Dear Experts! I am completely lost on this one. HELP! I have a standing seam metal roof brand new on an old house. A roof leak is somewhat unlikely but possible. We removed old insulation due to storm damage and installed 6" Icynene open cell foam last fall after the roof was installed. In the winter I noticed the R20 isn't as good as I was told. The upper floor (cathedral ceiling 10ft) wasn't getting warm enough in the mornings and the sound barrier qualities of the foam were also not satisfying. So I went to HD and bought a truck load of 2" pink board and installed it over the Icynene. I used spray foam from the can to seal between the pink boards and created a really nice seal. The noise from the street went away and also heating and cooling (now in summer) perfect! Up until then no problems at all apart from the Icynene smell which still doesn't go away. Now to the problem: after a recent thunderstorm the entire ceiling began to leak, water started to drop in 20 different places mostly at the peak. I suspected a roof leak and pulled down one or two pink boards. As expected they were wet behind and the Icynene foam was wet, too. However, I am not 100% it's a roof leak because the foam appears to be wet only on the surface and it dried within 24h. I don't want to take all those pink boards down again just to check of course. Questions: 1. Is this due to condensation? (Location: MD, we had 90+ F and over 90% RH that day) The idea of using 2" pink board was a. it's an awesome non-smelling insulator. I am starting to think the spray was a mistake... b. 0.5 perms should allow enough drying towards the inside, so I thought... 2.How can I determine it's a roof leak by experimentation or observation? I have a feeling moisture came through the open cell foam or through a fine gap between the foam and the rafters. When I opened a hole in the pink board hot steam came out. All this pressure implies there is some kind of air flow, doesn't it? What should I do next? Remove the pink boards again? Leave a gap at the top for venting? Thank you for all your advice!

As I understand it you have R20 in vapor-permeable half-pound SPF and R10 in XPS. If you're keeping the room at 70F and it's 90F outside with the roof in the shade, the temp at the XPS/SPF interface will be about 77F, so unless the dew point of the outdoor air was 80F (unlikely) you either have a lot of seasonally-accumulated moisture in the roof deck raising the dew point of the entrained air in the SPF to well above 80F or you have roof leaks (or both.)

But if the Icynene installation was slopping not making a good air seal it's conceivable that air-leakage is at least partly to blame. Air moves a LOT more moisture than vapor diffusion through 6" of half-pound foam. The dew point of 90F 90% RH air happens to be 86F, and the exterior surface of the XPS is below that even with a 100F roof deck, but it would have to be that muggy for awhile to get to the point of dripping unless the air leakage rate was large. A water leak higher up can move quite a distance both laterally and downward in the micro-space between SPF/XPS, and it would leave the surface wet, even if it didn't soak the SPF.

The fact that the problem exhibited itself after heavy rains points to a real water leak, not just air. Any moisture that finds it's way is in there it won't leave rapidly though 2" of XPS. It can pass through 6" half-pound foam at about 20 times the rate of 2" of XPS.

What's the underlayment and ventilation (if any) under the standing seam? (If mounted on purlins it may have decent drying capacity toward the exterior, but if it's a thin mesh or roofing felt, not so much.)

Sav;

If the ceiling is getting water in after a storm, I would say the roof is leaking and not a condensation problem, was there "Z" channel installed at the ridge? and was there tape seal used on the ridge /"Z"  ?

Pics would be nice...

I would also lean towards some type of roof leak. I assume we are talking about a standing seam roof on top of an OSB roof, with a vented attic, correct?

OK here are the pics. Two shots from interior. One is main area, the other above porch. All finished, sprayed with Icynene 6" under roof, 4" in walls, 2" pink board R10. The exterior shot was taken right above the wet piece of drywall

As you can see it was a lot of work putting up all these XPS. Was is a good idea after all? In the winter everything was fine. The arrow points to moisture condensating towards the interior. This is how it showed up initially. Then it was dripping pretty much everywhere from between the XPS boards. The first and second rows from the top leaked the most The standing seam apparrently has no z channel. You can see the gaps. The metal panel was simply folded up. It appears to be sealed only left/right but not at the top

the above shot was taken from west facing east (the 1928 house is perfectly aligned :) You can see the gaps and there's no z at the cap and no sealant

Chris, are you Greek by chance...Kavala is a city near my hometown! Check the pics below there appears to be no Z and no sealant up there. Is this done craftsman like / according to metal roofing standards? I think wind-driven rain should readily enter that gap. The thunderstorm did indeed come from south east and was quite strong.

Lbear, it's an unvented new metal roof. The interior has been gutted and finished with new insulation between rafters to create a cathedral ceiling. No vents the old vents have been closed and sprayed over. We have wood planks under roof. Three layers of HQ high temp underlayment (!) under the metal roof because I had to explain several times to those guys you are not supposed to staple it but use 1" plastic cap nails and cover with next sheet from above. So they ended up doing the job three times and I hope they didn't mess up the third layer.

Hi Dana1 the is no ventilation under the metal roof at all. 2 and 3 layers of new high quality high temp underlayment was used (see above why). Now when I opened a hole at the top I did feel hot steamy air (about 105 F according to laser thermometer) coming up relatively fast flowing as if pressured. Either there is air leakage through the SPF or the air comes from the uninsulated walls from the first floor below. I did spray under the XPS at the floor trying to avoid air going up but I guess that was definitely not a perfect job. When the foam was sprayed I left it as-is for two months in late fall. I didn't notice any imperfections...apart from the smell that drove me nuts and the flakes coming down from where the SPF was trimmed. Regarding the storm, it was the strongest one this year with wind driven rain and immediately followed by extremely muggy weather for 3 days. It's interesting you figured 77F at exterior side of XPS. That would mean I will be getting condensation there all the time in the summer if the dew point is above that. Should I leave a gap at the top, say 1 foot to allow vapor to mix with interior air faster? Would that help anything? Should I remove the entire XPS layer? Your analysis shows there is potential for moisture moving inwards so I wonder if XPS is causing the problems when the weather is extremely muggy

Hi all thank you so much for your advice. I can feel the expertise here is at a different level! Really appreciate your comments and suggestions and the time you take to help out. This forum rocks I am so happy I found it! Let me know what you guys recommend I should be doing next. I never imagined I would have to learn so much about dew points, roofing, etc. as a homeowner!!

one more thing. The temp at the exterior, ie. the temp of the metal panels that day was about 115F. The temp under the XPS on the SPF at the peak was around 102F.
This makes me wonder about what insulation actually does.
Since the XPS is better than the SPF, and the sealed XPS was facing the interior (A/Ced at 73F), could it be the exterior side of the XPS equals the outdoor temp after a while? Does that mean insulation only slows down the heat flow but eventually the temps will equal out?
I mean there shouldn't be much air flowing between the XPS and SPF, so there wouldn't be anything to cool off the SPF...? Perhaps it would have been better if XPS was under the roof instead but that was way too expensive unfortunately

"R-values are customarily given in units of British thermal units per hour per square-foot degrees Fahrenheit (Btu/h·ft²·°F)."
(from Wikipedia)
Insulation slows down heat transfer - it does not stop it.
Looks to me that there are two things going on here - the R value of the insulation may be too low for the application and you have air leakage between the two insulations. You should be able to stop the air leakage by removing the bottom sheet of XPS and somehow seal it tight to the Icynene Open Celll Spray Foam. you could install a 12" sheet of XPS and foam the gap between that and the OCSF. Then the XPS joints should be taped - Dow Weathermate tape works well. (Acrylic tapes like 3M 8067 are better but hard to find). Also if you have any gaps in the OCSF they MUST be sealed. I've avoided OCSF so I'm not familiar with it's characteristics.

It won't help you, but we always install a vented channel under steel roofing with a roof membrane under that to "catch" condensation & drain it off. - perhaps that is what your high temp underlayment's job is.

any chance you could tarp the roof to help determine if it is rain or condensation?

Thanks Bob for your input.
More questions for you:
* the channel you are talking about, is it like elevating the panel from the underlayment? How do you screw in the clips holding the panel, wouldn't it require like a taller clip?
* I used spray from a can to seal between the XPS, as shown above. Was that a bad idea? I also have some tyvek somewhere I could use taping over it
* is it realistic even to try to seal 100%. It was a fairly neat job but I don't think it could ever be sealed perfectly. Also, I think if the moisture does indeed come in from outside somehow (which it would have to if condensation is to blame, wouldn't it?) it would be trapped underneath the XPS, and it takes foreever to try inwards.
* what if we did the opposite: have a gap under the XPS, behind, at the bottom as well as the top to allow vapors to exit. Would that nullify the entire xps setup?

Sav - if you vent a cavity, you are nullifying the insulation, so you do not want to vent the XPS. Spray foam should work well enough if you can directly spray into the gaps. Turns out it is realistic to try to seal 100% - not that you'll ever quite get there, but you can get close. Moisture will migrate through some materials, like open cell foam, but much much more moisture travels on moving air. Stop the air flow - the infiltration & exfiltration and you can stop up to 90% of moisture. (check the chart on page 17 http://www.buildingscience.com/documents/reports/rr-0412-insulations-sheathings-and-vapor-retarders. Or, think of a bucket - how well does it hold water with a bunch of holes in it? So, if it is condensation, and if it is coming from the exterior (I'm familiar with cold climates) then it appears you have air gaps in the OCSF layer.

I agree with Dana1 in that since it happened after a heavy rain it is likely a leak in the roof. Climb up on the roof, or hire someone to do it and seal every possible joint with a urethane caulk like Geocell ProFlex. The leaks are probably higher than the highest water you observed but seal every suspicious place you can find.

Ok guys I did more research based on your input.


First of all yesterday we had 12 hours of tornado warnings and flash floods. The areas I exposed with the XPS removed ARE BONE DRY!


Now I believe with 95% certainty it is indeed condensation screwing me here. Luckily I didn't put drywall up yet... I will go outside and spray with the garden hose for an ultimate proof the roof is tight.


Now to the facts I gathered. Remember it's a mixed hot humid climated MD
According to this brilliant resource http://www.numericana.com/answer/gas.htm#perm
I found out that temperature doesn't actually matter when dealing with perms.
The Icynene open cell foam as 10 perms at 6", the XPS at 2" has 0.5 perms
Btw, please correct me if I am wrong with my analysis.

Now, the table shows that at a 23 deg C with 85 % RH difference from inside to outside, let's take that as worst case,
a total of 130 gram of water per 24h per square meter will go THROUGH the open cell foam.

That's 22 liters per day (!) for my entire upper floor ceiling and wall area which is about 170 sq meter surface and 226 cubic meters air vol.
The XPS however, lets through only 1 liter (exactly as dana said roughly 20x less) per day.
You see we're stuck with 21 liters of water per day and have no where to put them! Damn it!

Dana said the exterior side of the XPS will be roughly 80F if the interior is 70F and exterior 100F. If I understand correctly it's because I have R10(XPS) + R20(OCSPF)
hence the temps divide 1/3 to 2/3 and 80 is the point where XPS meets OCSPF.

Hence, I have to disgree Bob that air sealing the XPS is the wrong strategy for my climate.
The moisture coming through the OCSPF has to be evacuated or else it will "rain" inside as it did. Installing the XPS with a very low perm over a "high" perm OCSPF is a mistake in this climate.

I wouldn't want to call any product "crap" but the installer should not have recommended it for that exact reason.
Had we installed closed cell foam the perms would be in the 0.XX region and hence no issue. There would be so little moisture coming in, it would dry immediately.

Now if there wasn't XPS on my ceiling the vapor would still condensate. That day we had 90+F 90% RH outside, the DEW point is just 4 def F below that, which turns out is right 1-2" into the OCSPF. It will condensate in the foam no question!
BUT if you have good ventilation and very dry colder air underneath it may dry fast enough so that no drops become visible. Drywall has a very high perm rating but I don't think it would help. You would need to evacuate 18 liters, that's two buckets of water in 24h. As a result the interior RH rises dramatically.

Hence, you were right Bob, this wasn't the right kind of insulation for our climate.
I just wanted to tell people to look at the vapor permeability of the insulation material they choose not just the R factor.
I would also point out that the material with the lowest perm rating in a wall or ceiling WILL become the problem depending on the weather conditions.
The best system has to be the one that allows vapor to travel both ways, in and out. No matter what your climate is you cannot install a vapor barrier.
In a cold climate the risk may be lower to experience indoor condensation like that but a certain risk is still present.

In addition, I want to throw this into the discussion: maybe the problem isn't really related to unvented roofs. In a vented roof in a hot mixed humid climate, insulated with fiberglass you have the same problem, if not worse. You have a paper vapor barrier towards the interior followed by fiberglass which is totally permeable. Hot humid air brings the vapor through the soffits into the house, it easily goes through the fiberglass batt and then hits the cold paper. Boom you have condensation. Again if you have enough drying potential at the interior such as A/C or dehumidifier then you won't see it but it's occuring, regardless of R factor

The fiberglass I removed before I put the roof up was BLACK on the outside AND inside! An unvented asphalt roof was installed previously.

Since it's impractical i.e. freaking expensive to pull the smelly OCSPF down and replace it, what can I do at this point? I was thinking I could only add insulation over the OCSPF which has more perms, not less. Venting to the outside doesn't work because the problem COMES from the outside.
I was thinking the XPS needs to be removed then reinstalled with a gap away from the OCSPF, open at the top. Then hotter vapors may exit at the top fast enough to mix with dry interior air, which would HAVE to be generated by A/C and strong dehumidifier to cope with 22 liters per day minimum.

And there is another theory. Even if I had installed closed cell foam, it does end up with hairline cracks by the rafters, I see it happening under my kitchen floor (BASF CCF). Hence you need to deal with additional air infiltration as well. May be "minimal" but that really depends on the climate difference inside, outside, and the speed at which the air may mix. Drywall does sort of block air movement, too, so, it's tricky

All in all, the situation is total SH*T if you excuse my French, I put in so much effort and money and got s***** by the SPF reps.
Why did I not install CCF? It was too expensive but the main thing is they smell like rotten fish in the summer (I had two jobs done at the house with CCF). The Icynene on the other hand smells heavily sweet, more when it gets hot. My research leads me to believe it's isocyanates leftovers that generate this smell and they aren't the healthy kind. I am saying this for people to keep in mind it's a risk factor. The odors are real and usually dangerous substances. But other insulation materials come with their own risks, too, granted!

Sav; with no "z" flashing at the ridge it is open to the weather. Point the garden hose towards the ridge and be prepared for inside rain

Ideally you don't want high humidity levels in a home, period. Whether this is removed via ERV/HRV, dehumidifier or an A/C unit, you don't want more than 50% RH in a home. Here is an article from the US DOE: How Moisture Moves Through a Home

So even with a vented or unvented roof, the moisture has to go somewhere. More water vapor travels through a wall by air leakage than by diffusion. That is why it is so vital to get a home as air tight as possible. The tighter the home, the better. As YOU want to be in control of what air comes in or out, not the homes cracks and leaks.

Here is an article on Vapor Barriers


Remember, some roof leaks might not show themselves until a certain type of rain and wind speed/direction are present.

Chris, I am going to do just that as the ultimate proof. I am well aware that wind direction has a big effect. My old roof had it's own mind ;-)
It would leak with normal rain but when we had a hurricane and the rain came pretty much 90 deg, no water leaked in!

So definitely garden hose followed by a home run to check the foam. Nothing is 100% certain at this point. I would truly want to fix the roof with some caulk rather than having to deal with complex condensation issues inside

LLbear, I totally agree. Yesterday was one of those days I had to turn on the dehu inthe basement. What difference it made! 12h of severe thunderstorms and the humidy shoots up like crazy. Since it's an old house moisture comes in right from the slab too. I read cement has 10 perms, just like the Icynene at 5" by the way.
I think the basement floor is the perfect example. It's wet underneath, the floor temp is always 18 deg C and there's no condensation. Why? because I left it open unfinished so it doesn't condensate. However, the previous owner didn't know what he was doing.
He installed vinyl tiles and suffocated the floor. Carpet on the basement floor is another no-no, unless you've got thick and intact plastic sheet under your slab.

The basement stank all the time but when I pulled the vinyl the smell was so bad it reached the upper floors. Black mold all over!

So I covered the whole slab with hydraulic cement after removing all tiles and the smell as well the moisture is now gone. The plan is to put ceramic tiles down with a wide grout to allow for quick drying.

I think carpets and vinyl tiles are another example of an inappropriate vapor barrier at the other end of a house