musingsheader image
3 Helpful?

Heating a Tight, Well-Insulated House

For most superinsulated homes, a furnace is overkill

Posted on Nov 6 2009 by Martin Holladay

UPDATED on March 2, 2017 with information on the Dettson furnace rated at 15,000 BtuBritish thermal unit, the amount of heat required to raise one pound of water (about a pint) one degree Fahrenheit in temperature—about the heat content of one wooden kitchen match. One Btu is equivalent to 0.293 watt-hours or 1,055 joules. /h.

If you build a small, tight, well-insulated home — in other words, a green home — it won’t need much heat. Since typical residential furnaces and boilers are rated at 40,000 to 80,000 Btuh, they are seriously oversized for a superinsulated home, which may have a heating design load as low as 10,000 to 15,000 Btuh.

Builders have been struggling for decades with the question, “What’s the best way to heat a superinsulated home?” Your solution will depend in part on your answers to a couple of other questions:

Are you comfortable heating the house from a single point source? If you are, the best solution might be a wood stove, pellet stove, or a direct-vent space heater. These solutions work best in compact homes with open floor plans. Of course, the tighter the home’s envelope and the thicker the insulation, the more likely that indoor temperatures will remain fairly consistent from room to room.

Do you want an all-electric house? Green builders have diverging views on this question. Builders of net-zero-energy homes often avoid gas- and oil-fired appliances, preferring to balance energy loads with electricity produced on site by a photovoltaic(PV) Generation of electricity directly from sunlight. A photovoltaic cell has no moving parts; electrons are energized by sunlight and result in current flow. (PVPhotovoltaics. Generation of electricity directly from sunlight. A photovoltaic (PV) cell has no moving parts; electrons are energized by sunlight and result in current flow.) array or a wind turbine.

Of course, most homes still depend on grid-powered electricity, and if your local electric utility generates power from fossil fuel, then it makes little environmental sense to heat with electricity. From a carbon-production standpoint, it’s usually better to burn fuels on site rather than in a remote power plant.

What not to install
Before moving on to right-sized solutions, it’s worth mentioning that it rarely makes sense to install radiant-floor heat in a superinsulated house.

As Alex Wilson has explained, for well-built homes, an in-floor radiant system is usually overkill. Although these expensive systems are a good way to heat a poorly insulated house, they are usually a waste of money in a tight house. Moreover, many radiant floor systems include energy-wasting circulators that operate for 24 hours a day during the heating season.

A two-stage furnace may make sense
Some builders of superinsulated homes have concluded that there’s nothing wrong with a conventional natural gas furnace. The engineers at the Building Science Corporation in Westford, Massachusetts often specify a two-stage natural gas furnace (for example, the Goodman GMH95-045). Unlike a one-stage furnace, which has only two firing modes — on or off — a two-stage furnace with an ECM blower operates efficiently even under partial load conditions.

The Goodman GMH95-045 is available on the Web for between $839 and $1,050. According to Building Science principal John Straube, production builders can install a heating system using this furnace for about $2,500.

[Author's note: Since this article was written, a Canadian furnace manufacturer has begun selling a gas furnace rated at 15,000 Btu/h. For more information, see Finally, a Right-Sized Furnace.]


My paternal grandfather, William L. Holladay, was a refrigeration and cooling engineer who died in 2001 at the age of 98. In October 1948, a periodical called Engineering and Science Monthly published an article on heat pumps written by my grandfather. The article, predicted the development of ground-source heat-pumps, water-source heat pumps, and air-source heat pumps.

My grandfather wrote, "A third [potential heat] source [for heat pumps] is the earth. Much research on this idea is going on at present, but no firm statements about practicability should be made. For maintenance reasons it does not seem advisable to bury refrigerant coils, although some successful projects have been operated for periods up to several years in this manner. The best way to remove earth heat now seems to be through the use of a water coil, and the transfer of heat from water to refrigerant through an exchange of conventional design. ... To use earth heat, a hole must be dug, and the cost, while not always predictable, will surely be high: maybe from 25 to 50 per cent of the entire project...

"Some conclusions from field experiences with the heat pump may be summarized here. ... Buildings should be well-insulated, since good insulation reduces external load both in summer and winter, and tends to reduce also the difference between the two. Buildings should be shaded as much as possible, and windows should be shaded, particularly on south and west sides. ...

"More operating failures have been due to poor application and bad air distribution than to manufacturing and design errors. In other words, the pump isn't a piece of package goods like a radio or refrigerator, but an engineering project."

Right-sized heating systems
What if you don’t want a gas furnace — perhaps because your house lacks a gas connection, or because the smallest available furnace is still too big for your needs? Here are three other options:

  • Electric resistance baseboard heat.
  • A direct-vent space heater.
  • A ductless minisplit air-source heat pump.

Electric resistance baseboards
If your home has an impeccable envelope, and you’ve pared your heating loadRate at which heat must be added to a space to maintain a desired temperature. See cooling load. to the bare minimum, it may make sense to heat your house with electric resistance heat. That’s how Katrin Klingenberg, the founder of the Passive HouseA residential building construction standard requiring very low levels of air leakage, very high levels of insulation, and windows with a very low U-factor. Developed in the early 1990s by Bo Adamson and Wolfgang Feist, the standard is now promoted by the Passivhaus Institut in Darmstadt, Germany. To meet the standard, a home must have an infiltration rate no greater than 0.60 AC/H @ 50 pascals, a maximum annual heating energy use of 15 kWh per square meter (4,755 Btu per square foot), a maximum annual cooling energy use of 15 kWh per square meter (1.39 kWh per square foot), and maximum source energy use for all purposes of 120 kWh per square meter (11.1 kWh per square foot). The standard recommends, but does not require, a maximum design heating load of 10 W per square meter and windows with a maximum U-factor of 0.14. The Passivhaus standard was developed for buildings in central and northern Europe; efforts are underway to clarify the best techniques to achieve the standard for buildings in hot climates. Institute U.S., heats her 1,450-square-foot house in Urbana, Illinois.

It’s possible to buy a 1,500-watt (4,714 Btuh) baseboard heater on the Web for $90. If your house has a design heating load of 14,000 Btuh, you could heat it with three baseboard heaters ($270 plus installation). The main drawback to this solution is that in most areas of the country, electric resistance heat is expensive to operate.

Direct-vent space heaters
Many compact homes are easily heated by a direct-vent space heater. Installed on an exterior wall, these heaters require a wall penetration to accommodate two concentric vents. The outer pipe (the donut) brings in combustion air for the sealed-combustion burner, while the inner pipe is the exhaust flue. It’s possible to buy direct-vent heaters that burn natural gas, propane, or kerosene.

Direct-vent space heaters are affordable. A quick Web search reveals that a natural gas Empire MV145 heater (20,000 Btuh input) can be purchased for $500. Other options include the Rinnai 263 (9,100 Btuh output) for $909 and the Monitor GF1800 (16,000 Btuh output) for $1,150. If you want a heater that requires no electricity, you might choose the Robur TS2000 (7,400 Btuh input) for $509.

The first net-zero-energy house in the country — a superinsulated Habitat for Humanity house in Wheat Ridge, Colorado — is heated with a natural gas direct-vent space heater in the living room, supplemented by electric resistance heaters in each bedroom.

Ductless minisplit air-source heat pumps
If you’ve traveled to Mexico, the Caribbean, or the Mediterranean, you’ve probably seen ductless minisplit air conditioners. A ductless minisplit includes an outdoor unit (the condenser) and at least one wall-mounted indoor blower (a fan-coil unit); the two units are connected by copper tubing that circulates refrigerant.

A ductless minisplit is a type of air-source heat pump, so it can be used for heating as well as cooling. In recent years, Japanese and Korean manufacturers (including Daikin, Fujitsu, Mitsubishi, and Samsung) have significantly improved the heating efficiency of ductless minisplits. These units can now be used for heating in very cold climates — even in Vermont, where winter temperatures reach -20°F.

According to Andy Shapiro, an energy consultant from Montpelier, Vermont, Asian heat pumps are much more sophisticated than U.S.-made equivalents. “The Japanese are eating our lunch with these units,” Shapiro told me.

These ductless minisplits are so efficient and dependable that they are the heating system of choice for many low-energy buildings, including Passivhaus buildings. Shapiro used ductless minisplits for a heating system he designed for the newly completed net-zero-energy gymnasium at the Putney School in Putney, Vermont. He calculated that the capital cost for the building’s ductless minisplits and a photovoltaic (PV) array big enough to power them was cheaper than a ground-source heat pump system — even though the ductless minisplits require a bigger PV array. At the Putney School, the ductless minisplit system cost $100,000 less than a ground-source heat pump system, while requiring only $35,000 more in PV modules than would have been required to power the ground-source system.

According to Shapiro's estimate, which is based mostly on manufacturers' specifications, a ductless minisplit heating system can be expected to operate at an average coefficient of performance (COP) of 2.3 — that is, 2.3 times as efficient as an electric resistance heater. Shapiro's monitoring data reveal that the typical water-to-water ground-source heat pump system has an average COP of about 2.5. (Remember, the higher COPs claimed by ground-source heat-pump manufacturers are artificially high because they exclude pumping energy.) According to Shapiro, water-to-air ground-source heat pumps may have COPs as high as 3.5.

Specifying a ductless minisplit
The nominal heat output rating of a ductless minisplit is calculated at an outdoor temperature of 47°F. Since the unit’s heat output drops with the outdoor temperature, it’s important to check low-temperature performance before specifying a unit.

For example, the Mitsubishi Mr. Slim Hyper-Heat unit (model PUZHA36NHA) has a nominal heat output rating of 38,000 Btuh. According to the manufacturer, at an outdoor temperature of -13°F, its heat output drops only 21%, to 30,000 Btuh.

Another ductless minisplit, the Quaternity unit from Daikin (model FTXG15HVJU), has a heat output rating of 17,890 Btuh at 43°F. At -4°F, however, its heating output drops to 7,310 Btuh.

As long as the heating system designer sizes the unit so that it will meet the building’s heating load at the design temperatureReasonably expected minimum (or maximum) temperature for a particular area; used to size heating and cooling equipment. Often, design temperatures are further defined as the X% temperature, meaning that it is the temperature that is exceeded X% of the time (for example, the 1% design temperature is that temperature that is exceeded, on average, 1% of the time, or 87.6 hours of the year)., there’s no reason it won’t keep a home comfortable — even when the temperature drops well below zero.

Ductless minisplits In Massachusetts
Carter Scott used a Mitsubishi Mr. Slim system with a nominal heat output rating of 28,000 Btu/h to heat . The 1,232-square-foot home has a design heating load of only 10,500 Btuh. The installed cost of the ductless minisplit system — including one indoor unit downstairs and another one upstairs — was $5,250, a price that includes both heating and air conditioning.

According to the home’s architect, Ben Nickerson, at an outdoor temperature of 0°F, the Mitsubishi unit still cranks out 10,000 Btuh at a COP of 1.8.

Can I really depend on a heat pump?
If you’re worried that a heat pump won’t be adequate when the thermometer bottoms out in January, remember:

  • The coldest temperature of the year is reached for only a few hours a year.
  • Tight, superinsulated homes lose heat very slowly, even during power outages. As long as a cold snap doesn’t last for many days, most superinsulated homes won’t lose much heat.
  • In very cold temperatures, turning off the ventilation system will help a building stay warm.
  • If you’re really worried about prolonged subzero cold snaps, one or two electric resistance baseboard units provide cheap insurance.

This blog was adapted into a Fine Homebuilding article, Heating Options for a Small Home.

Last week’s blog: “Deciphering the Tax Credits.”

Tags: , , , , , , , , , , , , ,

Image Credits:

  1. Martin Holladay

Nov 6, 2009 9:44 AM ET

Radiant floor
by boone

Radiant floors when installed in finished concrete floors and connected to a solar thermal system which makes it eligible for tax rebates (NC=65% fed and state) makes them a very affordable option. We have installed them in small efficient houses with commerical on-demand water heater that do both heating and hot water for 2 floors and about the same price as a mini split. Does not include AC though which is minimal in our area. No wall mounted unit with air blowing and really comfortable even heat.

