Image Credit: Images #1 and #2: Kohta Ueno - Building Science Corporation Kohta Ueno installed a variety of sensors in several locations in double-stud walls in Devens, Massachusetts. It's safe to assume that the hole in the subfloor was sealed with caulk before the wall was insulated. Researchers from Steven Winter Associates (including Lois Arena) are monitoring the moisture content of OSB sheathing at a Massachusetts house that is only a short distance from the house being monitored by the team from the Buildling Science Corporation.
Image Credit: Image #3: Lois Arena - Building America This graph shows data points from Andy Shapiro's monitoring of the moisture content of board sheathing on a 12-inch-thick double stud wall in Vermont. Rather than using a data logger that provided a continuous stream of measurements, Shapiro visited the site periodically to record data using a Delmhorst meter and a multimeter.
Image Credit: Image #4: Andy Shapiro This graph reports data on sheathing moisture content collected by Kohta Ueno during the first winter of a monitoring study of double-stud walls in Devens, Massachusetts.
Image Credit: Image #5: Building Science Corporation
Most wood-framed buildings have no insulation on the exterior side of the wall sheathing. That means that the wall sheathing gets cold and wet during the winter.
Whether or not this common situation is a problem depends on who you talk to. Monitoring studies show that the moisture content of the OSB or plywood sheathing on some homes with 2×6 walls rises in February (especially on the north side of the house); fortunately, however, the sheathing dries out in March or April. As long as the sheathing stays dry for most of the year, it can usually endure a few weeks at an elevated moisture content without developing mold or rot. In most cases, sheathing won’t rot as long as the wall’s drying rate exceeds its wetting rate on an annual basis.
Not all builders are comfortable with this analysis, however. Some builders prefer to install rigid foam or mineral wool insulation on the exterior side of their wall sheathing, to keep the OSB or plywood above the dew point during the winter. Warm sheathing is dry and happy, so installing an adequate thickness of exterior rigid foam is one possible solution to cold sheathing worries. (The extreme version of this approach is called PERSIST. PERSIST homes put all of the wall insulation on the exterior side of the wall sheathing, and leave the stud bays empty.)
In theory, a very thick wall with lots of insulation on the interior side of the wall sheathing — for example, a double-stud wall — is riskier than an ordinary 2×6 wall, because the high-R insulation reduces heat flow through the wall, making the sheathing colder and wetter than ever. Some hygrothermal modeling programs, including WUFI, show that wall sheathing on a 12-inch-thick double-stud wall insulated with cellulose can have an elevated moisture content under certain conditions. (The sheathing moisture content shown by WUFI varies, of course, depending on the inputs made by the software user; variables include inputs for climate, wall orientation, and the interior relative humidity.)
Cold sheathing warnings
These computer modeling results have made some builders and designers nervous. Here are samples of warnings about cold sheathing on double-stud walls:
Having read these warnings for years, I set out to answer an important question: are these warnings based on actual measurements of the moisture content of installed sheathing, or are they based only on computer modeling?
Modeling studies show worrisome moisture levels
One modeling study that included somewhat worrisome findings, “Moisture Research — Optimizing Wall Assemblies,” was authored by Lois Arena and Pallavi Mantha. Arena and Mantha both work at Steven Winter Associates in Norwalk, Connecticut, where Arena is a senior building systems engineer and Mantha is a building systems analyst. Most of the research conducted by Arena and Mantha is funded by the U.S. Department of Energy, through its Building America program.
Some of the researchers’ findings sound alarming. Arena and Mantha used WUFI to calculate the predicted moisture content of several components of various wall assemblies, including 12-inch-thick cellulose insulated walls. The researchers wrote, “Results from WUFI indicate that condensation potential for the double [stud] cellulose walls is extremely high because the OSB in those wall assemblies is entirely outside of the insulation.”
“Realistically, we know that this is not true”
Delving deeper, I realized that some of the researchers’ warnings rested on a shaky foundation. The authors note, “WUFI offers several different methods for generating interior temperature and RH [relative humidity] levels. For this study, the interior conditions for all three wall types were generated using the ASHRAE 160-2009 method.
“It should be noted that, in all climates, the interior RH levels predicted by this method reach 90% even though cooling was assumed. Using these interior conditions, WUFI predicts that there is the potential for mold growth on the interior surface of the drywall in all climates. Realistically, we know that this is not true.”
What is ASHRAE 160?
ASHRAE 160 is a published standard that provides moisture design loads for those who perform hygrothermal modeling. These design loads are similar in principle to design loads used by mechanical engineers. Among the design loads in the standard are a set of “design indoor conditions” which (for example) establish the amount of moisture that modelers should assume is generated by a building’s occupants.
For some reason, the ASHRAE 160 values used by Arena and Mantha produced unlikely WUFI results. Arena and Mantha wrote, “Further research into the appropriateness of the ASHRAE 160-2009 interior conditions in moist climates is needed. Interior relative humidity levels generated with this method are higher than recorded in actual studies (Arena et al. 2010) and result in overly pessimistic predictions for mold growth on the interior of the assembly. Considering that almost every wall in this study failed the ASHRAE 30-day criteria, it is recommended that this threshold be reevaluated by industry professionals.”
Here’s the translation: the modeling results don’t pass the sniff test. Arena and Mantha clearly recognize that fact, and they accurately deduced that the anomalies stem from their use of the ASHRAE 160 values.
To find out how these unexpected WUFI results may have occurred, I spoke with Anton TenWolde, the building scientist who helped develop ASHRAE 160. TenWolde explained that the indoor moisture values first published in ASHRAE 160 needed to be tweaked. “We already fixed this,” TenWolde told me. “We have passed and published three addenda to the standard since the original publication. The indoor relative humidity is now capped at 70%. The standard is under continuous maintenance. We felt it was the best possible standard at the time of publication, but it was by no means perfect.”
TenWolde pointed out that the developers of the standard lack good data on indoor moisture levels. “When you look for the numbers on moisture generation in residences, you find that they are not there. All of the numbers floating around for the last 40 or 50 years are probably too high. It’s also important to remember that this is a design standard, so conditions are not average conditions. They are extreme conditions, and they are supposed to be extreme. You can’t uses averages for design, because it’s not the average house that fails.”
Since the ASHRAE 160 conditions used in the WUFI modeling performed by Arena and Mantha were flawed, the WUFI results from that study shouldn’t be used to make design decisions. Instead, it’s worth looking at data from monitoring studies of real walls.
Fortunately, we’re beginning to get results from a few field studies of double-stud walls insulated with cellulose.
Monitoring results from field studies
From September 2007 to March 2009, Andy Shapiro, an energy consultant from Montpelier, Vermont, monitored the moisture content of the sheathing on a double-stud wall in Vermont. The walls were 12 inches thick and insulated with cellulose. The house had a single occupant and was equipped with an HRV.
One thing about the house was unusual: it had diagonal board sheathing rather than OSB or plywood sheathing. “The house was sheathed with rough-cut local lumber in random widths — a mix of hemlock, fir, and pine,” Shapiro told me. “The siding was white cedar shiplap over Tyvek DrainWrap, which is very vapor-open. The interior was finished with painted gypsum wallboard with no interior vapor barrier.”
WHAT MOISTURE LEVELS ARE CONCERNING?
Before reviewing the field study data, it’s worth discussing how wet sheathing can get before we should be concerned about the possibility of mold or rot.Mold or fungi won’t grow on wood unless its moisture content is above 20% — a level that corresponds to 80% to 90% RH — so that is the moisture content level that usually sets off alarm bells for researchers (especially if the 20% moisture content level occurs during the summer, when temperatures are warm). Decay won’t set in unless the wood has a moisture content greater than 28% and unless temperatures are above 23°F. (Decay is quite slow when temperatures drop below 50°F.)
The sheathing moisture content was higher during the winter months than Shapiro expected. In February and March of 2009, Shapiro recorded moisture content levels over 30% in two locations. (For perspective on moisture content readings, see the sidebar to the left, “What Moisture Levels Are Concerning?”)
“The data isn’t complete,” Shapiro told me. “These measurements are really snapshots in time. But they are real moisture measurements. I chatted with John Straube about this, and given that this house has a moisture-open exterior, we’re not worried. But if we had the same readings with OSB sheathing, we’d be worried.”
Monitoring data collected by Lois Arena
In addition to co-authoring a paper on the use of WUFI modeling on high-R walls, Lois Arena has been monitoring the performance of some double-stud walls at a residential project in Devens, Massachusetts. Some of Arena’s data were reported in the September 2013 issue of Energy Design Update. Intrigued by the data, I called Arena and asked her to share up-to-date information from the monitoring study.
The monitored wall is an R-40 double-stud wall insulated with cellulose. The OSB-sheathed walls are clad with vinyl siding, and the interior finish is gypsum wallboard painted with a 0.5-perm vapor-retarder primer.
The research team installed temperature, RH, and moisture content sensors in a variety of locations in two stud bays — one on the south side of the house, and one on the north. “The OSB on the north wall is now [October 2013] down to a moisture content under 10%,” Arena told me. “In February, it was up to around 20%, but it came down to 10% in mid-April. The time at 20% was brief, and it doesn’t sound too alarming to me. I like the fact that the walls are drying out.”
According to Arena, one reason that the sheathing dried quickly in the spring was the fact that the cladding (vinyl siding) is well ventilated. I asked her, “Would you agree that builders don’t have to worry about the integrity of the OSB sheathing on a double-stud wall, as long as the wall has a ventilated rainscreen gap?” She answered, “That’s exactly right.”
Kohta Ueno’s data
At a another house in the same residential development in Devens, Massachusetts — at a house built by Carter Scott — a different research team is conducting a similar monitoring study. Kohta Ueno of the Building Science Corporation now has two years of data on several high-R walls, including a 12-inch-thick double-stud wall insulated with cellulose. Like Arena’s research, Ueno’s research is funded by the Building America program.
Some of Ueno’s data were reported in a Fine Homebuilding article called “The Future of Housing in America.” The article noted, “Twelve-in.-thick double-stud walls with cellulose insulation and [walls] with low-density foam were tested. In the stud bay insulated with cellulose, the moisture content of the sheathing on the north side of the house spiked to 28% in the winter. That’s high enough to raise concerns about mold and rot. In the summer, though, it dried out to 8%, and given that cellulose is treated with borates, no one seems too concerned.”
Ueno’s data (like Arena’s data) show that south walls stay dryer than north walls. When I telephoned Ueno, he told me, “The first winter, the cellulose wall peaked at 20 to 25% moisture content. Those are the numbers for the north walls; the south walls were boring. For a few months over the winter the moisture content was high, then everything dried over the summer. Everything in all three monitored walls dropped over the summer to 10 to 12%.”
The moisture content of the OSB sheathing monitored by Ueno is somewhat higher than the values measured by Arena. Ueno told me, “One of the key differences between the walls we’re monitoring and the ones that Lois Arena is monitoring is that there is vapor retarder paint on her wall, while we’re using straight-up latex paint.”
I asked Ueno whether he thought the moisture spikes during the winter were cause for concern. “I don’t think it’s an issue,” he answered. “As Mark Bomberg says, ‘We measure the moisture content of wood during the winter but it rots during the summer.’ Everything was drying down by March. During the second winter of this study, the house was occupied by a family of four, and they didn’t have the ventilation system running. The system was not hooked up correctly. So the interior RH was 40% to 50% during the first half of the winter. Later, after the exhaust fan was fixed, the RH came down to 30% for the second half of the winter. All of the walls on the north side had a high moisture content for a few months. But by the end of September, were back down to 10 to 12%.”
Ueno’s data show that an important factor in moisture accumulation in sheathing on double-stud walls (in addition to the reduced heat flow through the wall assembly compared to walls with less insulation) is wintertime vapor diffusion through the assembly from the interior to the exterior. For more information on this factor, see The Return of the Vapor Diffusion Bogeyman.
John Straube’s data
To complete my roundup of monitoring data, I telephoned John Straube, a professor of building envelope science at the University of Waterloo in Ontario. Straube was happy to discuss his reasearch.
“We’re doing a study now,” Straube told me. “We’ve got a test house here in Waterloo. The climate zone is similar to Burlington, Vermont. The interior humidity is kept between 30 and 40 percent in the wintertime. We’re testing a variety of walls, all with OSB sheathing: a typical 2×6 wall, a 2×6 wall with 3 inches of exterior Roxul, one with 2 1/2 inches of exterior XPS, one with 2 inches of foil-faced polyiso, a 9 1/2-inch-thick I-joist wall, and a 12-inch-thick double-stud wall insulated with cellulose. All the walls have Tyvek, furring strips, and fiber-cement siding. The study is entering its third winter.”
Straube’s measurements are similar to those made by other researchers. “The sheathing moisture content on the double-stud wall exceeds 20% in March, then hits a peak of 31% on the 26th of March,” he told me. “It dips below 20% in the middle of April.”
The walls with insulation on the exterior side of the sheathing are faring better than the double-stud wall. Straube posed a rhetorical question: “Maybe you’re wondering, ‘What is the moisture content of the sheathing on the walls with exterior XPS in February and March?’ Well, it’s 8%.”
One of Straube’s colleagues, Trevor Trainor, provided further information on the researchers’ protocol. “The peak moisture content levels occurred during a period of intentional, controlled air exfiltration,” Trainor explained. “We intentionally injected interior air into the walls at a rate that we determined as a realistic natural air leakage rate for a building that meets Canadian Energy Star standards (0.2 cfm per square foot at 50 Pa). We used a rate of 40 CFH per 4 X 8 wall panel (the equivalent of 0.2 L/sec/m2). Depending on the geometry of the house, this roughly translates to 2.5 ach50.”
Trainor added, “Obviously a house that is significantly tighter than this should perform better — depending on the distribution of the leakage.”
What do these measurements mean?
To understand whether the moisture spikes seen in double-stud wall sheathing are worrisome, it’s useful to compare these measurements with the moisture content of OSB sheathing on ordinary 2×6 walls.
“With an ordinary 2×6 wall, the moisture content of the OSB never gets as high as it does with these double-stud walls,” Straube told me. “On a north wall in a humid climate, it might reach the high teens, but not the low 20s.”
When researchers interpret data from double-stud wall studies, they often frame the conversation using a “good news/bad news” format. “We know that actually, most of our wood products are surprisingly resilient,” said Straube. “But these monitoring results show that we are right on the edge between risky and where we would be safe. These elevated moisture contents are lasting for several weeks to two months, depending on the study, into April. And in April, the sheathing may be at 50 degrees. That is right in the zone where you might get a problem, though it might take years for the problem to manifest itself. It’s in the gray zone.”
Should builders worry? “The monitoring studies say that these double-stud walls are on the edge, not obvious failures,” said Straube. “After all, where are the bodies? The failures happen only when we have a leaky window or an air leak or badly managed construction moisture.”
The message for builders of double-stud walls is simple: don’t screw up the details. “These walls are on the edge,” said Straube. “What that means is you don’t have a lot of room to move when it comes to making mistakes. When I compare a double-stud wall to a wall with 3 inches of exterior rock wool or 2 inches of extruded polystyrene, I find more risk.”
Straube advised, “Build a simple box with ventilated cladding, and do a good job with airtightness. Those are strategies to reduce the risk. Choose a sheathing that is more vapor-permeable than OSB — plywood, fiberboard, or DensGlass Gold.”
Martin Holladay’s previous blog: “Low-Road Buildings Are Homeowner-Friendly.”