GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X 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

Ideal load for a multispeed inverter-driven GSHP?

user-4435615 | Posted in Mechanicals on

In a planned “Pretty Good House” in Mixed Humid, Northern Climate Zone 4, I intend to use a 45 EER Ground Source Heat Pump with 3 inverters (loop pump, compressor, and fan) with a Cooling Capacity range from 9,000 – 24,000 Btu. I wish to have the lowest Peak Cooling Load, that utilizes the equipment most efficiently, while providing dehumidification and keeps short cycling to a minimum. Approximately what sould that theoretical Peak Cooling load be ?

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

    Your cooling load can't be guessed from the information you provided. You need to do a Manual J calculation to determine your cooling load. For more information on calculating a cooling load, see Calculating Cooling Loads.

    Q. "I wish to have the lowest Peak Cooling Load."

    A. If that's what you wish, there are many steps you can take: Make your windows small; minimize or reduce the size of west-facing and east-facing windows; choose windows with glazing that has a low solar heat-gain coefficient; shade your windows with porches; pay attention to air sealing your thermal envelope; include plenty of ceiling insulation; and choose energy-efficient appliances and lighting to reduce your internal loads.

  2. charlie_sullivan | | #2


    I'm not sure I understand your question correctly, but my best guess is different from Martin's interpretation. My guess is that you have chosen your HVAC equipment, and now you are designing your envelope, and you want to make sure your envelope results in a load that operates your HVAC equipment efficiently. You want to make sure you don't make your envelope so good that your HVAC equipment is oversized, because it would then operate inefficiently, would not effectively dehumidify, and would cycle on and off too much.

    That's backwards from the usual recommendation: to first make the envelope as good as possible and then choose the equipment to match the resulting load. Arguably, it's better to work on both together, because, for example, if you could have spent $200 more on making the attic insulation thicker, you might be able to change to a different HVAC system, and by doing so save $10,000. Perhaps that is actually what you are doing.

    You list three disadvantages of oversized equipment, and asking for us to think of it as disadvantages of a better envelope (I think). Of those, I think two are red herrings from that point of view. Cycling can in general reduce efficiency, wear out equipment, and degrade dehumidification. With an inverter system, the wear caused by cycling should be minimal, and the efficiency and dehumidification are already considered separately. So that leaves efficiency and dehumidification to consider. For efficiency, improving the envelope will always reduce energy consumption, even if the equipment becomes less efficient in the realm you move it into. That means that your return on investment for improving the envelope is reduced, but the better envelope won't hurt anything.

    So finally we have the dehumidification question. That is the one that I think is a real concern. I actually think it's an underappreciated issue with excellent envelopes in warm, humid climates. If the envelope doesn't let in much heat, the main job of the A/C is to dehumidify, not to cool. Even with properly sized equipment, you can end up needing to cool to the point of at least wasting energy if not also making it uncomfortable in order to dehumidify.

    That last problem could be solved by using a poorer envelope. At least in the summer, that's better than the common solution in commercial buildings where a reheat system actively adds the needed heat to prevent over cooling. But a better solution is to run an efficient dehumidifier to provide dehumidification without excessive cooling.

  3. user-4435615 | | #3

    Thank you both for your input. In brevity, I lacked clarity. I think Mr. Sullivan caught my drift. Initially I began with the envelope and looked for the best equipment for that plan. Now I want to work the other way and see what option I have for the optimum load for the best equipment I can find.
    In this 3200 sf single floor, slab home, room layout is constrained by the view, lot orientation, and ADA accessibility; a central open floor plan, but with other rooms further off to the sides. The plan calls for 0.1ACH50 (interim construction pressurized fog testing).
    My initial approach called for recommended levels of insulation for a High Performance Home for CZ-5 (though I'm in Northern CZ-4A). Alpen Windows. Broad Overhangs. Etc. Manual J Load Calculations gave me a total Peak Cooling Load of about 10,000 Btu/hr accounting, among other things, for Solar Heat Gain, Radiant Time Series, & Convective Time Series. A Zehnder ERV/HRV is planned.
    A Ductless/Ducted Mini-Split might seem to be the obvious solution. However, the house will be a bit spread out; the bedroom doors generally closed at night; we like the bedrooms cooler during summer nights; outside equipment would be less than ideal, etc. Zoning or multiple single mini-splits further decrease efficiencies, and increase costs. Mini-splits are less efficient than a 45 EER GSHP ( I'm aware that this does not take the loop pump into consideration.). Ducted versions encounter drops in efficiency, all of which will end up being offset by additional PV panels on the roof. In short, I'd prefer not to go the Mini-Split route, though I'm not closed to it. I may come back to it later.
    For now, I'd like to pursue the other approach using a 45 EER, 5.1 COP , 3-inverter GSHP (that includes a water heater that heats my water year 'round at a COP of 5). It's a higher initial cost, but with superb efficiencies, no additional hot water equipment, 30% rebate, lower PV costs, no outside equipment and, after the rebate, nearly competitive with zoned or multiple single mini-split systems.
    So, knowing that this approach is a bit backward.....the question for me comes down to: (Take as a given, for the sake of argument, the GSHP above.) "How much insulation do I peel off the various layers of my current, probably over insulated, house plan in order to reach a peak cooling load that will match the equipment .

    My question of you is: What Peak Cooling Load, in Btu/hr, would most likely induce the multi-speed inverter- driven GSHP, operating at it's lowest capacity of 9,000 Btu/hr (range = 9,000-24,000) in a "Pretty Good House" , to provide the greatest comfort through the entire cooling season with greatest efficiency; that will effectively allow that equipment to cool, dehumidify and minimize short cycling ?

  4. GBA Editor
    Martin Holladay | | #4

    I have to agree with Charlie. You are approaching this backwards.

    I was struck by one of your statements -- that a ground-source heat pump will lead to "lower PV costs." Strictly speaking, that's true -- but the cost of the GSHP is so high that your cost argument is undermined.

    Many engineers and builders have compared the following two approaches:

    1. A ground-source heat pump plus a PV array sized to handle the heating and cooling load; and

    2. Ductless minisplits plus a PV array sized to handle the heating and cooling load.

    Option 2 is always significantly cheaper.

  5. charlie_sullivan | | #5

    Martin is certainly correct that option 2 is almost surely cheaper. But I can understand your interest in enhancing comfort and efficiency beyond just cost considerations. You might also value the fact that the net result will contribute less to peak loads on the grid, which might someday be something consumers have to pay for in some way.

    You confirmed my suspicion that your plan is to deliberately reduce your insulation so that your A/C runs more, and thus provides the necessary dehumidification. That could work, and might not be the efficiency disaster it sounds like, if the GSHP's COP turns out to be awesome. Ideally you'd want to have insulation panels you could remove for the summer and put back in in the winter for that purpose. However, I really doubt that it would be more efficient than using a good dehumidifier, combined with running the GSHP only as needed for cooling.

    If your main interest in the GSHP over minisplits is being able to do zoning and being able to serve small dispersed rooms individually, another thing to consider would be the Chiltrix system:

    It has similar performance to minisplits, but with the ability to have more small heads because it distributes the heat hydronically rather than with refrigerant. It also uses significantly less refrigerant and is less likely to leak refrigerant, which I like because of the high global warming impact of refrigerant that escapes.

    If your main interest the the GSHP is the satisfaction of owning cutting edge awesome low-energy-consumption equipment, consider getting a solar desiccant dehumidifier, rather than removing insulation, to solve the dehumidification challenge that the oversized GSHP might have.

  6. user-4435615 | | #6

    Thank you both again for answering. I've quickly looked at the information suggested about the Chiltrix and the Novelaire and I will investigate them further. Though, finding anyone here locally, in what could be considered an energy conservation backwater, to sell, install, and maintain that equipment, could be a trick.

    Returning to my previous inquiry - I realize that the idea of reducing insulation goes against the grain of everyone who reads this site, and it does for me as well. However, I am trying to consider the advantages and disadvantages of as many options as possible to determine which one is best, in the circumstances unique to this particular situation, from the standpoints of economy, efficiency, comfort, ecology, etc. etc. I even appreciate the desire that others may have to protect me from myself and dissuade me from pursuing this vein. But, at this stage I'm not building, just following a line of inquiry, investigating one option among many. I'm considering the option of going from building a High Performance Home suitable for one Climate Zone to the north, to a High Performance Home or "Pretty Good House" for this CZ so often espoused by this site.

    Let me come at this question from a different angle.
    Presume that I ALREADY HAVE an incomplete framed, enclosed, but uninsulated home. It has a 45 EER, 5.1 COP (Note that the loop pump is also inverter driven) GSHP IN THE GROUND that produces Hot Water at a COP of 5 (which cools the home as it does so in summer, but does not cool the home during the winter). It performs at those efficiencies at all outside temperatures, never needs defrosting, can't be blocked by deep snow, etc. The house will have a moderate thermal mass; enough insulation suitable for a PGH; and prolonged Radiant & Convective Time Series. Pressure testing with fog came in at 0.1ACH50. The sub-slab and foundation insulation is in place and I am ready to begin applying the insulation to the walls and the roof of the Unvented attic. I know that the more insulation I add the lower my loads will be, but I don't know how much I should add. I belatedly come to the sage of misty mountain and advise her that my multi(I think it's 7)-speed, 3-inverter driven equipment's capacity is between 9,000-24,000 Btu/hr.

    The sage says.... "Both of my computer programs tells me that if you add 'X' amount of ocSPF to the roof sheathing and 'Y' amount of PolyIso to the outside of your wall sheathing, your Peak Cooling Load, at the design temperature for your area, will be __________Btu/hr. I would advise that you NOT add any more insulation than that because you are already at Pretty Good House load levels; and, if you add any more insulation the loads will be lower such that, though your equipment will continue to cool. it will actually begin loosing efficiency by short cycling and will dehumidify less well. Yup,.. I'd say that's your Ideal Peak Cooling Load, given your circumstances; that's where you'll operate most efficiently and be most comfortable for most of the cooling season...That'll be one chicken please".

    As esteemed experts, inveterate puzzle solvers, and fellow sages - what is the number that the sage advised ???

  7. charlie_sullivan | | #7

    I'll take a stab at your question instead of trying to divert your attention from it. Even if you had a 20kbtu/h peak load, there would be days when the load would be less than 9kbtu/h, so you'd already be cycling those days. If you improve insulation, you will decrease the number of days it can operate without cycling, until you get down to 9k load, at which point the only control option is cycling, and it works like a regular system that doesn't have modulation capability. That sounds like a loss, but the degradation due to cycling is at least partly offset by the increased efficiency running at 9k rather than higher. So without digging into the data for the particular unit and running simulations with climate data, I would assume that there is no change in efficiency worth worrying about down to that point.

    Now consider increasing insulation beyond that point. Now it is just a question of how much oversizing hurts. This report indicates that it is on the order of a 15% penalty for 2X oversizing. http://www​
    So if you make the envelope twice as good and drop the load to 4.5k, your electric consumption drops by "only" a 42% instead of the 50% you paid for.

    Bottom line: it doesn't matter as much as you might think, given how much ranting you hear here about the evils of oversizing. It's not a reason to stop improving the envelope.

  8. GBA Editor
    Martin Holladay | | #8

    Thanks for providing a helpful analysis.

  9. user-4435615 | | #9

    Yes, Thanks. And thanks for continuing this conversation with me. I read the research about short cycling. I had been a little concerned about the efficiency loss due to short cycling and that article is reassuring.
    There is still the larger and much more concerning problem of dehumidification for me yet to deal with. I'm sure that you understand, though you may not agree, with my pursuit. It's just an option, among others, that I wish to investigate to it's end. The efficiencies of the unit are so good that I can still meet a larger cooling load and still require less power to do than other options. In addition it can provide all the heat and hot water through the heating season at a COP far above most other options.
    In the 'old days' HVAC contractors used lousy rules of thumb to predict the needed capacity of equipment for a poorly defined load in leaky, poorly insulated homes with little thermal mass. Then they were only able to provide large single speed equipment in multiples of Tons. In my instance were dealing with a load that I can relatively precisely define , a tight house, a moderate thermal mass and I have a machine that can match the load precisely , to a point. Surely HVAC folks familiar with Pretty Good Homes have a reasonable feel for the ratio between the load and the ideal capacity of the equipment. An HVAC consultant might say: "If you have "X" load then the capacity your most ideal piece of equipment would be "Y". There must be a ratio X:Y. In my case I wish to reverse the question. I know what "Y" is and I wish to use that ratio to give me "X".
    It seems improbable that someone does not have a relatively good feel for that relationship, or is it that no one really does have a good feel for it?
    I have included an attachment that shows my calculated loads [more appropriately suitable for CZ-5] over a 24 hour Peak Design day of 93 deg F. The load varies between a low of 5,342Btu at 6 AM to a high of 9,948Btu at 4 PM (the 24 hour weighted average is about 5,850). CTS represents all conductive loads adjusted for the Conductive Time Series whereas the RTS represents the Solar Heat Gain adjusted for the Radiant Time Series, both for light to moderate thermal mass. The Yellow line, CTS+RTS, represents my Total Load. Note that the Solar Heat Gain represents half the load. On a SUNNY mid-summer day this would work out fine however on a cloudy design temp day the load would drop to just above half the lowest capacity of my equipment, 9,000Btu. Still OK. But get away from mid summer and more particularly in the shoulder seasons and the loads would be notably lower and the equipment run times may be reduced to 10 to 15 minutes per hour. It is my impression that cooling equipment needs to run for a minimum of 15 to 20 minutes continuously to dehumidify. I'm going to re-design my insulation anyway to at least conform to more appropriate CZ-4 levels, which will increase loads a bit. I may as well know the best levels of insulation to use for this hypothetical question which of course depends on the result of that question: Ideally, What is the load to capacity ratio for a Pretty Good, Tight, Light-Moderate Thermal Mass House ??? What Peak Design Cooling Load makes that 9,000 Btu capacity GSHP provide the most comfort throughout the cooling season. ???
    An alternative that I've been considering is the use of 3 Ductless Minisplits and a Heat Pump Water Heater to provide the desired characteristics outlined in Section 3 above. Or, possibly Minisplits and a compatable Air Handler. Though I've not priced that yet, I'm not sure that, with the Federal credit, the cost will be all that much different, and I think that the GSHP could possibly be more suitable.

  10. user-4435615 | | #10

    I flubbed the attachment above. Hopefully it'll make it this time:

  11. user-4435615 | | #11

    Or This

  12. user-4435615 | | #12

    Or This

  13. charlie_sullivan | | #13

    One big factor in how well dehumidification works with equipment cycling on and off is what it does when it cycles off. One approach is to keep the fan running for a little while after the compressor turns off, based on the theory that the evaporator is cold and you paid for making it cold and so you'd like to deliver that cold to the conditioned space by running the fan a little longer. But the evaporator is also wet, and keeping the fan running as it warms up tends to dry it off and deliver that moisture to the conditioned space. So when humidity is an issue, its better to stop the fan right away and leave the evaporator cold and wet, and hope that the moisture either stays on the surface or drips off, rather than evaporating.

    So you might see if the manual specifies that, or ask the company if they don't specify.

    Beyond that your question might be which dehumidification options uses less energy per liter of water removed: a high-performance dehumidifier, with the GSHP removing heat it puts in the space, or the GSHP running some extra hours, and overcooling prevented by having weaker insulation than you might want otherwise. Without crunching numbers on that, my intuition is that if you had an adjustable heat leak, so that you would run the heat pump just enough the get the dehumidification you want, and then leak in just enough heat to make it a comfortable, that would use less energy than the separate dehumidifier. But with fixed poor insulation, you have more heat than you need leaking in on hot days, and of course heat leaking out in the winter. So my intuition is that you'd e worse of overall on a seasonal average basis, unless you have some kind of moveable insulation that you could use to gain heat when the necessary run time for dehumidification makes it too cold inside. But I think it would be a lot easier and not much worse energy-wise to set up the GSHP as originally planned, insulate well, and use a separate dehumidifier as needed/if needed, perhaps a solar desiccant one.

    Did you get a chance to look at the Chilltrix system?

  14. user-4435615 | | #14

    I've reviewed all the information on Chilltrix that I could find on-line plus several threads that are on-going here in GBA. The system sounds impressive and could be a good solution for me. Thanks for the heads-up. I have a request for contact into the company and have numerous questions to ask.

  15. cnote75 | | #15

    Chilltrix seems to be the same technology as Multi-Aqua, but I don't see Multi-Aqua making any of the efficiency claims of Chilltrix. Makes me wonder why.

  16. Dana1 | | #16

    Multi-Aqua is a single speed compressor, Chilltrix used a DC-inverter drive modulating speed compressors & blowers. The part-load efficiency while modulating is much higher than a high fixed-speed. When operating at part load the lower compressor speed and lower blower speed use less power, and the heat exchanger coils are "oversized" for the load, all of which adds up to much higher average efficiency. Even running at maximum speed the Chilltrix may still be somewhat higher efficiency than the Multi-Aqua due to the higher efficiency DC motors, but the difference isn't as large as when operating at part-load.

    A more subtle difference is that the Multi-Aqua used R407C refrigerant, while Chilltrix used R410A. These are similar, but not identical.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |