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

How important is continuous wall insulation?

jollygreenshortguy | Posted in Energy Efficiency and Durability on

I’m questioning the practical need for continuous wall insulation in the situation described below. Am I off the mark and missing something important?

I’ve heard the importance continuous exterior wall insulation, especially to avoid thermal bridging. But it definitely adds some complications to detailing, such as at door and window openings, as well as adding steps in the work (for example 2 different types of insulation requiring 2 different installation practices and crew mobilizations). So I decided to compare 2 wall assemblies which both meet IRC2021, one with continuous insulation and one without.

I calculated the “whole wall R value” by using the method of calculating the u-value at the studs and the u-value at the insulated stud bays, and then using the ratios of the areas to determine an equivalent “whole wall R value”. See the attachment for the math.

By chance my 2 wall sections worked out to be identical in the final whole wall r-value result. Admittedly, my method may be considered somewhat simplistic. I did not use some fancy software to do a finite element analysis but I would be surprised if such an analysis would result in a significantly different result. Would it?

If my 2 wall sections provide an identical result am I really going to see a meaningful difference in thermal performance between the 2?

I did my analysis assuming IECC zones 4 through 8, with a code required maximum U-value of 0.045 per IRC table N1102.1.2. Here are the 2 assemblies:

Assembly 1:
exterior air film
ventilated rainscreen cladding (contribution to u-value ignored per Ashrae)
WRB (tyvek or similar, no contribution to u-value)
1/2″ thick sheathing, plywood or OSB, seams taped for air barrier
2×8 studs @ 24″ o.c.
dense pack cellulose at R3.8/inch
5/8″ drywall
interior air film

Assembly 2:
exterior air film
ventilated rainscreen cladding (contribution to u-value ignored per Ashrae)
R10 EPS exterior insulation, continuous, seams taped for WRB
1/2″ thick sheathing, plywood or OSB, seams taped for air barrier
2×4 studs @ 24″ o.c.
dense pack cellulose at R3.8/inch
5/8″ drywall
interior air film

Please look at the attachment before responding, as it may answer questions about my method and other details.

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.

Replies

  1. Expert Member
    Michael Maines | | #1

    It's hard to read your attachment but your overall approach and your results sound about right. There is nothing magical about continuous insulation, it just gets you more effective R-value in a thinner wall than cavity insulation alone.

    The only thing I noticed, and it's pretty minor, is that I use R-3.6/in for dense-packed cellulose. R-3.8/in is a good value for loose-blown. That's after reviewing the specs from every manufacturer I could find.

    1. jollygreenshortguy | | #2

      Thanks. I had mistakenly understood dense pack to have the higher R per inch. But I'll fix that. I'll also try breaking the attachment into 2 separate images which might make it easier to read.

      edit - I just checked the 2x8 wall with R3.6/inch and it still meets code but is slightly less than the other option of course, at a whole wall R value of 22.5 vs 23.4. So if I go this route I'll spec the lower figure and that will allow for contractors to go with either blown or dense pack, as they choose. If I could find an R26 cellulose batt that would work too, but I don't know if such a thing exists. I know I could get an R20 (5.5") and an R6 (1.5") and do it with 2 layers of batts. But that sounds like a hassle to install.

      1. Expert Member
        Michael Maines | | #3

        I have to think about it every time, as it seems more logical that higher density would be higher R-value.

        I haven't used cellulose batts but I would be surprised if they're available over 6" thick. Batts have binders and polypropylene fibers added; I prefer good old dense-pack when possible.

        1. jollygreenshortguy | | #4

          Dense pack it shall be! With loose blown as an option. Thanks again for your help.

          1. Trevor_Lambert | | #9

            I don't think you can do loose blown in a wall. It will just settle over time and leave a void at the top of the wall.

  2. maine_tyler | | #5

    11% framing factors seems quite low. I've never seen it that low. How about kings and jacks?

    You are using R 0.91 per inch for lumber?

    1. tdbaugha | | #6

      11% seems way low, especially when the wall has openings for windows and doors.

      1. jollygreenshortguy | | #11

        There is no reason to consider window and door openings as they would be the same for either wall configuration. I was only interested in comparing 2 wall section assemblies
        I included my math in the attachment. If you think it's wrong feel free to check the numbers. That's why I included it.

        1. maine_tyler | | #14

          Not the openings themselves, the framing to support. Jacks, headers, and off-layout cripples and kings. You are suggesting that won't change?

          1. maine_tyler | | #18

            I didn't see your post #10 re headers.

            "I could have calculated this thing to death with an entire wall section, replete with windows, jack studs, headers, blocking, ... and spent hours at it, only to find that it makes a minor difference."

            That's not the suggestion. The suggestion is to use more standard advanced framing factors. Yours are decidedly non-standard and it took you more work to generate them, not less. And despite your assertion about insulating headers, I suspect your 11% number is under reality. But you are right that this doesn't necessarily change things to a huge degree. I see no problem with your overall approach/logic.

          2. jollygreenshortguy | | #20

            For a specific wall with specific openings whether the studs are 2x4 or 2x8 won't change the surface area of framing to the total wall area. There will be the same number of studs, king studs, jack studs, etc.

            I really don't see that I need to make this any more complicated for my present purposes. All I'm trying to do is see how much difference there is in round terms, between 2 wall assemblies. What I've determined is that they are very nearly identical. If my results are off by 2-3% one way or the other it's not important. I don't see the need to be any more exact than that and waste hours pursuing this.

          3. maine_tyler | | #22

            I don't understand why people come on here with this attitude.

            You posted a question for review.

            Reread my post #18 and maybe you'll get it.

            Correct, it has nothing to do with 2x4 vs 2x8 studs. Your exterior CI covers those members, that's the difference.

            You've already done all the extra work to come up with 11% so, yeah, don't do any more work.

            If others in the future find your post and use it for guidance, it may be helpful for them to understand that your 11% framing factor is extremely uncommon and probably flawed... even if it doesn't change your own conclusion.

          4. jollygreenshortguy | | #24

            maine tyler, I seem to have offended you, entirely unintentionally. I'm sorry about that.

            I'll take a design I'm currently working on, take one wall from it and do a full, detailed area take off for framing area, stud bay area, and headers. I'll do it for 8'1" wall and a 9'1" wall heights.

          5. maine_tyler | | #26

            Sorry Jolly, I'm a bit short on sleep these days.

            I was not at all offended, merely frustrated that the point didn't seem to be getting through. For what it's worth, I don't disagree with your general conclusions; was just pointing at a possible numerical error.

    2. jollygreenshortguy | | #10

      In the attachment I included the calculations for a windowless 4' section of wall 8'-1" high. The actual figures were 89.9% and 10.1%, with better values for a 9'-1" wall. I rounded the 10.1% up to 11%, though the rounding doesn't amount to much. I will be using headers insulated to the maximum I can manage, such as rim joist headers when possible. So the headers above openings will be nearly at the level of the stud bays.

      I could have calculated this thing to death with an entire wall section, replete with windows, jack studs, headers, blocking, ... and spent hours at it, only to find that it makes a minor difference. Or I could try to use my time more efficiently to see quickly whether I'm on the right track. I chose the latter option for now.

      If I need to go into more detail I'll do that another time.

    3. jollygreenshortguy | | #13

      Yes, R0.91 for lumber is what I found in one reference. I had always assumed around R1/inch but decided to check and as 0.91 gave a more conservative result I went with that. Again, I don't see that a small increase to the lumber R value would make a significant difference in the result. But if it seems necessary I can always go into more detail.

      1. Expert Member
        Michael Maines | | #27

        JGSG, now that I can read your values, I agree that an 11% framing fraction is low. Even with "semi-advanced" framing, using a few sticks less than normal, when I have done calculations it usually works out to around 15% framing and some reliable sources suggest 20% or more. And after a lot of time researching R-values, I use R-1.2/in for framing lumber. I doubt these changes will drastically change your result, but if you want to be more accurate, you might run the calcs with different inputs.

  3. pnwbuilder | | #7

    Exterior insulation will help to keep sheathing temperature above dew point and prevent or at least minimize condensation. Depending on your climate zone this may or may not be important.

    1. Expert Member
      MALCOLM TAYLOR | | #8

      pnwbuilder,

      True. It would be interesting to look at both assemblies from a moisture management perspective for a couple of different climates.

      1. jollygreenshortguy | | #12

        I followed the IRC requirement as far as vapor retarders. This is shown in a note on the bottom of the second attachment. If I meet the insulation levels and vapor retarder requirements of the Code is there any reason I should expect poor performance? How would I analyze my wall sections for different climate zones to check for that?

        1. Expert Member
          BILL WICHERS | | #19

          You would need to work out the temperature of the condensing surface, which would be the interior facing side of the exterior sheathing. You can then work out how often that condensing surface will be below the dew point of the inside air, and your result will be something like number of hours or days that surface sisbelow the dew point, with the goal to minimize how much total time the surface is below that dew point temperature.

          BTW, I wouldn't allow for the air films. I've never liked counting those towards total R value, especially on the exterior, because they are affected by air currents. Basically on a windy day, that air film is pretty much gone on the exterior. You also list the air film on the exterior side of vented sheathing, and an insulating layer outside of the ventilation gap doesn't contribute to the overall R value of the wall. In your overall analysis, that is probably a pretty minor detail with a minimal effect on your results, but it's worth mentioning for completeness.

          Bill

  4. STZ_MA | | #15

    PTF. Considering two options myself (zone 5)

    #1: lap siding over 1x ventilation gap, 3” exterior rock wool, mento 1000, 1/2” cdx, 2x6 with dense cellulose, intello+, 5/8 gypsum

    #2: lap siding over 1x ventilation gap, zip sheathed 2x6 wall, rock wool batts, with intello+ inside then 5/8 gypsum

    Builder has suggested going with #3, which is #2 but with Zip-R12 as a middle ground.

    #2 meets code and is a much much better price point. Wondering how important it really is to do exterior insulation given the extra expense. Thinking along the lines of air tightness is more important than extra R-value.

    1. kyle_r | | #16

      I used BEopt to analyze a similar situation. You can see the post here.

      https://www.greenbuildingadvisor.com/question/beopt-modeling-of-ci-energy-savings

      1. STZ_MA | | #23

        Thanks for the link - this discussion is in line with my suspicions, and the build you describe is very similar in size & scope to my own.

        I am getting the feeling that with taped zip, aerobarrier, & intello+ cold sheathing and air sealing should not be much of an issue, and are details most contractors can execute at a price point that is reasonable.

        Comfort is a hard item to quantify. Personally I get a lot of comfort from having no debt and a large balance in my brokerage account. I am starting to question whether the additional comfort I would get from chasing my ideal building envelope would overcome the loss of comfort from spending my savings down and going deeply in debt. Still deciding.

        Right now I have proposed the #1 build to two different builders. One is a high performance designer / builder & building science expert, and the other is a local custom home builder with limited understanding of building science concepts.

        Both builders quoted me prices around $350/sf and up depending on finishes, not including design, permitting, land, septic, solar, or landscaping. I am really hoping to drive that number into the $250-$300 range, so all of my original details are on the chopping block. For reference, new 'energy efficient' homes in my area of similar size, all-in, are going for $300-$350/sf, so I would be wildly upside down if I spend $350+ on just the structure.

        1. kyle_r | | #25

          I think you will find the cost of rockwool batts very high, obviously region dependent. You may want to consider cellulose or plain old R21 kraft face fiberglass as well.

          1. STZ_MA | | #28

            Yeah dense pack cellulose is my back up plan, it saves a good amount on materials.

            I asked the local builder what he thought would be possible price wise if we were to build code minimum, no upgrades, with the most basic finishes. He told me, “turn-key I think we could get down to the low 3’s but probably not $300/sf.”

            Starting to lose hope! How can it be that expensive! A friend in Texas was complaining about $175/sf on his barndo, I would be so excited if I could get a quote sub $300.

            Or should I be building a barndo?

        2. jollygreenshortguy | | #30

          "dense pack cellulose is my back up plan"
          Something to consider is that cellulose is a lower carbon material than mineral wool.

          https://www.greenbuildingadvisor.com/article/choosing-low-carbon-insulation

  5. jberks | | #17

    Thanks for putting this together. it's good to see how thicker 2x wall assemblies can match a 2x4 w/ 2" of CI.

    If this exercise is just about straight R value, it's a good point.

    My initial thoughts go more towards hands on stuff. Like the extra 3¾" width inboard of the house the 2x8 takes or the extra eye rolls you'll get from framers on working with 2x8's or the extra labour or machinery needed to lift 2x8 walls.

    Also I like to use specifically Rockwool comfortboard for CI, as expensive as it is, because of its hydrophobic nature and vapour openess, I see it as an extra layer of protection for the sheathing. Also its fire resistance is a bonus and is pretty much a requirement in my dense urban parts. And animals hate it, including humans.

    Jamie

    1. Expert Member
      BILL WICHERS | | #21

      Exterior CI can get you more R value in the same physical space as a deeper wall insulated with batts (or equivalent) alone. that's probably the biggest difference on a new build. On a retrofit, you usually don't have the option to reframe the wall, so exterior CI is the only way to beef up the R value. With some retrofits you can change things around and squeeze some more depth in somewhere, but I think most people try to avoid doing that.

      I personally prefer CI to a deeper wall, but that's just my own personal preference. There are other ways to hit target R values, such as deeper walls, staggered studwalls, etc. I do think CI generally makes for a more robust wall in terms of moisture resistance though, since CI keeps the exterior sheathing warmer, where a deeper wall does not.

      Bill

  6. rockies63 | | #29

    I would watch this video from Christine Williamson of Acelab for Architects called "Why Exterior Insulation is the MVP of Insulations"
    https://www.youtube.com/watch?v=sV7yC4XGeTE&t=19s

Log in or create an account to post an answer.

Community

Recent Questions and Replies

  • |
  • |
  • |
  • |