What went wrong...?

[zoemac]

Hello folks I need some help.

At the start of December 2013 I had a Tarmac Tru Flo Screed floor laid over underfloor heating. I believe this is a gyvlon/gypsom/anhydrite screed.

I wanted to tile on the floor and was given advice by the screed company that either a Gypsom based adhesive was needed or if we wanted to use a cement based adhesive we would need to seal the floor well.

The floor was kept above 5 degrees c and was left to dry naturally until the mid Feb when the underfloor heating was switched on just to make sure it was dry.

The floor was swept and hoovered to remove any surface residue and dust and the my builder applied numerous coats of SBR primer to seal the floor.

The floor was then tiled with porcelain tiles using BAL Rapidset flexible. We didn't use the underfloor heating again till end of March to give the floor the chance to dry naturally (and because we were still working on other parts of the house and not actually living there)

The problem that I have is that all the tiles are lifting. I first noticed this in April 2014 and it has got worse and worse. Tonight we took up a couple of tiles and the adhesive is stuck really well to the tiles, but does not seem to be bonded to the screed. I am aware that there can be a reaction between the Gypsum Screed and the Cement based tile adhesive but the advice we were given was that this would not happen if the floor was well sealed.

What has gone wrong?

My builder came to take a look tonight and basically he thinks we need to re-lay the whole floor, and because the adhesive is so well bonded to the tiles we will not be able to re-use them. Its a BIG room so this will cost me a fortune - and I really can't afford to do this. What can I do??? Help please.

 

[Dougs Third Go]

sounds to me as though it could be laitence, was the floor sanded before sweeping/hoovering/priming?

 

[Gazzer]

As Doug says, the laitance should have been removed and hoovered up . have you removed a tile and if so what does the floor look like beneath and what does the adhesive on the tile look like?

 

[zoemac]

Thanks Doug. I did a bit of research before I got the floor laid. I believe there are a couple of different flow screed products. The one we used was Tarmac Tru Flow which I believe is slightly different from the Lafarge flow screed. The supplier told us that the tarmac one would not need sanding, it formed a thin crust which just needed to be removed by sweeping. I have double checked the data sheets we were sent at the time and they confirm this. I believe the LaFarge product forms a really tough shiny crystalline layer which needs to be sanded.

 

[CJ]

Also, it should have been an acrylic primer. 1st coat diluted, 2nd coat neat, but as mentioned above, if it wasn't laitence free, and thoroughly hoovered.......you were always going to have a problem.

It does sound like another "Builder error"

 

[Ajax123]

There's a good chance it was not dry. Natural drying would need around 20 degrees and 60% rh. The choice of adhesive puts you at risk and whilst SBR can be used it is sensitive to moisture as well. Sanding the floor to remove Laitance and any other surface detritus is essential for correct preparation. If the tiles are coming up clean it is a printer failure as that is the bit that is supposed to stick to the screed and the adhesive sticks to that in turn. You will need to lift the tiles and replace them. I would use a gypsum based adhesive. There are loads about and they are no more expensive than cement based these days.

 

[zoemac]

Sorry - as you will guess I am not all that clued up. The laitance is the layer that formed on top of the screed, yes? We definitely removed all of this - it didn't need sanding, came up easily as we brushed with a stiff brush. This is what we were told to expect by the screed people.

Yep - have removed a couple of tiles (didn't take much effort!) - the adhesive seems well bonded to the tile but not to the floor. The adhesive looked pale, chalky and was very brittle. There was still signs of the SBR on the floor. I'll take some pictures later on and post them here.

Again I'm really grateful for your help as I am clueless as to what could have gone wrong.

 

[zoemac]

Showing my stupidity no doubt - is SBR not an acrylic primer?

 

[The D]

who decided the floor was dry and did they test it or was it just that surfactant time had passed so the floor must be dry

 

[Ajax123]

who decided the floor was dry and did they test it or was it just that surfactant time had passed so the floor must be dry

I suspect no test was carried out but won't pre empt...

 

[zoemac]

We hired BIG electric heaters and had them onsite for laying the screed. These we used for around 6 weeks after so I'd be surprised it if wasn't dry. It then had another week or so with the underfloor heating on low.

I should add that the Screed Firm thought that this would me sufficient.

 

[zoemac]

Yes - no test was carried out - I think we all just presumed after over 2 months it would be dry.

 

[zoemac]

How do I put pictures up on here?

 

[zoemac]

https://www.facebook.com/media/set/?set=a.10155145852350113.1073741834.644845112&type=1&l=398ee69809

 

[zoemac]

I've just set up a facebook album with some photos - hope it works. https://www.facebook.com/mcleanzoe/...41834.644845112&type=3&viewas=100000686899395

 

[Dave]

Just to add , if the SBR wasn't diluted to correct ratio , it would simply skin on the screed surface ... Also SBR can be used as a primer but it is advisable to use an acrylic primer and always dilute to particular brand recommendations .

as above the trend now is to use a gypsum based adhesive ....

also so when you turned on the heating , did you do it gradually or just cranked it on ..?

 

[zoemac]

I just lifted a tile again, and I have added some photos to the Facebook album (link above) the floor underneath feels a little damp despite the underfloor heating being on - is this normal, or is this a sign that the screed was not dry enough when it was tiled.

The adhesive is very firmly stuck to the tile, but not at all to the floor. The floor is covered with a white substance which is dusty in places and peels in others. I think this is the remains of the SBR.

