SIP vs ICF
I know this is a topic that has been talked about a lot, but I do have a question we are mulling over in designing our house.
I have been pricing materials and it seems that both ICF and SIPS are going to cost relatively the same. (We are doing labor ourselves.) With the ICF we were going to get continuous R value of 24 and no air leaking or thermal bridges. We were going to bring this up to the roof where we would use SIPs that are 6 1/2″ for an R40. We were then going to either dense pack below the sips or add another couple inches on top to try and bring it to around R60.
The other option is to do sip walls that are 6 1/2″ thick and get an R40, but then there will be thermal bridging where the 2nd story floors sit on top of the Panels (rim joists).
What would be better for the same cost? High R value with more chance of thermal bridging or lower R value with no thermal bridging? Is there another way to build (without building an interior wall) to eliminate the thermal bridging with SIPs?
Thanks!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
The difference between R26 and R40 is pretty big. I think that will outweigh the thermal bridging. Suppose the thermal bridging is really bad and you get R15 for one foot of a 17 foot high, R40 wall. The net effective Rvalue for the whole wall is then 17/(1/40*16 + 1/15*1) = R36.4. I don't think that the thermal bridging will be that bad, but even if it was, you'd still be way ahead with the SIPs.
What's your climate? In a very few, the thermal mass in the ICF wall might be helpful. But more likely it's not.
R40 for a 6.5" thickness sounds pretty high. What kind of insulation are they using in the SIPs? I think of EPS as the normal insulation material for both SIPs and ICFs, but it sounds like it might be something else, in which case there might be other considerations regarding the different insulation materials and blowing agents used.
Jordan,
Q. "With SIP walls that are 6 1/2" thick, we get an R-40, but then there will be thermal bridging where the 2nd story floors sit on top of the panels (rim joists)."
A. They make joist hangers to address your concern.
.
Charlie,
I am located in Kansas City, MO. We are Climate 4a and not that far from 5a, but perhaps global warming will change things. We have a warm humid climate in the summer and it can get down to zero in the winter. The community I live in has a high tolerance for temperature variations. We keep the house around 82 in the summer and 62 in the winter. We are hoping to heat with just a wood stove and dehumidify (cool) with a ductless mini-split system.
The SIPs I found are from Thermocoremo. From their website they state:
"While some SIPs manufacturers use expanded polystyrene (EPS) insulation with an R-value of 3.5 to 4.0 per inch, Thermocore of Missouri’s structural insulated panels feature polyurethane foam, with an R-value of 6.0 to 7.0 per inch."
So they say that 6 1/2" gets R40. The nice thing is they are made not far from us. Are there questions I should be asking?
Martin, I appreciate these drawings. I did come across them in my google searches! Just for my clarification, I am rather new at this...is the Cap plate not a thermal bridge? Is there essentially 3 2x4s stacked on top of each other with foam tape and caulk on all sides? Those hangers are pretty cool.
Thanks for your help.
it can be designed with a one piece solid wall say 17-18 foot high.
I'm having a slot cut at the factory for a 2 by 10 to slip in on site then use typical joist hangers on the interior wall at the second story height.
no thermal bridge that way
62 to 82 is a pretty high tolerance! Given that tolerance and your climate, the thermal mass of the ICF won't buy you much.
Polyurethane foam has a few disadvantages compared to EPS:
1) The gas used to blow the bubbles in it is an extremely potent greenhouse gas, 1400X more potent than CO2. Just this year, alternatives with reasonable global warming potential have become available. You might ask the company if they offer that as an option, and if not, encourage them to consider it.
2) The high R-value is a result of the special gas inside the bubbles. But over time (years) that diffuses out, and gets replaced by air, reducing the R-value advantage. For a very thick panel, that will happen very slowly, but it will still happen. One rule of thumb is that you can't count on the long-term value being more than R5.5/inch. I'm not sure how thick the OSB is but if the actual insulation is 5" thick that would leave you with an R-value of 27.5, which is still better than the ICF. But it's not the R40 you paid for.
Dana is likely to post a more detailed answer, but you could also read what he wrote about it in comment number 3 on this thread:
https://www.greenbuildingadvisor.com/community/forum/green-products-and-materials/34547/thermocore-roof-sips-8-panels
I'd be inclined to go with the SIPS if you can get it done with the low-global-warming blowing agent, because you'll save some heating and cooling cost over the years it takes for it to settle to a similar R-value as the ICF. But you could also argue that the ICF gives you a more durable structure with fewer potential failure points and fewer ways to have hidden flaws in the implementation.
As for your question to Martin, yes, the cap plate is a thermal bridge, but the area is really small, so the heat loss is small. It might bring the net R-value of the whole wall down from R27.5 to R26.5, it's not a big difference.
Given the area is prone to termites and tornados internal ICF would be a great choice. Hate to have SIP's eaten by termites. When the wind is blowing an internal foam ICF is a lot safer. The climate your in it would be hard to justify a R40 SIP when a R20 will do just fine given your climate area.
Concrete is concrete. It's a time tested product that has stood the test of time for thousands of years. Each method will have its pros and cons but I believe ICF with a 6" concrete core is a bullet-proof house wall design (literally). You can use SIPs for the roof if you would like. ICF walls and a SIP roof.
A SIP roof is an expensive way to get R60, vs. cellulose on an attic floor. I would only use it if you are determined to have cathedral ceilings and a "scissors truss" design doesn't provide enough of a ceiling peak.
An R40 SIP has a closed cell polyurethane core blown with HFC245fa that has a global warming potential of nearly 1000x CO2. An equivalent 10-11" EPS core R40 SIP is blown with pentane, with a GWP of only ~7x CO2, and is a far greener product. Unless Thermocore will stipulate that they use an HFO124_ _ variant blowing agents, a 6" pu-core SIP is the opposite of "green".
Thermocore's hype completely understates the R-value of EPS. Any EPS that's only R3.5/inch would of such low density that it would be too weak to be structural. Most EPS SIPs are made with 1.5lb nominal density "Type-II" EPS that runs about R4.2/inch @ 75F average temp through the foam, or ~R4.5/inch at 40F average temp through the foam. At your average winter temp in K.C. you can expect mid-winter performance of about R4.3-R4.4/inch out of an EPS SIP.
In US climate zone 4A an R40 wall is complete overkill- you can hit PassiveHouse energy use levels in zone 4A with R30 walls. If you want a high performance house, design it. Don't get distracted by excessive wall-R. Anything over R25 may not necessarily be financially rationale over the complete lifecycle of the house. Download BA-1005 from this page, and take peek at Table 2, p10:
http://buildingscience.com/documents/bareports/ba-1005-building-america-high-r-value-high-performance-residential-buildings-all-climate-zones/view
ICFs are also usually Type-II EPS blown with pentane, with a GWP of only about ~7x CO2.
In a US Climate zone 4 A location it's possible to design & build Net Zero Energy houses with minimalist R21 ICFs (2.5" + 2.5" EPS). Unlike SIPs, it has no thermally bridging elements, and in a K.C. climate the thermal mass actually counts for something. With a lot of caveats about site conditions that will affect performance, an R21 ICF will deliver average thermal performance equivalent to low-mass walls in the mid to high R20s. (That's why even continuous R8 on the exterior of a concrete or CMU would meet code min in your climate: publicecodes.cyberregs.com/icod/irc/2012/icod_irc_2012_11_sec002.htm ) That performance is good enough to heat & cool with mini-splits in K.C.s climate.
In tornado alley an ICF is a much safer house to live in.
For equal money, forget the SIPs, build with ICF. A 2.5" + 2.5" EPS / R21 ICF is more than adequate for your thermal performance goals.
Like SIP vendors ICF vendors will also overstate the thermal performance in their hype, but even fairly simple modeling tools like DOE-2 , that underlies the slick BeOpt tool ( both "free" downloads courtesy of the US taxpayers funding the D.O.E.) model the mass effects well enough to be predictive of both peak load & energy use.