Insulating a old stone bothy - worth doing the floor or just walls and roof?

Did the same dilemma on a stone bothy in Perthshire a few years back before I converted my van full-time. The temptation is always to skip the floor because it feels like the least glamorous job — but honestly, it was the single biggest difference we noticed.

Stone floors lose heat to the ground constantly, especially in Scotland where the ground stays cold well into summer. We laid 100mm PIR board under a floating timber deck and the place transformed. Previously you'd get your walls and roof properly sorted and still wonder why it felt cold and clammy — that's the floor robbing you.

The other thing nobody mentions is vapour control. Old stone walls need to breathe or you'll get interstitial condensation and a mould nightmare within two seasons. Don't just slap Celotex directly against the stonework. Batten it out, leave an air gap, use something like wood fibre board instead if you want to be sympathetic to the building.

My rough priority order would be:

1. Floor (biggest gain per pound)
2. Roof/ceiling (heat rises, obvious)
3. Walls (do it right or don't bother)

What's the current state of it? Is there already a timber floor, or are you working with bare stone/earth? That changes the approach quite a bit. Also worth knowing if it's listed or in a conservation area — that can rule out certain materials immediately.

Anyone else tackled a traditional Scottish or Welsh stone building? Curious whether others found the breathable approach worth the extra cost versus standard PIR.

@CharlieThomas — don't skip the floor, full stop.

I learnt this the hard way with my cabin build. Did walls and roof first, felt smug about it, then spent two winters wondering why the place still felt like a meat locker by morning. The cold was coming straight up through the stone flags.

Laid 100mm Celotex across the whole floor the following spring, then tongue-and-groove over the top. The difference was genuinely shocking — heating times halved, and my little Victron setup wasn't getting hammered trying to compensate overnight.

Stone bothies lose a brutal amount of heat downward, especially if there's any ground moisture involved. Heat rises, yes — but cold radiates up through your feet and kills the whole room's thermal mass advantage you're trying to build.

Do the floor. Do it first if you can stomach the sequencing.

@MountainHermit is absolutely right, and I'll add the pedant's footnote nobody asked for: ground loses heat differently to walls. Walls lose heat via convection and conduction — the floor is almost entirely conduction, meaning you're essentially welded to a giant cold thermal mass all winter.

In my shepherd's hut (sits on wheels, so slightly different beast), I learned that cold feet = miserable nights, full stop. 75mm of Celotex under the decking transformed it completely.

For a stone bothy sitting directly on earth, I'd argue the floor is actually more critical than the walls, not less. Stone walls at least have some thermal mass working for you occasionally. The ground beneath? It's just endlessly, boringly cold. Like a particularly determined Scottish winter that's given up trying to impress anyone.

Do the floor. Do it first. Thank yourself in February.

@CopperWelder pedant's footnote appreciated actually.

Practical bit nobody's mentioned yet — vapour barrier placement on a stone bothy floor matters more than most people realise. Stone floors will wick moisture upward constantly. Get that DPM layer wrong and you're just trapping damp under your insulation, which then rots your boarding and kills your UFH if you've gone that route.

Did PIR board with a decent overlap up the walls before screed in my tiny house build. Faff at the time but dry as a bone underneath ever since.

Also — thermal mass of a stone bothy floor can actually work for you if you're heating the space regularly. Worth thinking about before you go full insulation mode underneath. Depends entirely on your usage pattern, intermittent vs continuous occupation changes the whole calculation.

@OffGridMax raises the vapour barrier point and it's worth lingering on that for a second, because I've seen people get it catastrophically wrong on boats — same physics applies to any structure sitting near damp ground.

Stone bothies are notorious for lateral moisture migration too. The floor insulation is almost pointless if you haven't broken the capillary path from the walls. I watched someone do a beautiful Kingspan floor installation in a similar Perthshire building, then scratch their head two winters later wondering why it still felt damp. The wall base was wicking straight past the DPM.

Do the floor, yes — but treat it as part of a system rather than a standalone job. Anything less and you're just shuffling the cold and damp around rather than actually dealing with it.