Floor Insulation

[canalsiderenovation]

For our existing floor and wall space of our bungalow we have made it clear to our architect and builders we want to improve the insulation as we will have wet UFH fitted with our ASHP. We are lucky in that we have higher than normal ceiling space in our existing bungalow (though it varies as our current layout is a bit like two bungalows so one is different than the other) and we will need super flat flooring for the gluedown LVT.

 

The builder was talking about 50mm or 75mm - anyone interpret these scribbles so I can get some idea of what he is talking about (still waiting for his quote yet) and which would be better?

 

We will also have wall insulation added and our windows will be triple glazed.

[PeterW]

That’s very thin. What is the proposed floor finish ..?

 

You could go 100mm insulation with one of the thin UFH board systems like the one from Wunda at 22mm on top, so your less than the 135mm and substantially better heat retention with a smaller loss of heat store from a screed. 

[Temp]

+1

 

We have a mixture but in some rooms we have 80mm PIR insulation and about 60mm screed with UFH. Wish we had more insulation. 

 

Our favourite rooms have 21mm engineered oak flooring on battens with PIR between and no screed. The top layer of PIR has foiled covered grooves for the pipes that spread the heat. A system like that would allow a greater depth of insulation.

 

 

[canalsiderenovation]

42 minutes ago, PeterW said:

That’s very thin. What is the proposed floor finish ..?

 

You could go 100mm insulation with one of the thin UFH board systems like the one from Wunda at 22mm on top, so your less than the 135mm and substantially better heat retention with a smaller loss of heat store from a screed. 

 

I'm a bit ignorant. Note that these measurements are only in relation to insulating our existing. The extension is bigger than our existing so this will be in line with building regs (I never thought to clarify what this will be as I was more worried on our existing). Our existing is a concrete pad (bunglow built in late 70s).

 

We are having a gluedown LVT if that's what you mean.

 

Edited March 16, 2020 by canalsiderenovation

[Ferdinand]

That's not enough imo.

 

My downstairs has about 80mm PIR or so ( extensive extend and reno of a bungalow, done in 2008),and I would far prefer it to be 125 or 150mm, since the ufh struggles.

 

Peter's idea is probably a good one.

 

Ferdinand

 

[canalsiderenovation]

13 minutes ago, Ferdinand said:

That's not enough imo.

 

My downstairs has about 80mm PIR or so ( extensive extend and reno of a bungalow, done in 2008),and I would far prefer it to be 125 or 150mm, since the ufh struggles.

 

Peter's idea is probably a good one.

 

Ferdinand

 

 

Thanks @PeterW it will be something to take forward with the builder. Having had no heating for 5 months now being warm and low heating bills is a priority, I'm so nesh.

[MikeGrahamT21]

Can you dig out the existing concrete?

[canalsiderenovation]

11 minutes ago, MikeGrahamT21 said:

Can you dig out the existing concrete?

 

Yes but due to cost we would prefer not to, especially as our ceiling height isn't an issue and all doors, skirting plaster etc being replaced....

[ryder72]

On a completely different note - 

 

We installed a kitchen with stone tops on an island on a floor that has 50mm PIR insulation and then screed on top. The worktop cracked at the hob in about a month and it was then discovered that the PIR was just not thick enough to take the point loading from the island. 

 

It turned out into a huge job for the builder to dig everything out and start again.

 

I am not an expert in insulation but 50mm may not be enough from a loading persective.

[canalsiderenovation]

15 minutes ago, ryder72 said:

On a completely different note - 

 

We installed a kitchen with stone tops on an island on a floor that has 50mm PIR insulation and then screed on top. The worktop cracked at the hob in about a month and it was then discovered that the PIR was just not thick enough to take the point loading from the island. 

 

It turned out into a huge job for the builder to dig everything out and start again.

 

I am not an expert in insulation but 50mm may not be enough from a loading persective.

 

Crikey!

 

I'm going to speak to the builder about the suggestion above about 100mm insulation with one of the thin UFH board systems like the one from Wunda at 22mm on top. 

 

If nothing else one thing I need to make clear is no less than 100mm insulation.....

[canalsiderenovation]

1 hour ago, PeterW said:

That’s very thin. What is the proposed floor finish ..?

 

You could go 100mm insulation with one of the thin UFH board systems like the one from Wunda at 22mm on top, so your less than the 135mm and substantially better heat retention with a smaller loss of heat store from a screed. 

 

@PeterW can you post links of what you mean so I can get my head around it. I just want to understand it before I try and communicate to the builder what I mean.....

[PeterW]

22 minutes ago, ryder72 said:

On a completely different note - 

 

We installed a kitchen with stone tops on an island on a floor that has 50mm PIR insulation and then screed on top. The worktop cracked at the hob in about a month and it was then discovered that the PIR was just not thick enough to take the point loading from the island. 

 

It turned out into a huge job for the builder to dig everything out and start again.

 

I am not an expert in insulation but 50mm may not be enough from a loading persective.

PIR has a large point load capability but I expect here the screed was too thin. 
 

Any info on the screed depth..?? 

[Mr Punter]

22 minutes ago, ryder72 said:

On a completely different note - 

 

We installed a kitchen with stone tops on an island on a floor that has 50mm PIR insulation and then screed on top. The worktop cracked at the hob in about a month and it was then discovered that the PIR was just not thick enough to take the point loading from the island. 

 

It turned out into a huge job for the builder to dig everything out and start again.

 

I am not an expert in insulation but 50mm may not be enough from a loading persective.

 

It would be the screed not thick or strong enough and / or the load from the kitchen units not spread far enough.

[PeterW]

Have a look at this video for LVT 

 

 

These are the boards

 

https://www.wundatrade.co.uk/product-category/home/overfloor-retro-fit-solutions/wundatherm-underfloor-heating-boards/

 

 

 

[ryder72]

5 minutes ago, PeterW said:

PIR has a large point load capability but I expect here the screed was too thin. 
 

Any info on the screed depth..?? 

I think the screed was 50-60mm.

 

The floor was covered in a vinyl tile finish.

 

What we found was that the screed had cracked where the stone slab ends of the island (1200 * 900*20) were in contact with the floor and the unit legs were passing the weight into the floor. This is very unusual and after digging the screed it was found that the PIR has sunk in the region.

 

The site structural engineer was of the opinion that 50mm PIR is not really suitable and it was a possible combination of insufficient thickness of PIR and screed that caused the issue.

 

 

[Jeremy Harris]

This isn't related to the PIR thickness at all, really, it's a consequence of the inability of the floor surface above to adequately spread the point loads.  One reason that passive slabs use reinforced concrete over the top of foam insulation is to increase the strength of the slab so that any imposed point loads don't cause the slab to fail locally.

 

The risk can be mitigated by making the floor stiffer and stronger, or it can be mitigated by reducing the magnitude of any point loads.  For example, if a kitchen island was built with wide, stiff, plates under the feet, then that would very significantly reduce the point load and remove the risk.  The same may apply to kitchen units in general, although usually there will be lots of feet to reduce the load imposed by each of them.  An island is probably the likely worst case, as it may only have a few feet to transmit a pretty hefty load into the floor.

[canalsiderenovation]

18 minutes ago, PeterW said:

Have a look at this video for LVT 

 

 

These are the boards

 

https://www.wundatrade.co.uk/product-category/home/overfloor-retro-fit-solutions/wundatherm-underfloor-heating-boards/

 

 

 

 

Thanks very much for this. A minimum of 100mm insulation is what I'm going to communicate to the builder/architects for the existing.... 

[ryder72]

8 minutes ago, Jeremy Harris said:

This isn't related to the PIR thickness at all, really, it's a consequence of the inability of the floor surface above to adequately spread the point loads.  One reason that passive slabs use reinforced concrete over the top of foam insulation is to increase the strength of the slab so that any imposed point loads don't cause the slab to fail locally.

 

The risk can be mitigated by making the floor stiffer and stronger, or it can be mitigated by reducing the magnitude of any point loads.  For example, if a kitchen island was built with wide, stiff, plates under the feet, then that would very significantly reduce the point load and remove the risk.  The same may apply to kitchen units in general, although usually there will be lots of feet to reduce the load imposed by each of them.  An island is probably the likely worst case, as it may only have a few feet to transmit a pretty hefty load into the floor.

Jeremy - the only thing different about this floor from hundreds others was the thinner PIR  or possible poor screed mix.

 

Nothing about the make up of the island indicated that it was unusually heavy or units overloaded.

 

So it stands to reason that something about the floor make up was not right. I am not sure under what circumstances should stiff plates be specified under a kitchen island.

 

[Jeremy Harris]

1 hour ago, ryder72 said:

Jeremy - the only thing different about this floor from hundreds others was the thinner PIR  or possible poor screed mix.

 

Nothing about the make up of the island indicated that it was unusually heavy or units overloaded.

 

So it stands to reason that something about the floor make up was not right. I am not sure under what circumstances should stiff plates be specified under a kitchen island.

 

 

 

I'll repeat, it has nothing at all to do with the thickness of the PIR (or any other insulation).  All foam insulation has a maximum compressive stress that determines how much it deflects when loaded.  Making the insulation thicker doesn't change this.  For example, the EPS under our floor has a maximum allowable compressive stress of 100 kPa for 10% compression.  In reality this equates to about 10 kPa for a more realistic 1% deflection .   It doesn't matter if the insulation is 50mm thick, 100mm thick or 300mm thick, the max allowable compressive stress remains exactly the same.  To get a higher value means changing the insulation for one with a higher max allowable compressive stress, like EPS300 that can take 300 kPa for 10% deflection.

 

The ability of the screed to work to spread a point load more evenly into the underlying insulation is key, as is the magnitude of the point load and the area over which it is imposed (which gives the imposed bearing stress on the floor immediately under the foot).

 

The best analogy I can think of is snow shoes.  Walk on thick snow in ordinary boots and your feet will sink in.  Spread the same load over a wider area by using snow shoes and you can walk on top of the snow.  If you have a floor where the screed isn't strong enough to take the point load, and so fails and crushes the underlying foam insulation, then spreading the load over a wider area will stop that from happening.

.

[ryder72]

1 hour ago, Jeremy Harris said:

 

 

I'll repeat, it has nothing at all to do with the thickness of the PIR (or any other insulation).  All foam insulation has a maximum compressive stress that determines how much it deflects when loaded.  Making the insulation thicker doesn't change this.  For example, the EPS under out floor has a maximum compressive stress of 100 kPa for 10% compression.  In reality this equates to about 10 kPa for a more realistic 1% deflection (any more than this might make the floor above crack). 

 

The ability of the screed to work to spread a point load more evenly into the underlying insulation is key, as is the magnitude of the point load and the area over which it is imposed (which gives the imposed bearing stress on the floor immediately under the foot).

 

The best analogy I can think of is snow shoes.  Walk on thick snow in ordinary boots and your feet will sink in.  Spread the same load over a wider area by using snow shoes and you can walk on top of the snow.  If you have a floor where the screed isn't strong enough to take the point load, and so fails and crushes the underlying foam insulation, then spreading the load over a wider area will stop that from happening.

.

Ok - so for a layman is it fair to assume in this case that the cause for the failure was the screed. Nothing usual in the makeup of the island and the load imposed by it.

 

[Jeremy Harris]

5 minutes ago, ryder72 said:

Ok - so for a layman is it fair to assume in this case that the cause for the failure was the screed. Nothing usual in the makeup of the island and the load imposed by it.

 

 

 

Most probably.  Any floor will have a maximum allowable point load capacity, so the solution is either to change the floor to one with a higher point load capacity or reduce the point load to remain within the floor limit.  It's very common to do the latter, thing like heavy water tanks can exceed the floor load capacity and the solution is to fit bearers, so the load is spread over a greater area.  I've seen a bed go though flooring for the same reason, that the point load imposed by the legs exceeded the capacity of the floor.

[ryder72]

1 minute ago, Jeremy Harris said:

 

 

Most probably.  Any floor will have a maximum allowable point load capacity, so the solution is either to change the floor to one with a higher point load capacity or reduce the point load to remain within the floor limit.  It's very common to do the latter, thing like heavy water tanks can exceed the floor load capacity and the solution is to fit bearers, so the load is spread over a greater area.  I've seen a bed go though flooring for the same reason, that the point load imposed by the legs exceeded the capacity of the floor.

 

Thanks. Thats useful.

 

Wonder how the structural engineer made the comments that he did.

[Jeremy Harris]

3 minutes ago, ryder72 said:

 

Thanks. Thats useful.

 

Wonder how the structural engineer made the comments that he did.

 

 

No idea, TBH. 

 

If you want to do a simple experiment to see what happens, then get a bit of foam insulation (or foam packaging), a bit of cardboard and a bit of strong plywood.  Put the cardboard over the foam (this simulates a relatively weak screed or floor) and then push on it with a narrow stick.  You'll find that the cardboard won't be able to spread the load and will fail, resulting in the stick sinking through it. 

 

Try the same with a bit of stiff plywood over the foam and you'll find it takes a great deal more force to try and push the stick through it.  In each case the foam hasn't changed at all, it's just that the surface above has, with the plywood spreading the load more evenly.

[ADLIan]

As above nothing to do with the thickness of the insulation, problem is the screed. Standard sand & cement screed onto insulation should be at least 65mm thick with light mesh reinforcement (or fibre reinforcement mixed in). The kitchen unit legs may have imposed too high a point load on the screed & insulation but this is unlikely.

Edited March 16, 2020 by ADLIan
Corrected screed thickness to 65mm

[Ferdinand]

2 hours ago, ryder72 said:

On a completely different note - 

 

We installed a kitchen with stone tops on an island on a floor that has 50mm PIR insulation and then screed on top. The worktop cracked at the hob in about a month and it was then discovered that the PIR was just not thick enough to take the point loading from the island. 

 

It turned out into a huge job for the builder to dig everything out and start again.

 

I am not an expert in insulation but 50mm may not be enough from a loading persective.

 

In that situation you could also have used a structural plinth (as they used to be done), which would spread the load, or even used an island with a flat bottom.

 

Or made the legs shorter and just sat it on a sheet of 18mm or 22mm plywood with the plinths round the side hiding it.

 

Ferdinand