Insulating a Low-Slope Roof
I’m the developer/owner of a new construction 12 unit multi family building in Minneapolis, Climate Zone 6. We’re shooting for R49-55 using rigid insulation on a low slope roof above the wood web trusses.
My background is in architecture and construction, so I’m in the middle of the design team in making decisions and weighing input from both contractors, designers and energy modelers. I’m following the ‘pretty good house’ / stealth passive house approach where I’m not pursuing any one rating system or standard. Instead, taking the best parts for all, noting that because I’m building a multi-family building the size of an extra large size house conventional green wisdom has so far been often off when compared to either SFH and 50+ multi family building wisdom. It’s a tweener problem for sure. However, this insulation issue seems pretty straightforward.
I’ve read through all of Martin’s suggested reading on rigid insulation. Because a lot of the discussion on this board is about single family houses, I found this article on multi-family design in the northeast very helpful. I noticed though that polyiso was used on all the designs.
Designing a Multifamily Building That Approaches Passive House Performance
Given the articles about thermal drift with XPS/Polyiso and the polyiso drop in R value at lower temps, I’ve largely been considering EPS. My energy modeler is an EPS fan and modeled 13″ of it for an R value of around 54 depending on your decimal place choice of per inch rating. We can go as low as 49 or a bit more as was done in the model. We WUFI modeled more than R55 and it really didn’t move the meter performance wise. This confirms, in my opinion, the thoughts shared in the linked article: Air tightness in MF structure is outweighing sheer insulation value. In short, we don’t need to over do it on the roof.
With the new Owen Corning FOAMULAR NGX XPS out (waiting to see if local lumber yards can actually get a hold of it) that would be a bit better alternative to lowering the depth, with the sticker amount reaching R50 for 10″. Even if you consider a discounted long term insulation rate of 4.5 R per inch, that still keeps us close to where we need to be.
The biggest push back of using these materials vs Polyiso (only need 7-9″ of that depending on how you want to rate the per inch R value in cold temps) is the longer anchor bolts needed to secure the insulation on the roof. I’m trying to get some clarity from the actual installer guys on how much a 7-8″ fastener costs vs say 12″ or 14″ one, as my roof is less than 3000 sf and so the added cost of just the fasteners can’t really be that much even if you are anchoring every 12-16″. My hunch is that the more expensive (and thus less tall stack of) foam will easily equal/outweigh extra metal on just a materials cost basis.
I’m looking for some advice on
A) Is the fastener length issue really a problem, or just roofers grumbling about how they have to do something slightly different and really pay attention when drilling so they hit the truss?
B) Anyone gotten a hold of the FOAMULAR NGX XPS this building season? And if I can get it, is that lesser GW gas a good tradeoff for cutting the depth by 3″ and feeling good about the world at the end the day. I’m okay with the upcharge (finding out how much soon).
C) Is the EPS, due to it being cheaper, just the way to go, since once you drill a hole 8-10″, what’s really another inch or two?
D) Have folks actually been installing GPS now since it has been out for a few years and anything useful to note about price or installation with that? The 0.3-0.5 boost per inch isn’t really going to save much depth compared to how much we need.
I’m awaiting the latest per sheet costs from our supplier for each product to put a full cost analysis together.
Look forward to any in the field thoughts from folks who have weighed and tested on this material choice issue for their projects on low slope installations.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
One follow up: With guidance from this article, I have considered some combination of batt or cellulose in the roof truss depth (which could be as much as 12-14") to also bring down the amount I need on the roof deck.
https://www.greenbuildingadvisor.com/article/combining-exterior-rigid-foam-with-fluffy-insulation
So open to the hybrid idea that wouldn't require all the insulation on the deck itself. It was just the easy way and less messy way long term if you need to open up the ceiling at any point.
Just one thing to suggest: you might be able to get NGX through home depot if you can't get it through local distributors. But I'm guessing that it will be hard to justify why you'd choose NGX. If EPS thickness is too hard, GPS probably cuts the thickness enough at a lower cost than NGX.
With all that wood, foam and fastener concerns, maybe just use a SIP?
Numbers I've seen indicate that GPS is +1 R over EPS.
> batt or cellulose in the roof truss depth
Do not leave a gap between this insulation and the foam.
In this type of building there is very little energy saved between an R30 and an R50 roof, going overboard with expensive rigid insulation is generally not worth it as there is no ROI based on energy savings (most likely neither on for CO2 emissions).
When you have expensive insulation, the best is to insulate to code min, with continuous you can also go for compliance based on U factor, which might mean only needing around R35 to R38 of rigid.
The temperature issues with polyiso are a bit overblown. Some newer formulations elimiated this effect. Some formulations that do loose effective R value at colder temperatures but this is only the first couple of inches of insulation, overall even when this is taken into account it is still thinner than equivalent EPS.
Since you have trusses for the roof, I'm wondering why not fill the trusses with loose fill cellulose and vent the roof? This is significantly cheaper, lower carbon footprint and you can get R60+ for very little extra cost. You do have to be careful with venting a flat roof as it needs much more than the typical 1.5" gap, but it can be made to work especially if you are careful with air sealing the ceiling bellow.