for water only...
by Allison A. Bailes III, PhD

That all seems perfectly reasonable, Martin, and I don't think those numbers surprise many of us who understand building science. If you're using a tankless water heater for both space and water heating, however, I think the payback would be more favorable. I wonder if the research team has any plans to look at that system.

Apr 7, 2012 5:31 AM ET

Low Cost Tankless
by Kevin Dickson, MSME

Of course, in new construction, the real estate that a tank uses is worth $300-$600.

And this Puerto Rican built model sure is cheap: The installed cost would be under $1800

Even so, I don't recommend them because they are nowhere near as rock solid dependable as a tank style gas water heater. Also, you have to flush them biannually with vinegar? How much does that cost?

Apr 7, 2012 5:50 AM ET

Edited Apr 7, 2012 5:52 AM ET.

Response to Allison Bailes
by Martin Holladay

Plenty of people have been experimenting with using tankless heaters to supply hot water for hydronic space heating systems, including researchers funded by the Building America program. One of those researchers, though -- Armin Rudd of the Building Science Corporation -- is not yet ready to recommend this approach to builders, because these systems have glitches that can cause major headaches.

At a meeting in 2011, this is what Rudd said: "Beginning in the mid-1990s, we began using water heaters for space heating. We had failed check valves in Taco pumps. We had scale. You get thermosyphon flow that gets past the check valve. On some systems, we got hot water flow to the coil during cooling. You get hot water heating systems and cooling systems fighting each other. To cut that short, most of those builders went to a different system because of the problems.

"Then we started using tankless water heaters. In 2006, decided to take another look at this. We had the low flow problem. The units wouldn’t turn on unless you needed .6 gpm. Then we had the frequent on/off flow problem. -- a lot of that. You end up with slugs of hot and cold water in the system.

"Since 2006, manufacturers have responded to these issues. They are now down to the .4 gpm threshold. ... We're working with NYSERDA on a pilot project up in Utica. We are now monitoring two systems, with and without a small storage tank. We're getting plugging of the inlet filter of the tankless heater. It happens a lot. The aluminum anode rod was corroding enough to plug the filter. It's a recirculation problem. If you have any recirculation, any decay of the anode rod, it will end up plugging your system. So we addressed that problem with a unit with a stainless-steel strainer. Now, it becomes an annual maintenance thing. Before that they plugged up in a matter of weeks. If you have to clean the filter once a month, that’s no good.

"We have builders asking if we are ready to recommend this type of system yet, and we are saying, “not really.” We are still worried about the high frequency of the cleaning of the inlet filters. Building America doesn’t want to recommend a system that will give us a black eye, and we want to be sure we get the savings that we are predicting. Should we be adding a buffer tank or not? If you don’t have that, what are the problems? We're not sure yet."

Apr 7, 2012 5:57 AM ET

Response to Kevin Dickson
by Martin Holladay

My guess is that the Puerto Rican unit from Marey is designed for high incoming water temperatures. It's probably designed for the tropics, and therefore wouldn't produce very much hot water with incoming water temperatures near 40 degrees F, like we get in the northern half of the U.S. Unfortunately, the spec sheet omits any information on the temperature rise or the incoming water temperature used to calculate gallons per minute ratings.

Also, the ignition system requires two D batteries. That's no good. "Honey, will you remember to buy some more flashlight batteries for the water heater? I'm getting cold showers again."

Apr 7, 2012 8:39 AM ET

by Philip Koepf

It seems like a lot of effort and money to make these tank less units more cost effective. If the average and more typical use is small amounts of hot water on and off through the day with long periods of down time is anyone looking at optimizing small tank type water heaters? What if you super insulate the tank and all of the piping? Is anyone making high efficiency tank type units? If you add a tank to a tank less water heater what is that?

Apr 7, 2012 9:54 AM ET

Response to Philip Koepf
by Martin Holladay

If you are choosing an electric-resistance tank-style water heater, thicker tank insulation will be reflected in a higher EF -- so when you're shopping, look for an EF above 0.93. I've been searching for an online database listing the EF ratings of electric resistance water heaters, but I haven't found one yet. If anyone knows of such a list (one that compares ratings of water heaters from several manufacturers), please share the link here.

Apr 7, 2012 5:16 PM ET

Apr 8, 2012 6:07 AM ET

Response to Richard Kinkel
by Martin Holladay

Thanks for the link -- I knew I had seen it before, I just forget where to find it.

If you are looking for a small, well insulated electric resistance heater, it looks like you can't get one that is smaller than 28 gallons. The best EF for that size appears to be 0.93.

Apr 8, 2012 8:36 AM ET

Edited Apr 8, 2012 8:51 AM ET.

EF and tankless hype
by Curt Kinder

If the EF isn't available divide 4423 by the annual kwh usage listed in the EnergyGuide - that seems to give a good approximation. I don't see EF on EnergyGuides anymore.

4423 is the number of kwh needed annually to heat the amount of water (64.3 gallons / day) from 58 to 135 per the DOE WHAM test. If a water heater consumed just 4423 kwh annually, it would have an EF of 1.00, but that would require perfect insulation (in the case of fired or resistive water heaters)

I think the tankless units are over-hyped, and I'm not suprised to read that they fail to attain published EF. I also wonder how long they'll last and how expensive to repair.

I also wonder how well they'll cope with hard water over time - the extremely high rates of heat flux across very small heat exchangers would seem to be vulnerable to scaling.

I grew up in a house with an oil fired indirect tank. During heating season that represents a simple solution operating at the efficiency of the connected boiler...if anything it increases the efficiency of the boiler since it spreads its standby loss across two loads (heat + hot water)

Efficiency plummets during summer, though - the house goes through half a tank of heating oil, 80+ gallons just to heat water for two people for 4-5 summer months.

I'd like to add a heat pump water heater for summer use up there, but haven't worked out the details. Cooling and drying the basement would be an advantage along with shutting down the furnace for the summer, but I haven't figured out a spring / fall switchover arrangement that Mom will be able to handle.

Apr 8, 2012 10:02 AM ET

Edited Apr 8, 2012 10:04 AM ET.

20 gal electric water heater
by Joel Cooper

Since an electric tank-style water heater is a viable option (especially in areas without natural gas service), how big of a tank is needed? We have been using a 20 gallon tank for 8 years (only 2 of us in the home and we did not have space for a 40 gallon tank) with no problems. Is there a measurable energy saving by going with a smaller tank?

Apr 8, 2012 3:07 PM ET

Response to Joel Cooper
by Martin Holladay

Q. "How big of a tank is needed?"

A. You've answered your own question: "We have been using a 20 gallon tank for 8 years with no problems."

Q. "Is there a measurable energy saving by going with a smaller tank?"

A. All other factors being equal, a smaller tank should have smaller standby losses, and will therefore save energy. But you have to be sure that the small tank is as well insulated as the larger tank you are comparing it to.

Apr 8, 2012 4:15 PM ET

extra insulation
by 5C8rvfuWev

Martin, does it make sense (with an electric dhw tank) to "add" insulation -- I had a blanket that slips over the top and it was nice. I'm wondering about adding even more, like wrapping/taping batts over the blanket. (keeping the control panels, etc. accesible of course) Does it make sense? Or am I going to damage something?

Apr 8, 2012 6:50 PM ET

Each hundreth of EF is worth
by Curt Kinder

Each hundreth of EF is worth about $5 per year in electricity at national average electricity rates.

It used to be fairly routine to add an insulating blanket around water heaters, but the last couple I installed recommended against it owing to improved factory insulation.

It is getting to the point where most of the loss is via tank fittings - 4 or 5 penetrations through the jacket and insulation.

Beware that really small tank water heaters (6-20 gallons) don't seem to have an energy guide or AHRI rating, so likely lack decent insulation.

The sweet spot seems to be 30 gallons / EF = 0.95

Apr 8, 2012 7:44 PM ET

by Keith Gustafson

Funny, I would have put one of these in my last house 15 years ago had they been a consumer item as they are now. HD sells a Bosch for 800 bucks that I think now has a water flow powered ignition. i am sure it would have cost little more than a standard to install but it was a previously unplumbed building.

I was going to complain that there was no way a retro would cost that much more to install, but then thought about replacing all that gas pipe, yeah, that would cost.

The new owners replaced the HD cheapo gas 40 gallon unit less than 2 years ago. One thermocouple in 13 years I guess it owes no one a thing.

Honestly the tankless units are reasonable enough, but plumbers do place a premium on stuff they are not used to. I have a hard time believing it would take an experienced plumber more than a day at 100 bucks an hour to replace any water heater.

The biggest knock on these that I have seen is that they do not work the way we expect, low flow gives them fits, temps vary.

That would give me pause more than maybe not getting my money back.

Apr 8, 2012 9:28 PM ET

I'd consider a gas tankless
by Curt Kinder

I'd consider a gas tankless only if space were at a premium and I had natural gas at the house. I'd lean toward a power vented high efficiency storage gas fired heater if at all possible.

We do routinely install electric trankless units in higher end new construction projects, not for the efficiency but for the ability to mount them near points of use. Obviously we have to plan for them in the electric service.

We feed the tankless from a preheat tank warmed with a refrigerant desuperheater. Geo units have a desuperheater as a factory option. We sometimes add a desuperheater to an air source heat pump / central AC in cases of families of four or more. For 3 people, a heat pump water heater starts to make more sense as the upstream supply to point of use tankless electric units.

Electric require service planning but avoid all the gas line sizing, carbon monoxide and venting issues as well as operating at higher EF.

Apr 9, 2012 11:31 AM ET

More to the story
by Daniel Beideck

My fear is that the take home from this will be reduced to the sub-title, "The energy savings you’ll get from a tankless water heater are usually too low to justify the high purchase price." There are two main issues I have with this simplified summation.

1) As others have pointed out in the comment section, the study was done under conditions that most favor the tank system. The fuel used, natural gas, is the cheapest of fuels typically used to heat water. Therefore, the dollars saved are fewer than if another energy source were considered. The other factor is that the study looks at retrofitting a home that has already been optimized for a tank system. Certainly, a situation of some practical utility, but not the same conditions found when considering new construction.

2) I believe is it false thinking to evaluate the two approaches based on economics alone. The tankless method saves energy. That means something more than just dollars! Let’s not lose sight of that.

Apr 9, 2012 11:55 AM ET

Edited Apr 10, 2012 4:46 AM ET.

Response to Daniel Beideck
by Martin Holladay

The advantage of a field study like the one mentioned in my article is that it measures real-world performance and reports real-world costs. Such figures are much more useful than hypothetical numbers pulled out of the air by equipment manufacturers. Of course, any field study can be criticized as unrepresentative; however, as long as the researchers accurately report the conditions monitored during the study, no readers will be misled.

The researchers in the Minnesota looked at natural-gas heaters; that certainly makes sense, since natural gas is one of the two most common fuels used for heating domestic hot water in the U.S. (The other is electricity.)

Electric-resistance tankless water heaters are much less common than natural-gas tankless water heaters.

Moreover, a study that examined homes that heat domestic hot water with propane would have limited applicability, since propane is usually more expensive than electricity.

Of course you are correct that tankless water heaters use less energy than tank-style water heaters. However, anyone who is interested in lowering their use of fossil fuels or reducing their carbon footprint still needs to consider whether any particular retrofit measure is worth the investment, because we don't have an unlimited number of dollars to invest.

The idea behind seeking the "low-hanging fruit" is that it's always better to choose cost-effective measures before investing in measures that are demonstrably not cost-effective.