Nov 6, 2009 9:54 AM ET

OK, Boone -- how much did the system cost?
by Martin Holladay

Leaving aside the question of tax rebates, please provide the actual cost to install a solar thermal system, on-demand gas water heater, and in-floor radiant distribution system in a house you have worked on.

Real numbers, please!

Nov 6, 2009 1:51 PM ET

mini split condenser adapted to a ducted system
by Kyle

I was wondering if any one has attempted to take a condenser unit from a mini split and size a evaporator coil to it so as to make a small output centrally ducted heat pump system? Seems possible if a correctly sized small evaporator coil possibly from another application could be found. Of course it would involve some fabrication to adapt an air handler but not too bad. It would seem a desirable solution to get the right size central unit for a passive house in the mixed-humid climate zone.

Nov 6, 2009 2:19 PM ET

Passive houses don't need to deliver heat through ducts
by Martin Holladay

There's really no reason to deliver the heat through ducts. The heat can be delivered as intended by the ductless minisplit manufacturer -- through one or more fan-coil units.

Nov 6, 2009 2:48 PM ET

by John Brooks

Daikin offers a ducted option.

Nov 6, 2009 2:57 PM ET

A Different Perspective
by Riversong

Martin, I just posted the following rejoinder to Alex Wilson's critique of radiant in-floor heating in a tight house:

Alex, you suggest radiant floor heating for a high heat-load, drafty house, but a floor at a maximum 85° surface temperature can deliver only 34 btu/hr·sf, which would not be nearly enough for most poorly-insulated (let alone drafty) homes.

In a tight house, it's true that the low heat demand would require a lower floor surface temperature, and that might mitigate against the "warm toes" effect. But any surface that is above room air temperature increases the mean radiant temperature (MRT) of a space, and the MRT is at least as important as air temperature in determining human comfort. Human skin has an absorptivity and emissivity of 0.97 - higher than almost any know substance, including matte-black metal - so we are very sensitive to slight variations in MRT.

A truly green house will not only be super-insulated but also small, and a small house (or any house,for that matter) benefits from the unobstructed wall area of a radiant floor heating system. While the initial installation costs will be higher for a central hydronic radiant system (compared to wall-mounted space heaters such as direct-vent or mini-splits), the uniformity of heat, the lack of room obstructions, and the higher MRT and human comfort can easily justify the additional expense, and some of that incremental cost difference can be made up in increased heating efficiency over the life of the system.

And a high-efficiency (95%) direct-vent modulating boiler can also efficienlty provide unlimited hot water through an indirect tank, reducing the cost of the domestic hot water system and somewhat offsetting the heating system cost.

As to your point that a superinsulated house will lose heat slowly during a power outage: the issue isn't how much heat (BTUs) it will lose (which, at design minimum temperature, will be maximum design heat load), but how quickly will it's indoor temperature drop. That's dependent not only on insulation levels but, more significantly, on the amount of thermal mass within the heated space. A house with a radiant concrete slab will maintain indoor temperatures much longer than a low-mass structure, because not only is there a great deal of thermal mass but the mass is warmer than room temperature and will continue to heat the space until it reaches thermal equilibrium.

In addition to the higher MRT that a radiant heating system provides, it eliminates the air convection and turbulence (and noise) and temperature stratification that contribute to occupant discomfort.

Nov 6, 2009 2:58 PM ET

Thanks, John
by Martin Holladay

John is right: Daikin offers an indoor unit designed for ducted systems. Their designation for the indoor ducted component is FDXS. Here's a brochure that describes both indoor unit options:

Page 10 of the brochure describes the FDXS option.

Nov 6, 2009 3:04 PM ET

Thanks Martin and John for
by Kyle

Thanks Martin and John for the responses!

Nov 6, 2009 3:07 PM ET

No single answer on radiant heat
by Martin Holladay

The debate over radiant heat can't be resolved, since some people want radiant heat regardless of the cost. If you can afford the $15,000 (or more), and you want the system, then you might as well go ahead.

If your house can be heated for a system that costs $5,000 or less, however, you might prefer to spend the extra $10,000 on better windows.

You're right that a house with lots of thermal mass will cool off more slowly than one with little mass. So if you want mass, include mass. The radiant heat loop is optional. Books make great thermal mass. I've got a lot of books -- more every week. A concrete slab also works fine.

As far as "air convection, turbulence, noise, and temperature stratification" are concerned, these are red herrings. My basic assumption is that we are talking about superinsulated houses with very low air leakage rates. These houses simply don't suffer from your catalog of horrors. They are easy to heat, have little temperature stratification, and don't suffer air convection or comfort problems as long as they are well designed.

Nov 6, 2009 4:08 PM ET

Minisplit heat pumps
by Marc Rosenbaum

I have a serious interest in minisplits because the they are cheap, are a packaged system so don't require much engineering design, are easy to install, and are a good match for low load buildings with renewable power. But at this point I don't know whether i'm working with a COP of 2 or 3! There seems to me to be a pretty big gap between the AHRI ratings and the manufacturer's engineering data.

What you should understand is that folks like Andy and I have no practical way to monitor COP in a system that delivers its energy via air (no simple BTU meter here) and is constantly varying the air flow and delta T of the indoor units (as well as varying stuff at the condenser). So the best we can do is an estimate of COP from how much energy our buildings use, which of course is subject to many factors of weather, setpoints, internal gains, blah blah. This is important IMO because the COP (HSPF really) in the AHRI ratings differ a lot from what the mfgrs claim.

I have a simple glass dome meter on our Plainfield School prototype classroom so I'll know how much energy the Hyperheat uses. But I have confounding factors of adjoining spaces, insulated heat pipes running through the room, setpoints, etc. to contend with when I go to calc COP. We'll have meters on the 8 affordable houses we're doing, and on a small dorm. We've just turned a new house over to the owners with a ducted system but I haven't got meters there yet. We'll have meters on projects under construction this season.

Nov 6, 2009 5:25 PM ET

by Gary K

I was surprised that there was no mention of Geothermal Heating and Cooling for a super insulated house. My 105 year old brick home with refurbished windows, foam insulation in all wall cavities (around doors and windows), foam insulation in the attic, and foam walls and roof in a 350 ft kitchen and sun room addition, uses about 20% of the power of my former home of the same age, 300 ft smaller, same design and directly across the street from the super insulated one.

Nov 6, 2009 5:59 PM ET

Ground-source heat pumps
by Martin Holladay

There is no doubt that ground-source heat pumps work, and that they can be efficient. The main drawback to systems like yours is their high capital cost. For a small, well-insulated house, a ductless minisplit (air-source) heat pump costs far less than a ground-source system. Since the heating load in a well insulated house is so low, it's hard to justify the cost of a ground-source heat pump.

Nov 6, 2009 9:08 PM ET

mini splits
by David Whte

Most of my PH projects use mini splits, and mostly ducted (for the sake of concealing the unit and distributing heating and cooling to each room). Distribution sized to around 0.2" static pressure.

Robert, the MRT and acoustical benefits of radiant are lost on superinsulated, super tight houses. In a house meeting Passive House standard, the floor will only be 2 F warmer than room temperature at peak load. And in these houses MRT is not of issue - even interior window surfaces are generally with ASHRAE 55 recommendations for asymmetry at design day.

Same for acoustics and draft. A forced air system that only needs to deliver 3,000 to 10,000 btuh can do so silently and without draft.

Marc, for monitoring mini splits, couldn't one get a good measure of flow rates (by garbage bag or flow grid) at installation? The fans are stepped speeds, so a logging voltmeter could indicate which speed is operating - that gives CFM at each time step. log supply temp, room temp, and outdoor temp. voila. or?

Nov 7, 2009 8:00 AM ET

Mini-splits and air distribution
by John Brooks

The mini-splits do seem to be very popular in Japan.
I doubt that they locate a ductless unit in every room or that they are using the ducted versions.
David White,
How do you think they are distributing heating and coolong to every room?
I had considered using a ducted minisplit system in North Texas ..but found the installed price to be extremely high.
I also noted that the SEER ratings for the ducted versions are lower than the non-ducted.
Any ideas for conditioned air distribution other than ducted or multiple units?

Nov 7, 2009 5:22 PM ET

Aesthetics of Minisplits
by Brett Moyer

I would think that most homeowners would not appreciate the look of wall mounted indoor blower in each room. I noticed in one of the pictures in the Daikin brochure referenced, a picture of what appears to be a supply grill above a closet. Is this a ducted system? Or was the wall mounted blower installed in the closet behind the grill. Would this be an appropriate installation? Can you hind the wall mounted blowers?

Nov 7, 2009 5:32 PM ET

Wall-mounted blowers
by Martin Holladay

In Passivhaus homes, I have seen a single indoor blower unit — installed in a hallway or stairway — that provides heat for the entire house. In a superinsulated house, it isn't necessary to provide heat to every room.

I have been living for 29 years in a 3-bedroom house with a single heat source (a wood stove in the living room). There is no need to provide heat distribution to every room.

Dec 29, 2010 12:29 AM ET

Edited Dec 29, 2010 12:29 AM ET.

Cost of Radiant floor
by Michael Chandler

It was a great question to ask about the actual cost of Radiant floor installations. Here's how I figure the cost of my projects

Radiant floor:
1/2" PEX tubing $0.30 / lf @ 2,000 lf`/ 2,000 sf slab =$600
flat plate heat exchanger from Solar Hot USA =$350
two circulator pumps (Taco 009 bronze or eq) @ $160 = $320
120 volt thermostat $15
slab prep labor and parts not including foam under slab =$600
Manifold and heat exchanger labor and parts $800
Total for 2000 sf single zone radiant floor cost before profit and overhead $3,285

Demand water heater w/ tempering tank for radiant floor heating
quietside 120ODW condensing demand water heater 94% eff 120 kBTUh $1,100
20 gallon electric water heater from Lowes for use as tempering tank $400
high speed pump (Taco 009 bronze or eq) w/ iso-purge flanges $300
Fittings and pump control wiring $400
Labor to install $600
Total heating system for endless 94% eff hot water and radiant floor heat before profit and overhead $2,900

Solar drain back system three panel 120 gallon
three 4' x8' solar panels $3,000
Solar H2ot Solvelux drain back kit $900
SS drain back tank $300
120 gallon solar servant tank $1,200
piping and roof labor $1,600
total cost of a high output solar drain back system to drive floor and domestic hot water $7,000

Obviously there are a lot of variables here but this is a good basic picture of the costs involved. with out an outdoor reset mixing valve which as I see it adds about $800 (and yes it does dramatically improve the system) $7,000 for the three panel 120 g drain back solar, $2,900 for the condensing water heater setup and $3,285 for a 2,900 sf radiant floor system = $13,185 total.


Solar radiant.JPG

Nov 7, 2009 11:30 PM ET

radiant floor costs
by boone

My costs are in line with Michael's.
I have installed a radiant floor system without solar on a 2 story 1300 sq ft house for $7000 and a one story 1100 sq ft with vac tube collectors and 120gal storage for $13,000. In 2 floors codes req at least one mini split on each floor =$6150
So the mini splits provide AC and cost $900 less for the 2 story design. In a one story the cost is more like $3000 less but not able to connect to solar thermal which supplies the domestic hot water and improves the energy payback and does not connect to the coal fire plant as back up is natural gas.
The mini splits do not heat water so that has to be added in to their costs.
With a 65% tax credit the $13000 system, that includes solar water heating for domestic water, comes down to under $5000.
One caveat is that we have 4 in finished concrete on the ground floor level and 2 inch concrete over wood frame on the second floor. Some people love the stained concrete and some don't like it but is an inexpensive ($2.50 sq ft.) durable, high thermal mass floor.

Nov 7, 2009 11:51 PM ET

Mini-split distribution
by Scott Heeschen

I went through a similar thought process a few months ago in my remodel. I had originally specced Warmboard subfloor, but when the time came, I realized that the savings from using ordinary subflooring vs. the Warmboard would just about pay for a minisplit heatpump. Seemed like a no-brainer decision to me - get heating and air-conditioning for less. Yeah, no warm feet, but way cheaper in my situation. Several manufacturers have multi-zoned minisplits. I've seen units with up to 4 zones, maybe there are some with more. Of course the price goes up, but you don't need a compressor per zone.

Nov 8, 2009 5:20 AM ET

Good to see real numbers
by Martin Holladay

Michael and Boone,
Thanks for sharing some real numbers.

My own comments:
1. Keeping costs low on hydronic radiant systems is possible; it helps to have experience with these systems and to keep the number of trades involved down to a minimum.

2. Many builders (and many owner-builders) have been quoted much higher figures. The more trades involved, the higher the numbers.

3. Even using your numbers, ductless minisplit systems (which include air conditioning) are cheaper.

Michael, I noticed that you recently estimated (in an answer to a Q&A question) that it costs $18,000 to install air conditioning in a tight, well-insulated 4,000-sf. North Carolina home.

Nov 8, 2009 5:22 AM ET

Hi, Congratulations to the
by gas heater

Hi, Congratulations to the site owner for this marvelous work you’ve done. It has lots of useful and interesting data.

Nov 8, 2009 12:31 PM ET

Capital cost of GSHP varies a lot
by Dick Russell

In a couple of the above posts, there are references to the capital cost of a ground source heat pump as being high. This would seem to be a misleading generalization, as the cost is very situation-dependent. In one particular situation, a new and very low heat loss home in a rural area, and for which a water well would be drilled anyway, the capital cost vs. conventional gas or oil-fired system could be a wash. Water Energy Systems (NH) estimates $4.5 to 5K per ton of heat for "inside costs" for equipment and duct work, numbers provided to me at a presentation by them last April. The standing column well design needs typically 80 feet of water column per ton. This means that the well needed for adequate domestic use may easily be deeper than is needed for supplying heat to a GSHP. At most, only incremental drilled depth cost is incurred. The water-to-air system then also supplies A/C for the warmer times, although for a superinsulated house in a heating climate the load would be small and dominated by dehumidification of the fresh air brought in by the HRV. As with so many things, the answer to which system is best is: "it all depends."

Nov 8, 2009 2:19 PM ET

4000 sf number
by Michael Chandler

We won't build a house over 3,000 sf so I was just pulling a number out of the air on that.

Generally on our 2,500 sf houses I'll be putting in a 15 SEER zoned-bypass system for dehumidification w/ MERV-11 filter for $12,000 to $14,000 in addition to the radiant floor and the solar hot water mentioned above. This is a luxury, not-so-big, aging-in-place clientele.

Nov 8, 2009 11:37 PM ET

Geothermal (geoexchange) is expensive
by Andrew Henry

Hi Martin,

When I factor in all the costs for the radiant / geoexchange heating system that I put in my home I come up with an embarrassingly large number. Which is why I wish I had known more about Passive House sooner before I set out on my home's addition and renovation.

That said, the one thing about the Japanese heat pumps is that their level of design and engineering is much more advanced than that found in North American geoexchange heat pumps. I wonder what the COP would be if geoexchange water-to-water heat pumps were as sophisticated as the japanese and korean mini-splits. And what they'd cost.

Also you point out one thing that I wasn't aware of when I put in my geoexchange system which is that system COP tends to be lot lower than the rated COP, the COP the manufacturer wants you to see. A friend of mine put in a water-to-air heat pump and noticed the system COP value in the heat pump manufacturer's software calculations that his contractor provided. When my friend asked his contractor why the system COP was lower than the rated COP, his contractor couldn't answer because he didn't know what the system cop value meant.

Finally, I set up a Google alert on Passive House and noticed this in my latest alert. A cautionary tale when considering your next geothermal/geoexchange system!



Nov 9, 2009 5:24 AM ET

System COP versus manufacturer's COP rating
by Martin Holladay

I agree that the COP ratings provided by ground-source heat pump (GSHP) manufacturers are always higher than actual system COP.

I wrote an article on the issue that was published in the April 2008 issue of Energy Design Update. The article reported on monitoring data gathered by Rob Aldrich and Andy Shapiro. Aldrich measured a system COP of 3.5 on a GSHP with a manufacturer's rating of 5.0. Shapiro measured an average system COP of 2.75 for three GSHPs with an average manufacturer's rating of 4.0.

The article said, in part:
“Shapiro mentioned a few problems with the GSHP systems he monitored. One of the heat pumps had a bad valve, said Shapiro, and ‘it took a long time for the heat-pump rep to figure it out.’ Although the Econar specifications claimed that the heat-pump unit had a maximum output rating of 130°F — a useful temperature for domestic hot water — the unit installed in Vermont ‘barely makes 118°F or 119°F water,’ according to Shapiro.

“On the plus side, since heat pumps are fueled with electricity, they are inherently compatible with on-site power generation (photovoltaic arrays or wind generators). While Shapiro expressed satisfaction that all three Vermont systems are working well, he told the Burlington audience that several issues ‘make me nervous.’ These include:

• There remains a dearth of good monitoring data, allowing some GSHP dealers to make inflated claims.
• The electronic components of GSHP systems may be vulnerable to damaging electrical surges from lightning.
• Unlike most residential HVAC equipment, GSHP systems require engineering.
• In most areas of the country, the GSHP industry is not mature; installations require interactions between several trade contractors who may be unfamiliar with GSHPs.
• The complexities of GSHP systems provide many opportunities for mysterious glitches.
• Some GSHPs are being installed in homes with high heating loads where insulation and air-sealing improvements would yield a better return on investment.
• GSHPs increase utility peak loads during cold weather.

Derating Manufacturers’ Listed COPs
“The system COP measured by Aldrich amounts to 74% of the manufacturer’s listed unit COP. Similarly, Shapiro found that the average system COP for the three monitored GSHPs was 69% of the average unit COP in the manufacturers’ specifications.

“The data suggest that, as a rough rule of thumb, builders should derate unit COPs by 26% to 30% to obtain heating-season COPs for GSHPs systems — at least until future data contradict that rule of thumb.”

Nov 9, 2009 11:10 PM ET

Smaller geoexchange heat pumps
by Andrew Henry

Hi Martin,

A couple of things I wanted to throw out there. The geoexchange contractor who put in my heat pump mentioned that he did the installation of some small 3/4 or one ton geoexchange heat pumps for one of the federal research labs here in Ottawa. So I would think there may be some small load geoexchange heat pumps coming down the line. Will these heat pumps be competitive with the mini-splits, probably not when you factor in the cost of the drilling, but you never know.

