best way to insulate unvented attic
Last Post 12 Sep 2011 11:58 AM by Dana1. 23 Replies.
Printer Friendly
Sort:
PrevPrev NextNext
You are not authorized to post a reply.
Page 1 of 212 > >>
Author Messages
zygote nycUser is Offline
New Member
New Member
Send Private Message
Posts:17

--
26 Apr 2011 09:43 AM
Guys,

 I'm in the process of doing a gut renovation of my old house. It a brick house built back in the 20's has an unvented flat roof on a slop and it has no insulation. I'm putting in a new TPO roof with 1 1/2" rigid foam then 1/2 dens deck under it. My roofer is asking me if I want it vented or not.

As for the attic, there's is an air space about 1/2'- 2' between the roof joist and ceiling. My spray foam guy suggested to put in 8" of open cell R30 under the roof between the joist.  That way I can keep the space unvented. Is R30 enough? Is this the best way to insulate this space? Do I need to worry about heat and moisture trap between the air space? I need your help in deciding.

Thanks,
Zygtoe
wesUser is Offline
Advanced Member
Advanced Member
Send Private Message
Posts:810

--
26 Apr 2011 12:13 PM
If you are in New York, I would think that R38+ would be better. How tall are your rafters/joists? If possible, it is better to completely encapsulate the rafters, and remove the thermal breaks that these can cause.
Wes Shelby<br>Design Systems Group<br>Murray KY<br>[email protected]
zygote nycUser is Offline
New Member
New Member
Send Private Message
Posts:17

--
26 Apr 2011 01:05 PM
Wes,

The joists are 3x8. should I add batts or cellulose in the space between the joist and finish ceiling?

thanks,
zygote
Dana1User is Offline
Senior Member
Senior Member
Send Private Message
Posts:6991

--
26 Apr 2011 02:20 PM
To go un-vented you need to have enough R value on the exterior relative to the 'tween-rafters insulation such that the binned-hourly mean temp of the roof deck in January doesn't drop below 30-40F (the dew point of 68-70F, 30-35% relative humidity conditioned space air.) If it's allowed to run much colder the moisture content of the roof deck & rafters rise, inducing a rot/mold risk.

The mean January outdoor temp in NYC is ~ 30F (see: http://www.climate-zone.com/climate/united-states/new-york/new-york-%28jfk-airport%29/ and http://www.climate-zone.com/climate/united-states/new-york/new-york-central-park/ ) So for the roof deck to average 40F with a 70F interior 1/4 of the total R needs to be in the foam above the roof deck. For the roof deck to average 40F with a 60F conditioned interior the ratio becomes 1/3. Splitting the difference, something like 30% of the the R as exterior foam would deliver some margin, but with some tweaks you could safely back off to 20% or even less:

Spraying the underside of the flat-roof deck with 1-2" of closed-cell spray polyurethane protects the roof deck by making the spray foam, not the wood the condensing surface. Closed cell foam won't wick moisture toward the wood, so it doesn't suck up the moisture the way it does when the wood itself is the condensing surface. Also, 1-2" of most closed cell foam (read the specs though), becomes a 0.75-1.25 perm vapor retarder, limiting the rate at which moisture reacheds the wood via vapor diffusion, but is permeable enough that the roof deck can dry toward the interior when outdoor temps are warmer.

Then, filling the rest cellulose as the rest of the cavity fill buffers seasonal moisture, and actively wicks moisture away from the wood, storing the seasonal load in the insulation, not the structural wood.

Assuming the rigid foam is XPS or iso (polyisocyanurate) the 1.5" of foam is worth ~R8 when it's cold out, if EPS, ~R7, to which the DensDeck adds another R0.5. For the sake of argument call it R7.5.

Without interior closed-cell on the roof deck, at the 25/75 ratio that would limit the interior R to ~R22, which would be ~6.5" of open cell foam or cellulose, but assuming you have full-dimension 3x8s that would be wasting 1.5" of space, and put the total R at ~ R30.

If you add an inch of closed cell to the interior that would add ~ R6 to the stackup, leaving 7" (~R25) for the cellulose, for R7.5 + R6 +R25=R38.5 total, which would be more like code, but with only ~20% of the total R on the exterior.

For more margin on wintertime condensation and higher R/lower cost, add an inch or two of EPS above the roof deck, which will also serve as a better thermal break against the substantial bridging of those 3x8 beams, and skip the interior spray foam, load it up with cellulose. At 8" the cellulose would be approaching R30, so you'd need a minimum of R10 above as foam,( R15 would be even better). At 2.5" EPS would deliver ~R10, at 3.5" it would be ~R14. Buying R-value as EPS is far cheaper per unit R than with closed cell foam, and putting it on the exterior does more good, since it adds to the thermal break over the rafters, whereas spraying between the rafters does not.

If 2.5-3.5" creates other issues, use iso. At 1.5" iso delivers ~ R8.5-9, and at 2" would be ~ R11-12, derated for 25F wintertime temps. Polystyrene increases in K-value with lower temp, so XPS @ 25F would be running ~ R8 for 1.5", R10.5-11 for 2", so you wouldn't need more than 2" to get there.

Also, forget-batts on an antique like this- cutting them to fit properly would be an endless chore. Wet-spray cellulose (at any density) or new-school high-density fiber such as JM Spider or Certainteed Optima (@1.8lbs density) using "blown-in-bag" or "blown in blanket" mesh would be far more effective, since it would fill in every knot-hole & gap eliminating losses in R value due to convection currents. Cellulose would outperform fiberglass somewhat in the summer cooling season due to it's higher thermal mass. Adding 7" of cellulose at 2.5lbs density would be like adding another 1/4" or more to the gypsum.

On the sloped roof/conditioned attic approach you can still get there, but again you CANNOT use open cell directly against the roof deck or it will be at higher risk of rot from wintertime moisture absorption. But with 1-2" of closed cell to protect it you can fill out the rest of the R with blown-in-blanket wet spray cellulose, or with open cell foam. Read this document carefullyt:

http://www.buildingscience.com/documents/reports/rr-1001-moisture-safe-unvented-wood-roof-systems

(In particular, look at tables 3 & 4 Boston is only ~2F colder in January than NYC, making it the best-comparison.)

In any of these stackups you should take care to keep the interior finish above 1 perm - no kraft faced batts, poly sheeting, oil paints etc. Latex & latex-acryilic paints are fine.
zygote nycUser is Offline
New Member
New Member
Send Private Message
Posts:17

--
12 May 2011 10:51 AM
Dana1,

thanks for the detail post. So I'm going to ask my roofer to put in 2-3" iso/eps if possbile. Than 2" close cell to cover all the surface under the joist . The rest of the space between roof and ceiling floor to be fill with cellulose. Please correct me if this is not best way to it.

Also for my surrounding walls, My foam guys suggesting 4" Icynene open cell. I'm wandering if it's better if I put in 2" heatlok 217 close cell if cost is not much difference. 

Regards,
Keith
Dana1User is Offline
Senior Member
Senior Member
Send Private Message
Posts:6991

--
12 May 2011 12:28 PM
On the roof stackup, you get more performance bang per foam $ putting more R above the roof deck. With 3" of rigid iso you'd have R17+ of thermal break over the rafters, and you could skip the interior spray foam, stuff it with 8" of cellulose for another R28, and have a true R40 (whole assembly R, with all thermal bridging included, not center cavity R) solution for less money.

With 2" of iso would be only R11 on the coldest hours of the coldest days and you'd be running ~R39 center cavity with an all-cellulose fill, which would still meet code, but your whole-assembly R would be in the low-mid 30s.

3" of EPS 2" of iso) above the roof deck would be ~R13+ then, and is much cheaper than iso. In that case a full -dimension 8" cavity fill of cellulose-only would give you R43 center cavity, and a whole-assembly R in the mid-30s, a slight performance boost over 2" of iso, for less money.

If you dropped back to only 2" on the rigid using EPS you'd then HAVE to use an inch of closed cell on the interior to protect the roof deck, since with only ~R8 on the exterior and R28 in the cavity your exterior to cavity R ratio would only be 22%, below your 25% minimum. Adding just an inch of closed cell on the interior of the roof deck that would raise your center-R by about R2.5 to ~ R38+, but your whole-assembly R by about R1-1.5. But it also puts about 1 perm of vapor retardency between the interior and the susceptible wood, sufficiently slowing the uptake of winter moisture while still letting it dry seasonally, and puts the condensing surface (the interior surface of the spray foam) at the 35%/65% point- where the condensing hours will be few, and the foam won't wick moisture toward the wood.

To get to the optimal solution on the walls we need to know, is there a vented gap between the brick and the studwall sheathing? And is there any sort of felt or rosin paper or ??? on the exterior of the sheathing, between the sheathing and the brick?
zygote nycUser is Offline
New Member
New Member
Send Private Message
Posts:17

--
12 May 2011 02:30 PM
Dana1,

If i go with the first suggestion, do i need to worry about moisture going up to the rafter and rot?

As for the walls, it's only 10-12" thick brick wall.  There's no insulation at all. I'm thinking of using 2" close cell may help keep water from coming in also.

Thanks
Dana1User is Offline
Senior Member
Senior Member
Send Private Message
Posts:6991

--
12 May 2011 02:43 PM
Posted By zygote nyc on 12 May 2011 02:30 PM
Dana1,

If i go with the first suggestion, do i need to worry about moisture going up to the rafter and rot?

As for the walls, it's only 10-12" thick brick wall.  There's no insulation at all. I'm thinking of using 2" close cell may help keep water from coming in also.

Thanks

With 2" of closed cell on the underside of the roof deck, 2 to 3 inches of either EPS or iso above the roof deck, and filling the rest of the rafter bay with blown fiber.

No, you don't need to worry about interior moisture drives ruining the rafters or roof deck.  A 2" shot of generic closed cell foam gives about 0.5-0.6 perms of vapor retarder to slow down wintertime vapor drives, and the place within the assembly that averages below the dew point of the interior air in January is within the foam, where the interior air can't penetrate.

Is the brick wall a cavity wall, with either a full wythe or a narrow cavity between the exterior wythe of brick and the rest, or is it solid?
Dana1User is Offline
Senior Member
Senior Member
Send Private Message
Posts:6991

--
12 May 2011 04:59 PM
After doing a bit of literature search and some rough calculating...

If you have a solid masonry wall UNDER NO CIRCUMSTANCES should you use lo-perm Heatlock 217, which has WAY too low a perm rating for this application (0.4 perms @ 1.25" see: http://www.goallfoam.com/pdf/HEATLO...20Data.pdf ).

If you go too low in perm rating (less permeable to water vapor) there's a higher risk of freeze/thaw damage to the brick & mortar and to the wood interfaces since the brick will not be able to dry toward the interior, and run higher moisture content (both winter and summer.) But if it's too permeable wintertime vapor diffusion will also cause damage- it's a balancing act, but between 0.5 & 2 perms works-mostly for climate zones 4-6 (NYC is on the edge of 4 & 5).

At 2" you may use other 2lb foams of higher perm rating though, but you still have to keep the area where wood members interface with the masonry above 1 perm, but below 2 perms. HeatLok Soy (see: http://www.demilecusa.com/Repository/File/HEATLOK_SOY_TDS_102709_2.pdf ) runs ~ 1.2 perms at 1"which means it'll be about 0.6 perms at 2", so you''d have to thin it out to an inch near any joist/rafter/window frame etc interface. Any more than 2" becomes too impermeable, putting the masonry at risk of freeze/thaw spalling and cracking. That would leave you at ~ R13 nominal, but with R6-ish spots everywhere, which is less than ideal.

A better solution is to instead use three inches of Icynene MD-R-200 (and not MD-C-200), which is a high-perm closed cell foam 1.3 perms @ 3". see:

http://www.icynene.com/assets/docum...ets-US.pdf

In $/R terms the high-perm Icyene is usually cost-competitive, but it's only R5/inch compared to R6+/inch for other closed cell foams. A 3" shot would give you R15 for about the same cost as 2" (~R13) of other vendors' goods, but you're giving up another inch of inteior space. But that R15 is also an UNBROKEN R15, (assuming you don't put studs between it, only narrow studs or furring on the interior to mount the interior gypsum), which is about the same whole-wall R value of R20 cellulose in a 2x6" 24" o.c. studwall.

A third alternative is to use 1" of HeatLok Soy (or similar ~ 1-1.2perm @ 1" foam) on the brick, and build an interior 2x4" studwall for hanging the gypsum, with spray cellulose filling the stud bays, and NO interior vapor retarder. In that assembly the R6.5 foam + R1 masonry puts R7.5 on the exterior of the ~ R13 center cavity R of the studwall, so in average January interior-air dewpoint layer occurs inside the foam, where that air can't reach. As long as the foam is air tight to the brick, there isn't enough moisture reaching the brick from the interior to create a problem via vapor diffusion alone, and the stud edges will stay above the dew point with just 1" of foam. This assembly would give you about R15-R16 for a whole-wall R for less money than 3" of high-perm closed cell Icynene.

If yours is a cavity wall with top & bottom vents to the outside ventilating the cavity gap you can relax the perm ratings a bit, but ~1-1.5 perms is still fine. The ventilation cavity keeps exterior wetting from wicking in, and gives both the interior & exterior wythes higher drying capacity.

Open cell foams are too high-perm (unless you're really go nuts on it, like 10" or more), and you'd have to back off the vapor diffusion with other methods, none of which seem as reliable long term as a 1, 2, or 3" (depending on foam type) closed cell directly on the brick.

Also note: At 2" HeatLok 217 would also be a bit too low-perm for the underside of the roof deck too, but it would fine at 1". HeatLok Soy or similar would be OK at either 1" or 2".

In general, closed cell foam is an EXPENSIVE way to go high-R, but it's great way to air-seal and adjust vapor permeance though. Anything less than an inch is too difficult to apply consistently, but an inch is still plenty for air-sealing. So when you don't HAVE to go thick, you can build a higher-R structure for the same money keeping it to an inch (which it appears you can.) At a buck a board-foot or more the difference between a 1" shot and a 2" shot is substantial.

zygote nycUser is Offline
New Member
New Member
Send Private Message
Posts:17

--
13 May 2011 01:31 PM
Dana1,

I'm thinking of not using 2" close cell under the rafter. Just 3" iso above and at lease 14" cellulose to fill up the space.

As for the wall it's solid brick.

Regards,
zygote
Dana1User is Offline
Senior Member
Senior Member
Send Private Message
Posts:6991

--
13 May 2011 02:42 PM
Posted By zygote nyc on 13 May 2011 01:31 PM
Dana1,

I'm thinking of not using 2" close cell under the rafter. Just 3" iso above and at lease 14" cellulose to fill up the space.

As for the wall it's solid brick.

Regards,
zygote

My bad- on the quick re-read of your roof joist dimensions I'd misconstrued that you only had 8" of rafter depth to fill, not 14".

8" of cellulose would be (not counting the iso) ~R28, which is where I had done the dew point calculation.

14" of cellulose would be ~R50- time for  a quick re-calc.

With 3" of iso+ 1/2" DensDeck above the structural deck you'll deck that would be at ~ R18 above the roof, deck, so you'd still have over 25% of the total-R with R50 in cellulose, but you'd be approaching it- the green-tinged edge of the yellow warning light.  With only 2" of iso, or if you used 3" EPS or XPS you on the orange tinged edge and a protective inch of spray would be called for.

At 14" cellulose at low density is going to run about 1.8lb per square foot,  if dense-packed it'll be over 3.5lbs per square foot, so supporting it with just gypsum you'd have bowing issues over the long term- a layer of 1/4"-3/8" OSB or ply might be in order, particularly if the spans are more than 16".   If you backed off a couple of inches in total depth to only 12" you could use a blown-in-mesh system to support the cellulose, and have ~2" of room to run your electrical etc. between the gypsum & mesh, and have a little bit of margin on the dew-point calc.  You'd still be at ~R60 center-cavity, and R55-ish for the whole-assembly R.

For a bit more margin, where possible, extend the rigid insulation down the exterior side of the walls to about the depth the fiber insulation, which will keep average temps at the edges of the roof deck & rafters above the average January dew point too. With cellulose fully in contact with that wood the risk is low, since the cellulose wicks condensation away from the wood, but low isn't zero.  Putting 3" of iso over edge of the fiber layer as well as the top brings the risk to zero:

http://www.buildingscience.com/docu...01_web.jpg

If that's not possible or if it's easier,  running your ~1-perm closed cell wall foam all the way up to the roof deck continuing onto the underside of the roof deck for 10-12" in will protect it, without having to foam the entire roof deck.

(See my previous post detailing wall insulation recommendations for solid brick walls.)




zygote nycUser is Offline
New Member
New Member
Send Private Message
Posts:17

--
14 May 2011 02:31 AM
Dana1,

for my brick wall, you are recommending either 3" open cell or 1" close cell heatlok correct? My foam guy told me the minimum is 2" for close cell due it expansion.

thanks.
Dana1User is Offline
Senior Member
Senior Member
Send Private Message
Posts:6991

--
16 May 2011 02:42 PM
Posted By zygote nyc on 14 May 2011 02:31 AM
Dana1,

for my brick wall, you are recommending either 3" open cell or 1" close cell heatlok correct? My foam guy told me the minimum is 2" for close cell due it expansion.

thanks.
Open cell products are not sufficiently vapor retardent do not use.

It's 3" of Icynene MD-R-200, which is a 2lb density closed cell product, and nearly ideal at 3"...

...or...

...1" HeatLok Soy and (definitely not HeatLok 217, which would be asking for trouble).  

With the Heatlok Soy you could use 2" everywhere and try to thin it out around where the wood hits the brick and still be OK.   At 2" it's going to be down around 0.5 perms or lower, which could lead to rotting out the wood due higher moisture content in the masonry.

Most of the time closed cell foam is applied in 2" lifts, so the installers are well tuned to shooting 2" foam, but 1" is possible, even if it's a bit inconsistent in depth.  If your foam guy really can't shoot 1" foam (most can), unfaced 1" XPS rigid foam sheathing against the walls and sealing the seams & edges with 1-part foam, fattening out the R value with a studwall with unfaced-batts or spray cellulose works.

Your goal here is to put a ~1-2 perm nominal vapor retarder and a modest amount of R (~R4 is enough) between the brick and an interior insulated studwall.   Going thicker with the foam the permeance drops, which limits the ability of the masonry to dry toward the interior.  But if you go with high-perm open cell you go too much moisture gets in during the winter putting the wood at risk.  In either case, when the annual average moisture content in the masonry goes up, any wood in contact with it at higher risk of mold/rot.  In winter a high moisture content in the brick can damage the masonry during  freeze/thaw cycles, forming cracks, spalling of the brick face, or mortar failure.

The high-perm closed cell Icynene is rated 1.3 perms @ 3", which is nearly ideal from a moisture point of view (but only R15 as opposed to R19 @ 3" for other 2lb foams.)

At 1" Heatlock Soy is rated 1.2 perms, also ideal, but too vapor-tight at 2".

Foamular 150 XPS sheathing is 1.1 perms, at 1" which is just fine, but also too tight at 2".

At 1.25" HeatLok 217 is rated 0.4 perms- already too low (and at 2" forget about it.)

JM Corbond closed cell foam is 0.6 perms @ 1.5", which works at 1" (or even 1.5") but is pushing the limit at 2".

Demilec Sealection 500 open cell is 6.6 perms at 3.5", (3.3 perms even at 7") which is way too high.

The slightly higher density Agribalance open cell is 5 perms at 5" thickness, also way too high.

Any number of 2lb closed cell polyurethane products come in at between 0.8-1.2 perms at 1", but only a very few are over 1 perm at 2".  The only 2lb foam I'm aware of that's still over 1 perm at 2" is MD-R-200.

Dana1User is Offline
Senior Member
Senior Member
Send Private Message
Posts:6991

--
16 May 2011 05:23 PM
Start your own thread rather than hijacking this one, eh? There's a whole forum on this site dedicated to ICF construction, with plenty of discussion, but they have no bearing on retrofitting brick building in NYC. (I'm a fan of ICF for foundations and for moderate-R walls on 1-story structures in hurricane zones. Lots of pluses, but it's not a cheap way to build or to go high-R.) See: http://greenbuildingweb.com/Forums/tabid/53/afv/topicsview/aff/4/Default.aspx
zygote nycUser is Offline
New Member
New Member
Send Private Message
Posts:17

--
17 May 2011 08:09 PM
Dana1,
My main objective is to get the house as tie as possible. is there's another options I can take without giving up too much space? One of wall for the stair to the second floor is only 32 inch wide. it's already very tie. So that one of my reason i try to keep it at 2" cc foam.
Dana1User is Offline
Senior Member
Senior Member
Send Private Message
Posts:6991

--
18 May 2011 03:44 PM
If you go with HeatLok Soy or another 1.2 perm @ 1" solution you shouldn't go MORE than 2", (but 1" would still be OK.)

At 2" you'd be at ~R13 for a whole-wall R (similar performance to a 2x6" timber-framed structure with R19 batts). Limit the areas where it's thinned out to just 1" as much as possible (eg: only the 6-12" around where structural wood is contacting masonry.)
Chloe TaylorUser is Offline
New Member
New Member
Send Private Message
Posts:89

--
23 May 2011 02:07 PM
Posted By zygote nyc on 26 Apr 2011 01:05 PM
Wes,

The joists are 3x8. should I add batts or cellulose in the space between the joist and finish ceiling?

thanks,
zygote
Surely a great way to provide insulation, no doubt....
<a href="http://www.capitalsteelbuildings.co.uk">Commercial Buildings for Sale</a>
jumpingspidermediaUser is Offline
Basic Member
Basic Member
Send Private Message
Posts:104

--
24 May 2011 06:42 PM
A roof system insulated with spray foam reduces energy several ways. Energy loss from ducts located in the attic is essentially eliminated.

<a href="http://www.londonccc.co.uk/">Carbon credits</a> | <a href="http://www.londonccc.co.uk/content/self-invested-personal-pension">Carbon investment</a> | <a href="http://www.londonccc.co.uk/content/what-are-carbon-credits">Carbon trader</a>
Chloe TaylorUser is Offline
New Member
New Member
Send Private Message
Posts:89

--
25 May 2011 07:57 PM
Posted By jumpingspidermedia on 24 May 2011 06:42 PM
A roof system insulated with spray foam reduces energy several ways. Energy loss from ducts located in the attic is essentially eliminated.


___________
SIPs

Surely building a roof is one of the best ways of providing insulation. But the question remains the same going for and building a roof out of what......? 
<a href="http://www.capitalsteelbuildings.co.uk">Commercial Buildings for Sale</a>
zygote nycUser is Offline
New Member
New Member
Send Private Message
Posts:17

--
26 Jun 2011 06:58 PM
Dana1,

I forget to mention my house is semi detach. Is it recommend to insulate the connecting wall also with foam? My foam guy told me it's not needed.

Regards,
zygote
You are not authorized to post a reply.
Page 1 of 212 > >>


Active Forums 4.1
Membership Membership: Latest New User Latest: hudson2000 New Today New Today: 0 New Yesterday New Yesterday: 2 User Count Overall: 34707
People Online People Online: Visitors Visitors: 161 Members Members: 0 Total Total: 161
Copyright 2011 by BuildCentral, Inc.   Terms Of Use  Privacy Statement