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

Sizing HVAC System for ICF House

Payton | Posted in Mechanicals on

I have an ICF home under construction.  The typical HVAC quotes I have received are based on sqft from what I can tell. (24,000BTU)

First time poster, but I’ve been reading up on this tread for my construction project since I started planning this a couple years ago.

This home is super unique. I will upload a few drawings.

Every Manual J test I have seen does not take into account Thermal Mass (Most walls are 10″ concrete with 2.5″ foam on both sides, the East wall and around the stairs is 8″ concrete) or allow me to enter 4″ concrete for the roof.

The main portion of the living space is 820sqft downstairs and has 23′ ceilings and a 5″ slab floor. There is a loft over this floor about 580sqft including the stairs. So I am testing with just 1 floor but at 23′ ceilings. Is that ok ?

My ceilings are 4″ concrete roof deck. They are flat and will have spray foam insulation to the required depth.

There are minimal interiors walls. And there are none that prevent air flow from one side of the home to the other. I also have a HRV Lunos e2 kit.
and the vents are on the East and west wall. Plus 2 ceiling fans.

I have many windows. 450 sqft for the South wall alone. All windows are Low-E. But the East and west only have 66sqft combined and North wall 1 window at 30sqft. I have great shade on the West side of the home with tall 50’+ trees close by

I am in Zone 4A.

I am wanting to do mini-splits and I’m not sure just 1 in the living room will do the whole home or if I should get 2 smaller ones and place the 2nd one in the loft.

My Architect thinks the HVAC design I have is twice what I really need.

Does anyone have experience with an extremely tight envelope and large thermal masses of ICF homes ?

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. paul_wiedefeld | | #1

    Why would thermal mass change your heat loss? And would it change it a meaningful amount?

    1. Payton | | #5

      Properties of the exterior mass walls that affect the energy use of the house include the type and thickness of insulation, thermal mass, and air infiltration. Wood and metal frame walls are considered low-mass walls. Heat loss through a frame wall is dependent on the amount of insulation and air infiltration. More insulation typically means less heat loss and less energy required for heating and cooling. The benefits of using more insulation are well publicized by insulation manufacturers and consumers understand these benefits. However, thermal mass also has a significant effect on the heating and cooling energy. The concept of thermal mass is less publicized and is poorly understood by consumers. Walls with high thermal mass, such as concrete walls, have the ability to store and later release heat. This ability tends to moderate indoor air temperatures, and reduces energy associated with heating and cooling.

      This is from the DOE

      1. Expert Member
        NICK KEENAN | | #8

        Citation please.

        Let me ask this: what are the units of this "thermal mass" of which you speak? What number should be put into a calculation?

        1. Payton | | #14

          I'm just a novice homeowner. This is just what I have come across in my research.

      2. ROBERT OPALUCH | | #16

        It is true that high thermal capacity materials which transfer heat readily, like concrete, can dampen interior temperature fluctuations IF the interior air is exposed to the concrete. Heat is readily absorbed whenever the interior air temp exceeds the temp of the concrete surface. Heat is also radiated back into the room whenever the interior temp is lower than the concrete. So interior temperature fluctuations are muted by interior solar gains, for example. And heat losses from windows can affect interior temps less, since you are radiating heat from the exposed concrete as the interior cools. But you still have to protect that concrete thermal storage from exterior heat losses UNLESS the exterior temps average out to be your desired interior temperature, which is not the case in your climate in winter. And you still need to calculate the heat losses given your levels of ceiling, ICF wall, and any foundation insulation, and your window U-factor (R-value). The concrete mass itself will not do much to stop heat losses, since the R-value for concrete is low.

        ICFs have insulation on both the interior and exterior sides of the concrete structure. The plans have spray foam insulation on the interior side of the roof deck/ceiling. So the effects of stabilizing interior temps would be muted or extended. The concrete mass cannot so readily absorb heat from the interior, since interior insulation stops that heat flow somewhat.

        Summer temps for your area seem to be okay for a high mass buildings in your location IF shading is used to avoid solar heat gains in late summer and early Fall, and windows are opened to bring in cooler air overnight or early morning for example. Its good that you have minimal east and west facing windows, which if unshaded, have high solar heat gains in June/July. South and north facing windows have minimal solar heat gains during June/July. As pointed out by DCContrarian, August becomes more problematic, as the sun’s lower position in the sky begins to create solar heat gain in south-facing windows, and the effect becomes greater in the Fall.

        On average, your winter temps would lead to an interior temp of perhaps 40F except you have tremendous gains and likely glare from the large expanse of south-facing glass if left unshaded. Plus heating from internal heat gains and mini split auxiliary heating. You likely experience interior overheating on fully sunny winter afternoons, but of course you can open windows to solve that problem. If your concrete was more exposed to the interior, more of that solar gain heating would be stored and radiated back overnight.

        1. Payton | | #21

          Thanks so much for such a detailed explanation.

  2. this_page_left_blank | | #2

    Definitely unique, even bizarre I'd say. The third floor is just a bathroom serving the rooftop balcony?

    I'm not an expert, but I predict if you used one minisplit on the main floor, you'll have massive overheating problems on the second and third floors. Pretty sure you have to treat that 23' ceiling area as if it was two levels, i.e. double the square footage.

    Thermal mass inside external walls isn't going to be of much benefit. As an aside, this is probably the least green design of a new house I've seen on here. Concrete is about the worst of the commonly used building materials, followed by foam insulation.

    1. Payton | | #6

      Yes, the 3rd floor is just a 1/2 bath.
      It was my only building option to achieve a pool installed on the rooftop other than a steel construction I was told.

      1. Expert Member
        NICK KEENAN | | #19

        How far along in the construction are you?

        Why do you feel the concrete is a better option than steel?

        1. Payton | | #22

          In the dry waiting for windows.

          My home designer is in the ICF industry. It was less expensive for me.

  3. Expert Member
    NICK KEENAN | | #3

    The Manual J process does not have an input for "thermal mass" -- whatever that is -- because it only includes factors that influence heat gain or loss.