Also, I wonder what the system COP starts to look like if geoexchnge systems are installed with good ECM circulators like the Wilo and or the ? Also ITT bought Laing so presumably we'll see these pumps here soon. That said these ECM circulators aren't cheap compared to PSC circulators, but they are bound to come down in price as ECM circulators start to displace PSC circulators from the market.

I got invited to a Wilo circulator demonstation (it's amazing what you can get invited to if you ask) where the presenter talked a bit about some of the stuff they're working on, and coming to market in Europe. Wilo has a really tiny circulator, the that could really open up a lot of design possibilities for radiant heating. And I just noticed this too!

Anyway, it still comes down to the fact that superinsulated houses don't have much of a heat load so it may all be moot. There may not be a place for radiant floors in a super insulated house but I wouldn't count radiant out entirely given how little energy some of these new circulators use and that they can be internally and independently controlled.



P.S. Martin, what are you doing up at 4:24 a.m. responding to comments?

Nov 9, 2009 11:12 PM ET

Here's the link to the Laing circulators
by Andrew Henry

The comments need a preview option!.

Here's the link to the Laing circulators...


Nov 9, 2009 11:51 PM ET

Insulation and Heating
by David Rouge

Thanks for the comment on superinsulated homes.

I am writing from New York City, which has an above-aeverage percentage of rental homes.
Multifamily buildings in NYC reflect the national pattern of including heat in the rent, which is characteristic of no other country in the world (I realize that there are subtleties here - small and electric-heated units in the US where renters pay for all utilities -- apartments in formerly socialist countries where there is a flat add-on fee for heat. But in the main this is true.) The biggest challenge in the US is 1) To meter all energy use, and 2) to place or increase taxes on all energy sources to reflect true costs.
Reducing heating costs for super-insulated homes is desirable, but not the big bang.

David Rouge

Nov 10, 2009 9:25 PM ET

GSHP can use domestic water well?
by Doug

All the closed-loop vertical-loop GSHPs I'm aware of use a dedicated well that's only for the GSHP, in fact it's grouted solid around the GSHP tubing. And, because these closed-loop GSHPs run their ground loop temps below freezing at times, even if you could fit the domestic water line & pump down the same well shaft, it would freeze solid at some point.
An open loop GSHP is not an option in much of the country due to water quality (any but the purest water damages the exchanger in the GSHP). I've heard that the added pump energy required by an open-loop configuration means higher overall energy use as well.
So, plan on a separate well in most scenarios, at full price.
Overall GSHPs could be great, but because mini-splits offer such high efficiencies without the first cost, I question the merits of GSHPs using today's technology.

Nov 10, 2009 11:42 PM ET

Natural Gas Floor Furnace
by Gayle

We have just purchased a nice smaller home in the Western part of Tennessee, brick and about 1400 sq feet. It is well insulated. It has a central split central unit using natural gas for heat. We are concerned about a not having a backup system should the electricity go off in the middle of the winter. We have had experience with a gas floor furnace installed in 1985 and were well pleased until when it was about 10 years old, we returned from a trip and smelled gas very stong in the house. We tried to have it repaired, but they never could get the leak stopped and we disconnected the unit and never used it again. We would like to purchase a floor furnace again for this house and try it again. Do you have any recommendations on what to install and have improvements been made in furnaces in the last 25-30 years?

Nov 11, 2009 7:08 AM ET

Heat during power outages
by Martin Holladay

For heat during power outages, you have several options.

1. The simplest and most common solution is a wood stove.

2. Several models of gas-fired space heaters are available that do not require electricity. These are commonly installed in off-grid homes. I have a propane-fired Empire heater in my own home that keeps my house above freezing when I go away for the weekend. In the article above, I also mentioned the Robur TS2000, which requires no electricity.

Nov 11, 2009 3:53 PM ET

GSHP can use domestic water well? by Doug
by Dick Russell

I have to wonder if Doug is confusing closed horizontal loops or bentonite-packed vertical bores, with antifreeze in the circulating water, when he speaks of running the loop below freezing. Standing Column Well (SCW) GSHP designs are quite numerous in the northeast, where groundwater is plentiful and usually of good quality. A SCW running to the freezing point obviously has been poorly designed. The website I referred to earlier [I believe the correct name is Water Energy Distributors, URL is] speaks of having done over 11,000 installations in the northeast part of the country. Operating data is available for many of these.

In terms of the carbon footprint of electricity-consuming GSHP vs. locally burned fossil fuel, even with an overall COP of only 3.0, with a delivered efficiency of the electrical power at only 33% of the fuel consumed at the power plant, the GSHP winds up at 99% overall efficiency, vs. low 90s for a high-efficiency NG or propane-fired boiler. If the electrical power comes from hydroelectric, the efficiency of the GSHP is quite large.

Clearly, GSHP (or any heating source) isn't best for every situation. However, generalizing serves no useful purpose.

Nov 12, 2009 2:58 AM ET

Once again, warm climates are forgotten
by Carl Seville

Martin- great piece, but with the exception of a very brief mention about mini splits being able to air condition, it would be nice to remember that much of the country lives in areas where air conditioning, and especially dehumidification is necessary for part of the year. The cold climate bias rears its ugly head (nothing personal) once again.

Nov 12, 2009 4:39 AM ET

Forgetting cold climates?
by Martin Holladay

I'll accept the criticism, but only up to a point. An article about how to buy fresh fruit shouldn't be criticized because the author forgot to mention pork chops.

No article can focus on every topic; this week's topic is heating systems. To the extent that heat pumps can provide both heating and cooling, that advantage was certainly highlighted. But I was focusing on methods of heating.

The article noted that available heating systems are usually oversized for green homes. Certainly that point applies where you live, in Georgia, as well as in Texas and Florida. If anything, the topic I addressed — choosing an efficient heating system when the heating load is very low — is even more relevant in your climate than it is in Vermont.

That said, I appreciate the reminder that hot-climate builders feel neglected. I'll try to return to topics that are more relevant to hot climates in upcoming blogs.

