Why did the roof fail but not the wall?
Hi all, trying to understand why a house that was fully wrapped (roof and walls) with peel and stick vapor barrier WRB (Ice and water shield) and no exterior insulation led to a roof condensation failure but not a wall failure? Climate zone 3B
The roof failed on the north side slope only, but the walls seem unscathed, even the north side walls, at least not in the areas where some drywall was removed for inspection.
Can anyone help explain why the roof was so much more susceptible to failure than the walls?
The roof will be repaired with exterior rigid foam insulation, but siding modifications is just not feasible at this point due to budget, complexity, and timeline.
Is it safe to assume that if the wall sheathing hasn’t failed in the 9 years since construction, then it should be fine for the long term?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
My guess: Walls are OK because of your relatively mild climate. The contractor was stupid, but you got lucky.
Roof sheathing failure might be due to "hygric buoyancy," but I'm not sure -- hygric buoyancy is a somewhat controversial explanation. (Of course, the north-facing roof cools on clear nights due to night sky radiation, and cool surfaces encourage moisture accumulation. The south-facing roof is warmed during the day by sunlight, driving out accumulated moisture.)
Why is hygric buoyancy somewhat controversial?
also for OP here is a 2016 article from BSC about buoyancy and ping pong water, which you see on that assembly with the damage at the peak. I'm not sure where the controversy Martin is suggesting comes from so maybe this is out of date?
https://buildingscience.com/documents/building-science-insights-newsletters/bsi-016-ping-pong-water-and-chemical-engineer
Freyr,
If you want to read more about the controversy I alluded to, see "High Humidity in Unvented Conditioned Attics."
I'll cut and paste the relevant paragraphs below:
Bill Rose, a research architect at the Building Research Council at the University of Illinois, doesn’t buy Lstiburek’s “hygric buoyancy” theory. “Imagine that you have a glass column 30 feet high, with a uniform temperature all around,” Rose told me. “Any amateur physicist should be able to describe the distribution of molecules, all of the molecules that are components of air, once the column of air has stabilized. In a glass column, the humidity ratio difference from the top to the bottom would show up at the fourth decimal place.”
Rose continued, “I don’t like the term ‘buoyancy.’ I think Joe [Lstiburek] really screwed up using that term. It makes the dew-point gradient appear unidirectional. … We know the kinetic behavior of molecules. If we had a glass column as tall as a house, with a uniform temperature all around, then high school physics would perfectly explain the distribution of water molecules in that glass column. There won’t be any stratification by molecular weight. If there is any notion of stratification by molecular weight, then industrial gases would become a whole lot cheaper — because it takes energy to separate gases. There is fairly uniform mixing. That’s kinetic theory. That’s how gas molecules behave. There isn’t going to be a border between the oxygen molecules and the argon molecules, with no movement of molecules. That isn’t going to happen.”
Rose said, “Of course, we don’t have perfectly closed columns in a house. We have temperature differences and sorption. But we had better be very careful that the building science we are describing is consistent with high school physics. If all we do is make measurements and say, ‘Ah ha! I have a dew-point gradient!’ then we haven’t done students any favors.”
When it comes to explaining the phenomenon of “ridge rot,” Rose sees no need to resort to a hygric buoyancy explanation. “I’m really comfortable explaining it in terms of how corners behave,” Rose said. “We get chilling at corners. There is a temperature depression at corners.”
Two days after our telephone discussion, Rose sent me an email to clarify his understanding of moisture buoyancy. “There is a real place for use of the term ‘moisture buoyancy,’” he wrote. “It refers to the buoyant movement of moisture-rich air within a drier air mass. Heck, cumulus clouds … result from moist air masses rising. Of course, the moist air mass diffuses unless it is captured in a balloon, but diffusion can take time and meanwhile it may be quite buoyant. By the Lewis relation, diffusion of moisture in air correlates closely (or perfectly) to the equalization of temperature in air. …
“And in fact, moisture buoyancy may be a component in ridge rot. If sun pumps water out of sheathing, there may be a film rising up the inclined sheathing. All of these examples are predicated on transient conditions of moisture-rich air masses or films, which dissipate to the normal mixed condition. This is where I differ with Joe and Kohta. The final, equilibrium, long-term condition is always the mixed condition, with no stratification.”
I asked Rose, “If these attics are humid for a reason other than hygric buoyancy, what’s the reason? Why is the air in these attics humid?”
“I’m not going to propose an explanation,” Rose said. “My measurements show a difference between basement and my second floor. It would be good for someone to explain these differences.”
Huh, his argument is not very convincing. I agree that gravitation stratification does not really exist, as it requires low pressures to see a stratification of partial pressures. But gas stratification does happen, perhaps based on concentration of the gases. For example, propane leaks will tend to accumulate lower, a helium balloon will rise. So if you have a close to 100% concentration of a particular density gas, ie humid, it will tend to stratify. there are also thermal factors to consider. To say he doesn't buy the explanation because it doesn't work in a completely sealed container where the conditions are not at all similar to a house seems to miss the point. Sure, if you have a glass house and don't introduce any outside heat, or gasses or molecules to the mix, they will tend to mix equally and not concentrate anywhere... that is not exactly what we are talking about.
It is not wrong to describe something as buoyant if it tends to rise in a specific environment, which is what moisture seems to be doing in houses, as shown with ridge rot and his own measurements. Why that happens is another question but a bit beside the practical point and not misleading if it happens.
DamionL,
Explanations aside, the order in which we see damage in buildings with risky assemblies is very often: North face of roof, other faces of roof, north face of walls, other faces of walls.
I hadn't thought of it that way but you're right, Malcolm.
In this case it was just a matter of time. The south-facing roof and the north-facing wall would (or will) show moisture damage soon enough, if no insulation is added.