Well, we have now had a significant failure. The owner of one straw bale house here (on which I served as a consultant and led the plaster and bale work) has removed his straw bale walls after only 2 years, because of moisture problems. I wasn't part of the deconstruction (this is both merciful and unfortunate) but I have a good idea of what was going on beforehand, and I've spoken to Jim McSweeney, the owner, about what he found. I'll attempt to convey it here. At the end of this email there will also be some questions—I hope you all will have some ideas to share.
To begin at the beginning:
Last February, after a particularly cold spell, I noticed a pattern of wetness at the outside top of the north wall, which appeared as if water had been poured along the top edge of the wall and allowed to run down. Jim removed the blocking at the end of the rafter bays, and found ice on the underside of the sheathing, and also on the underside of the proper vent that was installed at the upper part of the rafter cavity to maintain a ventilated air space. The extreme wetness on the surface of the plaster was caused by water (clearly condensate—the standing seam roof could not be leaking in every bay) running down the proper vent, hitting the blocking at the exterior of the wall, and running down the back side of the blocking into the outer 2" of the bale, and into the plaster. The outer 12-24" of cellulose was also quite wet.
This house is of post and beam construction, with bales wrapped around the structure, and dense-packed cellulose in cathedral ceilings. The walls had approximately 1" of lime stabilized clay plaster, with a lime finish coat, and limewash. Plaster and finishes were the same inside and out, except for the bathroom where a conventional paint was applied to the interior in place of the limewash. (I doubt that it was an especially vapor-retardant paint, though I'm not sure.) The ceiling is 1/2" drywall with latex paint, poly, and dense-packed cellulose. The cellulose ends atop the bale. Interior butt joints between plaster and timber framing were backed with 15 lb felt in an attempt at tightening the joint against air leakage.
In late February, Jim took a set of moisture readings which showed alarmingly high moisture levels in the outer 3-6" of the walls, often in the range of 25-40%. Unfortunately, I don't have any records of these readings; they seem to have been only loosely correlated to depth. I then took a set of readings (in different but nearby holes) in mid-June. By this time there had been considerable drying; while there were many readings still in the high teens and 20s, there were none in the 30s, and all high readings were in the very outer inch or so of the straw just behind the exterior plaster. The inner third of the bale was consistently at 8-12%, and the middle third at 8-15%. Even the outer third of the wall, except for that last inch, ranged from 10-20%. Jim had found that the highest exterior numbers were on the east wall, which is most exposed to rain, and receives some spray from an adjoining roof. This still seemed to be the case in June. I also pulled straw from the drill holes, and found some minor discoloration but no evidence of fiber breakdown.
In October, Jim tore down the walls. The patterns he found speak clearly of interior moisture sources, and also correlate to the long-held idea that areas of lesser density are in greater danger of moisture damage. The places where the straw was in the worst shape were at joints between straw and wooden members, and also at the corners of bales. In some of these areas the fiber breakdown was such that he could crumble the straw in his hand. Some of these areas would be places where air leakage was a possibility—around windows and timbers, for instance. (The wettest point was clearly one of these—a large joint between bale and timber that was imperfectly sealed with spray foam because it was impossible to plaster this area. Some plywood in the outer section of this wall was saturated and delaminating, and carpenter ants had moved into the structural timber at the inside of the bale wall.) But direct air leakage could not have been possible for depositing moisture in all of the problem places—many of the wood/straw joints and essentially all of the bale/bale joints would be buried in the field of the wall, with a continuous skin of plaster over each face. This seems to corroborate the theory that internal convection loops—or some pattern of air motion—are causing water to be deposited in the outer section of these less dense areas. By contrast, the main body of each bale (or at least the great majority of bales) was relatively intact, with only a thin layer of damage at the very outer surface. Additionally, the wall which received more wetting from rain did not show a greater degree of decomposition than the others.
For some time now, I have been musing on the possibility that the plaster system that we use—equally permeable on the two sides of the wall—may not be suitable for this climate. I know of three of our projects that show the annual damp patches in the upper areas of the exterior walls. I have been taking occasional moisture readings in one (buried wood block sensors) and these seem to indicate a pattern of increasing moisture levels through the wall section, from inside to outside, during the cold season. Of course, there are also many houses that show no visual (or olfactory!) signs of moisture accumulation. Only one is anything like the extreme of Jim McSweeney's house—but this was pretty clearly a different situation, a hilltop site where a combination of a failed finish coat and severe windblown rain caused water to enter the bales at the corners of the building. (Interestingly, in this house, it was only the areas of 30+% mc that showed any damaged straw; the 20+ areas felt damp to the touch but were bright and intact.)
I think the McSweeneys' house showed such severe signs so quickly because of two factors—
1—Higher than usual (for SB) moisture production within the house. The McSweeneys had not been venting showers, were drying laundry in the house(including diapers) with no ventilation, and had a pot of water on the woodstove. The McSweeneys reasonably felt that their practices were safe, because 3 humidity gauges in the house were reading consistently in the area of 40%. But the extra moisture had to be going somewhere. It seems that it was exiting in large amounts through air leakage at the tops of the walls, at the plaster/timber intersection. But the generally high moisture levels in the field of the wall also indicate that it was working through the plaster by diffusion, and that the resulting condensation at the back of the plaster was not drying fast enough to prevent accumulation. The fact that other houses are showing damp patches in late winter indicate that this is not an isolated case, even if it is extreme.
2—Weather. The summer in which this house was built (2003) was very wet and not very hot. The base coat of plaster took 2 weeks to stiffen, and weeks passed before the surface appeared dry on any wall other than the south. Finish coats were applied in late summer (exterior, in the rain) and in early winter (interior.) The owners then left for much of the winter of 2003/2004, leaving the building unheated. During the summer of 2004, the red oxide colored exterior limewash developed a splotchy character with whitish surface deposits. In retrospect it seems reasonable to think that this was caused by moisture migrating out through the walls. For a long time I have been saying that "plaster adds a totally unacceptable amount of water to bale walls." This is mostly a way of cautioning people against applying the interior finish coat too quickly, before the bales have a chance to dry off the water from the base coat. In this case, the walls may never have dried completely.
So now the big question—what to do? Jim McSweeney is understandably convinced that straw bale construction can't work in this climate. I tried on that idea, but it doesn't seem right. What is clear is that a system of equal permeability rates on both sides of the wall is not the best choice for colder and wetter climates. It appears to be working on lots of houses—and it's definitely working on at least two that have sensors installed—but it is just as clearly not working on the houses with damp patches. (From Jim McSweeney's experience I feel it is safe to conclude that seasonal damp patches are not OK—they are almost definitely causing some degree of deterioration of the straw.) The difference may have to do with usage patterns. But if straw bale houses are supposed to last indefinitely (250+ years, by the standards around here) we cannot reasonably expect every future family to be obsessive about a dry indoor environment.
I hate to admit this, but I will—for some time now, I have been thinking that with a clay/lime/ limewash or silicate paint exterior system, it would be wise to use a commercial vapor-retardant paint on the interior to reduce the permeability of the interior wall surface. But the problem is that no painted bale wall ever has the same timeless feeling as a limewashed bale wall—and this feeling is a big piece of what makes straw bale houses such wonderful spaces. So I've been resisting this idea. But what other realistic options are there? We aren't going to come up with a way of dramatically increasing the permeability of the exterior. Increasing the thickness of the interior plaster is a nice idea for a whole set of reasons, but applying ~1.5", as we currently do, is already a lot of work. Trying to double or triple that amount (would this dramatically change the permeability?—anybody know?) seems wholly impractical for any building larger than a shed.
Interestingly, we usually do paint kitchens and bathrooms because of the unusually high moisture production in the these rooms, and also for cleanability. In the two houses where I have taken readings, I have not found appreciably lower moisture readings in the walls behind these painted bathroom surfaces. I have assumed that this is because the exterior bathroom walls are typically pretty small, and so the moisture level will tend to equalize with that of adjacent wall areas. Also, I don't believe that either of these bathrooms is painted with a particularly aggressive vapor retarder. Any ideas, here?
I understand that there is also a danger in dramatically reducing the permeability of the interior surface—in the case of major water damage (roof leaks, etc) the ability to dry to the interior can be very important to the health of the walls. But we know, from good quality conventional construction, that "less permeable inside, more permeable outside" is what works for everyday conditions. Why should bales be any different? And aren't everyday conditions ultimately more of a driving concern than individual events?
Thanks for reading through this, and for any thoughts that you may have. I recall musing, at the time of Danny Buck's repair project, that failures mean straw bale construction is finally coming of age. Now I feel like I've aged a few years, as well—though not nearly so many as Jim McSweeney, who is paying quite a lot to have his walls rebuilt with studs and cellulose.
Thanks again for any thoughts you may provide.
All the best,
Paul
--
Paul M. Lacinski
Sidehill Farm
GreenSpace Collaborative
Mail: PO Box 107
Packages: 137 Beldingville Rd.
Ashfield, MA 01330 USA
+1 413 628 3800
I'm a bit confused by an assessment of the sources of the moisture. In the first part of your description, you describe an external source of an apparent good deal of water coming from condensation around the top of the roof: "The extreme wetness on the surface of the plaster was caused by water (clearly condensate—the standing seam roof could not be leaking in every bay) running down the proper vent, hitting the blocking at the exterior of the wall, and running down the back side of the blocking into the outer 2" of the bale, and into the plaster."
But then, after tearing open the walls, you remarked that: "The patterns he found speak clearly of interior moisture sources, and also correlate to the long-held idea that areas of lesser density are in greater danger of moisture damage."
It's clear that the bales were subjected to lots of moisture. First, during the wet summer of plastering, when apparently drying of the considerable moisture introduced through plastering was very slow; then during the following winter, when the house was unheated, and therefore unable to dry out; and finally because of the roof vent condition.
So I wonder whether what you are taking to be evidence of "interior moisture sources" isn't just an observation of the pattern of moisture migration (and collection) as the moisture laden walls (attempted to) dry out.
The relatively humid interior conditions, in this case, wouldn't be the cause of the excessive moisture but would have resulted in a sluggish drying process which, given the amount of water the bales had stored during construction and the winter, and the input of water from the roof, was insufficient to dry the bales before damage occurred.
When bales are moisture laden, the moisture will transport and migrate all around the walls, in and out, up and down, depending on 'climate,' even on a daily basis. So perhaps you are seeing evidence of the dominant pattern of moisture migration in the walls, but not of the introduction of moisture from the interior.
This would indicate that the lack of barrier on the interior is actually working to dry the wall: The gradation of humidity from dry at the interior to moist at the outside may be evidence that the humidity/temperature/pressure differential between inside and out is actually working to drive moisture out of the walls, but was insufficient to do the job before damage occurred.
But then again, I haven't seen the walls... what do you think?
Thanks for your thoughts. What you are saying about the stored water from construction as the primary source may well be correct. As I mentioned, I'm of the firm opinion that plastering adds too much moisture to the walls. But what are we to do about it? Not much that I've thought of, other than providing excellent ventilation (and heat if necessary) and hoping for good drying weather. Also—avoiding the application of interior finish coats until the bales (checked with a moisture meter!) have dried out from the deluge of the base coat.
I would certainly agree that any vapor retardant layer on the interior shouldn't be applied until the bales have dried fully. Once again, this would require testing—something that we (and, I suspect, 95% of bale builders) have been less than rigorous about.
Still, I see this house an extreme example of what seems to be a common pattern—the appearance of the damp patches. The houses that get them just seem to get them every year, both during and after the initial drying-out period. I don't yet have enough experience to correlate them to a source—is there a pattern of occupant behavior? Or some particular construction detail? I don't yet know how common they are, though I'm sure that the majority of houses don't get them; otherwise I would have heard from a lot more than 3 people, by now. I do know that they always appear at the upper part of the wall, and are most common right at the top of the bale wall, where the bale meets the roof insulation. They also appear in the same locations every year.
Though I think you understand, I'd like to clarify a bit on the water from the roof. From a design point of view, it's nothing to do with an external source. Whether the moisture was from construction or from internal production, it clearly made its way out the top of the bale, where it then condensed on the roof, and ran back down as liquid. Early on I was hoping that the source might be the roof ventilation air—we have an issue here with condensation from the ventilation air on the underside of metal roofs. But this is not a February phenomenon; it happens in spring and fall when days are warm and relatively humid, and nights are cold. It's like dew, but upside down; and it happens quickly because the metal roofs cool so quickly. You can get ice at night, and rain inside the under-construction building in the morning. The first time I experienced this, I was asleep on a stack of that first-generation cotton insulation that didn't loft enough to insulate well, but made a nice site mattress. The sun was up before me, and as soon as it hit the roof I got pretty wet. It's always nice to have a freakout in the morning, before even brushing your teeth.
I remember hearing Tim Owen-Kennedy talking about venting the upper part of bale walls to the outside to prevent moisture accumulation in the roof. At the time, I wasn't sure it made sense; it seemed to me that by leaving openings in the exterior plaster, you would be increasing the rate of air leakage from the interior of the building through the wall, and thus potentially increasing the rate of deposition. But I think I might be convinced.
So what if these questions were to be separated? First, we have a rather obvious (in retrospect) need to make sure the walls can dry off construction moisture, as soon as possible. But once that has happened, in a cold climate, should we be seriously (I still don't like that word) considering a vapor retarder on the inside, maybe in conjunction with Tim's vents?
Thanks again,
Paul
From: Rene Dalmeijer
Paul,
Thanks very much for your frank and open report on the McSweeney house. I see some parallels to a house I have just helped complete. Since plastering the exterior with a hydraulic plaster I have monitored the moisture level. Specifically because we noticed a damp spot on the exterior on the spot most exposed to weather that only went away slowly after removing the scaffolding. The spot dried in about 2 weeks; it is not visible anymore. After doing some hard thinking I realized it was caused by splash from water dripping from a roof drain on the scaffold and hitting the wall where the wet spot was only noticed after the scaffold was removed.
Now back to the parallel with the McSweeney house. The wet spot prompted me to regularly monitor the bales around the house. We only just plastered a few of the interior walls with earth plaster so most places were and are easily accessible from the inside. Instead of the moisture levels dropping, what I expected, They have been rising—not much, but up from 14% to 16%, and only on the very outside 30-50mm. The wet spot is slowly drying... albeit very slowly. The house is not occupied yet and is well ventilated.
Your mail has warned me though to keep a careful watch on the house. As I subscribe to your idea that the Scandinavian model of having less permeable finishes on the inside and more open finishes on the exterior is good practice in the Dutch type of climate. I would love to do a dynamic moisture transport simulation to better understand what is happening. Sven Eweleit of Andersehen might be able to help.
Rene
From: Paul Lacinski
Hi Rene,
We have also had the splash problem from the scaffolding—it's very easy at the end of a long day to convince yourself that the planks are just fine where they are, even if they are under the drip line of the roof. I'm not so surprised to hear that the moisture level is rising in that outer zone of your walls; if the plaster is still wet, the straw may still be absorbing moisture from it. My limited experience has been that that outer 30-50mm dries very slowly, and always experiences the most active cycling, from rain or from condensation.
On two projects we tried Tom Rijven's dip method, but it was so messy and labor-intensive, and made the bales so heavy and the floor so slippery, that we quit. But I wonder whether it isn't worth dipping the exterior face of the bales—because if that outer 30-50mm were impregnated with clay, I suspect that it would be a lot more durable. (Of course it now seems that you would want to let it dry before plastering—we did not do this before.)
I can tell you what I'd really like—I'd like to be able to buy bales that have been pre-impregnated with clay to a depth of ~75mm on all sides, then dried. The core would retain its full insulating value, while the exterior would become more moisture tolerant. Plus, the strings would probably not matter any more; you would likely be able to cut the bale open and lift out as long a section as you would like, without the usual pop. And if the ends were nicely squared off, we wouldn't have the funky corner intersections to deal with. It really wouldn't be hard—a very large greenhouse, some conveyor belts, a forklift, an efficient way of making and moving slip, and someone could be in the business of producing some beautiful eco-blocks. Maybe even in various, standard sizes.
I'd love to see a simulation—if you decide to look into it let me know if I can pass on any information. In your project, is the earth the finish material on the interior? And if so, doesn't that create the opposite permeability ratios from what we have been discussing?
Thanks and best of luck with your project,
Paul
From: John Swearingen
Hi, Paul...
Of course few of the sites where we build can be called extreme climates, unless you call them extremely mild, so we haven't seen the occurrence of wet patches you describe; but they do make sense. I've been impressed by the degree to which vapor migrates around inside the bale walls, and how warm vapor rises. If it encounters a cold surface, then naturally it will condense.
Vents would be one solution, though as you suggest they might be the cause of other problems. Another alternative is solid insulation at the top of the bales. We are fans of box beams (though they seem to be going out of style), and one of the many good things about them is that they provide a highly insulated and vapor impermeable layer on top of the bale walls, similar in effect to a top plate in stud construction. It would be unusual for the underside of a plywood box beam to ever get so cold as to trigger condensation.
In any event, I think we're reaching some sort of conclusion that the solution to the roof condensation condition lies in detailing the top of the wall...
I've been following this discussion with interest, empathy, and renewed appreciation for our California climate. As John says, top of the wall detailing does seem key. The amount of condensation you describe is remarkable, and it makes me think there must be air leakage, not just vapor migration; and perhaps the vents in this case are more of a condensation problem than cure.
How are the inside joints between wall and ceiling sealed? Is there an air barrier and/or vapor barrier between the insulation and vent space? I'm not sure what you mean by a "proper" vent—is that an impermeable plastic tray?
As with the wall, in that climate I think you'd want a less permeable interior ceiling surface and more permeable membrane to the upper vented space, if you have venting.
We recently did a bale house in Wyoming, and the cathedral ceilings were unvented with 1-2" of rigid foam sprayed under the roof sheathing, and the balance filled with batt insulation (as dense spray cellulose was unavailable). Here in California we typically now use dense pack cellulose in unvented rafter bays.
I know that venting is traditional, and should help, but I've also heard John Straube say the typical 1-2" of air space in a vented cathedral ceiling is too minimal for good air movement.
Thanks for describing the problems and conditions so thoroughly. It is certainly perplexing, and my take on it is from a decidedly drier side of the country.
Dan Smith
From: Dirk Scharmer—FSB
Rene,
I'll try to contribute a little to the moisture discussion by giving a short summary to our activities in this field. Sorry for not corresponding directly to the earlier messages.
In the last months we did some straw bale 'dynamic moisture transport simulation' with WUFI. You'll find some English description about WUFI at http://www.wufi.de/index_e.html. But I guess you know this software; it seems to be one of the worldwide standards systems for this task.
The 'Fraunhofer Institut Bauphysik' did the straw bale simulations. They also did a lot of publications to hygrothermal behaviour of walls. You'll find them on their (English) website: http://www.hoki.ibp.fhg.de/ibp/publikationen/publikationen_e.html. Take a closer look at the dissertation of Sedlbaur.
The report about two strawbale walls we assigned gives the following results:
1. In German standard climate (Holzkirchen), unprotected straw bales (but under roof) molds(!). This is merry nonsense, but it shows that this dynamic simulation software calculates to surely for our straw bales.
2. As expected after the first point above, our straw bale wall with 3cm wheat stabilized clay plaster, 45cm straw, 3cm interior clayplaster, molds at about 6000 hours in the year. In reality the wheat stabilized clayplaster works wonderful, for example at www.fasba.de >> projekte >> Strohpolis.
3. An alternative wall construction, which we don't prefer, behind-ventilated timber cladding, 2cm wood soft fiber, 35cm straw, 1.5cm OSB interior, doesn't mold.
Now we've at least one configuration which works under dynamic simulation done by computer. The reality is fortunately more friendly to our bales. In the next months, we've to find a way to prove the clayplaster option.
To the moisture problem of Paul Lacinski's client: I had a moisture problem with a lime layered clayplaster this summer, too. We had to renew the weathersided (western) wall of the straw bale house. You'll find it on www.fasba.de >> projekte >> Schier.
We assume that the 'resistance against vapor' (accurate expression?) of the lime layer was too high, but let too much driving rain coming in. I had not enough endurance to read all the partly very long messages from the moisture thread, but I guess the problem Paul Lacinski's client had is not similar to ours above.
We took at one time 10 straw samples from 5 buildings and analyzed their content of fungus. Two samples showed (amongst others the weatherside of House Schier) much above common house dust concentration of fungus.
Puh, several month since I had to write English...
(On Thursday the German licensing authority decides about our request for a general German approval of straw bales as infill insulation; our main challenge was to prove that straw bales don't mold.)
Great to see Dirk doing analysis and moving this forward. I am not sure what you mean by WUFI "predicts mold" though, and this could be important.
I have run thousands of different WUFI simulations on hundreds of different walls, including strawbale walls, which we correlated to field performance at the Ridge winery building. It is quite accurate for predicting the temperature and moisture conditions for strawbale walls if you get the input data correct—however, IT DOES NOT PREDICT MOLD GROWTH. I emphasize this because WUFI is widely used and people often apply WUFI-Bio (a separate program) or simple 80%RH thresholds to assessing mold growth. We have never been able to get correlation from field or lab measurements of wood and paper RH and temperature and mold growth. Almost all rules are VERY conservative
The interpretation of mold growth thresholds is, I believe, what explains Dirk's successful field results.
Dr John Straube
Associate Professor
Dept of Civil Engineering & School of Architecture
University of Waterloo
Waterloo, Ont. Canada http://www.civil.uwaterloo.ca/beg
From: Dirk Scharmer—FSB
Dear John,
The WUFI Experts (Dr.Martin Krus) of the Fraunhofer Institute indeed took WUFI-Bio for predicting mold growth. Therefore they classified straw in substrate class 1 (mold growth threshold). The substrate class 1 gives an isopleth of humidity and temperature for mold sprout and growth. For pure laboratory research this isopleth and WUFI-Bio seems to be well validated. Also they tested (very expensive!) our clayplaster to get the correct input data for WUFI.
I discussed the findings with Sedlbaur and Krus, but their belief in WUFI seems to be unlimited.
What is our back door? To find out a special straw isopleth?
What results did you get with comparable strawbale wall (3cm clay, 35cm straw, 3cm clay)?
By the way: The German licensing authority demands the mold-free evidence with WUFI. The bad alternative is the normal German mold test (DIN IEC 68 2 10) with nearly 100%RH and 30°C over 28 days by infecting the straw with a dozen mold spores. Under this condition straw will mold after 2 days and will fail the approval.
The goal of WUFI is to say: Is the straw bale construction SURELY staying for years free of mold? This is the sticking point we have to come over. How do you solve this?
How do you judge the mathematical model of WUFI-Bio? Isn't the way to calculate a single spore a little bit absurd?
Dirk (really excited to have found a Straw-WUFI-Expert out there)
From: Rob Tom
Paul and GSBN Amiguettes and Amigos;
I haven't had a close look at the long, long long (and very nice, thank you very much) report yet but my first response is that I doubt very much if vapour diffusion through the wall plaster is responsible for the wet walls or that it's even an issue. It never has been with conventional construction systems in cold climates and I think that it is even less likely with the thick wall plasters that are typically used with SBC.
Besides, the greatest vapour pressure will be at the ceiling and its vapour permeance (1/2" drywall) is higher than that of the wall plaster.
I suspect that the biggest culprits are rain wetting and air leakage.
The ice formation on the underside of the roof and the wet ceiling insulation and top of the walls are strikingly similar to a classic example of poor air sealing in the ceiling and poor attic ventilation. Cathedral ceilings are particularly prone to the above symptoms because they simply do not provide the attic volume that triangular trussed roof systems provide. The additional volume helps to alleviate the poor ventilation condition somewhat by dispersing the trapped moisture—but, of course, exhausting it altogether is preferable.
That the attic moisture then ran down to the bottom of the underside of the roof and then rained onto the walls is not at all surprising.
While it is mentioned that there was a poly sheet used in ceiling beneath the drywall and that there was tarpaper placed beneath the wall-ceiling junctions, I am more than a little suspicious that
(1) the poly was punched full of holes when installing the drywall
(2) there are penetrations for electrical boxes and wiring that weren't sealed
(3) the joints between sheets of poly weren't made air-tight or in short, the air barrier is not.
As for the 40% interior relative humidity in winter, that's not excessively high IMO.
I like to maintain at least that level in order to avoid the health problems that are associated with the too dry winter conditions that are typical of air-leaky buildings.
=== * ===
Rob Tom
Kanata, Ontario, Canada
From: Rob Tom
Scooz 'eh Moi;
In his previous message, that dodo Stronzo di Nord forgot to ask:
"Were the rafters
(a) solid lumber (or TJIs)
or
(b) Scissors or parallel chord trusses?"
and
"What sort of a ventilation space was provided over the insulation in the rafter bays of the cathedral ceiling and what arrangement was used to provide exhaust capability for the ventilation space?"
=== * ===
Rob Tom
Kanata, Ontario, Canada
From: John Straube
Rob Tom's analysis seems spot on. Diffusion can be ruled out except in cases with essentially no plaster on the interior and thick and/or impermeable plaster on the exterior. Air leaks are very very likely.
I would not run 40%RH in outdoor temperatures much below 30 F, and most windows (not good ones) will have condensation below this temperature at 40% anyway.
Dr John Straube
Associate Professor
Dept of Civil Engineering & School of Architecture
University of Waterloo
Waterloo, Ont. Canada http://www.civil.uwaterloo.ca/beg
From: Habib John Gonzalez
Greetings:
Early in our bale wall moisture monitoring I noticed the initial high moisture spike in the standard five moisture sensors immediately after the stucco was applied. The building grade bales generally arrived on site with a moisture content between 9% to 12%. A few days after plastering the moisture content at the sensors would be around 18% then begin to decrease steadily. Nearly all these homes were finished immediately after the stucco work and the families moved in.
Sensor readings after the initial Canada Mortgage and Housing Corporation study were then randomly taken as requested by the owners. This summer I tested one home prior to its sale. The house has been occupied since completed in 1997, the five moisture readings were between 7% to 9%. The house is featured on pages 166-171 of Catherine Wanek's "The New Straw Bale Home." It sold for $450,000CND.
Two years ago master plasterer Gordon Askey elbowed his way into our stucco work. His 35 years experience as a stucco contractor and innovator has proved invaluable to our clients. He has concerns about late season plaster work in unheated buildings due to the moisture load he feels the plaster puts on the building. I mentioned this discussion to him this morning and he was very clear about the need to gently heat the buildings in the fall and early winter, particularly when the night time temperatures drop to minus 5 degrees Celsius. Below 5 degrees he does not work outside. In the fall, if there is any delay in installing the heating system in the newly plastered bale home, he uses portable electric heaters and two dehumidifiers per house to protect the bales. He did this for his first two bale home projects and was astonished at the amount of water collected daily from the dehumidifiers. Gordon observes the steady drying of the exterior plaster, even in cool conditions, he worries about moisture trapped within the building.
The standard stucco product in this region by Lafarge; it has a composition of 70% portland cement and 30% lime. It dries significantly over night and often the walls need dampening before the second coat is applied. To avoid cracking of the second coat by the suction of the first due to the drier initial layer, Gordon "double backs" the plaster, a technique he learned as an apprentice. We divide the job into workable segments with expansion joints on the outside walls and use the interior posts as dividers. We spray this section with the carousel pump, level and edge it with trowels, scratch it with a broom, then stop for lunch. We respray the same wall section again in the afternoon, often with integral colour and float finish it. This system has drastically reduced the cracking in the surface layer.
Rain splashback off scaffolds has been a problem which we now avoid since we hang tarps from the eaves of the building to prevent extreme weather, either rain or direct sun from damaging the new plaster. Our experience with conventional stucco is it often dries quickly and has to be moistened between coats. This is not the case with earthen plasters, which often take weeks to dry between coats, allowing the bales to absorb increased amounts of moisture.
Yesterday an architect friend called from Washington state and I mentioned this discussion to her. She too has concerns about clients who do not heat their homes after the plaster is applied in the fall. She has seen earthen plasters mould over winter due to this excessive moisture.
Good ventilation, dry materials, safe placement of the building on the land, best construction practices, special considerations to the conditions we create with late season plastering, all important aspects of this work.
My thanks to you all.
Habib
**********************************
Sustainable Works
Habib John Gonzalez
RR#1, S-4, C-12
Crescent Valley, British Columbia
Canada, V0G-1H0
tel/fax 250-359-5095 www.sustainableworks.ca
***********************************
"Better the kindness of imperfection than perfection without kindness"
From: Rene Dalmeijer
Paul and all the others,
Last weekend just before the first Dutch SB house tour started, at the IJburg house I managed to do some measurements again. The moisture levels, besides the wet spot, are down again to normal; i.e., 14-16%.
The wet spot due to the scaffolding splash zone is drying out albeit very slowly.
>On two projects we tried Tom Rijven's dip method...
I have not built a real big house using this method, but did a test wall during a show and a very small outhouse using the dip method. The dip layer dried quite rapidly—about 1 hour after dipping. We completely finished the plaster in about 3 hours after stacking the wall (for the show wall and the small out house roughly the same). During both occasions it was very nice weather... although again on both occassions we also did have a short shower. I am really a big fan of the method; it is dirty but works very well for the reasons Paul mentions. I would like to add dipping also removes the need for trimming.
>In your project, is the earth the finish material on
>the interior? And if so, doesn't that create the
>opposite permeability ratios from what we have been
>discussing?
Yes, earth is the interior plaster; and yes, it is opposite to what I would like. A less permeable interior plaster with a more permeable exterior plaster.
I fully support Dirk's request for a more accurate prediction of damaging moisture levels in organic materials. Is this possibly a Phd research project? That would be great—taking away a serious constraint to the acceptance of (specifically in Germany) renewable regrowable building materials. We all know they perform better then the numbers tell. I wonder how wood fares within these simulations?
Thank you all for the replies.
From: Graeme North
Two things appear apparent to me.
The first is really obvious:
> The McSweeneys had not been venting showers, were drying
> laundry in the house(including diapers) with no ventilation,
> and had a pot of water on the woodstove.
What do you expect? Any house of ANY type would show trouble with these ownership practices.
I'm not sure what the roofing material is, or what the eaves overhangs are in relation to the rainfall/driving rain index in this climate. I assume that there are generous eaves, and that the roof goes over the tops of all the walls? Any significant external wetting of the walls can not possibly help.
However, the other source of real trouble seems to be condensation in the roof spaces running down the underside of the roofing material that has not been directed to the outside well clear of the tops of the walls.
In temperate New Zealand at least we would put vapour barriers on the warm (interior) side of skillion (cathedral) construction in the ceilings of wet rooms—e.g., bathrooms and laundry, possibly kitchen. Underneath roof cladding we install breathable absorbent kraft paper building wrap that will trap and direct any moisture that condenses on the underside of the roofing material into rain water gutters. I think that making interior surfaces more impermeable will only help trap moisture in the walls, and would be against that. Earth plasters, although brilliant at moderating humidity levels, are not infinite sponges and do need to allow absorption to be reversed with adequate ventilation.
Best,
Graeme,
Graeme North Architects,
49 Matthew Road,
RD1, Warkworth,
New Zealand 1241
Ph/fax +64 (0)9 4259305 www.ecodesign.co.nz
From: Graeme North
Hi
I received an offline email from one member off this group as follows:
"I read with interest your comments on the moisture issues currently being discussed by GSBN. They prompted me to look at your web site and I just read the paper you presented in 2002, STRAWBALE BUILDING GUIDELINES FOR WET AND HUMID CLIMATES. I would encourage you to reference that paper in this discussion as it articulates clearly many practices and potential problem areas that need to be spoken, rather than possibly assumed."
Sorry—a further comment—we are not seeing dew point formation in the strawbale walls are we?—I understood that this has never been detected in strawbale walls, but logic suggests that it could happen.
Graeme
From: John Straube
Sorry I have not had time to think about this together with you.
I am still not clear about the wetting observed. The one source mentioned, the condensation on the underside or roofing, is clearly due to air leakage and not protecting the top of the bale walls. The wetting around low-density gaps, the recurring wet spots, and the top of wall problems mentioned, are all unclear as to source or transport mechanism.
Some observations:
The discussion on drying out is very important. We cant make general conclusions about the viability of SB unless we are starting with the best practise in construction drying—heat the space generously in cold weather and ventilate. During operation, venting moisture sources plus general background ventilation is VERY important in cool to cold weather (less than 55 F or so). I think Habib's comments summed this up well, and we all likely know what should be done, and Paul's experience may emphasize how important it is.
Everything I know about straw, plaster, and buildings screams at me that vapor diffusion through the plaster skins is not a problem. Until we have clear evidence, I must warn against pursuing this. The exterior portions of walls during cold weather will have higher moisture content because moisture in the SB and the stucco will not dry in cold weather, and the average RH (which is what determines the MC of protected straw) rises in New England in the winter versus the summer.
The observations of moisture damage Paul is making at joints and low-density parts is compelling evidence of convective air loops and redistribution of vapor within the bales. This behavior is not unexpected, especially in walls exposed to more extreme temperature differences (say 40 or 50 F differences, due to cold air or hot sun). Hot air rises, cold air falls, and moist air at the same temperature as dry air is significantly lighter. If the bales are not well packed, and/or gaps not sealed, air can rise and fall and move around. This flow is what was observed during the bad ORNL R-value testing, where the flow transported heat. The same type of flow will transport tremendous amounts of vapor from wet parts of the wall to dry parts, and from warm parts to cold parts. It is quite conceivable for vapor to rise from the concrete slab base to the roof by this mechanism under daily temperature cycles: the sun heats walls, evaporates vapor, which rises due to its heat and moisture content reducing its density, and then vapor in it condenses as the exterior of the wall (and especially the roof) cools at night. Note: Roofs will often cool 10 to 20 F below air temperature on cool clear nights. The goal is to provide a void-free wall, and to maintain as many bits of the wall at constant temperature. Embedded concrete or steel lag or lead the temperature swings too much AND can't store vapor fast enough to prevent condensation when cooling occurs.
And finally. I see rotting failed houses all the time. Often for the same reasons failures occur in SB. Examples given on this list show, however, that even careful, intelligent, and skilled people will have a problem with SB, whereas the lack of same is often the problem in wood frame houses. That said, I know several great builders who have also problems that could not be foreseen. Check out this poor women's experience in Philadelphia—it is not uncommon... it is just rare to see it so on the web. http://imageevent.com/platow/defectivehouse
PS. I would also like to recommend Graeme's excellent paper.
Dr John Straube
From: Graeme North
Hi John, Rene -
The air tightness thing, and lack of loose fill or voids is something that has been hammered home to me in this latest discussion.
I'm inclined to agree with Rene that the interior plaster is often more important than the exterior.
By the way, the one instance I have seen of condensation inside earth walled buildings in NZ was a woman in a very cold climate who was in the habit of closing the windows and boiling soup all winter—not helped by cement plasters on the mud brick walls, with acrylic paint, and poor solar design—it this experience with earth as much as anything that makes me sure that the hygroscopic qualities of natural (unstabilsed) earth plasters, esp. on interior surfaces of straw walls will help immensely in keeping strawbale walls dry.
Interesting.
Cheers,
Graeme
From: Paul Lacinski
Rob, Dan and all,
Sorry for my slow response.
The rafters are regular old lumber (a); 2x12s if my memory is correct. The poly/cellulose is retained by strapping, to which the drywall is attached. The ventilation space is exactly what is created by Proper Vent—(not very permeable, thin foam tray) say, 1.5 inches?, with openings at soffit and ridge.
I should add that the ice was only on the outer 4 feet or so of the sheathing, and that it didn't occur at all in areas of the south roof that are over large windows with no bales above.
Also, Dan Smith said:
How are the inside joints between wall and ceiling sealed?
The answer is—not very well. A structural timber separates the wall and ceiling planes, and there was clearly massive air leakage at these joints. The joint between timber and plaster was backed with 15 lb felt, but it was not sealed in any effective way to either the ceiling poly or the beam. I'm pretty sure that most of the moisture in the roof cavity came from this point, though I'll be nothing short of thrilled if Rob can demonstrate otherwise. Nonetheless, this doesn't explain why the moisture level got so high throughout that outer 30-50 mm zone of the wall, even in areas that had very little exposure to rain, and why people are experiencing seasonal damp patches in houses that should be well beyond the drying of their construction moisture...
Thanks again for all of your thoughtful responses!
Paul
This document will be updated as new posts come in.