Apr 9, 2012 11:58 AM ET

Consumer reports 2008 "study" agrees with not worthwhile
by Brian Knight

Although its not nearly as through as the MN study, Consumer Reports had similar findings in its 2008 lab tests:

It also points out that not installing a water softener with tankless can compromise warranty claims. I think the hard water issues can have a wide range of problems from none to severe.

Ive always avoided tankless due to the air quality concerns in tight homes and the preference and simplicity of all-electric energy. PV prices are bolstering this strategy and it doesnt get any lower maintenance than the Marathon tank heaters that eliminate the anode rod.

Apr 9, 2012 12:24 PM ET

reply to Martin
by Daniel Beideck

I have no issue with the study itself. I do, however, believe we need to be careful and keep in mind that it not be misinterpreted and applied too generally to fuel sources other than natural gas or to new construction.

I also have no issue with an economic analysis or seeking the "low hanging fruit" first. I do have issue, however, when just economics alone are considered when discussing a particular technology. I presume after all that the "green" in refers to sustainability not the color of money. If that be the case, shouldn't sustainability warrant some mention when considering a tank versus a tankless system?

Apr 9, 2012 12:35 PM ET

Calculating sustainability
by Martin Holladay

Is it "sustainable" for American families to install tankless water heaters in their homes? Now there is a complicated question! I'm afraid that the ramifications of that question, and the algorithms necessary to solve it, are beyond most economists. They are certainly beyond me.

To be "sustainable," a practice must not lead to an economic or environmental collapse if it is maintained for 10,000 or 100,000 years. If I had to bet money on one side of the issue or another, I guess I'd place my bet on the side voting: not sustainable.

Most of the world's population still takes showers by ladling cold water over their heads -- and the sustainability of that practice is open to question, considering looming water shortages.

In short, I try to steer clear of any pronouncements involving the word "sustainable." For an in-depth discussion of this issue, see What Does ‘Sustainable’ Mean?

Apr 9, 2012 12:51 PM ET

How about green
by Daniel Beideck


So you don't like the word "sustainable". Fine. As the "green building advisor", shouldn't you be advising us on how green something is, and doesn't being green mean something beyond dollars? Yes, doing an analysis is easiest when only the economics are considered and becomes more complicated when other (green) factors are included. However, many of us come here because we care about more than just the bottom line, even if that's not always the easiest approach.

Apr 9, 2012 1:13 PM ET

On defining "green"
by Martin Holladay

I'm afraid I don't have a useful definition of "green" to fall back on. I'm sorry to disappoint you. I'm not trying to be a wise guy; I'm just trying to use accurate language and to encourage careful thinking.

Probably the most useful definition of "green" (at least for builders and architects) is the definition provided by the green building program that the builder or architect hopes to comply with. If you are enrolled in a green building program that gives extra points for installing a tankless water heater, then a tankless water heater is clearly green.

"Green" has no technical definition -- only a variety of programmatic definitions.

Apr 9, 2012 2:40 PM ET

It's not easy being green
by Daniel Beideck

Deciding what option is greenest is a difficult task in many cases. If often comes down to comparing apples to oranges. It is easier to just stick with the apples, i.e. dollars, and ignore the oranges, i.e. the other "green" factors. However, we shouldn't forget that the oranges exist too, in my opinion. For some readers the apples will be all that matters, but for others the oranges might change the balance.

Apr 9, 2012 2:56 PM ET

Apples and oranges
by Martin Holladay

It's hard to know which avenue is greenest -- an inexpensive water heater or a more expensive water heater that uses 37% less natural gas. I would usually prefer the more energy-efficient option, but the high price is an indicator that the tankless unit probably has higher embodied energy. It may also require more frequent maintenance, especially if you have hard water.

One possible downside is that a tankless water heater might encourage longer showers -- an example of the dreaded rebound effect or takeback effect. Of course, this problem only happens with some families.

In general, I am wary of the "gadgets will save us" philosophy. In most cases, families that want to reduce their carbon footprint can find many ways to reduce the amount of hot water they use -- and that's probably the greenest approach to lowering the negative environmental impact of the huge demand for domestic hot water made by American families.

Apr 9, 2012 4:22 PM ET

Combi space and water heating with tankless
by Jim Lutz

The group in Minnesota that did the tank vs tankless study is now working on a project to monitor the efficiency of combi space and water heating systems. An initial report of the study was presented at ACI.

Apr 9, 2012 4:45 PM ET

Response to Jim Lutz
by Martin Holladay

You're teasing us -- can you give us a two or three sentence summary?

Apr 9, 2012 6:49 PM ET

I'd just like to mention the
by Steven Leighton

I'd just like to mention the origin of these water heaters is the single water tube system from pre WW2 Europe.
Most houses had just one or two cold water taps in an average family house. One le(a)d pipe ran from the outside of the house to the kitchen then, when WC's were placed upstairs , the lead pipe ran up the interior wall to the WC/bathtub above. To get hot water to the kitchen and upstairs meant taking a short lead branch from the cold supply to the wall hung heater right at the point of use.
I've never thought of these old tech heaters as efficient and infact I doubt that you could buy one in northern europe.
The important fact is that they were point of use heaters so replacing a central boiler with one of these just could not work very effectively.
These heaters started to go out as copper pipe came in, in the late 1930's.

Apr 9, 2012 9:35 PM ET

...not so fast!
by Curt Kinder

Martin writes above "... natural gas is by far the most common fuel used for heating domestic hot water in the U.S."

I don't agree, and neither does ACEEE, per their late 2011 study of emerging domestic water heating technologies.

ACEEE writes that storage heaters command the vast majority of the water heating market, and storage heaters are about even split between gas and electric models.

Tankless offers a substantial bump in EF when compared with traditional center flue gas fired water heaters - those center flues lose heat 24/7, contributing to the abysmal EFs of that type of heater.

What irks me is the gas tankless industry's marketing hype that claims that a storage electric water heater also suffers large standby losses, and that is simply not true, as discussed above (new storage electric tanks routinely achieve EF in the 0.90 - 0.94 range)

Apr 10, 2012 4:50 AM ET

Response to Curt Kinder
by Martin Holladay

Thanks for catching my error; I have corrected my mistake. The point I was leading up to, however, is true: there are far more tankless gas heaters than tankless electric heaters, so if you are studying tankless heaters, it certainly makes sense to focus on natural gas models.

I also agree with your other point: that electric-resistance tank-style heaters have much lower standby losses than gas tank-style heaters. Electric-resistance water heaters are often a sensible choice.

Apr 10, 2012 7:31 AM ET

by Keith Gustafson

Of course at 15 cents a kw they can be blindingly expensive to operate, and with a very bad recovery rate you need a larger one than you would with a gas hell with it who needs hot water?

Apr 10, 2012 7:06 PM ET

Tank heater + drainwater heat recovery heat exchanger =...
by Dana Dorsett

...a higher net efficiency than a condensing tankless with no drainwater heat exchanger, in almost every 4-person family that showers rather than tub-bathes. As the family size gets smaller there's a crossover- the standby loss of the tank remains the same, but the return from the drainwater heat exchanger becomes less.

With a 4" x 48" or larger drainwater heat recovery heat exchanger and a ~30KBTU/hr burner on a tank heater the "endless shower" aspect of a tankless comes into play too, so like tankless, the extended showering capacity can peel quite a bit off (or even fully negate) any savings from the higher operating efficiency.

At home I use a ~0.82EF tankless as a modulating space heating boiler (only heating-system water in the tankless), and a "reverse indirect" heating system buffer tank with an internal heat exchanger, for the hot water, and a drainwater heat exchanger pre-heats the cold water feed to the indirect and the shower. I have it set up the heating system to run at domestic hot water temp (~125F), so the standby loss on the indirect is less than it is on those kept at much higher temps. The flow & delta-T through the tankless on the turn-off point of the indirect's aquastat multiplies up to less than my whole house heat load at the outside design temp, but under a showering load the temp in the tank drops to about 115F, ( a bit lower in mid-winter) where output of the tankless up to about 45-50KBTU/hr, with the rest of the heat for the shower supplied by the drainwater heat exchanger.

