GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Audio Play Icon Headphones Icon Plus Icon Minus Icon Check Icon Print Icon Picture icon Single Arrow Icon Double Arrow Icon Hamburger Icon TV Icon Close Icon Sorted Hamburger/Search Icon
Q&A Spotlight

Comparing High-Performance Wall Options

An owner-builder aiming for R-50 exterior walls ponders the best way of reaching his goal

This high-performance wall option comes from Passive House designer and architect Steve Baczek.

T. Barker is planning a high-performance house in Climate Zone 7A, where the number of heating degree days reaches 10,000 a year and the January design temperature hits 20 below zero.

Although he is still tinkering with his plans, Barker is leaning toward a design with R-50 exterior walls to make the house as comfortable as possible and to reduce the cost of the mechanical systems. The question is how best to get to that insulation value.

In a Q&A post, Barker asks, “Are there any good reports using actual construction experience comparisons for the cost to build double-stud 2×4 (or 2×6) walls compared to single 2×6 with exterior insulation?”

What reading he’s done so far suggests that most double-stud walls are framed on 24-inch centers. “I would think for the minimal cost difference to stay 16 inches on center you get better drywall finish and shear wall capacity for 2-story construction,” he says.

What does the most recent research show in terms of material costs, construction labor, and pros and cons of each of these two options? Further, are there insulation materials on the market with very high R-value that might be a good fit for his house, such as silica aerogels? Those are the questions raised in this Q&A Spotlight.

Double-stud walls are more practical

Reaching R-50 with a 2×6 wall plus exterior foam takes a lot of foam, says Dana Dorsett, something on the order of 6 inches. That makes construction awkward, and may be a good reason to choose the double-stud option.

“A double stud wall is far more practical, since the distance between walls is flexible — two feet thick isn’t substantially more difficult than a foot,” Dorsett writes. “You have to pay attention to the thermal bridging of the subfloors and band joists, etc., and dense packing cellulose gets harder, making fiberglass (which unlike cellulose won’t sag or settle if you don’t quite hit the target density) perhaps a better option.”

As to Barker’s questions about the most up-to-date research, Dorsett points him to a study by the Building Science Corp. Even though it’s not quite as recent as Barker would like, Dorsett says, it still offers sound advice.

But, he adds, forget about aerogel. “Aerogel is ridiculously expensive, and would be totally wasted as cavity fill,” he says, “but there are commercial products designed for framing edge strips, the primary market being commercial construction with steel-stud curtain walls.”

It’s a matter of builder preference

The choice between a double-stud wall and one with exterior insulation is a matter of the builder’s preference, says Michael Maines.

“The ones who prefer double stud walls and don’t like working with foam say that double stud walls are the best value, and that installing exterior foam is difficult,” Maines writes. “The ones who like using foam say that double stud walls are twice the labor and twice the material, and that installing exterior foam is not a big deal. I’ve had this conversation with many different builders and most of them fall into one camp or the other.”

With that said, Maines has found that exterior foam is the best value when the goal is in the R-30 range. For R-values of 40 or higher, a double-stud wall is a more economical choice. The cost difference between an 8-inch-thick wall and an 18-inch-thick walls is just in the extra insulation.

“There are many variables in the equation,” he adds, “so the answer will depend on your specific situation.”

Nor does the choice between 16-inch and 24-inch on-center framing have a clear answer. While 16-inch framing doesn’t use significantly more material, it does use more, and the extra studs don’t make the building substantially stronger. (Builders of three stories or more, however, are required by the International Residential Code to be framed on 16-inch centers.)

Making the case for R-50 walls?

“Can I ask the big question,” says Walter Ahlgrim. “Why R-50? Given the current prices of fuel, materials and labor in most places, an R-50 wall is unlikely to justify its cost in terms of dollars and cents.”

If Barker is shooting for performance numbers required in a Passive House building, R-50 might make sense. Or Barker may live in an area where the cost of fuel is unusually high. He suggests Barker spend some time with a no-cost computer program called BEopt, developed by the National Renewable Energy Laboratory, to learn more about this options.

Packing R-50 worth of insulation into the walls, and R-100 into the roof, will save about $500 a year in heating and cooling costs, Barker replies, in addition to the C$20,000 in mechanical systems. These savings over 15 years will add up to at least C$25,000.

“Additional costs to superinsulate and seal a 2,200-square-foot house will be approximately C$35,000,” he adds. “I can live with the extra cost for better comfort, a quieter house, and being green (never thought I’d hear myself say that — LOL).”

Fuel is indeed a problem. There is no natural gas available on the site, and while propane is a possibility, it’s expensive. Plus, Barker sees no sense in running gas lines into the house when the future is clearly electric.

Several GBA readers question Barker’s estimates of how much money he will save in HVAC operating and installation costs. Will he really save $25,000 over 15 years?

“I wasn’t clear,” Barker says, “but I’m comparing R-50 ‘superinsulated, super-tight’ against the standard that most decent homes would be built to in this area today, which is about R-24 walls, ACH50= 3.0, etc. This happens to be about the quality my current home was built to 20 years ago, so I use electrical and utility usage from that house to make some of my comparisons. Will it end up being R-35 or R-40 instead of R-50?  Maybe, but it will certainly be far more insulated than R-24.”

“At a whole-wall R-value of R-30 or more, the cost difference between R-30 vs. R-50 whole-walls or an R-50 vs. R-100 in the attic is often better spent on upgrading or fine-tuning the window options,” Dorsett adds. “Walter’s recommendation for maintaining a BEopt simulation as you adjust the design features is a good one. In most climates the heating and cooling loads can be cut to the level where point source heating/cooling can work without taking it to R-50 whole-wall and R-100 attic.”

Choices for exterior insulation

Barker has done some research on the cost of different types of exterior insulation and finds that mineral wool is the cheapest option. To reach R-50, he says, the cost of 12-inch-thick Rockwool would be about US$2.05 per square foot of wall area; EPS (13 inches) would be US$3.59 per square foot; XPS (10 inches) adds up to US$5.47 per square foot.

GBA editor Martin Holladay wonders why Barker might choose XPS for exterior rigid foam when most green builders prefer EPS or polyisocyanurate.

Barker says that he wants to stay away from polyiso because of its reduced R-values at low temperatures and, he suspects, some environmental issues.

Dorsett suggests he consider reclaimed polyiso. “Large industrial and commercial flat roofs are usually insulated with polyiso or EPS, which is often swapped out and upgraded during re-roofing,” he says. “The used foam is ‘gold’ for materials reclaimers — they make a good margin even when reselling it at less than 1/3 the price of virgin-stock foam.”

Plus, he adds, polyiso’s reduction in R-value at low temperatures is probably not as severe as Barker thinks.

Our expert’s opinion

GBA technical director Peter Yost adds this:

It’s surprising to have such a detailed discussion of wall performance without mentioning windows. This is especially true given the difficulty that windows introduce in either a 6-inch exterior foam walls or double-stud walls. What is their performance relative to the walls? What’s the increased difficulty of window installation in thick walls? There are lots of decisions to make, and some careful detailing to manage.

I decided to ask leading Passive House builder Steve Baczek what his preference would be for “superinsulated and super-tight” Climate Zone 7 walls. He provided the drawing at the top of this column, and added this:

“I’d love to tell you there is a ‘silver bullet’ wall assembly out there, but — sorry, I’m not aware of any.

“When it comes to enhanced wall assemblies, the choice is usually between a thicker frame (i.e. 2×8+) or a double wall assembly. I’ve done both numerous times and they each have their pros and cons. What I’ve found generally is that it really comes down to the builder’s perspective.

“I can make either assembly perform well, but builders apply the $ tag to it. If I asked 10 builders, I would likely get a 50/50 split in favor of each. We can debate a lifetime on this, but here’s one of my contentions: We worry so much about the R-value of the wall at the cavity, that we usually neglect the wall at the window.

“Understanding the impacts of whole wall R-value suggests heavy attention to exterior continuous insulation and, more importantly, to the windows. My stepping off point for a high-performance wall is usually a 2×8 24-inch on-center wood-framed wall with R-9 Zip  sheathing. Everything is a bit enhanced, but the number of parts is the same as a code-built wall. So I am essentially paying extra for enhancements, not additional labor. This wall for my climate gets me into a R-40ish+ range. It also usually leaves some money for that very important window upgrade to leverage my whole-wall R-value.”


  1. Doug McEvers | | #1

    This NREL case study may answer some questions. They found the traditional double stud wall to be the most cost effective.

    Harold Orr defined superinsulation at the 2007 Passive House Conference as annual heating degree days divided by 180 for wall R-value and 120 for ceilings. With this in mind R-50 would be a bit light for 10,000 hdd. With this formula and an ACH50 of 1 or less heating for a residence will be right around 1 Btu/sf/hdd.

  2. Malcolm Taylor | | #2

    A question occasioned by the accompanying illustration: Are there climates where the advantages of open rain-screens outweigh their downsides?

  3. Peter L | | #3

    It's the windows, stupid. To steal and rephrase a line from American politics.

    Who cares about R-50 walls if you get code standard windows! If you took a R-50 wall assembly and installed a 15% glazing ratio of dual pane (R-3) windows. That R-50 wall assembly would drop down to a whole wall average of R-15. That is a 70% reduction in R-Value. That's worse than a R-19 cheapo fiberglass batt installed wall without windows.

    Now, take that same R-50 wall assembly and install a 15% glazing ratio of a high-performing (R-8) windows. That R-50 wall assembly would now have a whole wall average of R-28. A significant increase in the average whole wall R-Value compared to the standard dual pane window.

    So I drive home the point. Why so focused on a R-50 wall only to reduce it to R-15 +/- due to code standard R-3 windows.

    Spend the $$$ on better windows than trying to achieve R-50 walls.

  4. User avater
    Dana Dorsett | | #4

    >"It’s surprising to have such a detailed discussion of wall performance without mentioning windows. "


    >At a whole-wall R-value of R-30 or more, the cost difference between
    >R-30 vs. R-50 whole-walls or an R-50 vs. R-100 in the attic is often better
    >spent on upgrading or fine-tuning the window options,” Dorsett adds.

    (Just sayin'... ;-) )

    An R30 wall isn't too tough to build without making excessively thick/awkward walls, and differences in window performance can exceed the energy use difference and peak loads than going from R30 to R50 on the wall performance.

    1. Peter L | | #9


      You are right, you did state that, I was more referring to the OP (Barker). Sometimes the focus is so much on the wall assembly and then little or no effort is ever made on the windows. I've seen my share of R-30+ walls and then they just install R-3 vinyl windows. Even when they have the $$ to upgrade, for some reason they don't. Either through ignorance or misguidance by the architect/builder.

      To me that's like getting humans into space but then not thinking about or designing the spacecraft to get them back to earth. That's just ONE part of the equation but a home is a whole and greater than then sum its parts. Doing a R-30 wall and getting R-8 windows is the better route than doing a R-50 wall with R-3 standard windows. Unless one plans on having less than a 15% glazing to wall ratio which is very unlikely as the average today is about 30-40% window to wall ratio. Last time I checked...

  5. User avater
    Michael Maines | | #5

    The question was about walls, not windows, so I don't understand the surprise... Obviously windows have a big impact, but that doesn't mean we should ignore wall assemblies--especially when that's what the poster was asking about.

  6. Roger Berry | | #6

    Just have to chime in with support for Peter L and Dana on the window thoughts. The question is obsessing about one parameter and seemingly confusing the whole house performance of an existing house with that one design point. It is actually believable that one could save $1666 in energy costs over a "normal" house given the climate zone, but as Peter and Dana have noted, the window question needs to be integrated into the design solution first. Given the amount of postings on wall design questions where thermal bridging from studs is still a point to be made, I would think it quite reasonable to nudge the questioner to the bigger picture re: energy efficient design.

    1. Malcolm Taylor | | #7

      Not to disagree, as I think you are right, but I guess it also depends in how you view the various components of the envelope. A case could be made that windows are replaceable, and that advances in window performance may make that a desirable thing to do sometime in the future, in a way that improvements to the wall and roof assemblies are not, being fundamental parts of the building structure. So coupling the R value of the walls and roof to present window performance might lead to an imbalance down the road.

      1. Peter L | | #8

        Double pane windows were developed in the 1930's and became popular in the 1970's due to the energy crisis. Fast forward almost 50 years later and double pane windows are still the mainstream window install in 95% of all homes built today. Only those building custom energy efficient homes are installing better windows (triple pane, heat mirror, etc). I highly doubt we will see any great advancement from the current R-8/R-10 triple pane/heat mirror windows that we have today. A window is still a window and only so much can be done with it to improve efficiency. A windows energy efficiency will never meet or beat a wall assembly. Not even Elon Musk can do it ;)

        Removing and replacing windows later on after the home is built usually costs triple of what it would have costed had they installed better windows in the first place. Cutting exterior stucco, removing drywall, removing siding, brick facade, re-flashing, etc. is something that adds a lot to labor and material costs. Then one hopes that the installer can flash the new windows properly to prevent water intrusion. A friend just got a window R&R quote and it came out to $25,000 to remove & replace two dozen or so windows. Always better to do it right the first time.

        1. User avater
          Dana Dorsett | | #10

          Most of windows in my house are 95 years old, just like the siding. (The storms are about 30 years old). Replaceable- yes, but at what cost/benefit?

          The cost effectiveness of replacing windows is TERRIBLE compared to the cost effectiveness of fine tuning the size SHGC & U-factors to optimize the energy performance of a house while it is still in the design phase. It'll never be cheaper or more cost effective than it is to do this right now, not 25, 50, 100 years from now.

          Spending extra on a high-R wall and ignoring the windows figuring that's a detail best left to the next generation (or a couple of generations beyond) is just silly. Better windows can be a cheaper performance upgrade than the next R20 of wall-R, even if one assumes re-glazing in 25 years.

          1. Peter L | | #11

            Well said!!

            Do it right the first time.

          2. Malcolm Taylor | | #12

            I completely agree. But just as it makes no sense to optimize your wall and roof design to match poorly performing windows, if the next innovation in glazing significantly increases its R-value, it would be a shame for the other building assemblies to then become the comparatively weak link.

        2. Alan B | | #13

          Not even Elon Musk can do it ;)

          Don't give him any ideas :D

  7. User avater
    Dana Dorsett | | #14

    >But just as it makes no sense to optimize your wall and roof design to match poorly performing windows, if the next innovation in glazing significantly increases its R-value, it would be a shame for the other building assemblies to then become the comparatively weak link.

    True. And?

    R30 whole-wall is already ~50% better than a code-min 2x6 R20 + R5 c.i. (which is about R20 whole-wall). Hitting 50% better on code for windows would call for U0.21 windows. In zone 7 that would have to be a triple pane (for wintertime condensation issues), not a double-low-E with a hard coat on surface #4, but "worth it" compared to bumping out the wall-R to R50. Smaller windows, with the SHGC tuned by the direction it's facing can make that affordable. Every square foot of U0.21 window replaced by R30// U0.33 wall reduces the heat loss of that square foot by 70%.

  8. User avater
    Jon R | | #15

    Most of the window/wall debate was addressed with:

    > He suggests Barker spend some time with a no-cost computer program called BEopt,

  9. sethw | | #16

    I'd also recommend considering simplicity of construction assembly, robustness of construction assembly, and ease of achieving air tightness. Other important factors will be: orientation, windows & doors (already much commented on), percent glazing area, roof insulation slab insulation, and others. The assembly drawn is pretty complicated and could present a bunch of field coordination issues including options for pour construction depending on the contractor.

  10. Blake Schatz | | #17

    Why not stagger the studs? 2x8 top & sill plates, with 2x6 studs staggered. You get a single wall assembly but will less thermal bridging at studs. You'd probably want spray foam insulation for the full cavity depth to avoid compressing batts at the reduced insulation thickness at the studs. You probably won't get to R-50, but with good air sealing this seems like an effective solution.

  11. sethw | | #18

    Have they considered SIPs?

    1. Peter L | | #19


      With polyurethane SIPS at 6.5" which would give you a R-40. They have roof SIPS at 8.25" which give you a R-50. Those are true #'s as there is no thermal bridging. Not sure if you can use roof SIPS on the wall??

      1. sethw | | #20

        Yeah. Peter.
        Think of how much simpler those construction wall assemblies would be with SIPs, an interior furr-out for rough-ins and some additional vapor-open insulation, and exterior WRB with siding? Boom. Done.
        As far as roof - there are options, but SIPs roof could work with possibly some added insulation on top depending on roof type.
        You always have to be careful of vapor diffusion but SIPs to do the majority of the work for insulation and wall/roof seems to simplify those construction assemblies.
        Simpler is usually better in my book.

  12. T. Barker | | #21

    Excellent article!
    I'm just catching up on some reading after an extended hunting trip, and then very busy at work.
    Obviously windows play an important role. The assumption is I will use at least very good windows (e.g. R-6), and possibly extremely good R-10 European type windows. That is a given.
    I really like Steve Baczek's solution above! I never thought about 2x8 walls, but very likely the way I will go. I have some concern about the inflated cost of 2x8's in some regions because they aren't as common, but hopefully I can sort that out.
    My only question regarding Baczek's detail is why 1" of Foam-Lok plus the Rockwool, instead of just 7-1/4" of blown-in Rockwool?
    Now that I feel more comfortable with a wall assembly that can get me to at least R40, I'll run some scenarios through BOpt.

    1. User avater
      Steve Baczek | | #23

      The 1" of Foam, does two things, 1. It warms the cavity a bit. 2. It provides a secondary air barrier to the exterior air barrier. Not necessarily required (moving towards warmer climates) but for a recent project it seemed the best solution.

  13. T. Barker | | #22

    The other thing I just thought about is the recommended minimum thickness of exterior foam insulation to avoid moisture issues in the winter. For my design conditions I believe a minimum of 4" of exterior foam is recommended. Or does the Zip Sheathing product eliminate this requirement?

  14. User avater GBA Editor
    Martin Holladay | | #24

    T. Barker,
    In Climate Zone 7, if you want to combine exterior rigid foam with fluffy insulation between the studs, the minimum R-value of the exterior rigid foam is as follows:
    R-10 foam for 2x4 walls
    R-15 foam for 2x6 walls
    R-20 foam for 2x8 walls

    With Zip-R sheathing, the first condensing surface is the polyiso adhered to the OSB on the Zip R -- and that surface needs to be warm enough in winter to prevent moisture accumulation. So the same rules apply.

    For more information on this issue, see these two articles:

    "Calculating the Minimum Thickness of Rigid Foam Sheathing"

    "Combining Exterior Rigid Foam With Fluffy Insulation"

    1. User avater
      Steve Baczek | | #26

      Martin - that's why I add the closed cell spray foam at the cavity - granted the assembly above I did for climate zone 5 not 7. If I did 7 I would up it to 2". Everyone always get s on the moisture train with exterior foam. I always question, "where is the moisture coming from?" If I maintain the RH inside via ventilation, and the wall assembly operates at 1.0 ach50 Pa or better, then where is the moisture coming from? I think alot of the fear is tied to bad wall assemblies that have air moving thru them like 7.o ach 50 Pa or worse. Those walls move moisture via air leakage, and yes they scare me too. But the wall above with controlled ventilation and virtually no air leakage, diffusion just doesn't scare me.

      1. User avater GBA Editor
        Martin Holladay | | #31

        I understand your rationale for the spray foam, and I hope you realize that I wasn't implying that your foam wasn't thick enough. I was just answering T. Barker's question with guidelines for walls in Zone 7.

        1. User avater
          Steve Baczek | | #34

          Martin - Love ya like a brother man!!! Keep up the great work......Long Live Our Buildings

  15. Andrea S | | #25

    Hi all,
    Doesn’t the proposed 2x8 assembly by SB with its total of 15.6 impermeable insulation Conflict with the R20 exterior foam requirement for zone 7? Also, isn’t the ratio of impermeable to permeable (38%\62%) off... ?
    Thanks much,


    1. User avater
      Steve Baczek | | #27

      Andy - the detail above was drawn for climate zone 5.

    2. User avater
      Jon R | | #32

      > the ratio of impermeable to permeable

      It's not a permeability issue, it's a sheathing temperature issue. For example, highly permeable external rockwool works even better than impermeable (eg, foil faced) external foam.

  16. Andrea S | | #28

    hi Steve,

    Got it! Thanks

    I am in zone 5 working with 2x6 walls thinking of hunter panel with r 18 (discounted to 16 for cold weather polyiso performance) on exterior plus zip structural sheathing as main air barrier... then 5.5” batt rockwool on the interior (r23). Total R 39 with 41% impermeable in the exterior. I would very much appreciate thoughts and comments. Thanks

    And yes... am using high performance windows 😁

  17. User avater
    Dana Dorsett | | #29

    When the polyiso is getting on to half the total R, in a zone 5 climate it doesn't really need to be derated. Most of the heating season it will meet or exceed labeled performance.

    In a typical zone 5 climate the mean outdoor temp even in January would be north of 20F. With your R18+ R23 ( R41 total labeled) stackup , 20F outdoors and 70F indoors the mean temp through the polyiso would be about 30F and it would underperform slightly. But whenever the mean outdoor temps hit north of 30F (which is most of the heating season in zone 5) the mean temp through the foam is north of 45F, where it will meet or exceed it's labeled performance. If it's under performing a bit for 15-20% of the hours and outperforming labeled-R the 75% of the time the average performance will be at least as good as advertised.

    From a dew point perspective even fully derated there is plenty of margin. From a peak heat load perspective the difference is "in the measurement error noise", almost impossible to measure, low enough in magnitude that the error bars on other factors are larger.

    1. Andrea S | | #30

      Hi Dana,
      I understand that I don’t have to underrate the polyiso. Are you saying that the assembly is a good plan?

      1. User avater
        Jon R | | #33

        Just in case future readers missed Dana's point, you need to use different de-rating values depending on if the concern is seasonal R value or reducing condensation on the sheathing. The latter, in cold climates , could need as low as R4/inch for polyiso (because of temp and very long term aging effects).

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |