What would you do?
December 8, 2012 1:06 AM   Subscribe

Is a heat pump worth it? Online calculators are a hugely mixed bag with obvious biases on many of them, and installers all want to sell their favorite gizmos so it's hard to get a straight answer from them.

Our current details: forced air propane furnace and water heater, both getting close to the end of their useful life. The furnace is 80% efficiency, and I have no idea about the water heater, though given the choices made in this house, I expect it's roughly the same. It's a 50 or 60 gallon unit.

Current prices: We pay an average of $0.10 per kWh for electricity. We pay $2.30 per gallon for propane.

Our house: roughly 1700 sq ft, two stories, but one end of the house is a two-story space. Build in roughly 2001, so reasonably insulated.

Our climate: This is us, in degrees F (avg high temp, avg low temp, avg temp):
January   | 45.3| 34.8| 40.0
February  | 48.5| 35.6| 42.0
March     | 52.6| 37.6| 45.1
April     | 57.1| 40.2| 48.6
May       | 62.6| 44.8| 53.7
June      | 66.5| 48.3| 57.4
July      | 70.8| 50.7| 60.7
August    | 71.9| 51.0| 61.4
September | 67.3| 46.9| 57.1
October   | 58.2| 42.1| 50.2
November  | 49.8| 38.1| 43.9
December  | 45.0| 35.0| 40.0
I would say in each of the four winter months there's a period of a few days where it's really cold. It's very unusual to have a prolonged cold snap, but it can happen. "Cold snap" around here is 25F for a few days.

Incentives: The local energy company will give us a $1,500 credit to switch from forced air propane to any type of heat pump.

Usage: The furnace comes on intermittently in October, is regularly on in Nov - March, and is usually off again by early May. We keep the house around 67 during the day and 64 at night.

There are maybe a few weeks over the summer when some cooling would be nice, but it's not necessary.

Other: I don't mind paying a bit more for a system, but cannot afford to spend, say 4x for a heat pump if it's not going to save us a lot over time. I'm confident a ground-loop system is well-outside our price range. We have ducting throughout the house so would prefer to re-use that for delivery of heat.

I mention the water heater as I don't mind it staying propane, but would have no objection to using the heat pump + electric or solely electric for that. We do keep the water fairly toasty, maybe 130 degrees, we never run short with our current tank, there are two of us here.

Those two energy choices are what we have available; natural gas is not around, solar is too expensive (but would be compatible with the heat pump, I guess, in the future if we somehow did that), and we do not have a wood source of heat and really don't want one. Environmental impact is something I like to think about but budget will intrude on anything grandiose.

Many thanks for any insight.
posted by maxwelton to Home & Garden (10 answers total) 6 users marked this as a favorite
I work at a sustainable energy charity making software models, including models of heat pumps, and it's certainly true that
  • Heat pumps are more annoying to evaluate than pretty much all other technologies
  • Comprehensive manufacturer data on heat pumps is quite hard to come by; the manufacturers tend to give simplified descriptions under the ideal conditions for their device (unsuprisingly)
Anyway, to do a quick explanation of a heat pump (skip down for the TLDR), open this picture (apologies for hasty crude drawing). In this image, the larger empty box represents your house, and the smaller box is the heat pump. The heat pump has two flows going through it, one at temperature Te (the external "cold-side" temperature), and one at temperature Th (the "hot-side" temperature). The hot circuit then goes into the little box marked R, which is the heat distribution system. In a wet heating system this would be some radiators, although it sounds like you have a warm air system already - I don't know a huge amount about these.

Next we can come to the power, Pout and Pin; Pout is fully determined by the specific heat loss H of your house, a value in watts / degree K, and the difference between your desired ambient temperature and the outside temperature, according to Newton's law of cooling:

Pout = H (Td - Te)

So, to keep your house at comfortable Td, the emitters have to put out Pout power (less the incidental gains in your house from sunshine and people and stuff, which we will ignore here). To emit Pout into the house, your emitters will need to have a certain hot-side temperature from the pump; for wet systems with radiators this is something following P = Area * alpha * (Th - Td) ^ 1.3. Not sure on warm air I'm afraid.

Finally this gets us to Pin - the heat pump's coefficient of performance is largely determined by Th and Te (although also by Pout, in a real heat pump, because at very low powers the baseload contributes relatively more and lowers the efficiency). In an ideal pump (which upper bounds the value for any real pump), the COP is

nu = Th / (Th - Te)

and Pout = Pin * nu

The key thing to observe about nu is that it gets bigger as Th approaches Te, meaning that when the hot-side and cold-side are close, the efficiency is improved, but when they are far apart, the efficiency is much worse.

This gets us to the bottom line and TLDR:
  1. If your specific heat loss is bad, you will need to deliver a lot of power into the house to stay warm
  2. If you need to deliver a lot of power, the circulating temperature must be higher in your distribution system
  3. Also, for radiators, if your radiator area is small you need a higher circulating temperature; I presume there is something similar about the heat exchanger in a warm air system but I don't really know.
  4. Having a high circulating temperature makes your heat pump less efficient
  5. Similarly, if the outside temperature is very low, your heat pump will be less efficient
Thus my recommendation would be:
  1. If you get a heat pump, do your level best to reduce your specific heat loss, i.e. insulate as much as possible
  2. Also, you may need to buy a different bit for your heat distribution system which has a bigger area and so can run at a lower circulating temperature, but I don't know how that works.
Unfortunately all the theoretical detail only tells you on which dimensions variation is important (specific heat loss, radiator size, external temperature distribution), but getting hold of actual COP data by these dimensions for a device which lets you do the cost-benefit is very hard.

One thing which might be useful is this website, which is a live feed from several houses in the UK which have Mitsubishi heat pumps installed, but these will almost certainly be wet heating system, and I don't know how the climate compares. They have lots of historical data for demand, output power, external temperature etc. To give an idea of how complicated real heat pump data is, here is a plot of the COP surface I made using data scraped from that website. It's all over the place, and way different from the theoretical model which is a nice smooth surface with a singularity where Th = Te.

In conclusion: it's really hard to work out whether heat pumps are a good idea or not - you need a lot of different data to get a good result, and some of those data are hard to come by.
posted by larkery at 2:17 AM on December 8, 2012 [15 favorites]

Oh, I forgot to mention another confounding factor - in my diagram I showed the case for an air-source pump where the external ambient temperature is equal to the cold side temperature for the pump. For a ground sourced system, the cold side temperature will be quite a lot warmer most of the time, because the earth doesn't get as cold as the atmosphere.

This makes ground sourced systems a bit better, and compounds the badness of air-source systems on a very cold day (power required is greater and cold-side temperature is less).

Also, on that mitsubishi website you have to click around a bit to get to the good stuff, for example this page. If you change the number at the end of the url you can get to the other logged houses without having to click around endlessly.
posted by larkery at 2:20 AM on December 8, 2012

In my experience the difference between the comfort of a forced air gas furnace and a forced air heat pump are huge. If you are used to a gas furnace you will not feel comfortable with a heat pump.
posted by jmsta at 4:33 AM on December 8, 2012

Just based on your stated temperature range, I'd say you could do a heat pump with little problem. At the coldest extremes, though, you might run into issues where the pump is running continuously and what heat it puts out to be pretty mediocre.

Throughout the rest of the range, a heat pump will work just fine.
posted by Thorzdad at 5:26 AM on December 8, 2012

If your water heater is at the end of its life, consider getting a tankless heater. We switched to one last year and its been great. We live in the Boston area, so it's not as cold as where you are, but the ground water temp does drop quite a bit. Also, why do you keep your tank so hot? Do you get a big drop in temperature at your faucets?
posted by reddot at 5:34 AM on December 8, 2012 [1 favorite]

I live in the southeast where our winter climate is similar to yours at the coldest end of the scale and the summer months are a lot hotter. I have had heat pumps for years and had no problem with them. Unless you really run your furnace hot you shouldn't tell any difference between a heat pump and other forms of forced air heating. On really cold days the auxillary/backup heat (usually electric coils) will keep the house plenty warm but with a big hit in efficiency. This usually happens around 35 degrees. Since you already have gas at your house a dual fuel heat pump might be an option if you are concerned about cold-weather performance. I replaced one of my heat pumps a year and a half ago and discussed the options extensively with my contractor (whom I have known socially for years and who I trust to give me good advice). I asked about ground-loop systems because they are so incredibly efficient and he said that they are very difficult and expensive to install properly and if a leak develops anywhere in the hundreds of feet of tubing buried in the ground the whole thing may need to be dug up to find it, an expensive repair. I ended up going with the high-efficiency Carrier unit (other top-end brands should be just as good, that is just what my contractor happens to carry). I have noticed some definite savings over my older, lower-SEER heat pump, as much as 20% in the summer. The new unit is also much quieter. Minimal modifications to the ductwork were required to make the switch. So I would definitely look into a heat pump in your position; I have been happy with mine. Advice from a trusted local HVAC person is invaluable.
posted by TedW at 6:25 AM on December 8, 2012 [1 favorite]

If your major thermal need is heat rather than cooling, another option you might want to run some numbers on is combined heat and power using fuel cells.
posted by flabdablet at 7:41 AM on December 8, 2012

I don't know if it's an issue where you live, but around here (Quebec) the consumer magazine has done investigations on heat pumps and has found that a LOT of pumps were shoddily installed, which would cause a huge drop in performance. You'll want to make sure that whoever is doing the job is doing it properly.
posted by Monday, stony Monday at 8:16 AM on December 8, 2012

You have a few advantages that might lean toward a heat pump. You have relatively low electricity costs and relatively mild winter temperatures that improve the efficiency of the heat pump.

For the cost of propane heating: (91,000 BTU/gal) * (gal/$2.30) = 39,565 BTU per dollar.
Lets say you install a new 95% efficiency furnace, that brings it down to about 37,600 BTU per dollar.

For electricity: (3412 BTU/KWH) * (KWH/$0.10) = 34,120 BTU per dollar.
Let's say you install a heat pump that gives you an annual average of 2 BTU of heat for 1 BTU of electricity so you get roughly 68,240 BTU per dollar.

So the heat pump annual bill will be approximately 37,600/68,240 = 55% of the gas furnace, roughly a 45% savings.

The one piece of information that you didn't give us was you current annual heating bill. Let's say it is currently $1000 per year. A new 95% gas furnace should cut that to about $850 per year. A heat pump will save 45% of $850 or about $380 per year.

Finally you have to compare full installation cost of the new 95% gas furnace versus the heat pump. Keep in mind that the heat pump cost might require an upgrade to your electrical service panel. And subtract the $1500 subsidy. Make sure the quotes include sales tax for both options. Take the installation price difference and divide by $380 and that is the number of years to payback the extra cost of the heat pump (ignoring interest).

Bottom line, for a quick seat of the pants estimate, take your current annual heating bill, multiply it by 40% (to account for the new 95% gas furnace) and that would be the annual savings for a heat pump vs a new 95% gas furnace.
posted by JackFlash at 2:06 PM on December 8, 2012

Response by poster: Many thanks for all of the information!

The home fuel cell is really intriguing, but judging from the fact that one of the companies selling them have a calculator which told me our annual savings would be $600 and it would pay for itself in 36 years, I'm guessing it's outside our budget.

We're definitely leaning towards a heat pump, but will continue with our research. This is a next-year's-heating-season problem, so fortunately we have some time.
posted by maxwelton at 12:39 AM on December 9, 2012

« Older # of recording artists per given area   |   34 year old dating 20 year old -very confused Newer »
This thread is closed to new comments.