With all zones calling for heat it'll still deliver a ~105F shower forever, which is why I set the occupancy sensor on the bathroom lights to ~10minutes or my kid could be in there for HOURS. :-)

It's probably more efficient in this application than in a straight-ahead domestic hot water use. With the buffer tank it can't short-cycle 50x/day on hand-washing cycles, and the minimum burn time is in excess of 5 minutes. Standby loss on the reverse-indirect is small compared to a tank heater, and the standby loss on the tankless is well below what it would be for a high-mass boiler, but the cycling and standby losses are still higher than a modulating condensing boiler with heat-purge. The electric power draw at idle is comparable to that of mod-con boilers too.

It's not perfect, but it's not terrible either. (Beats what it replaced by a good margin, anyway.) I could probably save ~8-10% replacing it with a condensing unit, but I can reduce the fuel use more by spending the money on more thermal retrofits to the building envelope instead. (If a ~1-2kw grid-tied cogenerator ever became available I might consider a bit o' re-configuring of the system- it would be pretty straightforward to integrate something like that in.)

Apr 11, 2012 12:17 AM ET

Low Tech Solution
by Jan Juran

Hi Martin: I was able to cut the net fuel consumption of my hot water heater by almost half for shower and faucet aerator usage, by swapping out the old 2.5 GPM showerheads for new Niagara Earth 1.25 GPM non-aerated showerheads (and 1.0 and 0.5 GPM faucet aerators). Neither home residents nor guests can discern any noticeable reduction of shower efficacy/experience. Rather than expend $2K to $5K for a new tankless installed cost, the new showerheads cost less than $10 each delivered. The life of my existing storage hot water heater should be significantly extended, since tank life is affected by the volume of new cold water (containing dissolved oxygen) drawn into the storage tank. BTW website offers the new Niagara Bi-Max showerhead (both 1.0 and 1.5 GPM, user selectable) for $10.25 and the Niagara Tri-Max (user selectable at 0.5, 1.0 and 1.5 GPM) for $10.50. Low tech solutions sometimes can be both functional and cost effective.

Apr 11, 2012 4:13 AM ET

Response to Jan Juran
by Martin Holladay

Thanks for your important reminder about the cost-effectiveness of simple conservation practices.

Apr 11, 2012 9:47 AM ET

In the spirit of conservation
by 5C8rvfuWev

I still wonder if the storage losses of an electric dhw tank can be cut safely and effectively by adding extra insulation over the shell of the tank. Bad idea?

Apr 11, 2012 9:59 AM ET

Response to Joe W.
by Martin Holladay

Modern water heaters have better insulation jackets than the water heaters of 20 years ago. In the old days, it made sense to install a water-heater blanket; these days, you'll see little savings. As a previous poster pointed out, most of the standby losses for an electric-resistance water heater occur at the fittings and pipes.

That said, I see no harm in adding insulation if you want to. Make sure that your P&T relief valve is unobstructed and is accessible for testing, and make sure that the electrical access panel and thermostat adjustment controls are also accessible. Start by installing pipe insulation; then wrap the tank with insulation, cutting away the insulation as necessary to provide access where needed.

Apr 11, 2012 10:22 AM ET

Insulate the pipes first, then the tank.
by Dana Dorsett

Putting 5/8" wall (~R4, not the 3/8" wall stuff found at most box-stores) closed cell pipe insulation over all of the near-tank plumbing, including the cold feed and the temperature & pressure outflow pipe does more to reduce standby loss than tank-blanket over a newer electric tank heater. Adding it over all of the accessible distribution plumbing is also cost effective in higher priced electricity markets, less so in 5cent/kwh land, (but you'd still spend less time waiting for hot water if you do.)

In an average home something like 15-20% of all hot water heating energy is abandoned in the distribution plumbing, mostly from short-draws where a gallon of water is pulled into distribution to serve a pint of 110F water at the tap for hand-washing, etc. Adding R4 to all distribution plumbing keeps the heated water in the distribution hot enough to be useful for about a half-hour, as opposed to less than 10 minutes for uninsulated pipes. See:

Apr 11, 2012 1:43 PM ET

response to Keith Gustafson
by Curt Kinder

I agree that 15 cent Klicks makes for expensive hot water from any resistance heater, but if the alternative is trucked propane at $4 per gallon, then the electricity ain't so bad.

Dual element electric tank heaters recover the top 1/3 of the tank fairly quickly.

Apr 11, 2012 6:04 PM ET

Tanks or no tanks
by Tom Gocze

I get this question a lot. I tell people that if they want unlimited hot water and want to pay for that, then a tankless makes some sense.
If they want to limit their teenagers or anyone else from spending too much time in the shower, they should install an electric hot water heater (for a small household) and consider a heat pump water heater.
I like the Geyser heat pump water heater since it is made here in Maine and it can be swapped from one tank to another.
The build quality is more industrial than consumer quality.
Recovery time is an issue and it needs to be in a reasonable sized basement.

Apr 11, 2012 6:43 PM ET

high cost of tankless HW system
by Dennis Pack

Why isn't someone asking about the high cost of the units? The same units sell for almost half the cost in Australia ... not more expensive in the US because of some lobbyist slapping on high tariffs, etc?
And why the high cost of installations? I had mine installed for a fraction of what is quoted in the article ... makes me think there are some contractors who just charge more because they think they can. In Australia I actually installed my own ... it's not rocket science ... actually quite simple in new construction.

Apr 11, 2012 7:34 PM ET

I totally disagree – 180 degree’s out
by Karl Henschel

I totally disagree – 180 degree’s out
I have a Bosch 125BS tankless (LPG) waterheater used with a closed loop solar hot water system
(2- 4’x8’ solar water panels, a sunwise controller, circulation pump and 100 gal storage tank)
Water coming out of the solar storage tank goes into the Tankless waterheater, if it is hot the tankless does not come on. It only comes on to bring the water up to temp (125 degs F).
This system has been in use for 6 years and works great, It paid for itself in 5 years of savings, over a propane fuel, tank waterheater of 40 gals.
We are getting 80 to 90 percent of our hot water free from the sun, and some of the new Hotwater Solar collectors are more efficient and can be used in the northern areas. We live in Sothern California.
And LPG is a hotter fuel than NAT so you will use less.

Apr 11, 2012 7:52 PM ET

Edited Apr 11, 2012 7:54 PM ET.

$49 gets you access to ASHRAE 2011 Combo Research
by Robert Bean

ASHRAE 2011 Annual Conference, Seminar 4-1 -- High Efficiency Space and Water Heating: The Combination Approach presenters: Rosalyn Cochrane, Martin Thomas, Thomas Butcher, Ben Schoenbauer,

Also good reading from Hoover, Schoenbauer and Kingston presented at ECW B4 2012 Conferences (working on making these available - stand by.)

I've also listed and linked for access over 36 topic related documents at this page: on domestic water heaters Many of these research papers are current and currently free to the general public. Get em' while you can.

Apr 11, 2012 8:49 PM ET

Edited Apr 11, 2012 8:52 PM ET.

Heat Pump vs Tank vs solar
by Jim Price

I reached the same conclusion about 8 months ago and have been looking for better options. The latest Rainbow seems to be the Heat Pump water heater. Has anyone done a fact based study between a traditional tank and the heat pump heater based on actual usage in the Northern part of the country? Florida doesn't count. In Northern climates the heat pump will raise the heating cost of the house by extracting the heat you just paid to produce. Also verses Solar using real numbers, real usage with actual complete installation costs, not just main component wholesale price costs?

Register for a free account and join the conversation

Get a free account and join the conversation!
Become a GBA PRO!