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

Sizing an Air-Source Heat Pump

acdc2 | Posted in Energy Efficiency and Durability on

I need help with air-source heat pump sizing.

My house is 1,870 sq ft ranch house with a 900 sq ft finished basement. The house is located in Westchester County, NY. The basement air is already conditioned with a 1-Ton ductless mini-split heat pump. I’m looking to replace the 50-year old central AC for the main level. The house has gas-based hydronic heating with baseboards. The gas boiler is a 120,000 BTU/Hr System2000 EK-1 boiler. It seems highly oversized for the house. The system cycles a lot and the temperature variation is quite high.

To increase the comfort level (keep the temperature steadier), I’m leaning towards using a heat pump to both cool and heat the house. In the winters, I would use the heat pump as the primary source of heat and when the outdoor temperature falls too low for the heat pump to handle, use the hydronic system as the secondary heat source. In the summers, the heat pump would be the only source of cooling.

I had a professional run Manual J and Manual D calculations for me. The Manual J report is attached. It shows total cooling required as about 30,000 Btuh and total heating required at about 47,000 Btuh. I want to go with a variable speed system in order to keep the indoor temperature study. I have narrowed down to the Carrier Infinity Greenspeed 25VNA4 series heat pump coupled with a Carrier FE4ANB005L air handler. My HVAC contractor is recommending the 4-Ton 25VNA448 heat pump (specs can be found at!/product/32814 and This is an overkill for summer months but meets the winter heating requirement.

My question for the experts:

1. Should I go with a 3-Ton 25VNA436 or the 4-Ton 25VNA448 heat pump? 3-Ton would be more than adequate for the summers. Given that I have reliable (though oversized) secondary heat source, would I be fine with a 3-Ton heat pump in the winters? One school of thought says to just buy what is needed for heating and let the heat pump adjust speed downwards to meet the lesser summer load. Does the fact that I have a reliable secondary source of heat impact this rationale?

2. Are there published efficiency figures on the variable speed heat pumps running at 25 or 50% of capacity? Without the efficiency ratings at low speeds, I don’t know whether a 4-Ton unit would be as efficient in the summers as a 3-Ton unit.

3. Would I have to wire the heat pump controls to the boiler controller or would it suffice to use thermostat settings to make the heat pump the primary heat source. E.g., I could set the heat pump thermostat to 70 degrees and the boiler thermostat to 65 degrees. Would that work?

I would greatly appreciate any input.

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. Expert Member
    Dana Dorsett | | #1

    >">The gas boiler is a 120,000 BTU/Hr System2000 EK-1 boiler. It seems highly oversized for the house. "

    D'ya think? :-)

    Even if the 47MBH indicated in the Manual-J were real it would be more than 2x oversized for the design load.

    And 47MBH seems inordinately high for an 1870' + 900' basement in Chappaqua NY (99% outside design temp ~ +10F), even if the basement load currently covered by the mini-split were real (and it probably isn't). I would anticipate a whole house load less than 40MBH if it's reasonably tight 2x4 construction with tight clear-glass storms over clear glass double hungs, low 30s if it's tight 2x6 with low-E windows. But without the full Manual-J it's hard to figure out where it may have gone off the rails. Was that Man-J from an HVAC contractor, or was it a qualified third party (like an engineer or RESNET rater)?

    Since you have a heating history on the place, first dig out some of last winter's gas bills and sanity check the Manual-J against a fuel-use per heating degree-day analysis, as explained here:

    >"Should I go with a 3-Ton 25VNA436 or the 4-Ton 25VNA448 heat pump? 3-Ton would be more than adequate for the summers."

    Take a look at this tool:

    Click on the Heating Capacity tab, set the design temp to +10F, select the Greenspeed air handler and compressor options in the lower left side, see where it takes you. The 3-tonner puts out about 25 MBH @ 10F with most air handler options, which may be enough to fully heat the 1870' of non-basement space.

    This series of heat pump is listed on the NEEP side, with a specified output at +5F (which is colder than your outside design temp, but probably happens at least once most winters.):!/product_list/

    Copy then past the compressor model number in the middle search selector.

