GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

What would our best heating option be?

julienAK | Posted in Energy Efficiency and Durability on

We live in Juneau, Alaska and our main heating options are electricity from hydro power (14 cent a kilowatt) or oil (diesel #2; $5 a gallon). We have an old oil fired hot water base board heating system we would like to replace with an energy efficient system. We can get a $10,000 rebate from AK Home energy rebate program on renewable energy sources (Air and ground source) and we are also looking at the federal tax credit. And either an air or ground source system reduces our electricity cost by about 1/2 for specialized rates. Our house is a 2.5 star rating, 1,500 sq ft, one level, single family home from 1970. We are worried we would have to come up with another heat source, if we chose air because we could be below freezing for several weeks. The bid we got for geothermal heat pump it would be $38,000 including duct work–with rebates and tax credits we would pay about $20,000.

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. GBA Editor
    Martin Holladay | | #1

    If I understand you correctly, you are eligible to buy electricity at 7 cents a kWh if you install a ground-source heat pump or an air-source heat pump. That's not a bad price.

    If your climate is mild enough to permit the installation of a Mitsubishi Mr. Slim Hyperheat ductless minsiplit system (a type of air-source heat pump), that's the route I would take. For that approach to work, you need to have a climate where the temperature rarely gets below -13 F.

  2. Expert Member
    Dana Dorsett | | #2

    What Martin said- despite a relatively high heating degree days number, Juneau is a fairly temperate climate in terms of the temperature extremes, with a 99% outside design temperature of +7F. (That's warmer than my design temp in Worcester MA, even though you have ~50% more heating degree days.)

    Almost any of the bigger mini-split heat pump vendors would have product suitable for your climate, you're not just limited to the Mitsubish H2i Hyper Heating series. Fujitsu, Daikin, Sanyo, LG all have decent models with rated performance at 0F or -4F/-20C, and will still be putting out a decent amount at temps below that.

    Almost all ductless heat pumps are fully modulating systems with variable speed blowers & compressors, which results in VERY stable and comfortable indoor temps, with a minimum of "wind chill"- the exit air temps are also above body temp, unlike most tepid-air ground source heating systems, that require careful consideration for locating the registers so that they don't blow on the humans. With the warmer exit air of ductless the primary siting issue is distribution, not wind-chill (though high-cfm heads can be pretty breezy at high-speed.)

    The mid-winter outdoor temperature in Juneau is about 28F, according to Weatherspark data:!dashboard;q=juneau%20ak%20USA

    Which will yield efficiencies from better-class ductless air source systems comparable to ground source heat pumps, at a fraction of the cost.

    Key to success for any heating system is to do a careful room to room heat loss calculation and sizing it right. An oversized heat pump will lose efficiency to cycling & standby, and won't be as comfortable due to higher air speeds and on/off cycling.

  3. julienAK | | #3

    Thanks for the input. We are seriously considering it, we are still a little concerned that we need to find a decent back-up. The climate here is usually above 0, but it is a really wet climate (temperate rainforest), and people in this area have had ice build-up on the systems. Has that been a problem in other areas of the country?

  4. wjrobinson | | #4

    Julie if ice build up is a problem... to me... don't duplicate that for sure. Dana?

  5. Expert Member
    Dana Dorsett | | #5

    The Mitsubishi H2i series (as well as many others) sense ice buildup and automatically executes a defrost cycle . It does this by reversing the refrigerant flow (and heat flow) for a few minutes until the ice clears, then resumes it's normal heating cycle. The defrost cycles are still very efficient, and as long as the delta between the outdoor air's dew point and the outdoor air is more than 10F there is only small hit in heating capacity relative to it's nominal rating. But there will be nights every winter where where the separation between outdoor temp & dew point is less than 10F, and during those hours it will deliver less heat than advertised, and possibly less heat than it takes to maintain indoor temp for those hours.

    How bad is this issue?

    Not as bad as you might think. For the Mitsubishi H2i series the rated output at +17F when the dew point is +10F (a 7F separation) is reduced to about 60% of full output. If that condition were to persist all night or all day it might be a problem, but that is NEVER the case- even in the foggy-dew cool maritime climates. When outdoor temps are below +25F the dew points will usually plummet, and periods where the delta-T is less than 10F persist only for a few hours. Those hours are usually in the pre-dawn hours when most of us are asleep. When it's below 10F outside the dew points are reliably below 0F with very little moisture content, and the capacity hit from defrosts fall away.

    At dawn the outdoor temp rises much more quickly than the dew point, resorting the temperature delta and boosting the heating capacity back to it's expected output at that temp. While a reduction to only 50-60% of output at +17F may SOUND like a hazard it really isn't. If you've size the thing EXACTLY for it's full output +5F output to support the load at that temp, the heat load at +17F is only ~80% of the +5F load, and the full-rated output at +17F with dry air is 125% of what you need to stay warm at +17F- it's normally still not running full blast- it's modulating. During the occasional periods where conditions compromise capacity, it ramps up trying to maintain temp, and even though it can't it's still supplying 70-75%, of what's needed to maintain 70F in the house. With heat at 70-75% of the actual load most houses would be losing less than a degree per hour for the first few hours, slowing as it approaches the temp at which the output matches the heat loss of the house.

    If the shortfall persisted for fully 4 hours, would it be a disaster to wake up to a 66-67F house instead of a 70F house? In practice most of the time the shortfalls never last long enough even to be noticed- capacity ramps up dramatically at dawn to make up the temp, and the impact on comfort is near-zero. The better insulated the house is and the more thermal mass there is inside the house, the lower the impact these events are.

    You'll see a lot more defrost in action when it's a few degrees above freezing with a narrow separation between dew point & temp (think "35F morning fog"). Under those conditions there is be copious icing and more defrost cycles, but the impact on capacity is quite low. This is in part because it's warmer outside than in the +17F example (warmer air always improves capacity), but also because the amount of heat that has to be given back to defrost the coil is a small fraction of the heat that had been taken in by the changes of state from water vapor to ice: The heat pump reaps the full benefit of the heat of vaporization of the water, but only has to give back the heat of fusion (which is smaller) to clear the ice. I have family in western WA heating with mini-splits where 35F & fog or rain is a common AM condition, yet they have to pay attention to notice the defrost cycles, even though they're happening at least once/hour sometimes more. I've yet to hear of a capacity issue at or near their design temp of +20F or lower, even though it's sized pretty close to the +20F heat load. (I'm sure at least some of those winter nights it's output was below the heat load for a few of hours though.)

    Even though it should never happen on a mini-split with auto-defrost, should it ever ice up the solution is to put in in air-conditioning mode for 10-15 minutes, then peeking to see if the ice has cleared on the coils before turning it back to heating mode. I suppose in mid-western style freezing-rain ice storms this might happen if a unit mounted in the open rather than protected by the eaves/rake of the roof (or some other shed-roof, etc), but won't happen in normal operation unless you picked a unit that doesn't have built-in defrost.

    See if you can't find somebody locally heating with mini-splits and ask them about frost-up issues- I'm betting the problems are few, and nothing like the misery some ducted heat pump systems with resistive-heat defrost go through in that kind of climate.

  6. Expert Member
    Dana Dorsett | | #6

    I need to correct a mis-conception I've been spreading about where & how mini-splits lose capacity. It's the wet-bulb temp not the separation between outdoor temp & dew point that's the determining factor. In the Mitsubishi spec the "rated capacity" number is the number and conditions at which their stated efficiency at that temp is valid, and the capacity is really the "maximum capacity" number.

    See the response from Marc Rosenbaum titled "Back to minisplit capacity and dewpoint" on this page:

    I find the corrected facts comforting, from a "will it keep up in my climate" point of view!

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |