Hello,

I am trying to figure out my roof and insulation strategy on a frame flat roof building in Jersey City, New Jersey. 

Does anyone have advice on:

1) Applying foam to underside of roof deck.  Amount? Closed or open cell?  (This tactic would allow me to create a lot more floor space in upper apartments, as there is a 5-8 foot  attic.)

2) doing a spray foam roof over a torch down? 

3) If there is a chance of water penetrating basement walls, can i still spray with 2" of closed cell foam?

Thank you!

It's usually far cheaper per unit R value to use rigid EPS under a fresh torch-down/EPDM membrane than it is to spray foam from the underside. R20-R24 (5-6") of rigid EPS would be the minimum to use with that approach, but if you're paying the utilities more (or attic floor insulation in addition to EPS above the roof deck) may still be cost effective. EPS vendors also make roofing insulation for the upper layer with taper built-in to design in improved drainage on flat roofs as well.

If you don't want to rip up & replace the roof deck, sealing the attic floor with 2" of closed cell foam and loading 6-12" of cellulose on top of that is probably cost effective. Under the roof deck 2" of closed cell foam would give you about R13, but more than that may run into issues with trapping moisture within the roof deck. (If it's metal, have at it, if wood, take caution.) At 4" you'd be at R25-ish, but with the thermal bridging of the structural framing it wouldn't perform nearly as well as R25 above the roof deck.

It's best to deal with bulk-water drainage issues first, but yes you can put a couple inches on the walls there too. If you have high hydrostatic pressure during flood conditions it can potentially separate the foam from the wall though. Closed cell foam is pretty waterproof- it can take days/weeks/months of submersion without taking on much water, where open cell stuff would saturate.

Hi Dana,

I'm not sure I can build up with the EPS because the deck doesn't sit more than say 3" beneath the height of the edge flashing... Then again, if I rip of the torch down(s!), I may have plenty of space. If I got up to say 6", what would you do under the roof? I'm going to have a space with 15-18 foot ceilings to heat/cool. The other option is the 2" foam roof (R14), plus open cell spray foam under...

Bisseti

Posted By Bisseti on 09 Mar 2010 12:56 AM
Hi Dana,

I'm not sure I can build up with the EPS because the deck doesn't sit more than say 3" beneath the height of the edge flashing... Then again, if I rip of the torch down(s!), I may have plenty of space. If I got up to say 6", what would you do under the roof? I'm going to have a space with 15-18 foot ceilings to heat/cool. The other option is the 2" foam roof (R14), plus open cell spray foam under...

Bisseti

If you wanted to spray more insulation on the underside of the roof deck in combination with EPS above you can, but it's more expensive per unit-R.   Open cell is preferred since it won't trap leak moisture at the roof deck (but roofs never leak, right? :-) ), and make locating leaks easier than with the more waterproof closed cell foam.

If the exterior walls in the attic space are uninsulated, air-sealing them with a 1-2" flash of foam (o.c. or c.c.) and and insulating to the same or higher R-value of the roof with cheaper fiber insulation works.  The only additional heating & cooling load to the "...15-18 foo ceilings to heat/cool" is the 5-8' of exterior wall. If the walls of the conditioned space below are insulated to R13 and have windows, etc, for a whole wall R value of R9-ish, insulating windowless walls of the attic space to R25+ pretty much takes it out of the heat gain/loss equation.  It's a load sure, but a small fraction of the gain/loss of the conditioned space below.

You can also split the R between the roof deck and the attic floor, using cheap fiber at the attic floor level.  When doing this it's important to not overdo the fiber (up to 50% of the total R as fiber is probably OK in your climate), since the intervening attic space becomes cooler, making it more susceptible to condensation.  If going this route the R-value of the exterior attic walls is an important factor, since the lower it is, the less you can get away with at the attic floor. If you make the R-value of the walls the same as the combined R of the attic floor & roof deck, going as high as 50% of the total as attic-floor fiber is fine.

In horizontal warm-side down applications not all fiber insulations are created equal, particularly at R25 or less. Cellulose has much lower intra-insulation convection than fiberglass or rock wool, both of which lose R value the bigger the difference is between conditioned space and attic space temperatures. But FG weighs a lot less than cellulose, and if there's any question of load bearing capacity, go with the lighter goods.  There's evidence that topping a blown fiberglass layer with 3" of cellulose is enough to pretty much "cure" the R-value loss issue.  Also, high end finer blowing wools available these days have less of the issue than the stuff that made the issue infamous in the '80s & '90s.  Avoid cheap low-density batts in this application though (or top 'em with at least 3" of cellulose.)

What type of insulation do you have now?  Is it under the existing roofing membrane?

Bruce

Hi Dana, Bruce,

As I'm leaning toward REMOVING the existing ceiling on the 3rd floor of this building, the conditioned space on the third floor will be 15 to 18 feet high and about 13 wide by 63 long. This is why I'm trying to find the best roof/rafter insulation strategy, instead of sealing the attic floor (which is essentially a lathe and plaster ceiling held up by widely spaced rough cut 2x3s).

I had a foam roof contractor out and she said we need to tear off the torch down... If we do the tear off we can add the EPS so we can reduce the amount we need under. Do I have a vapor/condensation issue with the open cell or icynene foam under the rafters if I either use the EPS or foam application above the deck? Other suggestions for under rafters?

I like the idea of doing the exterior walls of "flash spraying with one inch of foam" and then adding bats.



There is no insulation in the building now. It's a frame structure, built in 1907, framed with 4 inch walls, 2x8 floor joists, stone foundation, 5'6" basement, and a mildly sloped flat torch down decaying roof.

I cannot remove the walls in the 1st floor stores (beautiful tin and wood) so I won't be able to go the sealed envelope route. I've gutted the exterior walls, except for 2/3 of the store walls and the walls on about 2/3 of the 2nd story which are in good condition. I'm looking at blown in cellulose in walls I don't open. I've also looked at icynene foam injection but that's expensive...

No matter how much EPS you have above the roof you can safely use up to 2" of closed cell foam under the roof deck from a vapor control point of view, but you can go much deeper (and higher R) with open cell foam. Unless you're very space restricted at the roof deck the open cell is preferable, since it won't spread water in the event of a roof leak- it'll drip pretty much where the water is coming in, making it easier to find, and the higher vapor-permeability of open cell foam ensures that the roof deck can dry fairly quickly toward the interior (whereas with closed cell foam drying will take much longer.)

Both EPS and open cell spray foam are vapor-permeable yet form good air barriers. Without air being able to reach the roof deck, interior moisture only reaches if via permeating through the foam. In Jersey City the number of winter hours where the roof deck is below the dew point of the interior air are for the most part fleeting and few, and the more EPS you have above the roof deck, the fewer that is. There's almost NO chance of interior air moisture saturating the roof deck from below from vapor permeation, and with a semi-permeable interior insulation any moisture that does find it's way to the roof deck will dry toward the interior just fine as the season warms up, or even before, as even on a cold day that never exceeds freezing the sun can raise the exterior roof temp to WELL above room temp dew points. With the roof deck in some intermediate layer it'll be warmer still.

Average temps in Jersey City in January/February are in the low 30s, with average daily highs in the high 30s. The dewpoint of 30-35% relative humidity 68F air is 35-40F. As long as the wooden roof deck never stagnates for days on end at temps below 40F vapor diffusion through the interior insulation can't hurt it. If half the R value is above the roof deck, the average mid-winter temp of the roof deck will be in the ~50F range- plenty of margin. But risk of mold/rot begins to increase when the EPS is less than 1/4 of the total R, and the average winter roof deck temp is under 40F. The more R-value you have above the roof deck the better.

Rather than flash & batt, if you have a cellulose contractor already involved, wet-spray cellulose over the foam instead of batts will be higher performance for (usually) similar money. Unlike batts, sprayed insulation will conform to all irregularites and can fill all the odd sized (and even micro) cavites from odd-ball 100 year old framing to near-perfection. Wet spray cellulose needs a few days to dry, but will never settle due to the water-activated adhesives.

On your dry-blown cavites in the walls you won't gut, dense-packing is far preferable to 2-hole method. At 3lbs/ft^3 or higher cellulose won't settle, and the internal air flow/convection is cut by over 90% (it can be nearly as good as a foam seal.) Dense packing uses ~50% more material than 2-hole method, but it's not anything like 50% more expensive. R-values are about the same with either- you're buying greater longevity and lower air infiltration with dense packing. Specify "suflate free" or "borate only" material for the dry blow too. The sulfate fire retardents can corrode copper, some steels when wet. (Wet spray cellulose is always sulfate free, and can be dry-blown as well.)

Your desciption is helpful.

The idea of a foam roof in your application sounds appealing.  I will defer to Dana1 on the insulation issues, but I don't see a problem with an open cell insulation on the underside of the roof and a foar roof on top.

Bruce

Posted By Bruce Frey on 10 Mar 2010 04:42 PM
Your desciption is helpful.

The idea of a foam roof in your application sounds appealing.  I will defer to Dana1 on the insulation issues, but I don't see a problem with an open cell insulation on the underside of the roof and a foar roof on top.

Bruce

Nor do I- open cell is preferable to closed cell in this application. (Hopefully that came through in my verbose-mode-enabled prior post!?!) 

Open cell foam under a roof deck only becomes an issue in much colder climates and then only with little or no insulation above the roof deck causing the deck itself to stagnate at temps below the dew point of the conditioned space air.  The risk of diffused vapor condensation is fairly low in NJ, even with no foam above (but real in ND.)  But in any heating climate zone the greater fraction of the R-value installed above the roof deck, the warmer and drier the roof deck will be.  A 50-50 split reduces the risk to zero in NJ, but in the Canadian midwest 60-40 or 70-30 might be required.

Interior side vapor retarders can mitigate the issue in many assemblies, but most flat roof materials are too vapor retardent  (class-I or class-II) , so interior vapor retarders would make it a double-vapor barrier moisture trap, and can't be used.  With no insulation above the roof deck it would need to be ventilated between the roof deck and insulation layer to protect the deck (hard to do with spray foam, eh? ;-) ) But with insulation above the deck flat roofs can be sprayed, but not to a depth any greater than what forms a class-III vapor retarder (2" is the limit with 2lb foam, but IIRC you get about a foot with half-pound stuff), and in colder areas you  still need to keep the average temp of the roof deck above the dew point of the interior air so it can dry toward the interior.

lots of good advice...

I like the open cell idea