Using sand alone as a screed/heatsink for suspended timber floor

thewiseowl12

Hi,

Grateful for any thoughts/experience.

Looking at complete renovation of a 1940's concrete house (mass concrete or block, but no damp proofing) with suspended timber floors. I intend to externally insulate the house.

I have been looking into UFH on the ground floor as I imagine at some time in the future heat pumps will be the only heating option and I want to future proof the house. In the short term I will probably go with gas boiler which will last the next 10 years hopefully.

I will be pulling up the floor and replacing the timber joisting with deeper joists (say 12inch, currenly 7 inch) - what I'm wondering is would sand alone work as the screed for UFH, with no cement mix required? Say 150mm insulation and 50-75mm sand? The floor would be suitably sealed of course so that subfloor ventilation wouldn't interfere. The screed will not be load bearing as the joists will take the weight of the OSB+laminate floors.

If the above is not possible I will just go with radiators I imagine. It will be too much hassle/cost retrofitting concrete subfloors for a number of reasons particular to the property. The above also benefits from me being able to do most of the work myself if it is an option.

Lumen

You'd probably be better off keeping the joists you have and choosing appropriate insulation and a UFH system specifically designed for a suspended floor, e.g. (random example from a few seconds Googling)

https://www.nu-heat.co.uk/blog/underfloor-heating-options-for-first-floors/

...because replacing the joists (with posi joists or equivalent?) sounds like a right PITA.

But in any case, poking holes through continuous external insulation to ventilate a timber subfloor sounds like a bad idea.

thewiseowl12

Lumen wrote: »

You'd probably be better off keeping the joists you have and choosing appropriate insulation and a UFH system specifically designed for a suspended floor, e.g. (random example from a few seconds Googling)

https://www.nu-heat.co.uk/blog/underfloor-heating-options-for-first-floors/

...because replacing the joists (with posi joists or equivalent?) sounds like a right PITA.

But in any case, poking holes through continuous external insulation to ventilate a timber subfloor sounds like a bad idea.

Many thanks for that Lumen. The existing joists appear to have weakened from moisture issues as external vents were blocked and so they will most likely have to be replaced anyways.

There is a 3 ft subfloor below ground level and hence the vents for subfloor ventilation will be sufficently below floor level I hope.

I may look at mechanical ventilation of the subfloor with a humidity control fan which would hopefully brovide sufficient guard against reoccurence once vents are unblocked.

BryanF

Just fill in the floor with lots of insulation, install the ufh pipes, then screed ?

Mech vents for subfloor, joists weakened by moisture, screed over joists.. you are over thinking this..

thewiseowl12

BryanF wrote: »

Just fill in the floor with lots of insulation, install the ufh pipes, then screed ?

Mech vents for subfloor, joists weakened by moisture, screed over joists.. you are over thinking this..

Hi BryanF, thanks for your comments.

For the screed - Is a 50-75mm sand only (no cement) mix a runner?

Thanks.

Gileadi

I don't understand why you want to use a sand only layer? Whats the opposition to using a tried and tested approach?

thewiseowl12

Gileadi wrote: »

I don't understand why you want to use a sand only layer? Whats the opposition to using a tried and tested approach?

Hi Gileadi, you are right that I haven't provided enough context.

Generally sand and cement mixes appear to be used when the floor will take weight - I presume the solution goes semi hard with the addition of water in mixing them? Then flooring is placed on top on this slab.

As I will be placing the screed between timber joists, which will take the flooring, I was wondering if a sand only mix was sufficient and thus there is no need for the cement and additional water element.

sydthebeat

thewiseowl12 wrote: »

Hi Gileadi, you are right that I haven't provided enough context.

Generally sand and cement mixes appear to be used when the floor will take weight - I presume the solution goes semi hard with the addition of water in mixing them? Then flooring is placed on top on this slab.

As I will be placing the screed between timber joists, which will take the flooring, I was wondering if a sand only mix was sufficient and thus there is no need for the cement and additional water element.

how can you place the sand mixture "between" the joists, yet expect the joists to take the weight?

the sand is not going to weight any bit significantly less than if it was concrete

thewiseowl12

sydthebeat wrote: »

how can you place the sand mixture "between" the joists, yet expect the joists to take the weight?

the sand is not going to weight any bit significantly less than if it was concrete

Hi sydthebeat.

I am a little confused by your comment. As I mentioned in a previous post I will be replacing the joists to take any additional weight. I am not expecting that the mix will weigh less due to the exclusion of cement.

My question is simply, can sand only be used at a screed between the timber joists since the finish floor will not be resting on the screed.

sydthebeat

thewiseowl12 wrote: »

Hi sydthebeat.

I am a little confused by your comment. As I mentioned in a previous post I will be replacing the joists to take any additional weight. I am not expecting that the mix will weigh less due to the exclusion of cement.

My question is simply, can sand only be used at a screed between the timber joists since the finish floor will not be resting on the screed.

im confused too....

how exactly will you be installing the sand "between" the joists?

thewiseowl12

sydthebeat wrote: »

im confused too....

how exactly will you be installing the sand "between" the joists?

By pulling up the floor which needs to be replaced. Please see my first post.

I will be pulling up the floor and replacing the timber joisting with deeper joists (say 12inch, currenly 7 inch) - what I'm wondering is would sand alone work as the screed for UFH, with no cement mix required? Say 150mm insulation and 50-75mm sand? The floor would be suitably sealed of course so that subfloor ventilation wouldn't interfere. The screed will not be load bearing as the joists will take the weight of the OSB+laminate floors.

Old ground floor timber floorboards will be pulled up, the joists under them will be replaced. I can install UFH before the OSB and laminate flooring is put down.

sydthebeat

thewiseowl12 wrote: »

By pulling up the floor which needs to be replaced. Please see my first post.



Old ground floor timber floorboards will be pulled up, the joists under them will be replaced. I can install UFH before the OSB and laminate flooring is put down.

ok lets see if i have this clear please...

you want to put new 12" (300mm) timber joists on the existing ground, lets say at 600 mm centers? how do you intend on bracing these joists?

you then want to cut and put 150 board insulation between these joists, leaving say 75mm over this for the sand..... which means you need to install some kind of timber batten to both sides of every joists to keep the 150mm insulation up 75mm from the bottom of the joists (75mm raised, 150 insulation, 75 screed = 300mm joists)

and on top of this insulation you want to install UFH pipes in some kind of loop that would work without significantly damaging the structural integrity of the joists with all the notching that will be required.

the onto this you want to pour dry sand at 75mm depth to act as a heat sink screed between these joists, and suitably tamp down and hope no gaps for this sand to pour down under the insulation.

the fit OSB finishing board over

and all this down on top of a vapour barrier carried up the side of the walls?

and youd do all this instead of building up some filling, installing 150 board insulation, pinning the UFH pipes to this open space, and pouring a 75mm sand and cement screed over??

notwithstanding the likelyhood of you having significant problems trying to get the sand not to pour down like water, the idea of locating timber joists into a non ventilated build up like this just doesnt work. The joists would rot over time with the build up of condensation moisture.

it is not clear at all what particular problem you are trying to resolve with this suggestion.

thewiseowl12

Many thanks for your in depth response sydthebeat. However, you appear to have missed a number of details from my previous posts which mitigate some of your stated concerns.

sydthebeat wrote: »

ok lets see if i have this clear please...

you want to put new 12" (300mm) timber joists on the existing ground, lets say at 600 mm centers? how do you intend on bracing these joists?

you then want to cut and put 150 board insulation between these joists, leaving say 75mm over this for the sand..... which means you need to install some kind of timber batten to both sides of every joists to keep the 150mm insulation up 75mm from the bottom of the joists (75mm raised, 150 insulation, 75 screed = 300mm joists)

In essence, yes. I would center them at whatever the structural engineer deemed appropriate.

and on top of this insulation you want to install UFH pipes in some kind of loop that would work without significantly damaging the structural integrity of the joists with all the notching that will be required.

I am surprised by this in that I presume that the joists would be bored at sufficient depth not to undermine the strength of the timber and had presumed it would not be any more extensive than installation of a traditional radiator system.

the onto this you want to pour dry sand at 75mm depth to act as a heat sink screed between these joists, and suitably tamp down and hope no gaps for this sand to pour down under the insulation.

This is a good point - I would have thought some sort of breathable membrane would achieve this as quite obviously sand would pour away if the insulation layer was not sealed? That said, it would provide one reason for why a cement +sand mix may be more desirable.

the fit OSB finishing board over

and all this down on top of a vapour barrier carried up the side of the walls?

and youd do all this instead of building up some filling, installing 150 board insulation, pinning the UFH pipes to this open space, and pouring a 75mm sand and cement screed over??

In essence, yes. As I mentioned previously, there is a 2-3ft subfloor gap under ground level and that, in addition to a number of other reasons, means putting in a concrete subfloor is not desirable. As I previously stated.

notwithstanding the likelyhood of you having significant problems trying to get the sand not to pour down like water, the idea of locating timber joists into a non ventilated build up like this just doesnt work. The joists would rot over time with the build up of condensation moisture.

it is not clear at all what particular problem you are trying to resolve with this suggestion.

As I mentioned already, the existing ground floor has 2-3ft under it for ventilation purposes and. There is ventilation at this level so condensation of the joists should not be a concern if the subfloor is suitably ventilated.

Thanks

Lumen

Forget about 300mm solid joists. They're too heavy and expensive. Open web joists might have the gaps required to plumb the UFH through them, but I doubt you'll find any UFH installer prepared to do this because nobody does it like this.

The sand between each joist will weigh hundreds of kg. What stops that from dropping through the subfloor? Why use sand at all?

This is an entertainingly insane proposition.

Metric Tensor

It just doesn't work as a design OP.

If you absolutely want to stick with timber joists and have the depth you say you do, then:

1. Install the joists at a lower level than finished floor. (Get an engineer to size them and install appropriate vents at this level.)
2. Put a vapour layer/membrane on top of the joists.
3. Next put appropriately sized/selected plywood or OSB (be careful to get a professional to spec this)
4. Insulation on top of the plywood.*
5. Next put in a DPM on top of the insulation.
6. Install a floor screed of appropriate weight and depth on top of the newly created, insulated deck. This will have your underfloor heating in a concrete screed on top of insulation in a more traditional way.


*Note: It would be possible, subject to professional advice, to push the insulation between the joists if necessary. But if you have the 3 foot you say you do then it might be unnecessary difficulty when you can place it in a continuous layer on top.

thewiseowl12

Lumen wrote: »

Forget about 300mm solid joists. They're too heavy and expensive. Open web joists might have the gaps required to plumb the UFH through them, but I doubt you'll find any UFH installer prepared to do this because nobody does it like this.

Many thanks for the thoughts on this matter Lumen.

The sand between each joist will weigh hundreds of kg. What stops that from dropping through the subfloor? Why use sand at all?

I've done a little further research and it appears that in such situations what is used is a lighter 25mm screed. Each article mentions "suitably beefed up joists". This would make sense from a weight perspective.

The largest room is 3mx4m, so a screed to a depth of 25mm would involve .3m3 of sand which would weigh 500-600kgs(?) approx, over that area. The screed would involve a 10:1 mix of sand and cement (one link). My query around sand alone is that the mix is mainly sand.

Of course it now makes sense to me that cement would be incorporated to make the mix less... portable, as opposed to using some form of breathable membrane instead.

Here's three links.

https://www.theunderfloorheatingstore.com/confused/water-underfloor-heating-for-between-joists-or-suspended-floors

https://www.underfloorheatingsystems.co.uk/underfloor-heating-design/floor-construction/

https://www.thegreenage.co.uk/article/installing-underfloor-heating-with-suspended-timber-floors/

This is an entertainingly insane proposition.

To be honest I don't find this comment helpful, but many thanks for your other suggestions.

thewiseowl12

Metric Tensor wrote: »

It just doesn't work as a design OP.

If you absolutely want to stick with timber joists and have the depth you say you do, then:

1. Install the joists at a lower level than finished floor. (Get an engineer to size them and install appropriate vents at this level.)
2. Put a vapour layer/membrane on top of the joists.
3. Next put appropriately sized/selected plywood or OSB (be careful to get a professional to spec this)
4. Insulation on top of the plywood.*
5. Next put in a DPM on top of the insulation.
6. Install a floor screed of appropriate weight and depth on top of the newly created, insulated deck. This will have your underfloor heating in a concrete screed on top of insulation in a more traditional way.


*Note: It would be possible, subject to professional advice, to push the insulation between the joists if necessary. But if you have the 3 foot you say you do then it might be unnecessary difficulty when you can place it in a continuous layer on top.

Many thanks for the suggestions Metric Tensor, it provides more food for thought.

I guess in some respects I was wondering if UFH could be installed on the second floor too, where your suggestion is not really a runner, and between joist insulation may provide one avenue. However, the weight issue as discussed in other posts renders this idea redundant.

Thanks.

Metric Tensor

thewiseowl12 wrote: »

I was wondering if UFH could be installed on the second floor too.

I have seen this detail used on upper floors quite a bit. Insulation not needed between upstairs and downstairs because there's no heat lost from the house between the two so you have to drop your joist level 50-70 mm and possibly make them deeper. You will lose a bit of ceiling height below so it depends on those heights.


Compacted sand will probably be circa 1700 kg/m3 and concrete is 2400 kg/m3 - that's not going to make a huge difference to the structure in the grand scheme of things. It sounds to me what you are looking for is a "lightweight thermal store" that you can fit between your existing floor joists and somehow have heated by your heatpump. I have never seen such a product I'm afraid and sand is definitely not it.


It might be worth investigating low surface temperature radiators.

thewiseowl12

Metric Tensor wrote: »

I have seen this detail used on upper floors quite a bit. Insulation not needed between upstairs and downstairs because there's no heat lost from the house between the two so you have to drop your joist level 50-70 mm and possibly make them deeper. You will lose a bit of ceiling height below so it depends on those heights.


Compacted sand will probably be circa 1700 kg/m3 and concrete is 2400 kg/m3 - that's not going to make a huge difference to the structure in the grand scheme of things. It sounds to me what you are looking for is a "lightweight thermal store" that you can fit between your existing floor joists and somehow have heated by your heatpump. I have never seen such a product I'm afraid and sand is definitely not it.


It might be worth investigating low surface temperature radiators.

Thats a good synopsis Metric Tensor, as it happens there would be sufficient head height to drop the floor. I will investigate further.

Many thanks for your insightful comments.

thewiseowl12

Metric Tensor wrote: »

I have seen this detail used on upper floors quite a bit. Insulation not needed between upstairs and downstairs because there's no heat lost from the house between the two so you have to drop your joist level 50-70 mm and possibly make them deeper. You will lose a bit of ceiling height below so it depends on those heights.


Compacted sand will probably be circa 1700 kg/m3 and concrete is 2400 kg/m3 - that's not going to make a huge difference to the structure in the grand scheme of things. It sounds to me what you are looking for is a "lightweight thermal store" that you can fit between your existing floor joists and somehow have heated by your heatpump. I have never seen such a product I'm afraid and sand is definitely not it.


It might be worth investigating low surface temperature radiators.

One final query Metric Tensor if you don't mind.

One suggestion is that if I leave 2nd floor joists as are, for soundproofing between floors I could use gypsum plasterboards laid on existing floorboards with laminate placed on top as the finished flooring.

Just a thought - if I put in UFH heating pipework below this from below and insulate below this using rockwool or similar, replacing ceiling, it would presumably work as a cheap form of thermal store, in addition to providing some soundproofing benefits.

The density of 12.5mm plasterboard appears to be 640kgm3 as compared to 2400 as you pointed out for concrete.

Thanks.

Metric Tensor

Gypsum boards are not a good thermal store. They are both thin and lightweight. They absorb heat fast but lose it fast also. (I'm not convinced this plan would benefit you much in terms of sound either - but that's a story for a different day.) For thermally massive stores you need something "massive" - that's why old storage heaters were filled with bricks! Thin or light materials are "usually" not good for this duty although there may be specialist ones available.

On a related point - "standard" UFH pipes are specifically designed to transmit heat when embedded in a dense matrix like concrete. They will not perform correctly if just "attached" to something. I think there are specific types for other uses but not something I know a lot about.

Metric Tensor

Gypsum boards are not a good thermal store. They are both thin and lightweight. They absorb heat fast but lose it fast also. (I'm not convinced this plan would benefit you much in terms of sound either - but that's a story for a different day.) For thermally massive stores you need something "massive" - that's why old storage heaters were filled with bricks! Thin or light materials are "usually" not good for this duty although there may be specialist ones available.

On a related point - "standard" UFH pipes are specifically designed to transmit heat when embedded in a dense matrix like concrete. They will not perform correctly if just "attached" to something. I think there are specific types for other uses but not something I know a lot about.

thewiseowl12

Metric Tensor wrote: »

Gypsum boards are not a good thermal store. They are both thin and lightweight. They absorb heat fast but lose it fast also. (I'm not convinced this plan would benefit you much in terms of sound either - but that's a story for a different day.) For thermally massive stores you need something "massive" - that's why old storage heaters were filled with bricks! Thin or light materials are "usually" not good for this duty although there may be specialist ones available.

On a related point - "standard" UFH pipes are specifically designed to transmit heat when embedded in a dense matrix like concrete. They will not perform correctly if just "attached" to something. I think there are specific types for other uses but not something I know a lot about.

Again, many thanks for the considered thoughts!

sydthebeat

thewiseowl12 wrote: »

Again, many thanks for the considered thoughts!

your looking for solutions to a problem that doesnt exist.

your looking for a lightweight material that is a great thermal store.
these two things are practically oppositional.

UFH works best with a material that can absorb a large amount of heat, and then release said heat in a slow deliberate consistent manner. The lighter the material, the quicker the heat dissipates.

Dudda

Haven't read all the posts but I've done something similar to what the OP is looking for where I used sand with UFH on an existing floor. It's called a biscuit mix where sand and cement are mixed dry 1:8 approx.

Few reasons I did it. I needed to keep the existing floor height and timber which was appropriately sized. Systems exist where you can use aluminium panels or 20mm insulation that the pipes sit into. The problem with these is they distribute heat fast rather than a more controlled longer steady flow. As the house was insulated to passive standard I wanted a slow release of heat. The system was a lot of work but very easy to do myself saving a lot of money (but probably not time).

First I put airtight tape around the ends of all joists and parge coated the existing wall. Then I fixed timber battens to the side of the joists and then put in 50mm of insulation. This was kept down from the top of the joists by 50mm. Additional 100mm insulation was placed to the perimeter to line up with the drylining to form a continuous internal insulation layer. Then the UFH pipes were clipped to the floor insulation with small notches having to be removed from every second joist. This was infilled with a DRY mix of sand and cement. Next OSB was glued and screw fixed to the joists (paying particular attention not to screw through a UFH pipe.

It's worked out very well as the floor is very solid, has a slow release of heat and the weight really helps with noise reduction.

I did this on the first floor level where I wanted to keep the joists at existing floor height and have a slow release of heat. I don't see the need to do this on the ground floor in your case.

Photos attached.

Yellow_Fern

Lumen wrote: »

You'd probably be better off keeping the joists you have and choosing appropriate insulation and a UFH system specifically designed for a suspended floor, e.g. (random example from a few seconds Googling)

https://www.nu-heat.co.uk/blog/underfloor-heating-options-for-first-floors/

...because replacing the joists (with posi joists or equivalent?) sounds like a right PITA.

But in any case, poking holes through continuous external insulation to ventilate a timber subfloor sounds like a bad idea.

Does anyone have any experiences with these systems?

Lumen

thewiseowl12 wrote: »

It will be too much hassle/cost retrofitting concrete subfloors for a number of reasons particular to the property. The above also benefits from me being able to do most of the work myself if it is an option.

Out of interest, what are these reasons?

If you don't want to pour concrete, is there any value in considering an alternative insulating subfloor material, like cellular glass (foamglas or similar). Prices look to be about €150/m3.

(I'm just asking the question, I have no idea whether this is a good idea. If it was, lots of people would do it, right?)

BryanF

thewiseowl12 wrote: »

It will be too much hassle/cost retrofitting concrete subfloors for a number of reasons particular to the property. The above also benefits from me being able to do most of the work myself if it is an option.

Out of interest, what are these reasons?.[/QUOTE]

BryanF wrote: »

Just fill in the floor with lots of insulation, install the ufh pipes, then screed ?

Mech vents for subfloor, joists weakened by moisture, screed over joists.. you are over thinking this..

why are you opposed to traditional concrete screed?
In terms of Diy get a wacker and few ton bags of gravel, a load of sheets of EPS a, radon barrier and then do your screed say 100mm with a steel mesh.

I’ve done my own suspended floors with insulation and vapour barrier, but you have to change some joists? Make them air-tight, insulate and then add a screed to that?

I’ve installed the screed over joists on a floor over basement but in your case I don’t see the point of screed on upper floors, it’s not necessary.

thewiseowl12

Out of interest, what are these reasons?

If you don't want to pour concrete, is there any value in considering an alternative insulating subfloor material, like cellular glass (foamglas or similar). Prices look to be about €150/m3.

The reasons are that (1) with a 3ft subfloor I would have thought it a lot of hassle to build up dwarf walls and the subfloor to sufficient height to pour concrete on top. (2) I suppose also that there is no damp proof coursing in the external walls leaves me uneasy in terms of filling in this space. I know there are retrofitting damp course solutions but I suppose I'm happy to respect that they built the house a certain way for a certain reason and it seemed like the proposed solution using a biscuit type mixture, keeping a suspended timber floor, may be in keeping with that.

BryanF wrote: »

why are you opposed to traditional concrete screed?
In terms of Diy get a wacker and few ton bags of gravel, a load of sheets of EPS a, radon barrier and then do your screed say 100mm with a steel mesh.

I’ve done my own suspended floors with insulation and vapour barrier, but you have to change some joists? Make them air-tight, insulate and then add a screed to that?

Part of my reasons for keeping away from concrete pouring are as you mention in the next bit - a basement. The house is build on a slope and I am toying with the idea of installing a basement at some stage, if required, and if the costs add up. Consequently Im just exploring options other than the pouring of a permanent screed at present.

I’ve installed the screed over joists on a floor over basement but in your case I don’t see the point of screed on upper floors, it’s not necessary.

Many thanks for your thoughts in terms of the upper floors BryanF - I agree with you. How did you strengthen the joists before you poured the screed or what was your structure?

Thanks.

Lumen

thewiseowl12 wrote: »

The reasons are that (1) with a 3ft subfloor I would have thought it a lot of hassle to build up dwarf walls and the subfloor to sufficient height to pour concrete on top. (2) I suppose also that there is no damp proof coursing in the external walls leaves me uneasy in terms of filling in this space. I know there are retrofitting damp course solutions but I suppose I'm happy to respect that they built the house a certain way for a certain reason and it seemed like the proposed solution using a biscuit type mixture, keeping a suspended timber floor, may be in keeping with that.

Part of my reasons for keeping away from concrete pouring are as you mention in the next bit - a basement. The house is build on a slope and I am toying with the idea of installing a basement at some stage, if required, and if the costs add up. Consequently Im just exploring options other than the pouring of a permanent screed at present.

I suspect a fully costed proposal would quickly dent your basement ambitions.

I think your problem is that you're second-guessing engineering concerns with no qualifications or experience. I have the same type of house and the engineers recommendation was an insulated concrete subfloor. No hesitation, no equivocation. Like you, I expressed hand-waving concerns for preserving the original building design but they were met with a kind of blank "does not compute" incomprehension.

As I understand it, infill is the cheapest, best solution, particularly since you want UFH. If you don't believe me, get some professional advice.

thewiseowl12

Lumen wrote: »

I suspect a fully costed proposal would quickly dent your basement ambitions.

I think your problem is that you're second-guessing engineering concerns with no qualifications or experience. I have the same type of house and the engineers recommendation was an insulated concrete subfloor. No hesitation, no equivocation. Like you, I expressed hand-waving concerns for preserving the original building design but they were met with a kind of blank "does not compute" incomprehension.

As I understand it, infill is the cheapest, best solution, particularly since you want UFH. If you don't believe me, get some professional advice.

HI Lumen, many thanks for that - I wouldn't say that it is a case I don't believe you, I just wanted to hear about peoples experiences and explore a few options! I would say you are 100% correct on the basement costings.

In terms of mitigating rising damp etc, how did you do that in your own place?