Proportion insulation roof vs wall

[Mikael_L]

I suspect that it's just some kind of tradition that leads to newly built houses having 250mm insulation in the walls but 500mm in the roof. Not because it's calculated to be the best.

Or maybe it's just because you can insulate a lot there without too much hassle and the house would be "weird" otherwise.

But I don't know, so I thought I'd start a thread and hope that many people will chime in.

I'm cutting out a bit of what I and haavard wrote in another thread, where this topic became a bit OT:

But 350 mm of insulation in the roof is neither on the lower nor upper end of what is used today.

Yes, that's how it's usually built (insulated), but it's not the best, just because it's usually done that way.

Imagine you have a certain available amount of insulation, say 100 m3, and you have to insulate walls and roofs completely with this.
You could choose some extreme alternatives, like all in the walls - none in the roof, all in the roof - none in the walls.
It's clear to anyone that these two alternatives are crazy.

But what distribution is best then?
90cm in the roof, 10 cm in the walls?... nah...

The same thickness in the roof as in the walls? - I think that would be great.

And maybe even a little better with about 10% more in the roof than in the walls, as the indoor temperature might be about 10% higher at the roof than somewhere low on the wall.

Heat prefers to escape through the roof, then the walls, and finally the floor.
Today, most houses are built with 40-50 cm insulation in the roof and 23-25 cm in the walls. (I can imagine this is also because it's easier to add a bit more in the roof than in the walls.)

But if it should make sense to have twice as much insulation in one place, it should be motivated by having twice the temperature difference.

Like if it's +25° C at the roof, +5° C at the wall (at floor level, or maybe in the middle of the wall (average)), and -15°C outdoors.

That's how I think.
And I don't realize where I'm thinking wrong if that's the case.

And we had a little more conversation there, but in main features, it was this.
The thread is otherwise here

 

[KnockOnWood]

Interesting thoughts!

And a lot of tradition and opinions surely lie behind how we distribute the insulation.

We "believe" that the air is much warmer up at the ceiling than down at the floor, for example.
I myself think it feels a bit chilly on the feet when I walk around in socks on my wooden floor.
(I don't have underfloor heating.)

But what are the facts? It is now 22.4°C at chest height indoors.
A thermometer just below the ceiling and one lying on the floor just inside the outer wall show the same temperature within ± 0.5°C now that it's about 12°C outside. (What an Indian summer, by the way!)
The ventilation (FTX) naturally helps keep the temperature gradient floor-ceiling at a low level.
So, there is absolutely no "heat cushion" gathering up under the ceiling.

My conclusion is thus that the insulation should be about the same in the ceiling as in the walls.
Maybe a bit more in the ceiling if it is easier/cheaper to insulate there.

I have 310 mm of glass wool in the walls, 290 mm of glass wool in the attic floor + 100 mm of foam plastic on top of the ceiling.
That is probably quite well distributed.
And 300 mm of foam plastic under the floor slab. Unnecessarily much, KA thought, as I wouldn't have underfloor heating.

 

[Locke]

The fact that more insulation is used in the ceiling obviously has a physical background. Partly because there is a greater effect of convection in horizontal insulation, but also the largest temperature differences over the roof due to night radiation and the natural effect of the density difference between warm and cold air inside.

Technically and aesthetically, it is also possible to solve a thick ceiling. Walls have a limit which is a function of the window opening's width and the wall's thickness. The thicker the walls, the wider the windows are required for an acceptable viewing angle out.

 

[mycke_nu]

Heat does rise, as Locke writes, but in practice, it's often the case that bedrooms have a lower temperature on the second floor. Therefore, the reason to insulate extra due to higher temperature often disappears.

Radiation (when it's cloudless) is certainly a factor, but the question is how much this affects under the roof tiles where the outdoor air often circulates freely. I would argue that the temperature under the tiles is fairly independent of the radiation. A speculation on my part due to lack of facts, of course.

As Mikael suggests, I would say that the recommendations are rather broad rules of thumb where, by necessity, a number of variables such as the usage of the premises, roofing material, etc., are not considered.

From a socio-economic perspective, it is, of course, commendable to save energy, but when I insulated the attic here, it was nearly impossible to recoup the cost of thicker insulation than 20 cm in less than about 20 years. (This given that it is rarely or never even -10C here and that the house is terribly drafty.)

Therefore, one needs to think a bit. New construction with open ceilings to the ridge is not the same as renovating an old drafty house. Cheap energy (heat pump) makes it harder to justify insulation.

Not to mention the alleged global warming..

 

[Janus82]

As I see it, it's very cheap to insulate the roof as long as you have a cold attic. You just pour in as much insulation as will fit, and the only cost is the insulation itself. If you want to improve the insulation in the walls, you also have to invest in stronger studs for the insulation to fit, and the costs quickly escalate to unreasonable levels. My guess is that you quickly make an unprofitable investment if you try to fit 500mm of insulation in the walls compared to if you put it in the roof.

 

[Mell]

Yesterday morning I had -2c outside and -6c in the attic. I have about 170mm of insulation in the walls and 300mm in the ceiling and concrete tiles. So there is a certain inertia with weather changes and night vs day temperatures.

 

[KnockOnWood]

Yesterday morning, I had -2c outside and -6c in the attic. I have about 170mm insulation in the walls and 300mm in the roof with concrete tiles. So there is a certain inertia with weather changes and night vs day temperature.

But can't the inertia just as well go the other direction too? So that it's -2░C in the attic when it's -20°C outside.

 

[injonil]

It has been discussed previously on the forum and my conclusion is that Mikael_L is absolutely right.

There are several reasons why this misconception has spread, an important reason is probably that people have been told to add extra insulation to their attic. It is much cheaper to increase the insulation in the roof of a single-story house by throwing up some insulation in the attic compared to adding extra insulation to the walls. Otherwise, it is mostly a lack of knowledge.

 

[largab]

In the same way as insulation is poured onto basement walls below ground. But where the largest loss is, no insulation is placed (basement wall above ground).

 

[arkTecko]

The main reason is not that the insulation is better/more effective in the roof, but that it 'costs' less than in the wall. The cost here is lost living space for the same building area, awkwardly articulated façade, deep windows, etc.