Nov 12, 2009 3:13 PM ET

distribution not an issue?
by Dave Brach


Very timely topic. In my experience the critical issue always becomes heat/cool distribution, and I have to wonder about your claims that distribution is not that important in a Passive/super-insulated House. I have not been able to sell my clients on point source heat (although that’s exactly what I would do if I were designing my own home). I would love to see more scrutiny ( i.e. data) on this issue because it affects the cost so directly. If the room with the heat source averages 72 degrees what does the back bedroom average? Don’t forget that the bedrooms are supplied with fresh air from the HRV that is typically cooler than the room air.

You also mention that stratification is not an issue. Maybe in a one story building, but in the Passive House we just completed (basement plus two stories) we are already finding substantial temperature differences between the upstairs and downstairs.

Also, I’m surprised you did not even mention the hot water heat coil made by Ultimate Air. It can be hooked up directly to the ERV supply ducts, can supply 8700 Btu/hr, and I believe it costs about $800. This could meet the heat load in many climates, and could be backed up with electric baseboard in many others.

Nov 12, 2009 4:16 PM ET

I share many of your questions
by Martin Holladay

Like you, I'm eager to hear of other builders' experience with point-source heat. I'd like to know whether occupants are satisfied with such systems.

When I recently visited Urbana, Illinois, I toured two Passivhaus buildings with ductless minisplits; the heat was introduced at single location, although the heat was supplemented by electric resistance heat distributed through the ventilation ducts.

In my own house, which is heated with a single wood stove in the living room, two of the three bedrooms do get colder than the other rooms when the bedroom doors are shut. Whether or not this is a problem depends on the temperature one prefers for sleeping. If a bedroom door is open during the day, then the bedroom will be warm when one goes to bed. If the door is kept closed all night, the bedroom will be cool in the morning. The third bedroom is adjacent to the stone chimney, and stays toasty.

I'm delighted you mentioned the hot water heat coil for Ultimate Air ERVs. I have heard of it, but haven't been able to locate information on the Web. I just phoned Ultimate Air, and the woman who answered the phone didn't know anything about it -- although she promised to get back to me once she made further inquiries. I'd love to hear from builders who have installed the unit. I'm especially interested to learn whether it comes with controls and a circulator, or whether it's up to the builder to design these parts of the system.

Nov 12, 2009 8:59 PM ET

by Doug McEvers

In my experience with superinsulation and airtight (1 ach50) or less housing I did not see any measurable stratification with an open floor plan. By open I mean 2 story living areas and open stairwells. I believe the open plan allowed the natural convective loop to occur and the indoor temperature stayed quite even from floor to floor.

Nov 13, 2009 12:28 PM ET

More information on the ERV heat exchange coil
by Martin Holladay

I just got a phone call from Craig Kinzelman at Ultimate Air. He said the person who knows about the heat-exchange coil, Jason Morosko, is out of the office and won't be back until next week. I'll know more after I get a chance to speak with Jason.

Craig said the option is brand new. He said, "We've sold a number of them but we haven't delivered any yet." In other words, nothing has been shipped.

Craig said his best understanding is that the heat-exchange coil and the insulated cabinet to surround it cost about $325. That price does not include a circulator or controls, however.

To get the unit to produce 8,777 Btuh — the maximum output — you need to crank up the water temperature to 160°F and operate it at 200 cfm.

I'll post more information when I learn more.

Nov 13, 2009 5:52 PM ET

by Jesse Thompson

Here's a link to an ASHRAE paper by Duncan Prahl, Thomas Hartman, Bruce Coldham & Katrin Klingenberg from 2007. It has graphs of instrumented point source heated houses:

More here from Coldham & Hartman's Rocky Hill Cohousing:

Nov 13, 2009 11:25 PM ET

Air to Water Heat Pumps
by John Semmelhack

Coming soon is an air-to-water heat pump from Daikin, called Altherma. I'm told it's distribution is currently limited to he Pacific Northwest. Here's a link to what looks like the American brochure, thanks to the website of Nelson Mechanical Design -

It can handle space heating, domestic hot water and (optional) space cooling. Distribution system for space heating and cooling is up to you (water-to-air coil in-line with ERV, radiators, in-floor radiant, etc.). COP for heating is 4.34 (at 45F outdoor temperature). Annual "COP" for domestic hot water should be really high (if COP is 4.34 at 45F, perhaps it's close to 8.0 or 9.0 at 95F?). Cooling EER is 12.17 at 95F outdoor. I'm waiting on more data for its performance/capacity at other temperatures, especially its capacity at low outdoor temperatures.

Air to water heat pumps have been available in Europe and in Japan (where they're called "Eco-cutes") for the past several years. I would expect models from other manufacturers to follow soon.

Nov 14, 2009 7:38 AM ET

Great links
by Martin Holladay

Thanks for the great links to papers and presentations that discuss this issue directly. It's great to see temperature measurements.

I was struck by Bruce Coldham's conclusion, which exactly matches my experience living in my house for the last 29 years:

"In order for the heat to distribute evenly through the upstairs, bedroom doors need to be left open or ajar throughout the bulk of the day. If they are, one would expect a temperature differential of no more than 5 degrees F degrees during the coldest days. If upper floor doors are closed, a 10 degree F differential is to be expected." [Coldham & Hartmann presentation]

It's great to see researchers out there with their data loggers. But I can't help thinking what the reaction of most Vermonters when they read these papers: "Yup, they spent $20,000 of research money to show that you can heat your house with a wood stove. Oh, and they figured out that when you close the door, the bedroom is colder than the other rooms. I could have told them that."

Nov 14, 2009 7:44 AM ET

Eco-Cute water heaters
by Martin Holladay

I wrote an article on Eco-Cute water heaters for the June 2007 issue of Energy Design Update. The article noted, in part:

"Over the past few years, at least 12 Japanese manufacturers have entered the residential HPWH [heat-pump water heater] market: Chofu, Corona (Denso), Daikin, Hitachi, Kyocera, Matsushita, Mitsubishi Electric, Panasonic, Sanden, Sanyo, Shecco Technology, and Toshiba. As a marketing ploy, a coalition of electric utilities invited competing manufacturers to sell their HPWHs under a common brand — or, as several press releases call it, a 'pet name.' The utilities settled on 'Eco Cute' — a typical Japanese example of fake English dreamed up by the marketing department of Kansai Electric Power.