    With some air handler options the 3- tonner is good for 28MBH+ @ +5F:!/product/32806

    The 4 tonners hit the low 30s @ +5F, eg:!/product/32557

    Even though electricity is expensive in your area it's not insane to pick a size that's even a hair under your fuel-use calculated load, and add some heat strip to cover the shortfall rather than try to mickey-mouse the boiler as a (ludicrously oversized) second stage.

    Be aware that with the Carrier Infinity/GreenSpeed series you have to program it to NOT engage the auxiliary heat strip for re-heat during defrost cycles, which takes a pretty big toll on efficiency.

  2. acdc2 | | #2

    Hi Dana - thanks for the detailed response. Really appreciate it. I'll explore the resources you referenced. Thanks for the tip on programming the Carrier system not to engage the auxiliary heat strip for re-heat during defrost cycles.

    The Manual J is from a company that offers Manual J, S, D training and services for HVAC technicians. I was referred to them by my HVAC contractor. They used Elite RHVAC software which is ACCA approved.

    The house has 2x4 walls and 2x10 ceiling joists with fiberglass batt insulation from 50 years ago. I do need to upgrade the insulation. I'll first focus on insulating the attic properly and then get to the walls.

    Regarding the current heat source, the EK1 boiler, I just verified that it is indeed set to 120,000 Btu/Hr using .242 burner orifice (see the table on page 12 at I could change the orifice to the smallest size to configure it for 80,000 Btu/Hr but no less. In addition to home heat, this boiler is also used for an indirect water heater (80-gallon glass-lined water tank).

    How do I include links in my postings? The URL's in my original posting were not marked as links whereas the ones in your reply are showing as links.

  3. acdc2 | | #3

    Hi Dana - the above information was very helpful. It seems that if I don't improve the current attic insulation, I do need the 4-ton unit. If I improve the insulation, a 3-ton might suffice. Naturally, I'll work on insulation first.

    A BPI certified HVAC contractor is telling me that, for the current level of insulation, even the 4-ton unit won't suffice. He wants to put in two independent 2.5-ton or 3-ton systems to cover the living area and bedrooms. This sure looks way oversized. But another thing he said got me thinking. He said he installs a lot of Carrier Infinity Greenspeed systems with continuously variable speed compressor and air handler. However, he doesn't like variable speed air handlers. He says that when the air handler is operating at lower speeds, you don't get enough air flow in the rooms farthest from the air handler. Does he have a point or is this old-school thinking?

  4. Jon_R | | #4

    > I had a professional run Manual J and Manual D calculations for me

    I collect furnace burner on time data and if the season is right, a week of hourly data is more accurate than a Manual J or monthly fuel use data. The early morning hours are best (less internal load and solar effects). Such data logging is easy, low cost and underused as a sizing tool.

    > you don't get enough air flow in the rooms farthest from the air handler

    BTU balance does shift with airflow rate. How much depends, but it doesn't take much to cause discomfort.

  5. acdc2 | | #5

    Thanks, Jon. Great insight!

    Regarding the airflow balance shifting with airflow rate, would you recommend going with a fixed speed air handler coupled with a variable speed compressor? Or are you implying that you need to pay more attention to the design of the duct work?

    1. Jon_R | | #6

      For lower room-to-room variations, I'd use zoning. But done such that zones are slightly oversupplied (10%?) when open and then a partially closed zone damper creates a small decrease (20%?).

      Don't increase CFM/ton with AC.

  6. walta100 | | #7

    For me it is a question of dollars and what costs the least to operate after that I think about comfort.

    Part of the dollar calculation is will you be able to get completely off gas and save the 30$ or so in monthly fees it cost to be connected?

    Then there is some math to do about at what temperature it costs less to use gas if you decide to keep both systems.

    If you do decide to keep both systems I would be tempted to size the HP for the cooling load 2.5 ton and have the gas system wired as emergency heat that will be locked out above xx°. I am guessing the HP would do 90% of the heating.

    From a comfort point of view HPs need to have registers placed very carefully. If they are place where they blow on people they will hate the HP and quickly learn how to disable it and run the gas heat.

    I would not be surprised to find the current ducts are sized for the cooling loads and are way too small for a 4 ton unit to operate quietly.

    From a comfort point of view adding a boiler driven heating coil to the air handler that would run when the HP is in defrost mode for about 3 minutes every hour or so would be a great luxury.


  7. acdc2 | | #8

    Walter - thanks for your insights.

    All - I have the following proposal from a Building Performance Institute certified HVAC contractor. Instead of a single Carrier Infinity Greenspeed 25VNA448A 4-ton heat pump, the contractor is recommending two Carrier Enterprise 2-ton heat pumps with matching fan coils. Here are the model numbers.

    - 38MAQB24R—3 heat pumps
    - 40MBAAQ24XA3 fan coils

    There would be two independent zones with a heat pump and a fan coil for each zone - one zone serving the living-dining area, kitchen, and sunroom and another serving the bedrooms.

    The rationale for this approach is that despite the rating for both systems (the Infinity Greenspeed 4-ton single-zone system and the Carrier Enterprise two-zone system) is the same 4-ton, the model#'s above (Carrier Enterprise models) have a far greater heating capacity at 5°F than the Infinity Greenspeed model. Here are the NEEP ASHP specs for both systems.

    - Infinity Greenspeed 4-ton system -!/product/32814
    - Carrier Enterprise 2-ton system (double the heating and cooling BTU's for two systems) -!/product/33600

    The max heating Btu/hr for the 1st system at is 31,000 Btu/hr and the 2nd system is 54,000 BTU/hr. (2 x 27K). The Carrier Enterprise heat pump is designed for more heating at lower temps.

    I ran the heat load calculations using the actual natural gas consumption and 99% outdoor design temp and heating degree days for my area as detailed in this article reference above by Dana - Using data from the last two winters, the average heating load for my house is about 40,000 Btu/hr. The Greenspeed system has a max heat output of 31,000 Btu/hr at 5°F. According to my contractor, this is insufficient not only on average heat load basis but even more so for instantaneous heat load on a particularly cold day. That's why he is recommending going with the 2-zone/2-system solution in which the heat pumps are designed for high heat at low temps. Does this make sense?

    A few concerns I have about this 2-system design are:

    1. The fan coil in this design is not continuously variable; it's a 3-speed design - 588/765/882 rpm. I understand that the heat pump has continuously variable speed. How would the comfort of this system compare with one that has a continuously variable fan coil/air handler?

    2. I imagine that parts availability for this system might be a problem given that this system is not commonly installed in homes. I also believe that it might be harder to find techs that know this system well as opposed to Carrier residential system. Are these valid concerns?

  8. acdc2 | | #9

    Apologies in advance for the long post. I need help from the experts.

    My original post includes the summary page from the Manual J report for my house. It shows 47,064 Btuh of heat load. Using the method described at, I arrived at a design heat load of 34,000 Btuh. This is based on the actual natural gas consumption during the last two heating seasons, the boiler spec input and output Btuh, the 99% outdoor design temp for my area of 12°F from , and 65°F heating degree-day spreadsheets covering those dates for a nearby weather station from

    The original heat pump selection I mentioned in the original post, Carrier Infinity Greenspeed 25VNA448A, is rated for a maximum of 31,130 Btuh at 5°F outdoor dry bulb and 42,500 Btuh at 17°F outdoor dry bulb temperature.

    ConEdison, my local utility, is offering a rebate of $2,000 for every 10,000 Btuh of heat at 5°F outdoor dry bulb provided the heat pump can handle 90%-120% of the building’s heating load (BHL). See the last cell on the second row in table 1 at The document doesn’t specify how to calculate the BHL. However, the ConEd rep that my HVAC contractor is working with insists that the BHL be calculated using Manual J (per Manual J report attached to my original post, the BHL is 47,064 Btuh and using the method described at, the heat load is 34,000 Btuh). How do I convince them to use the latter method? If I’m unsuccessful in convincing ConEd to use the 34,000 Btuh figure, I would have to install the more expensive (to install and operate) 2-heat-pump system that I detailed on the preceding post. This 2-heat-pump system has the same overall rating of 4-Ton, but it consumes more than twice the amount of KW at moderate cooling and heating temps. Links to NEEP ccASHP specs are on the preceding post.

    Why would my utility force me to oversize the system (while both are 4-Ton systems, the 2-heat-pump system has a lot more Btuh heating at colder temps; it is also lot more expensive to install and to operate) in order to qualify for its rebate? How do I convince them that the 34,000 Btuh figure is the right BHL?

    1. Expert Member
      Dana Dorsett | | #11

      >"Using data from the last two winters, the average heating load for my house is about 40,000 Btu/hr."
      >"Using the method described at, I arrived at a design heat load of 34,000 Btuh."

      There seems to be a 6KBTU/hr conflict in those two numbers? Where did the 40K come from?

      >"The original heat pump selection I mentioned in the original post, Carrier Infinity Greenspeed 25VNA448A, is rated for a maximum of 31,130 Btuh at 5°F outdoor dry bulb and 42,500 Btuh at 17°F outdoor dry bulb temperature."
      >" How do I convince them that the 34,000 Btuh figure is the right BHL?"

      You need to do more analysis on the detailed version of the Manual-J to find the thumbs on the scale. A Manual-J calculated load that it 14KTBU higher than measured reality, a 47K/34K= 1.38x overstatement must have some easily indentifiable errors in the full Manual-J report. When the Manual-J is done correctly it usually doesn't overshoot by more than 15% from the fuel-use derived number, which would put it at 39-40K(calculated).

      The odds are pretty good (nearly 100%) that the Manual-J's infiltration numbers are inflated. Unless they did a blower door test they probably went with the tool's default for tight/leaky/average based solely on a guess as to how tight the house really is, which is very difficult to tell with any accuracy in a walk-through. If you have the full report (in addition to the 1 page summarized results) what do the infiltration and ventilation heat load numbers look like? I would guess that could be a large fraction of the error.

      The output is slightly different depending on which air handler is selected for the 4 ton GreenSpeed, but without pulling the extended temperature charts I'll assume the 4 tonner is still delivering about 40K @ +12F, which isn't an insane oversizing factor for a 34K load.

      Short of convincing them that the 4 ton GreenSpeed really covers it, you might be able to convince them that a pair of 1.5 ton Midea /Carrier units (rather than the 2 tonners) would work:!/product/33267

      At +5F they put out 19.5K each (x2 = 39,000 BTU/.hr), at +17F they deliver 23.7K each (x2= 47,400), so at +12F they're probably hitting around 90% of the calculated 47,064 BTU/hr (x 0.9= 42,358 BTU/hr, or 21,179 BTU/hr each.) I'm betting the 1.5 tonners would still put out more than 21.2K @ +12F.

      Whether a pair of 1.5 tonners is going to cost significantly less is still TBD.

      >"If I’m unsuccessful in convincing ConEd to use the 34,000 Btuh figure, I would have to install the more expensive (to install and operate) 2-heat-pump system that I detailed on the preceding post."

      More expensive to install, maybe but even with the 2-tonners the Midea/Carrier units probably have a slightly higher mid-range COP at your average winter temp (~40F) , and the 1.5 tonners would be even better than the 4 ton GreenSpeed at any heating season temperature. The 2 ton Mideas have a slightly higher COP than the GreenSpeed in the mid-to-high range at +17F, but the 1.5 tonners are about 10% more efficient at that cool temp, and 24% more efficient at +47F.

  9. user-2310254 | | #10


    Have you tried to talk with a technical expert or customer service supervisor at ConEdison? You probably need to talk with someone who isn't working off a call-script.

  10. acdc2 | | #12


    Thanks for the attention to detail. My first calculation of 40,000 Btuh heat load had an error. I used pencil and paper for this calculation. I’ve now entered by natural gas billing data, the HDD data, and the calculations into a spreadsheet and I have the following results.

    Using a 65°F balance point, the building heating load was
    • 34,255 Btuh for the last heating season (10/25/19 – 3/27/20) – I rounded this to 34,000 Btuh in the last post
    • 34,818 Btuh for the previous heating season (10/25/18 – 3/28/19)
    • 36,497 Btuh using only the billing month with the most natural gas consumption (12/27/18 – 1/28/19)

    Using a 60°F balance point, the building heating load was
    • 38,195 Btuh for the last heating season (10/25/19 – 3/27/20)
    • 37,974 Btuh for the previous heating season (10/25/18 – 3/28/19)
    • 38,964 Btuh using only the billing month with the most natural gas consumption (12/27/18 – 1/28/19)

    Which of these heating load numbers should I use to size the heat pump?

    I do have the full Manual J report. Infiltration numbers are as follows. This is for a house with 1,858 sq ft room area (I mentioned 1,870 sq ft in my original post when I was going by memory; 1,858 is the right figure), 8’ ceilings (and, therefore, 14,864 cu. ft. volume), and 1,660 sq. ft. wall area.
    • Infiltration Winter: 139CFM, Summer: 62CFM
    • Winter heat load due to infiltration: 9,162 Btuh

    Here's the breakdown of the total 47,064 Btuh heating load:
    • Subtotal for structure: 30,411 Btuh
    • Ductwork: 7,491 Btuh
    • Infiltration: 9,162 Btuh

    They didn't do a blower door test. They didn't even do a walk-through. The report was done remotely simply using the architectural drawings and property survey for my house and the window and door dimensions. The house is not drafty at all.

    The Manual J is showing building volume as 10x the floor area. Looks like they assumed 10’ ceiling height while I have 8’ ceiling height. They also have the wall area as 1,848 sq. ft. instead of the actual wall area of 1,660 sq. ft. Correcting these should bring down the heating load.

    Currently, I have an unconditioned attic with insulation on the attic floor; I plan to convert it into a conditioned attic with rigid insulation above roof deck when I redo the roof next year. I will also make the attic airtight when I redo the roof. While that would increase the building heating load due to additional volume of air that needs conditioning, it should get rid of the heating load due to ductwork losses, right? For now, the ductwork will be fully insulated. Is the heating load due to ductwork that is shown above reasonable given the current configuration of the attic?

    Your suggestion for two Carrier 1.5-Ton units is a great one. I will explore that. As always, thanks for your insightful and informative comments.

    Steve – what you are asking makes absolute sense except it's impossible to find someone at ConEd who is a technical expert and understands the above. NYSERDA – the NY State agency – developed the incentives program for heat pumps and handed it over to ConEd. I’ve called both NYSERDA and ConEd multiple times but I’m unable to find a technical expert. I know such people exist, but they are impossible to find.

    1. Jon_R | | #15

      FWIW, my measured balance point is 68F when the house is at 70F. Highly unlikely that a 60F balance point is the right choice.

      Then add some margin (like 15% but it varies) to a fuel-use load to account for higher than average wind load and solar gains. Do that and your corrected Man J vs fuel-use numbers are close.

  11. walta100 | | #13

    I am not a fan of third party zoning kits things get way too complicated. If your house has living spaces on different levels I think you are better off with 2 separate system yes it will cost more than a zoned kit will but the way I see it you are buying redundancy in that sooner or later the something will break and with 2 systems half the house will still be warm.

    My guess for the ConEd rebate your contractor submits a rebate form with manual J number he provides and equipment model numbers. A clerk at ConEd enters the data in a computer and your application will be approved or denied by the computer program based on the data provides by your contractor.

    I would not be surprised to find the contractors have electric backup heat coils hidden in the bids.


  12. BirchwoodBill | | #14

    Did you look at correcting the hydronics system and using an air to water heat pump with some hydronic air handlers? Space Pak or chiltrix comes to mind.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |