Sistering roof rafters to increase space for insulation

This is related to another thread, but is also it’s own topic. I’m currently renovating an unfinished attic or a front-gabled house.

The roof rafters are currently about 20" apart, made of 2X6’s (the house was built in 1915). I’ve been talking with the GC about approaches to try and increase the depth of the roof rafters to get in more insulation, while at the same time allowing a baffle for air flow (I have a slate roof).

One approach we’ve discussed is sistering each 2X6 with another 2X8, and then firring that out with a 2X3 to give enough depth to get R38 in there (you need 5.5" for the insulation, plus another 1-2" for air flow).

Another approach we’ve discussed is sistering the 2X8 “off” the roof, meaning the 2X8 is not overlapping the 2X6 for all 6", but maybe only for 4" or so. In this way, you eliminate the need for the 2X3. But, it seems you also reduce the strength of the sister, or is it not really necessary anyway?

You may want to consider applying a rigid polyisocyanurate insulation (typically R7 to 8 per inch) on the bottom of the rafters after installing baffles, continuous from eave to peak, and batt insulation between the rafters.

It works well to increase the R value and furring strips and then drywall can be applied. It reduces the ceiling height some but the added insulation value is worth it, IMO.

The thing that is nice about this method is the thermal break, the rafters don’t go all the way through.


Nathan, Larry is giving you good advice and there are different alternatives and solutions, and this would be an alternative;

Because warm air naturally rises, the attic or roof area of your home is your first priority for insulating. Insulation reduces the upward flow of heat, keeping it inside your home longer. That means you’ll stay warmer, and your heating system will not come on as often—reducing your utility costs! In Main’s cold climate, insulating existing attics to an R-value of at least .38 is recommended. R-value is the measure of an insulation material’s ability to resist heat flow. It’s measured per inch of material. For example, glass fiber batt or blanket insulation has an R-value of around 3.2 per inch, and the R-value of loose-fill cellulose is about 3.7 per inch. Both of these insulation types are commonly used to insulate attics. Twelve inches of the glass fiber batt insulation achieves R-38, and about 10 1/2 inches of cellulose will do the trick. How much insulation is in your attic? In this case, how much do you need?

As Larry has suggested would give you an R-factor of 19+ 2" of thermax insulation which is 14.4 would equal 33.4. The thermax insulation would have to be foiled faced.

Another alternative would be to just spray the whole underside of the roof with SPF. This polyurethan spray insulation closed cell would provide you with a 8 R-factor per inch. So again r-19 + 21 @3" = 40 Just add another inch and your at R-48 and what is nice is you do not need to add a vapor barrier. Using the open cell Icynene spray, you would get a lot less R-factor and would need the vapor barrier as well.

It is kind of you get what you pay for.

Hope this helps.

Marcel :slight_smile: :slight_smile:

There seem to be pretty divergent opinions on the issue, but the weak consensus from what I’ve read seems to be that one shouldn’t spray foam insulation directly to the underside of the roof - but that there should be an air gap. So, without sistering or firring out the rafters, I’m looking at only about 4.5 - 5" of depth, which is not much insulation whether you use batt or icynene.

Owens corning has a higher density fiberglass insulation for cathedral ceilings that achieves R38 with only 8.5" or so, I think. That seems like a potential option. Icynene is great for the airtightness, but I’m wary of the upfront expense.

Remember that insulation is affected by the “Law of Diminishing Returns”. The first R’s save you a lot more than the later ones. It’s no sense getting to R40 if the payback for the last R10 is 40 years away., especially if it’s a more expensive than the earlier R’s. Take that money and put it where it gives you a better return.

Going from R1 to R2= 50% reduction in heat flow
Going from R39 to R40= 0.06% reduction in heat flow (not even near another 1% more savings for this last increase of R1!!!

If you can get the R’s in cheaply using lower cost insulation like blown fiberglass or cellulose in an open attic blow, with today’s fuel prices, R45-50 may make sense! All the insulator has to do is hang around a little longer, blow a few more bags… no extra framing, labour and expensive insulation costs…just more of the same. He’s already at the site and set up, so the extra R10 is pure insulation/labour cost. This extra R value may pay for itself in as little as 5-6-7 years (depending on fuel types/costs) but as fuel prices go up in the future, the payback time decreases.

Which is one of the reasons there is so much ice damming on many of these older houses where the attic has been converted.

Why not simply sister the rafters with 2x12s? And please do NOT install those styrofoam baffles all the way up!!! I’ve seen that done, and the result was enough of a partial second vapor retarder that moisture condensed on the baffles and drenched the insulation almost permanently. Use the baffles only to maintain the air space where the roof and wall meet, and no further.

They have cardboard baffles as well, right? If we didn’t use something to hold off the insulation, I’d be afraid the installers might push it right up against the roof.

The GC seemed very hesitant about sistering the existing rafters with 2X10’s, let alone 2X12’s. He felt it would be too unweildly. It would be a lot of wood, too, I imagine.

I also thought about only sistering the portion that needs the insulation the most, which is the section that is not behind the kneewall, and below the ceiling joist.

For the first part of the roof, it is behind the kneewall, so insulation isn’t as important. For the top part of the roof, it is above the ceiling joists (and the insulation), so also not as critical. However, it’s the middle section that’s key.

How is a 2x12 any more wood than a 2x8 plus a 2x3? Same wood, give or take an inch, and half the labor. And only use baffles where they are necessary, so the insulation can vent. Trust the installers to get it right…after all, with the nailing flange on the insulation face, how difficult is it?

Why isn’t this conversation taking place with the GC and an engineer?

My thoughts exactly.

Hey!! Architects are people TOO!

Dang dude didn’t mean to snub the “line jockeys” :wink:

I’ll make it up to you next time now that I know.:shock: :mrgreen:

You have given out too much Reputation in the last 24 hours, try again later.

I don’t know that they’ll be able to use the nailing flange on the batts, since the width isn’t going to be standard (the rafters aren’t 16" on center). I’ve seen plenty of fiberglass insulation with all the R-value squished out of it, seems to be a common mistake.

The idea against the 2X12’s would be that it wouldn’t be necessary the whole run of the rafter (since behind the kneewall insulation isn’t as critical). But, it’s a good question. I assume the 2X12’s need to run the full length as one board, sistered to the rafter, to add structural integrity?


One of the worst items handled in an attic renovation/conversion is the air leakage at the attic kneewall down through to the ceiling below. This is usually never properly airsealed and the outer attic is usually warm. This system can be made worse by adding gable end vents for the outer attic as they allow more air movement from the attic floor system and house below.

Yes I know that Brian. I didn’t say the low insulation level in the cathedral was the only reason, but one of the reasons. Both convection and conduction heat loss contribute to the problem. You must address both to have a well performing assembly.

An update on the situation for those who are interested. After doing some research, my plan was to use baffles to allow for ventilation from the soffit vents up to the top of the roof and put in some gable vents. But, after going over the roof with some contractors over the weekend, there were some new learnings.

  1. The house has no soffit vents. Actually, none of the houses on my street have soffit vents, either (built in the early 1900’s when houses were drafty)

  2. There seemed to be a consensus opinion that gable vents are not worth anything at all. They don’t ventilate, and are more likely to allow snow or driving rain into the house than air to escape (well, I exaggerate a bit).

  3. The only good approach for ventilation would be to put in a ridge vent, unfortunately, we have a slate roof, which would make the ridge vent tricky and expensive.

So, the question then became, why worry about ventilation at all? The contractors suggested using foam insulation, directly to the underside of the roof. They said they’ve been doing it in all of their similar projects. It creates a real thermal and air barrier, and you don’t have to worry about ventilation.

It’s a dramatic step, and I’ve read a lot of research on the internet about it. However, has anyone seen any practical experience where this approach has led to problems? One question that no one seems to know the answer to is what happens if the roof leaks? Where does the water go? Do you even know it is leaking?

Since this product is closed cell, it will not allow water through and any leaks in the roof would take a long time to show up, so therefore, needs to be checked on a regular basis.

Unventilated space for attics is growing more popular all the time.

Just my opinion.

Marcel :slight_smile: :smiley:

Nathan, I agree with Marcel.

I have renovated many vintage home and I suspect the reason for no soffit vent is due to timber framing which will not allow for any venting. If this is the case then more than likely the ridge pole would not allow it either.

With that said your alternative are, gable vents or spray foam, the latter is probably the best choice.

If this is the same project you asked advise for before you really should seek advise from a design professional. This way everything is documented and done right will add value to your home, not take it away.

Nathan, the contractors do not really care about your house after a few years. They want to sell you a product/service and be on their way. Sad but true. Ventilation IS necessary, albeit maybe less so if you have a slate roof that “breathes” towards the outside (compared to torch down and other “sealed types”).
If I were in your shoes, I would create soffit vents somehow, make sure you get 1"-1.15" of space between underside of deck and insulation (regardless of what kind). In order to increase the “depths”, furr out existing 2x6" with a 2x2" or a 2x4" screwed and glued in place. It both allows for added space as well as stronger frame members. But, before doing this, allow the lumber to dry inside your house so it reaches a humidity level that is comparable to what is in the rest of the frame lumber.

If you plan on using fiberglass batts, make sure the ventilation baffles does allow the insulation to “breathe” into the moving air. Stay away from “Durovent” or other type of manufactured baffles out of styrfoam. Water vapor will only condensate on these baffles and eventually soaking the insulation. Custom made baffles out of “peg board” or some type of “mesh”/netting would be better. Only purpose is to hold insulation away from roof and maintain the needed air space.

Needless to sya, do not skip on the vapor barrier. Do NOT trust the paper on the insulation. Put on a proper PVC or similar type barrier with good overlaps at the seams.

The more I learn about American roofing systems, the more I am in favor for clay or concrete tiles that allow the roof to properly breathe. In (native) Sweden, there are very few problems with these roofs, in rain, or in snow, and they last 50+ years. Where I can see them less suitable are in eartquake prone areas since these roofs are very heavy (but still pretty common in CA).