Hello all,

I have been searching these forums and attempting to digest various building science articles as well as other resources and I cannot decide the best path. Here is the rundown.

1928 home, Region 6 (Mpls) 
Stick-framed with what appears to be 2x4 (can obtain photos)  lower pitch ~4-12 or 5-12.
Currently vented with soffit vents and active spinner.
blown-in cellulose ~4"-6" (uneven fill)
Air conditioner blower and vents in attic. (non-insulated) 

I am interested in tightening up the envelope of the house and will be starting with the rim joists and the attic spaces.  I am on the fence about the attic, specifically maintaining a vented configuration or going with unvented in my region.  The air conditioner and vents/hvac are not currently boxed with any insulation above/over them, also the ceiling penetrations (lights/fan fixtures) are not sealed. 

Given the low pitch I was considering going with a CC spray foam on the underside of the roof deck and then when I reroof, add taped xps above the roof deck, under a membrane.    I still have concerns re:snow load and ice damns in my region as well as load on the stick-framed roof.   I suppose an alternative to the XPS above the deck is to trim out the 2x4 to deepend joists (2x6/2x8) and run batts behind the CC foam.  Also concerned about roof framing loads.

The alternative would be maintian vented configuration, build enclosures for the air handler and venting and seal the ceiling penetrations, install stealth vents near hip-peak and remove the active vent (want to eliminate risk of active draw of conditioned air).  However, in this configuration there would still be at least some leakage from the access to the attic space.  I had thought about 1" CC followed by blown-in, but wonder if I should be concerned about ceiling loads.

As I mentioned it seems that in some publications, it indicates that in region 6 and above there is a need to maintain the venting in order to reduce any snow-laod/melt situations.  However, other publications seem to suggest CC foam is sufficient.

I understand that in the CC unvented roof humidity will have to be controlled.

Any thoughts on current "best-practice" or suggestions RE the attic space assembly?  Thanks in advance.

I'm assumeing rafters & joists are full-dimension 2x4s, not something something bigger? What type of span are we talking? Are they 16" on center or something else?

Rather than 2lb XPS above the roof deck you can get more R per POUND out of iso or EPS, either of which could live under a membrane. A couple inches of 2lb foam under the roof deck and sealing the venting would roughly double your current effective R value, but in your climate you're likely to need to then install a dehumidifier in the attic space to keep condensation under control in winter. If you later put R25-R30 above the roof deck the dehumidifier can be decommissioned, since even with the R15-ish attic floor insulation the attic would then stay above the dew point of the conditioned space air.

It may be possible to retro-engineer the rafters & joists into a truss system capable of handling more load, but you can also get a lot of mileage out of making the roof deck into a structural monocoque (which it probably is already to some degree, which is probably how it has survived 80+ years.) You might look into designing a monocoque of R25-R30 SIPs to go over the existing roof deck, but even if you used R25 nailbase iso panels like those from Atlas or Hunter properly through-screwed into rafters the nailerbase OSB would be adding significant structure. If you do the project in one fell swoop you could cheap out and use 3" of half-pound foam (~R10-R11) under the roof deck as your air-seal rather than 2lb foam, but 2" of 2lb foam would still protect the roof deck and rafters better. Consulting with a structural engineer ahead of time might be a worthwhile investment, but putting most of the R above the roof deck is likely your best option both from a structural and condensation-control point of view. Even 2" of cc foam under the roof deck would be HUGE from an ice-damming point of view, and would add a modest amount of structure.

No matter what you do with the venting, it's still worth air-sealing all of the penetrations through the attic floor, but that will be critically important from a heating economy point of view if left as an vented design, since the turbine vent is actively depressurizing the attic.

No matter what you do with the other insulation or venting, it's still worth insulating at least the supply ducts and air handler on the AC unit, and sealing the seams & joints of both supply & return ducts with duct mastic.

Thanks for the reply.  I just found the original blueprints for the circa-1931 house.  The attic joists are 2x6, 16"oc and the rafters are 2x4", 16 oc.  full-dimension.  The pitch is a more spacious 5-12, different than my prior recollection.  The main roof is a 24-8x 28-8 area.

I would likely have to do this project in one fell swoop as the roof is showing significant signs of age, including cupping of the shingles.  Thus, it really comes down to the vented/unvented choice.  I'm leaning heavily to the unvented design given that the air handler is situated in the attic spaces.  In the the unvented assemblies I was planning on removing the existing insulation as to move the boundary of the conitioned space to the underside of the roof deck.  Is this the right choice or would it be beneficial to leave the existing insulation in place? 

I had considered doing only ccspf on the underside of the roof deck, as per BSC rr-1001-moisture-safe-unvented-wood-roof-systems.  However, I am somewhat concerned with the snow-load ice-dam potential at the rafter edges/soffits, hence the idea of using the ventilated nail base.  However, I may be confused on the ice dam issues.  Additionally, I was somewhat concerned with a underside solution only as there would still be heat/moisture transfer at the wood/wood interfaces correct?

That being said, an underside-only solution would be simple to implement.  Thus, I would likely attempt to "use 2 passes (2 or 2” each) of ccSPF (about R24) followed by 6” of spray fiberglass (about R25) in a reasonably-sized space"  I would have to structurally support fur-outs to allow 6" of fiberglass against the foam on the underside of the decking. (strucutral engineer suggestion appreciated)

The concept of using full sip as a structural element is interesting however, I'm not sure about the ROI of that route?  It seems that the 4'x8' sheets of the ventilated nail base are not outrageously priced.  Thus, the other option I'm weighing is the 3.5" panels above the roof deck, which would cut down on thermal bridging, combined with a 2" ccspf under the deck for the air barrier. 

I suppose the cheapest method would be to spray 2" ccspf on the attic floor after boxing the air handler and install  blocking at the rafters and then fill the attic with loose fill up to the R49.  I would have to remove the turbine vent and would likely install the "stealth" vent near the hip ridge to insure sufficient ventilation.  However, with this route I am somewhat concerned with the attic access panel still providing a "hole" for conditioned air to flow into the attic.

Any additional thoughts/comments are appreciated.

Insulating the roof deck from below will still mitgate the ice damming issue by a huge factor, since the roof deck will run both cooler and with a more even temperature. In it's current state heat leaks from the interior are large and cause thinning/melt-out on warm sections of the roof deck causing a situation where it re-freezes only when the meltwater reaches the eaves. With even R12 applied directly to the roof deck from below the temperature differences between warmer & cooler parts of the roof are much lower.

The wood-to-wood interfaces are thermal short-circuits, but have a similar vapor permeability per inch as closed cell foam. With a well bonded cc foam job at the rafter there won't be a significant moisture transfer to/from the roof deck via the rafter itself. The 6" rafters are ~R6, which will show up as striping/runnels in the snowpack if you went with an under-deck only approach at R25-ish mid-bay, but would be fairly even with just 1-2" of ccSPF between them. But with as little as R10 panelized goods above the roof deck as a thermal break on the rafters there won't be much visible evidence even with R25 between the rafters.

Nailbase panels can impart significant structure to the 2x4 raftered roof in much the similar manner to (but lower loading than) a SIP. SIPs would be a custom order, and with a complicated roof structure it could be substantially more money than nailbase iso though. On simpler roof designs it can be comparable or even cheaper though. But the greater fraction of the R is exterior to the roof deck, the warmer & safer the rafters & roof deck will be in an unvented attic setup.

6" of spray fiberglass like 1.8lb Spider would add 0.9lb per square foot of dead-load to the roof deck, adding no structure. Lower density spray fiberglass is likely to allow too much convection to reach the cold SPF surface and condense. (1.8lb Spider has air-retarding qualities similar to 2lb spray cellulose). While it may be cheaper & easier to use sprayed goods from below than adding nailbase panels from above, you may be better off in the long run with panels above the roof deck, keeping the structural wood on the warmer edge of the insulation where it'll stay drier.

Aat ~2lbs/square foot an R49 loose fill of cellulose on the attic floor may exceed the dead-weight loading of 2x6 joists when combined with the weight of the air handler, etc. it depends on just how long the spans are between load bearing walls that are likely supporting it mid-chord somewhere (again, something to check out the engineering on before plowing on down that road.)

> I am interested in tightening up the envelope of the house

As a general comment, I'd put a fan in a window blowing outward and go around feeling (or use smoke) for the air leaks. You can also scan with an IR thermometer (on a cold day and let the fan run for awhile first).

> install a dehumidifier in the attic space to keep condensation under control in winter

I'm not sure where you are going to find a dehumidifier that will work in such cold temperatures. Maybe a desiccant based one.

Just a quick update. I was at the MN State Fair today in the Eco Challenge booth and there was a SIP installer there that told me (in response to my inquiry) that they are using a R49 "half-SIP" on retrofits now. The product is ~7" of foam with OSB on one side. He indicated that they use the long screws to attach to the roof deck/truss.

The panels are not ventilated, however, my guess is that with that much insulation, the ventilation is not needed, even at the eaves. I am having him come and check out my property. Any thoughts on using a straight-up R49 panel above the roof deck? it would bring my roof insulation to code in one fell swoop. I suppose it depends on cost...

Also, I understand that any standard shingle over such a panel may not have warranty support, however I also understand that some manufacturers warrant their shingles over unventilated spaces, just not installation....as would be expected. I also understand that the shingles may be hotter, but that the life will likely only be reduced 3-5% not a bad tradeoff for removing the active venting that is depressurizing the current attic.

I would go with the attic floor sealing route personally. You have access and it will provide the biggest bang for the buck.

Tighten up the envelope, build enclosures for the HVAC, and come over the top with blown in to an R-60.

That is going to be the cheapest round and eliminates the moisture concerns but source and by not changing the attic ventilation.

All of the ideas in this thread are really good. I just think that mine is the cheapest. 2lb foam is not cheap and as Dana mentioned, you will need to spray the rafters to eliminate the thermal bridging.

Posted By cmkavala on 29 Aug 2010 08:08 AM

ibillsi;

sounds like a nailbase panel, there are shingles that are warranteed over unvented roof, but even if the shingle life is diminished by 10% it is small compared the the energy savings over the same time period

From the description it's more than a mere nailbase panel, but rather a full SIP, a custom nailbase panel designed for a particular building/pitch.

To get R49 out of a 7 inch panel it's probably a high density polyurethane or high-density closed-cell polyisocyanurate foam too, either of which would impart substantially more structural strength to the roof than 4x8' nailbase panels.

Who was selling/making these half-SIPs?

Dana I believe that the manufacturer is http://www.extremepanel.com/ still working on getting an appointment.

All, I am still mulling the insulation options due to the relative lack of installers for the panelized goods, as in only a couple installers doesn't make for competition.

What are your thoughts about building enclosures into the attic space with a 1" gap around the air handler, then flashing the attic floor/enclosures with 2" of ccspf and backfilling with cellulose as much as possible? Talked to engineer who suggested the 2x6 16oc wouldn't have an issue (span 14' to load). I would redundant air-seal the scuttle hatch and back it with iso.


The main issue I have with this design is the external wall top-plate. I dont' have raised-heel truss on this 1928 stick. I would be retaining the venting, so the new soffits/roof would retain an air channel of at least 1". My rafters are rough 2x4 and joists rough 2x6. I am concerned with the seal/insulation at the top plates. I would likely attempt to place a iso blocker external to the top plate (just above external soffit) and extend up to 2" under upper roof deck (maintain air flow) and then spray CCspf at least 4" at the edges over the top plate when I'm laying down the 2" ccspf. Any comments/ideas would be great.

Posted By ibilisi on 01 Oct 2010 01:25 AM
Dana I believe that the manufacturer is http://www.extremepanel.com/ still working on getting an appointment.

All, I am still mulling the insulation options due to the relative lack of installers for the panelized goods, as in only a couple installers doesn't make for competition.

What are your thoughts about building enclosures into the attic space with a 1" gap around the air handler, then flashing the attic floor/enclosures with 2" of ccspf and backfilling with cellulose as much as possible? Talked to engineer who suggested the 2x6 16oc wouldn't have an issue (span 14' to load). I would redundant air-seal the scuttle hatch and back it with iso.


The main issue I have with this design is the external wall top-plate. I dont' have raised-heel truss on this 1928 stick. I would be retaining the venting, so the new soffits/roof would retain an air channel of at least 1". My rafters are rough 2x4 and joists rough 2x6. I am concerned with the seal/insulation at the top plates. I would likely attempt to place a iso blocker external to the top plate (just above external soffit) and extend up to 2" under upper roof deck (maintain air flow) and then spray CCspf at least 4" at the edges over the top plate when I'm laying down the 2" ccspf. Any comments/ideas would be great.

ExtremePanel's SIPs seem to have EPS type K values ( ~R4/inch), which would only get you ~ R28 out of a 7-inch half-sip, not R49, but maybe they're venturing into polyurethane SIPs and nailbase panels not yet listed on their website?

Boxing in the air handler in a sealed insulated enclosure makes it impossible to service, but that could work thermally.

Your approach of maxing out the R at the exterior wall top plates with ccSPF all the way up to the venting sounds rigtht, and you get nearly double the R there that you'd get with an equivalent thickness of open-blow cellulose.  If the engineer has signed off on the weight factor, air-sealing the rest with ccSPF and blowing a ton of cellulose in there buys you a very favorable R/$.