Does the fact that I can peel/scrape off bits of SBR mean that it skinned rather than soaked in?

When I scrape away the SBR the floor underneath looks good and clean - no other sign of residue or laitance. It is rough to the touch so I would have thought it'd be an ideal surface to tile onto.

 

[Dash J]

I think laitance is the problem here
i have seen exactly the same problem on 150m2 of granite with ufh
the floor was not scrabbled,laitance removed
it was exactly as you say,wet underneath and no adhesion to sub floor
white patchy areas,chalky its laitance

 

[zoemac]

Thanks Dash. What would the laitance look like on the floor before tiling? Do you not think this could be the SBR skinning or a reaction between the gypsum floor and concrete based adhesive? When the floor was laid we were left with a shiny grey surface layer which as we walked on it cracked and flaked off. Brushing with a stiff brush removed the remainder leaving a clean surface which looked just like the surface between the powdery/peeling layer which is under the tiles now. Can you see the pictures?

 

[Dash J]

Hi Zoe I haven't seen the pics,will have a look
just with what you are saying it sounds v much like laitance
it's the word powdery that makes me think this
also not commissioning the ufh or checking the screed is dry is a major error
it sounds v much a lift and re lay unfortunately
I don't know who blame lies with because the screed needs to be checked for moisture
the tiler could blame you,you could blame the tiler or the builder

 

[zoemac]

Sorry I think my use of the word powdery is a red herring.

I can peel off bits which is what makes me think its the SBR. There is a dusty layer ontop of the SBR but think this is from the tile adhesive which seems to be very chalky and brittle. I've added a new photo.

Under floor heating was commissioned before tiling - it was run on low for 8 hours on a couple of days.

 

[Spud]

From what I have read it sounds like you had laitence on the floor and it was still wet when it was tiled , I don't know of any screed where you can just brush the laitence off with a broom all the floors I have encountered have needed mechanical scarifying to remove the surface layer , running the heating on low for a couple of days wouldn't have achieved anything in regards commissioning the screed or help with force drying it

 

[Gazzer]

Another thing ! You say the floor heating was turned on but was it on when it was tiled ?

 

[judge]

Chalky and brittle adehsive doesnt sound to good either !

 

[Ajax123]

I just lifted a tile again, and I have added some photos to the Facebook album (link above) the floor underneath feels a little damp despite the underfloor heating being on - is this normal, or is this a sign that the screed was not dry enough when it was tiled.

no it is not normal if the screed is properly dry. The dampness is a clear indication that moisture has either gotten in or more likely was there in the first place

The adhesive is very firmly stuck to the tile, but not at all to the floor. The floor is covered with a white substance which is dusty in places and peels in others. I think this is the remains of the SBR.

this sounds like either Laitance or ettringite. Either way it is not good

Does the fact that I can peel/scrape off bits of SBR mean that it skinned rather than soaked in?

yes

When I scrape away the SBR the floor underneath looks good and clean - no other sign of residue or laitance. It is rough to the touch so I would have thought it'd be an ideal surface to tile onto.

i cannot see the pics from the link above but have answered your questions based on the information supplied. If the adhesive is powdery and brittle this would indicate moisture and possible ettringite. The tiles will unfortunately be unlikely to be save able and will need to be replaced

 

[Ajax123]

I think laitance is the problem here
i have seen exactly the same problem on 150m2 of granite with ufh
the floor was not scrabbled,laitance removed
it was exactly as you say,wet underneath and no adhesion to sub floor
white patchy areas,chalky its laitance

What you are describing does not sound like Laitance. It sounds more like segregation which is a screed defect. You should not scabble screeds. They should be mechanically abraded which usually means sanding.

 

[zoemac]

Thanks folks - it looks like possibly the floor was still damp despite having 2.5 months to dry some of which with heater and some of which with under floor heating commissioned. Also sounds like SBR could have been too thick and formed a skin.

Given the info from the screed supplier and the way the floor was prepped I just can't see that it laitance (but I could be wrong). It seems more likely that despite the SBR a reaction between the gypsum screed and the cement adhesive occurred which is why the adhesive seems really powdery and brittle.

Does that sound right to people?

Guess I just have to suck it up - unless anyone has any other suggestions?

So the plan of action would be - take up all the tiles, run the underfloor heating full belt for a bit to ensure it is super dry and then prep and seal the floor again and relay new tiles with a gypsum tile adhesive.

What do you think?

Can't afford to do this at the moment - so will be living with a cracked floor for a while. But at least I know what I now need to do

 

[Ajax123]

If you have run the heating for a couple of days on low then it has not been properly comissioned.
the correct procedure takes two to thee weeks.

this might help http://www.gypsol.co.uk/4_Underfloor_heating.pdf

 

[Ajax123]

If this was done by your builder or by a tiler then they should be footing at least some of the bill....

 

[zoemac]

No was left to dry for a while - probably 2 weeks and then was run on low for a short while and then underfloor heating not used till November

 

[zoemac]

Thanks Alun - the problem is proving anything - I guess I could blame the tile shop as I described exactly what I needed and they recommeneded SBR + cement adhesive rather than a gypsum adhesive. I could blame by builder who also did the tiling - was the SBR applied to thickly, was the floor dry enough to be tiled. I could blame the screed people - was there a fault with the screed? But how do I actually prove what the problem was and who was at fault? It is going to cost me at least £3500 to put right - money the I don't have right now - but I can't see the tile shop, builder or screed company volunteering to contribute.