GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

We want to install 2″ recessed junction boxes in our unvented roof but will this lead to rotting of roof sheathing?

steve210 | Posted in Energy Efficiency and Durability on

Martin Holladay in his 9-12-14 Musings of and Energy Nerd on How to build an Insulated Cathedral Ceiling said NEVER install recessed cans as they add heat and moisture to the roof cavity and can lead to roof failure.

We are building a flat-truss roof system with 9.5″ of open-cell foam between the drywall ceiling and the roof sheathing, covered by a tapered rigid foam of 4″ at center of the roof to 2″ at the edges, covered by 1/2″ ply and the roofing membrane. We are zone 5 climate in the Northeast.

We want to install downlights and mono-points on the interior drywall ceiling that would be connected to standard 2″ deep junction boxes mounted in the ceiling. All the light fixtures are LED. One-half of the fixtures will have individual transformers located in their j-box, while the others will have their transformer located remotely on a wall. So about one-half the fixtures will emit a bit a heat into the ceiling cavity. Those fixtures with the integral transformers (Glimpse Lamps), however, can be fairly well sealed in terms of transmitting the ambient room heat and humidity into the cavity.

Has anyone had experience with this type of installation?

The building designer is Lindal Homes and have so far not addressed this question, but only revert to the “NO CANS” in the ceiling position. It seems to us that j-boxes with LEDs are far less intrusive in terms of reducing R-value and introducing moisture than are recessed cans with incandescent lamps.

What do you think?

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.

Replies

  1. charlie_sullivan | | #1

    Open cell foam does OK at air sealing, and that's the main issue, so you should be fine. To be extra safe you can buy airtight junction boxes for just a few dollars more each. The heat from light fixture's power conversion system won't hurt the ceiling.

    Without checking, I'm not sure 2" thick foam above the sheathing is adequate. And it sounds expensive to get tapered foam. I'd think it would be safer to have 4" all the way down. You could also mitigate against potential moisture problems in the open-cell foam by putting MemBrain between the open-cell foam and the drywall.

  2. GBA Editor
    Martin Holladay | | #2

    Steve,
    There are several issues here.

    1. Do you have enough rigid foam above your roof sheathing? In Climate Zone 5, ideally you would want to have at least R-20 of rigid foam above your roof sheathing. Depending on how your local code official interprets the code, this could be seen as a code requirement. (For more information on this issue, see How to Build an Insulated Cathedral Ceiling.) The way I interpret the code, open-cell spray foam isn't air-permeable, so the requirement for R-20 of rigid foam doesn't apply. Moisture will be moving toward your sheathing by diffusion, not by air movement, so your sheathing will probably be safe -- especially if you take Charlie's advice and install MemBrain.

    2. Will you have air leakage problems with 2-inch electrical boxes if there is a layer of open-cell spray foam between the top of the box and the roof sheathing? The answer to this question is no.

    3. Does the roof assembly have enough insulation in total? The 9.5 inches of spray foam has an R-value of R-35. The rigid foam is about 3 inches thick. I'm not sure what kind of foam it is, but let's call that layer about R-12. In all, you have R-47. That's probably close enough.

    4. Should you worry that the insulation is a little thin above the electrical boxes? Probably not.

    5. Will the electrical boxes overheat? Maybe, if there are transformers in the boxes. I would contact the lighting manufacturer to get an answer to this question.

  3. steve210 | | #3

    Thanks very much Charlie and Martin. Would you recommend MemBrain for the walls as well? We have two floors with 2x6 frame and posts, open-cell foam, plywood sheathing, Tyvec, and cedar wood plank siding.

    Also, some of the lower floor walls are concrete foundation below grade, with 2 inches of exterior foam & waterproofing (and drainage system) and interior 2x3 metal frame wall. What would you advise for interior insulation for these walls? And I imagine MemBrain would be advisable on them as well, especially given the concrete foundation. The foundation was poured in September.

  4. charlie_sullivan | | #4

    Your walls can dry to the outside, so the mem-brain isn't needed there. And there's no need for it for air sealing as the foam does that too.

    Below grade, I'd recommend EPS or polyiso insulation. But preferably without metal studs creating a thermal bridge through the insulation board. Are the metal studs already up?

  5. steve210 | | #5

    Thanks Charlie. The metal studs are already up - about an inch or two away from the concrete foundation wall.

    Also: Please comment on Lindal's (the house manufacturer) email on the roof issue:

    "Looking at your design Lindal does not warrant the use of open cell insulation. I am by no means an insulation expert and would consult with your builder and insulation contractor for their advice as they would be responsible for the warranty on the insulation system not Lindal but there are a number of reasons for this why we do not recommend it's use

    Open cell insulation only provides about R-3.5 per inch of thickness, compared to R-6 per inch for closed cell. You are I believe is in climate zone 5 which requires R-38 insulation, even if it was possible to fill the full 9.5 inch cavity with insulation that would only give you about an R-34. Also open cell insulation, unlike closed cell, is not a moisture barrier so should only be used with a vented roof system with 2" of airspace above it, reducing your R-Value to about R-27, add to that that you have a flat roof and you would not be able to vent with standard continuous venting as you would not get the airflow draw over the length of the rafter bay so a mechanical venting system would be needed.

    As for your mechanical concerns if using closed cell spray foam to achieve the R-38 R-value that only requires about 6.5 inches of space leaving you a full 3" below for any of your 2.5" deep units and wiring, seems like this would be the way to go."

  6. GBA Editor
    Martin Holladay | | #6

    Steve,
    MemBrain for the above-grade walls isn't strictly necessary, but it adds to the wall's robustness by reducing the outward flow of moisture during the winter. Won't hurt, and could help.

    If you want to improve the R-value of your basement walls, which evidently are already insulated on the exterior with 2 inches of exterior foam, the best type of insulation would be closed-cell spray foam (installed to fill the gap between the steel studs and the concrete wall).

    The second-best approach would be to try to slide some rigid foam into the gap between the steel studs and the concrete wall. If you can successfully do that, that would be great.

    Adding fiberglass batts or mineral wool between the steel studs is a waste of material -- the thermal bridging through the steel studs is so severe that the fluffy insulation between the steel studs doesn't do much.

  7. GBA Editor
    Martin Holladay | | #7

    Steve,
    Regarding the disagreement between you and Lindal Homes: only you can decide whether you want to use an insulation method that Lindal Homes disapproves of.

    In general, I'm a fan of systems that include more rigid foam above the roof sheathing than the 2 to 4 inches of rigid foam that you plan to use -- especially for low-slope roofs. For more information on this topic, see Insulating Low-Slope Residential Roofs.

  8. steve210 | | #8

    Thanks Martin. Our roof is in and the EPS rigid foam is one inch at the outside edge of the exterior wall and 3-4 inches at the center (depends on which roof - we have 2 that are 12' square, 1 that is 18', and one that is 28' square).
    And we have 9.5" in the roof cavity.
    Given that my builder recommends open cell, and I am concerned about the off-gassing during install and after of closed cell, I am leaning to open cell for the 9.5" cavity, covered by MemBrain. I will ask my builder (or his insulation subcontractor) to provide a warranty against sheathing mold and rot.

  9. Expert Member
    Dana Dorsett | | #9

    In zone 5 you'd be far better off with 40% of the total R being above the roof deck as rigid foam, and 60% being open cell foam on the under side of the roof deck, which is far cheaper & greener than closed cell foam.

    IRC 2012 and newer code calls out R49 for ceilings/roofs, not R38 (but was R38 as recently as IRC 2009.) With fully 9.5" of half pound ocSPF with most current manufacturers' goods you'd be a bit above R35, which would require a minimum of R23 above the roof deck to be fully protective without MemBrain. That would be R58, total center cavity-R. If you limited it to 8" (~R30) on the interior side you'd be fine with R20 above the roof deck, and just make IRC 2012 code min.

    Closed cell foam can't be safely & reliably installed in lifts of more than 2", and it's one of the least-green insulating materials out there, due to the high polymer per R, and the very high global warming potential of the HFC245fa blowing agent (about 1000x CO2, 100 year potential.) Installing 6.5" closed cell foam to hit a paltry R39 is the opposite of "green building". Putting 5-6" of EPS or 4-5" of polyiso above the roof deck is much greener due to the low-impact blowing agent (pentane, ~7x CO2 warming potential) , with 7-8" of open cell foam below (blown with water, which has VERY low impact.) And it's more resilient, with an order of magnitude greater drying potential for the roof deck.

    If R38 is the local code, and you don't care to hit R49, 2" of closed cell foam on the underside of the roof deck (R12), and 7.5" of 1.8lb density blown fiberglass (Spider, Optima, L77 etc) under the roof deck adds another R32, bringing you to ~R44. That isn't terrible, if still 10% below current IRC code. At 2" the closed cell foam is fully protective of the roof deck at about 0.6 perms, but you'd need MemBrain to fully protect the fiberglass from collecting a bit of damp over the winter in a higher-humidity house, since you'd be at ~27% of the total R outside the air-permeable fiber insulation not 40%+.

  10. SteveRooney | | #10

    Thanks Martin, metal, for prior help. We are about to spray open cell into the 9.5" flat roof ceiling joist spaces and I am wondering how to resolve the area around the planned wood stove chimney to prevent humidity from entering the roof cavities.

    To recap: we are new construction and have installed numerous electric j-boxes in the ceiling. We have calked exterior of j-boxes (to block moisture) and will install Membrain over the insulation, with flaps into the j-boxes. We have very low-slope roof with 2-4" rigid foam over the deck

    We have three plumbing vents going thru roof, and we can tape the Membrane to their perimeters to block moisture. But what to do about the wood stove chimney? (I think the chimney needs a 2" clearance from combustibles.)

    My idea is to frame the roof hole, and drywall the 12' shaft, with Membrane under the drywall. Maybe use blue board instead of drywall? What do you think?

  11. GBA Editor
    Martin Holladay | | #11

    Steve,
    Chimney penetrations are made airtight with metal flashing and high-temperature silicone caulk. You may want to consult the manufacturer of your metal chimney before engaging in this air sealing work.

    For more information on this topic, see Air Sealing an Attic.

  12. Expert Member
    Dana Dorsett | | #12

    If you're going to insulate the flue chase (at the minimum 2" clearance) with rigid foam, use polyiso, not XPS (blue board) or EPS. Polystyrene melts while burning, which spreads the fire in many ways. Polyiso and polyurethane char in place while burning, and have a much higher kindling temperature.

    Better/safer than polyiso would be rigid rock wool, which is completely fireproof (but not an air barrier.)

Log in or create an account to post an answer.

Community

Recent Questions and Replies

  • |
  • |
  • |
  • |