"Most Eco Cute water heaters consist of two separate units: a heat-pump unit and an insulated water storage tank. Several manufacturers advertise Eco Cute appliances with a COP of 3.0; specifications note that the units can heat water to 90°C (194°F). A typical tank size is 370 liters (about 98 gallons); the relatively large tank size allows Japanese homeowners with time-of-use billing to take advantage of lower nighttime electric rates. ...

"The fast-growing market for Eco Cute appliances in Japan is particularly surprising in light of their very high cost. The typical Eco Cute unit sells for 700,000 yen, equivalent to $5,800; the Japanese government hopes that improvements in manufacturing efficiency will lower the price to about $3,300 by 2010. A subsidy of $665 is offered to any homeowner who replaces an existing water heater with an Eco Cute appliance; for new home purchasers, the subsidy drops to $416.

"So far, no Japanese manufacturer exports a Eco Cute unit to the US. ..."

Nov 23, 2009 1:25 PM ET

Natural gas vs electricity
by Garth Sproule

Interesting to note that the actual energy cost of gas is about one third the cost of electricity, at least in our market. But...because of the additional fixed monthly charges associated with natural gas, whether you use any gas or not, ($14.50/month here in Sask...over $30.00/month in Alta) it becomes much less of an advantage to use gas as the total energy use goes fact at some point, electricity becomes the cheaper alternative. The "greenness of electric depends of course on how it is generated...

Dec 24, 2009 4:25 AM ET

Significantly Cheaper than Minisplits: PTACs
by Kevin Dickson

PTACs are all-in-one heat pump units, 12-24,000btuh that have an installed cost of $500-$800.
Nobody makes one with a high a SEER and ECM motors yet, but who cares at that price? There are some other minor drawbacks, but in a small superinsulated house these are minimized because it's rarely on.

Here's a link to a builder who used one at

A little more theory at

And here's a discussion of using a PTAC in an "air to air geothermal system":

This thread is the near-term future of residential HVAC, but the average builder isn't dialed in. Some are (I believe) irresponsibly promoting Geoexchange for low energy homes such as:
Why would you spend $17k on a geothermal system to save $87/yr?

Mar 25, 2010 12:50 PM ET

HRV/ERV Use with Mini-Split
by Travis Kettering

I am interested in the use of a minisplit system. I was previously planning on using geothermal, but now I am wandering if it is worth the cost. If a minisplit system was used would you still use a HRV or ERV system? Wouldn't this help with only having one or two heat sources in the house?

Also have you heard any further information on the Ultimate Air with heat exchange coil? Thanks.

Mar 25, 2010 1:02 PM ET

Minisplits and HRVs
by Martin Holladay

A ductless minisplit system is used for heating and cooling. It has nothing to do with ventilation.

If your house is tight, you probably want to provide mechanical ventilation. As with any house, you have several options, including exhaust-only ventilation, supply-only ventilation, and balanced ventilation using an HRV or ERV.

UltimateAire sells the RecoupAerator ERV; as far as I know, it also offers a hydronic coil for insertion into the ventilation air stream. Once it is hooked up to a water heater or boiler, the hydronic coil can supply up to 8,700 Btuh. That's not much, but it might be enough for a very small, very well insulated house.

Mar 26, 2010 12:58 PM ET

living with my mini-split
by Sean Armstrong

One new development the article didn't mention was the combination of a air-source heat pump heating a large tank of hot water that can be used for both radiant heat and domestic hot water. That system effectively combines the ideal heat source with the ideal heat distribution system, and it's already available on the market--Daiken makes them, Sanyo makes them, etc.

I enjoyed the article in part because it affirms my own experience--I've foam insulated my 1100 sf, 100 year old farm house and installed a 9-12K BTU Fujitsu mini-split as the sole heating source, and all my guests comment on how warm and comfortable the house temperature is. I don't even use the ceiling fan with my 11' tall ceilings because the fan coil blows heat continuously, which seems to solve most of the heat stratification issues. So now I specify them for all my affordable housing projects, because going carbon neutral is important now, and we should stop pretending we have the luxury of proceeding with natural gas infrastructure as if global warming wasn't real.

Apr 7, 2010 8:57 AM ET

Radiant on lower level...what on upper level
by Kelly Gale

I am working on the plans for a tight, well insulated home. The walk out basement is living space and the owner is thinking about radiant down there from a ground sourse heat pump. He is also looking into ducted air on the main level of this home. Does this make sense? Full south facing longitudinal axis of home so solar is of great potential.

Apr 7, 2010 9:08 AM ET

I don't know your climate
by Martin Holladay

I don't know your climate, or the design of your house, so it's hard to comment. But as long as you asked your question — "Does this make sense?" — I'll answer: No.

You're planning on investing tens of thousands of dollars to buy:
1. A ground-source heat pump system. Let me guess -- about $20,000?
2. Radiant floor tubing.
3. Ductwork for the upstairs system, with heat provided by either a hydronic loop from the GSHP (I don't know if this would work, because GSHPs put out water with a maximum temperature of about 130°F) or a furnace.

That's a lot of equipment to heat "a tight, well-insulated home." What if you invested $15,000 in better air sealing and thicker insulation? Then maybe you could get away with $5,000 for ductless minisplits.
Any the homeowner would have much lower energy bills forever, because you would have upgraded the performance of the building's shell.

Apr 7, 2010 8:44 PM ET

Please convert Daiken Altherma to HSPF
by Kevin Hanlon

I'm looking @John S statement that the Daiken Altherma has a COP of 4.34 (@ 45F). Where is this information available, outside of the product manufacturers brochure? Is there a 3rd party like AHRI or ESTAR who validates this, and label it as HSPF, the seasonal value. I'm asking as I have a large project in Nashua that wants to go with this Daiken ASHP, It exchanges to water, then drops heat to the DHW tank for potable, and drops hydronic heat to the air handler for ducted space heating. Thanks for any help.

Register for a free account and join the conversation

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