GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Why is there such a bias against fiberglass?

user-987058 | Posted in General Questions on

My wife and I are adding a carrage house and I am interested in blowing in fiberglass behind a net. I am planning on building a carrage house and I have been researching the possibility of completing as much of the work as I can.

I am in climate zone 5 and it looks like I can use a product by Knauf called EcoSeal and then insulate the cavity with fiberglass netted. But as I read some of the other conversations going on I am not of the impression that fiberglass is well thought of.

The solution we choose needs to offer comfort and be as cost effective as possible while maintaining low bills as my wife and I expect to have a tenant renting out the space above the garage. Any help will be wonderful.

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. GBA Editor
    Martin Holladay | | #1

    Blown-in fiberglass performs much better than fiberglass batts, because blown-in fiberglass (like cellulose) fills every nook and cranny.

    That said, blown-in fiberglass is not quite as dense as blown-in cellulose, so it isn't quite as effective at reducing air leakage as cellulose. Your decision to use the EcoSeal system is a good idea; if properly installed, the EcoSeal will greatly reduce air leakage through your wall.

    As long as your contractors do a good job, you can expect good performance from your proposed wall assembly.

  2. Expert Member
    Dana Dorsett | | #2

    Batts have a bad rap due to installation issues and high air-permeability reducing performance at the temperature extremes. Even new-school fiberglass blowing wools have high internal convection losses when blown at low density, even if they don't have the compressions & voids that typify batt installations. To get air-retardency even on a par with low-density cellulose (which has almost no loss of performance to internal convection) most of the newer-better blown fiberglass needs an installed density of ~1.8lbs/cubic-foot, even though 1.0lbs is still considered a standard & acceptable density from a building-code point of view. (Most manufacturers publish R-values at 1lb & 1.8lbs.) At 2lbs+ density the air retardency of these products is quite good- comparable to 3.5lb dense-packed cellulose.

    But while 1.8lbs is sufficiently air retardent to mitigate performance loss at big temperature deltas, it's still quite air & moisture permeable, and to protect the structural sheathing from wintertime moisture accumulation will require either a sufficient exterior R from foam or rigid fiber products exterior to the structural sheathing foam (the minimum sheathing-R outlined in the IRC is R7.5 for 2x6 framing, R5 for 2x4 framing), or a vented air-gap aka "rainscreen" between the structural sheathing and the siding, or an interior side class II or class I vapor retarder.

    While the same code restrictions would apply for cellulose or open cell foam, cellulose has the capacity to buffer the wintertime moisture accumulation in the cellulose itself without damage or loss of performance, and open cell foam is essentially air-tight. Either makes for an assembly that is more resilient to minor air leaks at the gypsum or the exterior sheathing.

    If closed cell spray foam is used (even 1"), the fact that the foam is air-tight and will not wick condensation moisture toward the exterior sheathing, it's protection from winter moisture drives is at least as effective as exterior foam. But it's expensive, and most closed cell has significant greenhouse-gas issues related to the blowing agents used. At thicker than 2" it becomes highly vapor retardent (about as vapor retardent as kraft facers on batts), limiting the drying rate toward the interior.

    In general if an assembly can be designed such that it has decent drying rates without running into seasonal moisture accumulation in materials on the colder side it becomes much more resiliant to moisture related issues (mold, rot, etc). Blown fiberglass at dense-packed densities has a fairly stable R with temperature, and slows down the rate of moisture transported by air leaks, but it's not as protective as other insulation materials. At 1lbs density it's no better or worse than perfectly installed standard-density batts.

    Dense-packed to 1.8lbs+ the newer fiberglass blowing wools it have a slight center-cavity edge over cellulose or open cell foam, but when the thermal bridging of the framing is factored in that advantage is only about R0.5 for 2x framing, R0.7 for 2x6 framing- not enough to make it worth paying extra for compared to o.c. foam or dense-packed cellulose. It's slightly easier to dense-pack than cellulose since it has a more uniform consistency, so some installers will prefer it, but performance-wise it's basically a wash.

    Sealing the studs to the sheathing with highly flexible caulking such as acoustic sealants IS a good approach, and need not be part of a proprietary "system" as presented by Knauf Ecoseal. But if you're serious about air sealing the sheathing as a primary air barrier you'll put a bead under the stud-plates/floor/subfloor and between doubled-up top plates of the studwall too, since those leak points aren't addressed by simply sealing the the interior of the stud bay. As a DIY, get a powered large sized caulking gun that takes the larger-sized caulking tubes, and use acoustic sealant type caulk (Tremco, etc.)

    If dense-packing behind netting, spell out in the contract that it's to be rolled flat to the studs after installation, or it it won't be a very flat interior wall due to bowing.

  3. user-1050854 | | #3

    The apparent bias against fiber glass is based on install mistakes detailed above. Another bias is the source. FG is heavily advertised by the manufacturers. Cellulose is not. As an audit only home energy auditor I specify cellulose in my work scopes. About 1/2 go with Fiberglass, 40% with the cellulose and 10% with SPF. Generally, the SPF and cellulose projects perform better than those with FG. That is usually due to problems that crop up with install techniques. Those can be fixed after the problem is identified, but why not fix them first.

    You are considering Eco Seal - great! BIB fiberglass is a good option for you if installed to the correct density. I would still stay away from the batts. The sealing product will help what ever insulation you choose. I would recommend that you get a price on the FG @ 1.8 pounds per cubic foot and cellulose at 3.5 pounds per cubic foot. Then you can make your choice.

  4. user-1140531 | | #4

    I frequently hear objections to the performance of fiberglass batts that are stated as though they apply to the material itself when they actually apply to defects in installation of the product. Indeed, I would call that a bias. I see references that stipulate that batts require air barriers on six sides, and then criticizes batts for being air permeable when installed without the air barriers.

    One of the objections to fiberglass batts is that they suffer a performance degradation at lower temperatures due to the internal convection. However, my technical understanding is that the performance of fiberglass is accurately stated in the manufacture’s R-value, which takes into account the internal convection. This objection to batts suffering a performance degradation is often not stated with enough clarity to convey whether the batts are failing to meet the manufacture’s claim of performance; or whether they are simply not performing as well as an insulation with a higher R-value.

    Why is it considered an advantage for cellulose to be capable of buffering moisture content that it absorbs compared to batts, which do not have that capability, when the code calls for details that prevent water vapor from entering the insulation cavity in the first place?

  5. jklingel | | #5

    "Why is it considered an advantage for cellulose to be capable of buffering moisture content that it absorbs compared to batts, which do not have that capability, when the code calls for details that prevent water vapor from entering the insulation cavity in the first place?" I think that is because we do not live in a perfect world, and some vapor inevitably sneaks by. Vapor can also be sometimes driven from the outside; solar vapor drive. Either way, if the cellulose has a safety "what-if" valve, why not use it?

  6. GBA Editor
    Allison A. Bailes III, PhD | | #6

    Patrick, let me try to amplify what Martin said and simplify what Dana said. Fiberglass is not held in high regard by many green builders and building science geeks because it's often installed poorly and is contained in building assemblies without properly installed control layers.

    There's nothing wrong with fiberglass if you put it in completely enclosed wall assemblies and fill the cavities completely. As Martin says, it's easier to do that with blown fiberglass than fiberglass batts, but I've seen batts installed well, too. Martin and I have both written articles here at GBA about proper installation techniques (Martin's: & mine:, so you can use those as a guide. Of course, if you follow the manufacturer's instructions, you should do just fine, too.

    Controlling the flow of heat, air, and moisture is your objective in building, and you can do that effectively with a wide array of products and materials. You just have to understand which ones to use and how. See my article on being a control freak for more on that:

  7. user-1140531 | | #7


    Regarding your comment:

    "I think that is because we do not live in a perfect world, and some vapor inevitably sneaks by. Vapor can also be sometimes driven from the outside; solar vapor drive. Either way, if the cellulose has a safety "what-if" valve, why not use it?"

    I assume that moisture buffering refers to the ability of insulation to hold water within the insulation material like a blotter, which would prevent the puddling of condensed water in a way that could soak framing or other structural details.

    I am not sure if I follow the reasoning in concluding that cellulose is better than batts in this regard. Both materials hold water. If cellulose can be said to be capable of holding moisture within the insulation material itself, why cannot the same be said of fiberglass? I have experienced fiberglass batt that had become wet from wintertime condensation of outward moving vapor. When I squeezed it, it was like squeezing a wet sponge, and water ran down my arm. So, clearly, fiberglass batt has the ability to hold onto condensed water.

    Perhaps cellulose amounts to a finer sponge that could hold more water than fiberglass. I don’t know if that is the case, but if it is, is that really an advantage? How much water holding capacity is really needed?

    Furthermore, assuming that cellulose can hold more water than fiberglass, it would seem that the tighter, blotter-like mass of cellulose would tend to dry slower than fiberglass, which has freer air movement. So wouldn’t that mean that the slower-drying cellulose would sustain a longer spell of dampness within the insulation cavity? Wouldn’t that prolonged dampness encourage mold growth?

    And why wouldn’t the absorbed water in either type of insulation reduce its R-value? To absorb and hold condensed vapor, the insulation is holding liquid water within its air spaces, thus displacing the air that is required to function as insulation.

    I understand the attraction of a benefit that would partially backup the performance of an air or vapor barrier. However, this moisture buffering benefit seems weak, especially if you already have cold-side ventilation as a total backup for the warm-side vapor barrier performance. Furthermore, the moisture buffering seems capable of introducing new problems that would not exist without it. Therefore, overall, I would not consider it a reason to prefer cellulose over fiberglass.

  8. wjrobinson | | #8

    Ron, use this sites search feature for many of your questions.

    Try searching: cellulose Robert Riversong

    You can read up on "cellulose verses fiberglass batts" here and on the internet... You will learn about things like; hydrophobic and hydrophilic.

    Cellulose handles moisture better in many cases. It does not lose R value like fiberglass in many situations. Search and you will find. Riversong hugely loves cellulose and more natural materials.

    Unfortunately for me, Cellulose contractors cost as much near me as spray foam.

    Cellulose loose layed in attic floors is highly prefered around GBA.

    In the end, if you like whatever insulation, use it. Every product can work or fail. I have seen the most rot failures when rigid foam was used. I have not remodeled a spray foam home yet. The last home I built has by far the best energy performance. Open cell Icynene spray foamed with a conditioned attic.

  9. user-1140531 | | #9


    I have read about Robert Riversong’s ideas. But I am making specific comments here in response to specific comments by others here. Since we are focused on the details right here, it makes sense to discuss it and break it down right here.

    Often I hear people comment that any insulation product works fine if you use it correctly. And yet I hear a constant criticism of fiberglass batts as being a poor choice for insulation. Which way is it? Obviously the original poster has heard this too, as indicated by the title he chose, “Why is there such a bias against fiberglass?”

  10. GBA Editor
    Martin Holladay | | #10

    Q. "Often I hear people comment that any insulation product works fine if you use it correctly. And yet I hear a constant criticism of fiberglass batts as being a poor choice for insulation. Which way is it?"

    A. Both statements are correct. The solution to your apparent contradiction is simple: fiberglass batts are almost never installed in a way that corresponds to the manufacturer's instructions and best-practice recommendations.

    If a homeowner asks GBA for advice, we're not going to recommend a method that is so difficult that most practitioners get it wrong. We're more likely to recommend a method with a higher success rate.

    Researchers have confirmed the statements I have just made by a variety of means. One interesting study was one conducted by Bruce Harley several years ago. He looked at the blower-door test results for hundreds of Energy Star homes in New England. The homes insulated with fiberglass batts were by far the leakiest when compared to homes insulated with other types of insulation (mostly cellulose or spray foam). All of these homes were built by Energy Star builders aiming for good blower door results.

  11. user-1140531 | | #11


    I understand you point about criticizing fiberglass for being difficult to install. However, it often seems that criticism is directed at the fiberglass product itself when it is actually based upon an installation error or defect.

    Consider your statement quoted from your previous post:

    “Researchers have confirmed the statements I have just made by a variety of means. One interesting study was one conducted by Bruce Harley several years ago. He looked at the blower-door test results for hundreds of Energy Star homes in New England. The homes insulated with fiberglass batts were by far the leakiest when compared to homes insulated with other types of insulation (mostly cellulose or spray foam). All of these homes were built by Energy Star builders aiming for good blower door results.”

    Clearly the implication is that the leakiness is the fault of the fiberglass. Anybody reading that who is interested in energy efficient, airtight construction is going to conclude that fiberglass is the poorest choice of insulation for meeting the objective of airtightness.

    And yet you have advised in other references that airflow must be contolled with an air barrier, and in the case of fiberglass, it should have an air barrier on the interior and exterior side (or on six sides, as is often said).

    I assume that the air barrier that is required with fiberglass would be sufficient to satisfy the blower door test. If so, then how can you conclude that a higher leakage rate associated with fiberglass-insluated houses is not the fault of missing or inadequate air barriers? How can the problem possibly have anything to do with the fiberglass itself when it is understood at the outset that fiberglass is air-permeable and therefore requires an air barrier?

    When you conclude by saying that all of the homes were built by Energy Star builders aiming for good blower door tests, it certainly sounds like they knew what they were doing, and yet they got poor results. The one and only obvious implication is that the fiberglass let them down depite their expert work.

  12. GBA Editor
    Martin Holladay | | #12

    I think that your last paragraph is a good summary of the situation.

    No air sealing method is perfect. Seams that have been sealed with tape, gaskets, or caulk still leak air. (That's why we use blower doors -- so we can measure the air leakage rate.) Getting a good blower-door result is a function of close attention to a hundred details.

    It's much harder to get a good blower-door result in a house with fiberglass batt insulation than it is in a house with dense-packed cellulose insulation or spray foam insulation.

  13. wjrobinson | | #13

    Perfect batt use, starts to cost close to open cell. Perfect cellulose costs for me the same as open cell. Batts and rigid foam used together starts to cost as much as spray foam, with more moisture hazards and mire air exchanges.

    Superinsulating and building tight is not a cake walk. Lots of ideas, and lots of failures. Very important to control moisture and build fail safe. Materials that can handle some moisture cost more and perform better.

  14. BobHr | | #14


    No air barrier is 100% perfect because of that I am convinced that there is a primary and a secondary air barrier. Dense pack insulation acts as a secondary air barrier. I reviewed a report of blower doors tests at various stages of construction. The 1st test were taken when the sheathing was sealed. Another test after dense pack insulation and another after it was completed . Each step showed a reduction in air leakage.

    When Joe L was talking about his study of SIP failures he said he preferred a solution that had less to do with the installer doing things perfectly and more to do with a fail safe solution. Fiberglass batts fall in to the installer category. Dense pack would fall into the fail safe solution.

    I have read that with conventional framing up to 40% of the stud bays are not full width. Combine that with wiring or other obstructions. that makes it very difficult to do a good job when using bats.

    In reading publications at Building science you will see them say dense pack cellulose is not an air barrier but it comes close. If you look at the 3 methods moisture travels hopefully a builder built a water tight home with provision for water to drain should it get behind the cladding. There should be less chance for moisture to get into a wall when the wall is much tighter. Then what moisture that gets in the wall has a chance to be stored safely till it has a chance to dry.

    At building science you will also read about the change in building materials materials to hold moisture Solid wood products are can cycle more moisture through the year than plywood, osb, drywall etc. Cycling of moisture content through dry and wet season is natural and a good thing.

    When talking of insulation the air permeability of fiberglass bats and low density it is the highest of your insulation choices. Air permeability leads to degradation of insulation value unless its in an air tight enclosure. Wall cavities are not air tight so why not use an insulation that preforms better for the conditions.

    If you are starting new then you should build in the proper fail safes to make walls tight and to allow drying to either the inside or outside.

  15. user-987058 | | #15

    thank you all for your responses. I appreciate all of the responses as they have been very informative.

    So to summarize, it looks like fiberglass blown in at 1.8 (thank you John) and the EcoSeal product to stop the air movement are the correct items to consider. When we build this carrage house we will make sure to complete a blower door and hopefully post the results.

    Is there a number that we should expect to receive on the blower door or is there an area to research what good blower door numbers are?

