Underfloor Insulation

[RJH]

I'm thinking of insulating a suspended timber floor with 90mm insulation board. It's quite tricky to access, with pipes and wiring making it more difficult.

I'm thinking of getting some 1.2 x 0.6m seconds, and doing it in 60cm sections (the joists are about 40cm between centres), and joining/ sealing the ends with expanding insulation foam.

I'd guess that the performance of the foam, combined with an element of cack-handedness, would compromise the insulation to an extent - but would it be significant?

[The Natural Philosopher]

no, but airgaps will be.

[Andrew]

If there is any air-leakage between the warm side of the insulation and the floor itself, then you have wasted your time.

Stopping air-leakage is tricky when the cold side is in a ventilated and potentially draughty subfloor. This why packing the gap between the joists with rockwool or fibreglass supported on netting draped across the joists or stapled underneath, or sitting on bits of treated rough timber nailed under the joists with gaps for limited air movement is preferable. This still allows water vapour molecules to move through the insulation and escape.

If you successfully stop all air leakage between the junction of insulation, joists and flooring then you also prvent the timber from being ventilated. This is what gives dry rot a wonderful breeding ground.

Unless you can get into the crawl space, then it's going to be difficult without ripping up half the floor to start with.

[Tim Watts]

No - sounds fine. But you don't really need 90mm for a floor if you are talking about using celotex/kingspan PIR foam. 50mm will make a huge difference - a lot of the gain will be down to closing off the draughts.

75mm at the most would be super excellent.

I recommend using pink fire rated foam and a proper foam gun. Plain foam can be a bit flammable. The pink stuff burns but at least self extinguishes - about the same as the celotex you'll use should. And it's not much more expensive.

I would put in a line of screws into the sides of the joists for the board to sit down on and prevent slipping.

[The Natural Philosopher]

But you prevent any warm wet air from reaching it as well.

THii is how insulation works.

No, it isn't.

YOu should rip up the WHOLE floor, and stuff celotex between the joists, and foil tape OVER the joists, to prevent vapour from getting to the beams.

Ideally then lay more celotex over the rest, and re-lay a floating floor of chip .

Or omit the second layer and relay the original floor.

[RJH]

The main source of draught is at the walls due to gaps between the boards and walls, and wires/pipes. The floors themselves are pretty well sealed - either vapour barrier/laminate or underlay/carpet. But the insulation board would pretty well seal it.

As for thickness. I did calculate using the Kingspan app,and a PA ratio of 0.8, target U-value of 0.25. That comes out at 90mm.

That said, I did a section of floor using 25mm leftovers. That gives a measured temperature difference of 2C. So I'm not sure of the optimum price/benefit here.

Yes, thanks, good plan. I've had a foam gun for a while - top tool :-)

When I did the 25mm, I used a thin line of foam and waited for it to go tacky. But pinning with screws sounds like a good idea to hold them in place more firmly.

[RJH]

Yes, I'd agree.

Floor up isn't an option, alas. I'd have thought that the cold bridge would be at the lowest point, under the floor. But you're saying the tape should be set in the point where the joist meets the floor?

FWIW, Kingspan make no mention of cold bridging, or the need for tape:

[Andrew]

I suggest you look up 'interstitial condensation' that some of the insulation manufactures have apps for, to calculate what the effect of adding insulation willy-nilly, without thought or planning might cause.

It'll be the next 'big worry' for mortgage lenders surveyors to fret over in about 15 years time. Not helped if the weather is generally warmer and damper.

A relative caused thousands of pounds of damage to his roof structure by stuffing the 'eaves' with rockwool about 20 years ago. House was a 1960's row of attached properties where each property had a roof where the ridge went from front to back. That meant there was a valley along the top of every party wall. The bathroom was against this party wall and cold. Houses originally had warm air ducted heating with a huge electrically heated thermal store in the middle of the house.

The valley above the party wall and stair well was a notorious cold spot. Unfortunately, being unventilated, stuffing the angled junction inside the loft allowed condensation to destroy the ends of all the roof trusses. Easy mistake to make, but *very* expensive.

[RJH]

But the warm air has to pass through the floor and condense on a cold surface. I can't see how this can happen in my case.

I've no doubt. This house - a Victorian mid-terrace - was similarly treated, and the was close to some lasting damage. Surveyor didn't pick up on it.