    In general, your description includes a lot of factors that don't influence performance. In a proper Manual J there are a lot of inputs, but the important ones are design temperature, insulation levels of the walls and roof, surface area of the walls and roof, and window and door construction and area.

    The 450 square feet of south-facing windows are going to lose 13,950 BTU/hr with an outside temperature of 10F, an inside temperature of 72F and a U-factor of 0.5. You can add another 3,000 or so BTU/hr for the 96 SF of other windows.

    1. Payton | | #7

      I appreciate your input. So sizing will be more important on heating ?

      1. Expert Member
        NICK KEENAN | | #9

        Well, that's the next question: is that 24K number you were quoted for heating or cooling? What are the design temperatures for heating and cooling?

        As others have noted, that south-facing glass could impose a huge cooling load in the summer. But that's a more complicated calculation.

        1. Expert Member
          NICK KEENAN | | #11

          Just to give an example, the article that Robert links to below, ( ) has solar heat gain per day per month for windows of various orientations for Providence, RI.

          In August a south-facing window gains 355 BTU/SF/day and an east or west-facing one gains 387. With 450 SF of south-facing and 96 SF of east-west that's 196K BTU/day. Assuming that's evenly distributed over an 8-hour day that's just under 25K/hour. You'd then have to knock it down by the solar gain coefficient of your glass, but it's pretty hefty.

          1. ROBERT OPALUCH | | #17

            Since that article focused on northern USA locations, here's some data closer to the location of the home described in this Q&A. Really the homeowner should go to or other resources to get a more precise location data for the actual location in TN.

        2. Payton | | #15

          The installer did not convey that information. He said I needed 24k ( 2 12k units, I'm assuming 1 downstairs and 1 2nd floor. )

          1% cooling is 91
          99% heating is 20
          HDD/CDD is 0.8

          But the weather station is at the airport, 30 miles away and lower elevation.

          I would say drop those temps by 5 degrees for my exact location.

  4. ROBERT OPALUCH | | #4

    This house looks like it was designed for a cold and dry climate (colder winters than CZ4), and/or a very nice view towards the south. Could you tell us something about your location, so we have a better idea of the climate and cloud cover? What is the R-value or the depth of that spray foam insulation for the ceiling? Has the architect or an engineer done any solar heat gain analysis you could share? Could you say something about the shading of the extensive south-facing glazing (e.g., trees, other buildings, interior shades)? With predominantly south-facing windows (unless they are shaded), you potentially have a HUGE amount of unshaded daylighting and solar heat gain mid-winter, but minimal in June-July. Do you have a spectacular view to the south? All that concrete mass will reduce interior temperature fluctuations when windows are kept closed and shaded appropriately.

    Agree with Trevor that all this concrete and spray foam isn't environmentally sustainable. Far more concrete mass than seems functional, unless you are in a very high wind area but desire capturing nice views to the south. Or sometimes architects just want to make a unique building to make a statement or illustrate some concept. Has your architect stated some principles or architectural vision for this design, that you could share with us?

    The following 40 year old house is not ICF, nor “extremely tight” envelope by today’s standards, but has predominantly south-facing glazing in a CZ5 dry climate. The passive solar design exploits an insulated concrete slab to stabilize (absorb and radiate) solar heat gain to almost eliminate auxiliary space heating for the main floor, and significantly lower wintertime space heating upstairs.

    When designing this house, I did a LOT of calculations of heat losses (coldest winter days, average winter days, warmest winter days), solar heat gains (mid-winter solar heat gains on fully sunny, average, and overcast days), for hourly, daily, and multi-day periods. The calculations did predict actual performance quite well (since its elementary physics). You or your architect could do or may have done some similar analyses during design. Could you ask the architect to share any similar analysis with us? Maybe write a brief article?

    This article provides some general information about solar heat gain for various US climates:

    This might not be the answer you are seeking, but if I were you, I’d install the mini split the architect suggests. You can add a second one later (or use spot electrical heating or AC) if you find it necessary. Assume that you are incorporating fans, a ducted blower air distribution system, or ducts (perhaps with your mini split or ERV) to help redistribute air from the warmer upper area to the lower cooler area of your three-level building. Otherwise a minisplit head located high in the building for AC, and another head located low for wintertime heating might work well with typical multi-story heat stratification patterns.

    Although solar heat gain will be absorbed by your concrete structure, you could have warmer interior temps (maybe 10F+ warmer) in sunny winter afternoons vs. your interior temps at dawn. In a dry climate or controlled interior humidity, that can be within the typical comfort zone of typical 68F winter interior thermostat settings and 78F summer interior AC thermostat settings.

    I'd be very interested in reading an article summarizing your home's performance and your experiences after your first year or so of occupancy.

    1. Payton | | #12

      This is in the mountains of E TN. The elevation is 1,700ft.

      The ceiling has not been sprayed yet. Suggestions ? I have between the steel joist for 16" thick.

      The view is amazing, there is no potential for shading the South windows as the home sits on a cliff with no obstructions to the view. home orientation and window design was intended to take advantage of view. All the South facing windows are double-pane fixed low-e.
      The design will have windows awnings above each the south windows. About 2' extended from home and will be metal lattice. I am planing on having mechanical shades for all the windows inside, perhaps Lutron.

      I will show a pic of the view.

      On design of the entire structure, we had to build something very strong. We have a Modpool ( container pool ) and have it installed on the roof deck.

      The architect is a friend of mine and really is a "Home Designer". He did not design the home with HVAC in mind. Unfortunately he is not much help in this area. He designed ICF buildings but mostly commercial.

      I will post some updates to help record my experience with the heating and cooling during the first couple years.

      1. ROBERT OPALUCH | | #18

        You've got a very nice view there!

        For your ceiling/roof insulation: A diagram drawing (even pencil and paper notes) would helpful to make more practical suggestions for ceiling insulation alternatives. There is minimal ceiling/roof detail shown in these plans. The plan was to use spray foam between the steel I-beam joists? But no spray foam on the steel joists themselves? And what is below the spray foam, drywall? Is there any wood structure planned at the ceiling plane?

        Steel transmits heat very readily. Either the I-beam joists need to be outside the “thermal conditioned (interior) envelope”, or outside. If they touch both, they would be a thermal bridge that would allow a great deal of heat loss/gain. The architect knows this and did something.

        So an affordable alternative might be to use loose fill cellulose or fiberglass blown above a wood frame ceiling that is hung below the bottom of the steel I-beams. If space is an issue, polyiso board insulation has an R-value comparable to closed cell spray foam, without many of the negatives that lead some of us to oppose the use of site applied closed cell spray foam. It could be used instead of loose fill cellulose or fiberglass to save space (by about half the height for similar R-value).

        1. Payton | | #23

          Here is a photo of under the roof deck.

          I am waiting to see from the county if I can just paint the entire ceiling black after spray foam and leave it as is. If not, then it will have T&G Pine over the entire thing and can consider loose fill cellulose above the T&G

          The steel beams only go up in the concrete roof about 2" The rest (14") is inside the building.

          1. Expert Member
            NICK KEENAN | | #25

            A roof like that is usually insulated from above with sheet foam.

            You need R-38 in the roof, which is about 6 inches of closed-cell foam. Steel has an R-value that's so low it's indistinguishable from zero, so it messes with calculations, but if you're insulating from below every part of that ceiling needs to have 6" of foam facing down. Another option would be to do 6" of polyiso foam board below the joists.

          2. ROBERT OPALUCH | | #26

            Using spray foam between the steel joists could be almost useless in stopping heat flow through the uninsulated steel joists to the outdoors. But I don’t see the framing details in the plans. The plan must be to encapsulate the joists as well as the space between those steel joists. That’s a lot of linear footage to spray, and practically speaking, would require a ceiling below those spray foam joists. Seems to make more sense to span a flat ceiling with polyiso insulation board; or create a ceiling less than a foot below the bottom of the steel joists, and insulate above that new ceiling with blown in cellulose or fiberglass (cheaper insulation, but requires about twice the depth of polyiso or spray foam). Approximate R-values per inch are spray foam R-7, polyiso R-6, blown-in insulation of various types R-3+.

            Spray foam is flammable, even when cured, so has to be covered in habitable spaces. Flame retardant paint might be acceptable to the county, but you probably should do more than that for your own safety and better appearance. Just painting would appear a bit strange, as spray foam produces a bubbly shaped exterior appearance.

            Drywall is a typical thermal barrier used to cover spray foam, and would require a wood frame ceiling below the steel joists and installed foam. See this short article for more info and his additional references:
            See this article by a highly regarded supplier that is anti-foam. It shows tests of fires set to a room with exposed cellulose; another with exposed fiberglass; and the third with exposed spray foam, which performs very poorly in comparison. The 4 minute video is worth watching.

            It appears that a wood frame ceiling below your roof's steel framing can have enough space for blown-in cellulose or fiberglass (about R-3.3 per inch). If you want a higher ceiling, polyiso board insulation is a good choice. Note that polyiso approaches spray foam in R-value per inch, and varies a bit by manufacturer, but about R-6 per inch. Site-applied closed cell spray foam is not just environmentally damaging, it is expensive, and produces noxious off-gassing that decreases over time (but causes continued problems in rare cases). Blown-in cellulose and fiberglass are very cheap by comparison, and easy to install.

          3. Expert Member
            NICK KEENAN | | #27

            In the winter the interior dew point will probably be about 50F. That steel will be at outdoor temperature. So if any interior air reaches it condensation will occur. This will cause problems as the roof has no way of drying to the exterior. There needs to be enough impermeable insulation to keep interior air from reaching the steel. My back-of-the-envelope calculation is that the minimum safe configuration is four inches of polyiso and four inches of "fluff" -- cellulose or fiberglass. Or alternately, seven inches of foam.

            I think what I would try is putting half inch plywood on the joists. That gives something that the foam can mount to where you don't have to worry about hitting a small piece with a long screw. Then attach the foam in layers using progressively longer screws. Tape and caulk the final layer. Then run 1x3's to hold it all together and give something for the finish ceiling to hang from.

          4. Payton | | #28

            Robert, that was very informative. I did not know the fire risks associated with spray foam.
            The walls in this home will be all T&G. So if I need to cover the ceiling to hold the insulation as well as create a thermal barrier would T&G be an appropriate material ? I do not have any plans to use gypsum board or the like.

            How far below the metal beams do I need the ceiling ? The insulation can not be between the beams and the ceiling be flush to the bottom of the beams like in a wood joist system ?

          5. Payton | | #29

            Nick, I appreciate your insight.

            With the tops of the steel beam embedded in the middle of the concrete deck. Would the warm air below not reach the concrete to keep it somewhat warmer than the outside air ? I'm just wondering if that would be the case. Or is that not really the issue. I never thought that I needed to keep the conditioned space below the steel beams...

          6. Expert Member
            AKOS TOTH | | #30

            That type of roof deck is challenging to insulate.

            Those steel rafters are a huge thermal bridge. Say you spray 14" (~R90) of spray foam between them, that actual assembly R value because of the steel is around R15. The R value of the SPF is essentially wasted.

            You must have spray foam cover the steel completely.

            In your case, you want to cross strap the ceiling with 2x3s on edge perpendicular to rafters. Install enough spray foam for condensation control for your climate (2" to 4" depending on climate zone) against the roof deck and fully cover the rafters with at least 1.5" of SPF. Fill the rest of the cavity space with batts to bring it up to code.

            Getting this detail right is important not just for energy efficiency but for moisture. Any part of those steel rafters exposed to the hot moist air will likely start to sweat in the winter time.

          7. ROBERT OPALUCH | | #31

            I would assume that you want the finished ceiling to be as high as possible. However, you need to keep the ceiling low enough that the insulation is BELOW the steel beams, if those beams span to an uninsulated or under insulated area of the exterior walls (likely). If the steel structure at any point reaches near the outdoors or otherwise has very little insulation between it and the outdoors, it will become a thermal bridge. Heat moves very readily through steel, so it would become an easy way for heat to escape from the house in winter. I would assume that the steel beams terminate at or near the outside edge of the concrete in the ICFs. So minimal insulation exterior to the edges of the steel beams. However, the plans don't go into that detail, so we don't know how far the beams and steel above the beams extend to the exterior, above or into the ICF wall. (Is this steel "subfloor" above steel beams supporting the concrete roof during curing? Is there some insulation above that steel and below the concrete roof? No plan details to examine?)

            If the spray foam, polyiso, or cellulose insulation you add were only BETWEEN the steel beams, the thermal bridge would go from your finish ceiling (T&G board maybe 3/4" thick or R-1) then through the steel beams to the exterior through various paths, likely a very low R-value total, resulting in high heat losses during winter. Unless there is some ceiling insulation we can't see, the steel beams need to be isolated from the interior air space by R-38+ insulation, not just by a T&G ceiling alone.

            It would not be difficult to support a wood frame ceiling at the perimeter and from the beams above. Nick/DCContrarian suggested one way with plywood, then polyiso.

            There are several choices of insulation you could use. Seems the main choices would be polyiso (higher R-value like spray foam, but not the cheapest choice) and blown-in or batt insulation (e.g., blown in cellulose or fiberglass, or fiberglass attic batts). These are relatively inexpensive but lower R-value per inch, requiring more space (and a lower T&G ceiling with wood framing).

            Personally I would go higher than your minimum required (R-38?) ceiling/roof insulation. This has an impact on your decision for sizing mini splits, as well as comfort. A cold roof, windows or other thermal bridges can lead to convective drafts, even if your home is well sealed. Convective air movements are likely to occur with higher ceilings, as warm air floats higher and cool air falls down toward the floor at the bottom. There will be cooled air floating down that was cooled from proximity to windows on the south side. That cooled air flow would move across the floor, and back up along other walls to the ceiling, creating some heat stratification. A cold ceiling would create the same type of air movement draft (as well as adding to your supplementary heating load).

            (reply to Payton below...there's no reply button there!)
            These new illustrations and photo indicate that there is no insulation above the beams, and no insulation is planned above the roof slab.

            The steel beams and the lock bars, covered with the concrete roof, will allow a great deal of heat transmission to the exterior, unless there is insulation added below them (since none is planned above, and none was placed between the steel framing and the concrete roof). The beams and lock bars are steel and readily transmit heat through each other and into the concrete roof slab. So insulation needs to be installed to isolate the steel beams and lock bars from the interior air space to reduce heat losses through the roof. There are a number of ways to do that.
            1. Spray foam would best be comparable thickness across the entire ceiling, including steel beams/lockbars/any openings between these steel members. If you leave the steel beams uninsulated or under-insulated, a large proportion of the heat loss will occur at those places. Since spray foam is expensive to install, carries some risks of outgassing if not installed well, is environmentally damaging, most of us would recommend you avoid it.
            2. You can install polyiso board across the ceiling. The easiest and cheapest is a flat polyiso ceiling below the steel beams. You could install it the way Nick/DCContrarian suggested, or similar ways. Just like spray foam, it should be covered by drywall or other flame/fire protection. You could add T&G boards after drywall. If the boards are only 3/4" thick: Typically 2x lumber (1.5" thick), 3/4" plywood or 1/2" drywall continuously across the ceiling would qualify as firestop. Ultimately, everything depends upon the opinion of your building inspection dept as well as what you want to do with your home.
            3. You can install a wood frame ceiling a little lower to install blown-in cellulose or fiberglass, which is less expensive but requires more space. Cellulose insulation includes additives for insect and fire protection, and as you saw in the video, does much better if a fire occurs.

            Condensation could take place above these proposed ceilings. There will continue to be water vapor given off by the concrete ceiling, or migrating up through the proposed ceilings below the steel beams. It could condense on the steel. I haven't done concrete and steel buildings, or done concrete construction other than foundation walls, slabs and stairs. Others should address this issue, like your builder or others here.

            I assume you plan to work with your builder on resolving these issues, as he may have some other opinions, experience, analysis, or design solutions for the problems we've discussed. And some of these suggestions you could do later, if your house doesn't perform as well as planned. Like adding a second mini split, or adding an insulated interior ceiling at a later date. Best of luck with your new home and enjoy those views! We hope to hear back from you at a later date on your experiences, so we all learn something from each other. That's what the online GBA community is all about.

          8. Payton | | #32

            Thank you Akos,

            I was researching some more and did find a few examples where metal beams were completed covered in SPF.

            If I was to attach the 2x3's to the beams, would that inhibit the SPF from coverings the steel at the connection point ? Or would the wood provide protection where the foam would not cover ?

          9. Payton | | #33

            So I need more than to just spray the steel beams and then insulate the space between.
            In essence , the best option is to do a drop ceiling ( T&G ) and insulate above that to the steel beams, not going above the beams because that is ineffective ? or insulate the entire cavatiy ?

            You have seen the photos underneath the roof deck, here is some cross sections from the manufacture ( SpeedFloor USA ) and here is a pic prior to concrete pour.

  5. Expert Member
    Michael Maines | | #10

    Thermal mass can be somewhat effective when there are large diurnal temperature swings--i.e., when days are hot and nights are cool, and the Passive House energy modeling software does take it into account. But as others have said, the effect is small to negligible, and easily outweighed by insulation and air-sealing.

    Depending on your glass specs--SHGC and U-factor are the useful numbers--I wouldn't be surprised if your system is properly sized. Have you reviewed the Manual J report to see what they are listing for various values? The math isn't that complicated, if the inputs are correct.

    Also, as others have said, it's a travesty to build with the materials you have chosen, which are the worst contributors to climate change, when there are many benign materials available to choose from. But I don't expect to change your mind on ICFs--once people are sold on their "value" they seem to stop thinking logically.

  6. Payton | | #13

    I understand. Thank you for your insight.

  7. Expert Member
    NICK KEENAN | | #20

    The reason I ask up-thread how far along you are is that houses really should be designed as a system, the HVAC should be designed at the same time as the structure. This is particularly true if the house has a non-conventional construction. It's much harder to retrofit comfort and energy efficiency than to build it in from the onset. But if it's already built it is what it is.

    As Robert points out above, steel and concrete are highly conductive, if they're used they need to be entirely inside or outside the building envelope, they can't span it or they defeat the point of insulation. Usually that means putting them on the outside of the insulation layer.

    I recommend doing your own Manual J calculations. That allows you to see where the big energy flow items are and spend your time addressing them. You're going to run into a few issues:
    * With your construction, it's going to be tricky to make sure that you're actually getting the R-values you think you're getting with insulation. You really have to worry about thermal bridging.
    * Manual J is a worst-case calculation, making sure you have enough capacity for the hottest and coldest days of the year. What it doesn't help you with is average energy use. It may well be that over the course of the year those south-facing windows contribute more in heating than they lose from solar heating in the cooling season and energy leakage year-round. Manual J isn't going to tell you that, on the coldest night of the year and hottest day of the year those windows are no help and Manual J analysis is going to push you toward as few windows as possible.

    It sounds like several design choices were made in the thought that they would be energy efficient, but it turns out not to be the case.

    1. Payton | | #24

      Thanks Nick, yes, I wish the design had more involved the HVAC system but it didn't. This home was just designed to entertain and have an awesome view of the mountains. I was told that having the entire structure out of ICF, that I would experience great efficiencies in heating and cooling. But that doesn't look to be the case now. I will just need to install how ever much heating and cooling power is required to maintain a comfortable level in the home year-round.

      I will , however, report back , with what system we went with and how the system is operating for future knowledge bank.

  8. Expert Member
    NICK KEENAN | | #34

    I'm responding to post #33, but we've got too many levels of nesting going.

    You need to imagine the steel like it's a hole in the house. Steel is highly conductive, so much that its insulating value is indistinguishable from zero. Heat flows like water, except you can't see it. Would you have a roof with holes in it? That's basically what you're getting with uninsulated steel.

    Concrete has an R-value of 0.3 per inch, which while low is substantially more than steel. So if your roof is 4 inches that's R-1.2. That's about 3% of the R-38 you need, so if it's 20F outside the underside of your concrete roof is going to be at about 21.5F. So no, the concrete doesn't really do anything to keep the steel warm.

    What you're calling "beams" are called "joists" in the drawing, that's what they really are. You need R-38 of continuous insulation between the joists and the interior living space, both horizontally and vertically. To prevent condensation about 60% of that insulation needs to be closed cell foam or other insulation that is impermeable to air. For R-38 you're looking at six inches minimum of insulation, using closed cell foam. It could either go as a dropped ceiling below the joists, or you could box the joists with a box that goes six inches below and six inches to each side. Boxing is a lot more work but it may fit your esthetic better.

    1. ROBERT OPALUCH | | #35

      There's another factor here that needs to be taken into account. The theory of a "high (thermal) mass" building is that it can store and distribute a lot of heat capacity over time. So its theoretically okay if it loses a lot of heat, because it regains it later, and the high mass building averages it out, by the time the heat fluctuations reach the interior. The interior will see minor changes in temperature due to the dampening effect of that large "thermal" mass. I know you don't like the term "thermal mass", and might prefer some other terminology. But there's something at work here that distinguishes steel and concrete. Both have high weight mass and low R-value, but a building made of steel wouldn't operate thermally like a building made of concrete. High thermal mass (or whatever you may call it) will dampen and average out temperatures to provide a more stable indoor temperature. Attached is a table that notes the available thermal heat storage capacity of various materials, from:
      This complicates the heat loss and gain calculations that we do that focus on insulation R-value and not this thermal dampening effect.

      Given the summer temps (ignoring solar gain for now), this building probably would perform fine as a high mass building, to provide a stable indoor temp in May thru July, averaging the daily outdoor high and low temps (and avoiding solar gains from the south-facing windows those months due to the east, west and overhead position of sunshine May-July). I'm not a high mass building designer and don't know, but the designer friend has experience with this type of building design. But in winter, the outdoor temps get too low, and highs not high enough, to result in a comfortable averaged indoor temperature, so auxiliary heating would be necessary (e.g., mini split specified by the designer). The very high south-facing solar gains would have a major impact as well, particularly in winter months.

      You and I both agree that insulation should be added to the roof to decrease heat losses in winter. We both agree that insulation is key to an affordable, well-performing building in a cold winter climate. The designer friend apparently believes that the high mass effect would reduce the need for insulation, and I agree this is true to some extent (but probably not in winter), and agree with you that R-38 (or higher) ceiling insulation is a more efficient and safer bet than almost none (as planned).

      Of course, with the large expanse of south-facing glazing, solar gains would have a huge effect during winter. So I wouldn't classify this building as a pure "high mass" design, although the designer friend appears to have experience designing commercial buildings that are similar. Nor did it appear to be designed as a passive solar home either, since the thermal mass in the walls are covered with insulation on the interior side (ICFs). We don't know what solar gain and heat loss calculations, rules of thumb, or judgments based on this past experience, were made by the building designer friend. Payton could provide us some interesting data in the future.

      1. Payton | | #37


        Very interesting, this opinion of your designer friend.
        I'm just wondering if I have heat loss through the insulation of ceiling, would it not keep the concrete roof at a warmer temp than the cold outside air. So instead of the concrete being 22 degrees on the underside, when it's 20 outside, it could be 50 degrees on the under side? Now, would that transfer to the steel joists ? I don't know. Would they actually be warm and heat up to the concrete or would they be cold and penetrate that coldness to the lower part of the steel ?

        Which way would it be ? If metal conducts heat so well, wouldn't the steel stay warm in the winter if it wasn't insulated and not the other way around ?
        Yes, I would loose some heat energy out of the home this way, but how much ? The steel joist are only 1/8" thick.

        I'll definitely keep this post updated for future advise and experience.

        1. ROBERT OPALUCH | | #38

          Designer-friend: I was referring to YOUR designer, who you mentioned is a friend of yours.

          Concrete and steel readily transfer heat. When located next to one another, they readily transfer heat to each other. The warm air at the top of the building will be transferred through the steel to the concrete to the outdoors.

          I'd be interested to know what about this is confusing, or a better way to say this, or why you might want the steel or concrete in the roof to be warm. (Concrete only needs to be kept from freezing for a few days during curing.) I lead some building science classes so it helps to have clear explanations and address any confusion.

          The only reason to spend the extra money and effort to have ICF walls, rather than concrete walls with no insulation, is to keep the interior more comfortable despite your cold winter temperatures, and maybe to meet local building code requirements. Same goes for the roof. The purpose of adding ceiling/roof insulation is to keep your interior environment warm and comfortable during winter. You are paying extra for upgraded windows and ICF walls to reduce heat losses. But as we explored this design, it apparently has no insulation at the roof, which is unusual. I believe the requirement in your area was R-38 for ceilings, although what matters is what the local building inspection department will allow. They may have exceptions for high mass buildings, or they may not enforce any minimum insulation requirements. Your building permit was approved, so likely you are not required to add any ceiling insulation if you don’t want. You could add ceiling insulation later, if you are not satisfied due to cool drafts in winter, higher energy bills, or if the mini split recommended by your designer can’t provide enough heat (which you were concerned about).

          Ask your designer friend to see his heat loss calculations or estimates. You are supposed to complete a Manual J heat loss analysis to size the mini splits, but he may have used some other way to estimate the size of mini split he recommended. From what you said, he seemed confident in this estimate for mini split capacity.

          If you wish, you can use the following formula to calculate the expected heat loss of any part of your building envelope:
          BTUs of heat loss = Area X TempDifference X Time / R-value
          - Area = the square footage (ceiling, wall, window etc.)
          - TempDifference = difference between the desired indoor temp and the seasonal outdoor temp (e.g. 70F indoors, 31F average outdoor temp late January in Knoxville, or 17F design temp "coldest day expected in winter" to design your heating system capacity)
          - Time is typically 24 hours or one hour (results are BTU losses per day or per hour)
          - R-value is a measure of the insulation value of the building assembly (wall, ceiling, window, etc.)
          Assume your windows and doors are R-4 (U=0.25) if you don’t know. The ICF walls covered by T&G wood might be R-20 or your builder would know. The roof would be about R-2 as planned, or about R-40 if you add R-38 insulation. (Do this once for ceiling = R-2, and again for ceiling = R-40, and compare the difference. Adding R-38 ceiling insulation would cut your BTUs of roof heat losses by about 95%. I’m guessing it cuts your total home heat losses by about 20%.)

          Figuring out solar gains is more complex but not THAT difficult. Hopefully your designer did this too. It may be the reason he was not concerned about heat losses in the ceiling. But on overcast stormy days, you will get little solar heat gain.

          The thermal mass adds complexity as well but I don't know a good method to calculate its effects, just some rules of thumb and estimates based on experience. Your designer has experience to share, and no doubt incorporated it in your house design.

          1. Payton | | #43

            I misunderstood whom you were referring to.

            Can anyone elaborate on this company's claim to Closed-cell Foam.

            They are out of Oregon and I would believe their winters to be colder than mine.

            Are they only concerned about cooling in the summer in this article ?

            They state :

            Above is the Heat Flow Reduction chart that illustrates why there is no need to install any more than 3” closed cell spray foam. This amount provides a heat flow reduction of 96%. If we double that thickness, the heat flow only goes up a little over 1% which equated to an almost immeasurable/insignificant amount of energy savings. This is the diminishing rate of return of foam and why we only recommend 2” closed cell spray foam in the walls (94% heat flow reduction) and 3” closed cell spray foam under the roof.

            While you will receive a 94% heat flow reduction with only 2” of closed cell spray foam insulation, we recommend a minimum of 3” under and roof deck or in any attic assembly.

            This is due to 2 factors:

            A typical roof assembly is subject to a significant increase in temperature due to the radiant heat gain from the solar exposure of the roofing material. The 3” of Closed Cell Spray Foam will guarantee the 96% heat flow reduction which is needed for this increased heat gain.
            Large temperature swings in a roof assembly could cause an increase in moisture or vapor drive and the 3” of Closed Cell Spray Foam will almost eliminate the thermal drive which carries the moisture. (Vapor drive=latent moisture driven through an assembly by a thermal drive). This Vapor drive will vary from area to area based on the climate zones and the average heat days per year and is generally calculated by a WUFI scale for condensation.

      2. Expert Member
        NICK KEENAN | | #39

        The reason I get up in arms when people talk about "thermal mass" is that it's pseudoscience. From Wikipedia:

        "Pseudoscience consists of statements, beliefs, or practices that claim to be both scientific and factual but are incompatible with the scientific method. Pseudoscience is often characterized by contradictory, exaggerated or unfalsifiable claims; reliance on confirmation bias rather than rigorous attempts at refutation; lack of openness to evaluation by other experts; absence of systematic practices when developing hypotheses; and continued adherence long after the pseudoscientific hypotheses have been experimentally discredited."

        "Thermal mass" encompasses a set of beliefs that check all of those boxes. While I'm old enough to have seen at least three waves of thermal mass fads, what they all have in common is the belief that there is some magic property of concrete that makes it not follow the laws of physics.

        1. ROBERT OPALUCH | | #40

          That Wikipedia quote defines pseudoscience. It doesn't include anything about thermal mass, or high mass buildings. Claiming “thermal mass” is pseudoscience isn’t a logical claim, its your opinion without explanation.

          Read Michael Maines comment #10 above.

          Look at the popularity of adobe high mass buildings in the Southwest USA. Dry sunny climate with relatively mild winters. High mass buildings work fine there. They would NOT work well in a very cold climate. This house in TN is an interesting example as it is not a pure example of a high mass building, but the high thermal mass in walls and roof will have an effect on moderating internal temperature swings. That mass will have an effect on internal temperatures, it is not “pseudoscience.”

          Call it something besides "thermal mass" if you like. It’s the term that some designers use, and I’m sorry you don’t like that term being used. But concrete, masonry, water and other "high thermal mass" materials do absorb and store heat from surrounding air or from direct solar radiation on that mass, significantly more than low thermal mass materials. High thermal mass materials will radiate that heat back out to the surround air, when surrounding ambient temperatures are below the temperature of that thermal mass. This effect becomes more significant, usable, and noticeable as the temperature differences between the mass and the surrounding interior air increases. That is not pseudoscience. You don't understand how to use that effect in building design, and you don't need to. You don’t have to have high mass building components to design and construct an energy efficient building. But those of us who can design passive solar buildings in the cold dry areas east of the Rockies, or who design high mass buildings in the southwest USA, aren’t performing “magic” as you claim. We know something you don’t care to understand. We can design quantitatively to make it work if the climate is appropriate. Thermal mass dampens surrounding air temperature fluctuations. That’s a building science lesson you missed.

          REPLY TO NICK/DCCONTRARIAN post below (no reply button available below!)
          We already had this discussion before on another Q&A. Thermal mass allows heat to flow between the surrounding air and the thermal mass. You claimed that wood had more thermal mass, which is contradicted in the list of items posted in the attachment. I noted that you have to be able to transfer heat back and forth easily for this to work. Burning wood might release a lot of energy, but you can't unburn it to exchange energy back to store it. You replied that I needed to build two houses, one with thermal mass and one without, and note the differences. I replied that I already did that. The passive solar house I designed and built had a first floor and second floor of equal dimensions. The first floor had a tiled slab floor thermal mass, the second floor had carpet over ply subfloor, lacking much thermal mass. The first floor had daily wintertime temperature variations of about 10F, from 68F at dawn to 78F afternoon, without any supplemental heating, just solar heat gains. The upstairs had higher variation of low 60'sF to low 80'sF typically, yet it had less solar heat gains from less glazing, and less glazing heat losses. R-60 attic, R-25 walls, R-2 glazing (1980's). The tiled slab more readily absorbed heat, and more readily radiated that heat back into the room. Wood doesn't do that. Carpet doesn't do that. Drywall doesn't do that enough. Name it however you wish, everyone who designs passive solar thermal mass in homes uses some thermal mass (concrete, tile, brick, stone, water in containers) and calls it thermal mass, and it reduces daily heat fluctuations. Without it, a solar oriented home gets too hot on sunny winter afternoons, and can't retain enough heat for comfortable overnight temperatures. There is a community of passive solar designers, builders and architects who use this approach successfully. And there’s lots of failures by people who build without incorporating sufficient thermal mass, or overglaze without calculating solar heat gains, or who build without realizing the building must be a low heating load structure that’s airtight and super-insulated. Hard to argue with success, and claim its pseudoscience or magic. Its passive solar engineering and construction. You don’t do it, and apparently don’t understand how it works though its not complicated if you do the math and build to spec.

          I read everything I could find at the UCLA Architecture library, and took a course on passive solar and the ASHRAE Handbook of Fundamentals before designing and building. As I mentioned, I modeled the solar heat gain for fully sunny, average sun and overcast days; and conductive heat losses for average and 99% percentile January days. Calculated daily and hourly net interior temps for various conditions. Estimated convective heat losses via the crack method. Typically a 4” concrete slab floor works well for thermal mass. R-60 ceilings, R-25 walls and R-35 foundation insulation, with attached garage on the north side. So this isn’t just theory. The building performed as designed, and I was a bit surprised the daily interior temperatures tracked the predicted results as well as they did. But hey its elementary physics, thermodynamics, and attention to details when building the structure. And the house is built in a sunny, 5500 degree day cold climate with unshaded south-facing solar access mid-winter.

          This is not the same as a high mass buildings like adobe homes in the southwest US. They may also capture solar heat gains, but sun and diurnal heat variations get averaged out in their high mass walls. This is the theory behind the home under review. (Although there also will be tremendous solar heat gains on sunny winter days.) Seems an interesting case study. The designer ought to be in on this discussion. He could provide his own rationale for his design, ICF mass walls etc.

          Read Michael Maines comment #10 above. What's your response to that comment?

          1. Expert Member
            NICK KEENAN | | #41

            The physical quality you're talking about is "heat capacity."

            The implication is that ordinary construction has insufficient heat capacity, and that adding heat capacity makes for a better building. Neither of those statements is true.

            If you have conditioned space, the whole point of the conditioning is to prevent temperature swings in the interior. If you don't have temperature swings then you can't have heat flows.

            You're doing Payton, the original poster, a grave disservice by suggesting that because he has a concrete roof, by some magic the "real" R-value of that roof is somehow greater than the R-value of concrete that is measured experimentally. That kind of wishful thinking got him into the pickle he's in now, we're trying to work him out of it.

          2. Expert Member
            NICK KEENAN | | #42

            "But those of us who can design passive solar buildings in the cold dry areas east of the Rockies, or who design high mass buildings in the southwest USA, aren’t performing “magic” as you claim. "

            OK, to me "design" implies engineering calculations. What units of "thermal mass" do you use in your calculations when designing a building?

          3. Expert Member
            NICK KEENAN | | #48

            Comment #10 doesn't really support your position, he says the effect is negligible.

            Let me ask again: when you design a house with "thermal mass," what units of measurement do you use? How do you decide how much "thermal mass" is the right amount?

            The essence of the scientific method is that it is quantitative. Things that you can't measure are the hallmark of pseudoscience.

    2. Payton | | #36

      Thank you Nick ,
      I just contacted our local SPF contractor ( ) and asked him to come take a look. I feel more confident to ask more important questions and concerns. I will report back what they suggest.

      1. ROBERT OPALUCH | | #44

        It would be helpful if your designer were to join this conversation. He could communicate his vision and rationale for the design elements we are discussing. If he doesn't want to join the discussion, maybe you could ask him about it and summarize his views. High mass buildings are not discussed on GBA except this QandA thread.

      2. ROBERT OPALUCH | | #45

        In response to Nick/DCContrarian comment #41 above (no Reply option available there):
        You falsely stated "You're doing Payton, the original poster, a grave disservice by suggesting that because he has a concrete roof, by some magic the "real" R-value of that roof is somehow greater than the R-value of concrete that is measured experimentally. That kind of wishful thinking got him into the pickle he's in now, we're trying to work him out of it."

