January 10, 2014

The building science community has long known that interior vapor barriers are unnecessary and even harmful in some assemblies. So why do some codes still require them? In this Conversation, we discuss sections of the Canadian building code and how its requirements are influencing U.S. practices.

January 10, 2014: This issue came up because of the Canadian codes. Two folks called me recently about requiring an engineer-sealed letter with respect to insulating sheathing and spray foam. And I said “you’re kidding, I thought we solved this years ago.” And the answer was that apparently the building science community solved it but it never got memorialized in Part 9 of the Canadian code. So the best-performing walls and best-performing roofs are not available to the average Canadian home builder without having to go to a licensed engineer or a licensed architect. That’s just plain wrong.

January 10, 2014: Can you be more specific? I know that right now the Canadian code is messed up in the way they require vapor retarders. I understand they penalize vapor impermeable insulating sheathings if they are used on the outside, versus vapor permeable insulations. I think that’s the main issue here with the Canadian code interior vapor retarder requirements.

January 10, 2014: Normally, Achilles, because you and I are happily living south of the border, we wouldn’t care, but these really bad ideas are moving south. They’ve crossed the line into Minnesota, and they’re finding their way into Chicago. There’s people saying “my god, you’ve got to have plastic” in Chicago. In fact, I’ve found the plastic disease vector as far south as Texas last week. My god! It’s like the Alamo. We’ve got to stop Santa Anna from Quebec coming down to put plastic in American buildings. We’ve got to draw the line.

January 10, 2014: The clearest thing in the code is, number one, it doesn’t really matter what you do, you need a very low vapor permeance layer on the inside. Period, full stop. That is the big problem. It doesn’t matter if you’re building out of foamed metal panels, to tell you the most absurd situation. For a foamed metal-panel building – like a SIPS, made of metal sheets – they want a vapor barrier. That’s how stupid it is. That’s the big level. Then there is a secondary, lower level debate and concern that says that when you use a vapor impermeable exterior insulated sheathing, you have to use more R-value than if it’s vapor permeable. There’s actually some scientific basis for that. Their numbers are a bit questionable, but at least there’s a basis there. If it’s a foil-faced polyiso, we need to use more R-value on the exterior of wood framing than if it was, say, a mineral fiber insulation on the outside.

January 10, 2014: I agree with you John that it’s a little bit more reasonable on the vapour open vs. vapour closed exterior sheathing methodology, but not that much more. The last little piece that drove me crazy was that if you want to do calculations, they – the “authorities having jurisdiction” want you to pick 4% of the hours in January as your number, as your design temperature. Whatever happened to engineering judgement. At that point, when I heard that, blood shot out of my eyes and my head rotated while green stuff shot out of my mouth. It was almost as bad as when Boston beat Toronto in that 7th game…

January 10, 2014: So, problem number one is a requirement to use a vapor barrier everywhere even if it’s a SIPS, even if it’s an ICF, even if it’s a closed-cell spray foam. The answers that are completely obvious to the industry and to the professional community, the assemblies and materials and products that we know are totally safe because they have built-in vapor tightness, are not acknowledged by the Canadian code.

January 10, 2014: You guys recently did some R&D in the University of Waterloo BEG Hut about this, didn’t you?

January 10, 2014: Yes, this was for the spray foam industry, and I think we very definitively demonstrated that closed-cell spray foam doesn’t require a vapor barrier. We actually did a test to demonstrate that pretty obvious fact – because you know we like to take on all the tough challenges! – and yet it still only helps according to the industry maybe 1 out of 3 times. That’s how often they can show a code official the report and explain that it’s completely reasonable to do this and get listened to. Part of the problem is that the CCMC – the Canadian Construction Materials Centre – wants every individual product to be tested as to whether or not they need a vapor barrier. Even though it’s blatantly obvious, just from inspection, that some of them don’t.

January 10, 2014: I would say that what we have going on at CCMC is a blatant self serving way of funding CCMC. I know I can get away with saying that. What can CCMC do to me here in the United States?

January 10, 2014: They can push poly into Texas! But seriously, is this a problem that people in the States are worried about?

January 10, 2014: Actually, yes. And I blame bad WUFI calculations. Sorry Achilles, I couldn’t resist. When an amateur runs WUFI he or she can fix problems by specifying interior plastic vapour barriers as far south as Texas.

January 10, 2014: Touché Joe. Correctly run WUFI is our ally here, and I know you know that. John, to raise another point, there is concern out there that seems not to be fair. If you have a product that is 0.9 perms versus 1 perms, you are required by the Canadian code to go out and add a bit more insulation and a bit more insulation, and that’s making you less competitive.

January 10, 2014: Right now in the hut we have walls with extruded polystyrene and foil-faced polyiso, next to mineral fiber, and it definitely shows the difference between those products. Even though the polystyrene and the polyiso are seen as the same product, not surprisingly the aluminum foil facing does make it perform quite a bit differently. The aluminum foil-facing is a very low perm product, under 0.1 US Perms, whereas the extruded foam is just under 1 Perm: that factor of ten difference in permeance matters to how the products actually perform. Now, everyone gets that the mineral fiber is quite open, say over 10 Perms, and that’s going to make it act different, but assuming a product with a permeance of 1 acts the same as a product with a permeance of 0.05 is a bit silly.

January 10, 2014: I think what we need to do here is touch on the fact that you have airflow, and in Canada they’re trying to deduce the loading due to a small amount of airflow, and include that in the analysis. And they’re trying to do diffusion analysis, based on incredibly high interior loads.

January 10, 2014: Well, it depends on how high the loads are, but the end result is that in the assessment of exterior foam sheathing they do include airflow, whereas if you just use OSB, they don’t. So it’s a fundamentally unlevel playing field. I would say that Joe’s experience at BSC, my experience, a bunch of other people’s experience, all speaks to the fact that by and large, exterior insulated walls are less risky than walls without insulating sheathing. And that’s not what the code says; in fact it implies quite the opposite.

January 10, 2014: So who’s driving this? Let me offer a hypothesis. This has become a problem because more and more people are using continuous exterior insulation. Until more people started doing it, it wasn’t a problem – the language was there, but nobody cared. It didn’t affect them. The language was there simply because of old, outdated research – it wasn’t put there maliciously, it just ended up there, and now we’re noticing it because it’s starting to affect more of the business.

January 10, 2014: That said, in Ontario we’ve been using foam sheathing for about 20 years, in almost half the houses we build. So around here it might be more the issue of thinking that thicker foam sheathing is less safe, when in fact it’s more safe. But how do you explain the exclusion of closed-cell spray foam? That’s never been in the code. The code never says that the interior vapor retarder has to be a piece of plastic sheeting – it could in theory be plastic foam. But code officials do not accept that.

January 10, 2014: Isn’t it also the minimum ratio of the total thermal resistance outboard of the material to the thermal resistance inboard of the material? Isn’t that ratio a little bit disfigured in the way the national building code is using that in terms of requiring you to use different thicknesses when you have that less than one perm? I mean, I think it’s a myth, and I can claim being part of that myth, because in 1993 or ’95, when we were developing the air barrier guide at NRC – and I was part of the execution in terms of creating weather information – I can tell you that the initial analysis that was done for the national building code did not introduce any solar driven moisture, did not introduce any night sky radiation, did not introduce a lot of the fundamentals – liquid transport was not there, wind-driven rain was not there. I mean, a lot of the stuff that we do today was not even considered at that time. It was plain diffusion and lots of air leakage. All these numbers that we’re talking about were developed at that time. And now we’re in this mess in Canada and it is moving down south. The thing that appalls me – I can’t remember the exact numbers – I think you add 750 Pascals between inside and outside. So in other words, they’re recommending something in the order of 5.5 g/m3, and that is tremendously high. I mean, there is no basis whatsoever for doing something like that, and when you do it you have relative humidities that are 60% for a big chunk of the time in Toronto, I’m talking about over 150 days. And the average winter-temperature RH is 45%. Which, I find that excessive. So that needs to be changed. And then the other thing is that they’re requiring a mechanical pressure on the inside. And if I’m not mistaken, it was something on the order of 10 Pa mechanical pressure, all the time, from the inside to the outside. And I’m sorry, but that’s stupid. Big time stupid.

January 10, 2014: Of course it is. In a cold climate, that’s exactly what you want to avoid.

January 10, 2014: We carried out some work, and we’re more than happy to share that work, on what happens using different vapor resistances and introducing that air leakage component, and look -- we’re seeing that we’re getting the same results no matter what. Especially with high loads, I mean, my goodness.

January 10, 2014: I think that’s an important distinction to be made. Vapor barriers will not protect you in a high-humidity building in cold weather. Because such small levels of air leakage will totally swamp it. So you can’t actually test for high interior humidity loads because it’s going to fail anyway. So it’s really a fundamentally big point you make, Achilles. You can’t protect a house in Toronto, or Minnesota, or Chicago, with 60% RH when it’s 20 degrees outside. It won’t work, and a vapor barrier won’t fix it. If your analysis says it will, then your analysis method is by definition wrong. But I don’t know what the answer is, other than continuing to put on political pressure.