Smart Vapor Retarders for Walls and Roofs

musingsheader image

Smart Vapor Retarders for Walls and Roofs

Can a smart vapor retarder be used to make an otherwise risky assembly safe?

Posted on Jan 29 2016 by Martin Holladay

During the winter, when indoor air is usually warm and humid, most wall sheathingMaterial, usually plywood or oriented strand board (OSB), but sometimes wooden boards, installed on the exterior of wall studs, rafters, or roof trusses; siding or roofing installed on the sheathing—sometimes over strapping to create a rainscreen. is cold. Under these conditions, we really don’t want water vapor to move from the interior of our homes toward the exterior. That’s why builders in the 1980s installed polyethylene on the interior side of walls.

During the summer, on the other hand, outdoor air can be warm and humid, while our drywall is often cooled by the air conditioning system. Under these conditions, we want to limit the movement of water vapor from the exterior toward the interior. We also want to allow any moisture in our walls to be able to move toward the interior of our homes, unimpeded by a vapor barrier, so that a damp wall assembly can dry out. That’s why an interior-side vapor barrier works against us during the summer.

There are two possible solutions to this dilemma. The first solution is to install an adequate thickness of rigid foam on the exterior side of the wall sheathing. This foam layer prevents inward vapor drive during the summer, while also keeping the wall sheathing warm enough during to winter to avoid condensation or moisture accumulation.

In some types of wall and roof assemblies, it may be appropriate to consider a second solution: installing a “smart” vapor retarder — that is, a membrane with variable vapor permeance — on the interior side of the wall assembly.

The membrane’s vapor permeance rises and falls

When conditions are dry, a smart vapor retarder is relatively vapor-tight (in other words, it has a relatively low vapor permeance). When the air or the building materials adjacent to a smart vapor retarder get more humid, however, the membrane becomes more vapor-open — in other words, its vapor permeance increases. Under dry conditions, it acts like a vapor retarder; while under humid conditions, it opens up and allows moisture to pass through it.

In the U.S., the best-known smart vapor retarders are CertainTeed’s MemBrain and two Pro Clima products: Intello Plus and DB+.

According to CertainTeed, MemBrain has a vapor permeance equal to or less than 1 perm when dry, and 10 perms when damp.

According to Pro Clima, Intello Plus has a vapor permeance of 0.17 perm when dry and 13 perms when damp. In other words, Intello Plus is a more effective vapor barrier when dry than MemBrain, and is more vapor-open when damp than MemBrain.

According to Pro Clima, DB+ (a smart vapor retarder that is less expensive than Intello Plus) has a vapor permeance of 0.8 perm when it is dry and 5.5 perms when damp. That means that the vapor permeance range of DB+ is not as great as that of Intello Plus.

Lots of building materials have variable permeance

If you want to install a building material with variable vapor permeance, you don’t have to buy a smart vapor retarder. Many common building materials — including the kraft facing on fiberglass batts, asphalt felt, vapor-retarder paint, plywood, and OSB — also have variable vapor permeance. All of these materials restrict vapor flow when dry, but become more vapor-open when damp. (However, their range of vapor permeance may not match the permeance range of smart vapor retarders.)

Smart vapor retarders aren’t magic. In a recent article in the Journal of Light Construction, Ted Cushman addressed common misunderstandings and exaggerations that accompany the marketing of smart retarders. “In the field, you may hear salesman as well as contractors offering a whole grab bag of unscientific theories,” Cushman wrote. “Contrary to what you may hear, this class of vapor barriers is not a one-way gate that lets vapor pass through in only one direction. Vapor diffusion through a smart membrane, like vapor diffusion in still air, moves from more humid to less humid. … Products on the market today also don’t have ‘active vapor transport.’ That term refers to materials that can move vapor through the material against the direction of the vapor drive — what you might call ‘uphill’ — when voltage is applied to the material.”

The 475 approach

Unsurprisingly, the U.S. distributor of Pro Clima products, 475 High Performance Building Supply, heavily promotes the use of smart vapor retarders. The company recently posted a blog, “Insulating Unvented Roof Assemblies,” advocating the use of smart retarders to create a roof assembly that violates most building codes.

According to the two founders of 475, Ken Levenson and Floris Keverling Buisman, a roof assembly that is ordinarily considered risky — namely, an unvented roof assembly insulated with cellulose or fiberglass — can be rendered safe by installing a smart vapor retarder on the interior side of the assembly. Levenson and Buisman justify their recommendation with WUFI modeling. (WUFI is a software program that predicts the moisture content of different building components in wall assemblies and roof assemblies.)

Building codes require a ventilated air space between the top of air-permeable insulation materials and the underside of the roof sheathing for a reason: without the ventilated air space, moisture from warm interior air can accumulate on the cold roof sheathing during the winter. Because most roofing products are vapor-impermeable, damp roof sheathing can’t dry to the exterior in the way that damp wall sheathing can.

There are at least two possible problems with 475’s advice on unvented roof assemblies:

  • On its face, the assembly promoted by 475 violates building codes, and can only be used if the local code official certifies compliance based on evidence presented by the architect, contractor, or owner of the building. Obtaining code approval for this type of roof assembly is likely to be both time-consuming and uncertain.
  • This apparently risky roof assembly is justified on the basis of computer modeling and limited monitoring data, not long field experience. Many building scientists are doubtful that WUFI, the software used by 475 to justify their recommended roof assembly, provides results that are robust enough to justify using this unvented assembly without qualms. (For more information on this topic, see WUFI Is Driving Me Crazy and Hygrothermal Software Sometimes Yields False Results.)

Talking to the experts

In hopes of gaining perspective on 475’s advice, I decided to contact three engineers with field experience and building science expertise: Joe Lstiburek, John Straube, and Marc Rosenbaum. All three experts agreed that the roof assembly recommended by 475 is risky.

Joseph Lstiburek. Lstiburek, a principal at the Building Science Corporation in Westford, Massachusetts, is familiar with 475’s advice. (Lstiburek told me that the owners of 475 have challenged him in the past. “I don’t know why 475 keeps taking a run at me,” Lstiburek said. “They should stop kicking me.”) When I asked Lstiburek about 475’s recommendations on unvented roof assemblies insulated with cellulose, he said, “I think that it is risky and I wouldn’t recommend it. If they added a vapor diffusion vent at the ridge, I would have no problem with that assembly.” (A vapor diffusion vent is a new type of ridge vent invented by Lstiburek. It is an opening in the roof sheathing near the ridge that is covered with housewrap that is taped to the roof sheathing on all sides. This type of vent is airtight but vapor-permeable. For more information on vapor diffusion vents, see Can Unvented Roof Assemblies Be Insulated With Fiberglass?)

John Straube. According to Straube, a professor of Building EnvelopeExterior components of a house that provide protection from colder (and warmer) outdoor temperatures and precipitation; includes the house foundation, framed exterior walls, roof or ceiling, and insulation, and air sealing materials. Science at the University of Waterloo in Ontario, the 475 approach “does nothing to solve the potential for excess cold weather air leakage condensation. The smart vapor retarder (SVR) won’t help if you have a pencil-sized hole in the SVR and an outward leakage path (to exterior or interior) somewhere higher up the rafter bay. The past success of dense-packed cellulose is based largely on low interior winter RH values, and with the airtightness and smaller size houses we see today, interior RH is often too high. Will a SVR perform better than sheet poly or just painted GWB? Yes, and significantly better. Is it a low-risk? Not in my opinion.”

Marc Rosenbaum. Rosenbaum is a well-known energy consultant and the director of engineering at South Mountain Company in Chilmark, Massachusetts. When I asked Rosenbaum about the 475 approach, he said, “I think that those assemblies are risky. Given what we’ve learned, I wouldn’t do it.”

Examples of failed roofs

Rosenbaum’s advice is based on field observations of failures.

Rosenbaum told me, “In a Cape Cod house, I’ve seen dense-packed rafter bays where the cellulose has settled enough to see into the bays from the attic, with an air gap of about 3/4 inch in a 10-inch-deep rafter cavity. On the north side of the house, I could see water on the underside of the roof sheathing. This was in a house that was fairly airtight — about 1 ach50 — after a bunch of air sealing work was completed. Where did the moisture come from? Was this diffusion? Well, the drywall had vapor retarder paint, and the vapor retarder paint is close in performance to a smart vapor retarder. Somehow air was moving in all of those bays. The south side of the roof dried out, while the north side didn’t dry out. What about cellulose’s ability to redistribute moisture? That redistribution is actually limited. If there is a leak, it’s just wet near the leak. There is limit to that ‘capillarity magic.’ ”

Can we trust WUFI?

In my 2014 article on WUFI, I wrote, “Here’s my advice to architects: in general, be very wary of WUFI simulations. Although a handful of engineering companies in the U.S. probably have enough experience to provide useful WUFI results, it can be very difficult for an architect to separate valid WUFI runs from poppycock and horsefeathers.”

When it comes to using WUFI, building professionals tend to cluster into two groups. In one group are the professionals with limited job-site experience. Many of these are WUFI-loving architects. In the second group are the professionals with extensive job-site experience. Many of these are engineers, and they tend to be WUFI-skeptics.

“I would always be suspicious of people using WUFI to analyze a roof they have not measured, especially when the analysis ignores air leakage (which we know is the primary cause of unvented roof condensation failure),” Straube told me. “This is an example of a wonderful tool perfectly analyzing the wrong problem. Alas, we see this far too often.”

Less risky assemblies

The experts I spoke with explained that one dependable way to prevent damp roof sheathing is to install rigid foam insulation on the exterior side of the sheathing.

Straube noted that 475 customers could “make an unvented roof with a smart vapor retarder work by using some exterior insulation over the sheathing and an exterior air barrierBuilding assembly components that work as a system to restrict air flow through the building envelope. Air barriers may or may not act as a vapor barrier. The air barrier can be on the exterior, the interior of the assembly, or both. on the sheathing. This gets the airtightness levels up reliably and reduces the condensation risk dramatically by warming the sheathing. Or they can use rockwool outside and cellulose inside (since I know they have an irrational fear of both using tried-and-tested North American methods and foam).”

What are smart vapor retarders good for?

Of course, the fact that smart vapor retarders become more vapor-open during high-humidity events isn’t always desirable, and there are many problems that smart vapor retarders can’t solve. When asked about smart vapor retarders, Lstiburek noted, “These new products work predictably. But you have to control the interior relative humidity for these products to work. If you build a tight house, and if the interior moisture levels rise, then you open the smart valve, and the moisture passes through the membrane. But if you keep the interior relative humidity down to between 25% and 35% in the winter, it is a good technology.”

In some parts of North America, local code inspectors still insist on interior polyethylene — even though interior polyethylene can cause problems when a house is air conditioned during the summer. In these areas, it may make more sense to install a smart vapor retarder than to argue with the code official. In Cushman’s JLC article, John Straube is quoted as saying that in this type of installation, “the smart material really solved a building official problem, not a building science problem.”

Rosenbaum suggested one good application for smart retarders. He said, “A smart vapor retarder for a double stud wall is a good idea.”

Martin Holladay’s previous blog: “Passive Air Inlets Usually Don’t Work.”

Click here to follow Martin Holladay on Twitter.

Tags: , , , , , , , , ,

Image Credits:

  1. Photo #1: Alex Wilson
  2. Photo #2: Oliver Klein

Jan 29, 2016 12:59 PM ET

Great quote
by Charlie Sullivan

“This is an example of a wonderful tool perfectly analyzing the wrong problem."

One can learn a lot playing with WUFI, but there's also a lot one can't learn playing with WUFI.

Jan 29, 2016 8:42 PM ET

Vapor Permeance Range

Martin - Great article. The line," That means that the vapor permeance range of DB+ is greater than that of Membrain…" isn't quite right ,if your numbers are correct. You list DB+ at 0.8-5.5 and Membrain at 1.0-10, making Membrain have a slightly greater range than DB+.

Jan 30, 2016 5:56 AM ET

Response to Kevin Zorski
by Martin Holladay

Thanks for catching my error. I have corrected the text.

Jan 30, 2016 8:03 AM ET

Thanks for having a close
by william murray

Thanks for having a close look at this. The solution still seems to be a combination of a wall that dries to the outside as well as an interior wall that allows for a certain amount of humidity in it , but evenly, and when it needs to . I can't help but feel that polyethylene makes us feel like were living in a bag!

Jan 30, 2016 9:19 AM ET

Response to William Murray
by Martin Holladay

I'm not sure that I understand what you mean. But here's my take: From a moisture management perspective, it's hard to improve on a 2x6 wall with an adequate thickness of rigid foam on the exterior side of the wall sheathing.

Everything on the exterior side of the rigid foam dries to the exterior.

Everything on the interior side of the rigid foam dries to the interior.

The rigid foam prevents inward solar vapor drive during the summer.

The rigid foam keeps the wall sheathing warm enough during the winter to prevent condensation or moisture accumulation.

According to researchers who have monitored the moisture content of wall sheathing in a variety of wall types, this wall stays dryer than any other wall.

Jan 30, 2016 10:35 AM ET

If readers post questions here...
by Martin Holladay

I'll be on vacation until Saturday February 6, so I won't be answering questions between now and then. Feel free to post comments and questions here, though -- I'll address questions when I return.

Jan 30, 2016 10:36 AM ET

In content and tone,
by Ken Levenson

this is a surprising article. Subjects of such articles should be given the opportunity to comment - but we were not. We look forward providing clarifying information on our blog at We also look forward to seeing everyone at Better Buildings by Design this week in Burlington. Please stop by our 475 booth and say hello.

Jan 30, 2016 10:48 AM ET

Response to Ken Levenson
by Martin Holladay

Green Building Advisor is happy to provide you a forum for comment. You are most welcome to comment on this page, and provide any clarifying information you'd like. We welcome a dialog.

Jan 30, 2016 11:28 AM ET

More data on vapor barriers.
by Bill Dietze

I found that following interesting:
It shows the polyamide vapor barriers (a.k.a. Membrain) opening up to 40 perms in very high humidity environments.

Jan 31, 2016 1:02 PM ET

by Lucy Foxworth

I really think this article is an argument for redundancy - rather than relying on one building component to serve the sole air barrier function, you have a redundant system as a back-up. In other words, you tape the sheathing, tape your weather resistant barrier, and you have an interior air barrier as well. I went crazy in the house we built and made sure that each stud bay was sealed from the others as well.

I think a secondary air barrier on top of the roof is good insurance if you are filling the ceiling rafters with dense pack cellulose. That happened to be the only place we used foam in the house - 2 layers of taped and staggered 1.5 inch polyiso (upstate SC - zone 3).

Feb 2, 2016 10:30 PM ET

Edited Feb 10, 2016 4:09 PM ET.

475's Response
by Ken Levenson

475's response is now posted on our website We ask that corrections be made to this article as our response warrants. Also the response is here (pls go to our website to get all the links):

A regular platform of building science debate is the GreenBuildingAdvisor (GBA) website. We’ve been contributors to GBA and engaged in debate there – always with a seriousness of purpose. So it was with great surprise and frustration that we read Smart Vapor Retarders for Walls and Roofs on GBA. No questions were asked of 475 in preparing the article. The article is full of strawman arguments, unsubstantiated claims, gratuitous personalization and bait & switch tactics.


We have real-world experience and we have data. The Whitchurch Passive House Cottage, in Middlesex Vermont, built by Chris Miksic of Montpelier Construction with Passive House consulting by Indigo Ruth Davis and consulting by Bill Hulstrunk from National Fiber, for owner Greg Whitchurch, is one example of an unvented roof with INTELLO Plus inboard. The project was well suited up with data monitors during construction in 2014 – blogged by 475 here – and we’ve been getting updates. Greg’s readings from Feb 1st 2016, show moisture content at the outer roof sheathing of 8.1%.

Kohta Ueno, a Senior Associate at Building Science Corporation, did a presentation with Chris Corson of Ecocor, at NESEA BE15 – on the exactly the assembly the GBA article is focused on – with real world data: see slides 31-39 of the presentation here (pdf).

The 475 worldview is not divided between computer simulation and actual experience. 475 is committed to using all the tools at our disposal to lift the quality of building, through: writing and trainings, drawings, computer modeling, real-world experience and data collection.

The Vapor Curve

The vapor curve matters. Contrary to what the article would have readers believe, while many materials are vapor permeable the quality of the permeability curve is quite different from material to material. This difference matters. Vapor retarder paint is not close in performance to a smart vapor retarder. MemBrain has a less robust vapor curve than DB+, and a much less robust curve compared to INTELLO Plus.

We’ve careful laid out some of these differences in our blog post Why The Vapor Curve Matters. When Joe Lstiburek is quoted in the GBA article, that the limits of 25% to 35% interior RH in relation to smart membrane effectiveness, he is mistaken. INTELLO Plus works effectively at interior RH levels of 50%. Read the INTELLO Primer.


GBA never asked to see the 475 WUFI Pro models. GBA doesn’t know what’s in the WUFI models. Yet the article has further expert opinion telling us what’s in and not in our WUFI models. Strange.

“The 475 Approach”

475 always takes a whole systems approach to making high performance assemblies. While every component is important, how they work together is even more important. As insulation levels grow and expectations of comfort and efficiency rise, assembly drying capacity and reserves are increasingly critical. Experience is a tool. WUFI is a tool. Detailing continuity and training are tools. Blower doors are tools. We use all the tools at our disposal.

Smart vapor retarder, especially INTELLO Plus, can significantly increase drying reserves of assemblies. But contrary to what the GBA article would have readers believe, the addition of INTELLO Plus is not the end of the answer, rather, it is the start. The details must be worked through and the job site execution done properly. Consequently a pencil hole will not endanger the system.

We’ve gained our customer’s trust because we are focused on their specific requirements and solutions. 475 is committed to helping build the highest quality, highest performance, most robust and environmentally sustainable buildings. We’re proud to be part of teams executing some of the most exciting and ground-breaking work across the US and Canada.

Pushing and Push Back

475 recommends vented assemblies. See our blog post A high performance roof should be vented – how to do this properly.

The genesis for developing a safe (foam free) unvented roof came not from marketing research but in working with professionals on flat roof Passive House projects of high insulation levels, where vented roofs would not be effective. (See two in-depth 475 blog posts on this subject here, and here.) This is the case of the Whitchurch project in Vermont.

The other impetus for thinking about making a safe (foam free) unvented roof, was in hearing of home retrofits dominated by the application of toxic spray foam, where installers were looking for a safer viable alternative – like this leading insulation contractor in Massachusetts.

We don’t push products. We do push for more predictably energy efficient buildings, and a less toxic environment. Our North Stars are Passive House and Greenpeace. Both are where some of the the most interesting and meaningful work is getting done on the planet.

Is It About Foam?

We think the spray foam industry’s dominance in construction is a clear victory of chemical giant marketing over good science. (See our blog series Foam Fails.) And we understand foam isn’t going away – some of our products are foam. Our motto is Less is Best.


475 has big ambitions to help move the enormous US construction industry toward truly sustainable high performance. 475 is happy to challenge convention where we think it falls short, and engage in informed debate. Let’s make better buildings.

Feb 7, 2016 7:33 AM ET

Edited Feb 7, 2016 8:20 AM ET.

Response to Ken Levenson
by Martin Holladay

I'm grateful that you took the time to post your comments on GBA. This is an important dialog.

I'd like to start by emphasizing our points of agreement:

1. We both agree that monitoring building components -- especially, monitoring the moisture content of sheathing -- is a good idea. When builders and building scientists install monitoring equipment and share their data, everybody gains knowledge.

2. We agree that it's good that "475 recommends vented assemblies." That's an excellent recommendation.

We do have a significant disagreement, however. The disagreement concerns whether 475's recommendations for a "a safe (foam free) unvented roof" are, in fact, safe. Whether one calls these unvented assemblies safe or risky depends on judgment and a builder's appetite for risk.

Hoping to get a perspective on this question from three of the smartest experts in the industry, I interviewed Joseph Lstiburek, John Straube, and Marc Rosenbaum. Their verdict was unanimous: all three called unvented roofs insulated with cellulose between the rafters "risky."

I stand by my story. (After the article was published, John Straube sent me an email in which he called my story "very well done." Marc Rosenbaum also sent an email, noting, "nice job.")

That said, every builder has their own appetite for risk. I know for a fact that some owners and builders have been convinced by 475's arguments and have built this type of unvented roof. That's OK. Everyone's appetite for risk is different.

Feb 9, 2016 12:13 AM ET

Edited Feb 11, 2016 10:00 PM ET.

Clarification on Chris Corson's Roof
by Kohta Ueno

This comment is superseded by information below; left here for reference.

Kohta Ueno, a Senior Associate at Building Science Corporation, did a presentation with Chris Corson of Ecocor, at NESEA BE15 – on the exactly the assembly the GBA article is focused on – with real world data: see slides 31-39 of the presentation here (pdf).

I just want to make sure we're not talking past each other on this discussion, by making sure everybody understood all the layers of Chris Corson's roof (the one pointed out with low MCs). See the figure below, but to break it down, from exterior to interior:

  • Galvalume Roofing
  • 2x4 24" o.c. (laterally across roof)
  • 1x3 strapping 24" o.c. (vertically up roof
  • Proclima Solitex Plus (38 perms)
  • 16" TJIs @ 24" o.c., with dense pack cellulose
  • OSB sheathing (air barrier)
  • 2x10 24" o.c., with dense pack cellulose
  • 1x3 interior strapping (laterally across 2x10s)
  • Interior GWB

This is, no doubt, a seriously beautiful, moisture safe roof with boatloads of outward drying. It relies on a system of battens and counterbattens to create a ventilated cavity. Ventilation through this cavity bypasses the low permeance roofing materials (metal roofing in this case). I was not surprised that Chris had consistently nice and low moisture contents--the small amount of moisture that got into the system (air leakage or diffusion through the OSB air barrier) could zip right out through the 38 perm exterior membrane.

However, this isn't how the construction industry is building unvented roofs, as much as we would like them to. Instead, the assemblies typically involve an exterior made of asphalt shingles, roofing felt/paper, and structural sheathing (OSB or plywood)--a "sandwich" that has essentially no outward drying. This is why the folks cited above consider relying on an interior membrane (in combination with fibrous insulation) to be a risky strategy.

This is part of the reason for BSC's research on diffusion vents--assuming this limited outward drying roof assembly, the diffusion vent will allow some minimal vapor diffusion drying at the ridge (where moisture accumulates). Doing this in combination with interior air and vapor control is probably a very workable solution--we are pushing research on this now. But for now--here is information on the work to date.

BSI-088: Venting Vapor

BA-1409: Field Testing Unvented Roofs with Asphalt Shingles in Cold and Hot-Humid Climates

BA-1511: Field Testing of an Unvented Roof with Fibrous Insulation, Tiles, and Vapor Diffusion Venting

2016-02-04 Corson Roof.jpg

Feb 9, 2016 6:58 AM ET

Edited Feb 9, 2016 8:53 AM ET.

Response to Kohta Ueno
by Martin Holladay

Thanks for your comments. The details you have provided on Chris Corson's roof clearly demonstrate that this is a code-compliant vented roof -- an assembly that includes the code-mandated ventilation gap above the top of the insulation layer. It's no surprise that this roof assembly has no moisture problems -- and it seems disingenuous to me that Ken Levenson is using this roof as an example of "exactly the assembly the GBA article is focused on."

Contrary to Ken's assertion, I am not concerned with the performance of cellulose-insulated roof assemblies with a vent channel above the insulation. I am concerned about the performance of unvented roof assemblies insulated with cellulose or fiberglass.

The 475 blog has asserted that "while the ProPink Unvented Insulation System may provide a safe roof in zones 2B &3B, if we instead combine dense fiberglass insulation (batts or densepacked) with INTELLO smart vapor retarder and airtight membrane we can make safe and robust unvented roofs in climate zone 6!" [Emphasis added.]

Once the vent channel is eliminated, these roof assemblies become risky.

Feb 9, 2016 7:20 AM ET

Edited Feb 11, 2016 10:18 AM ET.

by Ken Levenson

Kohta and Martin,
See my #17 comment below.

Feb 9, 2016 8:14 AM ET

Response to Ken Levenson
by Martin Holladay

As I wrote before, I stand by my story. The story does not include misstatements.

Feb 10, 2016 8:03 AM ET

Update on Kohta's comment
by Ken Levenson

I was able to speak with Chris Corson yesterday, where Chris clarified that Kohta is actually mistaken about the assembly for which we were citing data - and that we were correct in our description. We will be adding back the comment regarding the presentation per our initial response based on Chris' confirmation.

Kohta, I kindly ask that you also confirm with Chris and update your comment accordingly. Thanks.

Feb 11, 2016 12:13 PM ET

Edited Feb 11, 2016 12:17 PM ET.

Clarifying what's going on here...
by Martin Holladay

Both Kohta Ueno and Chris Corson have promised to post comments here when they get a chance. Until that happens, I'll share what I know. I just got off the phone with Chris Corson. Evidently Kohta's image is an accurate representation of the roof that he monitored. It's a code-compliant vented roof, and it has no moisture problems.

At the conference Ken Levenson mentioned (the 2015 NESEA conference, BuildingEnergy 15), Chris Corson was a co-presenter with Kohta, and Chris reported on data from a different house -- one with an unvented roof assembly. Chris's monitoring shows that his unvented roof assembly is also performing well, with no evidence of moisture problems.

I'm grateful to Kohta and Chris for their monitoring work and their willingness to share data. As I noted in Comment #12, "When builders and building scientists install monitoring equipment and share their data, everybody gains knowledge."

I stand by my assertion that the issue here is a builder's appetite for risk. Unvented roofs insulated with fiberglass or cellulose do not comply with building code requirements, and several building experts -- including John Straube, Joe Lstiburek, and Marc Rosenbaum -- have labeled this approach "risky." It's good to gather and share data on these homes.

Builders are free, of course, to evaluate the data and estimate the risk associated with this type of unvented roof -- and then to choose the specifications for their next roof based on these evaluations and estimates.

Feb 11, 2016 10:33 PM ET

More Information from Chris Corson
by Kohta Ueno

I was able to speak with Chris Corson yesterday, where Chris clarified that Kohta is actually mistaken about the assembly for which we were citing data - and that we were correct in our description. We will be adding back the comment regarding the presentation per our initial response based on Chris' confirmation.

I corresponded with Chris, and he verified the construction of his monitored hot (unvented) roof assembly--he mentioned two assemblies, and provided the image below:

Roof 1:

  • GWB
  • 1x3 battens
  • Intello plus
  • 16 in dense packed I joist
  • 5/8 ZIP
  • EPDM
  • Floating Ipe deck tiles with built in drainage plane

Roof 2:

  • GWB
  • Intello plus
  • 14 " dense packed I joist
  • 3/4 AdvanTech
  • Grace Ice and Water HT
  • Standing seam galvalume roof nailed through the sheathing

He points out that he prefers the vented roof assembly discussed above--but sometimes he gets backed into a corner by circumstances.

The data he logged is shown in the other image. The roof is showing nice low wintertime MCs at the outermost (sheathing) sensor--peaking at 10-14%, or a very safe range.

He explains the circumstances of the ~20% MC spike in the first summer of operation, at the innermost sensor:

It's important to note the the WME spike you see in year one is the product of the windows being plasticed off, the vent being off, and the GWB being mudded in the summertime.

[KU editorializing] and of course the inward solar/temperature gradient--i.e., any stored moisture is blown out of the outer sheathing, and pushed inward.

Just to clarify--I don't think that anyone's saying that unvented roofs with a variable-permeance interior membrane can't work. It's clear that awesome practitioners like Chris... or other PassivHaus folks... or other folks who can pull off 1 ACH 50 regularly are likely to make this type of assembly work, assuming interior RHs don't get out of control. But I'd definitely be concerned releasing this assembly on the general construction market... i.e., the folks who are complaining about that newfangled code-required blower door 7 ACH 50 test. :)

2016-02-11 Corson Roof 01.jpg 2016-02-11 Hot roof.png

Feb 12, 2016 8:47 AM ET

Edited Feb 12, 2016 8:57 AM ET.

Response to Kohta Ueno
by Martin Holladay

Thanks very much for the further details. I agree with your assessment that Chris is an "awesome practitioner." He makes careful density calculations, figuring out the weight of the cellulose and the volume being filled, to ensure that enough cellulose is installed. He is also aware that very deep cavities (as are found in some double-stud walls) require cellulose to be installed at 4 pounds per cubic foot instead of 3 pounds per cubic foot. He often insulates wall and floor assemblies as panels that are manufactured indoors (in a manufacturing facility) before being transported to the job site. Finally, his quality control process includes the use of infrared cameras and opening up assemblies to check whether settling has occurred.

Chris Corson is unusual, and practices that work for Chris probably can't be generalized. Just this week, a homeowner posted a question on GBA, noting that he has had three cellulose insulation contractors visit his house, and only one of the three contractors knew what "dense-packing" is. The contractor who said he could dense pack a cathedral ceiling told the homeowner that flimsy polystyrene ventilation baffles are just fine -- they'll be able to resist the pressure of dense packing. This level of knowledge is, sadly, typical in much of the U.S.

Register for a free account and join the conversation

Get a free account and join the conversation!
Become a GBA PRO!