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

Heat pump efficiency

user-1106403 | Posted in General Questions on

I am designing 2700 sf 2 story house outside Sacramento, zone 3B. BeOpt suggests I will have about a 17,000 Btu heat need. AC is also a big need in summer, and since there is no natural gas, I figured on all electric. Propane appears to be quite a bit more money.
Design low temps appear to be well above 20*; it’s rarely below 29*, and only then for a few hours. Here is the question; my Title 24 consultant stated that heat pumps incur large electrical use when outside temp is cold. I thought newer systems were efficient down below 0*, not that Auburn has seen 0 since the last Ice Age. Are mini-splits different from a ducted heat pump? Should I consider an alternative heat source for cold nights? I will have R21 open cell walls with R4 foam board on walls, R50 non-vented attic in open cell foam, acH50 test specified at 3 or less. PS; my wife does not like air-handler heads, so mini-splits, unless ducted, are out.

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. user-1105741 | | #1

    Specify a 1.5 ton heat pump to be installed with deep buried, leakage free ( R-8 ducts, radial design, utilizing a mixed air distribution system, and backup heat only during defrost (if at all). Program your thermostat to maintain comfort set points with only modest night setbacks, and you will be fine in Auburn.
    The only problem I would see, with an even more severe climate up here in Redding, is that equipment manufacturers don't make equipment small enough for our needs. The 1.5 ton is oversized, but it's the best we can do.

  2. user-1106403 | | #2

    Thanks for the feedback. Would you put this in the non-vented attic, under the house, or run the ducts in between floors?

  3. GBA Editor
    Martin Holladay | | #3

    Michael McKinley,
    You can follow Mike MacFarland's suggestion, or you can use ductless minisplits (or even ducted minisplits). Your choice.

    Evidently Title 24 is a little behind the times when it comes to air-source heat pumps. As long as your local building inspector doesn't object, minisplits are a good way to go.

  4. Expert Member
    Dana Dorsett | | #4

    Title 24 allows ductless mini-splits, but there are inspection steps worded into the code for verifying the refrigerant charge that simply isn't possible with ductless systems- it's a slightly different paradigm. There is prototype language revision addressing that available on the relevant state websites for the version that goes into effect on 1 January 2014 if the inspectors balk.

    Right-sized ductless systems are great from a comfort point of view because they modulate over a wide range of outputs, leading to very long cycles and extremely steady room temperatures. A well designed high-R 2700' 2-story with a whole-house peak load of only 17KBTU/hr would probably be able to use a 2-head 1.0- 1.5 ton multi-split, or a pair of 3/4 ton mini-splits. In a house like that the early & late day solar gains during the cooling season are more likely to create room-to-room temperature balance/discomfort issues than during the heating season, but that would be primarily for doored-off rooms. With an open doorway the convection exchange alone would usually keep cooling peaks well bounded during the peak gain periods, and after dark it would fall to background. But both the room by room solar gains as well as the room-by-room heat loads have to be considered when optimizing the placement of the heads. Mini-ducted heads that split the output between pairs of adjacent rooms can fix a lot of those issues too, but take up a bit of space. It's common to locate them in the ceilings of closets, etc to minimize the visual impact.

    Ductless heads don't have to be some blob on the wall. Cassettes that look similar to ducted register grilles flush with the finish ceiling are available from several vendors. (eg: When going with anything but the standard wall-blob package be sure to verify compatibility between the cassette and the compressor unit though- HVAC contractors aren't always a diligent about that as they should be.

    I was talking to a guy over the weekend who has given up on Mitsubishi installers since the catalog listings for different cassette options are so endlessly long, and the contractors he had been using were so sloppy about verifying compatibility. It wasn't a performance issue, but rather a contractor lack-of-due-diligence issue- YMMV. With other vendors the options are fewer, but that makes the selection easier.

    I strongly suspect that your peak cooling loads will exceed your peak heating loads, but it's very design & site-specific. A lot of unshaded west-facing windows can throw the cooling loads way over the top if you're not careful.

  5. user-1106403 | | #5

    Martin; to clarify, it's not title 24 that states heat pumps kick in with big electrical loads; it is the engineer doing the calculation. Her lack of knowledge is discouraging, to say the least. Dana; thanks, I've got very few west windows and most are hidden under a 7 foot wide front porch. Is there any reason that 2 mini-splits with appropriate grills would work better that one 1.5 ton heat pump? The heat pump incurs duct costs, but is not much more money that smallest minisplit for the equipment. I would likely need two mini-split, 2 heads off 1 system downstairs, 1 head off a second upstairs.

