How to Avoid the Flaw in Structural Insulated Panel Construction
Before 2001, the biggest construction project I had ever taken on was a bookcase. My most complicated project was probably the mahogany lamp table I built in tenth grade shop class. (I still have it!) Then I decided to build a house. After weighing all my options, I decided to build with structural insulated panels (SIPs). With help from a SIP consultant and an experienced builder, I got the house framed and then finished it out over the next 21 months.
What are SIPs?
Structural insulated panels are sandwiches of oriented strand board (OSB) on each side and expanded polystyrene (EPS) rigid foam in the middle. You can see the ones we built with laid out on the floor on our first day of erecting the SIP walls in the photo below. After getting the basement done and a framed floor on top of it, we built all of the above-grade walls and the roof with SIPs.
Different companies have different methods for connecting the panels. Some eliminate all thermal bridging where the panels join by using connecting pieces (called splines) with foam in the middle. The house I built used solid wood splines between each two panels.
The standard width for the panels is 4 feet, though, so there’s a lot less thermal bridging with “studs” 4 feet apart rather than 16 inches apart. Using structural insulated panels for the roof requires wood splines for structural support, but again, they’re 4 feet apart. You could always put exterior continuous insulation on the outside of the walls or roof to reduce the thermal bridging from the wood splines.
The advantages of SIPs
The advantages of SIPs are reduced thermal bridging and easier control of heat, air, and moisture. Solid insulation embedded in panels means that air sealing should be easier. I had never seen or done a blower door test before building with SIPs, and I measured an air leakage rate of 1.7 air changes per hour at 50 Pascals (ACH50) when I tested the house upon completion.
Controlling liquid water is no different than doing so for the other types of structures. You can use house wrap, fluid-applied membranes, felt, peel-and-stick membranes, or another type of liquid water control layer.
The panels themselves have a low vapor permeability so you must make sure that the walls can dry from the panel to the indoors and also from the panel to the outdoors. The interior OSB can dry only to the inside of the house, so don’t put plastic under the drywall. The exterior OSB can dry only to the outside, so it really should have a gap to aid drying (i.e., a rainscreen).
SIP homes are generally stronger than many stick-built homes. They have survived hurricanes in neighborhoods where all the other houses were destroyed.
The flaw to avoid
The big caution for building with SIPs is that you must make sure all the seams, joints, and penetrations are air-sealed to the hilt. This is especially true at the top of the house because the stack effect will put pressure on any weakness in the air barrier. Some SIP houses have required extensive repairs to the sheathing after only a few years because of air leakage.My friend John Semmelhack of The Comfort Squad in Charlottesville, Virginia built a SIP house in 2008. As his family grew, he added onto the house in 2015 and discovered that some of the OSB on the roof and the upper part of the walls was damaged. The problem was not the overall airtightness of the whole building enclosure. He was almost Passive House tight at a little over 0.6 ACH50.
The problem was the concentrated air leakage through seams at the top of the house because of the stack effect. In winter, that put humid air from inside the house in contact with cold surfaces, and you know what happens then: accidental dehumidification. You can see one of the moisture-damaged seams on his roof in the photo above.
How to keep your SIP house from rotting
Semmelhack repaired his roof, but what could you do differently to avoid this problem from the start? One simple change could make the difference between needing repairs in less than 10 years and a SIP house that lasts for decades. The basic principle is to keep humid air away from cool surfaces.
When John and I built our SIP houses, the air barrier was the whole panel. The weak part was at the connections between panels. Before we connected two panels, we sprayed can foam on both sides to stop air leakage. As Semmelhack found out, though, even a few small areas of air leakage can damage the enclosure.One step in solving the problem is to put a continuous air barrier on the outside of the panels. Peel-and-stick or fluid-applied membranes work well as long as they’re vapor permeable. But that’s not be enough. Warm, humid air inside the house can cause problems through convective looping in the seams between panels (diagram above). Thus, the other important step in ensuring your SIP house won’t rot is to seal the inside.
Some suitable products for this are:
- SIGA Wigluv 100 (tape)
- Prosoco FastFlash (liquid-applied)
- ZIP System Liquid Flash (liquid-applied)
If I had it to do over again, I’d still spray can foam into the connections between panels. But I’d also make sure each seam on the inside and the outside of every wall and roof section was sealed with a good liquid flashing membrane or high-quality air-sealing tape like the ones listed above. The key is finding a good one that works with OSB.
Since John’s SIP house had moisture problems and mine used the same methods and materials, you may be wondering how mine has fared. It’s been 21 years since we framed and sealed it after all. You might think that mine has fared better because Carrollton, Georgia (IECC climate zone 3) is a warmer climate than Charlottesville, Virginia (IECC climate zone 4).
Like you, though, I also wonder if the house has had those problems. I got divorced and moved out in 2006, so I don’t know what problems the new owner may have had with it. But I know what to tell him if he ever calls to ask for advice.
Allison A. Bailes III, PhD is a speaker, writer, building science consultant, and the founder of Energy Vanguard in Decatur, Georgia. He has a doctorate in physics and is the author of a popular book on building science. He also writes the Energy Vanguard Blog. You can follow him on Twitter at @EnergyVanguard.
Structural Insulated Panels — An Easier Way to Build an Airtight Home
Two Rules for Preventing Humidity Damage
Air Barriers, Vapor Barriers, and Drainage Planes Do Different Jobs
Comments are moderated. Your comment will not appear below until approved.
This Post Has 27 Comments
Nice article Allison! Yes – since panels take care of “most of it (structure and insulation)”, it all comes down to the connections of the panels and the seam details between panels. Our assembly manual requires the interior seams of roof panel joints to be either taped or caulked – and this is on top of the two rows of sealant required at every panel joint during construction. This would include a foamed seal at the exterior skin. And many of our clients use Huber ZIP Roof sheathing for our roof panels, and so they tape that as well – but I question your comment about the “peel and stick” membranes over a panel roof. The only reason I question it is because I am not aware of vapor permeable peel and stick membranes. I don’t want to sandwich an organic between two inorganic materials (our closed cell foam and the membrane). If you are aware of some vapor permeable peel and stick membranes I’d love to know about them. For the ones I’ve seen they have not shared the data, so I figured that they were not vapor permeable.
Self Adhered “synthetic” underlayment were marketed in the 80-90s before the introduction in Florida Codes in 2007, though made optional in 2008. These were “pool liners” or vapor impermeable asphalt adhesive underlayment. Some have transitioned into more sustainable, butyl adhesives. Though with the advancements in pressure sensitive adhesive (PSA) with acrylic formulations, some have now advanced using a breathable PSA as the adhesive.
We still see the mythical demand for a roofing pool liner, considered maximum protection for water ingress, until the roofing is nailed into or a pipe penetration leaks, trapping water inside the pool liner.
Breathable synthetic underlayment entered the market just after 2000, though typically as mechanically fastened. VaproShield developed and marketed one of the first breathable underlayment in 2008, followed by a highly permeable self adhering version in 2010. In 2019, VaproShield released its new generation of breathable SA underlayment. A very durable, an air barrier, UV stable, PSA, breathable underlayment called SlopeShield Plus. Due to its black color and UV stability, its perfect for retrofitting and the self drying roof designs. Allowing the sun to heat the covered materials, and drying them through the vapor permeable membrane.
It is currently the underlayment choice for replacement of the Texas State capitol copper roof retro. Due to vapor diffusive drying through the membrane, it is the better choices for retrofitting, to dry the existing roof substrate/insulation. An excellent choice for roofing underlayment where a drainable and vented cavity above the membrane is designed.
Seal tight, ventilate right and allow vapor diffusion to dry intestinal moisture.
Thanks very much for the pointer to SlopeShield Plus Scott! I was not aware of it.
We have Pro Clima Solitex Extasana Adhero here in NZ
I understand its an international brand. Is that helpful?
SIP’s are a great system, but not without issues. One of the problems I have with them is the amount of adhesives/resins embodied in them. EPS sandwiched between two sheets of OSB.
I’ll be very guilty of straying from the topic of moisture movement in SIP joints, but Allison did name the article “How to Avoid the Flaw…”. I am just pointing out a possible flaw, worthy of further investigation.
Yes, OSB dominates single-family framing these days. When you enter a house that was just framed, you can smell it. You will smell it for a while after the house was finished. SIP’s are asking me to smell the OSB for a bit longer, because there’s twice as much of it. It also off-gases somewhat slower, presumably, because it’s stuck to EPS on two of the four available sides (ignoring the edges).
APA (American Plywood Association, or now The Engineered Wood Association) says “Structural engineered wood products manufactured for construction applications, such as structural plywood, oriented strand board (OSB), wood I-joists, laminated veneer lumber, and glued-laminated timber, are exempt from the regulation.” They’re talking about the 2018 EPA formaldehyde emissions standards that were based on the California CARB phase 1 and 2 standards. Those standards limit formaldehyde, the chief troublemaker in plywood and pressed wood products such as MDF, particle board, etc, used in things like furniture and cabinets.
The above exemption worries me a bit. The intent of the regulation was to oversee products that go inside of living/working spaces. Well, all that separates us on the inside from the OSB sheathing on the exterior walls is a flimsy layer of insulation (less flimsy if it’s spray foam, or rigid foam board), and a very porous layer of sheetrock. Similarly, all these I-joists (OSB) are “venting” mostly to the indoors, on all sides.
Hi Paul – I would say look for the manufacturers – and they are out there – like Huber – that say “no added formaldehyde”. Wood has naturally occurring formaldehyde, as you may know, and so concentrated wood products can run afoul of the intent of the rule, hence the exception. It was after 2005 and the debacle with the FEMA trailers and interior toxicity that most of the industry “got religion” and started removing formaldehydes from their binders. And some have done a better job of it than others.
Charles, in ideal world, yes, we would trust the manufacturers. But just like with food labels, where a manufacturer is able to say “trans fat-free”, because there is less than 0.something grams of trans fats per serving; or plumbing fixtures that are allowed to say “lead-free”, because they contain less than…, you get the point. And that’s just the regulated stuff.
I don’t understand your “concentrated wood” explanation. I think the exception is because these are exterior-use products, for the most part.
Hi Paul – yes – entirely agreed – “truth in labeling” and things like that – and people play games with labeling. Sad but true. Regarding “concentrated wood” – the southern spruce-pine-fir lumber has a specific gravity of 0.36 according to “EngineersEdge” website. OSB has a specific gravity of 0.62 – so much denser product (almost twice as dense) – hence that is what I am referring to “concentrated wood”. OSB is about 10% denser than plywood.
Paul Christopher Timusk’s 2008 Dr. thesis is a great reference for OSB. He describes the density as “…density profile is the result of several factors, which include how the mat is pressed into a panel during manufacture (the pressing cycle), the press platen temperature, the resin types, the species of wood, and the moisture content distribution within the mat.”
Note that water ingress into OSB is much faster then its drying. Thus the importance of keeping it dry to begin with and when wet for long periods of time its reference as vertical mulch, when on a wall.
Free drainage and ventilated drying of that gap or a “ventilated rainscreen” design, for walls or slanted walls called roofs.
Scott: I wrote about Chris Timusk and his dissertation in this article from 2013:
Moisture and the Quirkiness of OSB
When I said I don’t understand “concentrated wood” explanation, I meant it in the context of the EPA formaldehyde exemption for OSB. MDF as well as some particle boards (used for interior applications) can be even denser than OSB. In other words, the EPA exemption is there not because OSB is dense, but because its primary use is not indoors.
Well, until we have a SIP panel with one sheet of OSB separated from indoors by a mere 1/2″ of porous sheetrock.
When I first began building hurricane/disaster resistant homes back in 2010, I considered the three best non-stick framed technologies available: SIPS, ICF and Superior Walls precast insulated concrete panels. All offer increased insulation capabilities and resilience to high wind storms. I eliminated SIPS because of its’ vulnerability to moisture from poor flashings and other risks as stated here already. I eliminated ICF because of the high site labor factor and complexities in assembly that can often lead to issues and failures. Also concerns with termites migrating up the exposed foam insulation. That left me with my selection of Superior Walls. Factory built with quality control at the fabrication point. Site labor by qualified technicians in just a few hours so my framers could start the same day as the walls were set. Lastly, virtually no water issues to contend with. Have never looked back since.
Most pre-cast concrete wall systems – including the brand you mention – are terrible at insulating efficiency. And to their credit they do not claim a high amount of insulation. I have heard that in climates in Pennsylvania and north you can find higher insulation levels in Superior Walls – but I know some of the worst thermal bridging images I have taken with my thermal camera involve their precast wall panels. I advise our clients that if you are fine with not having a footer to support your home (and not everyone is), then they are great at keeping earth at bay for below grade applications. But they are only providing minimal insulation and they have significant thermal bridging – so be sure to add your own insulation on top of them. Remember – they are not selling a great insulated wall – they are selling a pre-cast concrete wall that is both strong and durable (and quickly set, btw – they do an amazing job with that).
I agree from a Passive House perspective, that you do not get that level of insulation. However, after building and living in only SW homes, my personal experience is that the U value they provide helps make up for lower insulation R values. Air movement over-rides insulation every time. My electric bills are very modest as are those of my owners. The 6″ interior wall cavities provided between the insulated studs provides a perfect space for adding R19 batts for those who would like some additional insulation. Very inexpensive to add. I think it is definitely more of a concern in extreme Nothern climates where temps drop into the teens and stay. I helped build my neighbors home with 3800 sf conditioned and it sized out for less than 2 tons HVAC. Put in a 2 1/2 ton American Standard that works great. As for the footer question, I have built using traditional concrete footers as well as the crushed stone that is common. Just remember, whenever there are soil compaction issues for footers, crushed stone is the engineered solution.
Nice article, Allison! Well done!
It’s a toss up whether OSB or EIFs are the worst construction material of the past 50 years. OSB rots like it was designed to rot – all it needs to rot is a glimpse of water.
Several homes in my neighborhood were built in the 1990s of OSB covered with EIFs – my God, have they rotted! The only fix is to remove all the EIF and rotten OSB – a job that takes almost as long as building the house. They finished one the end of 2022 – used stucco to replace the EIF. Took them about four months with a crew working every day.
You have great topics Allison, I enjoy your blog!
Stack effect is a big problem in my market (climate zone 7) and the main reason I avoid projects involving SIP panels. I’ve been in timber framed houses where the panel seams line up on the timber framing, impossible to create a good air seal at the seams and have to rely on sealing the panel to the timber. This can be avoided by offsetting the seams from the timber, but problem persists where the timber blocks access at the wall to roof and roof peak intersect. My preference is continuous exterior insulation using a more vapor open product like Rockwool or one of the newer wood fiber products (haven’t had the opportunity to try rigid wood fiber insulation yet, hopefully soon). My next option would be a nail base product (I have a friend that calls it an IP) after the structural sheeting is installed and sealed. These are all more labor intensive, but in my opinion, result in less risk. In the end, durability of the structure should trump ease of construction.
Great article Allison! Love the blog.
You mentioned adding exterior insulation on the outside to eliminate the thermal bridging of the splines. Would that eliminate the osb during the the outside?
Craig: I don’t understand what you’re asking. Are you asking if putting continuous exterior insulation on would eliminate the need for the exterior OSB on the panels?
Sorry, I miss typed that on my phone.
I was wondering if adding exterior insulation would trap moisture between the exterior foam and the structural panel?
There cannot be a convective air cycle within the SIP, as indicated in your diagram (from Lstiburek, Building Science Corp.) The polystyrene prevents conduction but also prevents convection (air movement). The actual moisture impact on the OSB is due to having an infinite source of moisture to the ‘exterior’ side of each slab of OSB – the interior side being sandwiched against the foam. The OSB is much more hygroscopic than the foam so any moisture absorbed from the air can only desorb from the OSB when there is a thermodynamic gradient that favors diffusion and evaporation.
I think the hygroscopic effects can be countered by one of two methods (or both in a carefully engineered design?) One can construct SIPs with a quarter inch vapor gap between the OSB and the foam layer with a pathway to remove accumulating moisture. The second method would be to use an insulating layer that has a significant hygrothermal buffer capacity, so that moisture doesn’t get trapped at the structure boundary but it is distributed within the OSB and insulation (similar to straw bale wall systems).
Todd: Yes, there certainly can be a convective loop. It’s not in the field of the panel, though. It’s at the seams between panels.
Allison, I do not have experience with SIPs, how big of a gap are we talking about? Per Joe Lstiburek, an air gap larger than 1/2″ will have convective loop losses.
SIP panels made with plywood do exist. I would prefer plywood over OSB. I would prefer an integrated system such as Zip or ForceField made with plywood. Why we don’t see them? I suspect the reasons are weaker tolerances on plywood, dimensional stability issues (some warping, expansion/shrinkage), and getting the weather barrier coating to stick to plywood with consistency. For SIPs made with plywood, I suspect there might be a bit more joint movement issues, and getting the interior joint tape to stick with consistency.
OSB has it’s own issues. Once wetted, it swells in all dimensions and does not fully return to its original specs (including thickness). It also does not fully return to its original structural properties. It also changes its moisture content profile after one wetting. Keeping it dry is important.
Resubmitted without less than symbols that interfere with http
Thanks for this article!
I am currently looking to build in the PNW and considering SIPS. Perhaps my anecdotal evidence will be helpful to some. I am not a stranger to SIPS as I previously used them back in the early 80’s when I read an article in Fine Homebuilding magazine on how to make your own “stress skin panels”. The article showed a 4’x8’ piece of 8 inch thick EPS, osb glued on one side, and drywall on the other side. The OSB and drywall were offset so the panels would interconnect. I believe some people were using OSB on both sides and later screwing drywall on it so the drywall was replaceable. That made no sense to me. I used several different configurations for different locations OSB/EPS/drywall, straight EPS, drywall/EPS/EIFS, OSB/EPS, etc. The EIFS worked fine where it was protected by overhangs, but not where rain water could freely flow over the surface (another story). I basically encased the house in foam roof/walls etc I even used my shop vac to “dig” a 4 inch space around the perimeter to encase the foundation. The house upgrade saved me probably a thousand per year times 40 years on heat bills.
So back then, the suggestion was to use canned foam IE “Great Stuff” to join things together. In areas that I could later access, I found over the years that the canned foam did not permanently fill the voids. Everything expands and contracts and so did the EPS. What started out as less than1/16 inch eventually became ¼ -3/8 inch and the canned foam pieces were loose in between. I am wondering if 40 years later the problem has been successfully dealt with. Canned foam is usually about like mixing 25% open cell with 75% closed cell foam. Some vapor transmission but different than that of EPS. Different expansion/contraction rate as well.
The reason I decided to again look at SIPS is that now for conventional 2’x4’ or 2’x6’ framing many experts recommend 2 layers of exterior insulation (like for instance polyiso 2 x 2” ~ R-12) to keep the dew point buried in the foam. Then a rain screen outside of that prior to the actual siding material. So it looks to me a bit like site-built SIPS. So why not just use SIPS to begin with? So how have they addressed the main problems with SIPS? Some of these are the same problems we have in conventional framing with exulation?
• Expanding/contracting EPS
• Expanding/contracting OSB
• Preventing Juneau Alaska problem at the ridge
• moisture intrusion and lack of drying
• OSB lack of being fully waterproof tendency to curl or warp around its edges
• OSB outgassing to interior – urea formaldehyde, phenol formaldehyde, Methylene diphenyl diisocyanate, acetaldehyde, acetic acid and other VOCs
So whatever we do, first and foremost the problem is moisture. Beyond vertical precipitation there is 1. wind driven moisture. 2. solar driven moisture, 3. occupancy generated moisture. So chances are pretty good moisture is going to enter any cavity. All roofs eventually leak, with the possible exception of the Great Pyramid of Gisa :-). So there needs to be a drying potential built in to walls, roofs, and floors. If we block vapor diffusion we also preventing drying.
Another anecdote worth mentioning. I once had a Hot Tub, more properly called a spa. It came with a thick 5” vinyl cover filled with large pieces of EPS. After a few years the cover seemed to be getting heavier, or was I getting weaker? I obtained a new cover and indeed it was ½ the weight. I unzipped the vinyl on the old cover and pulled out the EPS and it was indeed water soaked. Apparently years of steamy moisture below perhaps entering thru the zipper, and no escape due to being covered in an impermeable layer of vinyl caused the EPS to become water logged. The lesson is that there needs to be a way for vapor to escape.
So now SIPS manufacturer are using a mastic called “sip sealer” between the panels. It apparently hardens as they say to hurry to join panels before it cures. Not sure why they can’t use something like acoustic sealant that always stays pliable? There will continue to be cycles of EPS contraction and expansion. Wondering if there is any anecdotal evidence of what happens to these panels with sip sealer after a few years of EPS shrinkage.
I think manufacturers have solved the problem of OSB expansion/contraction by the use use splines in the joint along with “Sip seal” to counter moisture intrusion thru the external joint. The strips can also withstand the movement without apparent buckling.
It seems that SIPS installers now leave a purposeful gap at the ridge of the roof that they later fill with expanding foam. Below that they place an 18” wide impermeable tape on the ridge beam removing backing on the outer 3” on each side, thus blocking a majority of vapor drive from the habitable space. I have not seen many details preventing water intrusion from the top however. One method from 2008 Listiburek shows and aluminum cap immediately over the sips ridge. (page 125 BS2008 SIPS) Personally I think Siga Wigluv tape should be used over the top of the expanding foam ridge fill as it will allow vapor to release but still repel rain water, but I have not seen that detail anywhere. The foam fill that they use will have a different permeance and expansion rate than the EPS, or the OSB which I am a bit skeptical about after seasons of use. I believe the best practice for a roof would be to add a second layer of osb or perhaps taped Huber Zip over vertically oriented furring strips over a vapor open underlayment attached to the SIPS. ( SIPS/OSB-vapor open underlayment-furring/gap-Huber Zip-vapor closed underlayment like sharksin-metal roof.) This would be preferable to just the easier/cheaper furring strips on a 45 over the Sips and then a metal roof. Metal roofing manufacturers seem to discourage use of vapor open underlayment as it causes corrosion under the panels. The rainscreen gap between the SIPS osb and the underside of the Huber Zip could dry both sides, especially when summer roof temperatures hit their highest.
OSB was originally intended for exterior applications. As such it was not regulated for interior use. It appears that there has been an industry trend to decrease use of the more dangerous urea-formaldehyde (UF) and greater use of less toxic phenol-formaldehyde (PF) There will always be some chemical offgassing of any products. Oak wood itself emits 0.009 parts per million (ppm) of formaldehyde. Plywood and OSB both off-gas formaldehyde. It is expected that 4 months time removes the greatest amount of dangerous chemicals. But in 2019 the U.S. Environmental Protection Agency’s final amended formaldehyde rules state Structural plywood, oriented strand board (OSB) and other structural engineered wood products remain exempt from the EPA TSCA Title VI rules on formaldehyde emissions from composite wood products. There are some new-generation OSB panels that use isocyanate resins. Isocyanate-resin panels do not contain formaldehyde and are considered non-volatile when cured; they offgas less than PF panels. They also use edge sealers to prevent water infiltration at the side of the panel. It is difficult to find data on the actual offgassing of OSB and it is seldom mentioned by SIPS manufacturers. For more info on chemicals see Corinne Segura – My chemical free house.
Hope my anecdotes are helpful to someone.
Would installing an erv/hrv and passive vents to the attic work to prevent the roof rot you describe?
Installing an ERV is always a good idea in tight houses. More importantly putting in a dehumidifier near the peak is pretty vital in keeping the humidity in an acceptable level. Even a cheap dehumidifier will do the trick, just add some condensate drainage. A good idea would be to add some additional monitoring as a belt and suspender approach to confirm the levels. The solution for repairing the often mentioned Alaska SIPS failures was to cut in some “vapor diffusion ports” in the top foot of the peak thru the outer OSB layer, and covering them with a high permeance layer like Delta Foxx, and then a space for the roof cap. This is recommended for new construction in zones 1,2,3 only. A near example is demonstrated by youtuber Matt Risinger in his own home in his interview with Joe Lstiburek. Matt vented his 2 layers of polyiso overroof with ~1″ slots. The recommended minimum size of the vents is an area of 1/300 the area of the roof. So in normal ridge cases assuming the vent is along the entire ridge, 1/300 of the rafter length. So a 12 ‘ rafter length would require 1/300 *12 or about a half inch.