  16. GBA Editor
    Martin Holladay | | #16
  17. Expert Member
    Dana Dorsett | | #17


    The ASTM C 518 rated R value does NOT accurately state performance of fiberglass at the temperature extremes in many applications, even if it's fairly accurate at delta-Ts of 25-30F or less, provided it has snug fitting air barriers ON BOTH SIDES.

    In US climate zone 4 or lower the spec is "close enough", but in places like zone 6 or colder where you're looking at 70F inside, with an average January temp of 10-15F outside the AVERAGE delta-T is TWICE the ASTM C 518 test condition, and a the heating design temp it's even more.

    And it's enough to make a measurable difference for low-density batts like R11s or R19s, even if perfectly installed. An R19 batt is a pathetic abomination of a product, which doesn't even perform as-rated at ASTM test deltas when installed perfectly to the manufacturer's specs (and this is according to the manufacturers' own data). Compressed to exactly 5.5" in a perfect 2x6 wall-cavity, it's only R18, not R19. But it's tested in an ASTM C 518 test fixture fully de-compressed to 6.25", with the fixture's plates forming air-barriers on both sides, but they're allowed to label it that way.

    R22s are and even worse case of low-density mis-labeling hitting only R19 when installed perfectly and are OUTPERFORMED by high-density R21s (which strangely, perform at R21!)


    High density "cathedral ceiling" batts designed for rafters do a LOT better, even at climate-zone 6 extremes, but still not as well as dense-packed fiberglass.

    By contrast, cellulose is VERY stable in R-value across lower-48 temperature ranges at any density, though in wall cavities it needs to be installed at a density high enough to prevent settling. The required density will vary by wall-stackup and climate, since settling is a function of density and season moisture cycling (the parameters of which have been well documented by Torben Valdbjørn Rasmussen of Aalborg Unversitet in Denmark over the past 2 decades. His papers are available online if you really have the urge to see the real math on it. :-) )

    Low density batts between attic joists don't perform anywhere NEAR the ASTM C 518 labeling even with the test delta-Ts, unless there is a snug topside air barrier, yet it's still code-legal to pretend that they do. But with the higher air retardency of cellulose still meets its R/inch at any legal blown density with or without an air barrier.

    Designing wall stackups to be resiliant without strong vapor barriers is dead-easy for most US climate zones, but the moisture buffering aspect can add considerable margin- it's not just a hypothetical. Building assembly hygric-modeling tools such as WUFI are pretty good (it's now a freebie download-google-it ) and can show the real benefits/consequences of cellulose or poly vapor barriers. It's generally better to design in resiliance to the assembly without class-I vapor retarders when you can, and that's actually pretty easy in most of the US, and contrary to popular belief, vapor barriers are NOT a code requirement for most of the US (or even Canada, but it's tougher to design a code-compliant wall stackup without vapor barriers in Canada.)

    It may seem like I'm a shill for the cellulose industry, but I assure you I'm not. The facts regarding performance are readily researchable online. I'm a fan of dense-packed fiberglass too, but it's usually more expensive.

  18. user-1140531 | | #18


    In my comment in my post #11, responding to your post #10, I said this:

    "When you conclude by saying that all of the homes were built by Energy Star builders aiming for good blower door tests, it certainly sounds like they knew what they were doing, and yet they got poor results. The one and only obvious implication is that the fiberglass let them down despite their expert work."

    Above in your post #12, responding to my above comment, you said this:

    I think that your last paragraph is a good summary of the situation."

    I am not sure what you mean. A summary of what conclusion? My above paragraph means that I think it is illogical to blame the fiberglass product for an air test deficiency resulting from incompetent installation of the air barrier. When I say, “The one and only obvious implication is that the fiberglass let them down despite their expert work,” by that, I am referring to your implication in post #10 above, which I disagree with. In truth, the fiberglass did not let them down. If the installers were creating an air barrier only as good as needed for a less permeable insulation than fiberglass (such as cellulose), then they made a mistake. And being Energy Star builders, they certainly should have known better.

  19. GBA Editor
    Martin Holladay | | #19

    You wrote, "Being Energy Star builders, they should have known better." Fair enough. They probably shouldn't have been choosing fiberglass batts.

    Countless studies have shown that fiberglass batts are extremely difficult to install well. HERS raters know it. Bruce Harley's study confirmed the difficulty -- statistically.

    If you want, you can say that it's the builder's fault -- the builder should have done a better job. But builders don't. The main reason, in my opinion, is that it is extremely difficult. That's why fiberglass batts are almost never installed correctly.

    So that's why I advise builders to choose a type of insulation that is more likely to result in success, and less likely to result in failure.

  20. user-1140531 | | #20


    Thanks for your explanation. I am willing to consider all points on this subject. But generally, I find that most discussion leaves a lot of terms and conditions less than fully pinned down. I will read over your response again, because I am not sure what conclusion to draw from it. If I understand correctly, you and others are saying that the fiberglass insulation industry is defrauding the consumer by misrepresenting the performance of their product. If that is so, I would like to have the case presented in the simplest terms possible, so I could present it to the industry for their response.

    However, maybe you are not saying that the industry is defrauding the consumer, but rather, only taking advantage of loopholes in the performance rating protocol. If that is the case, I have to wonder if the same criticisms might apply to other types of insulation.

    What I am looking for is this:

    What is the R-value for a 14” layer of regular density fiberglass batt?

    From what I gather from your comments, there is simply no specific answer to that question. Instead, the R-value would slide on a scale according to temperature differential. Moreover, it sounds like the answer would vary depending on what batt thicknesses were used to make up the 14” layer.

  21. GBA Editor
    Martin Holladay | | #21

    R-value is defined by law; its definition is incorporated into the Federal R-Value Rule. The regulation requires reported R-values to conform to the results of ASTM test procedures.

    These test procedures are conducted at defined temperatures; the test procedures do not result in R-values that "slide on a scale according to temperature differential." R-value is fixed.

  22. user-1140531 | | #22

    Quote from post #14:


    No air barrier is 100% perfect because of that I am convinced that there is a primary and a secondary air barrier. Dense pack insulation acts as a secondary air barrier. I reviewed a report of blower doors tests at various stages of construction. The 1st test were taken when the sheathing was sealed. Another test after dense pack insulation and another after it was completed . Each step showed a reduction in air leakage."


    I agree with the wisdom of having redundancy in the form of one system backing up another. But I don’t like the idea of compromising the quality of both systems with the expectation that each system will make up for what the other lacks, in that each system is contributing its own share of the work of the total task. My idea of backup is that both systems are executed as perfectly as possible with the intention of each system being able to perform 100% of the task or very close to that portion.

    I keep hearing that you can’t produce a perfect vapor barrier, perfect flashing, perfect air barrier, etc. While that is true, the inability to produce perfection is somewhat of a red herring because true perfection is not necessary. All that is needed is adequate execution, and much of the system failures are miles below adequate. The citing of the inability to make anything perfect strikes me as excuse making for sloppy workmanship.

    I have seen what people can do in creating nightmares with the lackadaisical application tape and adhesive sealants with 6 mil film, and then ignoring subsequent film damage from following operations. You cannot excuse this type of failure by stating that perfection is impossible.

    I don’t like the idea of relying the density of cellulose to back up the air barrier in stopping airflow with the conclusion that neither one can be perfect because perfection is impossible. That seems like an excuse to not make a big enough effort to properly execute the installation of either the barrier or the cellulose. Instead of one system backing up the other, what will result are both systems working together with each one contributing a share of the total performance required. That is not backup.

    It is instead, cooperative performance. What you end up with are both systems working some of the time, one system backing up the failure of the other some of the time, and neither system working some of the time.

    What I would do is execute the air barrier, not perfect, but thoroughly adequate to do the job 100% all of the time. The true backup would be the cold-side ventilation system. And that too would be executed thoroughly adequate. It would be so well done that it could most likely do the job of keeping the wall cavities dry even if the vapor barrier or air barrier were omitted.

    I can see how cellulose could help in passing a blower door test, but the envelope leakage that it measures is not the whole issue. The finer point of air or vapor barriers is to keep moisture out of the insulation, and not just to keep air from leaking out of the envelope. So it seems wrongheaded to let the cellulose act as its own vapor retarder. That is like letting the fox guard the chicken coop by letting him live inside of it.

  23. wjrobinson | | #23

    Ron, your posts and thought process are concise and to me would work well in a courtroom in a legal argument.

    Myself and most likely many experienced builder remodelors and problem solver types, we know what "real world" materials work. How they fail, how well they can be installed,etc.

    You can legally somehow win your arguement. Then what?

    Anyway, really enjoying the banter, and interested in what your next step is Ron.

  24. user-1140531 | | #24

    Thanks AJ,

    I tend to believe that these things need to be taken to the precision of a courtroom argument in order to achieve clarity. Then if we disagree, we can at least agree on the terms of the disagreement. But I discuss these issues solely for the purpose of practical application.

  25. wjrobinson | | #25

    Ron, "But I discuss these issues solely for the purpose of practical application." You just lost the case Ron. Why? Stating your purpose you just summed up your oppositions point. The key word is.... PRACTICAL.

    What all are trying to explain to you is that experience in the field along with lots of scientific attempts at quantifying the good bad and ugly of building enclosures... the key in the end is to build what is practical is the best way to build with the men, the materials, the budgets we have today right now for todays demands right now and in the near future. The building codes are changing for the farther future In fact they have already changed drastically over the last ten years. So much that contractors are not able to meet the codes too often. Just read some of what Carl has been blogging. Any hoot...

    for now,

  26. user-1140531 | | #26


    Why would practicality be my opposition's point? I would not accept any scientific theory unless I made it work in the field, so I could see for myself. What makes you think I am being impractical?

  27. user-1140531 | | #27


    Regarding your information in post #21, "R-value is defined by law; its definition is incorporated into the Federal R-Value Rule. The regulation requires reported R-values to conform to the results of ASTM test procedures."

    That is what I have always understood and believed. But it seems quite common for people to contend that fiberglass batts do not deliver their rated R-value when the temperature differential increases to a certain point. If this is true, how can it be reconciled against the requirement of the Federal R-value rule?

  28. Expert Member
    Dana Dorsett | | #28

    Ron: It's not technically fraud as long as it has that performance at the sampled thickness in an ASTM C 518 test plate within the ASTM C 518 test conditions. The problems arise in the differences in real world construction vs. and ASTM C 518 test fixture, key of which are:

    1: The plates of the test fixture form perfect air-barriers, whereas real world installations are less than perfect, and in the case of open attic installations, (often) completely absent.

    2: Real world temperature ranges and delta-T can be well outside the ASTM C 518 range (and usually ARE at the temperature extremes.

    3: The tested batt thicknesses are always thicker than the depth of a stud-cavity, which leads to an over-rating of R for low-density goods like R19s or R23s even within the ASTM C 518 temperature range.

    The "R-value for a 14” layer of regular density fiberglass batt" is a complete unknown, and highly variable depending on how it's installed "Regular density" has no meaning, since batt (and blown) densities vary widely, and the operating delta-T makes a difference. If there are air-barriers on both sides, perfectly installed and a temperature difference is 30F or less, you can expect performance between ~R35 for barely-legal-density blown goods up to R56 for high-density batts. If there is no air barrier on one side the high-density stuff will still be about R56, but the performance of the low density stuff takes a serious hit, even with an R30 delta-T. How much of a hit depends on which side is open and which is the warm side: cold side up, open at the top the convection losses can cut performance of very low density goods by 20% or more even with a 30F delta-T. Warm side up, open at the bottom with no air currents, it'll still run around R35 since there is no convective drive.

    BTW: Cellulose is not a vapor retarder- it's extremely vapor-open but it IS a pretty good air-retarder at normal installed thicknesses, and in attic apps performs at about it's ASTM C 518 levels even without top-side air barriers. In dense-packed cavity applications it's as-good-as an air barrier from a practical thermal point of view, with nearly the air-retardency of high-expansion open cell foam. But it is still extremely vapor-open to water vapor at any installed density (as is dense-packed fiberglass, which also very air-retardent.)

    ANY sprayed or blown insulation has a near-perfect fit- better than can be achieved with batts even in fairly simple assemblies. When you start adding know holes, wiring & electrical boxes, plumbing etc the detailing required to get performance out of batts increases dramatically, but it doesn't change at all with most blown or sprayed goods.

  29. GBA Editor
    Martin Holladay | | #29

    While the R-value of a certain thickness of insulation is fixed, its thermal performance is not.

    Let's say that you have three pieces of insulation: an R-15 fiberglass batt, an R-15 piece of XPS, and an R-15 piece of polyisocyanurate. At 0°F, the thermal performance of the fiberglass batt and the XPS will be better than their thermal performance at 70°F, while the thermal performance of the polyiso will be worse.

    Although the thermal performance of these three types of insulation changes with temperature -- in different ways, as it turns out -- their R-values remain fixed.

  30. user-1140531 | | #30


    I do not understand what you mean by thermal performance, and it being distinct from R-value. I thought that R-value was an equitable rating system to enable the comparison of the thermal performance of insulation.

    I assume that the thermal performance you mention would translate into cost effectiveness of insulation, so if it varies independently from R-value, inconsistently between insulation types, how is a consumer supposed to distinguish the effectiveness of one insulation over another?

  31. GBA Editor
    Martin Holladay | | #31

    There is no substitute for study. If you think that the best way to choose insulation is by just reading the R-value label on the outside of the package, you're wrong.

    The thermal performance of insulation depends strongly on the skill and care exercised by the installer. And if you choose to install an air-permeable insulation without attention to air sealing, its thermal performance will be dismal.

    At very high temperatures, the thermal performance of many types of insulation is worse than at cold temperatures. If that fact concerns you, you should install more insulation.

    More information here: Understanding R-Value.

  32. Expert Member
    Dana Dorsett | | #32

    Labeled R value in the US is how it performs in an ASTM C 518 test fixture within a specified test range (IIRC one plate is kept at 75F and the heat flux is measured with the other plate varying between 75F +/-30F.) If the installed configuration had air barriers on both sides and the temperature range and delta-Ts stay within the ASTM range, all labeled insulation will perform similar to the abeled performance. The real world temp ranges don't fall within that tidy range though.

    And the R value is not constant across temperature or delta-T. Cellulose has a pretty stable over a wide temperature range, with performance rising slightly at sub-zero temps and big delta-Ts, as does high-density fiberglass, but low-density fiberglass performance falls at big delta-Ts due to convection currents within the fiber glass (sometimes by a LOT), even with proper air-barriers on both sides.

    Polystyrene (EPS & XPS) performance increases pretty dramatically at low temps, but also falls dramatically at high temps. Within the ASTM range 1.5lb density EPS averages about R4.17/inch, but it's not a constant. When the outdoor temp is +10F, indoor temp +70F (+40F average temp within the material) it'll average about R4.5/inch. But when it's +70F indoors and the sun-baked siding is +120F(average temp of 95F) it's average performance is about R3.6/inch. Closed cell spray polyurethane and XPS have derating curves in the same direction, but it isn't as significant as with EPS, and the derating curves vary depending on the blowing agents used.

    Polyisocyanurate has a derating curve in the opposite direction of EPS, but the slope isn't as steep. It performs at ~R6/inch in the ASTM temperature range, but with an interior temp of +70F, and an outdoor temp of -10F (+25F average) its performance is only R5.6/inch. On the higher temp range performance is pretty flat, without significant performance increases under hot roofing or siding, but no big decrease either.

  33. user-1140531 | | #33


    Regarding your post #29:

    My interpretation of it is that there is a discrepancy between the thermal expectation of an R-value rating and the actual thermal performance of the insulation.

    Regarding your post #31:

    Here you seem to be saying that the difference between the R-value rating and the actual thermal performance arises from poor installation.

    You said, “If you think that the best way to choose insulation is by just reading the R-value label on the outside of the package, you're wrong.”

    Do you mean that if you choose the insulation by just reading the R-value, you won’t get the promised R-value if you install it poorly? --OR-- are you saying that there is some other measure of thermal performance that differs from the R-value in the case of perfectly installed insulation? You seem to be making the former point in post #31 and the latter point in post #29.

    From the start of this discussion, I have tried to separate the actual performance of the insulation per se from the issue of insulation performance being degraded by poor installation. Introducing installation quality, proper air barriers, etc. clouds the issue of the actual performance of the insulation per se. Any type of insulation will fail to deliver its rated R-value if it is improperly installed.

    In the blog piece that you linked to post #31, you said:

    “To obtain the best performance from fiberglass insulation, the Energy Star Homes program requires most fiberglass-insulated framing cavities to be enclosed by air barriers on all six sides. While the recommendation is sensible, it’s hard to achieve in the field. If such a six-sided air barrier can be created, however, fiberglass insulation will meet the performance expectation promised on the product’s R-value label.”

    Your last sentence seems to conflict with what you said in post #29. But perhaps, as Dana has described, there is a performance expectation promised by the R-value rating, and another performance that is not promised by the R-value rating, and may conflict with it.

  34. GBA Editor
    Martin Holladay | | #34

    You wrote, "There is a discrepancy between the thermal expectation of an R-value rating and the actual thermal performance of the insulation."

    That is only true for those who haven't studied the issue. Most people know that thermal performance does not depend on R-value alone. R-value is only one metric that affects thermal performance.

    You wrote, "You seem to be saying that the difference between the R-value rating and the actual thermal performance arises from poor installation."

    I did not say there was a difference; you did. Thermal performance is not the same as R-value. Think of it this way: the outdoor temperature is not the same as the weather. It is one element of the weather -- but outdoor temperature will not tell you the wind speed or the amount of precipitation that is falling.

    You asked, "Do you mean that if you choose the insulation by just reading the R-value, you won’t get the promised R-value if you install it poorly? "

    I certainly think that is true.

    You asked, "Are you saying that there is some other measure of thermal performance that differs from the R-value in the case of perfectly installed insulation?"

    Well, yes. Even if insulation is perfectly installed, its thermal performance depends on many aspects other than R-value -- for example, air flow through the completed assembly as well as the outdoor temperature.

    You wrote, "Introducing installation quality, proper air barriers, etc. clouds the issue of the actual performance of the insulation per se." I don't see how you can separate the performance of insulation from installation quality. Installation quality affects performance.

    You wrote, "Any type of insulation will fail to deliver its rated R-value if it is improperly installed." I agree.

  35. user-1140531 | | #35

    If the insulation manufacturers’ claimed R-value does not address all of the real world conditions that affect thermal performance, why doesn’t GBA publish a table of information that shows how all of the real world conditions will be met by the various types of insulations? Why not provide the consumer of insulation all of the thermal performance criteria necessary to choose the product?

  36. GBA Editor
    Martin Holladay | | #36

    Your wish is our command. Here you go: Air Leakage Degrades the Thermal Performance of Walls.

  37. user-1140531 | | #37


    Here is what I am interested in:

    I have often heard people say that fiberglass batts do not deliver the R-value that the manufacturer claims. My understanding of this premise is that it is NOT based on batts installed incorrectly. Therefore, the alleged failing is inherently in the fiberglass product itself, and not related to poor installation workmanship, defective or missing air barriers, etc.

    Dana Dorsett has made this same allegation here, and elaborated by saying, “The R-value for a 14” layer of regular density fiberglass batt is a complete unknown, and highly variable depending on how it's installed.”

    Once again, I cannot interpret this because Dana is linking the problem to a dependence on installation. Dana has also mentioned another problem in that the R-value rating does not encompass all real world conditions. Therefore, that raises questions about the performance of fiberglass in conditions outside of its R-value rating.

    Thus overall, both you and Dana have seemingly implied or clearly said that the rated R-value for batts (when correctly installed) does not completely define the actual thermal performance of batts, and therefore batts cannot be compared to the thermal performance of other types of insulation on R-value alone.

    So, in the spirit of AJ’s courtroom metaphor, I will refer to the overall question of batt performance problems as outlined above as QUESTION EXHIBIT “A.”

    When I suggested that GBA publish a table of information that shows how all of the real world conditions will be met by the various types of insulations (as addressed by QUESTION EXHIBIT “A”), you referred me to your recent article “Air Leakage Degrades the Thermal Performance of Walls.”

    I do appreciate the article and found it to be very interesting, but it does not support the charge against fiberglass batts as embodied in QUESTION EXHIBIT “A.” If anything, your article refutes that charge when you say in the article this:

    First of all, if the walls were sealed and there was no air flow through the walls, all of the R-13 walls behaved the same. As Gertrude Stein might have put it, an R-13 wall is an R-13 wall is an R-13 wall. “When the nominal R-13 walls are sealed and tested, they have the same heat flow, plus or minus 4%,” said Schumacher. “They all perform roughly the same.”

  38. GBA Editor
    Martin Holladay | | #38

    I am not responsible for Dana Dorsett's statements. You quote Dana as saying, “The R-value for a 14-inch layer of regular density fiberglass batt is a complete unknown, and highly variable depending on how it's installed.”

    I disagree with Dana's statement. R-value is defined by law and is fixed. Depending on the density of the product chosen, the R-value of 14 inches of fiberglass is likely to be about R-49 to R-52.

    As I have already written, even though the R-value of an insulation is fixed, its thermal performance will vary widely, depending on how it is installed. This is true of all insulation products.

    As I wrote in my blog, Understanding R-Value, "A leaky wall assembly insulated with fiberglass batts will usually perform worse than a wall assembly insulated with spray foam having the same R-value as the batts. The performance differences are due to spray foam’s ability to reduce air leakage, not to any difference in R-value between the two materials."

    While it is certainly true that fiberglass batts deliver their labeled R-value when installed in a perfectly sealed wall with no airflow through the wall, that type of installation is extraordinarily difficult to achieve in the field, and was only accomplished in the BSC lab with careful attention to detail under very controlled circumstances.

    Talk to a RESNET rater. They will tell you that the perfectly installed fiberglass batt job is a kind of fiction. Manufacturers pretend it exists; builders pretend their homes have it; but blower door tests, thermal imaging tests, and hot box testing confirms that, like the unicorn, the perfectly installed fiberglass batt job is extremely rare -- and may not exist.

  39. user-1140531 | | #39


    Thanks for clarifying your position on this matter. I understand the issues of proper installation of batts. I am working on designs for superinsulated houses from the perspective of them being owner-built. I know I would achieve that elusive, perfect batt job, and am confident that I could guide others to do the same if they were motivated to do a good job because it was for their own house.

    For this approach to building, I prefer fiberglass batts because they are a manufactured and are subject to manufacturing level quality control. And if they are installed by a person who is motivated and knowledgeable enough to do a good job, the quality of installation will also be high. I can tick off the reasons why fiberglass batts are so often poorly installed. To do the job right, you just have to overcome that list of reasons. Some of them concern the building structure itself.

    The alternatives of blown cellulose and especially spray foam are subject to wide swings in quality depending on the skill and expertise of the installer as well as the quality of the product components. If I were building houses on a production basis, and found a cellulose or spay foam installer who proved to do reliable, quality work, I would consider those insulation materials. But for my application, I rule out the use of them. Batts are the only insulation I can install myself and know that the job is 100% effective. Cellulose is a possible second choice option.

    Regarding that 14” thick batt layer I asked Dana about: I come up with R-44.8 nominal for regular density batts.

    Compressing regular density batts is another option. My compression table lists values resulting in a 14.5” thickness; close enough to the 14” cavity design.

    If the 14” cavity were filled with two 9.5”-batts totaling 19”, and compressed to 14.5”, it would result in R-50.

    Compressing two 8” batts yields R-48 for a 14.5” cavity.

    Using high density batts is another option. I would have to price these and carefully analyze the economics before deciding which of these options I would use.

  40. user-1140531 | | #40

    Well gee AJ, if people here would differ with my theory, let the differing begin. Isn’t that what these discussions are for? And what theory are you talking about?

  41. homedesign | | #41

    Martin wrote:"Talk to a RESNET rater. They will tell you that the perfectly installed fiberglass batt job is a kind of fiction."
    High Quality Batt installation is rare...but I don't think it is fair to call it "Fiction"

    Allison Bailes posted a Grade I example done by a volunteer labor force (HforH)

  42. homedesign | | #42

    Daniel Ernst's Installation looks like non-fiction to me

  43. wjrobinson | | #43

    Ron, you are a smart handsome tall man. But yaa have no idea what you are talking about as to insulation. Good luck with the experiment. No sense batting this around anymore at this site as you know what you know and most here would just differ greatly with your entire theory.

    Fun thread to watch develop though. Thanks.

  44. wjrobinson | | #44

    There you go Ron, you have John. Good luck my man.

  45. homedesign | | #45

    Here's another example of Batts done well
    another Habitat for Humanity Project
    It's also a Building America Project
    I assume this was done by Volunteers with some adult supervision

  46. Expert Member
    Dana Dorsett | | #46

    Martin: The labeled R-value per the ASTM-C-518 is the letter of the law.

    But that's the LEGAL definition, not the physical-science definition, and it's the latter than really counts, eh?

    The R value as 1/U determined by measuring the actual heat flux depends on the installation. The as-installed performance of stacked batts is nothing like the ASTM-C-518 test configuration in MOST CASES, since they typically lack top side air barriers altogether, yet code allows it. So without more information on what product and how it's installed, I stand by "it depends" (on both the density and installation.)

    And low density R19s do NOT perform to their labeled value even when installed perfectly in wall cavities- they're closer to R18 due to the smaller than tested compressed thickness (and that's straight from Owens Corning's own test data.) Low density R23s are similarly run only ~R19 when installed in a 2x6 cavity they were nominally designed for.

    And Ron- there is no standard definition for "regular density". With the manufacturer's ASTM test data and published weight & dimensions in hand one could hazard a guess. But without a topside air barrier it's a woefully flawed installation that will underperform significantly in an attic application unless it's the highest density goods.

    Did you know that most R19 batts are the same weight per square foot as R13s? An R19 is a fluffed-up R13- same amount of material, different density, and dramatically different derating under a variety of "as installed" configurations. R13s are pretty good, but R15s are better, but R19s are about as crummy as you can get and chronically under perform in real-world applications.

    Density matters- it offers more forgiveness for the lack of perfection on the rest of it, and very few homes are built with perfect air barriers. But a buddy of mine's full gut rehab on a century old 3-family just tested out at 464cfm/50 last week, insulated with a combination of cellulose, spray foam, and taped rigid foam. (He's in a celebratory mood- that's only 9 square inches of total leakage area, equivalent to a 3.4" diameter hole!) Sure, he might have gotten there with batts, with a HELL of a lot of detailing and caulking on multiple layers, but the fit of sprayed/blown goods in real-world cavities are hard to match, even with the most meticulous of batt installers.

  47. user-1140531 | | #47


    The term, "regular density" is my designation to distinguish from batts labled as high density. I am not sure if the industry has their own term for what I call regular density. I notice that you use the term, "low density."

    You seem to suggest that the R-value of this low density batt product is unknown, but you seem to also be linking this conclusion to the product when installed without the proper air barrier on both sides.

    If we assume air barriers on six sides, why must one rely on hazarding a guess to determine the R-value, as you say?

  48. GBA Editor
    Martin Holladay | | #48

    You wrote, "The labeled R-value per the ASTM-C-518 is the letter of the law. But that's the LEGAL definition, not the physical-science definition."

    I disagree. I think the R-value of an insulation product is simply the R-value (as defined by law). If you want to re-define R-value, you'll get in hot water fast -- especially if you are an insulation manufacturer or a contractor. Inventing your own definition of R-value can (at least in theory) lead to federal charges.

    Any scientist who wants to talk about the thermal performance of an insulation at low temperatures or high temperatures is free to refer to the insulation's U-factor.

  49. Expert Member
    Dana Dorsett | | #49

    Ron: Because you specified neither an ASTM labeled R value nor an actual density to your 14" of fiberglass- any number I'd assign would be merely a WAG. With the density and air-barrier information I could take a better WAG at the as-installed R-value. With all 6 sides air-tight it should pretty much match it's ASTM C 518 performance within the test's temperature & delta-T range, but will deviate from that when outside those ranges.

    Martin: I suppose the split is between "what is the R value of the insulation" vs. what I'd presumed to be really "what is the R-value of the assembly". I was presuming (without sufficient facts) that the assembly would a typical attic rolled-out batt application with no top-side air barrier. But 1/U, it is THE definition, not MY definition, and it's independent of test method. The fact that the construction of an ASTM C 518 test fixture provides the necessary air-barriers (sometimes absent in real-world applications) is a problem, since those air barriers is all-important factor for highly air-permeable fiber insulation to achieve the tested 1/U is a source of real-world mis-calculation of performance. It's quite possible to test the R-value of batt performance without the air barriers using other methods such as ASTM C 236 (guarded hot box.) Those methods which deliver results consistent with ASTM C 518 if the air-barriers are applied, but different (yet realistic) numbers if the batts are missing an air-barrier on one side. This is a real beef that comes up with radiant-barrier vendors, since batts of all densities are installed in attics without benefit of top side air barriers. Building codes use the ASTM C 518 to specify R values, even when the installation methods clearly don't resemble the air-tightness of an ASTM C 518 test fixture, and ASTM C 236 with an air barrier on only one side in different orientations would provide a better indication of the as-installed performance. Even if ASTM C 518 is the legal labeling definition, that test number doesn't provide the 1/U R-value of the material in all legal assemblies, and that's a problem.

  50. GBA Editor
    Martin Holladay | | #50

    The situation is not as simple as your comments imply.

    First of all, testing assemblies with and without air barriers is far more complicated than you make it sound, which is why Building Science Corporation is years over budget, and who know how many tens of thousands of dollars over budget, on their much-ballyhooed attempt to slay R-value. If it were simple, everybody would be doing it.

    Coming up with a fixed definition benefits everybody. It benefits insulation manufacturers, because it is a known and well defined value. It benefits consumers, because they know that manufacturers can't play tricks when advertising R-value.

    The ASTM methods are reasonable and not too expensive to use for testing new materials. They don't tell the whole story, of course, but the methods are well defined and reproducible.

    Can you imagine the confusion if manufacturers were allowed to define R-value according to their own idiosyncratic method, or to report "R-value at 0 degrees F," or "R-value at 130 degrees R," or "R-value without an air barrier at 10 pascal pressure difference"? What a nightmare. No one benefits from this scenario.

    We all need to know a product's R-value. We also need to learn about the importance of air barriers, and to learn why air barriers are particularly important when builders choose to install air-permeable insulation materials. This is simply part of good builder education (and ideally, the education of architects).

    Just because knowing a material's R-value is insufficient information for determining the material's thermal performance doesn't mean that R-value isn't a useful metric.

  51. wjrobinson | | #51

    But, Martin, knowing the actual in use U-value of real world assemblies is The Most Important Knowledge Period.

  52. user-1140531 | | #52

    But isn't one component of a real world assembly the R-value of the insulation?

  53. wjrobinson | | #53

    Ron, r value is the absolute baby first step, as close to least important as is possible. Build and remodel a hundred homes and get back to me. Design and build a Passive House using the software, get certified, prove the numbers with followup energy use studies and come back to this thread my man with more to be concise about.

  54. user-1140531 | | #54


    The main issue in this part of the discussion is about whether fiberglass thermal performance somehow lies outside of its actual R-value rating in a way that makes comparing its R-value to the performance of other types of insulation like comparing apples to oranges. Dana Dorsett is making this point in two ways:

    1) The R-value testing standard does not fully account for real world conditions, and in the area of where the standard does not account for real world conditions, batts perform relatively worse than alternative types of insulation.

    2) Batts are tested at a thickness greater than their installed thickness, therefore, they’re R-value is overrated for the installed thickness. This deception does not exist with other types of insulation.

    Dana has, to some extent, quantified the variance of item #2. I would like to see a table of information that would quantify item #1.

    Dana has also said that the actual R-value of a layer of low density batts mismatches the rated R-value to the extent that the actual R-value cannot be known from the rating. Perhaps that goes back to item #2 above, but I need some clarification. Dana says that one can only guess. I find the rated R-value of low density batts layered to 14” to be R-44.8. Certainly, there is a tolerance factor, so in that sense, the indicated R-value is only nominal. However, isn't the nominal rating good enough to go by?

  55. Expert Member
    Dana Dorsett | | #55

    First, I accept the ASTM tested R ratings even for low density batts as long as it has complete air barriers and is at full loft. Provided there is reasonable air tightness on both sides accept Owens-Corning's data on the R-values of their batts when compressed thicknesses in rafter or stud cavities as well, and in many cases it differs from the labeled R value. See:


    Do you like how the R22 labeled batt only performs at R19 when installed into the cavity for which it was designed? Funny, the R21 performs at R21 in the same cavity- go figure!

    Performance with only partial air barriers is a known-unknown, but HAS been measured by researchers (despite Martin's protestation as to how diffcult that can be- the BSC project was taking it several steps further than what's been done elsewhere, with the introduction of 10 pascals pressure differential, etc.) And the results vary dramatically with the air-retardency of the material.

    Pray tell the method by which you come up with the statement...

    ..." I find the rated R-value of low density batts layered to 14” to be R-44.8"...


    How is that? You "found" that by testing it in your own ASTM C 518 test plate or something?

    I'm not saying it's an unreasonable or incorrect number, only that you haven't explained at all where it comes from, and how that number is substantiated. If you tested it, what were the test conditions & assembly? If somebody else tested it, how?

    And what means "actual R value", in your language? Does "actual R" refer only to legally labeled? Does it come from R=1/U in a specific test configuration & temperature range? Seriously!?

    Going back as far as the 1980s there have been several comparative in-situ testing of houses that compared the heating fuel use and summertime ceiling temperatures, etc, where loose-fill cellulose has been compared to batt or blown fiberglass without topside air barriers at equivalent R per ASTM C 518, and the results are generally pretty shabby for fiberglass. Texas A & M did a great deal of study measuring the temperature gradients through different attic insulations under hot roof decks too, finding that the IR translucency of fiberglass resulted in temperatures higher at 1.5-2" into the insulation than at the top side where it could convection-cool to attic air temps, an issue not found with rock wool or cellulose (but correctable in fiberglass with a radiant barrier.) Surf around the web- there has been a lot of study of these issues. More than most, the performance of fiberglass below some threshold density is sensitive to how it's used in assemblies, particularly when used with only partial air-barriers. And THAT's why there is a bias against it.

    And we'd all like to see that performance table with any arbitrary parameters we can think up, eh? ;-)

    Suffice to say, it ain't what you don't know that gets you into trouble so much as what you know to be true that just ain't so. (Apologies to Mark Twain.) If you rely merely on the labeled R value, you can find the performance significantly sub-par.

    My own personal bias has more to do with batts of any kind, or low density fiber of any kind, not fiberglass in particular. But the ubiquity of low density fiberglass batts, poorly installed makes it an easy target. But you don't have to install them badly, or use bottom of the density-range goods either. In cathedral roof apps where code requires a vent gap between the fiber and the roof deck it's worth the extra money for high-density goods over mid-density goods, and you'd easily see the difference in a infra-red image of the interior surface on a windy day, or with the sun beating down on it. In 2x4 construction I'd take R15s over R13s for similar reasons, but in the same assmbly I'd take 3lb cellulose over the R15 even though it's a lower nominal R, because the fit of blown goods is certain to be tighter, and the performance better-guaranteed. I like blown fiberglass at 1.8lbs density or higher too, since that too is highly air-retardent, and somewhat higher R than cellulose.

    It's good to have options, eh?

  56. user-1140531 | | #56


    My number of R-44.8 comes from Owens Corning. It is their number based on their average R-value of 3.2 per inch. I got the information off of the same spread sheet you linked above. I do not understand why you believe this cannot be accepted, rather than remaining unknown until it is actually tested by a person using their own laboratory apparatus.

    You say that you accept the ASTM ratings for low density batts. Then why don’t you accept R-44.8 for 14” of low density batts? You say that I have not said where I got the number. I will check back on what I said, but I thought I was making it clear that I was relying on the manufacture’s R-value rating. I thought that was obvious because you are the one telling me that I cannot rely on those ratings.

    I used the term “actual R-value” to distinguish between the true R-value and the manufacturer’s claimed R-value simply because you are telling me they are different.

    Regarding your statement, “Do you like how the R22 labeled batt only performs at R19 when installed into the cavity for which it was designed? Funny, the R21 performs at R21 in the same cavity- go figure!”

    Looking at that same spread sheet, I see what you are saying. However, why do you assume that the 6 ¾” batt, rated at R-22, is designed for 5 1/2” cavity? The different batt thicknesses are made to different densities. The 6 ¾” batt, at R-22, is R-3.259 per inch. So that would total R-17.9 for a 5 ½” cavity if filled with batt at that density. But compressing the 6 ¾” batt to 5 ½” increases its density, so it achieves R-19 versus R-17.9 had it not been made denser by compressing it.

    The 5 ½” batt, rated at R-21, is R-3.818 per inch. So it is made to a higher density than the 6 ¾” batt, which is 3.259 per inch.

    Therefore, I do not see any failing of the 6 ¾” batt by comparing it to the 5 ½” batt, as you suggest. They are two different products. If I were filling a 5 ½” cavity, I would use the 5 ½” batt. That way, I would get a higher R-value than compressing the less dense 6 ¾” batt down to 5 ½”.

    It would appear that the 5 ½” batt, and the R-15 and R-13 batts (3 ½” thick) are high density product, presumably at a higher cost per inch, with the significant compression already done in the manufacturing process. The R-15 batt is 4.286 per inch.

  57. user-1140531 | | #57

    This discussion about fiberglass density brings up some interesting points. Batts used to be available in only one density, so density was not mentioned. There was always the option to compress batts to achieve a higher density (although at a higher cost). However, compressing batts is confusing and misunderstood because of the universal admonition to never compress batts. Moreover, the benefit of compressing batts can reverse if they are compressed too much. Furthermore, in the 3 ½” and 5 ½” wall cavities, compression will be relatively high because of the standard thicknesses available. And not only can too much compression lower the cavity R-value, but also, it can interfere with drywall installation, and/or cause drywall bowing after installation.

    The need to consider compressing fiberglass arose with the need to insulate vaulted ceilings where the insulation cavity thickness is limited, and there is a need for relatively higher R-value, plus a ventilation space. So, Owens Corning responded with a table of batt compression values to guide consumers in this treacherous venture of compressing fiberglass. Since then, manufactures have developed higher density batt products. When I asked Johns-Manville for their guidance on compressing fiberglass batts, they advised against it because of the risks of interfering with the drywall. They advised using their high density batts instead of compressing low density batts, saying, “We have done the compressing for you,” (in reference to their high density batts).

    In superinsulated houses with double stud walls and scissors truss ceilings, there is ample room for low density fiberglass batts, so there is no need for compressing batts, or using high density batts to achieve sufficient R-value within a limited cavity depth. However, there is another issue to consider when installing low density batts in a deep cavity such as 14-30”.

    To illustrate, say you had a batt that is 30” thick and wide enough to “friction fit” the sides of the 30” deep cavity. If you force it into the cavity, by the time the far side is flush, the near side will be deeper than flush. This is because the batt will have compressed from being pressed in against its friction fit. As you push it in, the gathering friction will hold back the leading side, thus causing the batt to compress under the push force. So, to get the batt flush on the near side, you have to pull back on the near side to make it flush.

    However, this pull-back is difficult to control because all you can do is pinch the near side surface to get a hold of it. So you have to go over the batt near side and pinch/pull several points over the surface area. And although the near side surface ends up flush, the pull-back will probably leave the batt with varying density. The batt will tend to be bunched up near the far side, and the pullback will leave it “stretched” near the near side. Furthermore, the stretching will not be uniform because it must be accomplished by grabbing the surface here and there and pulling it back.

    I conclude that this basic problem is insurmountable when using low density batts in deep cavities and trying to achieve a 100% effective installation. One possible solution may be to use the high density batts because they will compress less when forced into the cavity against their friction fit. Another solution that does work is to use multiple batts thicknesses of low density product sized to overfill the cavity in total, and then to compress each layer during installation. This increases the density of each batt, resulting in a nearly uniform density throughout the cavity, and does not require any pulling of the near side surface to get it flush. That surface will be made flush as the final batt is pushed in.

  58. wjrobinson | | #58

    Ron my tall handsome fiberglass batt man... your last post will soon be used at Wikipedia for an alternative definition of... absurdity.

    Use are beating the batting battle to smithereens my man. For every idea yaa have you miss another tidbit... ie... scissors trusses... have 3.5" high bottom chords. So... any batt thickness that rises above 3.5" is doing nada for the most part as in nothing. Scissors trusses are difficult to superinsulate and should be avoided if one wants to superinsulate the batty way.

    Any hoot... carry on my man... enjoying the E ticket ride we're on.

  59. homedesign | | #59

    try to ignore the deragatory can't win...

  60. Expert Member
    Dana Dorsett | | #60

    At least he finally told us what it source for the number was, eh? OK, now we know, well, sort-of, he didn't exactly pick which batts he was stacking... R3.2/inch is a reasonable, number for a low-to mid-density batt sure, but it won't hit that performance without air barriers both sides due to convection losses. But R4/inch cathedral ceiling batts will (pretty much) if it has an air barrier on at least one side, since the higher density is air retardent enough to limit the convection loss.

    Batts designed for framing cavities are ALWAYS thicker than the standard lumber depth so that it is compressed when installed. This is by-design, to minimize compression-gaps and voids that would otherwise form thermal bypasses.

    I assume that R22 batts with 6-3/4" loft are designed for 2x6 framing for the simple reason that it states in the spec and in the advertising that they are for use in 2x6 framing.

    High density "cathedral ceiling" batts are designed for a 2" air gap between the roof deck and the insulation, and are far more resistant to compression, but are much closer to standard lumber dimensions for fitting into studwalls without the ventilation gap.

    Blown insulation always fits, no matter how uneven the bays are, no matter how odd-ball shaped the any anomalies such as electrical & plumbing (or even knot-holes and stud-twists) are. As long as it's blown with sufficient density it won't sag or create voids over it's service life. It's just a better way to go.

  61. user-1140531 | | #61


    I have not decided which batt thicknesses I would be stacking for walls layer of 14” and ceiling layer of 30”. I would consider either compressing low density batts or using high density batts. I would analyze several different options and price them. I am just not locked into a specific option at this time. But I did give two specific examples of compressed batts for the 14” cavity in an earlier post.

    You stated this:

    “Blown insulation always fits, no matter how uneven the bays are, no matter how odd-ball shaped the any anomalies such as electrical & plumbing (or even knot-holes and stud-twists) are. As long as it's blown with sufficient density it won't sag or create voids over it's service life. It's just a better way to go.”

    I would avoid uneven bays, oddball shaped bays, electrical and plumbing anomalies, twisted framing, and similar defects. Once I do that, I am not convinced that blown cellulose is a better way to go. You stipulated that it is a better way to go as long as it is installed at sufficient density. I do not want to assume that cellulose installed by someone else is at correct density. So when I add it all up, I see fiberglass as the better way to go for my purpose.

    You keep saying that the fiberglass will fail if not provided with proper air barriers. I don’t see that as a problem because I would provide the proper air barriers.

    The only point I was making about low density batts filling a cavity is that I would compress them. I would not, for instance, put two 12” low density batts into a 24” cavity.

  62. wjrobinson | | #62

    It's simply impossible to build with absolutely no imperfect batt width bays Ron. John and you are meant to be. Enjoying this endless saga. Never give up the ship.

  63. user-1140531 | | #63


    I would have bays here and there that would be narrower than standard batt width, and they would require custom cutting the batt. But I would not just freehand hack the batt with a hand knife by eyeball. Batts have to be accurately cut and fitted just like building a plywood cabinet. I think the world needs a better batt cutting table, and possibley a batt miterbox.

    But the custom cut batts would friction fit the bay with the same interference as the standard unmodified batts fit their bays.

    I expect the batts to fit the bays like a glass of milk fits the glass when filled right to the top edge.

  64. wjrobinson | | #64

    Well Dana, we can lead a horse to water but we can't make em drink. Please continue on Ron.

  65. user-1140531 | | #65

    So here is something I would like to see:

    Does anybody have a table that compares (without any bias) the price of the various types of insulation, so the cost can be compared on an apples-to-apples basis? It should indicate the cost of material-only for a unit of area of a given R-value. For simplicity, make it R-1 per 1 sq. ft.

    In addition, the average installation cost could be included in a second table of information. This too would be the cost per R-1 per 1 sq. ft.

  66. GBA Editor
    Martin Holladay | | #66

    Prices are all over the map, especially for cellulose and spray foam, with wide variations from one geographical area to another.

    In order to fairly compare the cost of spray foam to other types of insulation, you have to include the cost of installation labor -- and this cost is also extremely variable.

    So: get some bids and make your own chart for your particular geographical area.

  67. Expert Member
    Dana Dorsett | | #67

    Ron: In general there's usually some premium to be paid for blown fiber insulation over batts, be it blown fiberglass, rock-wool, cotton or cellulose, but usually worth the premium at any density.

    Dense-packing in netting can get fairly expensive when installed it in rafter bays, and open cell foam is often a cheaper option there.

    Damp sprayed cellulose into open stud bays can sometimes come in cheaper or the same as mid-density batts (R13s, R20s) , and always fits better.

    As a DIY you can take the time & enjoy the batt fitting nip & tuck sessions- it's not that bad if the obstructions are few and the bays are dead-even, but it's something that almost ALWAYS needs some amount of fixing before putting up the sheet rock when working with production builders. It's still worth using the high-density ~R4/inch batts on any exterior wall/roof. (I typically only choose lower density goods for use under radiant floors for zone separation. YMMV.)

    Dense-packing is more labor intensive, and that is reflected in the price. Whether it's worth the premium for dense packing depends on the wall stackup and climate, but for retrofits it's HIGHLY recommended, since it lowers the whole-house air leakage by quite a bit without having to rip open walls, strip siding or gutting the place to be able to achieve that air-tightness.

    I'd never use batts in an attic-floor app- open blown cellulose is both cheap and more effective by a number of measures.

  68. user-1140531 | | #68


    Generally, I am contemplating house designs with walls thick enough to achieve superinsulated standards without the use of foam. I lean toward fiberglass for all the reasons I have stated, but also consider the use of cellulose or blown fiberglass. As I mentioned, I am not looking at this from the viewpoint of a production builder. I have built one superinsulated house, and may build another one. I might also communicate my ideas to other owner/builders.

    What I want to avoid is having a contractor install the insulation, and then be less than 100% certain of the internal details of quality. In a superinsulated house, because there is so much riding on the insulation, I don't want any question about a compromise in performance due to quality issues.

    After discussing this here, I am inclined to give more consideration to high density batts. Previously, I leaned toward low density batts compressed. I have picked the taget R-values and thickness of the insulation, but I think a lot of careful consideration needs to go into the exact selection of batt thickness and density. Once you commit to double studs and scissors trusses, there is no need to jump to the highest R-value batts to maximize total R-value because it can be just as easy to increase the framing thickness. But cost and performance of the various batt combinations needs to be carefully studied, especially considering that the layers will be so thick as to permit a large number of possible configurations of the batt layer.

    In the past, I have used fiberglass board for an oven that I built. As you know, it has an R-value comparable to the foam board. I have always wondered if a variation of that product will one day be introduced as a building insulation. It seems like the industry might be being pushed in that direction by the developments of higher performance insulation systems.

  69. Expert Member
    Dana Dorsett | | #69

    There are rigid high density rock wool panel insulation available- usually used in commercial apps. They run about R4/inch and IIRC usually come with facers on both sides (=built-in air-barriers, at least on per-panel basis.) I'm not sure you'll like the $/R price though.

    Believe it or not, most cellulose installers actually know what they're doing, and provide better than DIY service. Sure, there are fly-by-nighters who "fluff" open blows, but if you use installers with a decade or more of experience & references you'll have filtered out most of them.

  70. user-1140531 | | #70


    I don't doubt that there are competent cellulose installers. I would not install cellulose myself, so I would not compare my DIY to the professional installers. But I would with batts. I am not that familiar with blown cellulose, so I am not sure how density is controlled. My understanding is that if the density is too low, the material will not yield the intended R-value, and may settle over time. I assume that the true post-installation density can be known by comparing the total cavity volume to product volume. If this is true, how can the fly-by-nighters you mention get away with a "fluffed" installation?

    I have also seen demonstrations where BIBS installations are checked for density after complete by cutting out a section and weighing it. In my mind, this raises the question of what you do if it is under-weight.

    Are there blowing machines that can weigh the product continuously as it is being blown so that the density could be correlated with realtime progess of filling a bay?

    If you have bays that are open to each other so the filling of one bay would spill into the adjoining bays, how would you control the density in that type of installation?

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |