Wall moisture monitoring
I’m diagnosing a tricky moisture issue. Part of the process has involved the use of wall cavity humidity and temperature probes. I have some data and I’d like to “peer review” some of the findings in broad strokes.
First off, this great article got me thinking:
https://www.greenbuildingadvisor.com/blogs/dept/musings/when-sunshine-drives-moisture-walls
The wall in question has cement fiber board siding that receives lots wind-blown rain, wind, fog, and sun exposure from 12 pm on. A lot of the time after a storm, it will be sunny the next day. The wall has SW exposure and is about 2250 sf. The moisture is consistent across the entire 75 ft length of the wall. Under the fiber board is Tyvek, then plywood, then Tyvek, then gypsum fire board, then studs/insulation, then interior wallboard.
As you know, California has been having huge storms recently. The initial data I collected showed that as temperature increased, RH increased after a big storm when the sun came out afterwards. This didn’t seem right until I started reading about reservoir cladding. Even on colder days the wall gets quite warm. The humidity in the wall is high at all times. I brought over a dehumidifier and have left it on 24/7. Now the RH rises as the temperature drops, as I would expect, and there is a downward trend in RH. I seem to be drying the wall, at least in spots. The plywood siding was pretty wet, but now shows lower moisture in the interior than the outside.
I can still smell the sweet smell of damp fiberglass through some cavity openings. I have performed an extensive leak search, to no avail. I’ve also blocked all reasonable air leak paths. There appears to me no air migration. I also believe 2×4 fiberglass insulation was used in a 2×6 wall. The Insulation does not have kraft paper. The cement fiber board siding was added after the building was initially constructed.
When I remove the dehumidifier in two weeks, I will take data as a base line and compare to data after another storm (hopefully there is one). My current thinking is the reservoir cladding and air space from the non-kraft paper 2×4 batts in the 2×6 wall are creating the issue. Does this seem reasonable? I’m hoping the wall system is polar and if I dry it enough, a normal drying process will resume. Probably wishful thinking.
Do any of you have any thoughts on a reasonable fix if RH rises once again? Perhaps a vapor retarder on the inside of the exterior wall?
Thanks for reading.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Sorry for the typos. I think there was a way to edit the OP in the past.
Where, through the thickness of the wall, are you measuring humidity? In the cavity? which side of the insulation?
And where in CA, so we can get an idea of what kinds of temperatures you are experiencing?
Finally, how old is the building?
Hi Charlie, thanks for the help:
San Francisco
I'm measuring in the cavity. The probes are between some roxul insulation I installed and the interior of the exterior wall, so between the roxul and gypsum fire board.
Rob,
You wrote, "The humidity in the wall is high at all times," but you didn't give us any acutal data.
1. Are you trying to measure moisture content of materials or relative humidity in the cavity?
2. What type of monitoring equipment are you using?
Measuring moisture content of materials or RH of air is tricky, and it's easy to get misleading readings. Inaccurate equipment or user blunders can give false readings.
You wrote, "The moisture is consistent across the entire 75 ft length of the wall," which is odd. If your roof has overhangs, the wall should be wetter near the bottom than near the top. If you can determine that the moisture readings show wet areas and dry areas, it would give you more confidence that your readings are accurate.
Here in New England, where our weather patterns are quite different from yours, high moisture readings in walls in February don't mean much, as long as the wall dries out in April and May. It's possible that your walls will be dry when you get a week of dry, sunny weather.
-- Martin Holladay
delete
Hi Martin, regarding tricky RH measurements, I'm looking for information to help hone my methods. The wall has been wet or damp for some time. Here is some more background information:
Located in San Francisco, CA.
I found soft gypsum board along the top 10 inches of a 16'8' wall, right below the ceiling. It was shot and had to be removed. The cavity side paper was falling off and there was some mold, but not as much as I expected, at all. There were water drops and moisture along the top 6 inches of wall, of varying degrees, for the entire 75ft length of the wall. I've checked other areas of the building with moisture probes and this wall seems to be the only visible problem. The heating system is a radiant floor heating system; about 950sf of it.
The tops of the wall fiberglass batts near the ceiling ranged from damp to wet to soaked.
Drywall screws with surface rust have been found from 1/3 of the wall height and up. There was bio on drywall I removed, and some on the tops of the studs, and bottom of the top plate. There really wasn't much. I was expecting a disaster. I looked for rot and didn't find any. Pretty lucky. The moisture I found explains why there's wall board cracks along joints of the interior wall board. Some tape pulled off very easily.
It seems there has been moisture in the wall to varying degrees, perhaps longer than the 6 years I've owned the place. Right before I bought the place, the cement board siding was installed. I saw a bad quality picture that appeared to show the original plywood siding (now below the cement board) to have peeling paint, all over. Perhaps the wall was very soaked when the cement siding was installed.
Above the wall in question is the flat roof and there is a 4ft parapet extending up above the flat roof along the perimeter on the 75x75 building. So to be clear, the wall in question turns into a parapet above the roof. The parapet cap is good and well sealed. I had a roofer check the flashing to the adjacent building and it's good. The adjacent building roof is 10ft below my floor. My floor level is about 30ft above grade. I'm in a zero setback part of town.
There are no unreasonable air leaks into the wall and I've verified this. Perhaps there is an unsealed pipe opening from below, but that would not explain the moisture across the entire 75ft length of wall. The North wall is very similar in exposed area and identical in construction, with no visible issues.
Starting at my floor, there are two bottom 2x6 sill plates, then 16'6" or so studs, with two 2x6 top plates.
At 8ft from the floor, there is a test hole where I used a wood moisture probe to check the plywood siding. The probe maxed out at 40% (the probe's max) before I started dehumidifying. Now the probe shows lower % on the interior surface and if I push it deeper, the % increases, so the dehumidifier is working.
I have two humidity/temp sensors that correlate in temperature and humidity. One is placed in the living space to monitor interior conditions. I used roxul insulation to block airflow in each stud bay. The other sensor is placed between the gypsum fire board and roxul insulation and the sensor on this probe is facing the roxul.
I was able to correlate the sensor readings by placing saran wrap over one of the stud bays. On sunny days, moisture condenses on the saran wrap and the sensors read 100% RH.
The data I have from last week shows RH rising as temperature rises. These were sunny days before the rain last weekend. During that time, 2/26 - 3/4, the cavity RH ranged from 100% to 51% and the cavity temperature ranged from 50F to 87F. I then started the dehumidifier last Saturday and the RH now drops as temperature rises. I've attached some pictures.
I hope I've answered everything. has anyone ever seen a wall take years to dry out? I suppose it's possible, but I wonder how likely.
Rob,
Usually, the wettest part of a wall is the bottom of the wall. It's hard to determine what's going on with your wall over the internet.
If this is condensation, it would indicate that the drywall is unusually cold. But again, the drywall near the floor is more likely to be cold than the drywall near the ceiling.
I'm guessing that the phenomenon you are observing has something to do with the parapet. Probably not a roof leak, because roof leaks are rarely uniform along the entire length of the wall.
Any readers want to guess?
-- Martin Holladay
Guesses appreciated! : )
I am looking for any ideas at this point.
I will most likely do the following:
1. Continue dehumidifying during the nice weather we will have at least for the next 8 days. I will turn off the dehumidifier before the next storm and take data through and after the storm to see what the trend is.
2. I’ll add 1” foam board or Roxul mineral wool board just under the top plate of the wall cavity in case there is an insulation defect between the top plate and the parapet wall above.
3. I may be able to somewhat easily reinsulate the entire wall if I remove a center horizontal band of 4x8 interior wall board and soft pack roxul insulation above and below there removed band. This would remove a decent amount of air in the wall and kill any air migration. This wall adds a ton of heat to the space in the summer as-is, anyway, and the 2x4 batts in the 2x6 wall could use improvement. The fiber cement board has a high specific heat and stays warm a long time.
4. Install probes at various points across the wall during the insulation work to monitor the situation. If conditions do not change, I will have to address the fiber cement board siding and house wrap.
Does this seem like a reasonable plan?
Here is my thinking:
Moisture across entire length of wall probably means it's not a leak, which I would generally think would occur in one point considering our new roof and construction.
Moisture across the entire length of wall could be related to a parapet negative pressure situation since this wall receives a lot of wind and moisture exposure. The Durolast roof seems to fold over the top of the parapet and the lip under the parapet cap is sealed. In addition, the original roof is under the Durolast roof. So, probably, it's not a negative pressure issue from wind sucking air into the wall. There are also seems to be no air paths from the wall cavity to the interior of the parapet wall.
The plywood sheeting is wet half way up the wall from the floor, and probably lower and higher. A roofer inspected the flashing between the buildings and said it was proper. This is the only portion I have not seen for myself, so I will take a look at this, soon. Hopefully there is some issue he overlooked; probably not. The plywood probably got wet either from a leak through the fiberboard cement board joints, condensation, or solar vapor drive. I am confident there are no major air leaks into the wall. I will check the fiber cement board butt and other joints again. Probably not likely, but it could be a problem consistent across the whole wall.
Condensation means, to me, there's too much air getting in the wall and/or there's a consistent insulation defect across the entire wall and trapped air is condensing against the cold exterior wall, then vaporizing when the wall is hot. I don’t think there are any unreasonable air infiltration paths. The 2x4 FG batts in a 2x6 wall stick out since they are consistent across the wall.
If the top plates are getting too cold, there could be an insulation defect above the top plate between the ceiling and roof/parapet. This could be consistent across the wall. It's unclear if this could generate the amount of water to saturate the plywood siding.
I was reading about the Tyvek Home Wrap we have under the cement board siding. Dupont says it is rated at 55 perms. I then checked the installation specs for Hardy cement vertical siding and another brand. My brand is unknown. Hardy recommends their Hardiwrap, which is rated at 15 perms. The other manufacturer (forgot the name right now) doesn’t say much except that they recommend a barrier (about 1 perm) based on ASHRAE specs “as required,” but don’t give a lot of information. This would also be a consistent factor across the wall. This would also be the worst case scenario.
Another idea is that the wall may have been soaked due to bad paint on the plywood siding before the fiber cement board was added and it could not dry out completely. How likely is this, and any of the above?
Here's some data I've collected from the probe in the wall. The week view plots ended yesterday. The month view is all the data I have. Some of the jaggies are me checking on the sensor. I started dehumidifying on March 2. There was rain on 2/26, 2/27, 3/4, 3/5, and 3/6:
Data:
Rob,
At the recent NESEA conference in Boston, I discussed your case with Kohta Ueno, who has monitored moisture in lots of walls in the course of his work at the Building Science Corporation. By email, I sent him a link to this thread. If we're lucky, he'll check in here in a few days and post a comment.
-- Martin Holladay
The only time I've seen wall sheathing that wet, consistently across an entire wall was from a large rainwater leak at a parapet.
John,
You may well be right. That said, I've seen a lot of roof leaks, and usually the water is concentrated in a few spots -- the water follows some element of the building assembly (usually framing) and stains the drywall in just one or two locations. What is odd here is the uniformity of the stains on the drywall -- such a distinct horizontal band.
-- Martin Holladay
Hey Martin, that is actually where I peeled paint off the joint. Along the entire wall paint and tape peels off very easily. Paint strips off easily from drywall compound, but is well adhered to paper. The drywall compound, perhaps, wasn't mixed well or high RH weakened it. Probably the RH since drywall screws have surface rust all along the wall. I wonder if the previous owner loved a humidifier.
Thanks for sharing the issue! I will redouble my leak search efforts again, just to be sure.
The next week will be dry and sunny, with temperatures that will bake the wall during the late afternoon- should get some ok data.
Rob,
Any evidence that your home was used to grow marijuana?
Not that I can tell- no weird power connections, etc.
Sorry I'm late to the party--per Martin's post, he chatted with me about this mystery; here are a few thoughts.
It seems like a rain leak from the roof (or parapet) is mostly--but not completely--ruled out. If you have a sunny day, some time, and a hose, an ideal test is a spray test--working your way from low to high, spray water on the roof-parapet interface, parapet, etc. Give each area maybe 15 minutes to soak in before moving on to the next area. It seems like you might be able to rule out bulk water leakage if you complete this test without seeing water inside.
Next--I think everyone might be on to something in regards to inward vapor drives; fiber cement does end up acting as a reservoir cladding, albeit a little less severe than brick, at times. A few questions: is this fiber cement lap siding? Or panel siding? Current Hardie recommendations are to put panel siding on 3/4" furring, which in addition to providing excellent rain control (via a rainscreen), also gets rid of inward drives by ventilation (see http://www.jameshardie.com/d2w/installation/hardie-reveal-panel-system-hz5-hz10-us-en.pdf). Also, what color is the cladding? Of course, darker claddings have greater solar absorption, will heat up more, and therefore will have greater inward drives. Is the fiber cement back-primed? That's BSC's recommendation (keeps the fiber cement from sucking up water from behing), but nobody ever does it.
For more on ventilated claddings, see:
BSD-013: Rain Control in Buildings
https://buildingscience.com/documents/digests/bsd-013-rain-control-in-buildings
For anyone looking for background on inward vapor drives and reservoir claddings:
Info-305: Reservoir Claddings
https://buildingscience.com/documents/information-sheets/reservoir-claddings
For a case study on reservoir claddings, and ventilating away moisture problems:
NESEA: Moisture Safe? The Writing is on the Wall
https://buildingscience.com/sites/default/files/2017-03-08_nesea_be17_ueno_for_pdf.pdf
One thing that is sticking in my head is the exposure of the wall: you said the floor is 30 feet up in the air, and the adjacent building is 10 feet below that, correct? Shielding vs. exposure can have huge effects--wind speed goes up with the cube of height, so I would bet that your wall sees a *lot* of rain. As an example--check out Figure 1 in this column by Joe. More rainfall = more loading up of that reservoir cladding.
BSI-095: How Buildings Age
https://buildingscience.com/documents/building-science-insights/bsi-095-how-buildings-age
The last item that sticks in my head is the use of 2x4 batts in a 2x6 wall. This creates an air channel that allows for convective looping of air in that empty space. For more on convective looping in wall cavities, see:
BSD-011: Thermal Control in Buildings
https://buildingscience.com/documents/digests/bsd-011-thermal-control-in-buildings
(see Figure 7: Convective air loops that reduce thermal control of insulation.)
http://buildingscience.com.678elmp02.blackmesh.com/sites/default/files/migrate/jpg/BSD011_Figure_07_web.jpg
Field Monitoring of Wall Vapor Control Strategies in the Pacific Northwest
https://buildingscience.com/sites/default/files/Field_Monitoring_of_Wall_Vapor_Control_Strategies.pdf
What might be occurring is that this creates a chimney for moisture buoyancy to operate. Believe it or not, moist air is less dense than dry air, so it will tend to rise. More on this topic here (especially check out BSI-016)--but key point is that when we see moisture problems in unvented roofs, they're often concentrated up at the ridge. This could explain why the worst moisture and condensation is seen at the top of the stud bay--that's a very tall (16 foot plus) high wall cavity that will "contribute" moisture.
BSI-077: Cool Hand Luke Meets Attics
https://buildingscience.com/documents/insights/bsi-077-cool-hand-luke-meets-attics
BSI-016: Ping Pong Water and The Chemical Engineer
https://buildingscience.com/documents/building-science-insights/bsi-016-ping-pong-water-and-chemical-engineer
Takeaways: using the pieces of information from the discussion thread, it seems *plausible* that this is inward vapor drives from the fiber cement reservoir claddings, which end up concentrating by the open airspace between the insulation and interior drywall (hygric buoyancy and convective looping).
So what to do? I'm not confident enough to say, "Sure, fill it with 2x6 insulation and it'll be fine!"--it seems like there might be worse problems than that. It *might* make it better... but not sure if it will be enough. A great solution would be to remove the fiber cement cladding, and add a rainscreen on top of the drainage plane, and provide ventilation... but cost wise, that's a non-starter.
Good luck, and keep us posted!
Hi Kohta. Thank you for taking the time to write that excellent summary. In addition, thank you to all who've contributed. This is one of my favorite sites. Martin & friends do a great job on all the content. I especially like the write up on house wrap. You guys don't mess around.
Back to the mystery. Kohta, I'm going to read the sources you cited. The insulation detailing is bad so I'll replace it with Roxul. Like you said, maybe it will work, maybe not. The current insulation level wasn't the best and the interior wallboard is in bad shape, so at least there will be some sort of upgrade in this process. It's been warm the last days in SF - high 70's! Yesterday, I removed all the FG batts and the interior RH jumped 20% very quickly. I could clearly feel the moisture in the space.
I then opened the ceiling to inspect the top plate. Picture attached. The ceiling insulation is also FG batts and it actually seemed pretty reasonable. The Batts were over the top plate, but if I built the place I’d do it better, but it seemed fine. Any thoughts? I also did not sense any moisture up in the ceiling.
The wall temperature yesterday was over 95F at the top of the wall. Today was the same. I also found the top of some batts right below the wall blocking were wet. The blocking is half way up the wall. Some of the blocking had drops on the bottom of them and the wall was hot! From what I can tell by the moisture I observed, and this is an educated guess, is the moisture develops in voids like you are saying. After discovering the drops on the blocking and noting the ceiling was dry, I am going forward with the assumption the ceiling and parapet wall is not an issue.
Another interesting note: I had plugged the top 3ft of the wall with Roxul. When I pushed the Roxul down, the FG batts (now removed) were pushed down. The transition from FG to Roxul was interesting. The FG seemed pretty damp, almost wet, at the top of the batt. The Roxul did not seem nearly as damp, but there was slight moisture in the Roxul. The Roxul may distribute moisture, or at least take up a lot more volume, limiting the maximum moisture that can be stored. Very unscientific, so, sorry about that!
The wall plywood sheeting is drying out based on the MC tool I have and I can sense it. There will be rain in about a week and until then it will be sunny and pretty warm. This will be the final test for me. If moisture appears to soak into the plywood again and / or after the rain I sense a lot humidity entering the building on a subsequent sunny day I will have the cement board taken off and replaced with a rain screen solution like you all recommend. I’ll probably have the North side done, too. I need to nuke this problem.
I took some pictures of the siding, the blocking, ceiling, and parapet. Take a look. It seems like the fiber cement panels are butted together on the short sides and there is drip edge on the long sides. The more I think about it, the more I think the butt joints seem bad.
Yes, my floor level is about 30ft up and the adjacent building is 10ft below, so roughly 30ft of exposure and about 2000sf on the West wall; same for the North wall.
I’ll report back when I know more.
In the meantime, what do you all think is the most economical and durable siding to use with a rain screen?
If you recall, the current assembly is fiber cement board, Tyvek Home Wrap, plywood sheeting, some kind of white sheet weather barrier, gyp-lap firecode core gypsum fire board, studs, interior gypsum board.
If I remove the cement board, would you guys add another wrap over the plywood considering there is some mystery wrap between the fireboard and plywood?
Kohta,
I really appreciate your taking the time to share your perspective. Thanks.
Rob,
Please report back here if your further investigations reveal anything new.
-- Martin Holladay
I agree--it seems like the problem is the wall, not the roof; there's not the same level of trickiness and vapor issues that are likely occurring in a membrane roof.
Yeah, this sounds consistent with convective looping in the stud bay concentrating moisture where it is stopped--the full-depth blocking, of course, with stop the air pathway, for the most part.
Question--at the drip edges on the long (horizontal?) sides--are the joints open? Or did somebody caulk them shut for some reason? Just looking for possible reasons why the moisture problems are as bad as they are. Of course, if the joints of the fiber cement panel are all caulked shut, there's nowhere for water to drain out of.
Honestly, any siding can be made into a rainscreen. I am betting that it will be difficult to remove the fiber cement panel without destroying it, but if the trades can do so, I'd have no problem re-using it in a ventilated rainscreen application.
For the "wrap" (water control layer/water resistive barrier/drainage plane): the key is to inspect what condition the existing wrap is in. If it is "wrecked," it's probably time to remove it and replace it. But if it looks mostly intact, there's probably little reason to replace it. The only wrap that really matters is the one that's directly behind the fiber cement, unless something really weird is going on (flashings directed into the inner "wrap.")
For rainscreen detailing--there are probably a bunch of articles on GBA showing the specifics.
Lastly, if it's an option to pad out the wall, adding ~1" XPS foam board on the outside would completely shut down any risks of inward vapor drives. Add furring strips on the outside, and you can install your cladding.
Hi Kohta, I'll take a closer look at the panel joints and get back to you.
I have all the FG out and all the interior wallboard off except the very top band. I'll get that off this weekend. There was for sure some good air movement in the wall. I can see moisture trails between bays and there are a couple trails through the middle of the sandwiched sill plates. For good measure, I'm thinking about caulking every corner of the stud bays, blocking, and electrical holes, etc. When the guys put the new drywall on I'll have them caulk the front of the studs. Also, there are a couple areas where the fire board has 3/8 horizontal gaps for a board or two, so I will fill them with fire caulking. Pretty poor detailing, overall.
It would be interesting to measure the effect of adjusting building pressure based on conditions. Air leaks (which all walls have) + rain driven against those leaks + wind pressure = rain in the wall. Depending on dewpoints (accounting for reservoir cladding) and temperatures: dehumidification + outward flow = drying in the air leak areas OR outward flow = lots of moisture deposited in the wall (inward would be drying). The latter can go way up in a multi story building.
Reasonable air leaks + enough wrong direction pressure = unreasonable amounts of moisture.
So, I have an update. On warmer days I've found two areas where water vapor is traveling through the bottom and upper sill plates. Picture of one attached. How should I seal the gap between the sill plates?
I went to the floor below and looked at the same wall. The electrical boxes there had a lot of air traveling into the room. I'm in the process of air sealing around all the boxes. It's unclear where the air is coming from. I'm guessing air is being introduced into the cavity, then vaporizes and enters my wall through the sill and other pathways. I don't think this explains everything that's going on in the wall above, but this seems to be a big component.
I've air sealed every joint in the wall on the second floor. You can see some in the picture. I'll have the drywall guys use caulk when they hang the boards and I'm about to install the Roxul batts.
Rob,
Q. "How should I seal the gap between the sill plates?"
A. With a high-quality caulk -- something a little better than ordinary latex painter's caulk.
-- Martin Holladay
Hi Kohta, it took me a while to gain access where I could take a close up of the horizontal joint detail. I did not see any caulk, but there is some paint. The second picture is of the flashing between buildings. I could not get all the way on the roof. It looks like the building next door painted up the vertical portion of the flashing and under the bottom of the fiber cement board. It's unclear to me if this would be a problem. There is plywood covered in building wrap that extends down the wall behind the flashing and down the building, I believe.
We had some rain a couple days ago and there is a clear correlation between rain and increased humidity from my sensors at the top of the stud bays with the new insulation and air sealing. I also have a moisture probe in an access hole sensing the moisture content of the plywood. Before noon it's generally 10% lower than the afternoon when the sun is on the wall on the same day. On dry days, the plywood is around 18%. After a rain the probe reads 30%. I am getting quotes to create a rain screen. Question for you all: what building wrap would you recommend? I would like to have air sealing if that makes sense because I want to seal any chance of air leakage. I have seen the adhesive wrap products and I would like to know your opinion. Also, are there any good methods to seal the wall below the flashing? Maybe roll the adhesive wrap using a long roller in the narrow gap between buildings?
Thanks again for all your help!
Gotcha--looking at your pictures, I don't see a "smoking gun" situation why the fiber cement panels would be causing major problems. The gaps at the bottom (between the fiber cement and the metal flashing) looks wide open, which should allow for plenty of drainage. Cladding ventilation, with the two materials smashed together... not so much.
My go-to products would be the many vapor-permeable self-adhered membranes--such as Henry Blueskin VP160, W.R. Grace Vycor® enV-S™, Cosella-Dörken Products, Inc. DELTA®-VENT SA, and Vapro-Shield SA. Also, the fluid-applied WRBs (see Martin's column, Housewrap in a Can) will also provide great performance and a robust air barrier.
Huh... not sure I understand the geometry, but yeah, anything that connects your air barrier to the next airtight layer would probably work, unless there's a bypass you're not considering.
Thanks for the wrap suggestions. Here is a bad picture of the gap between our building and the next. Our building is on the right. The flashing is above this gap.
Huh... that's a hard one Rob--there's not much access to get into that slot. Whatever you can do to make the connection will be helpful (I assume you're talking about the air barrier aspects). Of course, any water that is draining out of the wall above the roof should be drained out onto the lower roof.