        Payton's home designer friend is the one who is designing a high mass building. I don't design those type of buildings, and don't claim that mass replaces insulation, quite the contrary.
        I was showing Payton how to calculate the losses IF there was no insulation, since he asked why not skip the attic insulation and keep the steel and concrete roof warm.
        "If metal conducts heat so well, wouldn't the steel stay warm in the winter if it wasn't insulated and not the other way around ?
        Yes, I would loose some heat energy out of the home this way, but how much ? The steel joist are only 1/8" thick."
        My reply noted that the result of no roof insulation would be a massive increase in heat loss during winter through the steel and concrete roof. Insulation would reduce roof heat losses by 95% or whole house by maybe 20%. I never suggested mass replaces insulation. I never suggested that more heat losses would be a better choice. Read what I wrote and don't misquote me.

        1. Expert Member
          NICK KEENAN | | #46

          I was responding to your post #35, where you wrote:
          "You and I both agree that insulation should be added to the roof to decrease heat losses in winter. We both agree that insulation is key to an affordable, well-performing building in a cold winter climate. The designer friend apparently believes that the high mass effect would reduce the need for insulation, and I agree this is true to some extent (but probably not in winter), and agree with you that R-38 (or higher) ceiling insulation is a more efficient and safer bet than almost none (as planned)."

          I admit, I'm a little unfair. But you do say that you agree that the "high mass effect" (whatever that is) "would reduce the need for insulation." I'm a little puzzled by the caveat "(but not in winter)" because it's the same house year-round.

          1. ROBERT OPALUCH | | #49

            What I said was "The designer...apparently believes that the high mass effect would reduce the need for insulation..." (That's not exactly me endorsing his views.)
            What you claimed I said was "you agree that the high mass effect...would reduce the need for insulation."
            Then you say you are puzzled by what I actually said.
            I'd respectfully suggest if you claim someone said something, quote their actual statement. That way your inferences about what you thought they meant are not communicated, only what they actually said. Especially if you don't believe you understand what they are saying. Its easy to make incorrect inferences, and attribute to the speakers something they do not believe, or even did not say. Happens all the time. Language is not that exact in communicating concepts. These QandA blogs on GBA mostly have managed to avoid the hostile online atmosphere that’s common in faceless text communications on the Internet. I’m asking you to please temper your responses, to continue this open exchange of ideas, information and feedback on GBA.

            To answer your question:
            Yes its the same house, but the same house varies in performance seasonally. Design details that work in summer may not work in winter.

            Outdoors its colder in winter than our desired indoor temperature of about 72F, all day long, every day, even in TN, where this house is located. I don't see how high mass walls would do much good IN WINTER as far as overall heat losses for the wall building assembly. In contrast, its often warmer outdoors in summer, but sometimes a little cooler than our desired indoor temperature. And around June 1 and mid-September, the desired indoor temp of 72F happens to be the average DAILY outdoor temperature. (See the attached chart of outdoor temps for TN.). That's when a high mass wall might help as it dampens or "averages out" those diurnal variations, before the heat is conducted to the interior. So the interior might be more comfortable, and less interior temperature conditioning might be needed with that high mass wall, than a typical low mass wood frame wall, during early summer and early Fall. Otherwise, a mini split might be heating a bit overnight, or providing some AC in the late afternoon. But the effect would likely be minor anyway, especially with ICFs not a simple high mass wall, and typical wood framed walls would have some insulation. Not to mention internal heat gains, solar heat gains etc. So it’s the same house, but the mass in walls probably helps a little in warmer months and reduces AC usage somewhat, but I don’t see how averaging the cold and colder winter temperatures would have any useful effect on heating the interior in winter. In the end, I don’t know because I don’t design high mass walls, didn’t design or model the expected performance of this house, but the designer is not here to explain his rationale.

            I wouldn’t claim the designer is incorrect in his assumptions, since I don’t do high mass building design, and he does have experience with them. Nor would I recommend or design high mass walls myself. Too bad the designer isn’t in this conversation to explain his vision, engineering and expectations. But I don’t think there’s anything wrong with analyzing the building to see how it might work or why someone designed it that way. I am interested in finding out how this house performs in use.

            If you believe in science, when there’s a disagreement in the scientific community, experiments are done to evaluate theories. You told me to do an experiment building two houses, one with thermal mass, the other without, and compare results. I pointed out I already did that type of experiment with a 4” tiled slab thermal mass on one floor, and far less thermal mass upstairs. The results show daily winter temp variations of about 10F vs. 20F. You proposed the experiment and these results contradict your theory. How do you explain that? Or why are you not willing to accept the results of the experiment you proposed, and consider that maybe there’s something to this thermal mass dampening/heat storage effect? Or at least not attribute pseudoscience to this concept.

            I should let you know I happen to have bachelor’s degree in philosophy, philosophy of science and logic concentration; a degree in applied mathematics, and a PhD minor in methods of research in cognitive science. I’ve taught statistics and research methods courses and led lab sections. Some of my scientific research is published. I was asked by scientific journals to serve as a reviewer for studies submitted for publication. So I’m not as ignorant about science as you might assume. What’s your expertise in materials science or whatever leads you to believe in your superior understanding on these issues? Please show us why you concluded that materials listed by designers as “thermal mass” are incapable of dampening interior temperature swings, or storing and releasing heat energy in buildings. Please explain why the Int’l Energy Conservation Code IECC includes “Mass Wall R-value” as an entry in their insulation requirements table, with lower acceptable R-values than a wood frame wall.

        2. Expert Member
          NICK KEENAN | | #47

          You're really losing me when you talk about "high-mass buildings."

          I recently tore down a building, conventional construction built in 1928. The dumpster company invoiced me for the debris, it was 112,000 lbs, not including the foundation, which we left. Or 50,000 kg. Is that high-mass or low-mass?

          1. Payton | | #50

            This home has over 300,000lbs of concrete alone.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |