Posted By Sav on 04 Jun 2012 05:55 PM
Jeepster I hope it's a roof leak actually. it would be easier to handle; however, not easy to prove to my roofer. The National Weather Service says the max were 91F and 97% RH (at 10PM). Unfortunately I can't find an hourly graph to see how RH changed during the day

Heck, then the heat index would be over 127 degrees!!  I doubt those readings were at the same time.  It might have reached 91 at 3pm and hit the 97% humidity in the early morning when the low temperature existed.  I would doubt your humidity got much higher that 55% at 91 degrees, and that's still pretty damn uncomfortable.  The problem is, is if the 91 degree 55% RH air entered your home and hit a surface that was below 71 degrees you would get condensation.  Was your ceiling colder that 71 degrees?  I doubt it.  Even if it was, you'd need a bunch of air flow to cause that much condensation.

http://www.dpcalc.org/

http://www.nws.noaa.gov/os/heat/index.shtml

Posted By Sav on 04 Jun 2012 04:21 PM
Dana, I appreciate your help very much!

On that day I checked weather.gov for BWI it said 34 deg C, 90+ % RH at the moment
Maybe those stats websites you used show a day average or something.

Now regarding the quantity of moisture going through, I was relying on this resource
http://www.numericana.com/answer/gas.htm#perm

which says in the yellow table 135 grams of water can pass through in 24h at 11 perms per sq meter IF 85%RH -> 0% RH at 23 deg C.
I have 170 sq meters total so I figured up to 20 liters => 5 gallons per day is the max. possibility for the entire upper floor.
But the metal roof at the top does block air contact, I agree, so I would also expect less moisture to come in in that case.
I want to finally finish the cathedral ceiling with drywall so spraying even more OC over it seems difficult.....we would need foot long screws....

Show me the data- it simply can't be right.

Weatherspark.com uses weather data down to hourly readings for most weatherstations, but will sometimes interpolate between readings if they are separated by more than an hour.  The data for Baltimore MD are hourly.  The absolute record dew points for any random weather station in MD are all ~80F, which is pretty uncomfortable, but still nowhere near as torrid as tropical jungle climates with mid-80s dew points.

The vapor pressure difference across the vapor retarder and it's permeance determine the rate of moisture transfer.  It's highly unlikely that the vapor pressure on the interior was anywhere near zero (which it would be if the RH was zero).   Realistically it might have hit 40% RH @ 23C indoors (dew point of about 47F/9C), but not much lower humidity than that.  (Below 15% RH @ 23C discomfort from the excessive dryness is pretty miserable.) see: http://www.sugartech.co.za/psychro/index.php

The vapor pressure of (a realisitic) 80F dew point air is about 26mm Hg,  the vapor pressure of (also realistic) 47F dew point air is 8mm Hg, for a difference of 18mm Hg. (see: http://www.easycalculation.com/weat...essure.php )

5" of half-pound foam has a permeance of about 10US perms- a US perm is is ~0.659 metric perms, so for every square meter you'd be looking at (10 x 0.659= ) 6.59 grams of water x 18mm Hg = 119g per day.  Over the entire 170 meters of roof that's still (119ml x 170= 20.25 liters (comparable to your estimate).

But that assumes that the vapor retardency was zero for both the metal roof and the roof deck, which is clearly not the case:

A 3/4" wooden plank shiplap deck without roofing felt or other underlayments by itself runs about 1-2 US perms, so even if the metal roofing was vented between the metal and the roof deck (which it isn't) you'd be looking at more like 2-4 liters/day getting through the roof deck, were it fully exposed to the outdoor air.

But since the metal is more or less in contact with the roof deck except for about 10% of the total area (comprised by the standing seams) so now you're looking  at 100-200ml total water going through the roof via vapor diffusion through the entire roof for that worst-case day. That's not a very big latent load at all for a realistic worst-case day, and would add less than 0.25kwh to the cooling load for this worst-case muggy day- your latent loads for ventilation/infiltration will surely be higher.

It's the permeance of the wood deck itself, not the foam that determines how quickly bulk water can dry toward the interior from the space between the metal roofing and the roof deck as long as the finish materials on the interior are reasonably vapor permeable.  Gypsum is >10 perms, latex paint 3-5 perms, which is still more permeable than the wood roof deck itself.  But it's still better to fix any leaks before finishing it out.

If you mount 2x4s to the rafters laterally (either 16" or 24" o.c.) on the inside you can get another 3.5" of open cell foam in there cutting the thermal bridging down to just the cross points, and you'd have a center-R  that meets current code-min, and you'd be able to use standard length drywall screws for the gypsum.

Thanks Dana you are a phenomenon! The wood planks we have under the roof are each spaced about 1" from one another. I could see the underlayment from inside before the spray went up. Now the roofer had another idea: accumulated moisture due to condensation/vapor drive over time. Since XPS has 20x less perms than OCSPF, could that be? I figured the entire area of XPS lets through 1-2 liters per day, hence 20l would take 10-20 days to dry out.... Any validity to that theory? This is getting real interesting here, thanks again for your expertise!

Take the time to figure out Weatherspark - it's a very useful tool.  Select both "dew point" and "temperature" from the graphs pull-down tab, and you can zoom in  to as fine (or out to as coarse) a time period as you like. 

When zoomed out the graph uses blocky averages, but both record & usual-average maxima/minima are highlighted in the background.

Peak relative humidity on a given day is an incomplete datapoint without the corresponding temperature at the time it occurred.  Dew point is the measure of the absolute humidity, and far more relevant, since it doesn't change with temperature.

OK I checked again at weatherspark, great website!


Indeed they report around 50% RH when the weather was hot. HOWEVER according to this resource, temperature doesn't matter:

http://www.numericana.com/answer/gas.htm#perm

HENCE, if you look at weatherspark for May 27-30 for BWI airport, you'll see prolonged periods of 90+ RH in the evening hours at around 70F.
Could it be physicist at numericana.com is right, the temp doesn't matter and the vapor drive takes place at any temp. Only the RH differential counts?

Then the XPS slows down the vapor too much, hence it's stuck in the open cell foam and condensates say the next day when the A/C kicks in....?
Does the OCSPF turn into like a moisture battery when covered on both sides?

OK dew point was 70F the entire night May 28-29 more than 12h, it's on the edge. and let's not forget there could be local fluctuations as we are closer to the bay than BWI. I think we are around 3 miles from the harbor

The vapor pressure is the same for air of a particular dew point temp, independent of the temperature of that air. But it's the difference in vapor pressure ( dew-points) between the bodies of air on both sides of the structure that determine the direction and magnitude of the drive. The RH is always relative (to the temperature) whereas the dew point is not. That's why dew point is the more relevant number. To convert RH to dew point you need to know the temperature at which that RH was measured. Dew point reflects the absolute humidity level of the air independent of it's temperature, and it's absolute humidity that determine the vapor pressure.

Yes there are local fluctuations in dew point, but they're surprisingly small, except during on-shore wind flows near coastlines. Ocean air is in fact drier under hot weather conditions, since the ocean water is well below the dew-point of the overriding body of air, and is absorbing moisture directly lowering the dew point, whereas sun-heated soil is releasing moisture from the soil into the air raising it's dew point under hot weather conditions. The fact that you are closer to the bay than BWI means that there will be periods in hot weather when your air is measurably drier than at BWI, but the converse would only happen during months, when the land is both colder than the ocean and colder than the dew point of the overriding body of air.

The fact that 2" of XPS is relatively low permeance relative to the underlayment + roof decking that keeps the moisture from passing into the interior at a rate high enough to prevent condensing. But without the exterior moisture drive (and I'm convinced it's a bulk-water issue, not vapor drive from the outdoor dew points) it wouldn't much matter. Were it outdoor dew-point only the assembly would still be able to dry through the XPS without it reaching saturation levels (liquid water) within the assembly. If you replace it with 3" of o.c.SPF it would have more than 10x the drying capacity toward the interior than through 2" of XPS, and no condensing events would occur from summertime moisture drives, even incidental water under the metal roofing would dry fairly rapidly toward an air-conditioned interior.

But when you put latex painted gypsum over the foam the paint reduces that to about 5x the drying rate of XPS. It might still be enough to handle it, or not, depending on just how much water is getting under the metal roofing.

I will say the same thing that I said on the MRA....its a leak.

Dana is 100% dead on right here.

Yes I think so too. It's a roof leak 100% beyond a doubt.
I thought let's waste another weekend and took down the top rows of XPS everywhere. The very top was wet as expected but the second row 2ft down from the top was mostly dry.
The ridge at the top soaked up so much water, it drips when you push it lightly...

This morning it struck me why the water found its way it: When they installed the underlayment, they screwed up twice nailing and stapling it. Then the third attempt and layer went on fine. BUT then they left it exposed just like that for 2 weeks.
A tropical storm came and water started coming in. They sent a crew that nailed a blue tarp on the roof. After another week at least they came back and installed the metal roof.
BUT they didn't fix or patch the underlayment. So the underlayment has at least been punctured at least 100+ times

Is that an issue / can this be left that way?

How about we take the f****** roof out and replace it?

Do you want the bad news or the bad news?

The underlayment being riddled with holes is not a good think but the roof should be tight regardless.

Removing standing seam to repair these leaks is no small task and usually requires a full tear off.

OK, what about dew? can it form under the panels and leak inside? Would that be an issue you think?

The roof is direct to deck isn't it?

Very little should be forming at the end of the day.

I know this isn't a legal advice forum but

what would you do in my case considering all of the above we have been talking about?


Is this negligence?


What should I do next, legal action?

You should hire a competent inspector to look at your roof and write a report, no one on the forum can analyze your roof problem long distance. When you have a report that shows what the defects are , you can then send a copy to the roofing contrator via certified mail. giving them "X" (resonable) amount of days to make the repairs or you will take legal action. If they do not respond then seek an attorney.

Hey Chris, thanks a million for the suggestion! So what type of inspector should I use. Like if I open the YP there are a bunch of home inspectors. But if I think of the guy who inspected this house when I bought it, I get an "inspector aversion" ;-) So is there are particular kind of these experts I should be searching by name? The roofing company guy who was here at the site suggested I should hire an "condensation expert" to analyze the situation. But he had no recommendation either. Where can I find these experts in Baltimore MD, how would they be listed in the yellow pages for example?