Cost effective amount of exterior rigid foam insulation on Timberframe
alanferg
| Posted in Energy Efficiency and Durability on
Building a timberframe home in Northern Maine (climate zone 7) with 100% continuous rigid foam insulation (no batt/cavity insulation). A quick search reveals 2018 IECC code requires R-49 ceiling and R-20+5 or R-13+10 walls.
Assuming R-5/in for rigid foam, 10 inches of foam on the roof and 5 inches on the wall would meet code.
Roof: 6″ layer + 4″ layer, seams staggered and taped
Wall: 3″ layer + 2″ layer, seams staggered and taped
Is it cost effective to install additional foam or am I throwing money away?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Search and download construction details
Replies
Not sure how much reading time you have, but the following may be helpful.
https://www.greenbuildingadvisor.com/article/pretty-good-house-2-0
https://www.greenbuildingadvisor.com/article/how-much-insulation-is-enough
https://www.greenbuildingadvisor.com/article/how-much-insulation-is-too-much
https://www.greenbuildingadvisor.com/article/r-value-advice-from-building-science-corporation
https://www.greenbuildingadvisor.com/article/avoiding-the-global-warming-impact-of-insulation
Many here advocate for a 'Pretty Good House approach' which puts your wall-R closer to 40 and your roof R closer to 60.
If you wanted to be technical, and completely accurate, you would need to consider many inputs and crunch the data (i.e. what are you heating with, how warm do you plan to keep it, what solar gain do you expect, how air-tight will you build, etc).
The last article is food for thought regarding use of XPS foam (perhaps us EPS and/or Polyiso). Have you considered SIPS? For a timberframe with all exterior insulation, they seem like a decent path.
Tyler, thank you for the list of reads. I'm familiar with the 5-10-20-40-60 rule, and I wonder if those numbers are necessary with careful air sealing. Also, the global warming impact of rigid foam was more than I realized.
At this time I'm not considering SIPs mainly due to cost and what I consider to be less flexible.
Allan,
With continuous exterior insulation, you don't need R49, you can also meet code on U factor basis. I believe this is around R38 (~7" roofing polyiso).
Even in Zone 7, it is hard to make a financial case for insulating more than code with rigid insulation. For example in -10F weather on a 2000 sqft roof, you loose ~4kBTU through an R38 roof, and ~3kBTU for an R49 roof.
Generally, the best bang for your buck is making both your roof and walls the same effective R value, going overboard on either assembly on its own saves little energy.
It would probably be cheaper to strap out your roof deck with 11 7/8" TGI and go with 8"+3.5" of mineral wool batts (~R46 assembly assuming T&G ceiling over plywood). You can then cross strap this and install a standing seams metal roof without needing a 2nd roof deck.
"Generally, the best bang for your buck is making both your roof and walls the same effective R value"
?
Is the fairly pervasive recommendation for higher insulation values in roof assemblies explicitly not for financial reasons then?
My understanding is that the common recommendation for higher R-value in roofs is for a different financial reason:
When you are insulating a vented attic with loose fill insulation (cellulose or fiberglass), increasing from R-49 to R-60+ is relatively cheap, compared to a similar increase in wall R-value where you start having to deal with long fasteners through multiple inches of foam, etc.
If you're already wrapping your whole building in a thick layer of rigid foam (or using some other similarly complex method of insulation), Akos's suggestion to keep all your R-values similar to each other makes sense, and the usual wisdom of "dumping cellulose in the attic is cheap so you mind as well do it" doesn't apply.
Exactly.
In cold climate, when it comes to heat loss through insulation, the roof is just another surface, there is nothing special about it. If you are not using blown in insulation, it doesn't make any financial sense to insulate there more than anywhere else.
In sunny cooling dominated climates, it is a bit different. There are definite energy savings by insulating the roof more as the roof can get significantly warmer than ambient temperature.
>"In cold climate, when it comes to heat loss through insulation, the roof is just another surface, there is nothing special about it."
That's a common misconception, but it's not exactly so, and in meaningful ways.
The proximate air temperature and radiation temperature environment of walls are roughly the same. But the roof is subjected to radiating into deep space at temperatures well below the local air temperature, and the roof temperature at night is usually the dew point temperature of the proximate air, not the air temperature. Except during rime-icing fog conditions the dew point is below the air temp, and roof's average temperature over a winter will be well below the average outdoor air temperature.
It's a complex model, since during sunny days the roof temp can also experience temperatures above the air temperature, but there are more hours of clear night sky than sunshine in winter in most locations. The roof's solar reflective index would also have to be factored in, just as it's a factor for roof gains in cooling dominated climates.
But it's correct to go for IRC compliance on a U-factor basis rather than R-value when the only insulation is expensive continuous rigid foam. The IRC max is U0.026, which is the same as (1/U=) R38.5 whole-assembly, taking full credit for the R-value of the roof decking & ceiling gypsum , the interior & exterior (and intra-cavity) air films, etc. Most assemblies can squeak more than R3 out of the other layers, so R35-ish is all that's usually required of the foam layer.
But also, don't derate the foam for temperature- stick with the labeled R-value. For 2lb roofing polyiso that's about R5.7/inch, so it would only take about (R35/R5.7=) ~6", not ~7". The "payback" on the additional inch of foam is essentially "never", even if under peak load conditions the 6" is only delivering R30 rather than the labeled R35.
Similarly for walls, 3" of 1lb density foil faced polyiso would meet code performance on a U-factor basis (< U0.045, or R22 "whole wall") in most assemblies, with credit for the air films & wall sheathings, siding, etc. An empty studwall with half-inch OSB on one side and half inch sheet rock on the other is good for about R4 "whole wall".
Dana,
Is there a standard fudge factor that can be used for ballpark calculations for roof losses in colder climate?
Interesting. I guess I never gave this much thought, and took for granted the code and oft-cited recommendations.
It makes sense that one would want to pile on the R when its cheap (i.e. vented attic) and be more judicious when its not.
It's taking me a bit to recast my assumptions about the roof energy losses.
Convective heat rise and radiative roof cooling don't amount to an appreciable difference?
I notice that in the BSC article linked by Brendan their reccomended R-value for a 'compact' roof is lower than for attic (which makes sense) but still significantly higher than walls. Is the latter simply because even 'compact' roofs are assumed to allow cheaper R than walls?
If the answer is yes, than is the advice to shoot for the same wall and roof R sound, even with all exterior foam? What's typical of a PERSIST build?
Tyler,
There are two things, losses through air leaks and loss through insulation.
In a reasonably well sealed place with tall ceilings (12'-14'), you have around 3-4F difference between ground level and peak from stratification. That extra bit of temperature hardly budges the losses when it is 0F outside. At that temperature a slightly hotter ceiling looses maybe 5% more heat, which hardly justifies putting 2x insulation.
Air leaks ARE a big issues. This is why you see a lot of talk about air tight ceilings, stack effect can cause decent amount of air flow and a lot of heat loss. It also makes the place much more uncomfortable as air at ground will now be much colder and air by the ceiling much hotter. When you have a leaky ceiling, no matter how much R value you put in the attic, the house will still be cold.
This is why, it is always better to focus on air sealing the place.
In most colder climates, there is very little energy saved by going much above R25/R30. Also the law of diminishing returns also kicks in, going from R15 to R25 saves a fair bit of heat, going from R25 to R35 makes makes little difference.
An R35 wall is warmer, it might be worth it just form a human comfort perspective.
I hadn't considered making the R-value in the roof and wall the same. However, if I understand Dana correctly, it makes sense the roof and wall are subject to differing environmental factors (some of this info is a bit technical for me to wrap my head around!).
Since my original question, I found old, simple heat loss calculations I'd made on an earlier plan that never made it past the graph paper stage. It is helping me see where the cost of adding more insulation is not effective.
Also, I fully agree with careful and proper air sealing. I'd rather insulate to the code minimum and spend extra on preventing a leaky roof.
There are a lot of variables involved in the answer to your question of how much insulation will save enough energy to pay for itself.
The only way I know to have a real answer is to model your house in a computer program call BEopt watching the training videos and learning to use the program is a 40 hours commitment. The program is funded with your tax dollars so it will not cost you any money. https://beopt.nrel.gov/home
The program considers
1 your local weather
2 The size and shape of your house.
3 What direction you house faces.
4 The sizes of your window and door and what direction they face.
5 The amount of insulation on you wall and its cost per sqf.
6 Your heat source and current fuel costs.
7 The interest rate you will pay to finance the home and your prediction of future inflation of fuel costs.
Since no two homes will have the same answers to these questions there is no one correct R value.
The biggest factor is how long will you live in this house. Few people have a home built with a plan to sell it in a few years but life happens and plans change.
Walta
Interesting! Thank you for this resource... now to find forty hours.
Your reply reminded me of simple heat loss calculations I'd made a few years ago on a home I was hoping to build. The project never happened because I sold the land and decided to buy and build elsewhere. But I dug up the old file and found transmission and exfiltration heat loss formulas.
Transmission Heat Loss in BTUs/hour = Area ∙ U-factor ∙ Difference in Temperature
Q = A ∙ U ∙ ΔT
(Not sure where I found the formula and its validity or accuracy)
Outdoor Design Temperature: -13 °F
Indoor Design Temperature: 70 °F
Temperature Difference (ΔT): 83
Roof Area (A): 1,765 ft²
R-20 U-factor (U): 0.0500
R-40 U-factor (U): 0.0250
R-60 U-factor (U): 0.0170
R-80 U-factor (U): 0.0125
Transmission heat loss through roof:
R-5: 29,299 BTUs/hour
R-20: 7,324 BTUs/hour
R-40: 3,663 BTUs/hour
R-60: 2,490 BTUs/hour
R-80: 1,831 BTUs/hour
A quick glance tells me doubling an R-20 roof to R-40 cuts heat loss by half, whereas adding 50% insulation to an R-40 roof doesn't do much.
Although it's an old report, and of course pretty general (doesn't account for the particulars of your design, local construction costs, etc), there are some ballpark numbers for cost effectiveness by climate zone on page 10 of BSC's "BA-1005 High R Value Walls Case Study": https://www.buildingscience.com/file/5806
Whatever you do, for that much foam, look for reclaimed. You'll save many
thousands of dollars. Even if you have to pay for shipping.
Stephen, are savings generally the case? I was just reading about a builder looking for reclaimed foam and he found the costs to be the same as virgin foam. But thank you for the suggestion, it is worth looking into.
I can usually see reclaimed 2lb roofing polyiso at between 1/4 - 1/3 the price of virgin stock goods f.o.b. the distributor's lot. Factory seconds foil faced 1lb polyiso usually runs about half the cost of unblemished perfect product at distributors. YMMV.
Stephen, If I'm remembering correctly you're also in Maine.
Would you happen to have source(s) to share for EPS and/or reclaimed foam? I say eps because even that seems a bit hard to find (besides the foil faced low density stuff at Lowes etc)
Try Green Insulation Group in Worcester MA. Or Insulation Depot in Framingham MA, which is advertising on Craigslist Maine. It's selling a pallet of 2.5" polyiso (36 4x8 sheets) for $650. That's $18 a sheet, plus shipping.