Concrete floor - screeding - final brain check

6 views
Skip to first unread message

Tim S

unread,
Dec 28, 2008, 6:19:35 AM12/28/08
to
Hi again,

Quick recap - got to make up a floor with about 25-35mm screed. The current
surface is solid, but it's a bit of a "farmer's concrete" special, meaning
it was made with small round pebbles and in a few patches the pebbles are
not well attached. ie they're attached but it doesn't take much to loosen
them.

I suspect if we go digging the less well attached pebbles out we'll get to
earth fairly quickly. No vapour barrier either. I should add, that the
floor's been fine under quarry tiles so I don't see any need to go
replacing it - I just want to make sure the screed gets a good bond.


How does this sound:

1) Soak the floor in a stabiliser.

2) Prime the floor with SBR/cement slurry

3) Screed with 4:1 sand/cement + SBR

4) Paint damp proof membrane on top.

5) Finish (ceramic tile or engineered wood depending on location)

?

=====

What would be good for 1): Liquid SBR? I think PVA will bite the dust due to
the fact there's no vapour barrier.

Also, any good recommendations of a paint on DPC membrane?

Many thanks

Tim

The Natural Philosopher

unread,
Dec 28, 2008, 6:45:46 AM12/28/08
to

It may go slightly soft, but it wont dissolve back to liquid.
You probably could combine the two functions of stabiliser and DP by
using something a bit more epoxy resin-ish than PVA.

There are coatings for exterior walls designed to reduce driving rain
penetration.. I forget the name. I'd go that way.

Tim S

unread,
Dec 28, 2008, 7:05:26 AM12/28/08
to
Hi,


The Natural Philosopher coughed up some electrons that declared:

>
> It may go slightly soft, but it wont dissolve back to liquid.
> You probably could combine the two functions of stabiliser and DP by
> using something a bit more epoxy resin-ish than PVA.

Ah. OK - I'll do some research for epoxy products. I've seen some mention of
aqueous epoxy solutions - I'll have a dig on google.

> There are coatings for exterior walls designed to reduce driving rain
> penetration.. I forget the name. I'd go that way.
>

I saw those in Wickes yesterday. I could grab a tin and try it on a patch
and see how it goes.

Many thanks indeed :)

Cheers

Tim

Right, off to do more leccy box sinking now...

Bruce

unread,
Dec 28, 2008, 7:15:11 AM12/28/08
to
Tim S <t...@dionic.net> wrote:
>What would be good for 1): Liquid SBR? I think PVA will bite the dust due to
>the fact there's no vapour barrier.
>
>Also, any good recommendations of a paint on DPC membrane?


Some links here:
http://www.safeguardeurope.com/products/vandex_range.php
http://www.riw.co.uk/applications/tile_render_key.htm
http://www.ronacrete.co.uk/CMS/13.asp
http://www.palacechemicals.co.uk/Admixtures.htm

stuart noble

unread,
Dec 28, 2008, 8:09:14 AM12/28/08
to

If there's no damp, you shouldn't need SBR at all, but it would be an
extra precaution, at least for the stabiliser.
SBR doesn't mix well with sand cement mortars. It has a tendency to go
its own way and leech out if you leave it standing for any length of
time. Pva mixes in better and the mortar is easier to use because of that.
You could just use sand/cement for the screed and paint SBR on when it's
dry. It's penetrates like crazy, which is better than creating a film on
the surface

meow...@care2.com

unread,
Dec 28, 2008, 11:50:46 AM12/28/08
to
Tim S wrote:


I realise you dont want to be bothered, but now is your opportunity to
dig it up and put insulation down that will save you n times the cost
of doing the job. Where n is an unknown number that someone else might
have more idea about.


NT

Phil L

unread,
Dec 28, 2008, 12:15:29 PM12/28/08
to

The floor sounds like it's been botched up more than once in the past and
now you appear to want to do the same again, while you might save a few
quid, you are still left with a floor with no insulation and no damp
proofing, and believe me, trying to rectify this once the units are fitted
is nigh on impossible, not to mention astronomical in price.....doing it now
while the house is empty and in such a state is both economical and
sensible.

I would dig it out to a depth of at least 125mm, put down a vapour barrier
and 75mm of polystyrene, then 75mm of concrete....the concrete will
essentially be free because it will replace the cost of the sand/cement, the
polystyrene (jablite) is about £15 per 8X4 sheet.

The only other factors you need to think about are getting rid of the old
concrete and whether you want any ducting anywhere for cables, pipes etc.

--
Phil L
RSRL Tipster Of The Year 2008


jim

unread,
Dec 28, 2008, 5:47:36 PM12/28/08
to

ISTR that SBR mixes can be made (as long as enough SBR used) to be
waterproof in themselves - that's one of it's main uses isn't it? I
used it to construct two hearths that were previously damp being laid
on earth - worked well enough for me...

cheers
jim

jim

unread,
Dec 28, 2008, 5:49:57 PM12/28/08
to

Tim S

unread,
Dec 28, 2008, 6:49:23 PM12/28/08
to
Phil L coughed up some electrons that declared:


> The floor sounds like it's been botched up more than once in the past and
> now you appear to want to do the same again, while you might save a few
> quid, you are still left with a floor with no insulation and no damp
> proofing, and believe me, trying to rectify this once the units are fitted
> is nigh on impossible, not to mention astronomical in price.....doing it
> now while the house is empty and in such a state is both economical and
> sensible.

I don't think it has been botched - the quarry tiles were the originals
(1950's). I think it's just a rough old slab.

Having said that, there's never been issues with cracking or damp.



> I would dig it out to a depth of at least 125mm, put down a vapour barrier
> and 75mm of polystyrene, then 75mm of concrete....the concrete will
> essentially be free because it will replace the cost of the sand/cement,

> the polystyrene (jablite) is about Ł15 per 8X4 sheet.


>
> The only other factors you need to think about are getting rid of the old
> concrete and whether you want any ducting anywhere for cables, pipes etc.
>

If I did, I would follow your excellent advice (indeed, I'm doing it in the
shower room, partly because some of the floor had to come up for drains, so
it wasn't much effort to take the rest of 3m2 up).

Unfortunately, I have to draw the line there - it would be good to have an
insulated floor with UFH everywhere, but the house loses more heat in other
places (roof mainly) so I'd rather put the time and money towards that and
my budget's already tight.

But thanks for the suggestion (NT too).

Cheers

Tim

Tim S

unread,
Dec 28, 2008, 6:50:47 PM12/28/08
to
stuart noble coughed up some electrons that declared:

> If there's no damp, you shouldn't need SBR at all, but it would be an
> extra precaution, at least for the stabiliser.
> SBR doesn't mix well with sand cement mortars. It has a tendency to go
> its own way and leech out if you leave it standing for any length of
> time.

Oh. I thought SBR was the stuff for sand and cement. Never mind.

> Pva mixes in better and the mortar is easier to use because of that.
> You could just use sand/cement for the screed and paint SBR on when it's
> dry. It's penetrates like crazy, which is better than creating a film on
> the surface

That sounds like a plan. Think I need to go and read the SBR data sheets...

Cheers

Tim

Tim S

unread,
Dec 28, 2008, 6:51:43 PM12/28/08
to
jim coughed up some electrons that declared:


>
> www.flowcrete.com/images/product_downloads/file82.pdf
>
> jim

Thanks for that - I'll read it when I have my tea.

Cheers

Tim

Tim S

unread,
Dec 29, 2008, 7:36:03 AM12/29/08
to
Tim S coughed up some electrons that declared:

And it's extremely useful - thanks - I've saved that to my info collection.

All the details of the sealing, priming and screed are covered nicely.

BTW - What's Category 1 BS 13139:2002 sand? Does that mean "Builders sand"
as sold by B&Q or something a bit more specific?

Cheers

Tim

Phil L

unread,
Dec 29, 2008, 2:47:32 PM12/29/08
to

I'd use grit sand - sharp sand with very small stones in it, sets very
hard....mixed about 3 or 4 to 1 cement, grit sand is available in bulk bags
and is about the same price as other sand.

Builder's sand is no good for screed, it's too fine and will dust up.

I'd take that "BS 13139:2002" with a pinch of salt, I C&P it into google
only gives four results and they all pertain to the flowcrete site.

Tim S

unread,
Dec 29, 2008, 2:58:51 PM12/29/08
to
Phil L coughed up some electrons that declared:

>
> I'd use grit sand - sharp sand with very small stones in it, sets very
> hard....mixed about 3 or 4 to 1 cement, grit sand is available in bulk
> bags and is about the same price as other sand.

Cool - thanks Phil, I know the stuff...

Cheers

Tim

Tim S

unread,
Dec 30, 2008, 7:46:32 PM12/30/08
to
meow...@care2.com coughed up some electrons that declared:


> I realise you dont want to be bothered, but now is your opportunity to
> dig it up and put insulation down that will save you n times the cost
> of doing the job. Where n is an unknown number that someone else might
> have more idea about.
>

Well, an interesting discussion came about at work today, between me and the
boss.

He's an engineer and as engineers do, he'd been doing a casual study of
flooring insulation, and I'm sure he won't mind me repeating it here.

He has a kitchen, tiled onto concrete with no insulation. Using an accurate
temperature meter, he'd noticed a 4 celcius difference between air just
above the floor (colder obviously) and air at about 1m above the floor.

Did the same at work and the difference was 0.5C.

We work in a converted apple shed, so teh subfloor is farmer's concrete onto
the ground.

Over that is 25mm Jablite and chipboard panels.

================

Observation - although not upto building regs for newbuilds, even 25mm
Jablite (I assume it would be Jabfloor 70) makes a massive difference.

So I priced it up for my house, 2 ways:

Method 1:

25mm Jabfloor 70 overlaid with 18mm water resistant T+G flooring chipboard
as recommended by Jabfloor's data sheet = (for 95m2 ish, whole ground
floor) £731 inc VAT.

Same with 25mm TB3000 Celotex and 18mm chipboard = £888

Both methods need additional materials such as DPC barrier, glue etc.

=================

The Celotex method looks like it would achieve a U-value of about 0.3 which
doesn't seem bad for a loss in room height of 43mm. Including final
flooring, door frame headroom would be reduced to 1.91m which seems
acceptable.

UFH isn't an option, but I could believe the cost of materials might
actually be recovered through less heating. I'd have to do some
calculations to prove it though.

The only thing that bothers me is having chip under my nice tiles and what
would happen with water ingress (it's going to be impossible to seal the
top layer 100% and hard for the substrate to dry out).

What do folk think?

Cheers

Tim

Tim S

unread,
Dec 30, 2008, 8:14:11 PM12/30/08
to
Tim S coughed up some electrons that declared:

>
> So I priced it up for my house, 2 ways:
>
> Method 1:
>
> 25mm Jabfloor 70 overlaid with 18mm water resistant T+G flooring chipboard
> as recommended by Jabfloor's data sheet = (for 95m2 ish, whole ground
> floor) £731 inc VAT.
>
> Same with 25mm TB3000 Celotex and 18mm chipboard = £888
>

http://www.bathroom2u.com/varme-underfloor-heating-insulation-boards-20mm-p-516.html

Seems to be another option. Nothing rottable there AFAICS and everything's
glued to everything to no voids for water to collect.

Twice the price though (around 2k). But half the work to lay. Anyone know
anything about these "insulation backer boards"? They're new to me...

Cheers

Tim

Rod

unread,
Dec 31, 2008, 3:22:38 AM12/31/08
to

Living in a house with uninsulated solid concrete floors (on the ground
floor!), I would very much appreciate anything that would allow
insulation. The 6mm might just about be usable.

My own highly sensitive thermal detectors[1] show a difference between
warm and FF when stepping down from bottom stair to floor. So I strongly
concur that doing something is highly desirable.

Although I only looked quite quickly - I couldn't see the area of the
tile-backer boards! Is that per square metre or a 2.4x1.2 or what?

[1] - Slightly less noticeable with fluffy slippers on them.

--
Rod

Hypothyroidism is a seriously debilitating condition with an insidious
onset.
Although common it frequently goes undiagnosed.
<www.thyromind.info> <www.thyroiduk.org> <www.altsupportthyroid.org>

abuse@localhost

unread,
Jan 1, 2009, 6:47:20 AM1/1/09
to
On Wed, 31 Dec 2008 00:46:32 +0000, a certain chimpanzee, Tim S
<t...@dionic.net> randomly hit the keyboard and produced:

>Well, an interesting discussion came about at work today, between me and the
>boss.
>
>He's an engineer and as engineers do, he'd been doing a casual study of
>flooring insulation, and I'm sure he won't mind me repeating it here.
>
>He has a kitchen, tiled onto concrete with no insulation. Using an accurate
>temperature meter, he'd noticed a 4 celcius difference between air just
>above the floor (colder obviously) and air at about 1m above the floor.
>
>Did the same at work and the difference was 0.5C.
>
>We work in a converted apple shed, so teh subfloor is farmer's concrete onto
>the ground.
>
>Over that is 25mm Jablite and chipboard panels.

How far away from the external wall did you take the measurements?
Heat is only lost via the edges of a floor, so buildings with large
floor areas can have very little heat losses from the centre of the
floor. Most large buildings don't need to be insulated to comply with
requirements for heat loss, or if they do, only need to insulate the
slab close to the external walls.
--
Hugo Nebula
"If no-one on the internet wants a piece of this,
just how far from the pack have you strayed"?

Tim S

unread,
Jan 1, 2009, 7:26:01 AM1/1/09
to
Hugo Nebula coughed up some electrons that declared:

> How far away from the external wall did you take the measurements?

About 2m I think.

> Heat is only lost via the edges of a floor, so buildings with large
> floor areas can have very little heat losses from the centre of the
> floor. Most large buildings don't need to be insulated to comply with
> requirements for heat loss, or if they do, only need to insulate the
> slab close to the external walls.

OK - so if I understand this, once the slab and some of the ground under the
house has warmed up, net heat loss there is minimal?

Cheers

Tim

The Natural Philosopher

unread,
Jan 1, 2009, 7:41:40 AM1/1/09
to

That's not strictly true in the limit.

Especially with suspended floors.

Also the thermal mass implied by the load of wet soggy soil under a
solid concrete floor laid on it, is a very significant thing to take
into account.

The Natural Philosopher

unread,
Jan 1, 2009, 7:46:15 AM1/1/09
to

No, but *relatively* small.


The building regs have some complex tables for all this relating the
floor area to the external wall length etc.

The actual calculations are VERY complex. Lots of calculus IIRC..


> Cheers
>
> Tim

Tim S

unread,
Jan 1, 2009, 8:09:33 AM1/1/09
to
The Natural Philosopher coughed up some electrons that declared:

> Tim S wrote:
>> Hugo Nebula coughed up some electrons that declared:
>>
>>> How far away from the external wall did you take the measurements?
>>
>> About 2m I think.
>>
>>> Heat is only lost via the edges of a floor, so buildings with large
>>> floor areas can have very little heat losses from the centre of the
>>> floor. Most large buildings don't need to be insulated to comply with
>>> requirements for heat loss, or if they do, only need to insulate the
>>> slab close to the external walls.
>>
>> OK - so if I understand this, once the slab and some of the ground under
>> the house has warmed up, net heat loss there is minimal?
>>
>
> No, but *relatively* small.

OK. Now if the building is heated 10 hours out of 24, then would it be a
reasonable assumption that the ground floor would tend to reach a mean
temperature? Assuming a time constant in the order of days (or even a
couple of weeks) the floor mass might only manage to get upto 10-15C which
could explain the observations.

Cheers

Tim

The Natural Philosopher

unread,
Jan 1, 2009, 9:01:21 AM1/1/09
to
Tim S wrote:
> The Natural Philosopher coughed up some electrons that declared:
>
>> Tim S wrote:
>>> Hugo Nebula coughed up some electrons that declared:
>>>
>>>> How far away from the external wall did you take the measurements?
>>> About 2m I think.
>>>
>>>> Heat is only lost via the edges of a floor, so buildings with large
>>>> floor areas can have very little heat losses from the centre of the
>>>> floor. Most large buildings don't need to be insulated to comply with
>>>> requirements for heat loss, or if they do, only need to insulate the
>>>> slab close to the external walls.
>>> OK - so if I understand this, once the slab and some of the ground under
>>> the house has warmed up, net heat loss there is minimal?
>>>
>> No, but *relatively* small.
>
> OK. Now if the building is heated 10 hours out of 24, then would it be a
> reasonable assumption that the ground floor would tend to reach a mean
> temperature?

Yes. Although it might not be uniform across the floor.

> Assuming a time constant in the order of days (or even a
> couple of weeks) the floor mass might only manage to get upto 10-15C which
> could explain the observations.
>

Yes. Even an insulated concrete floor has a time constant of several
days - well mine has!

Add in a few hundred tons of soil underneath and well...its longer!

the whole efficiency of ground source heatpumps depends on the fact that
the time constant of subsoil below a meter is measured in MONTHS.

Hmm. got not much better to do..let's look at concrete specific heat and
do some calcs..

Assume 0.2 calories per gram per degree C..
Assume about 3 tonnes per cubic meter.
Assume a 100mm screed thickness.
Assume a 50mm polystyrene sheet under a block and beam floor - that's
what I have..

First calc. Assuming 100W/sq meter..thats a reasonable figure for a well
insulated house, how fast does it warm up?.

A square meter of 100mm thick screed is 1/10th of a tonne, so 100kg, so
will need 20 thousand calories to warm by a degree C..or 23 watt hours
roughly.

So at 100 watts per sq meter I should see, with no losses 4 degrees C
per hour..

OK, looking at the losses downwards the U value is about 0.6 for he
styrene sheet, so at say 10C difference (5C outside and under the
suspended floor) and 15 C in the floor, that's 6W per square meter loss.

Not too significant. But in terms of losses, the time constant is
definitely in the many hours level. 23 watts per hour per deg C is the
thermal inertia...0.6 watts per degree C is the losses.. so a native
time constant excluding losses to the room, of 23/0.6 hours.. a shade
over 38 hours.

Now add in the chimney stacks..haha.

If the slab were in contact with the soil itself..that 100mm thick might
easily be much higher. But I don't have a figure for the insulation of
soil, and the path through it to the cold air outside..

Not sure what all that proves, except that people with even well
insulated concrete floors have a lot of thermal inertia in their houses.

And why you need feed forward in UFH schemes to prevent temperature
overshoot, especially if you have a high power to area going into them.

Another reason why it's probably better if you don't, to leave em on 24x7.

> Cheers
>
> Tim

meow...@care2.com

unread,
Jan 1, 2009, 12:55:36 PM1/1/09
to


Do you really think a chipboard floor in wet conditions wont move or
rot in your lifetime? I wouldnt do it :)

There may be an easier & much cheaper way, thats to use an insulating
concrete slab. IOW the crete is made with a high percentage of
something like perlite, which insulates. So the slab itself is a
lattice of 3:1 mix and little air voids. Insulation wont be nearly as
good as polystyrene, but
a) it gets you a long lasting solid floor with no headroom loss
b) its cheap
c) its much less work/time to lay

I assume youre not wililng to dig out a few more inches.

I would also question whether if you've had concrete with no dpm for
years, you need a dpm now. I know its the fashion, but lets face it,
millions of us live in no dpm houses, and most are just fine. Your
risk, your choice.


NT

Tim S

unread,
Jan 1, 2009, 1:03:32 PM1/1/09
to

>> ground floor) ?731 inc VAT.
>>
>> Same with 25mm TB3000 Celotex and 18mm chipboard = ?888


>>
>> Both methods need additional materials such as DPC barrier, glue etc.
>>
>> =================
>>
>> The Celotex method looks like it would achieve a U-value of about 0.3
>> which doesn't seem bad for a loss in room height of 43mm. Including final
>> flooring, door frame headroom would be reduced to 1.91m which seems
>> acceptable.
>>
>> UFH isn't an option, but I could believe the cost of materials might
>> actually be recovered through less heating. I'd have to do some
>> calculations to prove it though.
>>
>> The only thing that bothers me is having chip under my nice tiles and
>> what would happen with water ingress (it's going to be impossible to seal
>> the top layer 100% and hard for the substrate to dry out).
>>
>> What do folk think?
>>
>> Cheers
>>
>> Tim
>

Hi

> Do you really think a chipboard floor in wet conditions wont move or
> rot in your lifetime? I wouldnt do it :)

Exactly - that's why I raised the concerned regarding chip myself and
started looking at insulated backer board instead (eg marmox). Maybe that
post got lost.

> There may be an easier & much cheaper way, thats to use an insulating
> concrete slab. IOW the crete is made with a high percentage of
> something like perlite, which insulates. So the slab itself is a
> lattice of 3:1 mix and little air voids. Insulation wont be nearly as
> good as polystyrene, but
> a) it gets you a long lasting solid floor with no headroom loss
> b) its cheap
> c) its much less work/time to lay
> I assume youre not wililng to dig out a few more inches.

Thing is, I don't want to dig out any inches. Only one area has been dug
out, and that will get proper insulation under the screed.

The other small area is rough due to breaking off tiles but the slab is OK.

I really don't want to hack any more of the floor (I have 95m2) - so it's
either what I can practically stick on top, or nothing.



> I would also question whether if you've had concrete with no dpm for
> years, you need a dpm now. I know its the fashion, but lets face it,
> millions of us live in no dpm houses, and most are just fine. Your
> risk, your choice.
>

This is of course true. DPM might be worse - I don't know these things for
sure at the moment.

Cheers

Tim

meow...@care2.com

unread,
Jan 1, 2009, 4:52:14 PM1/1/09
to
Tim S wrote:
> meow...@care2.com coughed up some electrons that declared:

> Hi


Perlite-crete sounds even more attractive then. No need to dig up
anything. Its full of half pea sized bits so will need some plain
mortar to smooth it over.


NT

Reply all
Reply to author
Forward
0 new messages