Stopping condensation in a steel-roofed garden building with spray foam

Have you thought about using steel insulated panels for the roof:

- even worse if you use open cell spray foam which is like a sponge (most of the stuff in cans is open cell, the two-part mixes are usually closed cell).

To apply the two part mix properly you'd need all the gear (two cans of chemicals, spray wand, mixing valve, probably some kind of propellant, respirator, ...) and if you mix it wrong you get toxic offgassing. That's why it's better to get a pro to do it, but the pros in the UK are hard to spot between the cowboys.

If you use those panels they're factory-made to meet the right insulation standard, and fairly easy to install:

Or just line with PIR - far easier than foam? Then some thin ply.

Fixing is the tricky bit. I'd think lots of glue and support at the edges if it's not too big a roof.

IME it doesn't seem to absorb anything - just stays 'foamy'. I've read that vapour can pass through open cell, so I suppose some is held in the (tiny) gaps. 'Sponge' sounds a bit dramatic?!

PIR tends to tear in tension - either the foil comes off or a layer of the foam pulls away. If you support it with boarding from underneath (screwed through from the upper roof deck) then you'll probably be OK as there is no tensile load on the PIR.

However you then need to worry about the load of the PIR and the boarding - may need additional rafters to take the weight otherwise it could sag, and cause the deck to sag too.

The nice thing about composite panels is they're also the roof structure so they don't need extra supports.

The problem with open cell is when people put it next to timbers imperfectly. The roof space is warmer (because you insulated it) and hence the air has more moisture. However the gaps get cold and there moisture condenses. It then absorbs into the open cell and is held against the timbers so they don't properly dry out, and so eventually they rot. This is what the UK 'spray foam scandal' is about.

If you have a good drying potential (ie the moisture can escape sufficiently fast one way or another to a well ventilated space, helped as open cell is somewhat vapour permeable) you're ok. If moisture gets trapped by impermeable layers then you're in for trouble.

Theo

Yes, it would need to be closed cell foam, but from a few searches this is easily available in gun-type containers - not needing any mixing.

Yes, coated panels were an option when I bought mine (from Thomas Panels) but I thought (perhaps wrongly) that the extra cost was unnecessary for a garden building (it doesn't seem right to call it a shed when it has blockwork walls).

Yes, good point. You *might* get away with the stickiness of the foil as the loads are quite light. But extra support would be much better.

The OP already has a roof, though.

Ah, yes, OK, I see what you mean now. Thanks.

OoI, can anyone recommend a gun grade one can closed cell foam?

The only explicitly closed-cell one I managed to find is this one:

It seems that other foams have varying percentages of open/closed cells - if you search for "percentage closed cells" and/or ISO4590 it brings up various datasheets. eg this one is 62%:

Meanwhile this is US-style two-part spray foam and they're 95% closed:

Yup I see what you mean - there seems to be a fair amount of "marketing" going on here as well. Many makers of single part foams appear to use "closed cell" to imply that the stuff sets to a rigid mass with a relatively impermeable skin, and open cell to denote something that remains "squishy" like a cushion foam.

That was the cheaper pack - some were over twice that price (and produced a greater volume of foam). Personally I would go with PIR boards rather than a spray foam anyway :-)

Thanks, seems quite complicated - and expensive - to use closed cell.

I can see why closed needs to be used in something exposed to water - so, most external uses. And it's stronger with better insulation properties.

But - say - filling gaps between joists when fitting PIR. Is closed cell necessary? If water were to condense on a poorly insulated or especially prone section, would the tiny amount of water retained in open cell foam be a significant problem?

It's not the slight porosity I have difficulty understanding, it's this idea that open cell foam retains large amounts of water. That's having used it - a lot. Not I'd hasten to add an understanding of the science!

Ive used ordinary externally.

By your criteria unglazed tiles aren't waterproof either...

Course it isn't necessary.

All you want is a barrier to draughts. That has low thermal conductivity

I don't believe it does. Maybe a little, That dries out next day...

It's only a problem where moisture in > moisture out.

Take the OP's shed. Suppose they built an airtight metal shed on bare earth. What then happens on a sunny winter's day is the shed is warmed by the sun, and the warm air inside evaporates water from the earth. The air inside is warm and humid, but when night falls the outside temperature drops rapidly. The metal surfaces become cold, and water from the humid air condenses.

The OP decides to use a thin layer of open cell foam to insulate the shed. Now what you have is foam with say 25C 80% RH air on one side and say -5C on the other, which means there's a temperature gradient across the foam, ie the middle of the foam is at 10C and it gets warmer towards the inside and cooler towards the outside.

At a spot in the thickness of the foam the temperature meets the dew point. In the case of 25C/80% it's 21.3C. Because it's open cell, air can diffuse through it somewhat. But at that spot in the wall the air cannot hold its moisture and so water condenses. That means you have *condensation forming inside the foam*, not on the surface as most people are familiar with seeing.

The inner part of the wall becomes wet. But that is still ok *if* there is a route for the water to get out faster than it comes in, ie the 'drying potential'.

In this case, the outside of the shed is metal so the water can't get out that way. It has to diffuse back into the room, but at night time it can't because the cooling room air is already overly humid. In the daytime maybe it can, but if new moisture is continually being supplied from the earth then perhaps it can't. Eventually the foam becomes wet, and wet foam next to steel starts to cause rust, or rot if there was timber.

There are three ways the OP could break this cycle. First is to add ventilation as they have done - that way the moisture has a chance of escaping from inside, and prevents moisture building up internally in the first place. The second is to add a vapour barrier on the surface of the foam - this means the vapour cannot migrate into the foam and avoids internal condensation. This is why PIR insulation typically has metal foil on the outside, as even closed-cell bare PIR is slightly porous.

The final way to avoid it is a thick enough layer of open cell foam - since the foam has a certain porosity, it means diffusion will drop off with distance from the surface. If you can keep the spot at the dew point far enough away from the surface by having a thick layer of foam, then not much moist air will reach there to condense. But it may be easier just to put a vapour barrier on the inner surface, eg to paint the foam with a non-porous paint.

When randomly throwing bits of expanding foam around to block up gaps etc, you are probably not in the situation where you have both a temperature gradient across it and a lack of routes for it to dry. It also likely doesn't see the same kind of temperature variations as in the shed example. It would also need to see these temperature variations a lot to see the moisture build up - if it was only one day it would be fine, but night after night during the winter and the moisture starts to build up.

Get any of these things right and you avoid the problem, but the OP's shed is a good example of where it's possible to get things wrong.

Theo