  6. Expert Member
    Dana Dorsett | | #6

    The 2-splits vs multi-split issue often comes down to what's easiest/best/possible from a refrigerant-line routing point of view for the optimum placments of compressor(s) & cassettes.

    It's the efficiency losses inherent in ducted systems (even well designed & implemented ducts) more so than the upfront system costs that make ducted systems less attractive. Zone dampers create back-pressure, drive air infiltration, etc. Two story homes rarely (never?) work well as a single zone in either heating or cooling mode, and zoning ducted systems takes the system-efficiency down yet another notch. While a few fully-modulating ducted heat pumps systems exist they're even more expensive, and likely to be even more oversized . A 1.5ton bang/bang single speed or 2-stage system is insanely oversized for a single zone, and will be nowhere near as comfortable as a reasonably-sized mini-split at any efficiency, and they're much louder to boot.

    Your low low loads are in the sweet-spot range for mini-splits, and are at or below the bottom of traditional split heat pump systems. Don't spend anything extra on a less well suited, less-comfortable equipment. You might think about spending a bit extra on lowering the loads for the doored-off rooms once you think you're really going ductless though. For example (and without reviewing the particulars of the design) while high-performance triple-pane windows might blow the budget if used everywhere it might be just the ticket for minimizing the load (and there by the temperature delta) on a doored-off room not directly served by the ductless head/cassette. Unlike the ducted paradigm, the house & system need to be considered together as a system a bit more carefully with ductless solutions, but they still work, and work quite well. It's always better to optimize it at the design phase than after you're living in it.

  7. user-1105741 | | #7

    The best place for the distibution system is within conditioned space, but not always possible. We live in a dry climate, and as such, have developed installation practices that allow us to optimize performance in any space- unconditioned attic, sealed crawlspace, and conditioned space between floors. It's too much to explain in a forum environment, but if you find someone in your area who happens to be a member of the previously mentioned leak-free duct society, your ticket to measured performance will have been punched.
    One of the issues we have here in the golden state with ducted (and ductless) mini splits is the inability to control our coil temperature (through high airflow) to ensure we are doing much higher sensible heat removal in AC. We design our systems with airflows around 500-650 CFM per ton- something not available in these mini ducted air handler units, and something that gets missed in these nationwide forums.
    If you want a quiet, very efficient, concealed system without stratification and with very consistent temperatures between rooms, I highly recommend the approach of an ASHRAE based, mixed air distribution system utilizing engineered supply outlets and common, everyday mechanical equipment that can be serviced and repaired by any service technician with the parts in his van.

  8. Expert Member
    Dana Dorsett | | #8

    High cfm still comes with higher air-handler power use. Even in dry AZ/CA I'm not seen convincing evidence of increased cooling efficiency or comfort in ANY ducted system. (But if you have in-situ data rather than a mere theoretical basis, I'd love to see it!) In tight high-R houses with way lower-than their neighbors' sensible cooling loads, even in dry climates the latent load can be a large fraction of the cooling load too.

    ASHRAE specs overventilate, and mixed air systems typically over-dry during the heating season when outdoor dew points bottom out. (I absolutely DETEST the mixed air systems at my office- it's far too dry for health & comfort most of the year, better in the shoulder seasons when the duty-cycle is minimal. I don't own the building, and only have thermostat control over the HVAC system.) It's far better to ventilate with separately controlled ducted heat recovery ventilation (HRV), and use it to compensate for temperature balance and stratification issues. High-R houses have very little room-to-room temperature differences as a rule, the exception being the high solar gain rooms in summertime.

    The NEEA study also includes eastern WA/ID which have comparable or even lower summertime dew points as all but the driest parts of CA. Looking at datasets, Boise has drier summertime air than Sacramento, and only somewhat lower sensible cooling loads. US climate zone 3 is still a heating dominated, not cooling dominated climate, and there's little question that mini-split will blow away ducted air source (and even many ground source) heat pumps on total system heating COP in that climate.

  9. user-1105741 | | #9

    Hi Dana,
    Yes, I absolutely have that data, as I don't believe in anything other than what I can measure. Last summer we explored exactly that presumption, and found some counter-intuitive results. If "high power use" is 129 Wh for 852 CFM, at 568 CFM/Ton AC, then yes. But at 6.6 CFM per Watt, compared to 2 CFM per Watt as the new home CA average, we think its pretty darn good. Especially with zero leakage, not 6 percent code allowable "pressure relief", and total external static pressure at 0.33 in W.C.

    We test every system we install, and ensure an "A" grade (90% or better) of delivered energy into the home (versus manufacturer specs). We do this regardless of system location.

  10. user-1105741 | | #10

    I just read your post further, and the mixed air distribution system I am referring to has nothing to do with the HRV system. Our installations are stand alone HRV's. I don't mean any disrespect by this, as I know you are very smart. Perhaps when I reference "mixed air systems", I am using incorrect terminology. But I am speaking, as previously referenced, of the ASHRAE definition (see Fundamentals 2009, 20.7) of "Mixed Air Systems."

  11. Expert Member
    Dana Dorsett | | #11

    I'm not as smart as some people seem to think- don't make too many presumption, eh? ;-)

    I don't have the the ASHRAE book in front of me to refer to. This building has a variable air volume system that mixes in ventilation air at some pre-determined fraction, that (apparently) does not adjust with season or load to compensate for humidity, nor does it actively humidify/dehumidify to a humidity setpoint. It's most comfortable in the shoulder seasons when the outdoor dew points are reasonable interior dew points, and it's in heating mode. Mid-winter is bone-dry, mid-summer it's usually comfortably dry but not always- it depends on the outdoor dew points and the sensible cooling load. Beats me how it really works- as I say, we're just tenants here, and the VAV mixed air buzzwords were mentioned when we moved in- so far I'm not overly impressed with it, but can't say it's dramatically worse than most R & D space I've been in, just slightly worse, primarily on the humidity end, and I suspect it's because it's dialed at ASHRAE ventilation volume requirements (that are arguably much higher than necessary for our low-indoor-air pollutant operation.)

    I have no means of analyzing the two-sentence performance description reported in post #9 relative to anything. Got a full-comparison writeup to publish with multiple in-situ datasets? The NEEA study monitored dozens, possibly hundreds (?) of units in-situ over the course of many months and compiled performance averages etc., as well as bench testing of a few representative units over a wide range of operation/load in a controlled lab setting (verifying or sometimes disagreeing with the manufacturers' stated performance of those models.) Got anything remotely like that to look at? (It's way more than a commissioning test during installation.)

    I'm still not sure how AC efficiency trumps heating efficiency in a very low-load house in a heating dominated climate, or how any non-modulating & ducted air source system is going to come close to better-class ductless performance on the heating end,.

  12. user-1106403 | | #12

    Dana and Mike; some of the lst few answers are over my head. I believe they pertain to the comments one HVAC person mentioned to me. She said even if you have a tight house and low BTU needs, you still need to have enough air movement, CFM, to be comfortable and move heat/cool around. She quoted a 3 ton and a 2 ton dual system, 60,000 BU for the above house. This seems like way too much system. PS she would not break down equipment cost vs install (22,000 total) so I said no thanks. If a service provider can't be transparent, why work with them?

  13. Expert Member
    Dana Dorsett | | #13

    Having been in houses heated and cooled solely by 1 & 1.5 ton mini-splits, I can say with some confidence that the notion that ...

    " still need to have enough air movement, CFM, to be comfortable and move heat/cool around..."

    ...has little bearing on the actual occupant-reported comfort levels in those homes. You don't get comfort out of a heating & cooling system, you get it out of a high-performance building envelope that doesn't NEED to "... move heat/cool around..." , which is what you have.

    Yes, on multi-story homes there can be significant temperature stratification, but that can be managed by multiple ductless heads (at least one per floor), if a ductless solution is where you're going. Oversizing the system 2-3x just for higher air movement seems like a mistake.

    Third party bench tested partial load cooling on mini-splits is VERY high, and the coefficient of performance even running flat out at highest fan & compressor speed is still about 4 at Sacramento's 1% outside design temp of 97-98F.,70,454

    At the 99% outside design temp of 30-32F a typical 1-ton mini-split can deliver the whole 17,000 BTU/hr at a COP just shy of 3, but you may be better off with a pair of 3/4 ton units or a 1.5 ton 2-head multi-split to deal with the stratification issues, (and to have a bit of margin during the next ice age. :-) ) When it's under +10F outside the whole system COP drops to about 2, but they still have substantial capacity down there in single digits F, eg, a pair of 3/4 ton Mitsubishi FE09s is good for about 22,000BTU/hr @ +5F. a pair of Fujitsu 9RLS would run about the same at that temp, maybe a tad more, and either would have PLENTY of capacity margin at your 99% outside design temps- about 27-28K @ 30F for a reported 17K load.

    That amount of oversizing isn't enough to cut into efficiency since they're fully modulating systems, but going to a pair of 1-tons might. A pair of name-brand 3/4 ton mini-splits would be about $5-6K (installed price) in my neighborhood. YMMV.

  14. user-1106403 | | #14

    Thanks, Dana, as usual, U the Man! Now to find an installer.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |