musingsheader image
6 Helpful?

Service Cavities for Wiring and Plumbing

Designers of superinsulated houses are developing new ways to disentangle utility runs from insulated wall and ceiling assemblies

Posted on Jan 20 2012 by Martin Holladay

Conventional wood-framed walls perform many functions. Exterior walls are supposed to support the roof load, resist racking, and provide insulation. They must also provide space for routing electrical cables and (in some cases) plumbing pipes or even ductwork. If the walls are built properly, they should also include an air barrierBuilding assembly components that work as a system to restrict air flow through the building envelope. Air barriers may or may not act as a vapor barrier. The air barrier can be on the exterior, the interior of the assembly, or both..

Timber-frame manufacturer Tedd Benson has criticized our conventional approach to building walls. He calls these multipurpose walls “entangled,” and he proposes a new approach to wall building — one that begins by disentangling the various functions of a wall.

According to Benson, “Wires and fixtures should be separated from high-performance wall and roof systems whenever possible. [At Bensonwood Homes,] we add a mechanical chase layer to our walls always, and to our roof panels when there is a heavy mechanical demand in that area. Otherwise, we have dedicated chases at the building peak, at the eave and other strategic locations when necessary. The point is that after sealing a building..., we need to also make changes and upgrades easy and provide a path other than one that violates the integrity of the envelope.”

There are many ways to disentangle utilities from the other functions of a wall. Approaches range from timber-framing, to PERSIST construction, to the use of Larsen trusses, to a variety of techniques adopted by Passivhaus builders in Europe. One element shared by many of these techniques is the service cavity.

What is a service cavity?

A service cavity (installationsebene in German) is a “wall-within-a-wall” — a secondary wall on the inside of an exterior wall. It can be framed conventionally, using vertical 2x4s or 2x3s, or it can be created by installing horizontal 2x2 or 2x3 strapping. The main purpose of a service cavity (also called a “service core”) is to provide room to run wiring, plumbing, and ductwork.

Most proponents of service cavities recommend the installation of an air barrier between the service cavity and the wall insulation. This air barrier can be a flexible membrane — for example, a “smart retarder” membrane — or rigid sheets (OSB or plywood) with taped seams.

Ceilings, like walls, can be equipped with a service cavity. For example, at the Freas house in Olympia, Washington, workers from the Artisans Group framed a ceiling service cavity with 2x4 ceiling joists that hang from the structural joists. The two joist assemblies — an upper (structural) one and a lower one for services — are separated by an OSB air barrier.

If you are a purist on the issue of disentanglement, you won’t fill your service cavity with insulation. Rather, the service cavity will remain empty of insulation for the life of the building, to facilitate the work of electricians when the home is remodeled.

On the other hand, some builders go ahead and insulate the service cavity once electrical and plumbing rough-in work is complete.

You'll have to give up some interior space, but there are benefits

There are two main drawbacks to installing a service cavity:

  • A service cavity entails additional cost in materials and labor, and
  • A service cavity takes up interior space.

Weighed against these drawbacks are several advantages:

  • A service cavity provides a way to locate a durable air barrier on the interior side of the wall, where it belongs.
  • Since most or all of the wall insulation is on the exterior side of the service cavity, plumbing pipes are protected from freezing.
  • Installing the air barrier on the exterior side of a service cavity is easier than the Airtight Drywall Approach, because there is no need to seal around electrical boxes.
  • If the air barrier is located on the exterior side of a service cavity, it will be less likely to be damaged by homeowner alterations.
  • If it is detailed properly, an air barrier on the exterior side of a service cavity can also function as a vapor retarder.
  • If you use plywood or OSB as the interior air barrier, the air barrier can provide shear bracing (which some types of exterior sheathingMaterial, usually plywood or oriented strand board (OSB), but sometimes wooden boards, installed on the exterior of wall studs, rafters, or roof trusses; siding or roofing installed on the sheathing—sometimes over strapping to create a rainscreen. , like thin fiberboard, may not be able to provide).

Five key questions

There are many ways to build a service cavity. If you want to include a service cavity in your next home, you'll need to answer five questions:

  • Should the service cavity be framed with vertical studs or horizontal strapping?
  • Assuming your walls have double or triple framing, which wall is the bearing wall?
  • Where do you want your air barrier: at the outside sheathing layer, directly behind the service cavity, or at both locations?
  • What’s the best air barrier material: OSB, plywood, or a flexible air-barrier membrane that comes in a roll?
  • Should the service cavity include insulation?

Vertical framing or horizontal framing?

For decades, many builders in New England and Canada have been installing 2x3 horizontal strapping on the interior side of exterior walls. (On the Breaktime forum at the Fine Homebuilding website, some posters referred to this type of wall as a “Mooney” wall, until other posters objected to one builder taking credit for a long-established practice.)

Although the two main reasons that builders install horizontal strapping are to interrupt thermal bridgingHeat flow that occurs across more conductive components in an otherwise well-insulated material, resulting in disproportionately significant heat loss. For example, steel studs in an insulated wall dramatically reduce the overall energy performance of the wall, because of thermal bridging through the steel. and to create a thicker wall that holds more insulation, many builders install a polyethylene air barrier behind the horizontal strapping. This allows them to run wiring without worrying about patching holes in the polyethylene air barrier.

When it comes to framing a service cavity, many European builders prefer horizontal strapping to vertical studs. Floris Buisman of New York City, a distributor of European building materials, notes that service cavity framing can be either vertical or horizontal. “This wall-inside-a-wall can be a structural 2×4 wall in lightweight construction,” Buisman wrote. “Or you can run the 2×2s (or 2×3s on edge) horizontally when the structure is on the exterior. This makes the electrician’s life pretty easy as he no longer has to drill through a large number of studs to get the outlets where they are needed.”

Albert Rooks of Olympia, Washington, another distributor of European building materials, wrote, “Try splitting a 2x6 to create the service cavity — 2 5/8 inch is just enough space and 2x6s (for us) are the best buy. Less interior space is lost to the cavity. Also: It takes some willingness for adventure, but think of framing the service cavity with split 2x6s horizontally rather than vertically. It’s faster to wire and it’s really only to hang the drywall. Once you’re just adding furring strips, there is no reason that they need to be vertical. They are only a nailing plane for the drywall.”

John Brooks, an architect from Dallas, Texas, points out that horizontal strapping “helps reduce thermal bridging and convection in the cavity.”

Lucas Durand of Neebing, Ontario, is including a service cavity in his new house. He wrote, “I will be installing 2x3s horizontally for my service core. Faster wiring (not boring holes) was what really appealed to me.”

A service cavity protects the air barrier

According to Patrick Haak, a sales representative for Swiss tape manufacturer Siga, a service cavity “protects the air barrier against penetrations caused by installation or later by the occupants. The cost difference for this additional layer is usually not significant.”

When it comes to installing an air barrier on the inside of a wall, many North American builders are most familiar with the Airtight Drywall Approach (ADA). While the ADA is a time-tested method of building a tight wall, some Europeans worry that drywall air barriers are vulnerable to abuse. That’s why Floris Buisman prefers to locate the air barrier behind a service cavity: “After the insulation and drywall is installed, the airtight layer behind the service cavity will be protected from nails to hang pictures or screws to hang shelves; and even wires for phones, cables and Internet, etc., can be easily installed without compromising the airtightness of the building for years to come.”

But in the eyes of some North American builders, the European worry about the vulnerability of ADA air barriers is overstated. According to Matthew Amman, a builder in Ashland, Oregon, “I understand the service cavity idea, but I don’t think it guarantees any less possibility of air barrier compromise. … Remodeling is due mostly to aesthetics, and judging Passivhaus buildings on this, these buildings will be remodeled for sure. This will most definitely interrupt the service cavity and the whole wall in general, but the whole idea that a building will be airtight is false, no matter how you build it. I haven’t seen any building with an ach of 0, and these tests are performed with the building in a state of closure that I am sure will not remain in daily use. As far as I concerned, any cutting and patching done in drywall is then taped and mudded, thus air sealing it once again. To imply that Passivhauses in America will not be remodeled, thus air barrier temporarily broken, is utterly ridiculous.”

However, Albert Rooks is quick with a counterargument. “I disagree about ADA being a good solution because it is repairable,” Rooks wrote. “The point of it being repairable is a bit foolish. If you’re building big thick cavity walls in cold climates, a service cavity is certainly the way to go. Put the pressure barrier away from the finish surface so that it doesn’t need to get repaired. … The ‘repairable ADA scenario’ requires an understanding by a homeowner that something needs to be repaired — a very unlikely occurrence. Whereas putting the issue out of reach is certainly more reliable. I of course have sympathy for the cost of adding a service cavity. These things have to move by progression and it’s tough to add cost.”

Air barrier material choices: OSB or a membrane sold in a roll?

Several European manufacturers sell air barriers that come in a roll. Some of these are exterior air barriers; in the U.S., these products are usually called water-resistant barriers (WRBs). Other products are interior air barriers; in the U.S., these products are usually called “smart retarders.”

Floris Buisman’s company, Four Seven Five, sells an interior air barrier called Intello Plus. Buisman wrote, “The larger stud walls behind the [air barrier] membrane should be filled with blown-in cellulose after installation of the Intello Plus, or any fibrous insulation can be installed beforehand.”

However, most North American builders who include service cavities install a layer of OSB directly behind the service cavity. Once the seams between the OSB are taped, the OSB serves as an air barrier. According to GBA reader Jan Nielsen, “After spending all this effort to get a super tight envelope, the last thing I want is a little electrical or plumbing work to compromise the envelope and cause moisture damage over time, so I really like the idea of OSB or plywood behind a service cavity.”

Albert Rooks points out that OSB can double as a vapor retarder. He wrote, “OSB is around 0.7 perm. This is enough to stop large amounts of interior vapor penetrating the cavities. And it’s still just vapor-open enough to let things dry to the interior when the levels are higher in the cavity than the interior. It’s a failsafe feature.”

Should a service cavity be insulated?

When the PERSIST construction method was developed in Canada in the 1960s, stud bays were always left uninsulated. The stud walls held up the roof and provided a space for wiring, but they were not intended to be filled with insulation.

Many Passivhaus builders have adopted the PERSIST approach to service cavities. However, others like to see the cavities filled with insulation. Garth Sproule, a GBA reader from Canada, wrote, “Maybe the best way to stop convection in the service cavity would be to throw in some cheap batt insulationInsulation, usually of fiberglass or mineral wool and often faced with paper, typically installed between studs in walls and between joists in ceiling cavities. Correct installation is crucial to performance. before drywall — boosts the total R-valueMeasure of resistance to heat flow; the higher the R-value, the lower the heat loss. The inverse of U-factor. a little as well.”

Another fan of insulated service cavities is Albert Rooks, who wrote, “I like adding insulation into the service cavity. It’s an available cavity that is ready to fill. … To take the idea a little farther, for the Passive HouseA residential building construction standard requiring very low levels of air leakage, very high levels of insulation, and windows with a very low U-factor. Developed in the early 1990s by Bo Adamson and Wolfgang Feist, the standard is now promoted by the Passivhaus Institut in Darmstadt, Germany. To meet the standard, a home must have an infiltration rate no greater than 0.60 AC/H @ 50 pascals, a maximum annual heating energy use of 15 kWh per square meter (4,755 Btu per square foot), a maximum annual cooling energy use of 15 kWh per square meter (1.39 kWh per square foot), and maximum source energy use for all purposes of 120 kWh per square meter (11.1 kWh per square foot). The standard recommends, but does not require, a maximum design heating load of 10 W per square meter and windows with a maximum U-factor of 0.14. The Passivhaus standard was developed for buildings in central and northern Europe; efforts are underway to clarify the best techniques to achieve the standard for buildings in hot climates. designer, if they don’t insulate the cavity, it counts as treated floor area in the PHPP. [Editor's note: Floris Keverling Buisman disputes Rooks's statement; see posted comments below for more information on this topic.] Not that it’s all about Passive House, it’s just illuminating that it counts either as insulated assembly, or as heated floor area. It’s not something that can be set aside and ignored. It has a value in either retention or consumption of the heating/cooling load.”

What about buildability?

It’s easy enough to draw a complicated wall sandwich on a piece of paper. If a designer wants, a wall can include three stud walls — two walls to define the insulated portion of the wall, and a third wall as service cavity — with an OSB air barrier in between.

However, in addition to adding up the cost of such a complicated wall, it’s important to consider buildability. Whether the air barrier consists of OSB or a flexible membrane, seams will need to be taped, and the air barrier will need to be sealed at the top and bottom. Insulation must be installed in a way that doesn’t interfere with the air-sealing steps. Before deciding on a wall assembly, a designer has to think through each step required for insulation and air sealing.

While OSB is a durable air barrier, many cellulose installers prefer to blow cellulose behind a membrane like InsulWeb — because poking the InsulWeb provides real-time feedback on cellulose density. If InsulWeb is used, how will it be attached to the studs? Will the wall’s air barrier be installed before or after the cellulose is installed?

According to Albert Rooks, the air barrier seams can be taped before the service cavity wall is raised. “Place OSB on the backside of the interior stud wall and tape it. Tilt it up onto gaskets (Resource Conservation Technology).”

Since there are many ways to build a service cavity, no universal rules apply except this one: the wall must be buildable.

Weighing the pluses and minuses

Many experienced designers of superinsulated houses have decided that service cavities make sense. According to Thorsten Chlupp, a builder in Fairbanks, Alaska, the service cavity framing “is now part of your conditioned spaceInsulated, air-sealed part of a building that is actively heated and/or cooled for occupant comfort. and is utilized as your utility chase — electrical, mechanical, and plumbing, which opens many more options on how you can run your utilities. It is your structural building core. It can be insulated if additional R-value is needed and the 1/3 to 2/3 insulation ratios in-between you pressure boundary is maintained. I actually think 40/60 is still a safe ratio. The insulation is wrapped around the exterior of the structure and keeps everything warm and is protected with a weather barrier.”

Chlupp continued, “The key in both assemblies is the placement of the airtightness layer within the assembly, where it can be continuous and (most importantly) protected. If you place your poly/air/vapor layer on the inside frame face you just can’t get there. Even if you did the best job (and hardest) to air seal you be surprised to see how fast this can deteriorate a few years with homeowners taking over. And you fight your installations of electrical, mechanical and plumbing — now, today and in the future. An installation layer is simply invaluable. Combine this with a foolproof pressure boundary and you got a system which just works.”

According to Garth Sproule, service cavities have both pluses and minuses. “I really like how using a service cavity makes it so easy to create an air barrier that is easy to build and requires no maintenance,” he wrote. “It also allows rewiring and re-plumbing without disturbing the pressure boundary. The costs, however, are not insignificant. Apart from the obvious costs of the materials and labor involved, we must not forget that we are giving up significant interior floor space. This unusable space must still be conditioned the same as the rest of the interior. For example, on a 30' x 40' building with a 3.5-inch service cavity, we have to give up about 40 square feet of floor area.”

Taking all these factors into consideration, Lucas Durand concludes that service cavities make sense. “I think the issue of the extra cost of a service cavity is a little overblown,” he wrote. “The loss of interior space is a fact. But what wall system doesn’t involve a compromise of some type?”

Last week’s blog: “Fukushima’s No-Entry Zone.”

Tags: , , , , , , , , , ,

Image Credits:

  1. Image #1: Martin Holladay
  2. Image #2: Alex Wilson
  3. Image #3: Justin - Wicklow House -
  4. Image #4: Wolf Haus
  5. Image #5:
  6. Image #6:

Jan 20, 2012 8:01 AM ET

Edited Jan 20, 2012 8:22 AM ET.

by John Brooks

I agree with Thorsten and Lucas... we need something more durable than drywall for the continuous interior PRESSURE boundary.
I'm not so sure we need to give up interior space (or add footage) for a wall service cavity around the entire structure just for a few wall outlets. Of course plumbing walls should be isolated and be interior of the PRESSURE boundary.

pressure2.pdf 188.92 KB

Jan 20, 2012 10:56 AM ET

No loss in space
by Ray Smith

When I lived and built homes back in Ireland I always used service cavities. If you use the finished internal dimensions as fixed, making the walls wider, then there is no loss of space. Internal wall board was Fermacell which stiffened up any walls a treat, Intello+ was the air/vapour barrier and, as your article states, not compromised in any way. Also, Fermacell was easy to patch if any electrical outlets etc were ever added at later dates.

Jan 20, 2012 11:11 AM ET

Response to Ray Smith
by Martin Holladay

I understand your point, but thick walls have at least two possible disadvantages:

1. In a small lot with setback requirements, or in a remodel job, thick walls may indeed mean a loss of interior space.

2. In many parts of the U.S., property owners pay taxes based on the exterior measurements of their home -- and so homes with thick walls have higher taxes per square foot of interior space than homes with thin walls.

Jan 20, 2012 12:48 PM ET

Thanks for the article Martin
by Brian O' Hanlon

It seems to be a quite comprehensive run-down of the options and design decisions that need to be made. I do confess, service cavities in domestic construction has always been an area I wondered about, and your article will provide me with an excellent first base, from which to think about this subject.

Thanks again for all the care in writing it, BOH.

Jan 20, 2012 1:25 PM ET

Edited Jan 21, 2012 8:26 AM ET.

Two pressure boundaries...
by Lucas Durand - 7A

No 3D airflow networks likely for that detail...
Very robust.

Assuming the interior wall is load bearing...
Maybe a vapour permeable fibreboard sheathing on the exterior would be better than plywood/OSB?

Struggling to think outside my "climate box"...
For hot/mixed humid climates maybe it would be better to make the exterior wall load bearing so that a more vapour permeable sheathing could be used on the interior side?

Edit to say:
Never mind - sorry.
The thought behind my question about hot/mixed humid climates was less than half-baked.

Jan 20, 2012 1:34 PM ET

Service Cavities
by Pat Murphy

As I recall, the first houses that Katrin built in Urbana used 2x3s inside the insulating wall.

If framing for the service cavity is installed horizontally, how are the edges of the sheet rock nailed? Does it imply vertical blocking where the sheetrock will fall?


Jan 20, 2012 1:37 PM ET

Response to Pat Murphy
by Martin Holladay

Gypsum drywall only has to be fastened every 16" or 24", so seams don't have to be fully supported.

Jan 20, 2012 4:49 PM ET

Sheetrock corners
by floris keverling buisman

Regarding TFA in Passive Houses, you measure sheetrock to sheetrock - so no added TFA from a service cavity, Doesn't matter if you insulate the cavity or not. It is all about conditioning the space you can actually use.

Note on materials: it is much easier to put INTELLO Plus on a ceiling than mounting OSB overhead and then tape + fur out a service cavity - see picture. This can be done by one person instead of 3 and provides you drying potential as a bonus, which is especially important for flat or un-vented roofs.

Beyond Passive - service cavity sloped ceiling installation of INTELLO2.jpg

Jan 20, 2012 4:53 PM ET

Edited Jan 20, 2012 4:53 PM ET.

An explanation of Floris Keverling Buisman's abbreviation
by Martin Holladay

For those of us who haven't yet been invited to the inner circle, and are scratching our heads over "TFA" -- I think I figured it out.

Buisman is apparently referring to the "treated floor area" input for the Passive House Planning Package software.

Jan 20, 2012 5:38 PM ET

Three questions
by Richard Patterman

If you put up your OSB barrier, then add your horizontal strapping, then add your wiring and services, when does the building inspector inspect the framing of the exterior wall?

Other than wiring and MAYBE one kitchen wall, what services are you guys putting in your exterior walls? I've always thought it best to put plumbing and mechanical in INTERIOR walls.

I understand using horizontal stapping as a thermal break and understand wanting 1.5 to 3.5 inches more insulation, but I do not understand creating this space and NOT insulating it.
Won't you lose some of the benefit of the thermal mass of the drywall if it has an uninsulated space behind it?

Jan 20, 2012 5:48 PM ET

Response to Richard Patterman
by Martin Holladay

Q. "What services are you guys putting in your exterior walls? I've always thought it best to put plumbing and mechanical in INTERIOR walls."

A. If the service cavity is on the interior side of the wall insulation and the air barrier, there is no reason whatsoever why plumbing pipes can't be located in the service cavity. The service cavity is interior, heated space.

Q. "Won't you lose some of the benefit of the thermal mass of the drywall if it has an uninsulated space behind it?"

A. No. In either case, the drywall should experience interior conditions.

Jan 21, 2012 12:09 AM ET

More detail for the simple-minded
by Lucy Foxworth

Ok, say you are a layperson who reads GBA daily to help renovate your house and to plan building another house. And you've never built a house before. You guys have written about service cavities before and it makes sense that it is something to consider for a new house, but the details are pretty overwhelming.

Do you have more detailed drawings of what is entailed? I have to see stuff to understand it.

Does the wall cavity have to be about 3" deep for the electrical boxes? If you install a service cavity in the ceiling, do you still use insulated canned lights?

Thank you. Lucy

Jan 21, 2012 6:35 AM ET

Edited Jan 21, 2012 6:37 AM ET.

Response to Lucy Foxworth
by Martin Holladay

Q. "Do you have more detailed drawings of what is entailed?"

A. The photos on this page should help you visualize several approaches to service cavities. In the first comment posted on this page, John Brooks included a detail of a ceiling service cavity. If you click some of the links in the article, you'll be able to read discussions on GBA pages; many GBA readers have posted details that include service cavities.

In the future, GBA hopes to develop and share more details that include service cavities.

Q. "Does the wall cavity have to be about 3 inches deep for the electrical boxes?"

A. Standard wall boxes are 3 inches deep, but shallower boxes are available. The electrical code has requirements on the minimum volume of electrical boxes; volume requirements vary depending on the number of wires entering the box and the number of receptacles or switches. It's certainly possible to use shallower 2 inch or 2.5 inch deep boxes for receptacles and switches.

Most ceiling boxes are about 2 inches deep, but (as with wall boxes), different depths are available, including shallow boxes that are merely 1/2 inch deep -- just deep enough for the Romex. Obviously, these 1/2-inch boxes can only be used for a limited number of applications.

Q. "If you install a service cavity in the ceiling, do you still use insulated canned lights?"

A. If you want to install recessed can lights in a service cavity, be sure that the cavity is deep enough. You'll probably need to frame the cavity with 2x6s or 2x8s if you intend to install recessed can lights.

If the service cavity has no insulation, you don't need to use recessed cans that are rated for insulation contact.

Jan 21, 2012 1:00 PM ET

The Alaskan study to me says
by aj builder, Upstate NY Zone 6a

The Alaskan study to me says insulation in the cavity may promote mold.

Jan 21, 2012 2:53 PM ET

Edited Jan 21, 2012 2:58 PM ET.

An important topic
by albert rooks


Thanks very much for such an in-depth look at the service cavity. This is really great. I think GBA is showing itself as "far ahead of the curve" by treating this issue as one that is worthy of discussion by "all" of us, not just those who are living at the "bleeding edge".

Considering a service cavity represents a far bigger step for US designers & builders than one would think. It means that we are really thinking about the longterm overall health of the envelope and what forces affect it.

From a cold climate, dense packed fill insulation perspective: I look at this as a progression of increasing quality standards. It's my hope that all of the building industry becomes aware of, and considers it's relative position on the "quality progression". I think the path starts with airsealing.

Air-sealing the envelope by creating a pressure boundary at the sheathing is the first relatively simple step to separate conditioned space from the exterior. The focus is on an airtight envelope only. The line of the air-barrier may be a bit "squiggly" at first: Not following the exterior lines but just where you can initially reach the leaks to seal them.

Stepping up and adding the Air-tight Drywall Approach (ADA), means that now your moving the pressure boundary from the exterior to the interior. in addition to separating conditioned air from the exterior air, this will attempt to stop moisture laden air from entering the wall. An important step when the wall has become thicker and the exterior sheathing is now so far from the interior heat sources that it will not stay warm. Stopping moisture laden air at the drywall is taking the first step in preventing rot and decay.

If you typically build thick -super insulated- walls in mild to cold climates, then keeping that moist air out of the wall at the interior surface becomes a larger issue the more you become aware of it. More so with our typical North American assemblies that use OSB or plywood that is not vapor permeable on the exterior. Speaking for myself, when I really look at these assemblies I come to the conclusion that the long term health of the building, preventing rot and mold depends on the insulated cavities being protected from air and water on both sides. North American Builders got that "point" early and developed ADA. if you stay on that "path" and believe that these interior pressure boundaries are the key to long term health, then you have to "painfully acknowledge" that the finished wall will get punctured in many many places over what we hope will be a 200 year life span of this new structure.

So if your serious about the building and occupants health and life span, you have to protect the interior pressure boundary: Today that means placing it inside the wall and away from meddling hands. Once a builder takes this big step, a host of new choices open up: Both the builder and designer can start to entertain designs where the shear loads aren't picked up by the exterior sheeting. it's now possible to move the structural sheathing from the outside to the inside and build entirely new assemblies.

With the structural sheathing now on the inside and doubling as the air and vapor pressure boundary, this means that you can build the classic european wall that has an increasing diffusion gradient to the exterior. The assembly that essentially will follow exterior relative humidity and can always dry since the exterior sheathing is replaced with something vapor permeable.

Using a membrane for the pressure boundary is equally attractive. it's a similar build to standard framing and probably assembles faster. There are two groups to work with: fixed or variable permeance membranes. The fixed permeance membranes work just like sheet goods: The vapor permeance is fixed.

Moving to a variable permeance membrane means (to me) you do need to be careful and really think about your climate and the rest of the layers of your wall. This really leads to another discussion. In short, variable membranes are variable because the permeance increases as the relative humidity increases. This means that the membrane can open up during periods of high indoor relative humidity. These levels change with macro and micro climates, occupation levels and activities. What needs to be carefully considered as you develop a wall assembly is the fact that the pressure boundary is there to limit air and water vapor from entering and accumulating in the insulated cavity. When the membrane becomes variable you lose the ability to rely on low permeance. if you're doing WUFI studies on an assembly with a variable permeance, you need to acknowledge that one case study possibly won't do it. WUFI can only model a single set of values. You might need to run a few cases that change both the indoor humidity as well as the membrane permeance. Having a variable permeance is a good tool when it's planned for and fits the application.

It may seem like this is over complicating things, but if the goal is to honestly maximize the health and longevity of super insulated energy efficient buildings, the ability to begin using these new assemblies starts to change dramatically when you introduce a service cavity; however you do it.

I really appreciate that you started this with Tedd Benson. His shop is the "Perfect Storm" for this kind of thinking: Timber frame folks (I'm one) are descended from a world-veiw that buildings last in the many many hundreds of years, not decades. Envelope quality is an entirely different conversation in that world -it's all about long term durability. Much like where Alex Wilson is going with the idea of Resilient Design, but deeper into the skin of the envelope. Add the fact that Tedd's shop has builders who are well versed in european assemblies where service cavities are the norm, and you've got a company with both the "knowledge" and the "guts" to take on the realities of what makes envelopes fail... And what you can do about it.

sorry for the long windedness. It's a passionate subject for me.


Jan 21, 2012 2:53 PM ET

depth of cavity and insulation of it
by floris keverling buisman

the depth of the cavity can be dictated by the fact that the electrical inspector/NEC requires wires to be > 1.25" from edge of the stud it is passing through. If it is less you need a metal plate to prevent shorting out these wires by nails/screws used for sheetrock, picture hanging etc. If you use a 2x2 and INTELLO Plus you can fish the wire behind it and comply with this code w/o metal plates or drilling through studs. Shallow 1.5" boxes would be needed in that case.

If you intend to insulate the service cavity it will probably mean the cavity will be deeper so you can use off the shelf batts and this also allows you to drill/run wire through the studs without needing metal plates. Please note that you would still want >2/3's of the insulation on the outside of the airtight layer to prevent vapor diffusion/condensation issues in winter.

Jan 21, 2012 3:31 PM ET

Edited Jan 21, 2012 3:34 PM ET.

The Alaskan study to me says
by aj builder, Upstate NY Zone 6a

The Alaskan study to me says this wall design is a set up for mold.

The frame of the home here in the north needs to be fully in or fully out (of the insulation and vapor barriers) to best control mold. When OSB and drywall are are in between cold and or hot, they when moist grow mold.

So sandwiching OSB in a built up wall to allow for one wire to sm outlet is nuts. Put outlets in adjoining walls and in baseboards and floors. No plumbing or other items need to be in exterior walls.

Albert and others are right that for inside insulated and vapor barriered walls we need good permeable exterior components.

So if yaa like this chase idea be careful to think thru where mold and OSB may end up being its downfall.

Jan 21, 2012 3:43 PM ET

Response to Albert Rooks
by Martin Holladay

You wrote, "There are two groups to work with: fixed or variable permeance membranes. The fixed permeance membranes work just like sheet goods: The vapor permeance is fixed."

I disagree. Plywood is a smart vapor retarder. According to John Straube, the vapor permeance of plywood sheathing varies from 0.5 perm when dry to 20 perms when wet.

Jan 21, 2012 3:50 PM ET

Response to Martin
by Lucas Durand - 7A

Plywood is a smart vapor retarder. According to John Straube, the vapor permeance of plywood sheathing varies from 0.5 perm when dry to 20 perms when wet.

I forgot about this...
Thanks for reminding me.

Jan 21, 2012 4:03 PM ET

Response to AJ
by Lucas Durand - 7A

So if yaa like this chase idea be careful to think thru where mold and OSB may end up being its downfall.

Did you read Martin's blog?
Did you see Thorsten Chlupp's comments towards the end?

Jan 21, 2012 4:54 PM ET

I get conflicting reports on Plywood
by albert rooks


Yes I have been told that plywood permeance changes with saturation. The numbers I've heard about are more in the .5 to 1.4 - 2.0 perms. Not 20 perms. That's a new number to me. Can you refer me to where it's posted? I'd love to understand the conditions that it was measured.

Further... If the Plywood is used as an air and vapor pressure boundary at the interior of a super-insulated assembly, how does it get saturated enough to affect permeance that much? I'm not sure that this adds up when it's now inboard rather than at the sheathing layer.

Jan 21, 2012 5:14 PM ET

Edited Jan 21, 2012 5:15 PM ET.

Response to Albert
by Lucas Durand - 7A

20 perms sounds high to me too...
Here is a snip from info-312 that shows wet cup permeance of 10 perms.

Although the thickness of the plywood isn't mentioned...


Jan 21, 2012 5:28 PM ET

Wait a sec...
by Lucas Durand - 7A

This snip from info-500.

More revealing...


Jan 21, 2012 6:10 PM ET

Edited Jan 21, 2012 9:43 PM ET.

Wet cup vs "saturation"...
by Lucas Durand - 7A

If the Plywood is used as an air and vapor pressure boundary at the interior of a super-insulated assembly, how does it get saturated enough to affect permeance that much?

Assuming that we don't include wetting from bulk water intrusion...
Isn't the wet cup value most appropriate for consideration under these circumstances?
Or (under the right conditions) can plywood act as a sort of reservoir that accumulates moisture through vapour transmission over time?

Jan 22, 2012 4:13 AM ET

Conflicting permeance ratings??
by albert rooks

Nice work Lucas,

I followed your links and found APA's Publication J450 (attached here). I also attached where I had gotten my original low perm rating for Plywood: NRCA Roofing & Waterproof Manual (the first 4 of 21 pages attached) See my highlights on this. It show a static 0.7 perms at .25". That number would decrease as the thickness increases to 0.5".

The two sets of data vary widely. Hmmm. Martin right (again)? Hmmm.

The number Martin was quoting varied so widely from what I read that I thought I must be misunderstanding something and that the permeance increase was measured when wetted by bulk water: A soaked board in an exterior application where the cells have expanded due to liquid saturation rather than RH. Hence my wondering about the position in the assembly...

In Publication J450, the even the OSB rating are higher than expected.

NRCA Waterproof Manual.pdf 559.85 KB
APA J450.pdf 251.49 KB

Jan 22, 2012 5:40 AM ET

Response to Albert Rooks
by Martin Holladay

Thanks for attaching the APA document, “Water Vapor Permeance of Wood Structural Panels and Wood Wall Construction,” which confirms my statement. (The document even has a section with the heading, “Plywood and OSB are ‘smart vapor retarders.’”)

I never claimed that a plywood air barrier would get wet enough for it to have a permeance of 20 perms in an actual wall. Obviously, if your plywood is that wet, something is wrong with your wall assembly. I merely provided the measured values for the vapor permeance of plywood, to point out that I disagreed with your view that the permeance of plywood is "fixed."

Vapor permeance of plywood and OSB - APA.jpg

Jan 22, 2012 10:47 AM ET

Conceptual shift?
by James Morgan

Albert, thanks for taking a step back and looking at the bigger picture. We definitely have a tendency to look at a newly completed home as a finished object which will never be messed with. Resilient building systems need to allow for a measure of less competent interventions further down the line.

Re: 'losing' floor space - this is not a sensible metric for evaluating a particular strategy. 40 sq. ft. may sound like a lot but in most rooms 2"-3" off the width or the length will make no practical difference whatsoever (the exceptions are the few tiny spaces like a compact bathroom). In many homes, bad planning wastes far more space than small increments in wall thickness.

Property taxes: any environmental upgrade has a cost - the marginal increase in property tax burden of a thicker wall will be a small part of that, to be measured in the normal way against the proposed benefit.

Fitting into a tight lot: where the lot is reasonably generous Ray Smith's approach is correct - start with the required interior space and let the wall be as thick as it needs to be. If the lot is extremely restricted a balance will need to be struck, but I'd point out that in the UK where lots are tiny, exterior wall thicknesses for good reason are commonly 10" - 14".

Jan 22, 2012 12:35 PM ET

OSB sandwiched between
by aj builder, Upstate NY Zone 6a

OSB sandwiched between insulation equals mold. Worst idea ever dreamed up. Alaska study is your proof.

My conclusion, do not insulate in the proposed chase if for some wrong reason you build a chase. I think the idea is useless. Most homes will never have this.

Worried about a picture being Hung on the drywall, really???

Just sayin

Jan 22, 2012 12:43 PM ET

Edited Jan 22, 2012 12:46 PM ET.

Response to AJ Builder
by Martin Holladay

You wrote, "OSB sandwiched between insulation equals mold." That's not true, AJ.

To get mold, you need several conditions:
1. The OSB has to be cold; and
2. There has to be a source of humid air that introduces moisture to the cold surface, and
3. There must be limited drying ability.

There are many ways to prevent these problems, including:
1. Adding insulation on the exterior side of the OSB to keep it warm,
2. Installing an interior air barrier to limit the escape of warm, humid air, and
3. Ensuring that the materials on the interior side of the OSB are vapor-permeable, to facilitate drying.

Jan 22, 2012 7:12 PM ET

Response to Martin Halladay
by albert rooks


Per your: "I never claimed that a plywood air barrier would get wet enough for it to have a permeance of 20 perms in an actual wall. " - "I merely provided the measured values for the vapor permeance of plywood, to point out that I disagreed with your view that the permeance of plywood is "fixed.""

I have to admit, The plywood seems "smarter" than I do at the moment. :)

Sure, things would be amiss if the inboard plywood was at 8o% RH for any length of time. While I was struggling to imagine what would set that up in a cold climate, I was more caught off guard about my own ignorance of the plywood being able to reach such high permeance. Even in the 45 to 60% RH range. It's a far more permeable than I was used to thinking. Interesting...

I have to admit, I'm certainly still a student (not a master) on this topic. I expect that I'll be one for a long time...

Thanks for the clarification.

Jan 23, 2012 1:23 AM ET

The best and only way
by aj builder, Upstate NY Zone 6a

The best and only way superinsulated walls should be constructed is with one insulation/air barrier layer. OSB is not a durable product in superinsulated assemblies.

Buildings should NOT be designed so exacting that merely hanging pictures is a concern as raised earlier. Real poor reason to build out a wall into a chase. And saving the time to drill for a piece of romex, really? Electricians can drill an entire house for wire in a couple hours.

This whole thread is not logical. Adding cost, time, material, mold trap, loss of square footage.

Not even a bad idea. Worse than that IMO.

Jan 23, 2012 5:46 AM ET

Response to AJ
by Martin Holladay

Concerning your points:

1. "The best and only way superinsulated walls should be constructed is with one insulation/air barrier layer."
I'm sure you've seen moldy walls, but you are overgeneralizing from the failures you have seen. The mold was likely due to too few air barriers, not too many. And plenty of durable walls have more than one insulation layer. Each wall assembly has to be assessed on its merits.

"OSB is not a durable product in superinsulated assemblies."
That's a defensible statement. But when the OSB is on the warm side of the wall, it's likely to last a very long time.

"Buildings should NOT be designed so exacting that merely hanging pictures is a concern as raised earlier."
That sounds like an argument in favor of installing a service cavity.

Jan 23, 2012 11:02 AM ET

Edited Jan 23, 2012 11:06 AM ET.

Martin, sometimes you sound
by aj builder, Upstate NY Zone 6a

Martin, sometimes you sound like you never owned a hammer and yet I know that's not true.

We will have to agree to disagree. The more you defend this chase idea the more I think it is down right silly.

Next idea.....

Jan 23, 2012 11:11 AM ET

Edited Jan 23, 2012 11:12 AM ET.

Rephrasing my answer
by Martin Holladay

OK, I'll say the same thing, this time holding a hammer in one hand.

AJ: Martin, I don't think that OSB is a good idea. It's gonna get moldy.

Martin: Naw -- it'll be fine.

AJ: I don't like it. Insulation on both sides.

Martin: Yeah, but there's more insulation on the outside, so it'll stay warm and dry. Don't worry so much. Who's getting donuts today?

Jan 23, 2012 1:46 PM ET

Lol, good your holding that
by aj builder, Upstate NY Zone 6a

Lol, good your holding that hammer otherwise I'd take a trip over to VT and take it away from you till you come to your senses.

Jan 23, 2012 1:50 PM ET

Service chases aren't exactly new to energy efficiency
by Andris Auzins

Thanks Martin for your article about service cavities.

It was interesting to read the responses and comments and somewhat differing viewpoints.

It got me thinking about what used to be one of the ways to build exterior walls in "superinsulated" homes in the 1980's. I built my family's home using this method. In those days, "double wall" construction enabled the inclusion of greater thicknesses of insulation since the interior stud wall could be offset from the exterior stud wall using plywood "plates" at the top and bottom. The total wall thickness was established by the width of the plywood plates tying the inner and outer stud walls together.

One of the construction "features" of this approach was the placement of the vapour barrier inside the wall construction (on the exterior side of the interior stud wall, usually protected by waferboard). In essence, the interior stud wall was the "service cavity". The poly was out of harm's way, and wiring and plumbing could occupy the interior stud space. Insulating the interior stud space was a "no brainer" and was done as a matter of course. The total cavity available for insulation in my walls was 14 inches.

That was then. Now my renovation of an old stone farm house has re-introduced the use of "service cavities".

Use of a service cavity makes sense when trying to achieve higher insulation levels and protect the integrity of vapour/air barriers. A service cavity also makes sense where it would be inadvisable to run services through construction members. In the case of my farm house renovation, the roof construction has been rebuilt and clay tile roofing installed by a local contractor.

The rafter dimensions and spacing are skimpy to say the least. Rafter cross-section is 2 3/4" x 6" and spacing 24". Clay tile is a very heavy roofing material - for this cottage total weight is over 10 tons. Thus, I certainly am not going to drill or notch any of the roof members for electrical or any other installations, to not reduce their bearing ability. Further, as the framing material for this project was fresh milled and not necessarily dry when installed (and also, no doubt, due to the weight of the clay tile), The rafters are not in one plane, due to sags and some twisting.

The solution here is to build the "service cavity", inside the living space, outside the bearing construction. Not only is there a place to put the electrical and plumbing, the inner wall and ceiling that will create the service cavity will be trued and in one plane. The space in the rafter cavities will be filled completely and covered with vapour barrier before proceeding with the "service cavity".

All the best,

Andris J. Auzins


Jan 23, 2012 4:13 PM ET

Edited Jan 23, 2012 10:17 PM ET.

Unavoidable Penetrations?
by Kevin Dickson, MSME

Before you run off thinking that service cavities are the ultimate solution, remember there are still many wall penetrations required:

1. Sillcocks
2. Exterior electrical outlets as required
3. Exterior lights at all entry doors and patio doors
4. Dryer vent
5. Range hood
6. Bath vents
7. Plumbing vents
8. Gas appliance vents
9. Main service panel/electric meter
10. Gas service
11. Telephone and cable service
12. Air conditioning lines and wires.
13. Doorbell

This seems like enough penetrations to eliminate most of the gains from a service cavity.

What to do? Just a bunch of tape and foam? Or can we eliminate some of these? (e.g. condensing dryer) Most of them are required by existing codes.

Jan 23, 2012 5:25 PM ET

Time has proven that high
by aj builder, Upstate NY Zone 6a

Time has proven that high quality plywood is a great invention. Every existing home I have worked on has for too many reasons led me to the simple conclusion that OSB falls short of plywood and is not worthy of the tiny savings in overall project cost.

Even Advantech turns to mush if left too long in the weather. A 2000sqft unfinished deck here was totally shot in three years. Not a fair test true but.....

For me time has shown me that OSB is worthless.

Nobody owns me so, that's how I feel about OSB.

Jan 23, 2012 7:00 PM ET

Response to Kevin Dickson
by Lucas Durand - 7A

I agree that service cavities aren't an "ultimate solution" - they are a design compromise.

However, I think that if very low infiltration rates over long time scales is a goal, then a service cavity strategy has much to offer.

I think the total number of penetrations is largely influenced by where the pressure boundary is located in an assembly.

Some of the penetrations in your list (exterior boxes/fixtures for example) are not a concern if the pressure boundary is located far enough to the interior side of the assembly - cold climate bias.

As for sealing penetrations...
I think it depends on the components involved...

I am sealing penetrations to plywood and site-fabricating seals from a sheet of neoprene.
The neoprene seals very tightly around conduit, pipes and vents.
None of my electrical boxes/fixtures need to be sealed at all.
The "flange" part of the seal sits flush against the plywood and tapes easily.
The "flange" could probably also be glued down...

Running several wires through a piece of sealed conduit that penetrates the pressure boundary is easier that sealing individual wires.
Still working out the best way to seal the inside of the wire filled conduit...

Daniel Ernst...
Have you figured this one out yet?

Jan 24, 2012 1:33 AM ET

Back to the basic issue
by albert rooks

AJ, Kevin and all of us thinking that the service cavity is not realistic...

Again, all that is going on here is that we are moving the air/vapor barrier back away from the finish layer. That's really the issue at hand. Not what the barrier is made of.

AJ: I don't know why your are caught up on the OSB concept so much. This air/vapopr barrier is just a barrier. What that barrier is being made of is up to you. I like OSB or a low perm membrane. Floris likes a variable membrane. Thorsten Chlupp likes plywood. All of those materials are one heck of a lot better than drywall. These just happen to be our preferences. I think you are in favor of air barriers. Right?

I spent some years as a motorcycle mechanic in a BSA/Triumph shop when I was young. Beautiful things these bike were. But geez, here comes an electronic ignition to replace the age old Lucas distributors and you'd think we were all on our way to heck. It was a suspicious change for many. However it stuck. We get better ideas. Things improve.

Kevin: All those penetrations on your list still have to happen no matter where you place the pressure boundary. No. Nothing is perfect. It's all an effort to make these assemblies a little better, step by meager step. The issue with the boundary is not sealing indoors to outdoors (perhaps I didn't follow your comment properly) it's sealing a cold climate wall from having warm humid air entering by infiltration. The assembly has to have penetrations and they should be sealed by however, and at best effort.

All of this pressure boundary issue is based on a wall that will manage it's moisture. If you're assembling a wall that won't dry and will rot the material... Don't do that!

There is no requirement that exterior sheeting has to be low perm OSB. Floris produced a good drawing that looks to dry well, Martin has reminded us about using exterior foam to "warm things up" (ext sheathing wise) and I'm a fan of high perm sheathing. And finally Herr Chlupp up in Alaska -makes me look like a ninny with his ultra high perm "membrane only" sheathing. His is most likely the most durable of all of them.

I think moving that pressure boundary back from the finish layer is a real step up. Sure it's more work, but I feel much better thinking about putting an emphasis on creating a more durable envelope than a few more interior ft2 or fancier finishes.

Jan 24, 2012 3:13 AM ET

A few thoughts...
by Dan Whitmore

Good article and responses.

After building a Passive House without a service cavity (instead substituting a baseboard chase for outlets) but with the Air Barrier/ Vapor Retarder as OSB directly behind the sheetrock, now whenever someone inquires I always recommend not following my path. Unless necessary I wouldn't do it again for any wall. I spent the majority of the building phase biting my nails and micromanaging all the trades to ensure they didn't compromise the layer. Now it's time to constantly annoy the occupants... for years and years.

A thought on horizontal framing for the cavity: check with your sheetrocker first. Wallboard is usually hung horizontally also, necessitating 1) a good layout to catch the seams and 2) structurally strong enough fastening to keep the sheets from sagging.

Ceilings are a possible location to not use a service cavity (an added benefit of not having one there: getting rid of annoying can lights.) There's always a wall somewhere nearby.

Lastly, after sheathing another roof in plywood, I'm firmly a proponent of using OSB where appropriate. Sadly I imagine the Ply industry is working hard to meet the price point of OSB, but closely inspecting each & every sheet or having to remove sheets just installed because of defects is very costly and annoying. Yes it was only 4-ply 15/32", not the better 5-ply or 5/8" but that's even more expensive. No I won't use OSB for roofing applications, but for an interior air barrier/structural layer, absolutely. Also, in order to reach the air infiltration levels we're achieving, not being able to trust a plywood barrier to perform on account of imperfections just is a headache I don't want to fight.

Jan 24, 2012 8:29 AM ET

Magic Plywood
by John Brooks

Plywood is more than "smart" it is magical

snipit taken from "Mind the Gap"


Jan 24, 2012 9:39 AM ET

Response to Albert
by Lucas Durand - 7A

Kevin: All those penetrations on your list still have to happen no matter where you place the pressure boundary.

I'm not sure this is universally true.

In my case, the pressure boundary has holes for neither interior nor exterior boxes/fixtures.

Jan 24, 2012 10:50 AM ET

Edited Jan 24, 2012 10:53 AM ET.

Response to Lucas
by albert rooks

Most of the penetrations on the list were venting or services: Dryer, range, bath, gas appliance, elect, gas cable, phones service... All of those would cross the envelope from inside to outside (or vice versa), and therefore cross the pressure boundary. My point was: Whether there was a service cavity and the pressure boundary was moved back from the finish layer has no affect on these either way (all joint sealing being equal). They still exist regardless of where you place the pressure barrier. -Generally speaking.

Jan 24, 2012 11:29 AM ET

Dan, what would you do now?
by Lucy Foxworth

You said that you wouldn't recommend anyone doing what you did. What would you recommend now? A full wall service chase - is that what you are saying?


Jan 24, 2012 1:01 PM ET

Edited Jan 24, 2012 1:05 PM ET.

Dan? Seriously? This topic
by aj builder, Upstate NY Zone 6a

Dan? Seriously? This topic had you biting your nails???

Martin, your hammer please..... Forget it, have four in the truck.

This thread has been the most entertaining since our last years building/religion/politics wars.

More posts .... Enjoying this. Why I may even chase something today.... Wish I had a tail right now.

(This post is meant to be completely in fun and not meant to be harmful to anyone!)

Jan 24, 2012 1:07 PM ET

Nail biting
by Martin Holladay

AJ, Dan is a conscientious builder. I'm glad he bites his nails.

We could use more conscientious builders like Dan who care about the details and worry over getting things right.

Jan 24, 2012 2:18 PM ET

Edited Jan 24, 2012 2:22 PM ET.

My two cents
by Garth Sproule 7B

Just to further Kevin's point about penetrations...(btw he might also include the two large pipes necessary for a HRV on his list) All these penetrations have to be sealed...permanently. Rather than fuss with neoprene or expensive tapes to seal behind a service cavity, where it is not inspectable, I personally would rather deal with all penetrations at the drywall layer, where I can seal it with cheap caulking if I want and inspect and repair as necessary. Simpler, faster, less costly, inspectable, and repairable.

Sure, someone can kick a hole through the drywall and ruin the barrier. Sorry, but I have no sympathy for someone who would do that and not repair it.

The main benefit to a service cavity, as Lucas points out, is that the electrical boxes themselves do not have to be sealed.

Jan 24, 2012 2:20 PM ET

Edited Jan 24, 2012 6:39 PM ET.

I vote no on service cavities
by Doug McEvers

I am glad I was on vacation, reading this has given me a headache. Production builders will not be doing service cavities soon, it is expensive and not needed. We have been building superinsulated homes for over 30 years in a cold climate with a warm side air barrier and they are doing just fine. These homes test in the 1 ach50 to 1.5 ach50 range with comprehensive air sealing details. A thick wall has a lot of diffusion capability for the very minute amount of air and moisture passing through. Too many people are relying on what they hear or read as to the durablity and performance of walls. The proof is in the results, revisit some built homes to learn about what works.

We might not all be building superinsulated homes in the near future but they are proven to be durable and are quite easy to build.

Jan 24, 2012 2:25 PM ET

Question for Doug
by Garth Sproule 7B

What are you using for an air barrier in the homes you describe...drywall or poly?

Register for a free account and join the conversation

Get a free account and join the conversation!
Become a GBA PRO!