musingsheader image
0 Helpful?

Banish These Details From Your Plans

Some features and methods are so difficult to implement that designers might want to avoid them entirely

Posted on Jun 27 2014 by Martin Holladay

Is it possible to disassemble old shipping pallets and glue the pieces of lumber together to make furniture? Of course it’s possible; some woodworkers have used this method to make beautiful tables and chairs. There’s a fly in the ointment, however: while it’s possible, it’s not very easy.

Many commonly used construction methods, design details, and materials fall into a category I would call “possible but not easy.” I decided to create a list of items that fall into this category.

The unfortunate twist to this category of difficult methods is an apparent paradox: though hard to do well, these methods are actually fairly easy to implement if you are willing to do a sloppy job. (That’s why many of the methods are common.) The methods are only difficult for builders who want to build a tight, high-performance home.

My list is deliberately provocative. Some readers will respond, “I’ve used many of the methods on your list, and I always do a good job.” To which I will answer, “That’s great. But most builders aren’t as conscientious as you are.”

I’m addressing this article to architects and designers. Since these methods are hard to implement well, it’s best to leave them off your plans and specifications. Use other methods.

Martin’s ‘possible but not easy’ list

Look your client in the eye and just say no. It may take a little practice, but you can do it.

Crawl spaces. Most crawl spaces are damp, leaky, poorly insulated, and rarely visited. Creating a dry crawl space is possible, but it’s rare for a builder to get all the necessary details right.

Basements can be damp, too — but at least they are easier to access. (Easy-to-access spaces are more likely to be frequently visited, and frequently visited spaces are less likely to have problems that go unnoticed for years.)

Slab foundations have drawbacks — but nowhere near as many pitfalls as crawl spaces.

For more information on this issue, see Building an Unvented Crawl Space.

Dormers. Occasionally, an architect will design a dormer for a vented unconditioned attic, in hopes of creating a roof that looks lively and exciting from the curb. (Needless to say, this is nuts. It’s the kind of detail you would see on a Hollywood set.)

In most cases, however, dormers are specified when the space under the roof is conditioned spaceInsulated, air-sealed part of a building that is actively heated and/or cooled for occupant comfort. . The components of this type of dormer are part of the home’s thermal envelope, and it’s hard to get the details right. The thermal envelope has to be continuous. That means that you need to have insulation that follows the roof slope; when you pop a dormer, you need to make sure that all of the barriers — the 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., the water management barrier, and the insulation barrier — are uninterrupted, and that the barriers on the dormer cheeks are continuous with these barriers on the main roof.

Let’s face it — the air barrier ends up like Swiss cheese. Moreover, the framing used for dormer cheeks and dormer rafters is never thick enough for an adequate insulation layer, and the roofer’s step flashing details at the base of the dormer cheeks are often faulty.

So — don’t do dormers.

Bay windows. Bay windows are almost as difficult to detail well as dormers. In most homes, this type of bump-out is an energy nose-bleed.

If you insist on building a bay window, make sure that you have developed good details to limit air leakage and to provide a thick layer of insulation.

Garrison overhangs. The second floor of a Garrison-style home has a greater area than the first floor. This magic is made possible by cantilevering the floor joists a foot or more, creating an exterior soffit.

You can guess what’s coming: I don’t like Garrison overhangs. These bump-outs are hard to seal, so they usually leak abundant quantities of air.

Windows in showers. You want a window in your shower, don’t you? Maybe you live in the country. You don’t have neighbors, and you enjoy the little frisson of excitement you feel when you shower in front of your window. Or maybe you just want to look at the trees or enjoy the natural light.

Can you install a window in a shower in a way that keeps water out of your wall? Sure. With enough copper flashing, butyl gaskets, and peel-and-stick tape, you can do almost anything. Do most window installers get these details right? No.

Fiberglass 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. . Yes, it's possible to install fiberglass batt insulation in such a way that it achieves the R-valueMeasure of resistance to heat flow; the higher the R-value, the lower the heat loss. The inverse of U-factor. advertised on the package — but it doesn't happen often. If you care about the thermal performance of the building you are designing, it's best to specify a different type of insulation.

To learn more about this issue, see Installing Fiberglass Right and Should Batt Insulation Be Outlawed?

Recessed can lights in insulated ceilings. Many builders have tried and failed to make recessed can lights airtight, using a combination of caulk, spray foam, and site-built covers. They've also tried to heap enough insulation on top of the fixture to meet minimum code requirements, only to be stymied by insufficient clearance to the roof 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. or difficulties keeping the mounded-up insulation in place.

Fortunately, most designers who focus on energy efficiency have sworn off recessed cans. The biggest remaining problem is that designers are weak; they need to develop a backbone to better resist homeowners who plead for recessed lighting.

For more information on this issue, see Recessed Can Lights and Ban the Can.

Brick veneer. Brick claddingMaterials used on the roof and walls to enclose a house, providing protection against weather. is durable, low-maintenance, and attractive. Traditionally, however, bricks were not intended as a facing for structural wood framing. The idea of marrying masonry cladding to structural wood framing is a relatively recent development, and it is fraught with problems.

How many ways are there to screw up brick veneer? Lots. The builder can use the wrong wall sheathing; the builder can install the water-resistive barrierSometimes also called the weather-resistive barrier, this layer of any wall assembly is the material interior to the wall cladding that forms a secondary drainage plane for liquid water that makes it past the cladding. This layer can be building paper, housewrap, or even a fluid-applied material. (WRB) with reverse laps; the mason can forget the weep holes at the bottom of the wall; the mason can fail to install durable flashing near the weep holes; the weep holes can end up below grade after landscaping work is completed; the mason can fail to install proper flashing and weep holes above a bay window; and the mason can clog up the drainage space behind the bricks with gobs of mortar.

Is there a good way for the architect and builder to make sure that all of these details are done right? Not really — not without a webcam.

Including brick veneer on this list is something of a stretch, I know — and I'm going to get pushback on this issue, especially from builders in Texas and Georgia. But designers and builders who specify brick veneer need to be aware of the pitfalls of this cladding, and need to have a plan for supervising the work carefully.

Adhered manufactured stone cladding. Adhered manufactured stone cladding over OSB sheathing is probably the riskiest cladding option for a wood-framed building. Hundreds — perhaps thousands — of homes with this type of cladding have had serious problems with moisture intrusion and rot, dragging the homeowners, builders, designers, and their insurance companies into protracted construction-defect negotiations.

If a designer and builder can come up with details that include a ventilated rainscreenConstruction detail appropriate for all but the driest climates to prevent moisture entry and to extend the life of siding and sheathing materials; most commonly produced by installing thin strapping to hold the siding away from the sheathing by a quarter-inch to three-quarters of an inch. gap between the cladding and the OSB, along with bulletproof flashing details, this cladding option might work. If you want to go that route, you have a stronger stomach than I do.

For more information on this issue, see All About Wall Rot.

Ground-source heat pumps. There are two hurdles that need to be crossed if you want a residential ground-source heat pump (GSHP) that performs well: system design and equipment installation. Both are difficult. Energy experts have learned that both steps are often muffed.

For more information, see Air-Source or Ground-Source Heat Pump? and Are Affordable Ground-Source Heat Pumps On the Horizon?

Ducts. Most ducts are undersized. Most duct seams are leaky. Many ducts are convoluted and are located outside of a home’s thermal envelope. To overcome a duct system's static pressure, furnace fans and ventilation equipment must expend a significant amount of energy.

If you are designing a home, one solution to these problems is to choose systems (for example, ductless minisplits or Lunos ventilation fans) that don’t require ducts.

If you decide to install equipment that requires ducts, make sure that:

  • The duct system is designed according to standard engineering principles;
  • The ducts are installed as shown by the system designer;
  • Duct seams are sealed with mastic or high-quality tape to make them as airtight as possible;
  • All ducts are located inside the home’s conditioned envelope.

For more information on this topic, see:

Can all of these features be avoided?

It’s perfectly possible to build a house that has none of the features on this list. Such a house will be easier to build and will be more likely to perform well than a house that includes several of these features.

If you are an architect who wants to include some of the features on this list, my opinions are unlikely to dissuade you. Just make sure that you provide your builders with plenty of detail drawings. And it might be a good idea to stop by the job site frequently to make sure that everything is going according to plan.

What did I forget?

If you think I've left off a few important items from this list, feel free to set me straight in the comments section below. After all, adding to the list is fun.

So far, the suggested additions to the list include:

  • Beams or joists that penetrate the thermal envelope.
  • Stucco.
  • Obscenely complicated rooflines and ceiling height variations.
  • Roofs with stingy overhangs.
  • Chimneys.
  • Cathedral ceilings.
  • SIPs.
  • Skylights.
  • Glass that extends to the floor.
  • Sliding doors.
  • Decks over living spaces.
  • Flat roofs.

The above bullet list includes the most defensible suggestions posted below by readers of this article. Although I have cataloged the suggestions as a public service, I am not ready to defend all of them. To learn more about the reasons why these items were proposed for inclusion on the list, read the comments below.

Martin Holladay’s previous blog: “How Balanced Ventilation Systems Become Exhaust-Only.”

Click here to follow Martin Holladay on Twitter.

Tags: , , , , ,

Image Credits:

  1. Window in the shower: Juhan Sonin

Jun 27, 2014 8:27 AM ET

One on, one Off?
by Malcolm Taylor

I'd add structural elements that penetrate the building envelope. Beams and joists that create thermal bridges and are also difficult to flash. Cantilevered decks being one particularly poor detail.

In return I'd like to make the case for crawlspaces. While slabs may be a preferred default method, there are many sloping sites where they aren't feasible, demanding too much fill, very tall foundation walls or compromises to the floor plans. Basements are often finished living space or at least regarded by their unsuspecting owners as providing dry storage - both bad ideas below grade. A crawlspace, detailed as part of the conditioned envelope, doesn't pretend to be something it isn't. Unlike a slab or basement, a house with a crawlspace provides access to the services, is easy to renovate or add to, and suffers little damage if the space is compromised by floods originating either inside or out. Like attics they do suffer from being largely unseen and thus neglected by the builder, but that isn't an inherent fault of theirs. Built with care I don't think we should write them off yet.

Jun 27, 2014 8:42 AM ET

Second on cantilever
by Dan Kolbert

Monster beams in general are a problem - often taking up space where insulation should go.

And while not perhaps in the same category, be careful that improved window performance doesn't lead you to keep putting bigger ones in. As my colleague Wes Riley says, the best window is still a lousy wall.

Jun 27, 2014 9:34 AM ET

Edited Jun 27, 2014 9:39 AM ET.

Response to Malcolm and Dan
by Martin Holladay

Malcolm and Dan,
Thanks for your comments and additions.

"Beams or joists that penetrate the thermal envelope" definitely belong on the list. (I recently included a rant about this type of penetration in an article for Fine Homebuilding, "Second-Story Balconies." In that article, I wrote, "For years, some designers have created second-floor balconies by cantilevering floor joists (or a floor slab) through the thermal envelope of the building. However, there are two serious problems that are associated with following this practice: thermal bridging and moisture management.")

One of my favorite "penetrating beam" photos is one I took of a concrete beam at Harvard University. (Perhaps the architect was a Harvard alum?)

[Click to photo to enlarge it.]


Thermal bridge at Harvard University - low res.jpg

Jun 27, 2014 9:46 AM ET

Another suggested addition
by Martin Holladay

Allan Edwards has already started a thread on the JLC website, linking to this article. Allan suggests that stucco belongs on the list.

When I was drafting the article, I considered including stucco -- after all, I listed all the reasons why stucco is risky in my 2010 article, To Install Stucco Right, Include an Air Gap -- but then I decided to be sensitive to the feelings of all of the stucco-lovers from from California.

I thought, "I've already written about stucco. Maybe it's time to alienate readers from Texas. I'll write about brick veneer instead."

Jun 27, 2014 9:48 AM ET

Crawl spaces don't belong?
by Martin Holladay

You make a good point about crawl spaces on sloped lots; thanks for pointing that out.

Jun 27, 2014 10:15 AM ET

Good list
by Carl Seville

Martin - I pretty much agree with all your points, however it is going to be a long, slow haul to get architects and homeowners to give up those obscenely complicated rooflines and ceiling height variations. The last house I built a while back (over which I had very little design control) had 35 different roof planes and almost as many individual exterior wall sections. Lots of people are resorting to spray foam insulation to make these crenellated boxes sort of perform, but it really gets back to planning for high performance from the beginning instead of trying to layer it on after the design. I do not hold out much hope that there will be significant change in the industry in this regard, at least not until energy prices skyrocket, which, given the political climate isn't very likely. We will just have to keep chipping away at the problem, piece by piece. I do suppose the slow pace of change does provide us with some job security at least.

Jun 27, 2014 10:19 AM ET

Response to Carl Seville
by Martin Holladay

I've added "Obscenely complicated rooflines and ceiling height variations" to my new bullet list at the bottom of the blog. Thanks.

Jun 27, 2014 11:11 AM ET

by Malcolm Taylor

Spare a thought for the designers at your local truss plant. My salesman went on a rant on this topic the last time we talked. Packages with not one common truss. dead valleys, individual trusses that end up looking like silhouettes of dinosaurs - they look back to the past with a fair amount of nostalgia.

Jun 27, 2014 11:15 AM ET

Edited Jun 27, 2014 11:16 AM ET.

Response to Carl and Malcolm
by Martin Holladay

I agree with both of you. For more ranting on the topic, see Martin’s Ten Rules of Roof Design.

The photo of the old Garrison at the top of the page is instructive. The cantilevered joists may be problematic, but the old designers got the roof details right.

Jun 27, 2014 12:22 PM ET

Edited Jun 27, 2014 12:25 PM ET.

Not all Concrete Beams and Cantilvered Decks Are Bad
by Peter L

Building Science has caught up with the concrete cantilevered decks. I just finished designing a home that was thermally broken by utilizing Schock Isokorb Thermal Breaks on the 2nd floor balconies:

So concrete slabs that transition from the exterior to the interior and carry loads can be thermally broken by utilizing Isokorbs. Martin, you can put an asterisk * next to the concrete balcony because there are now ways to fix the thermal bridging problem. Isokorbs have been used in Europe for 20 years and they work.

So that concrete beam in your photo can be thermally broken & insulated but of course it would have to be done during the construction and not after the concrete has cured. There are no deep energy retrofits for that detail.

Jun 27, 2014 12:47 PM ET

wood rot
by Debra Glauz

For long term durability, structural elements made of wood that are integral to the engineering of the home should be completely protected by overhangs and not subject to constant wetting or dew. Wood placed beyond the protection of an overhang should be considered sacrificial and replacement should be expected and will be part of the maintenance of the home.
This probably will not make the list but flat roofs with no overhangs are much riskier than simple sloped roofs with large overhangs.
Hey, you’re hurtin my feelings. I’m from California. Take stucco off the list. Stucco is here to stay. Stucco doesn’t burn. Done correctly stucco is less problematic than wood siding in the area where I live. There comes a point when new methods must be taught. Builders and their employee’s often have the skill but they haven’t been taught and are often not aware of the latest and best methods of their trades.

Poke To bad we don't have an arbitrary metric that will slow designers, architects and homeowners from designing homes with those obscenely complicated rooflines and ceiling height variations.

Jun 27, 2014 12:54 PM ET

Response to Peter L
by Martin Holladay

I agree that the Schöck Isokorb system is a good way to thermally isolate a concrete balcony from an interior slab.

But I disagree with your terminology. You wrote, "Not all concrete beams and cantilevered decks are bad." But if you build a balcony with the Schöck Isokorb, you (fortunately) don't have a cantilevered balcony; what you have is an interior concrete deck which is attached to the Schöck Isokorb hardware, along with a separate exterior concrete balcony which is attached to the opposite side of the Schöck Isokorb hardware.

Fortunately, these are two separate pieces of concrete -- so nothing is cantilevered. (See the illustration below; click on the image to enlarge it.)


Schöck Isokorb.jpg

Jun 27, 2014 1:04 PM ET

Response to Debra Glauz
by Martin Holladay

Thanks for your comments. I have added "Roofs with stingy overhangs" to the bullet list at the bottom of the blog.

Concerning your stucco comments -- "Hey, you’re hurtin' my feelings. I’m from California. Take stucco off the list" -- I will note in my defense that stucco was not on my original list; it was suggested by Allan Edwards.

As I noted in Comment #4, "I decided to be sensitive to the feelings of all of the stucco-lovers from from California."

Jun 27, 2014 1:06 PM ET

Not all fiberglass batts hit their labeled R, even if perfect.
by Dana Dorsett

Low density R19s are only R18 when compressed to 5.5", as in a 2x6 wall assembly.

Low density R22s are only R19 when compressed to 5.5" in a 2x6 wall assembly.

And this is straight off the manufacturers' compression charts: (Owens-Corning's chart.)

High density R21s can hit their labeled performance, which means the outperform the R22 labeled batt by 10%, despite the lower number.

This is a significant flaw in the labeling & testing regulations IMHO, since some batt products can, and others simply can't meet the labeled performance, even when installed perfectly per manufacturer's instructions in the framing for which they were designed. R22s would rightly be labeled R19, and R19s labeled R18. Given that R19s & R13s are the most widely used wall insulation products in the US, labeling at it's true performance seems reasonable, even if many installers screw it up and undercut performance even further.

Of course if you just lose the batts you lose the problem, eh? :-)

Jun 27, 2014 1:18 PM ET

Edited Jun 27, 2014 1:19 PM ET.

Response to Dana Dorsett
by Martin Holladay

I understand your point. But there is, I think, a historical explanation for the oddities you note.

You wrote that some fiberglass batts "simply can't meet the labeled performance, even when installed perfectly per manufacturer's instructions in the framing for which they were designed." Here's the key: the manufacturers will note in their defense that the batts you highlight were designed for attics, not wall cavities -- and therefore are intended to be installed at their labelled thickness.

I know it's hard to believe, but I remember the day when walls were insulated with 3.5-inch batts and attic floors were insulated with 6-inch batts. (Yes, I grew up with rotary-dial phones and a black-and-white TV.)

Those 6-inch batts were never intended to be crammed into 5.5-inch-deep stud cavities (say the manufacturers in their defense).

Yes, times have changed. Our building codes no longer allow R-19 batts to use used as ceiling insulation (unless, of course, you install multiple layers of batts).

I agree with you that it's time for the labels on insulation packages to reflect how these batts are actually used.

Jun 27, 2014 1:36 PM ET

Edited Jun 27, 2014 1:44 PM ET.

That's too sheer a fig leaf ( response to Martin)
by Dana Dorsett sheer that they're as good as nekkid on that one. Read the packaging on kraft-faced R19s:

Seems the font for "2 x 6 WALLS" is a bit bigger than "floors" on the R19 Owens-Corning, the only listed assembly on the J-M or Owens-Corning R22, and first listed application on the Certainteed R19 bag.

That's a labeling problem.

R19 batts still cuts in for state codes built on IRC 2006, even in places as cold as Minnesota, but they are really only R18.

Jun 27, 2014 1:46 PM ET

Edited Jun 27, 2014 1:48 PM ET.

Response to Dana Dorsett
by Martin Holladay

Even though my little history lesson was accurate, I have to admit that those labels are misleading.

(And, it should go without saying, I was never defending the fiberglass manufacturers...)

Jun 27, 2014 3:48 PM ET

Edited Jun 27, 2014 3:53 PM ET.

by Peter L

Funny you should bring that up. I was just talking to Schock and my engineer and even though the Isokorb is attached to both sides of the "cantilevered" balcony, it still is technically cantilevered and engineers & Schock views it as so. One can thermally break a SUPPORTED concrete balcony and an UNSUPPORTED/CANTILEVERED balcony and Schock can provide Isokorbs for both applications.

In addition, Schock can thermally break an exposed concrete beam with their Isokorbs. So it's not just balconies but beams and concrete joists and columns.

Maybe we are splitting hairs but technically with the Isokorb it is still a cantilevered balcony since the slab is "free floating" with no support from underneath. The loads are transferred through the Isokorb to the interior slab. Even without the Isokorb, the loads are transferred the same way. The Isokorb is the high-tech product that will transfer the loads while thermally breaking the slabs.

Jun 27, 2014 4:21 PM ET

Response to Peter L
by Martin Holladay

Thanks for the explanation. Perhaps I would put it this way: the load is cantilevered, but the concrete slab is not cantilevered. But, as you said, this discussion is an attempt to split hairs.

The most important point is that if you use the Schöck Isokorb product, you don't have a slab penetrating the thermal envelope -- and that's good.

Jun 27, 2014 5:37 PM ET

by Peter L

When I requested to utilize the Schock Isokorb on my design the US engineers had no clue what I was talking about. They never knew that such a product existed. The German engineers knew it very well but the US engineers were thrown for a loop. Amazing product and it allows cool looking designs without affecting thermal efficiency.

I guess that is why in Germany you can buy triple pane windows at the big box store while here in the USA the big box stores would think you are crazy if you asked for triple pane windows. We are behind the curve, that is for sure. Europe gets the cool energy stuff before we see it here. Don't get me started on why Europe has dozens of heat pump dryers available while we are still struggling to get one by 2014.

Jun 27, 2014 5:58 PM ET

As an owner of neglected rental homes
by Kevin Dickson, MSME

EIFS, vinyl siding, all chimneys, T-111, cathedral ceilings.
SIPs don't have high enough R value anymore.

Jun 27, 2014 8:32 PM ET

Edited Jun 27, 2014 8:33 PM ET.

by Malcolm Taylor

Stucco almost completely disappeared here in the PNW after the widespread building envelope failures we had - as our climate isn't as amenable to using it as yours is. There was a real gap as builders looked for something comparable to use, but recently Hardi panels with metal trim strips are becoming the material of choice. They have most of the attributes of stucco, minus the cracks, but do have to be re-painted. I can't say I miss the old style stucco with its sharp aggregate. My shoulder still has scars from sliding along a wall returning home late one night.

Jun 27, 2014 9:54 PM ET

Exterior Masonry Chimney's
by terry grube

I would add masonry chimney's installed on exterior walls.

Jun 28, 2014 5:51 AM ET

Edited Jun 28, 2014 5:53 AM ET.

Response to Kevin Dickson (Comment #21)
by Martin Holladay

Just to be sure that everyone understands this list: it's a "possible but not easy list."

It's not a "I hate these materials" list.

EIFS? I happen to think that EIFS is the best system for insulating an old building with multi-wythe brick walls.

Vinyl siding? It may look ugly to some, but it performs better than almost any other siding -- because it is naturally well drained and ventilated.

All chimneys? This suggestion will end up on the list, I think. (I wrote a blog in 2009 called Farewell to the Chimney?) Once chimneys are on the list, however, you have to be ready to hear complaints from people with wood stoves.

T-111? Like your suggestion to include vinyl siding on the list, I think this belongs on the ugly list, not on the "possible but not easy" list. Panel siding is fairly easy to detail well.

Cathedral ceilings? Sure. I'll put them on the list.

SIPs? You'll get an argument, but I'll put SIPs on the list, too -- because of the difficulties inherent in sealing the seams, and because of the susceptibility of the outer OSB facing to rot unless there is a venitlated air gap on the exterior.

Jun 28, 2014 6:01 AM ET

Response to Terry Grube (Comment #23)
by Martin Holladay

When it comes to chimneys, Kevin Dickson beat you to it -- and his suggestion was even more categorical than yours.

For more information on chimney placement, see All About Wood Stoves.

Jun 28, 2014 11:49 AM ET

Edited Jun 28, 2014 11:54 AM ET.

by Peter L

Ironic that cathedral ceilings and SIPs were just put on the list since cathedral ceilings can be properly done if you utilize SIPs. The comment about SIPs that, "because of the difficulties inherent in sealing the seams, and because of the susceptibility of the outer OSB facing to rot unless there is a ventilated air gap on the exterior." The problem I have with the comment that sealing the seams is an "inherent difficulty" is that it's simply not true if you know what you are doing. Any seam if not sealed correctly poses a problem and that includes wall seams, window seams, door seams, etc. If sealing seams was a reason for banishing then everything would be banished.

As far as the OSB rot goes. Creating roof gap vents is a solution to the potential of SIP roof rot. Other remedies exist and like anything else, it's all climate dependent. Living in the desert SW vs Juneau Alaska are worlds apart in regards to how one would approach the details. Stucco is everywhere in the desert SW and performs quite flawlessly, not to mention it is fireproof. To banish stucco in the southwest makes no sense.

I agree that certain building conditions are bad for each and every climate but banishing things like stucco, SIPs and cathedral ceilings (which can be done properly with SIPs), is just "throwing the baby out with the bathwater" type of approach. Building codes are regional dependent. What works in the desert might not work in the Pacific Northwest but that doesn't mean we should just discard it because it is not universal.

I believe roof SIPs do work when like anything else, they are done properly. An 8" urethane panel has a R-50 value and it can resolve a host of problems associated with stick frame roof trusses. One can lay a SIP over the wall and create a 24" overhang. With roof trusses you create complex transitions and issues with air sealing, insulating, thermal bridging, venting, etc. One can have a SIP roof done in 3-4 days on a home while it would take weeks to install trusses, sheath the roof, insulate the roof, build soffits, etc. A cathedral ceiling done with a SIP is a match made in building heaven.

Jun 28, 2014 11:57 AM ET

Response to Peter L
by Martin Holladay

Thanks for your votes.

I'm maintaining the list, not necessarily endorsing it.

Jun 28, 2014 12:04 PM ET

by Malcolm Taylor

Wouldn't the distinction between the problems of sealing SIPs seams and others be that air leaks in SIPs cause an inordinate amount of damage that compromises both the building envelope and the structure of the building?

Jun 28, 2014 3:15 PM ET

Edited Jun 28, 2014 3:18 PM ET.

by Peter L

Air leakage is bad on all fronts, both in structures (SIPs, wood frame) and of course in terms of energy loss. SIPs are a great technology and offer a lot of advantages over stick frame roof trusses. As with any product it has some drawbacks and that drawback is if you don't properly seal the joints, you can experience problems.

A properly sealed SIP is 100% sound and will last a lifetime. Sealing a SIP is not rocket science either. Here is what the industry has concluded:
* Use a T&G Joint SIP
* Flexible rubber gaskets are installed at each T&G Joint
* The interior/bottom side of the roof SIP is sealed with peel & stick tape like SIGA or similar

Following the above will ensure a completely air sealed SIP and provide a worry free SIP roof. You have a triple layer of protection. So it's beyond "belt & suspenders", it's "belt, suspenders, and a second belt"

Jun 28, 2014 3:33 PM ET

easy to well
by aj builder, Upstate NY Zone 6a

T1-11 easy and new homes can look great with it in a wooded setting. Price is right... Dual function... I love it.

Vinyl... Same but great in the burbs.

Jun 28, 2014 8:35 PM ET

Edited Jun 28, 2014 8:41 PM ET.

Possible But Not Easy
by Kevin Dickson, MSME

It's possible to use painted T-111 siding but not easy to keep it looking good. It stays looking much better with well-maintained stain (see below).

It's not easy to keep vinyl siding looking good for more than 15 years. Buyers are told "you never have to paint again" but the truth is they are much better off with painted siding in the long run.

It's not easy to keep EIFS looking good for more than 20 years. It's not that hard to repair though.

12" roof SIPs are R48. That doesn't meet the IRC 2012 requirement, and future requirements may be higher. Detailed right (which IS easy), they are the best way to do low slope roofs, cathedral ceilings, and generous overhangs. R60 SIPs (when available) should get them off this list.


Jun 28, 2014 11:03 PM ET

Edited Jun 28, 2014 11:09 PM ET.

by Peter L

An 8.25" urethane SIP is rated at R-50 and it meets the 2012 IRC requirements in Zones 1-8. I know this and verified this through my building department and I am located in Zone 4. Now Zones 5-8 would benefit from a higher roof R-Value but a 8.25" urethane SIP will pass 2012 code.

With that being said, even though it passes code, I will bump my R-Value by adding 3" of Roxul Batts on the underside of my ceiling SIP. This will bring the R-Value to around R-60 and add a fire protection layer to the underside, along with the sheet rock that will go beneath it.

My design has cathedral/vaulted ceilings in some areas and 24" overhangs. My pitch is 5:12 and I can have the entire SIP roof installed in 4-5 days. Doing it as a stick frame would prove much more difficult and time consuming. Not to mention I would have to vent the soffits/overhangs and vent the roof, which is a BAD idea in wildfire country. A gable, soffit, attic vent in pine country is an invitation for flying embers and for your house to burn down. 90% of wildfire houses that are burned to the ground are caused by embers landing inside the attic via vents. Nothing GREEN about having your house burn to the ground.

Jun 29, 2014 11:50 AM ET

90% caused by embers entering thru attic vents??
by John Brooks

Peter L wrote:"90% of wildfire houses that are burned to the ground are caused by embers landing inside the attic via vents"

That's a "fantastic" statistic ... do you have a source?

Jun 29, 2014 4:03 PM ET

Edited Jun 29, 2014 4:04 PM ET.

by Peter L

No fantasies here. I live in wildfire country and I've personally been up close to last years fire that claimed the life of 19 firefighters. Embers can travel over 2 miles in the air from the source of the fire. Also a fact and not a fantastic stat.

Embers are the #1 leading cause of home loss during a wildfire, More houses burn due to flying embers than any other cause. Even the 1/8’’ mesh screening for vents will allow embers in and laboratory studies proves this. IBHS has tested this and verified it. Vented attics and soffits in a wildfire puts you at a high risk for having your home burn to the ground in wildfire country, even when defensible space is being employed.

Jun 29, 2014 4:18 PM ET

Edited Jun 29, 2014 4:21 PM ET.

source of the 90% claim
by John Brooks

Hi Peter, I don't doubt the danger of embers....
I am surprised by the "90%" via attic vents claim....
I browsed the links you posted and could not find the "90%" statistic

Jun 29, 2014 10:36 PM ET

What about skylights?
by kim shanahan

If windows make lousy wall then skylights make even worse ceilings. Natural light? Great for GBA writers working out of their homes who don't want to turn on bathroom lights to pee, or retiring baby boomers wandering around the house looking for a project to keep themselves occupied. But when half of the year we get up in the dark and go to work and then get home in the dark after work, who are they serving? Especially since during those dark months the energy is flowing out of those R-1 bubbles faster than we can produce it. I used to love skylights. I used to love bell-bottoms jeans also. No more. Great for commercial works spaces and offices where natural light is supposed to enhance productivity, but not in homes. Turn on the motion-sensing
full-spectrum LED powered by your PV panels.

Jun 30, 2014 6:57 AM ET

Response to Kim Shanahan
by Martin Holladay

I have added skylights to the list. Thanks.

Jun 30, 2014 8:27 AM ET

Response to Peter L (Comment #29)
by Martin Holladay

You wrote, "Here is what the [SIP] industry has concluded: Use a T&G joint SIP; flexible rubber gaskets are installed at each T&G joint; the interior/bottom side of the roof SIP is sealed with peel & stick tape like SIGA or similar."

The method you describe is a good one, but I am not aware of a single SIP company in the U.S. that recommends such an approach. Instead of using EPDM or rubber gaskets, they all still recommend that seams be sealed with canned spray foam. For example, see:

Insulspan Installation Guide: "All gaps between wall and roof panels may be filled with approved Insulspan two-part polyurethane foam sealant before siding and roofing."

Timberline Panel Installation Guide: "The Timberline Panel Company prescribes the Drill and Fill method to seal the seams between wall and roof panels where routed foaming channels occur."

Jun 30, 2014 1:33 PM ET

Edited Jun 30, 2014 1:33 PM ET.

by Peter L

That is true that there is no general consensus in the SIP industry for these details but SIP companies like ThermoCore are ahead of the curve and they recommend the above recommendation of T&G, gaskets and tape. As with any newer green building material, R&D involves time and the GBA forum has been a tremendous help in this realm. I learned that air infiltration in the panel joints is what causes the SIP rot issue as based on what Joe concluded from his Alaskan SIP rot study. Here are some details showing the T&G and gaskets:

These 3 techniques to sealing a SIP roof would solve 95% of all the SIP roof issues but sadly the industry is slow to change. SIPA for some reason is slow to change and like building codes, they drag their feet for years before any changes are made.

Jun 30, 2014 1:52 PM ET

Reponse to Peter L
by Martin Holladay

I'm glad that at least one SIP company has moved away from using spray foam for sealing seams. The Thermocore installation instruction describe their gaskets as "foam gaskets," so it doesn't sound like it's EPDM. Still, it sounds like a step in the right direction.

But clearly, the SIP industry as a whole isn't on board with this idea yet.

Jun 30, 2014 3:50 PM ET

Edited Jun 30, 2014 3:51 PM ET.

Recessed Can Lights
by Eric Habegger

Martin, I agree that recessed can lights don't have a very good history with air tightness. But, I think you may be throwing the baby out with the bath water. The more recent LED downlights are one piece construction that may eliminate many of the problems that used to exist. The old fluorescent and incandescent bulbs required a separate ceiling (not sealing) ring and there was really no good way to air seal between that ring and the box above the ceiling. You were forced to seal from the attic around the whole box.

I don't think that is necessarily the case with the newer Crees. There may be other manufacturers by now also copying that construction. These are built with the ring as part of the bulb assembly. It seems to me that one only needs to put a weather stripping gasket on the ring now. Also a dab of silicone on the small tangs that poke through holes on lense to hold it on wouldn't hurt.

I haven't tried it myself yet but I have plans in the works to do just that. I think the success or failure just depends on developing the proper technique more than anything else. This is no different than air sealing door or windows, in fact it would be much simpler. You do have to remember to not install even these lights less than about 3 feet from any perimeter wall to allow sufficient insulation over the top of the box in the attic. I've included a picture.


Jun 30, 2014 4:00 PM ET

Edited Jun 30, 2014 4:01 PM ET.

Response to Eric Habegger
by Martin Holladay

If you insist on using recessed can lights, those units are probably the type to choose. (I do wonder, however, whether those radiator fins -- clearly intended to dissipate excess heat -- play nice with insulation.)

There are still many pitfalls for the unwary: some installers won't caulk around the trim, so the installation won't be airtight. Some installers won't heed your warning about installing them in locations where there is plenty of room for enough insulation above the units. Some of the units will end up being little "hot spots" in the insulated roof assembly, leading to ice dams.

So I still don't like them.

Jun 30, 2014 4:33 PM ET

Edited Jun 30, 2014 7:06 PM ET.

I meant cans, not boxes, in previous comment
by Eric Habegger

Martin, being from California I don't worry about ice dams. But the lights I chose only consume about 7 watts. That's not much. It seems to me that we're quibbling here. As long as you still install them in cans rated for insulation contact, as they are meant to be installed, and if you install them correctly with at least a foot of insulation over the top of the can ... then no problem. In my first comment I meant cans when I said boxes, so that may be a source of confusion.

I think with new technologies comes new procedures for installing it. It can be disconcerting when an old taboo becomes something conditional. Sooo, I guess we'll have to agree to disagree on this.

Jul 1, 2014 9:50 AM ET

Response to Carl Seville (Comment #6)
by Martin Holladay

Carl (and Malcolm Taylor),
This photo of a beautiful roof was just posted on Twitter by Chris Laumer-Giddens.

If I were the engineer at the truss company who took the order for this monstrosity, I might just announce my retirement.


Complicated roof.jpg

Jul 1, 2014 11:44 AM ET

Is the alphabet long enough
by aj builder, Upstate NY Zone 6a

Is the alphabet long enough to name all the different truss sections?

Jul 2, 2014 5:03 PM ET

Air Ducts? and Plumbing??
by Billy Ware

If you are going to include air ducts as a detail that can be left out of a house, then you might want to consider indoor plumbing as well, at least on the supply side. It is often fraught with problems and leaks and can cause major, major issues when not installed correctly. You could argue that most plumbing systems are installed with at least a few potential problems. Most do not leak immediately, but there will be leaks in the future.

My real argument is that the quality of the installation is critical on both of these and few of the other items listed above. This goes for most everything, but in the case of no ducts, we typically have to tell the owners that they have to leave doors open to get proper ventilation and comfort in a house. This typically is just not acceptable considering the amount of money they would be ponying up for a mini-split kind of system and I am a fan of mini-splits.

Thank you for putting the article together, it is a good reminder about the challenges we face everyday.

Jul 2, 2014 7:15 PM ET

Response to Billy Ware
by Martin Holladay

You wrote, "If you are going to include air ducts as a detail that can be left out of a house, then you might want to consider indoor plumbing as well, at least on the supply side. It is often fraught with problems and leaks and can cause major, major issues when not installed correctly."

Your comparison between ducts and plumbing pipes is interesting. In fact, almost all new homes have leak-free plumbing systems, while almost all new homes have leaky ducts. If the average HVAC installer could achieve what plumbers routinely achieve -- leak-free systems that are tested under pressure for 24 hours -- we could all sing "Hallelujah" and retire.

Jul 3, 2014 4:57 AM ET

I expected more on windows!
by Nick Grant

Great list and discussion

Glass to floor (except for doors) is on my black list and makes me really popular with Architects in design meetings! Extra summer gain and winter heat loss without the benefit of daylight. Requires more overhang to achieve summer shading and that can cut out daylight. More difficult to install thermal bridge free. No wiggle room for aligning with floor finishes. Lost opportunity for a window seat or storage. Needs to be toughened or laminated glass. Gets splashed with dirt etc . . .

Double, folding and sliding doors all leaky and or expensive and hard to adjust and maintain draft free.

Re cantilevers, yes there are clever products that reduce the thermal bridge but it is still difficult and expensive and hard to detail to keep water out.

Also I heard of a (skilled) contractor who mixed the connectors up on site and used one designed for shear not bending on a cantilever balcony. When the props were removed the balcony fell off. Luckily no one hurt.

Agree cathedral ceilings can be be more expensive but they can be good value where wall or ridge height is limited and they feature in most projects I work on although I try to avoid SIPS!

Jul 3, 2014 5:05 AM ET

Edited Jul 3, 2014 6:29 AM ET.

Response to Nick Grant
by Martin Holladay

Thanks for your suggestions. I will add "glass that extends to the floor" and "sliding doors" to the list.

In response to the disappointment expressed in your title -- "I expected more on windows!" -- I suggest that you read my article, Study Shows That Expensive Windows Yield Meager Energy Returns.

Jul 3, 2014 10:41 AM ET

What is a cantilever?
by Carl Mezoff

To beat a dead horse (comment #12) a little more, in my experience a cantilever is generally defined by where the support is provided, and not by the particular hardware used to make the cantilever sturdy. If a beam or deck or even the garrison overhang you show in your lead photo is supported along only one edge, it is generally considered to be a "cantilevered element," irrespective of the makeup of the element, or whether in passing over the point of support there is a change in material.

But your list of troubled details wonderful food for thought! Thanks!

Register for a free account and join the conversation

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