GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter 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

Closed-cell spray foam versus dense-packed cellulose – Noise control in a PassivHaus

LenMinNJ | Posted in PassivHaus on

This is a comment rather than a question.

In our pre-certified PassivHaus in northern NJ, we have 10-inches of densepack cellulose and 2-inches EPS in the walls, and 10-inches of closed-cell spray foam in the roof.

We’ve noticed that the noise attenuation through the walls is much greater than through the roof.

Most of the noise getting into the attic comes through the roof, which is asphalt shingle over plywood decking on 2×12 rafters, with the space between the rafters filled with 10-inches of the closed-cell spray foam.

Lots less noise comes through the walls, which are double 2×4 frames on a 10-inch baseplate, covered with Zip-system sheathing, 2-inches of Type I EPS foam, and Certainteed Monogram vinyl siding.

If noise control is important to you, and you have a choice of insulation products, you might prefer the densepack cellulose over spray foam.

(We were told that we couldn’t use densepack cellulose in the roof due to building code restrictions.)

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.


  1. Expert Member
    Dana Dorsett | | #1

    You can use dense-packed cellulose in the roof as long as you have sufficient insulation on the exterior of the roof deck for dew-point control. In the coldest parts of NJ (US climate zone 5) o that would be a minimum of 40% of the total R on the exterior.

    You have about R60-R65 of very expensive and climate-damaging (due to the HFC245fa blowing agent) closed cell foam in the roof. With 6" of exterior Type-II EPS (R25) and 10" of dense packed cellulose (R35-R37) in the rafter bays you would be at about R60-R62 center cavity, but with lower thermal bridging (since the EPS covers the rafter elements), and less than 1% of the global warming of a 10" closed cell foam solution.

    You could even safely dense-pack the full 11-1/4" of the 2x12 rafters (R40-R42) if you used 3" of polyiso next to the roof deck (nominally R18, but performing at about R15-R16 in mid-winter in this proposed stack-up), and 3" of Type-II EPS above that (nominally R12.6, but performing at about R13-14 in mid winter.) The center-cavity R would be about R70.

    What's more, the roof deck would then have a reasonable drying path toward the interior, whereas currently it's a moisture-trap between the low-permeance (and often wet) roofing and the low-permeance 10" of foam.

    In the US climate zone 4 parts of NJ you only need about 30% of the total R to be exterior to the roof deck for dew point control With R40-42 dense-packed in the 2x12s that takes only R20-R21 on the exterior, which can be achieved with 5" of EPS, or 2" polyiso + 2" EPS. Again you'd be in the ~R60 range center cavity, but with the reduced thermal bridging it would modestly outperform what you have.

  2. Expert Member

    Len, A third choice which people might want to consider if sound attenuation is important to them is rock wool. Roxul's Safe and Sound batts perform remarkably well in wall and ceiling assemblies. Then again, insulation is only part of any sound attenuation strategy.

  3. onslow | | #3

    Insulation choices aside, I think the sound differences are being driven by the structural choices for wall and roof. The walls are essentially two disconnected surfaces with soft fill between. This is similar to fairly standard sound attenuation tactics for interior walls. A 2x6 plate with staggered 2x4 framing alternately set to each edge is often "woven" with unfaced batts like a slalom course before drywall is applied to both sides. Some people will set R-11 in each bay on both sides and ignore the slight compression. In either case the wall's ability to mimic a drum head is reduced, some frequencies more muted than others.

    If you remember your old movies, people eavesdropped on the neighbors by pressing their ear to a glass held to the wall (or door). Bone conduction primarily makes the sound connect to your ear. The 2x4 "bones" in the wall connect the drywall "drum heads" attached to both sides of the 2x4's in wall shared with the neighbors. Interrupt the connection between vibrating surfaces and sound energy will be reduced. Apartment buildings will sometimes make gapped double CMU fire walls between units to help reduce sound transmission.

    Unfortunately, the roof structure described, if open to the interior rather than an attic, will likely behave more like an apartment wall. The type of insulation in this case might even be increasing the relative stiffness of the overall plane and therefore enhance its behavior as a simple diaphragm. I would suspect that lower frequency noise with strong pressure waves are the most notable, i.e. truck engines, wind vibrations, aircraft...

    One possible solution if the problem is really annoying might be installing a second layer of drywall over a sound mat material meant for floor to floor isolation in concrete buildings. However, it might not have much affect on low frequency sound. I think it is meant to reduce the "high heel" noise of sharp tapping on concrete floors.

    Perhaps someone on this site has a background in acoustics and can suggest how to find or detect the node points on the roof plane and some way to dampen them. Barring that, I have read that people building sound studios can help decouple existing finished drywall from conducting to adjacent rooms by using a very gooey caulking like substance between additional layers of drywall. The effect is to increase the total mass with elastic coupling between mass layers. The screws of course all go the same shared 2x4's so the attentuation is different from the physically separate 2x4's in the staggered stud wall described above. I am guessing lower energy, but higher frequency sound is probably better attenuated with this method. Think clarinet, flute, voice. This method would likely be too heavy for ceilings.

    I do vaguely recall hearing that SIP panels on timber frame can exhibit diaphragm like behavior. I would be curious to know if people that have chosen the TJI Larson wall variant find sound induction from outside to be a problem. If I understand the method correctly, there would be fairly direct connection from membrane to membrane as one goes from siding to sheathing to drywall.

  4. Expert Member

    Before suggesting any solutions it would be useful to know what type of noise is being heard in the attic. Is it structure borne, like rain, or street noise?

  5. KeithH | | #5

    I'd like to echo Roger's comment. If you are trying to design a stick system for noise attentuation, you'll want to disconnect the interior surface from the exterior surface. In a single stud system, the studs connect these two surfaces. In a double stud or staggered stud system, you disconnect the two drum surfaces (plywood and drywall) from each other. This page explains the STC ratings for different interior wall assemblies. The same principles can apply to exterior walls and roofs.

    I also want to echo Malcom's comment. We replaced some fiberglass in one of our walls with Rockwall. We believe that we can perceive the difference in sound transmission. If you search online you'll find people who disagree with that assertion but I trust my own senses.

  6. LenMinNJ | | #6

    Thanks for everyone's comments.

    Dana: The roof framers and the roofers were unhappy about having to deal with foam on top of the deck (don't ask what we went through trying to find a siding contractor who would install over 2-inches of EPS), so we decided to do a simple roof with all of the insulation on the interior in the 2x12 rafter bays.

    During construction, the only contractor we dealt with who was completely comfortable with a thick layer of exterior foam was the EIFS contractor who finished our above-grade poured concrete walls. The exterior of the foundation walls have 4-inches of EPS. (The interior has 6-inches of fiberglass bat.)

    The roof will have a PV system installed, and the lack of exterior foam should make that install a bit easier too.

    If I remember correctly, there's a layer of ice & water shield between the shingles and the plywood deck. All of the deck seams are taped.

    You can see photos in our Facebook group: Passive House NJ . See the post for October 29, 2013.

    Malcolm: It's both rain and street noise. Anything going on outside can be heard coming in through the roof rather than through the walls. The noise comes in evenly across the entire roof. (I am an audio engineer.)

    Roger: Good advice. We already have one room in the house designed for sound attenuation, with double-framed staggered walls filled with fiberglass with a double layer of sheetrock; the outer layer is Quiet Rock with Green Glue between the two sheetrock layers.

  7. Richard Beyer | | #7

    Aside from insulation choices, there are other dynamics of a building which affect noise transmission. Attached below is some good reading material. Personally, if it was my house, my first and last choice would be Roxul brand products for sound and fire protection.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |