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

Choosing a heating system for a new build

krom | Posted in Mechanicals on

Hey all. I’ve been plugging away at designing my new house, and now that I’ve been running running Beopt, and a Manual J spreadsheet for a while, its time to decide how to heat/cool the place.

Build will be in upstate NY zone 6, zip code 13439 design temp -6F electricity $0.13 kwh

House will be 28×40 with 192 ft^2 of glass facing east (bottom of attached floorplan) basement will be walkout on east end, Roof is 12/12

Originally I had planned on doing radiant floor heating with a combi of some sort, then an acquaintance of mine tried to sell me on a GSHP (researching them is how I found this site almost a year ago)

I am still thinking about radiant because I would like to finish the basement sometime in the future, and it would seem to be the most comfortable way to heat a basement to prevent cold feet.

I don’t think that a GSHP makes financial sense, but FWIW I was quoted $25k for a 3 ton system, before any rebates (quote was before any heat load numbers where ran)

I looked at daikin altherma (haven’t seen any pricing though) and even ran there simulation software (nice package) Unfortunately for my location the biggest heat load they can support is 20 kbtu (even with full aux heat)

I recently found thermo matrix out of Canada, but they don’t list heating capacity for any lower than 5F (their 3 ton has out put of 21.6kbtu at 5F, and 32.4kbtu for the 5 ton)

I am going to have a small woodstove because I like sitting next to a fire from time to time, but I don’t want to be stuck relying on it to keep the house above 60.

My manual J spreadsheet, and beopt both pretty well agree on my heat load being roughly 25kbtu with anderson A-series, or 21kbtu with intus upvc (I’ve read on how the european numbers don’t directly compare to nfrc, but didn’t know how to work around this so I used u=0.128 for the intus and .290 for anderson)

I plan on building this as tight as possible, and will blower door test a couple times throughout the build, and plan on using an HRV, (mentioned because ventilation and infiltration is calculated to be almost 1/2 of the heat loss)

I am looking for help deciding, and suggestions if I have missed anything.
If more info is needed please let me know.

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. iLikeDirt | | #1

    Just to throw this out there, you can get windows with much better specs than u-0.29 without breaking the bank for the pricey European triple panes. I just replaced all the windows in my house with nice double-panes that clock in at u-0.26. In NY, there are a lot of manufacturers of really good vinyl windows like Okna, who can sell you u-0.25 doubles and u-0.16 triples (not associated with the company).

  2. GBA Editor
    Martin Holladay | | #2

    Here's the short version: it's really hard to beat ductless minisplits or ducted minisplits.

    Here's the basic principle: it's usually better to spend more money on a good thermal envelope and choose a simple heating system than it is to spend a lot of money on an expensive heating system and end up short-changing your thermal envelope.

    Here's a link to an article that explains why you probably don't need a radiant floor: All About Radiant Floors.

    Here's a link to an article that explains why you probably don't need a ground-source heat pump: Are Affordable Ground-Source Heat Pumps On the Horizon?

    Here are links to two articles that give advice on choosing heating equipment:

    Heating a Tight, Well-Insulated House

    Heating Options for a Small Home

  3. krom | | #3

    I'm all ears for suggestions on window companies, especially if it means I can keep some money local, instead of shipping it overseas.

    I've been looking at mini's and assuming that I'd need one on each end of the house and a third for the basement (does this seem like overkill? and will the upstairs room stay warm without adding another head for it?) I attached the room by room heat load

    Will they be able to keep up when it drops below 0 (the daikin software showing that the alltherma couldn't keep up made me a little nervous)?

    I'm pretty surprised by the heat loss to ventilation, is there any way to reduce it?

  4. user-2890856 | | #4

    Trevor ,
    Nice , well thought out house . If you want radiant heat , don't be deterred by the high cost hype , . Many believe radiant MUST be expensive because the people whom have designed many systems just did not possess the proper knowledge of how these systems should be designed and installed . Have you ever heard of radiant ceilings ? Check them out . If you would like you can e mail me at [email protected] to discuss , it's free . Feel free to come back here and discuss or ask about anything I tell you . Not all houses are candidate4s for low cost radiant , however all low load houses usually are , just need the right people looking at them as opposed to the guys who designed early , crazy expensive systems . By the way , radiant floors done right will not overheat a room . Wives tale , no way a floor with 80* water running through it can add heat to a room with an above 80* slab due to SHG , in fact it would tend to keep the room from overheating , hot goes to cold you know , no way around it .

  5. user-2890856 | | #5

    Martin ,

    Although some may not need a radiant floor , ceiling , wall . Many people want these systems for their comfort , health , and efficiency . I still wish you would send me a PGH or Passive plan so I can prove my point and have a constructive discussion with you about this . First cost is not everything and is often misinterpreted . Please look up the Infosys report and cost analysis of their building in Hyderabad India . Pretty illuminating when you do ALL THE MATH

  6. GBA Editor
    Martin Holladay | | #6

    Q. "I'm pretty surprised by the heat loss to ventilation."

    A. I assume you are talking about the result of your heat loss calculation. There is no way for us to comment on the ventilation heat loss unless we know more about your input numbers. What ventilation rate (in cfm) did you input? Did you assume that your system will include heat recovery?

  7. fitchplate | | #7

    Plug the current retail prices into this calculator to figure out the price by fuel type.

    National Grid is not 13/kwh, its 30 cents. Divide your kwh into your total bill.

    Unless you have a net return on your PV system (to run your GSHP or mini splits), wood in your area will be the lowest cost for heating: $65/face cord split and delivered.

    In zone 6 air to air heat pumps are not a effective at those low design temps.

  8. STEPHEN SHEEHY | | #8

    Trevor: in 2005 we had a substantial addition put our existing house. We went with radiant heat. The problem is that between substantial solar gain through south facing windows and use of a small wood stove, the radiant heat is typically not running during the day. It ends up coming on while we're sleeping and by the time the house warms up, we're getting the solar gain. We're in the middle of building a new house and will heat with two mini-splits.

    I would not use radiant, especially in a tight house. It was expensive to install and not cheap to operate. We're in Maine, zone 6, design temp 0°F.

    We're using Intus UPVC windows. A big selling point was the availability of a local dealer, Performance Building Supply in Portland. They're sending someone out to help the GC with the first window installation.

  9. user-2890856 | | #9

    Stephen , Imagine one having the vision to put those rooms with solar heat gain on a constant circulation strategy using a circ that would use 9 watts . That circulator could harvest heat from the floor and mix it down to heat other rooms and areas and even preheat water before it reaches the water heater .
    Why did the radiant take so long to respond to the call for heat after SHG was not available ? I would say , knowing what I know about moving BTUs that you were the victim of the kind of designer I spoke about in my prior post . If you had excess solar heat gain in the floor it should be able to respond rapidly on a call . The floor should have been easily brought to a temp sufficient to heqat the room since it had stored BTUs in it . Maybe the wrong thermostat was used , such as a run of the mill thermostat that must drop a good 3 -4 degrees before closing the relay and beginning to charge the slab . Another fact is that mass in a floor application is a really poor application for radiant because of the mass . Radiant in a tight house as in any house also has to be zoned properly and this is a science to insure even temps in all areas . Since the home is so tight I would expect sub 90*F supply water temps .
    You should not allow your bad or not as expected experience to influence others who have a real chance of having properly designed systems of any sort installed in their homes . Unfortunately , as in any industry , about 85 - 95% of the guys who say they are experienced are not experienced enough to offer what the minority offer .

  10. krom | | #10

    The ventilation numbers all came from the manual J spreadsheet (off of acca website) it is from table 8A in the manual.

    Flitch mentioned that an airsource heat pump probably wouldn't be able to work for me, is this true for mini-splits as well?

  11. GBA Editor
    Martin Holladay | | #11

    Flitch is wrong. If you choose a Mitsubishi or Fujitsui minisplit that is designed for cold climates, it can certainly work down to -13 degrees F or colder.

    I don't know what the default ventilation settings are for your software. The software could be assuming a high ventilation rate, and it could be assuming that your equipment won't be capable of heat recovery.

  12. user-2890856 | | #12

    What type of construction are you planning for with this home ? I would venture that Martin is right about the ventilation rates and such .

  13. fitchplate | | #13

    Trevor … you’re getting lots of kudos but frankly I don’t get your design. Why put so much glass on the east end?

    Why use concrete at all? Why not use a well-insulated preserved wood foundation? Then you don’t have to manage and compensate for the undesirable effects, characteristics and discomfort of the concrete mass. And in a super-insulated, very airtight house, the concrete mass will be uncomfortable (in-floor radiant or not).

    I say have two heating systems for seasonal and zone control; and give yourself flexibility by building in multi fuel options (wood, gas, electricity) to allow you to optimize as the market plays with prices.

    Stephen … don’t rule out “radiant” just because radiant in concrete floors is not so good. You kind of dismissed all forms of radiant in one fell swoop. For one thing, a wood stove is “radiant”, and so are all forms of baseboards and wall hung radiators.

    The problem with radiant in-floor is a problem with concrete, not radiant.

    The issue you raise is important to the decision. How fast does one need the house to warm up (the system to respond) and what is the heat hold and release characteristic of the mass. Staple up and transfer plates in a wood framed floor, or baseboards, and or wall hung radiators are all responsive and controllable in a very tight and well insulated houses.

    The problem you raise is due to the high mass radiator (i.e concrete), which is virtually impossible to control. There are no thermal breaks in concrete so zoning is ineffective. And the rate of heating and cooling is glacial; inappropriate for a high performing envelopes and fully conditioned spaces.

    I have a multi-zoned layout, using European wall hung radiators in my very tight, very well insulated house. We are delivering about <90f of hot water supply temp, and get an instant heating response when we crank it up, with outdoor reset. can drop the temp quickly in radiators but .75ach50, r-60 cathedral ceilings, r-45 walls r30 floors, building holds residual heat for hours at design temp. a wood stove provides atmosphere boost; boiler lower threshold winter serves as back up to spring fall.

    I think air to air heat pumps/mini-splits are fine if you want forced air heating; but the human body likes radiant heat. And warm forced air likes to stay high in the building, increasing stack effect and convection in winter when you don’t want these effects.

    Radiant heat in a concrete basement floor is probably out of date now, with such well insulated and air sealed envelopes. We don’t need the mass for heat retention, response time is terrible, and heating-cooling swings put the concrete radiator out of sync with the house and outdoor temp changes. If you use a wood stove or heat pumps or low mass radiators (wood floor, wall, baseboard), you will probably find your in-floor concrete radiator will rarely turn on by itself and you will be disinclined to turn it on yourself.

  14. STEPHEN SHEEHY | | #14

    Response to Flitch:
    My present floor isn't concrete. It's slate over plywood, with aluminium plates under the floor. I know now that the design was probably faulty. Nevertheless, I plan to use minisplits in the new house, which does have a concrete floor. In floor radiant heat, which is what most people think of when someone mentions radiant, has no appeal. I may install a few radiant cove heaters as supplements to the minisplits for the few winter days when it gets way below zero.

  15. Expert Member
    Dana Dorsett | | #15

    Taking off a bit on Flitch Plate's last comments...

    Too much east facing glass becomes a summertime cooling problem, since the solar gains can't be cut with overhangs on either the east or west sides. Some AM gain can be useful for wintertime comfort, but ~200 square feet is a problem- you can cook yourself out of those rooms even in winter!

    Mini-splits pull air in from the top of the indoor unit, and distribute it laterally & downward. Stratification is not really an issue with mini-split solutions unless you have a ridiculous 22' tall "great room" to deal with.

    Comfort is more about average radiant temperature than air temperature, true, but better windows are as big a factor in average radiant temp than radiant floors or panel radiators. In cold climates even if there isn't always an energy-cost rationale for U0.20 or lower windows, there can be a comfort rationale. (Both Harvey and Paradigm make some reasonably cost-effective triple panes in the U0.18-U0.22 range, but they'll still be more expensive than anybody's U0.25 double-pane.) Radiant floors & ceilings are great when you have crummier windows or low-R walls with large convective loops and one fairly cold radiant-temp wall, but in a higher-R house with way better windows the additional boost in comfort is pretty marginal, and has to be factored against the up front cost.

    The "radiant" fraction of fin-tube baseboard is pitiful even with high-temp hydronic systems and like taller finned convector units, really shouldn't be considered "radiation", since they barely budge the needle on the average radiant temperature of the room. They might more correctly be called UN-forced hot air heating, since nearly all of the heat transfer is to the air via convection.

    But low-temp panel radiators can be pretty nice, with a much larger fraction of the output being radiated (though they are still primarily convectors). It's a good compromise should you go with a hydronic solution rather than ductless heat pumps, and much cheaper than radiant floors & ceilings. Cast iron baseboard is also pretty good at raising the average radiant temperature of the room, but also pretty pricey compared to low temp panel radiators (unless you buy it at scrap yards.)

    The cold climate ductless heat pumps running full tilt at -10F outdoor temps still deliver a COP of nearly 2, and can be correctly sized for loads at design temps that low or lower. At the January mean temperature of 20-22F in Richmond Springs they will deliver a COP better than 3 at part load, with a seasonal average efficiency in the low 3s. Even at 30 cents/kwh that's cheaper heat than condensing propane in most markets.

    And at 30 cents/kwh it's cost effective to put up a ton of grid-tied solar. which at recent pricing has a lifecycle cost less than half that. SolarCity will be offering 30 year/4% financing in NY beginning next year, and Vivint (their largest competitor in the residential scale PV market) will be chasing that market soon, despite leasing being at the core of their business model to date:

  16. KeithH | | #16

    $ will vary a lot by size and type of window but I wanted to throw out a few numbers.

    I recently got some Pella 350 windows in triple pane. The 350 has is a thick framed vinyl window (be sure to get the foam insulation in the frame). I was able to get triple paned XOX slider assemblies for about $13/sf in my area. A similar assembly in all operable egress hardware casement was closer to $29/sf. The performance was pretty good by my terms.

    Straight low-e argon glass package with a U-factor of 0.18, which is excellent for a 'commodity' window.

    Or you can also get a south facing glass that has a U of 0.23, an SGHC of 0.46, and a VT of 0.56. Pella calls this glass package Natural Sun but it is probably Cardinal 180 glass.

    I'm certain there are better windows out there but if you are looking for more commodity pricing and vinyl is acceptable to you, you might look at the Pella 350 vinyl line. BTW, these windows do have bulkier frames so if you have narrow openings, you might not like them. I have larger openings with factory mulled assemblies so these work fine.

    As for how much improving your window U factors might improve your heating load, I don't know.

  17. krom | | #17

    Lots of info here.

    Thanks for the calculator. Electricity would come from NYSEG and I calculated .13.kwh using a years worth of electric bills from my mothers house. (5220 kw, for $678.40 works out to $0.1299/kwh)

    I looked at the spread sheet now that I am home, and it is assuming 92cfm for ventilation rate, which is supposed to be what is prescribed by manual J table 8A

    I'm planning a poured foundation with exterior foam, and foam under the slab. Walls will be and "arctic wall" or double stud wall, or larsen truss (all seem pretty close to the same to me) inside out: GWB, structural 2X wall, 1/2" cdx, taped and sealed, dense pack cellulose, membrane for WRB (tyved brand, or fancy european stuff) 3/4" air gap, siding (fibercement or vinyl something low maintenance) Total R40.
    Roof will be similar GWB, 3-1/2" service cavity, 1/2" cdx taped and sealed, structure (I joists, or parallel chord trusses) membrane for WRB, air gap, roof sheathing, shingles or standing seam. R60 (hoping to qualify for PGH :)

    The windows all face east for the lake view, there is a hill to the west, nothing but trees to the north, and the property line is pretty close to the south, so for now there is nothing but trees, but there could be neighbors in the distant future.
    Honestly the basement is concrete because thats what basements are mad out of around here, and I've never seen or heard of any one using PWF, I'm not comfortable using wood.

    The east facing glass is in a a ridiculous 22' tall "great room" (its actually 23' to the ridge lol, 8' ceiling height, 1' thick upstairs floor, 12/12 roof=14')
    Is there any chance a single head minisplit will keep that end of the house comfortable?

    I am still open to any one who has suggestions for a manufacture. I have quoted anderson from the local lumber yard, (roughly $10k) Intus upvc ($16k plus shipping) and zola upvc (17k plus shipping and technician visits) before coming on here, I had never heard of them or any other euro windo company
    With all the east facing glass I would like to use good windows for comfort reasons, but I have to determine if the 15% reduction in heat load helps make up some of the huge price difference (suggestions on figuring ROI for something like that?)

  18. fitchplate | | #18

    I stand corrected on the efficiency of the mini-splits. Its taken a few months to sink in.

    This video is enough to make grown man cry (if he heats with anything else): Mitsubishi Zuba

    Martin ... they are saying the mini-split can work down to -22 for ductless and -30 for ducted.

  19. Expert Member
    Dana Dorsett | | #19

    With 23' tall great room w/200 square feet of east facing glass you have a shot at staying comfortable you'd have to go with a soft-coat low-E heat-rejecting window glazing with a very low SHGC, the type usually sold in cooling dominated climates. A ceiling fan can easily break up the stratification issue even at low speed but the cooling loads can be pretty high with that much unshaded glass even in winter. You can probably get there with 1.5-2 tons of cooling, (what kind of peak numbers are you getting out of Manual-J & BeOpt for that room?).

    That could be served by a ductless system, but it's kinda silly to have to run it in air conditioning mode in the AM to keep from cooking on a bright winter or shoulder season morning, since only a few hours prior you had to be running it in heating mode, and will have to return to heating mode again in the afternoon. An AM "sauna effect" solar blast might feel OK in the middle of winter, but probably not so much come July. Unlike south facing solar gains, east facing glass gains are not manageable by roof overhangs.

  20. krom | | #20

    Honestly I didn't put a lot of though into the east facing gains, ( I didn't even turn on my air conditioning this summer, so it kinda slipped my mind) I did make sure that the south windows where shaded until September. I concentrated on the heating season since its seems so much longer around here.

    Any effective way to take the hot air from the peak of the ceiling and move it to the basement lol

    Just for giggles here some screen shots, for June at 11 am, first is with a 2' overhang, second is a 3' overhang.

  21. krom | | #21

    After more reading on here, I found this paper on ventilation, from BSC:

    The formula in it suggests 37 cfm. this drops the block heat load down to 17kbtu

    Any more thought on the possibility of a single minisplit working without removing a few windows?

  22. Expert Member
    Dana Dorsett | | #22

    With a 23' room height you'd need a ceiling fan to break up stratification no matter what you were heating/cooling with. Whether a ductless would be a realistic solution depends a lot on your peak cooling & heating loads.

    It's possible to move hot stratified air from the top to other places with ducts & small air handlers, but it needs to be a fairly big delta-T to be really efficient. It doesn't take a super-breeze high-speed ceiling fan to break up the stratification, but it adds at least some wind-chill factor for mammalian occupants. In summer that would probably be a good thing, in winter not so much.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |